With the F11 alpha coming out in a few days, I think now is a good time to list a few things that we should look at for the desktop spin in F11:
- Empathy. When we switched back to pidgin as the default im client late in the F10 cycle, we agreed to reevaluate empathy around the time of the F11 alpha, and see if the concerns that prevented it from being the default in F10 have been addressed.
- Brasero. Upstream GNOME is moving towards replacing nautilus-cd-burner with brasero in 2.26 (following a move that some distros have already made). Some integration issues clearly still have to be worked out, but we should consider if we want to follow this in F11, or stay with nautilus-cd-burner for the time being. Even if we decided to do that, we'll have to look at possible conflicts if n-c-b and brasero are installed at the same time.
- CD size. Jeremy informed me that the recent flurry of font packaging changes seem to have freed up a considerable amount of space on the desktop spin, to the point where we might even consider shipping OOo (sans langpacks, of course) instead of abiword. We should look at the various options for filling the space we have.
Matthias
On Mon, 2009-01-26 at 11:32 -0500, Matthias Clasen wrote:
With the F11 alpha coming out in a few days, I think now is a good time to list a few things that we should look at for the desktop spin in F11:
There's also the case of the fingerprint integration. Right now, I don't think anything is pulling fprintd-pam into the media. Could you please make sure it gets added to the disc?
If this just requires some changes in comps, I'll be happy to do those myself.
FYI, for testers with a bit of free time: https://fedoraproject.org/wiki/Features/Fingerprint#Release_Notes Also be sure to read the known issues section: https://fedoraproject.org/wiki/Features/Fingerprint#Known_issues
Cheers
On Tue, 2009-01-27 at 01:14 +0000, Bastien Nocera wrote:
On Mon, 2009-01-26 at 11:32 -0500, Matthias Clasen wrote:
With the F11 alpha coming out in a few days, I think now is a good time to list a few things that we should look at for the desktop spin in F11:
There's also the case of the fingerprint integration. Right now, I don't think anything is pulling fprintd-pam into the media. Could you please make sure it gets added to the disc?
If this just requires some changes in comps, I'll be happy to do those myself.
making fprintd-pam installed by default in whatever comps group it is should bring it on the live cd. Alternatively, we can force it in via the kickstart file, but I think the comps way is preferable.
Matthias Clasen wrote:
- Empathy. When we switched back to pidgin as the default im client late
in the F10 cycle, we agreed to reevaluate empathy around the time of the F11 alpha, and see if the concerns that prevented it from being the default in F10 have been addressed.
Won't comment about this, myself I am forced to use Yahoo as the primary IM transport, so Empathy is useless for me, but I know I am a corner case.
- Brasero. Upstream GNOME is moving towards replacing nautilus-cd-burner
with brasero in 2.26 (following a move that some distros have already made). Some integration issues clearly still have to be worked out, but we should consider if we want to follow this in F11, or stay with nautilus-cd-burner for the time being. Even if we decided to do that, we'll have to look at possible conflicts if n-c-b and brasero are installed at the same time.
Brasero was recently upgraded in F10 from 0.82 to 0.84 and I find its new user interface much better, from an user point of view it would not be bad as a default. I personally prefer it as a default instead of n-c-b, but that is mostly because I need some of its advanced features.
- CD size. Jeremy informed me that the recent flurry of font packaging
changes seem to have freed up a considerable amount of space on the desktop spin, to the point where we might even consider shipping OOo (sans langpacks, of course) instead of abiword. We should look at the various options for filling the space we have.
If we have the space available, instead of one single application (well, OOo is not just a single application but...) we can include a few more smaller ones, like Inkscape and something else. Or some freely licensed multimedia content.
On Mon, 2009-01-26 at 11:32 -0500, Matthias Clasen wrote:
- Empathy. When we switched back to pidgin as the default im client late
in the F10 cycle, we agreed to reevaluate empathy around the time of the F11 alpha, and see if the concerns that prevented it from being the default in F10 have been addressed.
One of the major concerns was file-transfer support. That was recently merged into trunk and is in the current 2.25-series releases. My understanding is that this is still Jabber/XMPP-only though (Gabble backend); but one thing that comes to mind is that Haze (and protocols through it, such as MSN) currently still does not support this well.
