This might come off as ironic, but the proposal for type hinting illustrated to me quite well what kind of restraint people have put on themselves in order to work together for the greater good of releasing 5.3.0. Let me explain: To me even the thought of talking about adding a language syntax feature right after 5.3.0 is out is so inconceivable that I cannot even begin to express how shocked I am that a long time RM like Ilia would bring this up and then moan about discussions being needles bickering and preventing him to commit the patch within less than a week. Even more surprising is that apparently so man people thinks its a good idea to do so. This is not because I think that strict type checking is not a good fit for PHP, but simply because putting this in CVS now means, that we have no time to let the feature mature in CVS for a while as we need to make a bug fix release fairly soon. Adding a language syntax, even if its an optional syntax, should never be pushed into a stable release so quickly imho.
Now, given this displayed attitude about releases processes, I am now starting to see the level of constraint that people were giving Johannes and myself over the last few months. I was starting to get a bit annoyed at complaints about the short commit freezes we put into effect. But overall people still swallowed it down and so the main stumbling blocks in the last months were more time constraints of key people, which is simply a fact of life and nothing I can or will complain about. So I want to thank you all for having been so willing to work with Johannes and myself doing this release and accept what we felt was best to get the release out. I appreciate the amount of trust you have given us!
I still think that PHP could benefit from a rethinking of the release process, specifically to make sure that we do not need these commit freezes anymore (RM has their own branch and can easily cherry pick patches) while still ensuring that really big stuff like doesn't get delayed again and again because people get frustrated over smaller, yet non trivial, patches not making it into a stable release quickly enough.
However the feedback I got from my "competing" RFC to Ilia's proposal once again illustrated to me that I simply do not have what it takes to write a clear and easy to understand RFC (maybe I should have realized that back when I was trying to document the namespace decisions). Plus in recent months I felt my patience for humanity slowly fade away, which is not however a direct cause of me RMing 5.3. This certainly has causes through out my entire professional and personal life. Its just a phase I am going through. As a result, I will leave it to others to decide if they want to do some formal adjustments to the current implicit PHP development and release process.
Seems like internals@ is not as insane as I began to fear and now people have finally realized that Ilia suggested this for 5.3. Now this does not reduce the fact that I am thankful for how well we all pulled together towards the end.
Release early, release often.
Branch PHP5.4 and release it in 3-6months? :)
With svn coming soon, implementing your suggested release process should be much easier to introduce. Especially if the RM is using something like git to pull from svn and merge.
5.4 in 3-6 months?
-__toScalar magic Method
[-multithreaded PHP-CLI *SCNR*]
Its realistic since most of them already have working patches.