RE: [PHP-QA] RE: [PHP-DEV] Wierd CVS Build problem with libmysql

Looking forward to seeing the RFC. (I removed the MySQL guys from the cc. Poor guys are getting spammed ;) Andi At 09:23 PM 11/6/00 +0000, James Moore wrote:

Andi,

IMHO, this looks useful. Late last week I thought Zeev was asking for last minute do or die stuff, with a view to putting out 4.0.4. Is this going to affect that, time wise? Maybe it should wait until after 4.0.4?

I think it should wait for after 4.0.4 because I understand this is a non-tested MySQL patch? Also it would mean that our MySQL extension update would also need a bit of time to stabilize (even though it probably would not be a big patch). I think it would be good to get 4.0.4 out the door by next week. Having spoken to zeev yesterday about the release process I think it is important to get somthing slightly more formalised in place for our release cycle to give the QA guys a fighting chance to find showstoppers and test properly. I have been working on a RFC about this and will post it to php-dev & php-qa in the next day or so on a possible cycle. I think that it would foolhardy of us to think that this cycle will take any less that 10 days but I am also suggesting a branching on the CVS at RC1 (Dont shout yet see the rfc first). At this point then anything can be added to the main tree without effecting the RC. There are various reasons that I feel that it is valid to branch the cvs but the main one is to stop any rogue code appearing in the CVS between RC1 and actual release. So next week might not be attainable for the actual release but I do think we should start the process by next week. Although you and zeev are both full time none of the QAers are and most of our work is done in evenings and weekends so we need more time than the two of you to get our work done. The reason why it has been taking so long because there were quite a few open issues raised (most of them have been fixed by now). Are the build problems with the new MySQL client all solved now? I'm also cc'ing the PHP QA guys who might have a better idea of the issues which were raised. James

--- Andi Gutmans <andi@zend.com> http://www.zend.com/