Re: Re: Metaphone function attachment
От | Joel Burton |
---|---|
Тема | Re: Re: Metaphone function attachment |
Дата | |
Msg-id | Pine.LNX.4.21.0105041339290.30777-100000@olympus.scw.org обсуждение исходный текст |
Ответ на | Re: Re: Metaphone function attachment (Peter Eisentraut <peter_e@gmx.net>) |
Ответы |
Re: Re: Metaphone function attachmenty
|
Список | pgsql-hackers |
On Fri, 4 May 2001, Peter Eisentraut wrote: > Joel Burton writes: > > > I think, as little things in contrib/, it's easy for people to miss > > these. With a project page, some discussion, etc. (& a place in contrib/), > > more people would be able to use these. > > Most of the extension functions and types in contrib should, in my mind, > eventually be moved into the core. contrib is a nice place for things > that we don't really know how/whether they work, but once we're confident > about the quality we might as well offer it by default. Yeah, but things do seem to languish there for quite a while. (soundex(), for instance, was in contrib when I first looked at PG). Also, some things are in contrib/ that seem a bit out of date (I think there was still some early RI stuff in there last time I went through it) I understand the need not to stuff PG full of *everything* -- and perhaps stuff like soundex(), metaphone(), etc., shouldn't go into the core *. But I think if we leave them in contrib/, after a while, it feels like there's an implied comment on the quality/soundness of the code. Would it work to have a different mechanism for distributing proven yet out-of-the-mainstream stuff, like soundex(), etc. * - soundex(), in particular, should go into the core, though. Many other DBs have it built in, so users could reasonably have the expectation that we should have it. -- Joel Burton <jburton@scw.org> Director of Information Systems, Support Center of Washington
В списке pgsql-hackers по дате отправления: