As requested, some high-level thoughts: * Would you comment the code that deals with the catalog cache more thoroughly? It took me a couple of passes to parse cache_fmri(), and even now, I don't understand why the structure is so complicated.
* If we're spending all of this effort to render the text-based catalog into an easily cachable binary format, why not just make the catalog a binary format and be done with it? * Why did you add the additional call to symlink in the install_file logic? A comment explaining this would be nice. -j On Mon, Jun 16, 2008 at 08:45:59AM -0700, Danek Duvall wrote: > http://cr.opensolaris.org/~dduvall/pkg-newlist/ > > It's rough -- there are still some things I want to do, like see if I can't > reduce inventory() and get_matching_fmris() to a single function, but it's > fast, passes the testsuite, and I'm generally pretty happy with it. > > This is against the changeset prior to Shawn's cherrypy putback. I don't > anticipate much changing because of that, but I haven't tried the merge > yet. > > This also has a few lines of Stephen's code which puts links to all the > installed files in a single directory. I don't plan on taking his change > in with this one, but I wanted the extra speedup to get a fuller flavor of > how fast list would end up being. > > K, I think you probably have the most skin in this change, so I'd > appreciate it if you at least could give it a high-level pass to make sure > I'm not stomping on your bits too much. > > I may be on a bit during today, but in general will be in flight, so > there's no rush. But I figured people might like to start looking at this. > > Thanks, > Danek > _______________________________________________ > pkg-discuss mailing list > [email protected] > http://mail.opensolaris.org/mailman/listinfo/pkg-discuss _______________________________________________ pkg-discuss mailing list [email protected] http://mail.opensolaris.org/mailman/listinfo/pkg-discuss
