From scottp at dd.com.au Wed Jan 4 19:58:44 2006 From: scottp at dd.com.au (Scott Penrose) Date: Thu, 5 Jan 2006 14:58:44 +1100 Subject: [Melbourne-pm] Perl Mongers next week - Wednesday 18th of January - 6:30pm - Lightning Talks Message-ID: <820E60B7-97C7-4F9F-BB90-31AC44B453E8@dd.com.au> Hey Perl Mongers. Happy new year and all of the seasons greetings too you. I hope you have all hauled your way through the silly season and back into the swing well rested. This month we were planning on doing Lightning Talks. A few volunteers have already made contributions, but you are all welcome to come up and do a talk. The rules - 7 minutes is your maximum - including setup time. If you have two talks, you can do both, but don't expect them to run side by side (and part 1 of 2 is not allowed). It should be fun. Please bring along your ideas, email them to the list if you like, and turn up next Wednesday week at 6:30 (the 3rd of the month). Where: myinternet Level 8, 14-20 Blackwood Street North Melbourne When: 6:30pm Wednesday 18th of January 2006 What: Lots of lightning talks ! After: Dinner :-) Scott -- * - * http://www.osdc.com.au - Open Source Developers Conference * - * Scott Penrose Welcome to the Digital Dimension http://www.dd.com.au/ scottp at dd.com.au Dismaimer: Contents of this mail and signature are bound to change randomly. Whilst every attempt has been made to control said randomness, the author wishes to remain blameless for the number of eggs that damn chicken laid. Oh and I don't want to hear about butterflies either. Please do not send me Word or PowerPoint attachments. See http://www.gnu.org/philosophy/no-word-attachments.html Microsoft is not the answer. It's the question. And the answer is no. -------------- next part -------------- A non-text attachment was scrubbed... Name: PGP.sig Type: application/pgp-signature Size: 186 bytes Desc: This is a digitally signed message part Url : http://mail.pm.org/pipermail/melbourne-pm/attachments/20060105/25fb14fa/PGP.bin From benji at arsimagica.net Wed Jan 4 20:56:57 2006 From: benji at arsimagica.net (Benji Wakely) Date: Thu, 05 Jan 2006 15:56:57 +1100 Subject: [Melbourne-pm] Simple(?) question involving backticks and die Message-ID: <200601050456.k054uvAn003428@arch.localnet> Greetings all. I'm confused by this behaviour... When I look at the two lines of code: my @list = `/bin/ls /home` || die ("Couldn't list files"); die ("Couldn't list files") unless my @list = `/bin/ls /home`; they to be equivalent to me. The perl interpreter begs to differ, however, because: The first invocation generates: $list[0] = "alfred betty charlie" and the second invocation fills the list up one element at a time, eg. $list[0] = "alfred"; $list[1] = "betty"; $list[2] = "charlie"; ...I was just curious to know what actually causes this difference in behaviour. Any takers? -- Benji Wakely +614 0958 8566 benji at arsimagica.net From tconnors+pmmelb at astro.swin.edu.au Wed Jan 4 21:27:29 2006 From: tconnors+pmmelb at astro.swin.edu.au (Tim Connors) Date: Thu, 5 Jan 2006 16:27:29 +1100 (EST) Subject: [Melbourne-pm] Simple(?) question involving backticks and die In-Reply-To: <200601050456.k054uvAn003428@arch.localnet> References: <200601050456.k054uvAn003428@arch.localnet> Message-ID: On Thu, 5 Jan 2006, Benji Wakely wrote: > Greetings all. > > I'm confused by this behaviour... > > When I look at the two lines of code: > > my @list = `/bin/ls /home` || die ("Couldn't list files"); Aren't you meant to use "or" rather than "||" when using die, for precedence reasons? Could this be what is happening? `` becomes scalar context when combined with the || ? Dunno, haven't tried. -- TimC > As you know, Linus took the word the penguins kept saying over and > over again, rot13'ed it, and used that as the name of his OS. So, who's going to record this .au file: "Hello, my name is Yvahf Gbeinyqf, and I pronounce yvahk, yvahk." -- Anthony de Boer && Michel Buijsman in ASR From benji at arsimagica.net Wed Jan 4 21:33:03 2006 From: benji at arsimagica.net (Benji Wakely) Date: Thu, 05 Jan 2006 16:33:03 +1100 Subject: [Melbourne-pm] Simple(?) question involving backticks and die In-Reply-To: Your message of "Thu, 05 Jan 2006 16:27:29 +1100." Message-ID: <200601050533.k055X3Fd003826@arch.localnet> On Thu, 5 Jan 2006 16:27:29 +1100 (EST), Tim Connors wrote: >> my @list = `/bin/ls /home` || die ("Couldn't list files"); > >Aren't you meant to use "or" rather than "||" when using die, for >precedence reasons? Could this be what is happening? `` becomes scalar >context when combined with the || ? That checks out, and now I'll happily use 'or' rather than '||' in this situation. ...Wish I knew why it was so, though. -- Benji Wakely From tconnors+pmmelb at astro.swin.edu.au Wed Jan 4 21:38:28 2006 From: tconnors+pmmelb at astro.swin.edu.au (Tim Connors) Date: Thu, 5 Jan 2006 16:38:28 +1100 (EST) Subject: [Melbourne-pm] Simple(?) question involving backticks and die In-Reply-To: <200601050533.k055X3Fd003826@arch.localnet> References: <200601050533.k055X3Fd003826@arch.localnet> Message-ID: On Thu, 5 Jan 2006, Benji Wakely wrote: > > On Thu, 5 Jan 2006 16:27:29 +1100 (EST), > Tim Connors wrote: > > >> my @list = `/bin/ls /home` || die ("Couldn't list files"); > > > >Aren't you meant to use "or" rather than "||" when using die, for > >precedence reasons? Could this be what is happening? `` becomes scalar > >context when combined with the || ? > > That checks out, and now I'll happily use 'or' rather than '||' > in this situation. > > ...Wish I knew why it was so, though. Why? Because what you were trying to do was to || the backticks with the die, and then assigning the result into the array. || forces the backticks to become a scalar, and the result is true, so die doesn't evaluate to anything. Hence, the scalar ends up in the first element of the array. The alternatives, are to bracket things correctly, forcing the evaluation of the backticks in array context, then assigning that to the array, and then evaluating whether to die. Or use "or", which has a lower precedence than "||", as it is meant to be used -- the proper idiom is to use: $blah=blah() or die() rather than $blah=blah() || die() Read up on precedence in a perl or C book somewhere. -- TimC Disclaimer: Contents of this email are to be considered secret. You must not read this email, and you must dispose of your computer immediately, should you receive this email accidentally. I cannot take responsibility for any incorrect information that this email may contain, especially the bit about the monkeys and John Howard. Should anyone say otherwise, they are obviously lying. From scottp at dd.com.au Wed Jan 4 21:41:02 2006 From: scottp at dd.com.au (Scott Penrose) Date: Thu, 5 Jan 2006 16:41:02 +1100 Subject: [Melbourne-pm] Simple(?) question involving backticks and die In-Reply-To: <200601050533.k055X3Fd003826@arch.localnet> References: <200601050533.k055X3Fd003826@arch.localnet> Message-ID: <1E92C16A-D0AC-4FD1-AF24-F7191ADA2A4D@dd.com.au> Precedence really. By doing this... "blah" || "blah" You are saying you want one or other the scalars above. Therefore it returns in scalar context. By using "or" you are lowering the precedence and thus it no longer treats it in scalar, but rather array context. It is generally safer to use "or" unless you want to return the value. eg: my $b = executethis() || "default value"; if you want to die then my $b = executethis() or die("ahh, it broke!"); Scott On 05/01/2006, at 16:33, Benji Wakely wrote: > > On Thu, 5 Jan 2006 16:27:29 +1100 (EST), > Tim Connors wrote: > >>> my @list = `/bin/ls /home` || die ("Couldn't list files"); >> >> Aren't you meant to use "or" rather than "||" when using die, for >> precedence reasons? Could this be what is happening? `` becomes >> scalar >> context when combined with the || ? > > That checks out, and now I'll happily use 'or' rather than '||' > in this situation. > > ...Wish I knew why it was so, though. > > -- > Benji Wakely > > _______________________________________________ > Melbourne-pm mailing list > Melbourne-pm at pm.org > http://mail.pm.org/mailman/listinfo/melbourne-pm -- * - * http://www.osdc.com.au - Open Source Developers Conference * - * Scott Penrose VP in charge of Pancakes http://linux.dd.com.au/ scottp at dd.com.au Dismaimer: If you receive this email in error - please eat it immediately to prevent it from falling into the wrong hands. Please do not send me Word or PowerPoint attachments. See http://www.gnu.org/philosophy/no-word-attachments.html Microsoft is not the answer. It's the question. And the answer is no. -------------- next part -------------- A non-text attachment was scrubbed... Name: PGP.sig Type: application/pgp-signature Size: 186 bytes Desc: This is a digitally signed message part Url : http://mail.pm.org/pipermail/melbourne-pm/attachments/20060105/c756232b/PGP.bin From stephen at sydney.pm.org Wed Jan 4 21:42:10 2006 From: stephen at sydney.pm.org (Stephen Steneker) Date: Thu, 5 Jan 2006 16:42:10 +1100 Subject: [Melbourne-pm] Simple(?) question involving backticks and die In-Reply-To: <200601050456.k054uvAn003428@arch.localnet> References: <200601050456.k054uvAn003428@arch.localnet> Message-ID: <894E9B6C-FAB9-4B32-8104-D9D95A25ED59@sydney.pm.org> > my @list = `/bin/ls /home` || die ("Couldn't list files"); > > die ("Couldn't list files") unless my @list = `/bin/ls /home`; > > they to be equivalent to me. Hi Benji, In your first example, the precedence of the binary or ('||') is higher than the assignment so coerces the backtick results into scalar context for comparison. What you end up with is the equivalent of: my @list = (`/bin/ls /home` || die ("Couldn't list files")); ... so you really want to be using the lower precedence "or" : my @list = `/bin/ls /home` or die ("Couldn't list files"); Although for fun you could always make the brackets explicit: (my @list = `/bin/ls /home`) || die ("Couldn't list files"); Cheers, Stephen * http://perldoc.perl.org/perlop.html#Operator-Precedence-and- Associativity From benji at arsimagica.net Wed Jan 4 21:47:47 2006 From: benji at arsimagica.net (Benji Wakely) Date: Thu, 05 Jan 2006 16:47:47 +1100 Subject: [Melbourne-pm] Simple(?) question involving backticks and die In-Reply-To: Your message of "Thu, 05 Jan 2006 16:42:10 +1100." <894E9B6C-FAB9-4B32-8104-D9D95A25ED59@sydney.pm.org> Message-ID: <200601050547.k055lmtB004125@arch.localnet> Thank you all for making a simple door very happy. -- Benji From jarich at perltraining.com.au Thu Jan 5 15:27:29 2006 From: jarich at perltraining.com.au (Jacinta Richardson) Date: Fri, 06 Jan 2006 10:27:29 +1100 Subject: [Melbourne-pm] Perl Mongers next week - Wednesday 18th of January - 6:30pm - Lightning Talks In-Reply-To: <820E60B7-97C7-4F9F-BB90-31AC44B453E8@dd.com.au> References: <820E60B7-97C7-4F9F-BB90-31AC44B453E8@dd.com.au> Message-ID: <43BDAB61.8030406@perltraining.com.au> G'day Scott, Next week isn't the 18th of January, according to my calendar. Is the meeting next week on the 11th, or the week after on the 18th? Jacinta -- ("`-''-/").___..--''"`-._ | Jacinta Richardson | `6_ 6 ) `-. ( ).`-.__.`) | Perl Training Australia | (_Y_.)' ._ ) `._ `. ``-..-' | +61 3 9354 6001 | _..`--'_..-_/ /--'_.' ,' | contact at perltraining.com.au | (il),-'' (li),' ((!.-' | www.perltraining.com.au | From scottp at dd.com.au Thu Jan 5 15:32:22 2006 From: scottp at dd.com.au (Scott Penrose) Date: Fri, 6 Jan 2006 10:32:22 +1100 Subject: [Melbourne-pm] Perl Mongers next week - Wednesday 18th of January - 6:30pm - Lightning Talks In-Reply-To: <43BDAB61.8030406@perltraining.com.au> References: <820E60B7-97C7-4F9F-BB90-31AC44B453E8@dd.com.au> <43BDAB61.8030406@perltraining.com.au> Message-ID: <34BF89A1-6992-4E0B-AD49-B600BD4C7067@dd.com.au> On 06/01/2006, at 10:27, Jacinta Richardson wrote: > G'day Scott, > > Next week isn't the 18th of January, according to my calendar. Is > the meeting > next week on the 11th, or the week after on the 18th? Sorry yes. I got the dates and stuff right, but wrote next week by mistake in the subject. Definitely the 18th ! Not next week, the one after. Thanks Jacinta Scott -- * - * http://www.osdc.com.au - Open Source Developers Conference * - * Scott Penrose Open source developer http://linux.dd.com.au/ scottp at dd.com.au Dismaimer: Open sauce usually ends up never coming out (of the bottle). Please do not send me Word or PowerPoint attachments. See http://www.gnu.org/philosophy/no-word-attachments.html Microsoft is not the answer. It's the question. And the answer is no. -------------- next part -------------- A non-text attachment was scrubbed... Name: PGP.sig Type: application/pgp-signature Size: 186 bytes Desc: This is a digitally signed message part Url : http://mail.pm.org/pipermail/melbourne-pm/attachments/20060105/13798707/PGP.bin From pjf at perltraining.com.au Mon Jan 9 16:51:50 2006 From: pjf at perltraining.com.au (Paul Fenwick) Date: Tue, 10 Jan 2006 11:51:50 +1100 Subject: [Melbourne-pm] PerlNet: Australian Perl Portal Opens Message-ID: <43C30526.8010302@perltraining.com.au> G'day Melbourne.PM, As some of you may be aware, PerlNet, the Australian Perl Portal, has been in incubation for almost a year. However the time has finally come to bring it out into the light. Melbourne.PM already has a page at http://perl.net.au/wiki/Melbourne_Perl_Mongers . Be bold in your edits and enjoy! The full press release follows: == PerlNet: Australian Perl Portal Opens == 10th January, 2005 (Australia) PerlNet (http://perl.net.au/), the Australian and New Zealand Perl portal, is announcing its doors open to the world. PerlNet is focused on all aspects of Perl in Australia and New Zealand, including user-groups, business, education and development. The content of PerlNet can be freely edited and distributed, and the site uses MediaWiki, the same software that powers Wikipedia. PerlNet's goal is to be a resource for all users of Perl, and to bring the community together to collaborate and share ideas. Businesses that support Perl are encouraged to provide listings. User-groups are encouraged to host their pages and content. Individuals are encouraged to write about their projects and experiences. Stephen Steneker, the leader of Sydney Perl Mongers, has already moved his group's content across to the new site. "PerlNet just makes it much easier for our members to get involved," said Steneker. "It's easy for members to write meeting reports, plan new events, and discuss ideas and projects." While PerlNet is aimed primarily at the Australian and New Zealand communities, it's free for anyone to use and edit. "We've been seeing a wide range of contributors," said Paul Fenwick, the founder of PerlNet and director of Perl Training Australia. "We've had material submitted