Shawn Walker wrote:
Brock Pytlik wrote:
Here's the preliminary/alpah/pre-alpha/"should never see the light of
day" webrev representing the current state the codes in.
The API code has been cleaned up a tiny bit, but basically the code
is in the state it's in when I develop it.
Things I know I need to fix: error handling, formatting, removing
commented code, making it work with the test suite, and probably lots
more I can't think of at the moment.
Things that might yet be changed: Basically everything. I think the
api is converging, but I spent today doing a fairly major rewhack so
I might rethink things again as I get more distance.
I'll be updating it in place as I make changes to either the client
or server API.
John, Michal, Padraig, please let me know if there are pieces of
information missing from the API as is.
If anyone else feels like diving in, comments are of course welcome
(though ones about formatting and commented code will probably be
ignored for a while).
http://cr.opensolaris.org/~bpytlik/prelim_search_v1/
One concern I have is the direct manifest loading by the search engine
instead of going through the image object to retrieve manifests when
an image object is available. Given Bart's pending change to dump
items not specific to the current client arch, it seems like you might
end up indexing a lot of content you don't want to index.
Se the follow up I've added to the bug and let me know what you think is
the right direction to go.
Instead of "include_pkg" in modules/fmri.py:get_fmri, I would name
that "include_scheme".
Will you be changing remote_search to be the default for the client?
Yep
Finally, it seems the ply library is in /pending now.
Does that make a difference for how we want to install it? I would've
thought that until it gets into the wos, we can't use a normal
dependency since we can't depend on a user having the pending authority.
Brock
Cheers,
_______________________________________________
pkg-discuss mailing list
[email protected]
http://mail.opensolaris.org/mailman/listinfo/pkg-discuss
