On 15.2.2005 17:29 Uhr, Zeev Suraski wrote:
At 07:26 15/02/2005, Rasmus Lerdorf wrote:
D.Walsh wrote:
On Feb 14, 2005, at 23:49, Adam Maccabee Trachtenberg wrote:
Well, that's below 2.5.11, which is what we currently require, so those folks are already out of luck.
Meanwhile, Mac OS 10.4 is at 2.6.16, so that's okay. I don't have a 10.3 machine with me here at LinuxWorld, so I can't check that.
OSX 10.3 is at 2.5.4
At the same time, why would people on older operating systems who are obviously quite conservative when it comes to upgrading suddenly try to upgrade to the very latest PHP? As long as we don't move the goalposts beyond the latest releases of the various main operating systems I think we are fine.
At the end of the day it's a fact that lots of people use the latest version of PHP on exceptionally old systems. Some of the reasons I can think of off hand:
- Because they're new comers to PHP, and the first version they try may already be too 'demanding' for their server
- Because of a (somewhat justified, but not quite) perception that the latest (major) PHP version is more secure, and because PHP is much more visible to them than some library(*)
- Because they want specific features in the latest version of PHP, and the same (*) applies here too.
There are probably other reasons I didn't think about. People who follow PHP don't necessarily follow all of PHP's dependencies. In my experience they rarely do.
Zeev
(*) The average PHP user will easily tell you that he's using PHP, but he's much less likely to know that he's using libxml2, and will almost definitely not know which version of libxml2 he's using. We can't assert that because someone is upgrading to new major versions of PHP, he shows the same level of interest in OS/library updates.
And for that reason, it's maybe a good thing, if we require a more recent libxml2. Because not all things can be supported with 2.5 and therefore people will get confused, when a feature doesn't work on their system... Furthermore it's a pita for reusable-code-writers, if it doesn't work on some installations (and they would have to test their scripts against 2.5 and 2.6).
But basically, for me it boils down to the points I made before. Testing and maintaining. Each time we add a new feature or change some things, we have to compile and test it against 2.5 and 2.6 . Looking at the very limited amount of people working on the libxml stuff (it's mainly Rob currently) and the disinterest of me and Rob (don't know about the other XML people) in 2.5, this isn't really the way to go for me.
Besides the fact, that some things just can be done nicer and easier in a 2.6-only environment.
And it's not that we want to change the requirements from 5.0.3 to 5.0.4, but only for 5.1 (just to make that clear).
chregu
-- christian stocker | Bitflux GmbH | schoeneggstrasse 5 | ch-8004 zurich phone +41 1 240 56 70 | mobile +41 76 561 88 60 | fax +41 1 240 56 71 http://www.bitflux.ch | [EMAIL PROTECTED] | gnupg-keyid 0x5CE1DECB
-- PHP Internals - PHP Runtime Development Mailing List To unsubscribe, visit: http://www.php.net/unsub.php