The account-management window has been cleaned up quite a bit too. Especially, many more of the login details for creating a new account are guessed based on the account type and automagically filled in.
Setting one's personal avatar image as well as calling with VoIP are still a bit hit-or-miss, unfortunately. All told, it has definitely made a significant amount of progress (thanks, Xavier and team!) but it may still be a bit too rough around the edges, so to speak, to be the default in F11.
I propose that we *do* have it as default for the Alpha and Beta and if it comes to a point, similar to the F10 trials, where it's still too rough, that we can simply fallback (again) to Pidgin for another release. I sincerely hope that this is not the case, though. ;-)
- Brasero. Upstream GNOME is moving towards replacing nautilus-cd-burner
with brasero in 2.26 (following a move that some distros have already made). Some integration issues clearly still have to be worked out, but we should consider if we want to follow this in F11, or stay with nautilus-cd-burner for the time being. Even if we decided to do that, we'll have to look at possible conflicts if n-c-b and brasero are installed at the same time.
I've not followed the upstream discussion on this as closely as I'd have liked. However, from what I understand, most of the work has gone into replacing the backend n-c-b library with libbrasero-media and friends. It stills seems rather incomplete though. According to the patches in various upstream GNOME bug reports (such as the one for Rhythmbox [1]), it merely creates the appropriate playlist file and passes that the Brasero application.
For the record, I **HATE** [2] having a seperate application for burning CDs/DVDs (a la Windows and some OS X programs). This type of functionality just does away with the whole "do one thing really well" Unix model to me. The applicable capabilities should be inherent to the application being used for that media. For example, with Rhythmbox currently, burning a CD is as easy as dragging the songs you want to their own playlist and then selecting "Create Audio CD..." from that playlist's context menu (or "Burn" in the main toolbar).
I would love it if it is intended to be used as a backend library, such that applications could call to it for CD-burning stuff (a la what I think libburn tried to accomplish); but it should not be a do-all burner application in its own right IMNSHO.
[1] http://bugzilla.gnome.org/show_bug.cgi?id=536732
- CD size. Jeremy informed me that the recent flurry of font packaging
changes seem to have freed up a considerable amount of space on the desktop spin, to the point where we might even consider shipping OOo (sans langpacks, of course) instead of abiword. We should look at the various options for filling the space we have.
As beneficial as OO.org may be on the disc, perhaps we could instead add a few games or "killer apps" which showcase the potential of F/OSS?
Though, offhand, I have no idea what those would be...
Thanks, and regards.
On Mon, 2009-01-26 at 11:32 -0500, Matthias Clasen wrote:
- Empathy. When we switched back to pidgin as the default im client late
in the F10 cycle, we agreed to reevaluate empathy around the time of the F11 alpha, and see if the concerns that prevented it from being the default in F10 have been addressed.
Another huge issue with the transition was that there was no easy way to migrate accounts from Pidgin, In recent releases, there is now an "Import Accounts" button to the accounts dialog which automagically imports login information from Pidgin. (I have tested this successfully on Fedora 10 with rawhide Empathy/Telepathy packages, and AIM, Yahoo! IM, Jabber, and GoogleTalk all were imported nicely).
That should definitely make things a lot simpler for new users who would be hesitant because of the necessity to re-create a lot of account info.
On Tue, 2009-01-27 at 09:45 +0200, Nicu Buculei wrote:
Won't comment about this, myself I am forced to use Yahoo as the primary IM transport, so Empathy is useless for me, but I know I am a corner case.
Yahoo! IM _should_ work with the Haze telepathy backend (conversations and such, at the very least). If it doesn't, Bugzilla is the place to be! :)
Peter Gordon wrote:
On Tue, 2009-01-27 at 09:45 +0200, Nicu Buculei wrote:
Won't comment about this, myself I am forced to use Yahoo as the primary IM transport, so Empathy is useless for me, but I know I am a corner case.
Yahoo! IM _should_ work with the Haze telepathy backend (conversations and such, at the very least). If it doesn't, Bugzilla is the place to be! :)
I know that, but Haze uses the Pidgin core, libpurple, so then why not use Pidgin instead? It has a better coverage of features anyway.