from editors in Europe, China, Israel, North America, as well as locally." All contributions to PerlNet are made under a creative-commons license, allowing everyone to use, distribute, and make derivative works. Registering as an editor takes only a few seconds, and both a mailing list and IRC channel exist to help further discussions. -- Paul Fenwick | http://perltraining.com.au/ Director of Training | Ph: +61 3 9354 6001 Perl Training Australia | Fax: +61 3 9354 2681 From pjf at perltraining.com.au Mon Jan 9 21:29:46 2006 From: pjf at perltraining.com.au (Paul Fenwick) Date: Tue, 10 Jan 2006 16:29:46 +1100 Subject: [Melbourne-pm] Wikification of melbourne.pm.org? Message-ID: <43C3464A.3090508@perltraining.com.au> G'day Melb.PM, The current melbourne.pm.org webpage looks rather plain, and is a far cry from the exciting site we had running under YACMaS. I'm like to see the site returned to its former glory, with calendars, talk descriptions, member profiles, summaries, reviews, and all the rest. Now that perl.net is out in the open, we have a grand tool for making it easier for community editing to occur. There's already a Melbourne.PM page at http://perl.net.au/wiki/Melbourne_Perl_Mongers and even a meeting history at http://perl.net.au/wiki/Melbourne_Perl_Mongers/Meeting_History . However I believe this is just the tip of the iceberg; there's much more information that could be added regarding Melbourne.PM that hasn't. I'd like to put forth the proposal that we turn melbourne.pm.org into a redirect to the Melbourne.PM PerlNet page. That means that all members can then contribute and improve the pages as they see fit. If that goes ahead, I'd also like to see if we can dig up the old YACMaS database and extract the information from it. I'm sure that many hands will make light work in converting it over to its new format. What's my motivation for all this? Well, I'd like the Melbourne.PM website to be great, but most importantly, I want it to be better than Sydney.PM's pages. ;) All the very best, Paul -- Paul Fenwick | http://perltraining.com.au/ Director of Training | Ph: +61 3 9354 6001 Perl Training Australia | Fax: +61 3 9354 2681 From blm at woodheap.org Mon Jan 9 23:46:56 2006 From: blm at woodheap.org (Ben Marsh) Date: Tue, 10 Jan 2006 18:46:56 +1100 Subject: [Melbourne-pm] Wikification of melbourne.pm.org? In-Reply-To: <43C3464A.3090508@perltraining.com.au> References: <43C3464A.3090508@perltraining.com.au> Message-ID: <20060110074656.GA6373@amygdala.woodheap.org> * Paul Fenwick (pjf at perltraining.com.au) wrote: > G'day Melb.PM, > The current melbourne.pm.org webpage looks rather plain, and is a far cry from > the exciting site we had running under YACMaS. I'm like to see the site > returned to its former glory, with calendars, talk descriptions, member > profiles, summaries, reviews, and all the rest. What is YACMaS? > Now that perl.net is out in the open, we have a grand tool for making it easier > for community editing to occur. There's already a Melbourne.PM page at > http://perl.net.au/wiki/Melbourne_Perl_Mongers and even a meeting history at > http://perl.net.au/wiki/Melbourne_Perl_Mongers/Meeting_History . However I > believe this is just the tip of the iceberg; there's much more information that > could be added regarding Melbourne.PM that hasn't. Granted I am only new here but I like the idea of a seperate site. Doesnt it look better for Melbourne Perl Mongers to have their own site? Don't get me wrong. I like wikis and I like perl.net.au. Admittedly the current Melbourne.pm.org maintainer seems very busy on other things. I actually had the crazy thought that I might help with/do the melbourne.pm.org web page. I do envisage that such a website requires information to be donated from lots of people. But then so does a wiki. > > What's my motivation for all this? Well, I'd like the Melbourne.PM website to > be great, but most importantly, I want it to be better than Sydney.PM's pages. ;) > Me too. :-) If everyone thinks that the best way for that is the redirect to the Melbourne_Perl_Mongers page then so be it. But I like the idea of a separate web site more. Ben Marsh From jarich at perltraining.com.au Tue Jan 10 00:09:46 2006 From: jarich at perltraining.com.au (Jacinta Richardson) Date: Tue, 10 Jan 2006 19:09:46 +1100 Subject: [Melbourne-pm] Wikification of melbourne.pm.org? In-Reply-To: <20060110074656.GA6373@amygdala.woodheap.org> References: <43C3464A.3090508@perltraining.com.au> <20060110074656.GA6373@amygdala.woodheap.org> Message-ID: <43C36BCA.8090903@perltraining.com.au> Ben Marsh wrote: > What is YACMaS? Yet another content management system. We used to have a dynamic site until pm.org got hacked one too many times and they decided to host static web pages only. > Granted I am only new here but I like the idea of a seperate site. > Doesnt it look better for Melbourne Perl Mongers to have their own site? I really can't imagine it looking bad for Melbourne.pm.org to redirect to anywhere else, but that might just be me. > Admittedly the current Melbourne.pm.org maintainer seems very busy on > other things. I actually had the crazy thought that I might help > with/do the melbourne.pm.org web page. I do envisage that such a website requires > information to be donated from lots of people. But then so does a wiki. Scott, Paul and I are the Melbourne.pm.org maintainers. I wrote the current content mid last year after I got sick of no content at all (due to the pm.org issues). I'd be *delighted* to add more content if anyone wants to volunteer any. I'd even be willing to write further content on various topics if anyone indicates that they *want* it. Of course one of the big joys of having a wiki is that I don't have to be involved, you can all just add content as you go. You'll see that I've been keeping a meeting history. Feel free to add to it. > Me too. :-) If everyone thinks that the best way for that is the > redirect to the Melbourne_Perl_Mongers page then so be it. But I like > the idea of a separate web site more. Realising that we'd keep the Melbourne.pm.org domain, and it would just redirect (possibly after doing something cute in the same way that http://sydney.pm.org does): * why do you like the idea of a separate website more? * what would you like to appear on that separate website? All the best, Jacinta -- ("`-''-/").___..--''"`-._ | Jacinta Richardson | `6_ 6 ) `-. ( ).`-.__.`) | Perl Training Australia | (_Y_.)' ._ ) `._ `. ``-..-' | +61 3 9354 6001 | _..`--'_..-_/ /--'_.' ,' | contact at perltraining.com.au | (il),-'' (li),' ((!.-' | www.perltraining.com.au | From pjf at perltraining.com.au Tue Jan 10 00:25:50 2006 From: pjf at perltraining.com.au (Paul Fenwick) Date: Tue, 10 Jan 2006 19:25:50 +1100 Subject: [Melbourne-pm] Wikification of melbourne.pm.org? In-Reply-To: <20060110074656.GA6373@amygdala.woodheap.org> References: <43C3464A.3090508@perltraining.com.au> <20060110074656.GA6373@amygdala.woodheap.org> Message-ID: <43C36F8E.5010209@perltraining.com.au> G'day Ben, Ben Marsh wrote: > Admittedly the current Melbourne.pm.org maintainer seems very busy on > other things. I actually had the crazy thought that I might help > with/do the melbourne.pm.org web page. I do envisage that such a website requires > information to be donated from lots of people. But then so does a wiki. One of the problems with the existing pm.org system is that it's not quite as simple to make changes as one would like. Altering content is done using authenticated WebDav, which I don't use for anything else. Adding new webmasters means a lengthy registration process. In short, the barrier to making changes is high, and the number of people who can make changes are few. The wiki advantage is that everything's provided out of the box. Previewing, change control, collaborative editing, a simple mark-up syntax, RSS feeds and other goodies. People who aren't going to go to all the effort to submit (and chase) a request to a melbourne.pm webmaster will happily fix a wiki page. I will admit that I'm doubly-biased towards wikification of melbourne.pm.org. Not only do I think it will result in greater and better written content, I also think that it will mean less work for me personally, as I'm one of the few people with change access to the existing melbourne.pm.org website. Cheerio, Paul -- Paul Fenwick | http://perltraining.com.au/ Director of Training | Ph: +61 3 9354 6001 Perl Training Australia | Fax: +61 3 9354 2681 From guy at alchemy.com.au Tue Jan 10 01:24:32 2006 From: guy at alchemy.com.au (Guy Morton) Date: Tue, 10 Jan 2006 20:24:32 +1100 Subject: [Melbourne-pm] Wikification of melbourne.pm.org? In-Reply-To: <43C36BCA.8090903@perltraining.com.au> References: <43C3464A.3090508@perltraining.com.au> <20060110074656.GA6373@amygdala.woodheap.org> <43C36BCA.8090903@perltraining.com.au> Message-ID: <43C37D50.2020500@alchemy.com.au> If there is a problem (as there often is) with getting enough content to keep one web site fresh, it doesn't make sense to then try and maintain TWO sites. The way the Sydney site redirects is cute, as you say - damn those flamboyant Sydney-siders! Maybe we could redirect visitors with a random perl haiku? ;-) Guy Jacinta Richardson wrote: >Ben Marsh wrote: > > > >>What is YACMaS? >> >> > >Yet another content management system. We used to have a dynamic site until >pm.org got hacked one too many times and they decided to host static web pages only. > > > >>Granted I am only new here but I like the idea of a seperate site. >>Doesnt it look better for Melbourne Perl Mongers to have their own site? >> >> > >I really can't imagine it looking bad for Melbourne.pm.org to redirect to >anywhere else, but that might just be me. > > > >>Admittedly the current Melbourne.pm.org maintainer seems very busy on >>other things. I actually had the crazy thought that I might help >>with/do the melbourne.pm.org web page. I do envisage that such a website requires >>information to be donated from lots of people. But then so does a wiki. >> >> > >Scott, Paul and I are the Melbourne.pm.org maintainers. I wrote the current >content mid last year after I got sick of no content at all (due to the pm.org >issues). I'd be *delighted* to add more content if anyone wants to volunteer >any. I'd even be willing to write further content on various topics if anyone >indicates that they *want* it. > >Of course one of the big joys of having a wiki is that I don't have to be >involved, you can all just add content as you go. You'll see that I've been >keeping a meeting history. Feel free to add to it. > > > > >>Me too. :-) If everyone thinks that the best way for that is the >>redirect to the Melbourne_Perl_Mongers page then so be it. But I like >>the idea of a separate web site more. >> >> > >Realising that we'd keep the Melbourne.pm.org domain, and it would just redirect >(possibly after doing something cute in the same way that http://sydney.pm.org >does): > * why do you like the idea of a separate website more? > * what would you like to appear on that separate website? > >All the best, > > Jacinta > > > -------------- next part -------------- An HTML attachment was scrubbed... URL: http://mail.pm.org/pipermail/melbourne-pm/attachments/20060110/6ff04966/attachment.html From alfiejohn at acm.org Tue Jan 10 05:26:22 2006 From: alfiejohn at acm.org (Alfie John) Date: Wed, 11 Jan 2006 00:26:22 +1100 Subject: [Melbourne-pm] Wikification of melbourne.pm.org? In-Reply-To: <43C36BCA.8090903@perltraining.com.au> References: <43C3464A.3090508@perltraining.com.au> <20060110074656.GA6373@amygdala.woodheap.org> <43C36BCA.8090903@perltraining.com.au> Message-ID: On 10/01/2006, at 7:09 PM, Jacinta Richardson wrote: > > > Ben Marsh wrote: > >> What is YACMaS? > > Yet another content management system. We used to have a dynamic > site until > pm.org got hacked one too many times and they decided to host > static web pages only. > >> Granted I am only new here but I like the idea of a seperate site. >> Doesnt it look better for Melbourne Perl Mongers to have their own >> site? > > I really can't imagine it looking bad for Melbourne.pm.org to > redirect to > anywhere else, but that might just be me. > >> Admittedly the current Melbourne.pm.org maintainer seems very busy on >> other things. I actually had the crazy thought that I might help >> with/do the melbourne.pm.org web page. I do envisage that such a >> website requires >> information to be donated from lots of people. But then so does a >> wiki. > > Scott, Paul and I are the Melbourne.pm.org maintainers. I wrote > the current > content mid last year after I got sick of no content at all (due to > the pm.org > issues). I'd be *delighted* to add more content if anyone wants to > volunteer > any. I'd even be willing to write further content on various > topics if anyone > indicates that they *want* it. Maybe at the next meeting we could run a requirements analysis session on what we want PerlNet's wiki to do for us. Wiki content can grow at an enormous pace meaning maintenance nightmares arhh! A couple of guidelines and templates could help things go a bit smoother. Alfie > > Of course one of the big joys of having a wiki is that I don't have > to be > involved, you can all just add content as you go. You'll see that > I've been > keeping a meeting history. Feel free to add to it. > > >> Me too. :-) If everyone thinks that the best way for that is the >> redirect to the Melbourne_Perl_Mongers page then so be it. But I >> like >> the idea of a separate web site more. > > Realising that we'd keep the Melbourne.pm.org domain, and it would > just redirect > (possibly after doing something cute in the same way that http:// > sydney.pm.org > does): > * why do you like the idea of a separate website more? > * what would you like to appear on that separate website? > > All the best, > > Jacinta > > -- > ("`-''-/").___..--''"`-._ | Jacinta Richardson | > `6_ 6 ) `-. ( ).`-.__.`) | Perl Training Australia | > (_Y_.)' ._ ) `._ `. ``-..-' | +61 3 9354 6001 | > _..`--'_..-_/ /--'_.' ,' | contact at perltraining.com.au | > (il),-'' (li),' ((!.