Re: Re: cvs: php4 /ext/curl config.m4
| From: | Sterling Hughes | Date: | Sat, 09 Nov 2002 16:44:40 +0000 |
| Subject: | Re: Re: cvs: php4 /ext/curl config.m4 | ||
| References: | 1 2 3 4 5 6 | Groups: | php.dev |
| Request: | Send a blank email to php-dev+get-90705@lists.php.net to get a copy of this message | ||
> On Sat, 9 Nov 2002, Sterling Hughes wrote:
>
> >> On Sat, 9 Nov 2002, Sterling Hughes wrote:
> >>
> >> >>
> >> >> There is no such release of Curl yet. This makes testing
> >> >> the RCs quite a hassle now so either you revert those changes
> >> >> or get the curl folks to release this 7.10.2.
> >> >>
> >> >> --Jani
> >> >>
> >> >
> >> >7.10.2-pre4 is currently out, it should be released quite soon.
> >>
> >> But it's not 7.10.2 release. Revert the changes.
> >> Or I will.
> >>
> >
> >err... no.
> >
> >7.10.2 will be out long before PHP 4.3. We haven't even had the first
> >RC for PHP, and its a lot longer before PHP releases to come out. If
> >you want to test it for QA purposes, you can easily install one of the
> >prereleases. 7.10.1 might also work, not sure....
>
> You're now missing the point here. I'm NOT doing QA for Curl.
> I'm not going to install ANY curl pre-release here.
>
> You can (if you wouldn't be so lazy) add some ifdef's around
> the new stuff instead of everytime requiring people to update curl!!
Its not laziness... I'm not having that many '#ifdef's in the code.
Some cases it makes sense for cURL to have #ifdef's, in some cases it'll
just make the code both unreadable and unmaintainable.
The solution I'm working on is to autogenerate from the curl.h and
multi.h files. However, that's after i get the multi interface in, and
the cURL extension is hopefully up-to-date.
Anyhow, I get the point. I just don't find it valid, if you're
doing QA, than upgrading to a pre-release is not that much harder than
upgrading to the release itself.
PHP-Curl is _not_ that often tested in the QA process, simply because
most features rely on a server being available, and with PHP/QA there is
none. Therefore, very few regression tests.
-Sterling
--
Sterling Hughes <sterling@bumblebury.com>
Did I help you? Consider a gift: http://wishlist.edwardbear.org/