Peter Gordon wrote:
On Mon, 2009-01-26 at 11:32 -0500, Matthias Clasen wrote:
- Empathy. When we switched back to pidgin as the default im client late
in the F10 cycle, we agreed to reevaluate empathy around the time of the F11 alpha, and see if the concerns that prevented it from being the default in F10 have been addressed.
Another huge issue with the transition was that there was no easy way to migrate accounts from Pidgin, In recent releases, there is now an "Import Accounts" button to the accounts dialog which automagically imports login information from Pidgin. (I have tested this successfully on Fedora 10 with rawhide Empathy/Telepathy packages, and AIM, Yahoo! IM, Jabber, and GoogleTalk all were imported nicely).
That should definitely make things a lot simpler for new users who would be hesitant because of the necessity to re-create a lot of account info.
Account information is usually small and can be recreated. What about chat logs, preferences etc?
Rahul
On Tue, 2009-01-27 at 14:15 +0530, Rahul Sundaram wrote:
Account information is usually small and can be recreated. What about chat logs, preferences etc?
Those are not yet being imported, from what I can tell through the source code (src/empathy-import-pidgin.c and friends). It's only login details at the moment.
However, the fact that they *can* be quickly recreated should not make that a necessity when they're already stored, just in a different application. (I, for one, used only Gaim/Pidgin for so long with my AIM account until I found Empathy that I had completely forgot what my password was, and had to reset it through the web/email interface.)
On Mon, 2009-01-26 at 20:29 -0500, Matthias Clasen wrote:
On Tue, 2009-01-27 at 01:14 +0000, Bastien Nocera wrote:
On Mon, 2009-01-26 at 11:32 -0500, Matthias Clasen wrote:
With the F11 alpha coming out in a few days, I think now is a good time to list a few things that we should look at for the desktop spin in F11:
There's also the case of the fingerprint integration. Right now, I don't think anything is pulling fprintd-pam into the media. Could you please make sure it gets added to the disc?
If this just requires some changes in comps, I'll be happy to do those myself.
making fprintd-pam installed by default in whatever comps group it is should bring it on the live cd. Alternatively, we can force it in via the kickstart file, but I think the comps way is preferable.
Done in comps.
On Mon, 2009-01-26 at 11:32 -0500, Matthias Clasen wrote:
- Brasero. Upstream GNOME is moving towards replacing nautilus-cd-burner
with brasero in 2.26 (following a move that some distros have already made). Some integration issues clearly still have to be worked out, but we should consider if we want to follow this in F11, or stay with nautilus-cd-burner for the time being. Even if we decided to do that, we'll have to look at possible conflicts if n-c-b and brasero are installed at the same time.
Last time I tried brasero (it was F9 I believe) it kept trashing my DVDs (CDs were fine), n-c-b does not suffer that issue. So I hope I won't loose my last resort for safely burning data DVD by this switch.
And yep, as someone noted in this thread as well, I'd like it to be just a burning backend - I'd like to keep buring data DVDs using nautilus, audio CDs using rhythmbox,... It's much more convenient for me than one know-it-all burning app...
Martin
Martin Sourada wrote:
On Mon, 2009-01-26 at 11:32 -0500, Matthias Clasen wrote:
- Brasero. Upstream GNOME is moving towards replacing nautilus-cd-burner
with brasero in 2.26 (following a move that some distros have already made). Some integration issues clearly still have to be worked out, but we should consider if we want to follow this in F11, or stay with nautilus-cd-burner for the time being. Even if we decided to do that, we'll have to look at possible conflicts if n-c-b and brasero are installed at the same time.
Last time I tried brasero (it was F9 I believe) it kept trashing my DVDs (CDs were fine), n-c-b does not suffer that issue. So I hope I won't loose my last resort for safely burning data DVD by this switch.
And yep, as someone noted in this thread as well, I'd like it to be just a burning backend - I'd like to keep buring data DVDs using nautilus, audio CDs using rhythmbox,... It's much more convenient for me than one know-it-all burning app...
It appears with the current changes, it can be both. You should look at the changes they have made recently and try it out.