-' | www.perltraining.com.au | > > > _______________________________________________ > Melbourne-pm mailing list > Melbourne-pm at pm.org > http://mail.pm.org/mailman/listinfo/melbourne-pm From scottp at dd.com.au Tue Jan 10 14:37:06 2006 From: scottp at dd.com.au (Scott Penrose) Date: Wed, 11 Jan 2006 09:37:06 +1100 Subject: [Melbourne-pm] Wikification of melbourne.pm.org? In-Reply-To: <20060110074656.GA6373@amygdala.woodheap.org> References: <43C3464A.3090508@perltraining.com.au> <20060110074656.GA6373@amygdala.woodheap.org> Message-ID: On 10/01/2006, at 18:46, Ben Marsh wrote: > * Paul Fenwick (pjf at perltraining.com.au) wrote: >> G'day Melb.PM, >> The current melbourne.pm.org webpage looks rather plain, and is a >> far cry from >> the exciting site we had running under YACMaS. I'm like to see >> the site >> returned to its former glory, with calendars, talk descriptions, >> member >> profiles, summaries, reviews, and all the rest. > > What is YACMaS? > >> Now that perl.net is out in the open, we have a grand tool for >> making it easier >> for community editing to occur. There's already a Melbourne.PM >> page at >> http://perl.net.au/wiki/Melbourne_Perl_Mongers and even a meeting >> history at >> http://perl.net.au/wiki/Melbourne_Perl_Mongers/Meeting_History . >> However I >> believe this is just the tip of the iceberg; there's much more >> information that >> could be added regarding Melbourne.PM that hasn't. > > > Granted I am only new here but I like the idea of a seperate site. > Doesnt it look better for Melbourne Perl Mongers to have their own > site? > Don't get me wrong. I like wikis and I like perl.net.au. > > Admittedly the current Melbourne.pm.org maintainer seems very busy on > other things. I actually had the crazy thought that I might help > with/do the melbourne.pm.org web page. I do envisage that such a > website requires > information to be donated from lots of people. But then so does a > wiki. Ironically I believe Paul and Jacinta were maintaining it :-) but yes it is too much work without a wiki. A radical suggestion - Can we possibly run melbourne.pm.org on the same server as perl.net.au ? > >> >> What's my motivation for all this? Well, I'd like the >> Melbourne.PM website to >> be great, but most importantly, I want it to be better than >> Sydney.PM's pages. ;) >> > > Me too. :-) If everyone thinks that the best way for that is the > redirect to the Melbourne_Perl_Mongers page then so be it. But I like > the idea of a separate web site more. Yes. Melbourne.pm.org should be separate from perl.net.au - but it may be able to share. So when I say separate I think the URL should be separate and have its own wiki, but maybe we can still work on it together. Scott -- * - * http://www.osdc.com.au - Open Source Developers Conference * - * Scott Penrose VP in charge of Pancakes http://linux.dd.com.au/ scottp at dd.com.au Dismaimer: If you receive this email in error - please eat it immediately to prevent it from falling into the wrong hands. Please do not send me Word or PowerPoint attachments. See http://www.gnu.org/philosophy/no-word-attachments.html Microsoft is not the answer. It's the question. And the answer is no. -------------- next part -------------- A non-text attachment was scrubbed... Name: PGP.sig Type: application/pgp-signature Size: 186 bytes Desc: This is a digitally signed message part Url : http://mail.pm.org/pipermail/melbourne-pm/attachments/20060110/8f995ea2/PGP.bin From scottp at dd.com.au Tue Jan 10 14:41:17 2006 From: scottp at dd.com.au (Scott Penrose) Date: Wed, 11 Jan 2006 09:41:17 +1100 Subject: [Melbourne-pm] Wikification of melbourne.pm.org? In-Reply-To: <43C36BCA.8090903@perltraining.com.au> References: <43C3464A.3090508@perltraining.com.au> <20060110074656.GA6373@amygdala.woodheap.org> <43C36BCA.8090903@perltraining.com.au> Message-ID: On 10/01/2006, at 19:09, Jacinta Richardson wrote: > >> Granted I am only new here but I like the idea of a seperate site. >> Doesnt it look better for Melbourne Perl Mongers to have their own >> site? > > I really can't imagine it looking bad for Melbourne.pm.org to > redirect to > anywhere else, but that might just be me. I am afraid I disagree - I think it looks very second rate. It also looses our independence - which is very important and must not be lost or swallowed by larger groups, especially if it ends up with some commercial interest. Scott -- * - * http://www.osdc.com.au - Open Source Developers Conference * - * Scott Penrose Welcome to the Digital Dimension http://www.dd.com.au/ scottp at dd.com.au Dismaimer: Contents of this mail and signature are bound to change randomly. Whilst every attempt has been made to control said randomness, the author wishes to remain blameless for the number of eggs that damn chicken laid. Oh and I don't want to hear about butterflies either. Please do not send me Word or PowerPoint attachments. See http://www.gnu.org/philosophy/no-word-attachments.html Microsoft is not the answer. It's the question. And the answer is no. -------------- next part -------------- A non-text attachment was scrubbed... Name: PGP.sig Type: application/pgp-signature Size: 186 bytes Desc: This is a digitally signed message part Url : http://mail.pm.org/pipermail/melbourne-pm/attachments/20060110/e38af76a/PGP.bin From scottp at dd.com.au Tue Jan 10 14:44:11 2006 From: scottp at dd.com.au (Scott Penrose) Date: Wed, 11 Jan 2006 09:44:11 +1100 Subject: [Melbourne-pm] Wikification of melbourne.pm.org? In-Reply-To: <43C36F8E.5010209@perltraining.com.au> References: <43C3464A.3090508@perltraining.com.au> <20060110074656.GA6373@amygdala.woodheap.org> <43C36F8E.5010209@perltraining.com.au> Message-ID: On 10/01/2006, at 19:25, Paul Fenwick wrote: > G'day Ben, > > Ben Marsh wrote: > >> Admittedly the current Melbourne.pm.org maintainer seems very busy on >> other things. I actually had the crazy thought that I might help >> with/do the melbourne.pm.org web page. I do envisage that such a >> website requires >> information to be donated from lots of people. But then so does a >> wiki. > > One of the problems with the existing pm.org system is that it's > not quite as > simple to make changes as one would like. Altering content is done > using > authenticated WebDav, which I don't use for anything else. Adding new > webmasters means a lengthy registration process. In short, the > barrier to > making changes is high, and the number of people who can make > changes are few. > > The wiki advantage is that everything's provided out of the box. > Previewing, > change control, collaborative editing, a simple mark-up syntax, RSS > feeds and > other goodies. People who aren't going to go to all the effort to > submit (and > chase) a request to a melbourne.pm webmaster will happily fix a > wiki page. > > I will admit that I'm doubly-biased towards wikification of > melbourne.pm.org. > Not only do I think it will result in greater and better written > content, I also > think that it will mean less work for me personally, as I'm one of > the few > people with change access to the existing melbourne.pm.org website. Agreed. I think a wiki is perfect for Melbourne PM. Also we can point the melbourne.pm.org domain to any IP number no problem, it is only the services provided by pm.org that are static, a redirect is not necessary - a separate site is though. If PTA could run the melbourne.pm.org site along side perl.net.au - then it would be the same to administer, theoretically no extra load (since you were talking about a redirect anyway) and should have a link at the bottom saying that PTA and perl.net.au are hosting it. I think it is a little disturbing that we are effectively forming a splinter faction rather than trying to use pm.org like we should, and I think Sydney has made a mistake. Scott -------------- next part -------------- A non-text attachment was scrubbed... Name: PGP.sig Type: application/pgp-signature Size: 186 bytes Desc: This is a digitally signed message part Url : http://mail.pm.org/pipermail/melbourne-pm/attachments/20060110/96bac4c5/PGP.bin From scottp at dd.com.au Tue Jan 10 15:06:57 2006 From: scottp at dd.com.au (Scott Penrose) Date: Wed, 11 Jan 2006 10:06:57 +1100 Subject: [Melbourne-pm] Wikification of melbourne.pm.org? In-Reply-To: References: <43C3464A.3090508@perltraining.com.au> <20060110074656.GA6373@amygdala.woodheap.org> <43C36F8E.5010209@perltraining.com.au> Message-ID: On 11/01/2006, at 9:44, Scott Penrose wrote: > > I think it is a little disturbing that we are effectively forming a > splinter faction rather than trying to use pm.org like we should, > and I think Sydney has made a mistake. > Sorry, I wanted to add - this is only in reference to the Perl Mongers - not the rest of Perl.net.au - which rocks !!! Scooter -- * - * http://www.osdc.com.au - Open Source Developers Conference * - * Scott Penrose Welcome to the Digital Dimension http://www.dd.com.au/ scottp at dd.com.au Dismaimer: Contents of this mail and signature are bound to change randomly. Whilst every attempt has been made to control said randomness, the author wishes to remain blameless for the number of eggs that damn chicken laid. Oh and I don't want to hear about butterflies either. Please do not send me Word or PowerPoint attachments. See http://www.gnu.org/philosophy/no-word-attachments.html Microsoft is not the answer. It's the question. And the answer is no. -------------- next part -------------- A non-text attachment was scrubbed... Name: PGP.sig Type: application/pgp-signature Size: 186 bytes Desc: This is a digitally signed message part Url : http://mail.pm.org/pipermail/melbourne-pm/attachments/20060110/5d6fc546/PGP.bin From scottp at dd.com.au Tue Jan 10 16:54:57 2006 From: scottp at dd.com.au (Scott Penrose) Date: Wed, 11 Jan 2006 11:54:57 +1100 Subject: [Melbourne-pm] Perl Mongers next week - Wednesday 18th of January - 6:30pm - Lightning Talks Message-ID: <96C83FBB-2D37-475F-99D6-800FA597B41D@dd.com.au> Hey Perl Mongers. Reminder that Melbourne Perl Mongers is on again next week - Wednesday 18th of January at 6:30pm. This month we were planning on doing Lightning Talks. A few volunteers have already made contributions, but you are all welcome to come up and do a talk. The rules - 7 minutes is your maximum - including setup time. If you have two talks, you can do both, but don't expect them to run side by side - and part 1 of 2 is not allowed because that would be cheating :-) It should be fun. Please bring along your ideas, email them to the list if you like, and turn up next Wednesday week at 6:30 (the 3rd of the month). Where: myinternet Level 8, 14-20 Blackwood Street North Melbourne When: 6:30pm Wednesday 18th of January 2006 What: Lots of lightning talks ! After: Dinner :-) Scott -------------- next part -------------- A non-text attachment was scrubbed... Name: PGP.sig Type: application/pgp-signature Size: 186 bytes Desc: This is a digitally signed message part Url : http://mail.pm.org/pipermail/melbourne-pm/attachments/20060111/eb78491c/PGP.bin From simon at unisolve.com.au Tue Jan 10 19:11:13 2006 From: simon at unisolve.com.au (Simon Taylor) Date: Wed, 11 Jan 2006 14:11:13 +1100 Subject: [Melbourne-pm] Wikification of melbourne.pm.org? In-Reply-To: References: <43C3464A.3090508@perltraining.com.au> <20060110074656.GA6373@amygdala.woodheap.org> Message-ID: <200601111411.13278.simon@unisolve.com.au> On Wed, 11 Jan 2006 09:37 am, Scott Penrose wrote: > Yes. Melbourne.pm.org should be separate from perl.net.au - but it > may be able to share. So when I say separate I think the URL should > be separate and have its own wiki, but maybe we can still work on it > together. The sydney.pm re-direct is rather cute, but I also believe that we should present the melbourne.pm site as a stand-alone entity. However, making it a wiki would be a positive step. - Simon -- Unisolve Pty Ltd - Melbourne, Australia +61 3 9568 2005 From stephen at sydney.pm.org Tue Jan 10 19:16:16 2006 From: stephen at sydney.pm.org (Stephen Steneker) Date: Wed, 11 Jan 2006 14:16:16 +1100 Subject: [Melbourne-pm] Wikification of melbourne.pm.org? In-Reply-To: References: <43C3464A.3090508@perltraining.com.au> <20060110074656.GA6373@amygdala.woodheap.org> <43C36F8E.5010209@perltraining.com.au> Message-ID: Hi Scott, Having some familiarity with the decision to move Sydney.pm to a shared wiki, felt it was worth sharing longer comment on some of the concerns raised... > I think it is a little disturbing that we are effectively forming a > splinter faction rather than trying to use pm.org like we should, and I > think Sydney has made a mistake. I'd have to strongly disagree with your feeling of "splinter faction". I actually see this as a healthy unification of the AU perl community (mongers and otherwise) which allows everyone the opportunity to contribute rather than relying on a few "leaders" who are generally keen but often pressed for time. We are not changing the identity of our perl mongers group, or moving away from pm.org services like mailing lists and group listings. In fact, the reason pm.org provides these services is for groups who do *not* have access to their own infrastructure: http://www.pm.org/faq/. Well established groups like london.pm.org manage their own hosting and mailing lists. The intent is to create a perl.net.au wiki resource for perl, not a perl.net.au user group organisation. If Hobart decides to form a user group, they will be registering via pm.org. > A radical suggestion - Can we possibly run melbourne.pm.org on > the same server as perl.net.au ? Hey, isn't that asking perl.net.au to be a splinter faction offering services that should be provided by pm.org ? :D The benefits of a shared wiki are: - we can easily link, share, and search information between different groups - by using a common environment, we can share templates and assist other groups - we foster a sense that there is an active perl community which is not limited to isolated pockets of activity - we encourage a wider group of contributors with an easy way to get involved - we become more aware of other user groups, developers, and companies who are doing great things with perl closer to home - visiting speakers may consider more than one user group when making a trip to Australia (hint: how many overseas visitors to OSDC presented to user groups outside of Melbourne?) You can (and definitely should!) maintain a Melbourne perl monger identity within the larger community .. melbourne.pm.org should be the advertised entry point. In actual fact, I would eventually love to see a wiki.pm.org! Unfortunately, given the general disrepair of perl monger group sites (e.g. the need for an annual "are you alive" survey!) I think it would be very difficult to convince anyone that a centralised site would have much more chance of success. I envision the perl.net.au effort will help convince folks that pm.org user groups can rise above their general apathy and isolation to form a strong community. > It also looses our independence - which is very important and > must not be lost or swallowed by larger groups, especially if > it ends up with some commercial interest. An obvious concern, and one which must be kept in mind. At the moment I trust that those involved (particularly Perl Training Australia) are doing so out of a genuine interest to help the perl community. A good central resource can help user groups, employers, and