PostgreSQL Anniversary Summit a success (NewsForge)
[Posted July 10, 2006 by ris]
NewsForge has a report
from the PostgreSQL Anniversary Summit. "This weekend marked the
10th anniversary of PostgreSQL's posting as a public, open source
project. To celebrate, the PostgreSQL project held a two-day conference at
Ryerson University in downtown Toronto, Ontario, Canada."
(Log in to post comments)
PostgreSQL Anniversary Summit a success (NewsForge)
Posted Jul 11, 2006 6:29 UTC (Tue) by alvherre (subscriber, #18730)
[Link]
There's more coverage of the Summit in Planet PostgreSQL, including a group picture of (almost) all attendants. Some people actually wondered that it may be dangerous that Oracle gets this picture; I personally consider myself out of danger.
btw
Posted Jul 11, 2006 14:47 UTC (Tue) by ccyoung (subscriber, #16340)
[Link]
Happy Birthday PostgresQL!
Certainly my favorite, most trusted, most integral part of AMP of the LAMP stack. PG is so refreshingly sane, purely vinalla with a few nuts thrown in for those who like to crunch.
btw
Posted Jul 11, 2006 16:54 UTC (Tue) by nix (subscriber, #2304)
[Link]
Vanilla? I'm not sure how I'd describe the rule rewriting system, or table inheritance, or the completely-extensible type system, but 'vanilla' is not it.
I think it's whizzo in a pinstripe suit. :)
vinalla vs nuts
Posted Jul 11, 2006 19:35 UTC (Tue) by ccyoung (subscriber, #16340)
[Link]
vinalla core - compare to, for example, text types in MS-SQL or MYSQL. compare to, for example, joins in Oracle.
as to the nuts, should you choose to go there:
yes, rules are beautiful and an may instances literally rule
table inheritance has never been fully implimented, eg, cannot inherit constraints or triggers or about anything else except a few columns
extensible types are cool, and need to be played with once out of principle (paula), but for sanity and maintenace seldom used in production
or more widely used, subtle extensions to SQL like update from, which are quite nice and easy to parse by anyone SQL literate
Aesthetically, PG does not leave a bad taste in my mouth, which I cannot say for any other SQL.
vinalla vs nuts
Posted Jul 11, 2006 19:41 UTC (Tue) by nix (subscriber, #2304)
[Link]
In the sense that it makes the easy things easy and the hard things possible, I suppose that it is 'simple'.
(I'm not sure what to call most other RDBMSes, in that case: Neanderthal? Or is that offensive to Neanderthals, who were quite sophisticated really?)
AIUI, inheritance of constraints is implemented and will be in 8.2.
And personally I use extensible types all the time. They make SQL usable :)
I sit corrected
Posted Jul 12, 2006 1:27 UTC (Wed) by ccyoung (subscriber, #16340)
[Link]
I think you are much more of the student of SQL than myself (personally limited to PG, Oracle, MS-SQL, and MySQL). I and probably others would enjoy an overview of your assertions. Do you have anything in writing you can share?
Did not know that full table inheritance was taking place - this is really good news and I look forward to revisiting - have done many elaborate work-arounds.
I also sit corrected on extensibility: my point is that extending a single app among a dozen or so is, for me, asking for trouble.
Lastly, SQL is, after all, SQL. Every young Turk who sees it (of whom I have been one) decries is stodginess and oblique expressiveness. At this point in my professional career I'm (happily) too busy finishing projects to re meta-think SQL, although I'm still sensitive to inelegant code.
Also, most modern programming (Java, .Net) tends to demote the database to a simple repository, trying to capture the business rules in the middle tier with communication objects for example. SQL is a bad impedance match for this. Some of us (me) prefer use instead-of views for user interfaces and triggers to keep the database aligned. OK but not great.
Again, please share your enlightenment.
Extensible types
Posted Jul 11, 2006 21:11 UTC (Tue) by flewellyn (subscriber, #5047)
[Link]
Extensible types aren't used much in production? Woah there. I think the many happy users of PostGIS, including myself, would beg to differ on that score; the entire library is basically a massive exercise in type extension. And very much used in production, in many places!