[00:00] stdin, what logs? where? [00:00] woody86: click the Open Week link in the topic, then click on the class name to get the log [00:00] stdin, oh, thanks :D === steve___1 is now known as mysterio === PrivateVoid_ is now known as PrivateVoid === PrivateVoid_ is now known as PrivateVoid === F4ilure is now known as Gonzofox === Gonzofox is now known as Defaultnick^ [05:50] dmesg [05:51] ... [05:52] - -! [05:52] nat3: hello, thailand [06:05] relax guy look over there [07:24] 0000000 === monsterb is now known as Brotherred [10:35] <|sundown|> Äîáðûé äåíü! Ìîæåò ëè êòî-íèáóäü ìíå ïîìî÷ü â âîïðîñå êîíôèãóðèðîâíèÿ âèðòóàëüíûõ ñåòåâûõ èíòåðôåéñîâ? [10:36] <|sundown|> Good day! Can anyone help me in the matter konfigurirovniya virtual network interfaces? === Rafik_ is now known as Rafik === jcastro changed the topic of #ubuntu-classroom to: Current Session: Polishing a Package | Welcome to Openweek, questions in #ubuntu-classroom-chat please || Session details here: https://wiki.ubuntu.com/UbuntuOpenWeek [14:58] Welcome to day 3 of Ubuntu Open Week! [14:58] Schedule is here: [14:58] https://wiki.ubuntu.com/UbuntuOpenWeek [14:58] we'll begin in a few minutes [14:59] \o/ [15:00] OK. Welcome to the OpenWeek session Polishing a Package. [15:01] I'm Emmet Hikory, a MOTU interested in package quality, amoung other things. I tend to prefer an interactive discussion, so please feel free to interrupt with questions if you have any. [15:02] ignore #ubuntu-classroom CRAP NOTICES SNOTES CTCPS JOINS PARTS QUITS KICKS MODES WALLOPS NICKS DCC DCCMSGS CLIENTNOTICES CLIENTCRAP CLIENTERRORS HILIGHTS [15:02] sorry for bad "ignore" [15:02] So, the goal of this session is to demonstrate some of the resources available to help improve the quality of a package. [15:03] Most packages in Ubuntu come from Debian, and those tend to have a maintainer, who generally does a pretty good job of keeping the package in shape. [15:04] While these techniques may also be applied to such packages, it's generally better to use them for packages that don't have a maintainer, at least until all those are in perfect shape. [15:05] The Ubuntu External Health Status pages try to keep track of packages that don't have Debian maintainers, and the status in relation to upstream. You can take a look at http://qa.ubuntuwire.com/uehs/ [15:05] Of most interest are the packages without a watch file or packages not updated. These are good candidates for some investigation and polish. [15:06] For today's example, I selected the ikvm package, which is shown as out of date on http://qa.ubuntuwire.com/uehs/no_updated.html [15:06] We have version 0.34.0.4 and upstream has 0.36.0.11, which probably contains some useful improvements that users might want. [15:08] So, looking at debian/copyright for the ikvm package, we can see that the upstream homepage is at ikvm.net. Some packages also have this information in the debian/control file, some in a README, etc. Depending on the state of the package, you may need to hunt a bit, although the contents of debian/watch may be helpful. [15:09] So, the first step is to grab the new upstream, and prepare the update. Doing that alone would make the package drop from the list of non-updated packages, but it's not always enough. [15:10] Next would be to take a look at the bugs outstanding against ikvm : you can see those from https://bugs.launchpad.net/ubuntu/+source/ikvm/+bugs [15:11] As you can see, there's a few bugs that indicate some adjustments are required. At least making sure the pathnames are correct and that it can be used as a jvm seem reasonable requests, so it's worth trying to fix those as part of the update. [15:11] persia: can you join #ubuntu-classroom-chat ? [15:12] Did I miss a question there? [15:13] < ian_brasil> QUESTION: what is the best way to package for hildon?..pass --enable-hildon into configure.ac or some other way? [15:13] Well, that's a little outside of the scope of polishing packages, but it very much depends on the upstream sources. [15:15] So, once we've cleaned up those issues, it's worth checking to see if the package is in Debian. I usually do that by visiting http://packages.qa.debian.org [15:16] Entering the package name there will bring you to http://packages.qa.debian.org/i/ikvm.html which provides a nice summary of things needing doing. [15:17] We already knew about the new upstream, but it's worth checking the lintian warning. Apparently the manpage needs to have the charset defined, so that's another thing we can do before uploading. [15:18] There's also a link to the Debian bugtracker available from this page, so we can see the bugs in Debian. [15:19] One of these (Provide java-virtual-machine) looks a lot like the "ikvm is not installed as a viable java alternative" in Malone, so that's already done. [15:19] The other three are worth looking at in a bit more detail, and worth fixing. [15:21] Next, we'll take a look upstream. The "Report Bugs" link on the upstream homepage goes to the sourceforge bug tracker, and shows 1 open bug, which seems different than any of the other bugs we've seen. [15:22] This one is a bit more complicated, as it's hard to reproduce, so we'll let it go, but for a different package, you might find easy bugs, or bugs with patches available that you might want. [15:24] It's also worth checking to see if the package is in other distributions, and if they have any good patches that we might want. While this can be done