Mail list predators

Tkil tkil at
Mon Jul 22 20:04:41 CDT 2002

>>>>> "Jason" == Jason Annin-White <jasona at> writes:

[out of order, but this needed to be answered first.]

Jason> I monitor this list as well, if you think you're burnt out by
Jason> newbie requests on this list, you should try joining some of
Jason> the lists I contribute to.

*shrug* I've put in my time on usenet and irc.  While I don't do those
anymore, I do help out on various Perl and Linux lists.  I wouldn't
refer to myself as a wizard, but I do have relatively deep knowledge
of a few fields.  When a hard question in one of those fields comes
up, I try to contribute usefully.

Jason> You can't be too burnt out if you have the time and energy to
Jason> line-by-line comment through my email.

Believe it or not, I was trying to be courteous.  I happen to find the
point-by-point style much more useful than having the original and
response in two different chunks; the interleaved style used to be the
majority, until various MUAs and newsreaders started changing their
default behavior.

Jason> You don't teach how to fish Tkil, you dangle the fish and say
Jason> you might teach somebody how after you've had sufficient
Jason> entertainment watching them thrash around knee-deep in water.
Jason> Get off your high horse and do something useful.

Hm.  I disagree with this assessment, and I suspect that some of the
people who knew me back when I had more energy would also disagree.

[back to the original order]

Jason> I was not implicating either Randal or Tom in my unpleasant
Jason> experiences.  I hope it didn't sound that way.

I apologize for coming to this erroneous conclusion; Tom's response
was the only one I saw publicly, and it did sound a bit sharp to me
(particularly given that Tom is, in my experience, very patient and
mellow with people).  Sorry about that.

Jason> None of them are doing what I am trying to do, which is
Jason> straitforward in most other languages.  

I think that this touches very close to what set me off with some of
your questions:

 - They don't do *exactly* what you're doing.  Well, what can you
   generalize from them?  What small-scale experiments did you run, to
   try to determine the behavior of perl, to verify your assumptions,
   try different solutions to smaller and more focused problems?

   The sample program I included in a different message is the sort of
   small-scale experiment that I'm referring to.

 - "straightforward in most other languages".  That's a loaded phrase;
   which other languages are you referring to?  Object serialization
   is hardly trivial in traditional 3rd-gen languages (C, C++); Java
   has it, and it is indeed straightforward there.  Visual Basic?
   Lisp?  Prolog?  Python (which does have a very nice freeze/thaw
   system, IIRC).  Assembly?  Ha.  :)

   And when you say straightforward, does that take into account
   portability between architectures, little-/big-endian, network byte
   order, end-of-stream signaling, database persistence...?

In this instance, it seems that the fundamental issue (after Randal
steered you towards Storable as being superior to Data::Dumper) was an
unclear understanding of references and objects in perl.  That's
nothing to be ashamed about -- but without us asking some questions
that can sound hard at times, we never would have found out where your
code was really having the problem.

Jason> I can't `perldoc typedef` or `perldoc reference` so I dropped
Jason> in here to ask for help in accessing my restored objects.

perlref was probably the one you were looking for (and also perlobj;
there are corresponding tutorials, "perlreftut" and "perltoot").
These are listed if you do "perldoc perl".

Jason> I didn't come here begging for fish, I came here asking to
Jason> watch a pro fillet one, which is what the sign over the door
Jason> offers.

To continue this analogy, I saw you asking "how do i filet this nice
bit of trout here?"  Randal suggested "lengthwise"; due to a mis-
communication, you went ahead cutting it straight across, then asked
us how to remove the bones.

Jason> I get tired of reading through what somebody has already done
Jason> (especilly since it wasn't working) 

Ah, but the history of what you've tried (as well as an idea of what
your overall comfort level is) will help us figure out what the real
question is.  Most importantly, it tells us *what* you're trying to
do, not *how* you're trying to do it.  (The former is obviously much
more valuable than the latter, since the latter isn't working -- would
you rather help someone by understanding their real needs, or by
reverse-engineering code that you know isn't really doing what they
want anyway?)

There's a phrase we used on the IRC channel: "the X-Y problem":

    They ask how to do X, but that's because they really want to do Y
    but think X will get them there

Jason> when I recieve hundreds of messages per day.

Might I suggest a better mail-reading setup?  I see about 500 messages
a day (as well as following linux-kernel on a private newsgroup feed,
so that's another 150-ish per day).  A combination of procmail and a
newsreader does wonders for dealing with high volumes of mail.

Jason> I formulate a question which is strait forward and includes
Jason> relevent details and code examples.  Easy to read, easy to
Jason> understand, easy to descide if you have the answer or not.

At least one of your replies clarified things (the bit about trying to
preserve the format/structure of an existing codebase, and the
implicit desire to alter the structure of that code as little as
possible while adding the necessary functionality).  That was
completely missing in the original post.

Anyway.  I didn't mean to pick on you in particular; I was just trying
to explain why some people tend to fly off the handle at some
questions, and maybe some ideas that you could incorporate to get
better responses (both less hostile and more informative).  If you are
based in the PDX area and are interested, maybe I can stand you a pint
the next time I'm in the area.


More information about the Pdx-pm-list mailing list