Re: Proposed revision to ingres extension
| From: | Pierre | Date: | Tue, 11 Sep 2007 19:52:28 +0000 |
| Subject: | Re: Proposed revision to ingres extension | ||
| References: | 1 | Groups: | php.pecl.dev |
| Request: | Send a blank email to pecl-dev+get-4863@lists.php.net to get a copy of this message | ||
On 9/11/07, Grant Croker <grant.croker@ingres.com> wrote:
> Hi,
>
> I work on the the pecl/ingres extension and have been working on
> bringing it up to date. Unfortunately some of the changes have required
> a change in the order of parameters to certain functions. The main
> driver for this change was the addition of support for multiple
> concurrent result sets.
>
> Previously the driver would close any active statements when
> ingres_query was called. In order to manage this change there is a new
> resource that is used for a result set returned ingres_query(). Where as
> before the ingres_fetch_{row|object|array} functions would take the link
> resource to fetch the result, now these functions require the result
> resource be passed. In addition the link resource was an optional
> parameter to these functions, it is now mandatory.
>
> After asking on #php.pecl what to do about this BC breakage, a new
> extension name was suggested, ingres2. I now have code ready to submit
> into pecl, see http://croker.net/php/ingres2-0.9.0.tgz.
>
> If there are no objections can someone create a pecl/ingres2 directory
> on cvs.php.net and give me karma?
Why not simply bump the version to 2.x? That's how we define possible
BC breaks in PHP (when we break BC :) and it is also well known in
PEAR.
Doing so you can simply create a branch for 1.x (INGRES_1 for example)
and use HEAD for 2.x. The CVS history is also kept (and that's a real
good thing :)
Cheers,
--Pierre