manually, I tend to be lazy, and just check Harvest to use the automated patch collectors people have writted. [15:24] Harvest is available from http://daniel.holba.ch/harvest/ Checking the Sourcepackage list, we see that ikvm doesn't have any outstanding opportunities, so there's nothing to get from there. [15:25] Note that right now, Harvest is only pulling patches from Fedora : I'm sure the Harvest team would appreciate it if anyone wants to write a feed for other distributions. [15:26] So, at this point, we've investigated everything, and hopefully put all the fixes into the package. [15:27] Next is cleanup : if we found a bug not reported upstream, and not fixed with the new upstream, and needed to create a patch, it's worth reporting that to the sourceforge tracker with the patch (and linking to the bug in Launchpad). [15:28] Similarly, if patches were prepared for the Debian bugs, those patches should be attached to the relevant Debian bugs, and we should add bug watches for the BTS bugs where we found them to match bugs already in Launchpad. [15:29] At this point, we're almost ready : it's worth running lintian a last time locally, just in case there's anything that happened as a result of the patching, and the package can be uploaded. [15:29] QUESTION: Because of the different freezes in current ubuntu release for package to go into repos, there's no real point in updating packages for intrepid for example? (except for release critical bugs right?) [15:30] For intrepid, no. For Jaunty, there's great value in updating and polishing the packages. Most packages have open bugs that don't get fixed because they aren't so important to the people usually working on the package (or maybe nobody works on the package). [15:31] While none of these would qualify for a stable release update (targeted at intrepid today), fixing them in the development release (jaunty today) means that when the next release happens, those users affected by the bugs are likely to be very pleased. [15:32] QUESTION: how can one get rights to upload a package? or should a updated package be sent to the maintainer? [15:34] Well, none of the packages listed at UEHS have an explicit maintainer : generally it's either one of the Ubuntu packaging teams or the Debian QA team. In cases where there is a Maintainer, it's best to communicate the patches in the Debian BTS, but if there's enough oustanding, or a good reason to update in Ubuntu (and you plan to track the package), one doesn't have to wait for the maintainer to upload. [15:35] If one doesn't yet have permission to upload, it's best to attach either a debdiff (no new upstream version), or the resulting diff.gz from the updated package to a bug, and subscribe the relevant sponsors team (ubuntu-main-sponsors or ubuntu-universe-sponsors) to request review and upload. [15:37] Any other questions? Is the process by which one finds various reported bugs or minor issues clear? [15:40] QUESTION: would harvest benefit from integration into maemo or are the patches there automatically pulled into ubuntu-mid [15:42] Well, I'm not sure what "integration into maemo" would mean for harvest, but any additional feeds to pull patches from known reliable patch authors for harvest would probably be good, as this helps to provide a complete picture of available fixes. [15:43] The package I picked didn't happen to be listed there, but if you look at other packages in the sourcepackages list, you can see that some have quite a few available patches, perhaps from Fedora, or perhaps in the bugtracker. [15:44] The more sources from which patches are available and shown, the easier it is for those cleaning up packages to make sure they are perfect. [15:46] QUESTION: there's a package with a new version in debian, but also a new (at the top) upstream version. What should we do? packaging it for debian, then packaging it for ubuntu? === graaf_ is now known as Nameless [15:48] Well, if you pick a package without a Debian Maintainer, and you prepare a new upstream version, it's worth checking with the Debian QA team to see if they would appreciate the update. Generally they advise that you might want to become the Debian maintainer, but rarely turn down help. [15:48] If such an update is accepted into Debian, and the Ubutu repositories are not yet frozen, that can be synced directly into Ubuntu, so there's no value in packaging it separately for Ubuntu. [15:50] If there's some reason that it can't be updated, or significant delay, and a freeze is approaching, or you really want to, it's worth sending it to Ubuntu directly, although usually this is just a change to the first line of the changelog, rather than a significant packaging difference. === sourcode_ is now known as sourcode [15:51] Also note that it's not always worth grabbing the very newest upstream : it may be that upstream has a stable release and a development release, and doesn't recommend the development release for most users. It may also be that a new upstream is incompatible with something else in some way, which may have to be resolved before it can be updated. [15:53] For packages with Maintainers, it's especially desirable not to update in advance of Debian, unless you're personally willing to be responsible for the package until Debian catches up: most Maintainers will update soon enough, and it's best not to duplicate work (as there are plenty of packages without maintainers). [15:54] meaning we