libsoup Roadmap
Maintainers:
What are your plans for GNOME 2.26 (next 4 months, before feature and UI freezes)?
Already done:
- New convenience methods:
- Methods for working with multipart/form-data bodies (HTML forms with file uploads), both client- and server-side.
- Methods for working with the Content-Type and Content-Disposition headers
- Support for partial GET requests, both client- and server-side
- Added a new libsoup-gnome library, with additional dependencies and features:
- Currently, the primary feature is automatic proxy support, via libproxy (or directly via gconf)
Added SoupCookieJarText, for storing cookies in mozilla-style cookies.txt files, and SoupCookieJarSqlite (in libsoup-gnome) for storing cookies in Firefox 3-style sqlite files.
Definitely in progress:
GResolver (async/cancellable dns resolver for glib), Bug 548466. If this doesn't make it into glib in time for GNOME 2.26, I'll be dropping it, egg-style, into libsoup itself. (Insert jokes about "egg-drop soup" here.)
Possibly still to happen:
libsoup bugs with milestone "GNOME 2.24". (Um, yeah well.). Specifically:
- Gnome-keyring integration (very likely; the code is basically done)
- Content-Encoding (maybe)
- Better SSL support (probably not this release)
- Caching (likely if the Epiphany developers push on it, unlikely otherwise)
What are your plans for GNOME 2.28?
Stuff that doesn't get done for 2.26, other stuff in LibSoup/ToDo, other feature requests that come up.
Do you have any goals from 2.24 that were not achieved? Why?
Yes, see above. As for "why", the list of goals was just a lot bigger than I had time to do myself, and I didn't get as much help from Epiphany/WebKit people as I'd been expecting when I did the 2.24 roadmap entry. (It turned out that most of the projects that were looking at that point like they'd end up using WebKit+soup ended up using WebKit+curl instead, or in the case of Epiphany, not using WebKit yet.)
Likewise, I don't actually expect to achieve all of the above 2.26 goals for 2.26.
Is there something that is really missing in our infrastructure or platform that would help you?
git. I do most of my libsoup hacking in git now, in a repo that I push to gnome.org so other people can pull from it. Even just having easier repo setup, gitweb, etc, without actually migrating svn.gnome.org to git, would make things a little nicer for me. (Eg, I've set up a second git repo in gnome.org/~danw/ now for my glib GResolver hacking.)
Do you have plans to work on other modules not maintained by you? What are they?
I'm only bothering to list libsoup-related stuff here.
the GResolver work in glib
I'm keeping an eye on libsoup's bindings in gir-repository; in particular, libsoup bindings are important for the javascript environments (gjs and Seed) and for Vala, since none of them have a native http library.
I'll probably send out patches to other libsoup-using modules to make use of SoupProxyResolverGNOME before 2.25.3, to make sure that new functionality gets used.
Do you have any GNOME-wide goals suggestions for the next releases?
Do you have any cool ideas for GNOME 3? What are they (both GNOME-wide and about your module)?
LibSoup/ToDo is pretty much the extent of my ideas for libsoup. I'm hacking on GnomeShell right now, although that's mostly other people's cool ideas for GNOME 3.