companies offering perl-related services. The current site incarnation is not overwhelmed by Perl Training Australia, but does provide some context for themselves (and others) to promote their services. If things become uncomfortably commercial or unworkably complex, anyone is free to relocate their content elsewhere (and there are tools to easily sync pages with mediawiki for local backup and editing). This is also a community that *you* can participate in and help keep it on track. Interestingly enough, the camel which is favoured by user groups is actually a commercial trademark of O'Reilly. The perl foundation designed a new logo which is meant to replace this: http://www.perlfoundation.org/legal/trademark.html In any case, hope this has been helpful feedback .. ! Cheers, Stephen From scottp at dd.com.au Tue Jan 10 19:42:44 2006 From: scottp at dd.com.au (Scott Penrose) Date: Wed, 11 Jan 2006 14:42:44 +1100 Subject: [Melbourne-pm] Wikification of melbourne.pm.org? In-Reply-To: References: <43C3464A.3090508@perltraining.com.au> <20060110074656.GA6373@amygdala.woodheap.org> <43C36F8E.5010209@perltraining.com.au> Message-ID: <186DAAD5-DF06-46D4-BAF6-6C2597B0CDBC@dd.com.au> On 11/01/2006, at 14:16, Stephen Steneker wrote: > > > A radical suggestion - Can we possibly run melbourne.pm.org on > > the same server as perl.net.au ? > > Hey, isn't that asking perl.net.au to be a splinter faction > offering services that should be provided by pm.org ? :D No absolutely not ! Unfortunately, and I blame the brevity of mail in this case, you have completely missed my point. I don't mind support and help from any group or company, as long as the content and appearance remains independent. I can spend quite a bit of time explaining just why such an affiliation is bad, and site you many examples of where these types of things have failed in the past, but in the end I probably won't convince you with the words I write here - just like you have not convinced me at all, probably in some ways strengthened some of my concerns. Arguments are good and important to have on topics but sometimes we have to realise that not everyone will be convinced. Do you know we actually lost members when we became an association. Even though it is required now by government, some people didn't agree - we can't make everyone happy :-( Anyway... This is all about appearance - and control - I never mentioned having to run code on pm.org - you have made that bit up :-) Thus my comment about mail being easy to misunderstand. PM.org servers can be run anywhere - we have many times thought of running our own. I would expect that the company running that server would have a sponsorship link on it. I think pm.org needs a new system - like PHP - a system that is shared across all groups ... thus > The benefits of a shared wiki are: > - we can easily link, share, and search information between > different groups > - by using a common environment, we can share templates and > assist other groups > - we foster a sense that there is an active perl community > which is not limited to isolated pockets of activity > - we encourage a wider group of contributors with an easy > way to get involved > - we become more aware of other user groups, developers, and > companies who are doing great things with perl closer to home > - visiting speakers may consider more than one user group > when making a trip to Australia (hint: how many overseas > visitors to OSDC presented to user groups outside of > Melbourne?) provide all these advantages (which I agree is good) across PM groups. perl.net.au is for all perl in Australia, not all PM - there is a huge difference - and so there should be. PM is not about companies. perl.net.au has filled that niche very nicely. One of the many things groups get wrong, especially large communities, is not focusing on a small enough area. I believe perl.net.au is about bringing all perl communities and business together across Australia and NZ - as it says on the site. It is therefore larger and out of scope for Perl Mongers. This type of broad areas causes the small groups to be swallowed up. > You can (and definitely should!) maintain a Melbourne perl monger > identity within the larger community .. melbourne.pm.org should > be the advertised entry point. If you redirect - as you are - that will not stay that way. You can't control that, it is just the way of the internet. Eventually, especially if we all do it, the PM groups in Australia will loose their identity and become Perl Net meetings. > In actual fact, I would eventually love to see a wiki.pm.org! Exactly my point. I would be happy if this discussion was about bringing the pm.org groups together - I think it would be daft to have a single wiki though - even MediaWiki splits its' content up. > Unfortunately, given the general disrepair of perl monger group > sites (e.g. the need for an annual "are you alive" survey!) I > think it would be very difficult to convince anyone that a > centralised site would have much more chance of success. Yes - although the PHP group - who are central have the same problems - in the end what is important is that it is easy to maintain. The fact that it is separate or grouped together will not produce any more likelihood of updates (see any of the big communities). > I envision the perl.net.au effort will help convince folks that pm.org > user groups can rise above their general apathy and isolation > to form a strong community. Yep, I fully support that, and really hope that it does. This discussion neither weakens nor strengthens that at all. > > It also looses our independence - which is very important and > > must not be lost or swallowed by larger groups, especially if > > it ends up with some commercial interest. > > An obvious concern, and one which must be kept in mind. At the > moment I trust that those involved (particularly Perl Training > Australia) are doing so out of a genuine interest to help the > perl community. A good central resource can help user groups, > employers, and companies offering perl-related services. > > Interestingly enough, the camel which is favoured by user groups > is actually a commercial trademark of O'Reilly. The perl > foundation designed a new logo which is meant to replace this: > http://www.perlfoundation.org/legal/trademark.html If you read carefully and understand - that is a trademark for a good reason. For the same reason you keep copyright in the company that created software even if it is released as GNU - same thing. The camel is registered to protect it, like you copyright your code to whom ever you wrote it for/with/yourself to protect it. > In any case, hope this has been helpful feedback .. ! Absolutely. About 90% of what you write I agree with :-) I am not asking for something very different, just separate. Thanks Scott -- * - * http://www.osdc.com.au - Open Source Developers Conference * - * Scott Penrose Open source developer http://linux.dd.com.au/ scottp at dd.com.au Dismaimer: Open sauce usually ends up never coming out (of the bottle). Please do not send me Word or PowerPoint attachments. See http://www.gnu.org/philosophy/no-word-attachments.html Microsoft is not the answer. It's the question. And the answer is no. -------------- next part -------------- A non-text attachment was scrubbed... Name: PGP.sig Type: application/pgp-signature Size: 186 bytes Desc: This is a digitally signed message part Url : http://mail.pm.org/pipermail/melbourne-pm/attachments/20060111/5c41438b/PGP.bin From blm at woodheap.org Wed Jan 11 02:10:24 2006 From: blm at woodheap.org (Ben Marsh) Date: Wed, 11 Jan 2006 21:10:24 +1100 Subject: [Melbourne-pm] Lightning Talks Message-ID: <20060111101024.GC28717@amygdala.woodheap.org> Hello all, I am only new to perl mongers. However I would still be interested in doing a lightening talk. However I dont want to bore everyone to tears. Is there anyway to find out what people would be interested in hearing about? The ideas that I was working on were: * Translating VBScript and JScript to Perl and Win32::OLE There are many opurtunities for automation and other scripting on the Win32 platform. However much of the documentation and examples on the Internet deal with VBScript implementation. This talk will describe with examples the Perl implementation of such scripts. * What I use Perl for. The things that I have done with perl. Automation of common Win32 server tasks. Also hobby based linux scripting and web scraping. Thats the best I can do at the moment. If these are uninteresting let me know. Also if there are things that you want to know about but dont have enough time to research yourself maybe you could let me know? Any thoughts appreciated, Ben Marsh From david_dick at iprimus.com.au Wed Jan 11 02:16:42 2006 From: david_dick at iprimus.com.au (David Dick) Date: Wed, 11 Jan 2006 21:16:42 +1100 Subject: [Melbourne-pm] Lightning Talks In-Reply-To: <20060111101024.GC28717@amygdala.woodheap.org> References: <20060111101024.GC28717@amygdala.woodheap.org> Message-ID: <43C4DB0A.1000806@iprimus.com.au> Ben Marsh wrote: > Thats the best I can do at the moment. If these are uninteresting let > me know. Also if there are things that you want to know about but dont > have enough time to research yourself maybe you could let me know? I've been using Win32::OLE to access WMI services at work. In my spare time i've been trying to figure out how to use ADSI to install/update web applications on IIS???? :) One of the chief interesting things i've found about WMI at least is the difficulty in enumerating all the properties of different objects. I'm still at the stage of googling for "WMI script + xxxx" and adapting the resulting vb code to perl. hope this helps -Dave From jarich at perltraining.com.au Wed Jan 11 02:42:30 2006 From: jarich at perltraining.com.au (Jacinta Richardson) Date: Wed, 11 Jan 2006 21:42:30 +1100 Subject: [Melbourne-pm] Lightning Talks In-Reply-To: <20060111101024.GC28717@amygdala.woodheap.org> References: <20060111101024.GC28717@amygdala.woodheap.org> Message-ID: <43C4E116.505@perltraining.com.au> Ben Marsh wrote: > Hello all, > > I am only new to perl mongers. However I would still be interested in > doing a lightening talk. However I dont want to bore everyone to tears. One of the great things about lightning talks is that they only take 7 minutes. This means it's easy to keep the audiences' attention for the majority of your talk. At worst, even if someone is an appallingly bad speaker, or is talking about a dreadfully boring topic, all of the audience knows that it'll be over in just a few minutes. :) As a result, you won't be able to bore us to tears unless you try really hard. So fear not! > Is there anyway to find out what people would be interested in hearing > about? I collected some requested talk ideas over on http://perl.net.au/wiki/Melbourne_Perl_Mongers/Meeting_History a couple of months ago, but I'm note sure that they're all appropriate for lightning talks. If anyone wants to add to that list, it would be great. If we get a lot of additions, we should put it on its own page. > The ideas that I was working on were: > > * Translating VBScript and JScript to Perl and Win32::OLE Sounds short and sweet. > * What I use Perl for. Likewise. I look forward to hearing them. Jacinta -- ("`-''-/").___..--''"`-._ | Jacinta Richardson | `6_ 6 ) `-. ( ).`-.__.`) | Perl Training Australia | (_Y_.)' ._ ) `._ `. ``-..-' | +61 3 9354 6001 | _..`--'_..-_/ /--'_.' ,' | contact at perltraining.com.au | (il),-'' (li),' ((!.-' | www.perltraining.com.au | From ajsavige at yahoo.com.au Wed Jan 11 04:50:52 2006 From: ajsavige at yahoo.com.au (Andrew Savige) Date: Wed, 11 Jan 2006 23:50:52 +1100 (EST) Subject: [Melbourne-pm] Wikification of melbourne.pm.org? In-Reply-To: Message-ID: <20060111125052.36560.qmail@web36111.mail.mud.yahoo.com> --- Scott Penrose wrote: > I think it is a little disturbing that we are effectively > forming a splinter faction rather than trying to use pm.org > like we should, and I think Sydney has made a mistake. As a Sydney.pm member, I applaud what Stennie did. I'm always short of time and I've found perl.net.au to be a perfect fit for my needs, allowing me to quickly and easily contribute to Sydney.pm web site while keeping up to date with what other Antipodean Perl Mongers are up to. Given most of us are so short of time, I just don't see the sense in maintaining two sets of information about Sydney.pm. Let's wait and see. I hope that the benefits of unification, sharing and sense of Antipodean community will far outweigh any perceived loss of individual group identity. /-\ ____________________________________________________ Do you Yahoo!? Never miss an Instant Message - Yahoo! Messenger for SMS http://au.mobile.yahoo.com/mweb/index.html From alecclews at gmail.com Wed Jan 11 04:50:56 2006 From: alecclews at gmail.com (Alec Clews) Date: Wed, 11 Jan 2006 23:50:56 +1100 Subject: [Melbourne-pm] Lightning Talks In-Reply-To: <20060111101024.GC28717@amygdala.woodheap.org> Message-ID: <43c4ff3f.77d929e1.4aad.096d@mx.gmail.com> Ben, > * Translating VBScript and JScript to Perl and Win32::OLE > > There are many opurtunities for automation and other scripting on the > Win32 platform. However much of the documentation and examples on the > Internet deal with VBScript implementation. This talk will describe > with examples the Perl implementation of such scripts. This sounds very useful, there are many of us stuck on Windows, and looks worth more than 7 minutes. Perhaps it can start as a lightning talk and then get expanded for later in the year or even OSDC 2006? Regards -- Online Contact details: Alec Clews, Skype ---------------- alecclews Melbourne, Australia MSN -------- aclews at serena.com Jabber -- alecclews at jabber.org.au http://alecc.blogspot.com/ Yahoo ---------------- alecclews http://www.geocities.com/alecclews/ GTalk ---------------- alecclews Voice: +61-(0)425-770-886 AIM ------------- alecclewsaus From pjf at perltraining.com.au Wed Jan 11 17:13:46 2006 From: pjf at perltraining.com.au (Paul Fenwick) Date: Thu, 12 Jan 2006 12:13:46 +1100 Subject: [Melbourne-pm] PerlNet Message-ID: <43C5AD4A.30405@perltraining.com.au> G'day Everyone, It looks like we've had a very active discussion regarding both PerlNet, and the Melbourne.PM website in general. I'm glad to see this is getting so much interest, and I hope to use this opportunity to clarify a few matters. There's a page at http://perl.net.au/wiki/PerlNet:PerlNet_and_User_Groups which describes PerlNet and what it can and cannot do for user-groups. I hope that it answers a number of questions that have been raised on the list. If you have a question that it doesn't answer, then please feel free