apr-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Garrett Rooney <roo...@electricjellyfish.net>
Subject Re: [PATCH] Configurable read/write buffer sizes
Date Fri, 14 Oct 2005 16:39:44 GMT
On 10/14/05, Colm MacCarthaigh <colm@stdlib.net> wrote:
> On Fri, Sep 09, 2005 at 02:57:40PM +0100, Colm MacCarthaigh wrote:
> > O.k., I've done some new patches and a few read throughs to think the
> > implications of a large buffer size out, and implemented a flags-capable
> > variant of the stdin/out/err openers.
> These look simple enough to commit, the actual patch is;
>     http://people.apache.org/~colm/apr-IO-patches/001-flags-stdio.patch


> The variable buffer size patches on the other hand; while I have it
> tested on unix and win32, they will likely break on netware and os2, as
> my patches there are based on guesswork.
> I'd like to avoid knowingly breaking trunk for any platform, so I think
> creating a branch is a simple enough workaround, and when it works for
> all platforms, merge it into trunk. Looking at the subversion repository
> though, we don't seem to use branches for anything but the specific
> versions of apr. Is this intentional? Any problems with creating a
> branch for this? (or any problems with the idea of variable buffer sizes
> in general?).

I don't have any objection to the patch.  Historically, we haven't
used branches for things like this.  I think the general way things
have gone is that platforms that aren't easy for most developers to
hack on get temporarily broken, and before the next release their
developers clean up the mess.  I don't know if that's the policy, but
it's what I've observed in the past.

I'd say as long as nobody objects in concept to the patch, and you can
write a test for it so that the OS/2 and Netware people can confirm
that it works, there's no reason not to commit it to trunk.


View raw message