Rahul
On Tuesday, January 27 2009, Nicu Buculei said:
Matthias Clasen wrote:
- CD size. Jeremy informed me that the recent flurry of font packaging
changes seem to have freed up a considerable amount of space on the desktop spin, to the point where we might even consider shipping OOo (sans langpacks, of course) instead of abiword. We should look at the various options for filling the space we have.
If we have the space available, instead of one single application (well, OOo is not just a single application but...) we can include a few more smaller ones, like Inkscape and something else. Or some freely licensed multimedia content.
Inkscape, though not that large on its own, has a remarkably large dep chain from what I remember
Jeremy
Jeremy Katz wrote:
On Tuesday, January 27 2009, Nicu Buculei said:
Matthias Clasen wrote:
- CD size. Jeremy informed me that the recent flurry of font packaging
changes seem to have freed up a considerable amount of space on the desktop spin, to the point where we might even consider shipping OOo (sans langpacks, of course) instead of abiword. We should look at the various options for filling the space we have.
If we have the space available, instead of one single application (well, OOo is not just a single application but...) we can include a few more smaller ones, like Inkscape and something else. Or some freely licensed multimedia content.
Inkscape, though not that large on its own, has a remarkably large dep chain from what I remember
Large, but I believe not as large as OpenOffice.org.
However, if the size is a concern, I would like to see a little bit of multimedia, to showcase Fedora's capabilities.
On Mon, 2009-01-26 at 23:55 -0800, Peter Gordon wrote:
I propose that we *do* have it as default for the Alpha and Beta and if it comes to a point, similar to the F10 trials, where it's still too rough, that we can simply fallback (again) to Pidgin for another release. I sincerely hope that this is not the case, though. ;-)
If you're going to do that again, you're really going to need to nail down a set of criteria for "pass" vs "fail" so that QA et al can better help make that judgment call. It was entirely too wishy washy last time around.
On Tue, 2009-01-27 at 00:06 -0800, Peter Gordon wrote:
Another huge issue with the transition was that there was no easy way to migrate accounts from Pidgin, In recent releases, there is now an "Import Accounts" button to the accounts dialog which automagically imports login information from Pidgin. (I have tested this successfully on Fedora 10 with rawhide Empathy/Telepathy packages, and AIM, Yahoo! IM, Jabber, and GoogleTalk all were imported nicely).
That should definitely make things a lot simpler for new users who would be hesitant because of the necessity to re-create a lot of account info.
Users with existing account information and settings and history via say pidgin, will likely just continue using pidgin, as they'd be doing upgrades. I don't see this as a big issue, and instead concentrate on the new user experience.
On Tue, 2009-01-27 at 09:18 -0800, Jesse Keating wrote:
On Mon, 2009-01-26 at 23:55 -0800, Peter Gordon wrote:
I propose that we *do* have it as default for the Alpha and Beta and if it comes to a point, similar to the F10 trials, where it's still too rough, that we can simply fallback (again) to Pidgin for another release. I sincerely hope that this is not the case, though. ;-)
If you're going to do that again, you're really going to need to nail down a set of criteria for "pass" vs "fail" so that QA et al can better help make that judgment call. It was entirely too wishy washy last time around.
For starters, it would be nice if the people who rejected it the last time around would log their reasons for doing so on the feature page...
On Tue, 2009-01-27 at 15:01 -0500, Matthias Clasen wrote:
For starters, it would be nice if the people who rejected it the last time around would log their reasons for doing so on the feature page...
It was the feature driver himself who decided to not have it by default last time around.
On Tue, 2009-01-27 at 12:06 -0800, Jesse Keating wrote:
On Tue, 2009-01-27 at 15:01 -0500, Matthias Clasen wrote:
For starters, it would be nice if the people who rejected it the last time around would log their reasons for doing so on the feature page...
It was the feature driver himself who decided to not have it by default last time around.
I don't think the logs at http://bpepple.fedorapeople.org/fesco/FESCo-2008-11-05.html support that narrative.
Anyway, in that meeting, Colin said
i haven't checked in on it in the last month or two; i guess my thoughts are that if we decide not to go with it, we should have a rationale writeup we can give to the empathy developers
but I don't think any rationale was written.