to add it to that page, mention it on the discussion page, or even raise it with myself directly. I want to stress that all content on PerlNet is governed by an open source license (http://perl.net.au/wiki/PerlNet:Copyrights). Everyone is welcome to use PerlNet content and contributions for their own sites and projects. If Melbourne.PM members would like to develop content or ideas with the intention of these appearing on the main Melbourne.PM or other sites, then PerlNet provides a simple way of doing this right now with a minimum of effort and overhead. If Melbourne.PM does not wish to use PerlNet as it main page, then Melb.PM and its members are still welcome and encouraged to take advantage of the other features of the site. Melbourne.PM is also very welcome to use PerlNet as a main page on a temporary basis, to provide easy collaboration and content until the final Melbourne.PM site is ready. Pages can be generated without the wiki navigation by adding '?printable=yes' to any URL, and there are both Perl modules and other tools for extracting and exporting wiki content. For everyone who has contributed to PerlNet already, thank-you! There's been a few suggestions that Perl Training Australia could provide separate hosting facilities for Melbourne.pm. I consider this to be a different discussion to that of PerlNet. Using PerlNet in any way does not preclude or exclude any future hosting arrangements for Melbourne.pm in any way. There was a suggestion that hosting two wikis on the same server would be little more effort than hosting one. Unfortunately, in my experience this isn't as easy as it sounds. Two wikis usually means separate maintenance and monitoring for each, and very often the separate development of tools, policies, templates, and contributors. All the very best, Paul -- Paul Fenwick | http://perltraining.com.au/ Director of Training | Ph: +61 3 9354 6001 Perl Training Australia | Fax: +61 3 9354 2681 From scottp at dd.com.au Wed Jan 11 17:38:58 2006 From: scottp at dd.com.au (Scott Penrose) Date: Thu, 12 Jan 2006 12:38:58 +1100 Subject: [Melbourne-pm] PerlNet In-Reply-To: <43C5AD4A.30405@perltraining.com.au> References: <43C5AD4A.30405@perltraining.com.au> Message-ID: <68ED0EFA-4E03-4560-871A-FF00B380EC04@dd.com.au> Thanks Paul. You and Jacinta have made a great site. Scott On 12/01/2006, at 12:13, Paul Fenwick wrote: > G'day Everyone, > > > > It looks like we've had a very active discussion regarding both > PerlNet, and the > Melbourne.PM website in general. I'm glad to see this is getting > so much > interest, and I hope to use this opportunity to clarify a few matters. > > There's a page at http://perl.net.au/wiki/ > PerlNet:PerlNet_and_User_Groups which > describes PerlNet and what it can and cannot do for user-groups. I > hope that it > answers a number of questions that have been raised on the list. > If you have a > question that it doesn't answer, then please feel free to add it to > that page, > mention it on the discussion page, or even raise it with myself > directly. > > I want to stress that all content on PerlNet is governed by an open > source > license (http://perl.net.au/wiki/PerlNet:Copyrights). Everyone is > welcome to > use PerlNet content and contributions for their own sites and > projects. If > Melbourne.PM members would like to develop content or ideas with > the intention > of these appearing on the main Melbourne.PM or other sites, then > PerlNet > provides a simple way of doing this right now with a minimum of > effort and overhead. > > If Melbourne.PM does not wish to use PerlNet as it main page, then > Melb.PM and > its members are still welcome and encouraged to take advantage of > the other > features of the site. Melbourne.PM is also very welcome to use > PerlNet as a > main page on a temporary basis, to provide easy collaboration and > content until > the final Melbourne.PM site is ready. Pages can be generated > without the wiki > navigation by adding '?printable=yes' to any URL, and there are > both Perl > modules and other tools for extracting and exporting wiki content. > > For everyone who has contributed to PerlNet already, thank-you! > > > > > > There's been a few suggestions that Perl Training Australia could > provide > separate hosting facilities for Melbourne.pm. I consider this to > be a different > discussion to that of PerlNet. Using PerlNet in any way does not > preclude or > exclude any future hosting arrangements for Melbourne.pm in any way. > > There was a suggestion that hosting two wikis on the same server > would be little > more effort than hosting one. Unfortunately, in my experience this > isn't as > easy as it sounds. Two wikis usually means separate maintenance > and monitoring > for each, and very often the separate development of tools, > policies, templates, > and contributors. > > > > All the very best, > > Paul > > -- > Paul Fenwick | http://perltraining.com.au/ > Director of Training | Ph: +61 3 9354 6001 > Perl Training Australia | Fax: +61 3 9354 2681 > > _______________________________________________ > Melbourne-pm mailing list > Melbourne-pm at pm.org > http://mail.pm.org/mailman/listinfo/melbourne-pm -- * - * http://www.osdc.com.au - Open Source Developers Conference * - * Scott Penrose Welcome to the Digital Dimension http://www.dd.com.au/ scottp at dd.com.au Dismaimer: Contents of this mail and signature are bound to change randomly. Whilst every attempt has been made to control said randomness, the author wishes to remain blameless for the number of eggs that damn chicken laid. Oh and I don't want to hear about butterflies either. Please do not send me Word or PowerPoint attachments. See http://www.gnu.org/philosophy/no-word-attachments.html Microsoft is not the answer. It's the question. And the answer is no. -------------- next part -------------- A non-text attachment was scrubbed... Name: PGP.sig Type: application/pgp-signature Size: 186 bytes Desc: This is a digitally signed message part Url : http://mail.pm.org/pipermail/melbourne-pm/attachments/20060112/9e8b1971/PGP.bin From blm at woodheap.org Sun Jan 15 04:13:39 2006 From: blm at woodheap.org (Ben Marsh) Date: Sun, 15 Jan 2006 23:13:39 +1100 Subject: [Melbourne-pm] Lightning Talks Message-ID: <20060115121338.GA23562@amygdala.woodheap.org> Will the internet be accessible for the speaker at the Lightning talks? Thanks, Ben Marsh From scottp at dd.com.au Sun Jan 15 15:16:54 2006 From: scottp at dd.com.au (Scott Penrose) Date: Mon, 16 Jan 2006 10:16:54 +1100 Subject: [Melbourne-pm] Lightning Talks In-Reply-To: <20060115121338.GA23562@amygdala.woodheap.org> References: <20060115121338.GA23562@amygdala.woodheap.org> Message-ID: <5D64A232-B075-4CE4-B183-B7FC6F1BDB77@dd.com.au> On 15/01/2006, at 23:13, Ben Marsh wrote: > Will the internet be accessible for the speaker at the Lightning > talks? Yes. I have a screen switch - I will setup one machine with internet access that can stay there, and people can plug in a laptop if they want too :-) See ya Wednesday. Scooter -------------- next part -------------- A non-text attachment was scrubbed... Name: PGP.sig Type: application/pgp-signature Size: 186 bytes Desc: This is a digitally signed message part Url : http://mail.pm.org/pipermail/melbourne-pm/attachments/20060115/512ac559/PGP.bin From david_dick at iprimus.com.au Tue Jan 17 03:40:56 2006 From: david_dick at iprimus.com.au (David Dick) Date: Tue, 17 Jan 2006 22:40:56 +1100 Subject: [Melbourne-pm] Deploying perl web applications to end users Message-ID: <43CCD7C8.9040209@iprimus.com.au> For a while now, i've been building an application at home as a testing ground for a lot of ideas that i can use for my job. It's a perl web application with a database backend, so fairly basic stuff. I'm at a stage where i think i might be only a couple of years out from having a product that actually does something useful :$ so i'm thinking about deployment to potential customers. Two possibilities here (afaik) that boil down to the same issue (afaik). Possibility one is customer installs on a local machine. Issue: need to ensure a quick, easy install to make sure s/he actually uses the software. Possibility two is we admin server farms for the application. Issue, need to ensure quick, easy install to remove the possibility of stuff ups. Looking at two known perl web applications, Bricolage and Movable Type, they have installation instructions at http://www.bricolage.cc/docs/current/api/Bric::Admin and http://www.sixapart.com/movabletype/docs/mtinstall.html#initializing%20the%20system that oddly enough to me, don't seem to be all that smooth as far as a "one click" sort of installation goes. Has anyone had an experience deploying perl web applications in large numbers? Did you use Wix, rpm, apt-get, Module::Build, movable type/Bricolage type instructions? How did it go? From scottp at dd.com.au Tue Jan 17 14:31:46 2006 From: scottp at dd.com.au (Scott Penrose) Date: Wed, 18 Jan 2006 09:31:46 +1100 Subject: [Melbourne-pm] Deploying perl web applications to end users In-Reply-To: <43CCD7C8.9040209@iprimus.com.au> References: <43CCD7C8.9040209@iprimus.com.au> Message-ID: <79E61B24-11E6-4358-89A9-E7AD2DA5DF6D@dd.com.au> On 17/01/2006, at 22:40, David Dick wrote: > For a while now, i've been building an application at home as a > testing > ground for a lot of ideas that i can use for my job. It's a perl web > application with a database backend, so fairly basic stuff. > > I'm at a stage where i think i might be only a couple of years out > from > having a product that actually does something useful :$ so i'm > thinking > about deployment to potential customers. > > Two possibilities here (afaik) that boil down to the same issue > (afaik). > Possibility one is customer installs on a local machine. Issue: > need > to ensure a quick, easy install to make sure s/he actually uses the > software. Possibility two is we admin server farms for the > application. > Issue, need to ensure quick, easy install to remove the > possibility of > stuff ups. Two things to review. TWIKI has a untar, go to URL type installation. But if you want to be really transparent, try considering PAR - Think TAR or JAR only Perl - it is designed to distribute an entire system in a single file, you could even have multiple versions of the code installed there, and it allows data files to be included. Tell us how you go, because this is definitely an issue. One of my problems is that CPAN asks too many questions - I can't remember the time I didn't just press ENTER on any question. This is a problem that Debian has on install too - and is starting to be realised by the community. CPAN is trying to fix this though with the META files, so that a tree of dependencies can be built in advance, rather than one node at a time, like it currently is. Good luck Scott -- * - * http://www.osdc.com.au - Open Source Developers Conference * - * Scott Penrose Open source developer http://linux.dd.com.au/ scottp at dd.com.au Dismaimer: Open sauce usually ends up never coming out (of the bottle). Please do not send me Word or PowerPoint attachments. See http://www.gnu.org/philosophy/no-word-attachments.html Microsoft is not the answer. It's the question. And the answer is no. -------------- next part -------------- A non-text attachment was scrubbed... Name: PGP.sig Type: application/pgp-signature Size: 186 bytes Desc: This is a digitally signed message part Url : http://mail.pm.org/pipermail/melbourne-pm/attachments/20060117/b50a2aad/PGP.bin From jarich at perltraining.com.au Sun Jan 22 16:05:49 2006 From: jarich at perltraining.com.au (Jacinta Richardson) Date: Mon, 23 Jan 2006 11:05:49 +1100 Subject: [Melbourne-pm] Details of last meeting Message-ID: <43D41DDD.7060907@perltraining.com.au> For those who missed our lightning talk meeting, you can see a short description of the fun talks you missed out on at: http://perl.net.au/wiki/Melbourne_Perl_Mongers/Meeting_History_2006 For those who presented these talks, or paid better attention than I did, feel free to correct any mistakes. All the best, Jacinta -- ("`-''-/").___..--''"`-._ | Jacinta Richardson | `6_ 6 ) `-. ( ).`-.__.`) | Perl Training Australia | (_Y_.)' ._ ) `._ `. ``-..-' | +61 3 9354 6001 | _..`--'_..-_/ /--'_.' ,' | contact at perltraining.com.au | (il),-'' (li),' ((!.-' | www.perltraining.com.au | From dgolembo at unimelb.edu.au Sun Jan 22 19:58:56 2006 From: dgolembo at unimelb.edu.au (Daniel Golembo) Date: Mon, 23 Jan 2006 14:58:56 +1100 Subject: [Melbourne-pm] SOAP::Lite Serializing Complex objects. Message-ID: <43D45480.20403@unimelb.edu.au> I have a problem serializing complex objects while using SOAP::Lite. The issue is, if I have a complex object, that iteslf contains a complex object, the second object does not get correctly encoded in the SOAP envelope. For Example, I have an object that needs to be sent as an argument in a SOAP call: my $expression = bless( { 'descriptor' => bless( { 'sum' => '1000', 'description' => 'Adding 900 to 100' }, 'sample_Descriptor' ), 'a' => '900', 'b' => '100', 'dateAdded' => '2006-01-23T03:32:03.781Z' }, 'sample_SumExpression' ); and I have SOAP serializer functions to serialize sample_SumExpression and sample_Descriptor like: sub SOAP::Serializer::as_sample_SumExpression { my $self = shift; my($value, $name, $type, $attr) = @_; my $a=SOAP::Data->new(value=>$value->{a}, type=>'string'); my $b=SOAP::Data->new(value=>$value->{b}, type=>'string'); my $dateAdded=SOAP::Data->new(value=>$value->{dateAdded}, type=>'dateTime'); my $descriptor=SOAP::Data->new(value=>$value->{descriptor}, type=>'ns1:sample_Descriptor'); my $rval=[$name, {'xsi:type'=>'ns1:sample_SumExpression', %$attr}, {a=>$a, b=>$b, dateAdded=>$dateAdded, descriptor=>$descriptor}]; return $rval; } and sub SOAP::Serializer::as_sample_Descriptor { print "Firing SOAP::Serializer::as_unimelb_merlin_Descriptor\n"; my $self = shift; my($value, $name, $type, $attr) = @_; my $description=SOAP::Data->new(name=>'description', value=>$value->{description}, type=>'string'); my $sum=SOAP::Data->new(name=>'sum', value=>$value->{sum}, type=>'int'); my $rval=[$name, {'xsi:type'=>'ns1:sample_Descriptor', %$attr}, {description=>$description, sum=>$sum}]; return $rval; } When SOAP call happens, the envelope that is generated by SOAP::Lite is: HASH(0x9c7d55c)9001002006-01-23T03:32:03.781Z So, the Top level object, sample_Expression, gets encoded correctly, but the lower level object, sample_Descriptor, is not. I have also tried using encode_object function of SOAP::Serializer, but this resulted in sample_Descriptor being encoded as an array. I wonder what I am doing wrong or what SOAP::Lite is doing wrong. Please help. From Kathryn_M_Grant at national.com.au Sun Jan 22 20:28:38 2006 From: Kathryn_M_Grant at national.com.au (Kathryn_M_Grant@national.com.au) Date: Mon, 23 Jan 2006 15:28:38 +1100 Subject: [Melbourne-pm] SOAP::Lite Serializing Complex objects. In-Reply-To: <43D45480.20403@unimelb.edu.au> Message-ID: Hi Daniel I have encountered the exact same problem with nested complex objects, and never managed to get the damn thing to work. Hope you have better luck! Kat Daniel Golembo To Sent by: melbourne-pm at pm.org melbourne-pm-boun cc ces+kathryn_m_gra nt=national.com.a Subject u at pm.org [Melbourne-pm] SOAP::Lite Serializing Complex objects. 