httpd-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Jim Jagielski <>
Subject Re: slight API change
Date Sat, 02 Mar 1996 20:15:05 GMT wrote:
> 	What about the peer review? Don't you still need to get peer
> review before you commit?

Doesn't sound like it. Instead it sounds like once I'm able to commit,
I can do whatever I want to the development tree; If it causes YOUR OS
heartburn, oh well, it's up to YOU to remove the offending code. Also,
if I add some code that happens to work like a charm on MY OS, but causes
every else's to crash and burn, then I can do that as well and you won't
know until you try to run the latest dev-version, and then it'll be up
to YOU to figure out just which of the many changes caused the problem.
Once that's done, then someone will need to pull off the offending code
from the tree... All because I didn't bother to have other people check
out the changes before committed. 


	The Devil's Advocate

PS: Not that we don't need some sort of version control, but the current
    setup appears to make things VERY murky. Instead of there being
    an official baseline and then various patches there is instead a
    development system that includes patches THAT THE SUBMITTER DECIDES
    IS WORTH INCLUDING (committing) at which point it's now part and
    parcel of the new development baseline. I don't trust MYSELF that much!

    With the older method, there was a baseline and, for example, 12
    patches. We could test each diff-file out and then vote on whether
    it should be included as the next phase. If someone's patch didn't
    work for me, I could ignore it and fold in the one's that did. I
    would also know for _sure_ which patch/diffs I would vote against.

    Now, as I understand it, once I get the latest development version,
    it includes all the patches that people have committed already. Thus
    the version on Monday may be "quite" different from Fridays and a patch
    committed on Wednesday my not work for me. I have no way of knowing that
    until it's compiled and it dies. Then I need to backtrack and figure
    out which commits between Mon and Fri was the culprit.

    Maybe I'm just not too clear on how this is being implemented and
    am showing my utter ignorance on the direction this is going, but if
    the above is at least partly true, there's potential for mega amounts
    of confusion. I don't want someone who has the ability to simply start
    making additions and improvements willy-nilly with no control over
Jim Jagielski  << >>   |      "That's a Smith & Wesson,
  **  jaguNET Access Services  **      |       and you've had your six" 
      Email:          |             - James Bond
++         +++      Voice/Fax: 410-931-7060       ++

View raw message