should look for ophans packages ? [15:55] Well, I recommend starting with the packages in UEHS, or the packages in Harvest. Either is a good source of packages needing work. I usually recommend UEHS more at the beginning of a development cycle, and Harvest later, once things start to settle. [15:56] Every package in UEHS is either orphaned in Debian, or doesn't exist in Debian, and so has no specific individual maintainer. [15:56] Many packages in Harvest have a Debian maintainer, but in Ubuntu there are no specific maintainers generally, so most packages are fair game for basic updates. [15:57] There's no need to look specifically for an orphaned package. Of course, if there's nothing in UEHS, and nothing in Harvest, then maybe it's worth it, but there's a few thousand packages that need a bit of polich first :) [15:58] I think we have time for one more question before the end, if anyone has one. [15:58] weboide> persia: ok, so the most important would be UEHS, Harvest, and then orphans (debian or not?) packages [15:59] Well, once UEHS and Harvest are clear, I'd suggest just looking for packages with a few bugs in Launchpad, and using the same techniques to bring them up to date. Doesn't really matter if they are orphaned. [15:59] And with that, I'll turn you over to the Ubuntu Netbook Remix crew. Thanks for attending. [16:00] ty persia === jcastro changed the topic of #ubuntu-classroom to: Current Session: Ubuntu Netbook Remix: https://wiki.ubuntu.com/MeetingLogs/openweekintrepid/NetbookRemix | Welcome to Openweek, questions in #ubuntu-classroom-chat please || Session details here: https://wiki.ubuntu.com/UbuntuOpenWeek [16:00] Ok guys, take it away! [16:01] Hi all [16:01] There is a presentation that goes along with this talkk [16:01] you can find it linked in the topic [16:02] \o/ [16:02] I am Pete Goodall, the product manager for the OEM Services group at Canonical. And joining me are Neil Patel and Bill Filler, also with OEM Services [16:02] weee [16:03] bfiller: hello everyone :) [16:03] hi [16:03] I'll give you a quick agenda: [16:03] * Canonical and OEM Services Group [16:03] * What is Ubuntu Netbook Remix (UNR)? [16:03] * Ubuntu Netbook Remix pre-installed [16:04] * Installing UNR [16:04] * Language Support [16:04] * Contributing to UNR [16:04] If you have the presentation, let's go to slide 3 "Canonical" [16:05] For those of you that don't know Canonical is the commercial sponsor of the Ubuntu project. We employ many of the developers that work on Ubuntu. [16:05] [16:06] Neil, Bill and I work for the OEM Services group at Canonical, and we are the group that is responsible for customising Ubuntu for device manufacturers [16:07] By "customising" we mean making sure that all the hardware components work, integrating custom interfaces (ala Dell) and providing on-going maintenance. [16:07] [16:08] [16:08] Before I get into "What is Ubuntu Netbook Remix" I wanted to define what we mean about what a netbook is. [16:09] This is a term that is mostly coined by Intel, but refers to all these devices coming out that look like little laptops. [16:10] They usually have a resolution of 1024x600 (particularly problematic), and a low amount of local storage. They also tend to be low cost devices and have an Intel Atom processor. [16:10] However, the Intel Atom processor is not a hard requirement ;-) [16:11] [16:11] Ubuntu Netbook Remix came about because there was a massive amount of interest in running Ubuntu on these netbooks [16:12] We wanted to create a simple interface that was both easy to use, but put everything you needed within your view. [16:12] UNR is really bringing Linux to a new class of user, that typically doesn't know what Linux is [16:13] they just want to get their e-mail or blog, or chat with friends [16:14] We based UNR on Ubuntu 8.04 because it is an LTS release (long term support), and we tried to streamline things a bit by reducing the amount of applications preinstalled. However, that doesn't mean you cannot install whatever you want. [16:14] [16:15] Hopefully you all have the slides, but if you don't I just want to explain about the various elements of the UNR home screen, and what they do. [16:16] There are several components created specifically for UNR that create the look, feel and functionality. [16:17] The main thing the user sees when they start UNR is the Home Screen. On the left are application categories. In the middle are application launchers, and on the right are (generally) places and a way turn the device off. [16:17] It's nitpicky but battery and network labels are backwards, sorry for the interruption [16:17] nb72, np. I'll have to look at that. ;-) [16:18] hi [16:19] Anway, most of this will be familiar to you if you already use Ubuntu, but one thing that is slightly different is the "Favourites" cateogory. This allows the user to group the applications they use most, and this is the default category shown when you first login. [16:20] In the upper left of the screen, you see the Ubuntu logo and that is the "Go Home Applet". Clicking on that always brings you back to the home screen. So while you are using Firefox, you click on Go Home and start something else, like Pidgin. [16:21] Next to that is the Window Picker Applet. The icons to the right of the Ubuntu logo are the currently running applications and the title to the right is the application in focus. In this case "Home", but otherwise it would be the title bar for the