23/01/2006 02:58 PM I have a problem serializing complex objects while using SOAP::Lite. The issue is, if I have a complex object, that iteslf contains a complex object, the second object does not get correctly encoded in the SOAP envelope. For Example, I have an object that needs to be sent as an argument in a SOAP call: my $expression = bless( { 'descriptor' => bless( { 'sum' => '1000', 'description' => 'Adding 900 to 100' }, 'sample_Descriptor' ), 'a' => '900', 'b' => '100', 'dateAdded' => '2006-01-23T03:32:03.781Z' }, 'sample_SumExpression' ); and I have SOAP serializer functions to serialize sample_SumExpression and sample_Descriptor like: sub SOAP::Serializer::as_sample_SumExpression { my $self = shift; my($value, $name, $type, $attr) = @_; my $a=SOAP::Data->new(value=>$value->{a}, type=>'string'); my $b=SOAP::Data->new(value=>$value->{b}, type=>'string'); my $dateAdded=SOAP::Data->new(value=>$value->{dateAdded}, type=>'dateTime'); my $descriptor=SOAP::Data->new(value=>$value->{descriptor}, type=>'ns1:sample_Descriptor'); my $rval=[$name, {'xsi:type'=>'ns1:sample_SumExpression', %$attr}, {a=>$a, b=>$b, dateAdded=>$dateAdded, descriptor=>$descriptor}]; return $rval; } and sub SOAP::Serializer::as_sample_Descriptor { print "Firing SOAP::Serializer::as_unimelb_merlin_Descriptor\n"; my $self = shift; my($value, $name, $type, $attr) = @_; my $description=SOAP::Data->new(name=>'description', value=>$value->{description}, type=>'string'); my $sum=SOAP::Data->new(name=>'sum', value=>$value->{sum}, type=>'int'); my $rval=[$name, {'xsi:type'=>'ns1:sample_Descriptor', %$attr}, {description=>$description, sum=>$sum}]; return $rval; } When SOAP call happens, the envelope that is generated by SOAP::Lite is: HASH(0x9c7d55c)9001002006-01-23T03:32:03.781 Z So, the Top level object, sample_Expression, gets encoded correctly, but the lower level object, sample_Descriptor, is not. I have also tried using encode_object function of SOAP::Serializer, but this resulted in sample_Descriptor being encoded as an array. I wonder what I am doing wrong or what SOAP::Lite is doing wrong. Please help. _______________________________________________ Melbourne-pm mailing list Melbourne-pm at pm.org http://mail.pm.org/mailman/listinfo/melbourne-pm From etaylor at realestate.com.au Mon Jan 23 14:56:58 2006 From: etaylor at realestate.com.au (Eric Taylor) Date: Tue, 24 Jan 2006 09:56:58 +1100 Subject: [Melbourne-pm] web application programmer position Message-ID: <6FA6124E9378214883CA93A54AED601D0255B57E@EMAIL.win.int.realestate.com.au> Realestate.com.au is looking for keen Perl programmers to work in the Richmond office. To apply, please email etaylor at realestate.com.au with your CV. Details of the position(s) available are as follows. duties: - development and documentation of web based Perl code - designing and documenting technical specifications - autonomous problem solving and time management skills: - Perl 5 experience - object oriented libraries - general Internet knowledge - effective communication - technical writing - SQL/HTML/XML experience - time estimation/management expectations: - participate in team environment - pro-actively suggest/implement improvements - follow existing standards and best practice -- Eric Taylor Developer Team Leader realestate.com.au the biggest address in property T: 03 9843 4208 Warning - This e-mail transmission may contain confidential information. If you have received this transmission in error, please notify us immediately on (61 3) 9897 1121 or by reply email to the sender. You must destroy the e-mail immediately and not use, copy, distribute or disclose the contents. Thank you. From mathew.robertson at netratings.com.au Mon Jan 23 17:09:19 2006 From: mathew.robertson at netratings.com.au (Mathew Robertson) Date: Tue, 24 Jan 2006 12:09:19 +1100 Subject: [Melbourne-pm] SOAP::Lite Serializing Complex objects. In-Reply-To: <43D45480.20403@unimelb.edu.au> References: <43D45480.20403@unimelb.edu.au> Message-ID: <43D57E3F.909@netratings.com.au> Hi there, I haven't used SOAP::Lite, but I had a quite look at the module code -> it appears that you shouldn't be using 'ns1' as a prefix. Not sure if this helps, Mathew Daniel Golembo wrote: >I have a problem serializing complex objects while using SOAP::Lite. >The issue is, if I have a complex object, that iteslf contains a complex >object, the second object does not get correctly encoded in the SOAP >envelope. > >For Example, I have an object that needs to be sent as an argument in a >SOAP call: > >my $expression = bless( { > 'descriptor' => bless( { > 'sum' => '1000', > 'description' => 'Adding 900 to 100' > }, 'sample_Descriptor' ), > 'a' => '900', > 'b' => '100', > 'dateAdded' => '2006-01-23T03:32:03.781Z' > }, 'sample_SumExpression' ); > >and I have SOAP serializer functions to serialize sample_SumExpression >and sample_Descriptor like: > >sub SOAP::Serializer::as_sample_SumExpression { > my $self = shift; > my($value, $name, $type, $attr) = @_; > > my $a=SOAP::Data->new(value=>$value->{a}, type=>'string'); > my $b=SOAP::Data->new(value=>$value->{b}, type=>'string'); > my $dateAdded=SOAP::Data->new(value=>$value->{dateAdded}, >type=>'dateTime'); > my $descriptor=SOAP::Data->new(value=>$value->{descriptor}, >type=>'ns1:sample_Descriptor'); > my $rval=[$name, {'xsi:type'=>'ns1:sample_SumExpression', %$attr}, >{a=>$a, b=>$b, dateAdded=>$dateAdded, descriptor=>$descriptor}]; > return $rval; >} > >and > >sub SOAP::Serializer::as_sample_Descriptor { > print "Firing SOAP::Serializer::as_unimelb_merlin_Descriptor\n"; > my $self = shift; > my($value, $name, $type, $attr) = @_; > > my $description=SOAP::Data->new(name=>'description', >value=>$value->{description}, type=>'string'); > my $sum=SOAP::Data->new(name=>'sum', value=>$value->{sum}, type=>'int'); > my $rval=[$name, {'xsi:type'=>'ns1:sample_Descriptor', %$attr}, >{description=>$description, sum=>$sum}]; > return $rval; >} > >When SOAP call happens, the envelope that is generated by SOAP::Lite is: > >xmlns:ns1="http://..." >soap:encodingStyle="http://schemas.xmlsoap.org/soap/encoding/" >xmlns:soap="http://schemas.xmlsoap.org/soap/envelope/" >xmlns:soapenc="http://schemas.xmlsoap.org/soap/encoding/" >xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" >xmlns:xsd="http://www.w3.org/2001/XMLSchema" >xmlns:tns="http://sampleWebServices.wsdl">xmlns="sampleServices">xsi:type="ns1:sample_SumExpression">xsi:type="ns1:sample_Descriptor">HASH(0x9c7d55c)xsi:type="xsd:string">900100xsi:type="xsd:dateTime">2006-01-23T03:32:03.781Z > >So, the Top level object, sample_Expression, gets encoded correctly, but >the lower level object, sample_Descriptor, is not. >I have also tried using encode_object function of SOAP::Serializer, but >this resulted in sample_Descriptor being encoded as an array. > >I wonder what I am doing wrong or what SOAP::Lite is doing wrong. >Please help. >_______________________________________________ >Melbourne-pm mailing list >Melbourne-pm at pm.org >http://mail.pm.org/mailman/listinfo/melbourne-pm > > From ts at meme.com.au Fri Jan 27 18:22:12 2006 From: ts at meme.com.au (Tony Smith) Date: Sat, 28 Jan 2006 13:22:12 +1100 Subject: [Melbourne-pm] Opening hours representation Message-ID: A client who has what I am sure must be a common need to represent the hours that a business, venue, et al is open each week in a manageable format. At first glance, neither search.cpan nor Google came up with anything which separates the idea of regular weekly opening/operating hours from more general calendaring and that misses the point. Basically what we want is to be able to query whether and maybe what for a business is open at a particular time, e.g. if a restaurant closes at 10:00 pm it would probably be wrong to suggest it is open for dinner at 9:30 pm. I don't really want to reinvent the wheel, particularly when it comes to defining a data representation format, and if it is going to take a new module it would be good to make it as widely useful as the problem would suggest it should be. Any thoughts? Tony Smith 0405 499 718 TransForum Developer http://www.transforum.net/ From raphael at mitija.com Sat Jan 28 20:50:53 2006 From: raphael at mitija.com (Raphael Alla) Date: Sun, 29 Jan 2006 15:50:53 +1100 Subject: [Melbourne-pm] security hole Message-ID: *Let's condider the following perl cgi script. I cannot find a practical way to use it as a security hole. Any suggestion? *#!/usr/bin/perl my $sub = $ENV{QUERY_STRING}; &{$sub}; -- Raphael Alla Mitija Australia +61 4 15 678 576 Premium open source accounting for Australia http://www.thetravelingaccountant.com -------------- next part -------------- An HTML attachment was scrubbed... URL: http://mail.pm.org/pipermail/melbourne-pm/attachments/20060129/de7c5337/attachment.html From raphael.alla at gmail.com Sat Jan 28 20:51:48 2006 From: raphael.alla at gmail.com (Raphael Alla) Date: Sun, 29 Jan 2006 15:51:48 +1100 Subject: [Melbourne-pm] Security hole?? Message-ID: *Let's condider the following perl cgi script. I cannot find a practical way to use it as a security hole. Any suggestion? *#!/usr/bin/perl my $sub = $ENV{QUERY_STRING}; &{$sub}; -- Raphael Alla Mitija Australia +61 4 15 678 576 Premium open source accounting for Australia http://www.thetravelingaccountant.com -------------- next part -------------- An HTML attachment was scrubbed... URL: http://mail.pm.org/pipermail/melbourne-pm/attachments/20060129/79f69094/attachment.html From david_dick at iprimus.com.au Sun Jan 29 01:27:32 2006 From: david_dick at iprimus.com.au (David Dick) Date: Sun, 29 Jan 2006 20:27:32 +1100 Subject: [Melbourne-pm] Security hole?? In-Reply-To: References: Message-ID: <43DC8A84.5040206@iprimus.com.au> Raphael Alla wrote: > *Let's condider the following perl cgi script. I cannot find a practical > way to use it as a security hole. Any suggestion? > > *#!/usr/bin/perl > my $sub = $ENV{QUERY_STRING}; > &{$sub}; maybe i'm just a little tired after the weekend, but this script would appear to not do anything apart from cause an error. What are you trying to achieve? From raphael.alla at gmail.com Sun Jan 29 02:33:01 2006 From: raphael.alla at gmail.com (Raphael Alla) Date: Sun, 29 Jan 2006 21:33:01 +1100 Subject: [Melbourne-pm] Security hole?? In-Reply-To: <43DC8A84.5040206@iprimus.com.au> References: <43DC8A84.5040206@iprimus.com.au> Message-ID: If this script is called security.cgi, and you invoke it like this: hostname/security.cgi?sub_name, then it will call the sub sub_name (it is a call of sub by reference). Potentially a user can call any sub from any module which it has available to it. It is seen as a bad habit to give the right to anyone on the internet to execute any sub on your system, yet I fail to find a practical threat coming from this specific construct. R. On 1/29/06, David Dick wrote: > > > > Raphael Alla wrote: > > *Let's condider the following perl cgi script. I cannot find a practical > > way to use it as a security hole. Any suggestion? > > > > *#!/usr/bin/perl > > my $sub = $ENV{QUERY_STRING}; > > &{$sub}; > > maybe i'm just a little tired after the weekend, but this script would > appear to not do anything apart from cause an error. What are you trying > to achieve? > > -- Raphael Alla Mitija Australia +61 4 15 678 576 Premium open source accounting for Australia http://www.thetravelingaccountant.com -------------- next part -------------- An HTML attachment was scrubbed... URL: http://mail.pm.org/pipermail/melbourne-pm/attachments/20060129/9de06c57/attachment.html From alfiejohn at acm.org Sun Jan 29 02:54:45 2006 From: alfiejohn at acm.org (Alfie John) Date: Sun, 29 Jan 2006 21:54:45 +1100 Subject: [Melbourne-pm] Security hole?? In-Reply-To: References: <43DC8A84.5040206@iprimus.com.au> Message-ID: <6826B54C-538C-4FDC-B435-245C0E7833A2@acm.org> Hi, On 29/01/2006, at 9:33 PM, Raphael Alla wrote: > If this script is called security.cgi, and you invoke it like this: > hostname/security.cgi?sub_name, then it will call the sub sub_name > (it is a call of sub by reference). > > Potentially a user can call any sub from any module which it has > available to it. It is seen as a bad habit to give the right to > anyone on the internet to execute any sub on your system, yet I > fail to find a practical threat coming from this specific construct. > Pointing to 'http://localhost/test.cgi?what_is_my_password' would do the trick. --- 8< --- #!/usr/bin/perl print "Content-type: text/html\n\n"; my $sub = $ENV{'QUERY_STRING'}; &{$sub}; sub what_is_my_password { print q{your password is '0wn3d'}; } --- >8 --- However, is this a practical threat? Not really since you are doing sanity checking. Aren't you ;) Alfie > R. > > On 1/29/06, David Dick wrote: > > > Raphael Alla wrote: > > *Let's condider the following perl cgi script. I cannot find a > practical > > way to use it as a security hole. Any suggestion? > > > > *#!/usr/bin/perl > > my $sub = $ENV{QUERY_STRING}; > > &{$sub}; > > maybe i'm just a little tired after the weekend, but this script would > appear to not do anything apart from cause an error. What are you > trying > to achieve? > > > > > -- > Raphael Alla > Mitija Australia > +61 4 15 678 576 > > Premium open source accounting for Australia > http://www.thetravelingaccountant.com > _______________________________________________ > Melbourne-pm mailing list > Melbourne-pm at pm.org > http://mail.pm.org/mailman/listinfo/melbourne-pm -------------- next part -------------- An HTML attachment was scrubbed... URL: http://mail.pm.org/pipermail/melbourne-pm/attachments/20060129/a12ea977/attachment.html From jarich at perltraining.com.au Sun Jan 29 03:00:48 2006 From: jarich at perltraining.com.au (Jacinta Richardson) Date: Sun, 29 Jan 2006 22:00:48 +1100 Subject: [Melbourne-pm] Security hole?? In-Reply-To: References: <43DC8A84.5040206@iprimus.com.au> Message-ID: <43DCA060.6070000@perltraining.com.au> Raphael Alla wrote: > If this script is called security.cgi, and you invoke it like this: > hostname/security.cgi?sub_name, then it will call the sub sub_name (it > is a call of sub by reference). > > Potentially a user can call any sub from any module which it has > available to it. It is seen as a bad habit to give the right to anyone > on the internet to execute any sub on your system, yet I fail to find a > practical threat coming from this specific construct. > > *#!