httpd-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Aram Mirzadeh <>
Subject Re: patch - vote
Date Sat, 10 Feb 1996 23:03:16 GMT
> >  -1 72a.mod_imap_overhaul This didn't work for me... I had to patch it in by
> >  			 hand, and then it complained about index being 
> >  			 redeclared. 
> Sorry, but I don't think this is criteria to veto a patch. It's impossible
> to make changes that are guaranteed to work on every OS that you *don't*
> have access to. If it's a showstopper compile problem, then we need to
> be flexible enough to accept a last minute patch to fix the problem.
> *Especially* when there are this many patches.

Sure it is... if a patch doesn't work, and I can't fix it just by looking 
at it.  There isn't much I can do about it... I tried the fully patched 
version uploaded, but then I couldn't tell which patch was at fault.  If 
you want that to change you have to change the guidelines.. I think that's
why we have the week... to fix the small problems. 

> >  +0 99.bind.patch	Doesn't compile on linux alone / doesn't patch with
> >  			other patches 
> >  
> >  http_main.o(.text+0xd0f): undefined reference to `FD_ISSET'
> >  http_main.o(.text+0x1270): undefined reference to `FD_ZERO'
> >  http_main.o(.text+0x129c): undefined reference to `FD_SET'
> Again.

Again what?  I have no freaking idea what the devil is FD_ISSET?! or which
library it's supposed to go into.  You can't expect me to say yeah it's okay
what happens when the release is out, and it doesn't compile on half of the
OS's out there?

> >  -1 100b.os_port.patch	I think we should build the next release, and make the
> >  			OS/2 port available as a patch.... I think it's more 
> >  			the other patches that don't allow OS/2 to patchin/
> >  			compile cleanly.
> Garey has been waiting patiently for nearly 6 months now. I think that is
> long enough. The changes appear to only effect the __EMX__ defined code.
> I see no reason to keep vetoing these patches.

Hey, I have no problem with Garey, that's why I'm proposing the build... 
not the official one, but take a set of patches that everyon agrees one, 
build a release... let him Patch it up with OS/2... ( which I personally 
use ) and then do another test on all other platforms... this way we know
OS/2 stuff is going to work.

> >  I propose we change the rules of accepting/declining a patch as well... we
> >  should change the decline of a patch from a single -1 vote to a minimum of 2
> >  or 3 people, more and more people are getting involved, and a +5 -1 vote 
> >  shouldn't make it a bad patch... if it's a OS thing okay, but just on marrit
> >  no.
> That, or just remember to be somewhat flexible as we hash out this voting
> stuff.

-1 doesn't mean no way no how.. it means either I couldn't patch it, didn't
work on my platforms, or I don't think it's a good idea, to do this. 

the 3rd part is where I have the problem.  If it passes the first 2 and
9/10 like it, then the last person shouldn't rule the castle.  I don't think
majority rules works either, but maybe a 5/6 or a 7/9 would do... 


			       Aram Mirzadeh,

	43rd Law of Computing:
        	Anything that can go wr
	Segmentation violation -- Core dumped

View raw message