Re: pear installer in PHP 5.1 RC1
| From: | Clay Loveless | Date: | Sun, 14 Aug 2005 17:10:24 +0000 |
| Subject: | Re: pear installer in PHP 5.1 RC1 | ||
| References: | 1 | Groups: | php.pear.core |
| Request: | Send a blank email to pear-core+get-3535@lists.php.net to get a copy of this message | ||
Pierre-Alain Joye <pierre@dotgeek.org> wrote:
> It is useless and does not bring any additionnal flexibility.
Actually, it brings a lot of additional flexibility. One of the biggest
complaints about PEAR out in the wild is the inclusion of bloated code. Even
if someone can include PEAR/ErrorStack.php specifically rather than the
entire PEAR.php file, developers who wish to leverage PEAR_ErrorStack must
either manually unbundle the ErrorStack code, or package up the whole PEAR
installer.
Again, this is a bloat issue. Enabling developers to use ErrorStack
functionality alone, without having to worry about the rest of the PEAR
installer in their applications, is a huge benefit, is _very_ useful and
adds considerable flexibility.
> But it brings one more check when we bundle, for example.
I don't see how this is a big deal. Given that Greg is the lead developer on
PEAR 1.4.x and PEAR_ErrorStack, I don't think checking on this external
dependency will be a major inconvenience for anyone.
> If there is a critical bug in PEAR, we just release it another PEAR 1.4.x+1
> for example, no problem with that.
What about non-critical bugs? By breaking things down into smaller
components, it's easier to make smaller, incremental enhancements -- as you
already know, I'm sure.
> PEAR_Error is in pear, the ErrorStack is something like a version
> two and belongs to PEAR. As it was defined during the proposal
> stage. It's a "core" source file and must be in the main package.
Surely it's conceivable for a developer to change positions on a package and
it's most appropriate place after the proposal stage?
You may want to review this thread:
http://marc.theaimsgroup.com/?t=109607852600004&r=1&w=2
Greg expresses a desire to release updates to PEAR_ErrorStack more often
than the typical release cycle for PEAR core allows, which would be good for
everyone who is interested in the benefits that ErrorStack offers.
I'm a strong +1 on breaking ErrorStack out as proposed in the above thread.
-Clay
--
Killersoft.com