httpd-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Brian Behlendorf <>
Subject Re: all clear
Date Sat, 08 Jun 1996 01:09:36 GMT
On Fri, 7 Jun 1996, Randy Terbush wrote:
> > The first two (vhost docroots, mime.types) I will +1, the second I still 
> > want to somewhat stew on before releasing it.  I think putting that 
> > conglomeration outside the server might be a better idea (like into a 
> > perl script).  
> I disagree. SIGHUPing the server would then require you to rerun
> your perl script. It moves a fundemental part of the server's
> operation to an external program that is susceptible to it's own
> problems. It's a simple little piece of code that is quite useful
> for config management and cannot effect the servers ability to do
> it's job if it fails. A trivial change like this should not take
> a month to get into the code base. 

Every feature addition to the core code should be carefully considered.  
Ease of implementation is not an issue - ease of support is.  How many 
people will come to rely upon this, and kick and scream when we have to 
say "there's a better way"?  I guess I am the one at fault for not being 
more elucidating about my objections, and since I can't really think of 
any beyond "maybe there's a better way", I acquiesce and won't veto it.  
All your points above are correct... but similar to how we have resisted 
more creative logging solutions since it would complicate core code "and 
can be done anyways using a logging process", I was resisting this.  

> > I can't think of any other bug-related patches - 
> > 
> > > We're also gonna want to sanity check builds from this release before
> > > we actually go and announce it as 1.1b3, I think. I can do several of
> > > them tomorrow.
> > 
> > Me too.  stars.patch and spath.patch can wait until after b3 imho.  
> We had enough votes to include them though. Why put it off?
> They just get lost in the shuffle.

I'm sorry if it seems like I'm being a bottleneck here, but our *primary* 
interest at this point should be to flush out bugs, not add new 
features.  stars.path I see as not quite a feature addition, not quite a 
bug, and I suppose I can acquiesce on that.  spath.patch I want to stew 
on a little bit more... just to consider how we would use it, or if its 
inclusion would affect us at all.  If for *no*other*reason* than it 
affects the core, not just a module.

But, if three +1 votes come in for that patch, then by the apache rules 
(which I agree with) my sloth is voided.


--=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=--  |  We're hiring!

View raw message