Best database to use with Python

Tim Lavoie tim.lavoie at mts.net
Thu Apr 27 18:04:31 CEST 2000


In article <1255280638-66426487 at hypernet.com>, Gordon McMillan wrote:

>Sometime ago linuxplanet did a comparison (sorry I can't 
>check the link - my ISP is hosed right now):
>http://www.linuxplanet.com/linuxplanet/tutorials/1224/1/
>
>As I recall, Postgres was the most complete DB; MySQL the 
>least complete. MySQL was very fast at some operations 
>(probably related to lack of transaction support), but the most 
>startlingly fast one was bulk load, which does you no good in 
>a CGI situation. Again, from memory, MySQL was fast at very 
>simple queries, but had a very limited SQL dialect and fell 
>apart with fancier queries.

Probably the simpler answer here is for the original poster to have a quick
go at both, he'll form his own opinions soon enough.  :-)

I did a work project involving a search engine of sorts, with a crawler
creeping through the data at night, and a CGI front-end to look for one or
more words. The first iteration used Postgres, but was slow.... The crawler
took forever on inserts, and since the data changed often, there were a lot
of them. Queries for single words were acceptable, but multi-word 'OR'
searches were painfully slow. In this case, MySQL fit the bill very well;
data was added very quickly, and queries were extremely fast. Also, we
didn't need transactions or anything of the sort, since at the very worst
we'd crawl the whole works again.

P.S. This .sig was a coincidence, but I thought it fits anyway...

-- 
Give a small boy a hammer and he will find that everything he encounters
needs pounding.



More information about the Python-list mailing list