/usr/bin/perl > > my $sub = $ENV{QUERY_STRING}; > > &{$sub}; This code in and of itself won't work. It doesn't matter what $ENV{QUERY_STRING} is, there aren't any subroutines to be invoked. Assuming you have subroutines that you are willing to invoke, there's a better solution (better for all sorts of reasons such as obviousness, maintainability, preventing accidents...) using a dispatch table. #!/usr/bin/perl -w use strict; my %dispatch = ( sub1 => \&sub1, sub2 => \&sub2, # ... ); my $sub = $ENV{QUERY_STRING}; if(exists $dispatch{$sub}) { $dispatch{$sub}->(); } else { die "unknown sub $sub"; } or something like that. I haven't tested the above, nor slept for 2 nights. Remove the if/else to get something more like your code. Sure the code takes up more space, but it is strict compliant, limits the subroutines that can be called and generally makes people who have to look at your code happier. As for specific security threats, it depends on what your subroutines do. If your code can be tricked into doing something that you didn't intend it to do, there's a bug. If you have *private* subroutines, which generate email, or print reports, or display sensitive information, then this is a security hole. Relying on your attacker not to guess the names of those subroutines, is security through obscurity, and consequently a bad idea. Ideally you should code as if your attacker has access to read your source code. They might. If you don't have private subroutines, and you really are willing for *all* subroutines that this file has access to, to be called; then the threat is likely minimal. Of course, someone will later go and add a private subroutine to your code, and get confused when it gets called in strange circumstances... Some security focus must be spent on future proofing your code, from a maintainer who is focussing on something other than how clever you have been. All the best, Jacinta -- ("`-''-/").___..--''"`-._ | Jacinta Richardson | `6_ 6 ) `-. ( ).`-.__.`) | Perl Training Australia | (_Y_.)' ._ ) `._ `. ``-..-' | +61 3 9354 6001 | _..`--'_..-_/ /--'_.' ,' | contact at perltraining.com.au | (il),-'' (li),' ((!.-' | www.perltraining.com.au | From scottp at dd.com.au Mon Jan 30 16:01:33 2006 From: scottp at dd.com.au (Scott Penrose) Date: Tue, 31 Jan 2006 11:01:33 +1100 Subject: [Melbourne-pm] Document Management System Message-ID: <261620F4-DF4D-4A51-889F-BA392CD5AED3@dd.com.au> Hey Guys I have been asked to put together a simple document management system (not a CMS !) on a web site. Simple stuff really. Does anyone know of any open source, preferably perl, document management systems? Scott -- * - * http://www.osdc.com.au - Open Source Developers Conference * - * Scott Penrose Welcome to the Digital Dimension http://www.dd.com.au/ scottp at dd.com.au Dismaimer: Contents of this mail and signature are bound to change randomly. Whilst every attempt has been made to control said randomness, the author wishes to remain blameless for the number of eggs that damn chicken laid. Oh and I don't want to hear about butterflies either. Please do not send me Word or PowerPoint attachments. See http://www.gnu.org/philosophy/no-word-attachments.html Microsoft is not the answer. It's the question. And the answer is no. -------------- next part -------------- A non-text attachment was scrubbed... Name: PGP.sig Type: application/pgp-signature Size: 186 bytes Desc: This is a digitally signed message part Url : http://mail.pm.org/pipermail/melbourne-pm/attachments/20060131/92720320/PGP.bin From pjf at perltraining.com.au Tue Jan 31 04:54:23 2006 From: pjf at perltraining.com.au (Paul Fenwick) Date: Wed, 01 Feb 2006 01:54:23 +1300 Subject: [Melbourne-pm] security hole In-Reply-To: References: Message-ID: <43DF5DFF.2040804@perltraining.com.au> G'day Raphael, Raphael Alla wrote: > *#!/usr/bin/perl > my $sub = $ENV{QUERY_STRING}; > &{$sub}; This code warrants that every subroutine from every module and library you have loaded is perfectly safe to be called without arguments by a hostile attacker. That's a very big warrant. &{$sub} does not in any way restrict you to your own package. If your subroutine specifies a subroutine in another package (eg: 'Dangerous::Package::Kaboom') then that *will* be called. To make matters worse, the use of &{...} syntax results in the contents of @_ being passed implicitly to the subroutine, something which not many people expect. This code has two fundamental problems, even if there are some circumstances where you may not be able to exploit them: * It results in action from a distance. Any subroutine from any module could be called, making it *very* hard to determine all possible execution paths. This is not only very bad for security, it also makes debugging and maintenance difficult. This is reason enough to never ever use symbolic references. * It is the antithesis to 'deny by default'. Any potential hole elsewhere in the program is magnified greatly by the code above. I personally would never allow such code past review, let alone run in a security sensitive context. All the best, Paul -- Paul Fenwick | http://perltraining.com.au/ Director of Training | Ph: +61 3 9354 6001 Perl Training Australia | Fax: +61 3 9354 2681 From raphael.alla at gmail.com Mon Jan 30 17:40:34 2006 From: raphael.alla at gmail.com (Raphael Alla) Date: Tue, 31 Jan 2006 12:40:34 +1100 Subject: [Melbourne-pm] security hole In-Reply-To: <43DF5DFF.2040804@perltraining.com.au> References: <43DF5DFF.2040804@perltraining.com.au> Message-ID: Hi Paul, my query was actually generated from the material coming from your own web site! I think I understand better the security issues now. I have thought of some other solutions than using a hash table. Having to maintain the hash table can be cumbersome to maintain or not applicable to all situations. Solution 1: use a namespace convention such as &("namespace_name_" . $sub}(); But this leaves the call by reference. Solution 2: Use tags to flag which one are the subs which can be called from the outside and build the hashtable "on the fly" at execution time. This seems to be similar to what Catalyst does. for instance: #!/usr/bin/perl use safe_call; sub can_be_called: Callable { ...... } my $sub = $ENV{QUERY_STRING}; safe_call($sub); and in safe_call.pm my %symcache; my @declarations; my $identified = 0; my %callable; #from Attribute::Handlers sub findsym { my ($pkg, $ref) = @_; return $symcache{$pkg,$ref} if $symcache{$pkg,$ref}; $type ||= ref($ref); foreach my $sym ( values %{$pkg."::"} ) { if (*{$sym}{$type} && *{$sym}{$type} == $ref) { $callable{ *{$sym}{NAME} } = 1; #print "added ", *{$sym}{NAME}, "\n"; $symcache{$pkg,$ref} = \$sym } } } #just stores the declarations for later identification sub MODIFY_CODE_ATTRIBUTES { my ($module, $ref, @attributes) = @_; my @other_attr; foreach (@attributes) { if ($_ eq "Callable") { push @declarations, [ $module, $ref ]; } else { push @other_attr, $_;}; } return @other_attr; } sub identify { foreach (@declarations) { findsym @$_[0,1]; } $identified = 1; } sub safe_call { identify() unless $identified; $sub_name = shift; if ($callable{$sub_name}) { &{$sub_name}; } else { die "Trying to call unallowed sub $sub_name \n"; } } 1; I am sure that the above can be improved. R. On 1/31/06, Paul Fenwick wrote: > > G'day Raphael, > > Raphael Alla wrote: > > > *#!/usr/bin/perl > > my $sub = $ENV{QUERY_STRING}; > > &{$sub}; > > This code warrants that every subroutine from every module and library you > have > loaded is perfectly safe to be called without arguments by a hostile > attacker. > That's a very big warrant. > > &{$sub} does not in any way restrict you to your own package. If your > subroutine specifies a subroutine in another package (eg: > 'Dangerous::Package::Kaboom') then that *will* be called. > > To make matters worse, the use of &{...} syntax results in the contents of > @_ > being passed implicitly to the subroutine, something which not many people > expect. > > This code has two fundamental problems, even if there are some > circumstances > where you may not be able to exploit them: > > * It results in action from a distance. Any subroutine from any > module could be called, making it *very* hard to determine all > possible execution paths. This is not only very bad for > security, > it also makes debugging and maintenance difficult. This is > reason > enough to never ever use symbolic references. > > * It is the antithesis to 'deny by default'. Any potential hole > elsewhere in the program is magnified greatly by the code above. > > I personally would never allow such code past review, let alone run in a > security sensitive context. > > All the best, > > Paul > > -- > Paul Fenwick | http://perltraining.com.au/ > Director of Training | Ph: +61 3 9354 6001 > Perl Training Australia | Fax: +61 3 9354 2681 > _______________________________________________ > Melbourne-pm mailing list > Melbourne-pm at pm.org > http://mail.pm.org/mailman/listinfo/melbourne-pm > -- Raphael Alla Mitija Australia +61 4 15 678 576 Premium open source accounting for Australia http://www.thetravelingaccountant.com -------------- next part -------------- An HTML attachment was scrubbed... URL: http://mail.pm.org/pipermail/melbourne-pm/attachments/20060131/add5692d/attachment.html From jens at porup.com Mon Jan 30 17:42:58 2006 From: jens at porup.com (Jens Porup) Date: Tue, 31 Jan 2006 12:42:58 +1100 Subject: [Melbourne-pm] Document Management System In-Reply-To: <261620F4-DF4D-4A51-889F-BA392CD5AED3@dd.com.au> References: <261620F4-DF4D-4A51-889F-BA392CD5AED3@dd.com.au> Message-ID: <20060131014248.GB4528@localhost.localdomain> On Tue, Jan 31, 2006 at 11:01:33AM +1100, Scott Penrose wrote: > Hey Guys > > I have been asked to put together a simple document management system > (not a CMS !) on a web site. Simple stuff really. > > Does anyone know of any open source, preferably perl, document > management systems? I know of one, mostly cuz it's the one I wrote. ;) It was written for a very specific client's needs, however, so if you give me a better idea what you need, I can tell you if it will help. I suppose it also depends on what you mean by a document management system. I mean, a filesystem is a pretty good dms, some sort of DAV scenario might also solve your problem. The problem that led to my little project was the obvious limits of email for transferring large files to and from a particular client, and the $BigCorp firewall issues with using ftp. What my code does is let you communicate large files to and from a client over HTTP. It uses IP-based auth, that is, if you're inside our local network you have all permissions, if you're outside, you have none. You can set an expiry on the file, cancel the file, blah blah, and a link is then generated for the client to use to download the file you're sending them. It's a smallish app, written in perl. We haven't released it, but I understand we were thinking of open sourcing it anyway. Let me know if this sounds like something you can use. Cheers, Jens From scottp at dd.com.au Mon Jan 30 18:16:47 2006 From: scottp at dd.com.au (Scott Penrose) Date: Tue, 31 Jan 2006 13:16:47 +1100 Subject: [Melbourne-pm] Document Management System In-Reply-To: <20060131014248.GB4528@localhost.localdomain> References: <261620F4-DF4D-4A51-889F-BA392CD5AED3@dd.com.au> <20060131014248.GB4528@localhost.localdomain> Message-ID: Sounds very close yes. Basic requirements of the system are: * File based restrictions - Who can view - Who can edit (upload / replace) * Simple version control - eg: Keep old versions - Mark current version (so you can be working on a draft) * Simple MetaData - Title, Description - Last modified and who * Simple organisation - eg: Directories * Checkout - Not essential, but nice to know someone is working on it This is for a group of about 25+ people, geographically diverse (all around Australia) and all based on Web and simple system. I will avoid DAV at this stage and stick to HTTP Post. Scott On 31/01/2006, at 12:42, Jens Porup wrote: > On Tue, Jan 31, 2006 at 11:01:33AM +1100, Scott Penrose wrote: >> Hey Guys >> >> I have been asked to put together a simple document management system >> (not a CMS !) on a web site. Simple stuff really. >> >> Does anyone know of any open source, preferably perl, document >> management systems? > > I know of one, mostly cuz it's the one I wrote. ;) > > It was written for a very specific client's needs, however, so if > you give me a better idea what you need, I can tell you if it will > help. > > I suppose it also depends on what you mean by a document management > system. I mean, a filesystem is a pretty good dms, some sort of DAV > scenario might also solve your problem. > > The problem that led to my little project was the obvious limits of > email for transferring large files to and from a particular client, > and > the $BigCorp firewall issues with using ftp. > > What my code does is let you communicate large files to and from a > client over HTTP. It uses IP-based auth, that is, if you're inside > our local network you have all permissions, if you're outside, you > have > none. > > You can set an expiry on the file, cancel the file, blah blah, > and a link is then generated for the client to use to download the > file you're sending them. > > It's a smallish app, written in perl. We haven't released it, > but I understand we were thinking of open sourcing it anyway. > > Let me know if this sounds like something you can use. > > Cheers, > > Jens > -- * - * http://www.osdc.com.au - Open Source Developers Conference * - * Scott Penrose VP in charge of Pancakes http://linux.dd.com.au/ scottp at dd.com.au Dismaimer: If you receive this email in error - please eat it immediately to prevent it from falling into the wrong hands. Please do not send me Word or PowerPoint attachments. See http://www.gnu.org/philosophy/no-word-attachments.html Microsoft is not the answer. It's the question. And the answer is no. -------------- next part -------------- A non-text attachment was scrubbed... Name: PGP.sig Type: application/pgp-signature Size: 186 bytes Desc: This is a digitally signed message part Url : http://mail.pm.org/pipermail/melbourne-pm/attachments/20060131/0b3c0bb6/PGP-0001.bin From pjf at perltraining.com.au Mon Jan 30 18:38:21 2006 From: pjf at perltraining.com.au (Paul Fenwick) Date: Tue, 31 Jan 2006 13:38:21 +1100 Subject: [Melbourne-pm] security hole In-Reply-To: References: <43DF5DFF.2040804@perltraining.com.au> Message-ID: <43DECD9D.10602@perltraining.com.au> G'day Raphael, Raphael Alla wrote: > my query was actually generated from the material coming from your own web > site! I think I understand better the security issues now. Good... I hope... ;) Would this be from our sample chapter from the Perl Security course? The one that discusses that if you turn off strict then taint mode won't save you from calling subroutines by symbolic reference? > Solution 1: use a namespace convention such as > > &("namespace_name_" . $sub}(); > > But this leaves the call by reference. Indeed, and if 'namespace_name' is a package (the most obvious choice) there's still a risk that a module that has been use'd has exported subroutines into that package. > Solution 2: Use tags to flag which one are the subs which can be called from > the outside and build the hashtable "on the fly" at execution time. This > seems to be similar to what Catalyst does. [snip] The trouble here is making it 'obviously correct'. Luckily, using Attribute::Handlers is a good way of going about this. Attribute::Handlers allows you to write subroutines which will then act when something with that attribute is declared. The following *untested* code, with no warranty, provides an example of how we can catch all subroutines declared with the 'Callable' attribute, and add them to a hash of (name => coderef) pairs. #!