RE: [PEAR-DEV] general questions

Tomas, I am very much with you .. actually all the way :-) From what I gather the reason why metabase was slow was because of the way the fetch was implemented .. maybe a couple things here and there can use some tweaking .. but all in all this seems to be the major difference. As I said this can easily be changed (I already did it) I should probably take the benchmark on the ADBODB site and run it after my little modifications and see if that helped at all. Then of course right now there are a ton of functions in the core part. I mean in the end all the DB abstraction layers call the same functions. Then they all provide for a way to fix syntax differences in the supported RDBMS through the use of a few functions that return the proper syntax. Actually aside from coding a ton of features the main trick in the DB abstraction layer lies in making sure that everything works ... I do not see major differences in the design. Obviously there are little things that mostly boil down to differences in taste. But do share your ideas. I must admit I also did not find anything new over on the ADODB sites explanation on why it rules. I assume he is saving it all for his book :-) So what Tomas is suggesting (and I agree with this is): Write the goals -> write the tests -> write the API -> write the code I will add: -> provide wrappers Obviously not the same people will have to do this. Lets get going! I will write up my goals tomorrow. Best regards, Lukas Smith smith@dybnet.de _______________________________ DybNet Internet Solutions GbR Alt Moabit 89 10559 Berlin Tel. : +49 30 83 22 50 00 Fax : +49 30 83 22 50 07 www.dybnet.de info@dybnet.de _______________________________