> Let's talk about this at the Zend conference.
>
I agree with that, the four of us will be ine one place at one time, I just 
wanted to start the topic in order to get thoughts from those who won't be 
there or won't have time.

> I've only had time to skim your email, but what I had in mind is more
> C than A or B on your list.
>
Nod, I like this one best (of the options so far) too.  Based on talking to 
Ilia I realize the INI option is unnecessary.  From what I understand Ilia 
would like to see the user required to specify the encoding via a driver 
parameter, though I think in many cases we can allow the driver to have a 
default which can be overridden via driver param only if necessary.

> the PHP 5.2 branch is where it's at for PDO, as the unicode APIs were
> changing too wildly to make it feasible for me to keep everything in
> sync.
>
More than understood.  Things have settled a bit in terms of the APIs, but 
there's still a few tweaks going on.

> Given my lack of free time, I'm still in favour of finding a way to
> avoid branching and merging PDO too much while we maintain the PHP 5
> branch, as there are so many PDO extensions that it will become very
> easy (for me at least) to forget a vital merge.
>
I would think we'd make these changes as ifdef'd bits in whatever branch is 
the primary PECL release branch (5.2).  They'll be compartmentalized enough 
that I think we can avoid making it hard to work around, and it would allow 
andrei's preview-release plans to move forward without forcing a new branch 
on PDO maintenance.

Basicly, we'd have the same situation we have now.  Any release of HEAD 
requires bulk copying 5.2's PDO code into HEAD's cvs tree (either via 
backend hackery or megacommit to sync it up).

-Sara

-- 
PHP Internals - PHP Runtime Development Mailing List
To unsubscribe, visit: http://www.php.net/unsub.php

Reply via email to