[Boulder.pm] Boulder-pm Digest, Vol 39, Issue 15

Kate rise at knavery.net
Tue Aug 28 09:59:10 PDT 2012


On Tuesday, August 28, 2012 at 8:40 AM, Ed Dow wrote:  
> Or I've found a more common issue is not allowing time for the version control to take place. Management doesn't want to hear it's going to take an extra day because of a check in and check out process.  

Either you're using the wrong tool or that's not actually the time the _version control_ part of the process takes!  :)  

More seriously, assuming that's actually the time for reviews, testing, sign-offs, etc. yeah, it's a hard sell but potentially easier to sell framed as something else?  QA & testing of any format usually end up getting the short end of the stick[0], but when trying to explain to a major customer doing a high-risk migration that they really didn't want us to run it in an hour and put it live I found it useful to frame things in terms of risk tradeoffs and their tolerance.  It doesn't help at all with the technical debt part unless you can get that idea across as increasing risk of schedule slippage, etc. in the long run, but sometimes it'll buy a lot more understanding of trying to test thoroughly and roll out carefully if you've let them come to their own conclusions about risk.  On the other hand sometime you just get people being bozos and they'll decide retroactively that the risks were more than you told them.  *sigh*

On that downer note, perhaps it's time for the eternal version control system flame war:

RCS or SCCS?  Clearly if it's good enough for BitKeeper SCCS is good enough for anyone!


[0] Weird idioms, hurrah?

—  
Kate



More information about the Boulder-pm mailing list