On Tue, 2009-01-27 at 16:19 -0500, Matthias Clasen wrote:
I don't think the logs at http://bpepple.fedorapeople.org/fesco/FESCo-2008-11-05.html support that narrative.
Anyway, in that meeting, Colin said
i haven't checked in on it in the last month or two; i guess my thoughts are that if we decide not to go with it, we should have a rationale writeup we can give to the empathy developers
but I don't think any rationale was written.
Hrm, from one of the discussions i could have sworn bpepple had claimed he was the feature owner for empathy in F10. I'll have to find where I thought this claim was made.
On Tue, 2009-01-27 at 19:10 +0530, Rahul Sundaram wrote:
Martin Sourada wrote:
On Mon, 2009-01-26 at 11:32 -0500, Matthias Clasen wrote:
- Brasero. Upstream GNOME is moving towards replacing nautilus-cd-burner
with brasero in 2.26 (following a move that some distros have already made). Some integration issues clearly still have to be worked out, but we should consider if we want to follow this in F11, or stay with nautilus-cd-burner for the time being. Even if we decided to do that, we'll have to look at possible conflicts if n-c-b and brasero are installed at the same time.
Last time I tried brasero (it was F9 I believe) it kept trashing my DVDs (CDs were fine), n-c-b does not suffer that issue. So I hope I won't loose my last resort for safely burning data DVD by this switch.
And yep, as someone noted in this thread as well, I'd like it to be just a burning backend - I'd like to keep buring data DVDs using nautilus, audio CDs using rhythmbox,... It's much more convenient for me than one know-it-all burning app...
It appears with the current changes, it can be both. You should look at the changes they have made recently and try it out.
It only does data burning through nautilus. It will brings up its ugly interface if you wanted to burn through Rhythmbox for example (which is why the patch isn't in Fedora or upstream yet).
On Tue, 2009-01-27 at 14:27 -0800, Jesse Keating wrote:
On Tue, 2009-01-27 at 16:19 -0500, Matthias Clasen wrote:
I don't think the logs at http://bpepple.fedorapeople.org/fesco/FESCo-2008-11-05.html support that narrative.
Anyway, in that meeting, Colin said
i haven't checked in on it in the last month or two; i guess my thoughts are that if we decide not to go with it, we should have a rationale writeup we can give to the empathy developers
but I don't think any rationale was written.
Hrm, from one of the discussions i could have sworn bpepple had claimed he was the feature owner for empathy in F10. I'll have to find where I thought this claim was made.
No, I just maintain the majority of the telepathy stack and co-maintain empathy in Fedora.
Later, /B
On Mon, 2009-01-26 at 23:55 -0800, Peter Gordon wrote:
One of the major concerns was file-transfer support. That was recently merged into trunk and is in the current 2.25-series releases. My understanding is that this is still Jabber/XMPP-only though (Gabble backend); but one thing that comes to mind is that Haze (and protocols through it, such as MSN) currently still does not support this well.
Not entirely correct. Currently, file transfer support is only provided for Local-link XMPP (via telepathy-salut). Telepathy-gabble just recently added support for SOCKS5 bytestreams in Tubes (which isn't file transferring). The plan is to use Jingle for file-transfers (and also Tubes), but I'm not sure that is going to be finished before F11.
Later, /B
On Tue, 2009-01-27 at 15:01 -0500, Matthias Clasen wrote:
On Tue, 2009-01-27 at 09:18 -0800, Jesse Keating wrote:
On Mon, 2009-01-26 at 23:55 -0800, Peter Gordon wrote:
I propose that we *do* have it as default for the Alpha and Beta and if it comes to a point, similar to the F10 trials, where it's still too rough, that we can simply fallback (again) to Pidgin for another release. I sincerely hope that this is not the case, though. ;-)
If you're going to do that again, you're really going to need to nail down a set of criteria for "pass" vs "fail" so that QA et al can better help make that judgment call. It was entirely too wishy washy last time around.
For starters, it would be nice if the people who rejected it the last time around would log their reasons for doing so on the feature page...
I've added some of my concerns to the discussion page, but I'll try to fleshed it out a bit more this weekend.
Later, /B
desktop@lists.fedoraproject.org