Re: breaking out DB dsn parser
| From: | Alan Knowles | Date: | Thu, 19 May 2005 06:11:31 +0000 |
| Subject: | Re: breaking out DB dsn parser | ||
| References: | 1 2 3 | Groups: | php.pear.dev |
| Request: | Send a blank email to pear-dev+get-37750@lists.php.net to get a copy of this message | ||
You better hope I get more time to work on DBDO_pdo (the version which
uses's pdo's backends as drivers.. ) ... ;)
Regards
Alan
On Thu, 2005-05-19 at 00:17 -0400, Lukas Smith wrote:
> Robin Ericsson wrote:
>
> >> Some background infos:
> >> PDO uses a dsn string format that is very much based on the idea that
> >> one should use the native dsn string format where possible and use
> >> something that is as easy as possible to parse and stick into each
> >> specific driver. Therefore the format is very different for each
> >> RDBMS. As a possible remedy for this situation we agreed at the PDO
> >> meeting in Cancun to add a function that will accept an array and will
> >> return a dsn string specific to either one of the chosen drivers.
> >
> >
> > We not make it accept a dns-string instead of an array? The dsn has been
> > around for a very long time.
>
> the problem is that Wez pushes several core philosphies that I
> personally dont agree with. He also seems to have little interest to
> worry at all about the actual migration experience for either php
> database extension or PEAR users and instead choose to create an API
> that is very much based on the first few native API's he messed around
> with when writing the respective PDO drivers. He also believes in
> keeping the number of classes and methods to a minimum and prefers to
> make things context sensitive. So it goes. Like I said I have it one
> last shot at changing his POV and a function that generates PDO style
> dsn strings from arrays is the most I got.
>
> regards,
> Lukas
>
> PS: dont go running down Wez door. He will be announcing the meeting
> results shortly (I already mailed him my summary). This summary will
> likely explain his principles while designing his API. Since I disagree
> with him still (but have accepted that this will not change) I obviously
> have a hard time presenting his side of the story in a convincing manner.
>