Weekly Edition
Return to the Press page
| |||||||||||||
|
| |||||||||||||
A CTO gives thanks (InfoWorld)
InfoWorld CTO Chad Dickerson
shares some things he is thankful for. "Open-source development
continues to make my job easier by providing solid solutions that work both
technically and financially. During the boom times, a common job benefit
and recruiting tool of the CTO was to sanction work on open-source projects
during company time. Now that the economy has tightened, I don't hear much
about this kind of arrangement any more, yet the quality of work in the
open-source community continues to be impressive because of the passion of
developers committed to improving technology despite a harsh macroeconomic
environment. Linux may be old hat, but the kernel continues to improve in
important ways for large businesses." Thanks to Lenz Grimmer
(Log in to post comments)
Why people don't use the best tools? Posted Nov 28, 2002 10:00 UTC (Thu) by leandro (subscriber, #1460) [Link] People are still acting in herds instead of doing their own research...The guy talks about the Linux 2.6 scalability, and then in the next paragraph about MySQL... never mind PostgreSQL is simpler, faster, more scalable, standards-compliant, a real SQL DBMS, and MySQL is none of this.
Why people don't use the best tools? Posted Nov 28, 2002 12:14 UTC (Thu) by mrness (guest, #8271) [Link] This is the first time when I hear someone who claim that postgres is faster than mysql.Have you tested both? If not (which is pretty obvious) maybe is better to keep your opinions/prejudices for yourself. I'm not contesting the fact that postgres is RDBMS and mysql is not.
Why people don't use the best tools? Posted Nov 28, 2002 14:25 UTC (Thu) by Wol (guest, #4433) [Link] Faster at what?MySQL was *designed* for *fast* *retrieval*, and as a general purpose database it just sucked. But as a web server it blew everything else away. It's now catching up on the general purpose speed stakes - I don't know where it actually stands in the field nowadays. But then, nothing beats Pick/MultiValue for sheer speed, especially on underspec'd hardware ... Cheers,
Why people don't use the best tools? Posted Nov 29, 2002 1:59 UTC (Fri) by job (subscriber, #670) [Link] Please stop repeating that nonsense. MySQL is a "real" RDBMS (for any sane definition of "real") nowadays, but there's a slight performance hit compared to the old ISAM-style tables.
Why people don't use the best tools? Posted Nov 29, 2002 9:35 UTC (Fri) by mrness (guest, #8271) [Link] R from RDBMS stands for "relational".Are mysql capable to define relations between tables (as in enforce data integrity, cascade updates/delete and so on)? I think not.
Why people don't use the best tools? Posted Nov 29, 2002 20:18 UTC (Fri) by tizzyd (guest, #8304) [Link] Two things:1) thread is a bit off topic, and 2) nothing in a relational requires cascaded updates or referential triggers. R simply stands for relational, that is, using the relational syntax to identify, relate, update, and retrieve data. IMHO, doing a table join qualifies a relational database. Now, is it a relational database engine?! That's another matter. DBF files qualify as a RDMBS, as you can perform a join (using a local memory based engine). I think what you are hoping for is a relational engine, something that PostgreSQL and MySQL both have. PSQL has more to it, but both are relational databases. Lastly, just a side note, many of that *crap* people say they need--referential triggers, etc.--are just ways of hiding business level detail in the database layer. That approach was fine a while ago, but with today's architectures, I don't want to have to let a transaction fail because of bad data if the app could have enforced good data going in. An ounce of prevention is worth a pound of cure. ;-)
Why people don't use the best tools? Posted Dec 5, 2002 12:21 UTC (Thu) by Peter (guest, #1127) [Link] many of that *crap* people say they need--referential triggers, etc.--are just ways of hiding business level details in the database layer. As opposed to hiding business level details in the applications? Face it, you have to hide your business level details somewhere. I don't want to have to let a transaction fail because of bad data if the app could have enforced good data going in. An ounce of prevention... And I, on the other hand, don't want to corrupt the database state because the app developer wasn't able to keep track of the necessary integrity constraints. Really, you should have both - the app should do its own sanity checking before trying a transaction, so that it can do the right thing with regards to bad user input. But, given that applications will have bugs, it is better for the transaction to fail then to succeed and produce an inconsistent data state. Another reason referential integrity is important is that you often have multiple applications that use the same database. Enforcing exactly the same integrity rules in each app is a losing game, then. (It should still be done, for reasons mentioned above, but it's a lot easier to develop correct apps when the developer sees transaction failures instead of silent corruption.)
Why people don't use the best tools? Posted Aug 7, 2003 23:03 UTC (Thu) by leandro (subscriber, #1460) [Link] > MySQL both have. PSQL has more to it, but both are relational databases. None are. MySQL violates both the SQL standard and the relational model more than PostgreSQL, and it also violates the minimum DBMS requirements. BTW, SQL itself violates the relational model, remember?
|
|
||||||||||||