Shawn Walker wrote:
So, at this point, I think it's just best to ditch the catalog cache
completely.
I can revisit making other scenarios faster (such as all installed
packages or other scenarios we discussed) later.
Thoughts?
After further discussions and research, it seemed better to me to keep
the catalog cache for these reasons:
* the slower the system, the more benefit the catalog cache provides (my
desktop system saw a 400-1200 milliseconds difference typically between
cache and no cache while my laptop saw a 1-3 second difference typically)
* significantly more analysis is needed to determine whether other code
(such as fmri parsing) can be improved further to the point where the
cache isn't needed
* there was some uncertainty surrounding the reasons what specific cases
the catalog cache was necessary for originally
Cheers,
--
Shawn Walker
_______________________________________________
pkg-discuss mailing list
[email protected]
http://mail.opensolaris.org/mailman/listinfo/pkg-discuss
