This is the mail archive of the guile@cygnus.com mailing list for the guile project.
Index Nav: | [Date Index] [Subject Index] [Author Index] [Thread Index] | |
---|---|---|
Message Nav: | [Date Prev] [Date Next] | [Thread Prev] [Thread Next] |
Jim Blandy <jimb@red-bean.com> writes: > > For a practical scripting language, you don't need "best". > > Okay, this is not so cool. I don't see any reason to embrace sloppy > interfaces, just because we're an interpreter. Sometimes people do > want to get something exactly right, even in scripts. I'm not saying you shouldn't go for "good", or even "as good as you can make it", but from a practical engineering viewpoint, "now" is of value, especially when accompanied by "pretty good", and "useful". But basically we all agree, I think. The DBI people have been thinking about this for a while, and they have something which they feel works, so it makes lots of sense to look at that before starting to think from scratch. (Especially since the details of some of this varies from database to database---if you design an interface for a database which doesn't allow precompiled queries, then that's not going to be so efficient on a database which does.) Having said that, the last time I used DBI, it didn't seem too clean to me. Perhaps such messiness is required, but perhaps more radical rethinking could produce something better. I don't know anything about ODBC, but perhaps an interface based on that would be better?