/usr/bin/perl -w use strict; use Attribute::Handlers; my %callable; # Remember subroutines marked as 'Callable' and place them into the # %callable hash declared above. This code does *not* examine the package # into which such subroutines are declared. $_[0] contains the package # name if this is required. sub Callable :ATTR { my (undef, $glob, $sub_ref, undef, undef, undef) = @_; if (ref($sub_ref) ne 'CODE') { croak q{'Callable' attribute set on non-subroutine}; } if ($glob eq 'ANON') { croak q{'Callable' attribute set on anonymous subroutine}; } # Find the name of the subroutine (technically the typeglob # it was entered into). Note this does not provide the package # name. my $name = *{$glob}{NAME} or die "Internal error: subroutine with no name"; # Now plug our name -> coderef into our hash. $callable{$name} = $sub_ref; return; } __END__ Cheerio, Paul -- Paul Fenwick | http://perltraining.com.au/ Director of Training | Ph: +61 3 9354 6001 Perl Training Australia | Fax: +61 3 9354 2681 From raphael at mitija.com Mon Jan 30 18:57:42 2006 From: raphael at mitija.com (Raphael Alla) Date: Tue, 31 Jan 2006 13:57:42 +1100 Subject: [Melbourne-pm] security hole In-Reply-To: <43DECD9D.10602@perltraining.com.au> References: <43DF5DFF.2040804@perltraining.com.au> <43DECD9D.10602@perltraining.com.au> Message-ID: Hi, your code will not work in all circumstances. In particular, the way attribute work is that the MODIFY_*_ATTRIBUTE function is called during the BEGIN stage of the compilation process at a stage in which subroutines have not yet been given a name. The Attributes::Handler module works around that by calling the "find_sym" function during the CHECK compilation phase to identify what are the name of the syms. The function they use is the same one I have reused. Now, if you put your code in a module, let's say safe_call.pm, then the identification of module will take place during the compilation of safe_call.pm, and not during the compilation of t.pl. All subs will be identified as ANON. file t.pl: #!/usr/bin/perl use safe_call; sub this_is_ok: Callable {...} __END__ In the code which I propose, the find_sym sub is called at the first access to the safe_call() sub, which is after the compilation time. Your code would work if the sub this_is_ok is part of the file for which the function sub Callable: ATTR is defined. Part of the difficulty stems from the fact that the MODIFY_*_ATTRIBUTE function is not called with a glob as a parameter but with a reference to the code instead. The glob must be found by doing a search through the symbol table, which is cumbersome. This could be the subject of an interesting, yet technical, lightning talk! Cheers R. On 1/31/06, Paul Fenwick wrote: > > G'day Raphael, > > Raphael Alla wrote: > > > my query was actually generated from the material coming from your own > web > > site! I think I understand better the security issues now. > > Good... I hope... ;) Would this be from our sample chapter from the Perl > Security course? The one that discusses that if you turn off strict then > taint > mode won't save you from calling subroutines by symbolic reference? > > > Solution 1: use a namespace convention such as > > > > &("namespace_name_" . $sub}(); > > > > But this leaves the call by reference. > > Indeed, and if 'namespace_name' is a package (the most obvious choice) > there's > still a risk that a module that has been use'd has exported subroutines > into > that package. > > > Solution 2: Use tags to flag which one are the subs which can be called > from > > the outside and build the hashtable "on the fly" at execution time. This > > seems to be similar to what Catalyst does. > > [snip] > > The trouble here is making it 'obviously correct'. Luckily, using > Attribute::Handlers is a good way of going about > this. Attribute::Handlers > allows you to write subroutines which will then act when something with > that > attribute is declared. > > The following *untested* code, with no warranty, provides an example of > how we > can catch all subroutines declared with the 'Callable' attribute, and add > them > to a hash of (name => coderef) pairs. > > #!/usr/bin/perl -w > use strict; > > use Attribute::Handlers; > > my %callable; > > # Remember subroutines marked as 'Callable' and place them into the > # %callable hash declared above. This code does *not* examine the package > # into which such subroutines are declared. $_[0] contains the package > # name if this is required. > > sub Callable :ATTR { > my (undef, $glob, $sub_ref, undef, undef, undef) = @_; > > if (ref($sub_ref) ne 'CODE') { > croak q{'Callable' attribute set on non-subroutine}; > } > > if ($glob eq 'ANON') { > croak q{'Callable' attribute set on anonymous subroutine}; > } > > # Find the name of the subroutine (technically the typeglob > # it was entered into). Note this does not provide the package > # name. > > my $name = *{$glob}{NAME} > or die "Internal error: subroutine with no name"; > > # Now plug our name -> coderef into our hash. > > $callable{$name} = $sub_ref; > > return; > > } > > __END__ > > Cheerio, > > Paul > > -- > Paul Fenwick | http://perltraining.com.au/ > Director of Training | Ph: +61 3 9354 6001 > Perl Training Australia | Fax: +61 3 9354 2681 > -- Raphael Alla Mitija Australia +61 4 15 678 576 Premium open source accounting for Australia http://www.thetravelingaccountant.com -------------- next part -------------- An HTML attachment was scrubbed... URL: http://mail.pm.org/pipermail/melbourne-pm/attachments/20060131/1532be61/attachment.html From raphael.alla at gmail.com Mon Jan 30 18:58:41 2006 From: raphael.alla at gmail.com (Raphael Alla) Date: Tue, 31 Jan 2006 13:58:41 +1100 Subject: [Melbourne-pm] security hole In-Reply-To: <43DECD9D.10602@perltraining.com.au> References: <43DF5DFF.2040804@perltraining.com.au> <43DECD9D.10602@perltraining.com.au> Message-ID: Hi, your code will not work in all circumstances. In particular, the way attribute work is that the MODIFY_*_ATTRIBUTE function is called during the BEGIN stage of the compilation process at a stage in which subroutines have not yet been given a name. The Attributes::Handler module works around that by calling the "find_sym" function during the CHECK compilation phase to identify what are the name of the syms. The function they use is the same one I have reused. Now, if you put your code in a module, let's say safe_call.pm, then the identification of module will take place during the compilation of safe_call.pm, and not during the compilation of t.pl. All subs will be identified as ANON. file t.pl: #!/usr/bin/perl use safe_call; sub this_is_ok: Callable {...} __END__ In the code which I propose, the find_sym sub is called at the first access to the safe_call() sub, which is after the compilation time. Your code would work if the sub this_is_ok is part of the file for which the function sub Callable: ATTR is defined. Part of the difficulty stems from the fact that the MODIFY_*_ATTRIBUTE function is not called with a glob as a parameter but with a reference to the code instead. The glob must be found by doing a search through the symbol table, which is cumbersome. This could be the subject of an interesting, yet technical, lightning talk! Cheers R. On 1/31/06, Paul Fenwick wrote: > > G'day Raphael, > > Raphael Alla wrote: > > > my query was actually generated from the material coming from your own > web > > site! I think I understand better the security issues now. > > Good... I hope... ;) Would this be from our sample chapter from the Perl > Security course? The one that discusses that if you turn off strict then > taint > mode won't save you from calling subroutines by symbolic reference? > > > Solution 1: use a namespace convention such as > > > > &("namespace_name_" . $sub}(); > > > > But this leaves the call by reference. > > Indeed, and if 'namespace_name' is a package (the most obvious choice) > there's > still a risk that a module that has been use'd has exported subroutines > into > that package. > > > Solution 2: Use tags to flag which one are the subs which can be called > from > > the outside and build the hashtable "on the fly" at execution time. This > > seems to be similar to what Catalyst does. > > [snip] > > The trouble here is making it 'obviously correct'. Luckily, using > Attribute::Handlers is a good way of going about > this. Attribute::Handlers > allows you to write subroutines which will then act when something with > that > attribute is declared. > > The following *untested* code, with no warranty, provides an example of > how we > can catch all subroutines declared with the 'Callable' attribute, and add > them > to a hash of (name => coderef) pairs. > > #!/usr/bin/perl -w > use strict; > > use Attribute::Handlers; > > my %callable; > > # Remember subroutines marked as 'Callable' and place them into the > # %callable hash declared above. This code does *not* examine the package > # into which such subroutines are declared. $_[0] contains the package > # name if this is required. > > sub Callable :ATTR { > my (undef, $glob, $sub_ref, undef, undef, undef) = @_; > > if (ref($sub_ref) ne 'CODE') { > croak q{'Callable' attribute set on non-subroutine}; > } > > if ($glob eq 'ANON') { > croak q{'Callable' attribute set on anonymous subroutine}; > } > > # Find the name of the subroutine (technically the typeglob > # it was entered into). Note this does not provide the package > # name. > > my $name = *{$glob}{NAME} > or die "Internal error: subroutine with no name"; > > # Now plug our name -> coderef into our hash. > > $callable{$name} = $sub_ref; > > return; > > } > > __END__ > > Cheerio, > > Paul > > -- > Paul Fenwick | http://perltraining.com.au/ > Director of Training | Ph: +61 3 9354 6001 > Perl Training Australia | Fax: +61 3 9354 2681 > -- Raphael Alla Mitija Australia +61 4 15 678 576 Premium open source accounting for Australia http://www.thetravelingaccountant.com -------------- next part -------------- An HTML attachment was scrubbed... URL: http://mail.pm.org/pipermail/melbourne-pm/attachments/20060131/eec4427f/attachment-0001.html From jarich at perltraining.com.au Tue Jan 31 15:43:36 2006 From: jarich at perltraining.com.au (Jacinta Richardson) Date: Wed, 01 Feb 2006 10:43:36 +1100 Subject: [Melbourne-pm] Discount on upcoming Perl Training Australia courses Message-ID: <43DFF628.3040402@perltraining.com.au> G'day everyone, Get up to two days of free training with our latest special. Perl Training Australia would like to invite you and your colleagues to join us on our next set of courses in Melbourne. We're running the following courses: Course Date -------------------------------------------------------- Programming Perl 21st - 22nd March 2006 Object Oriented Perl 28th - 29th March 2006 Databases and Perl 30th - 31st March 2006 Web Development with Perl 5th - 6th April 2006 Perl Security 7th April 2006 Programming Perl is the amalgamation of our two most popular courses, Introduction to Perl and Intermediate Perl. Mention Melbourne Perl Mongers to get a 5% discount off any full price course. Book on multiple courses to get a 25% off each subsequent course, with our "Bundle and Save" special. For example, book on the full set (Programming Perl, Object Oriented Perl, Database Programming with Perl, Web Development with Perl and Perl Security) and you can *save $1100* per person! That's a two day course for free! Even better, if you book 3 or more attendees on each of these courses, you'll also be eligible for the "Group discount" of an additional 5% off all courses. A potential saving of *$1408* per person! To book on these courses visit http://perltraining.com.au/bookings/Melbourne.html Please don't hesitate to contact me for further information. All the best, Jacinta -- ("`-''-/").___..--''"`-._ | Jacinta Richardson | `6_ 6 ) `-. ( ).`-.__.`) | Perl Training Australia | (_Y_.)' ._ ) `._ `. ``-..-' | +61 3 9354 6001 | _..`--'_..-_/ /--'_.' ,' | contact at perltraining.com.au | (il),-'' (li),' ((!.-' | www.perltraining.com.au | From jarich at perltraining.com.au Tue Jan 31 21:36:20 2006 From: jarich at perltraining.com.au (Jacinta Richardson) Date: Wed, 01 Feb 2006 16:36:20 +1100 Subject: [Melbourne-pm] Mugs for sale: $6 Message-ID: <43E048D4.3040902@perltraining.com.au> G'day everyone, At the last Perl Mongers meeting we were approached with the request to buy a few of the Perl Training Australia regular expression mugs. These were provided to OSDC 2005 conference attendees. The ones currently at myinternet are *not* for sale as they have been set aside to go to the OSDC 2005 sponsors. If you missed out on attending the conference, or broke yours or who want (another) one for some reason (good gifts for colleagues) let us know. We're very happy to provide these to Melbourne Perl Mongers at $6 each. To see the image on the mugs please visit: http://perltraining.com.au/~jarich/mug.jpg To see an image of the finished mug please visit: http://perltraining.com.au/~jarich/mugphoto-small.jpg Please let us know what quantities you'd like and whether you require a tax invoice. Orders before the 8th of February (next Wendesday) will be brought along to the OSDClub meeting (unless you request otherwise). All the very best, Jacinta -- ("`-''-/").___..--''"`-._ | Jacinta Richardson | `6_ 6 ) `-. ( ).`-.__.`) | Perl Training Australia | (_Y_.)' ._ ) `._ `. ``-..-' | +61 3 9354 6001 | _..`--'_..-_/ /--'_.' ,' | contact at perltraining.com.au | (il),-'' (li),' ((!.-' | www.perltraining.com.au |