currently running application. [16:22] The element you do not see is "Maximus". This is a daemon that runs in the background and make sure most applications always open maximised. [16:23] The reason we do this is to save screen real estate. Maximus also removes the title bar from the application window and puts it in the panel. [16:23] In addition, there is a gconf key to add application exceptions. We hope to make it easier in the future to do that. :-) [16:23] * repete_ looks at Neil ;-) [16:24] [16:25] Since UNR is based on Ubuntu 8.04 it still includes the applications current Ubuntu users (and Linux users in general) will be familiar with. These applications are all open source, and we always use open source applications when possible. [16:26] [16:27] All the components we created for UNR are open source software. [16:27] available at http://launchpad.net/unr [16:27] I have already mentioned most of the, except for two: The Human Netbook Theme and Desktop Switcher. === gouki_ is now known as gouki [16:28] The Human Netbook Theme is a theme, based on the standard Ubuntu Human theme, that goes better with UNR's components and layout. [16:29] The Desktop Switcher is a control centre applet that allows the user to switch back and forth between the UNR and the standard Ubuntu interface (termed as the "Classic" interface) [16:29] Thanks for the link bfiller [16:29] [16:30] As a value add for the device manufacturers, we include various third party components to make a more complete product. [16:31] These are commercial components (read: somebody has to be paid for their use), so we cannot include them in the publicly available offering. [16:32] But they are things that generally make things easier for those users that are not familiar with Linux and don't know how to install stuff like Adobe Flash, or legally licensed audio and video decoders. [16:33] [16:33] So as I mentioned before we work with device manufacturers to get Ubuntu and Ubuntu Netbook Remix on their device. This has a number of advantages. === artir_ is now known as artir [16:34] First, this means that all the hardware components work out of the box. That means you don't have to download and compile the latest madwifi or search user forums for which Windows driver to use with ndiswrapper. [16:35] Second, suspend and resume works...more than once :-) [16:35] In fact we test these devices to a much higher standard because those are the standards of the consumer electronics industry. [16:36] for example, suspend/resume tests are carried out between 300-400 times in a row before they are deemed a "pass" [16:36] Third, you can actually get support from the manufacturer. No more, "I'm sorry, but we don't support Linux... and I don't actually know what that is..." [16:36] Thx njpatel [16:37] Fourth, is what I talked about before. We include stuff like audio and video decoders that are legally licensed. [16:37] it also means that all the buttons on your keyboard work from the get-go :-D ! [16:38] And of course we test that all this stuff works. [16:38] If you haven't already got a netbook and are looking to get one, the Sylvania g Netbook Meso and the Toshiba NB100 are both available with UNR pre-installed. [16:39] [16:39] and the Dell Mini 9 ships with a UNR variant [16:39] So you may be saying to yourself, "I already bought a netbook and I just want to know how to run Ubuntu on it." [16:40] Just as a note, that screenshot on that slide is thanks to Gimp, not the watermark in the lower left. :-) [16:40] [16:41] Many community members have been asking for an installation image for UNR. We have not had one to date because UNR isn't really a new edition. [16:41] However, it is obvious this is something that people want because they already have a netbook and want Ubuntu one it. [16:42] Soon (this week I hope) we will downloadable images for you to install UNR and you won't even need a (non existent in many cases) CD-ROM. [16:43] Otherwise, your best bet for now (if you are impatient) is to use the instructions in the next slide. [16:43] [16:44] As I mentioned, products with UNR will be based on Ubuntu 8.04, but we have made the packages for UNR available in Ubuntu 8.10. [16:45] One thing to note is that the "ume-launcher" package has been re-named in 8.10 to "netbook-launcher" [16:45] the central place for installation instructions are at http://wiki.ubuntu.com/unr (instructions for image, hardy, intrepid, umpc-image) [16:46] The packages in 8.10 do not have the latest and greatest features, because we had to respect the feature freeze for 8.10. However, if you want to use the latest and greatest in 8.10 you can use the PPA. [16:46] (linked later) [16:46] also important to note, the unr pacakges can be installed on any computer running Ubuntu 8.04 or 8.10 (i.e. not just netbook, but laptops and desktops too) [16:47] another good point bfiller thx. [16:47] [16:47] Since UNR was first announced, we have built a vibrant community and have already taken both patches and suggestions from the community. [16:48] If you are interested in getting involved in UNR you should go to the UNR page on Launchpad, http://launchpad.net/unr [16:49] That page is a "super group" that includes all the packages for UNR. You can checkout the code using bzr. [16:49] I don't know if it has come up in the questions already, but the UNR Home Screen is written in Clutter. [16:50] If you are not already familiar with clutter... Use Google. :-) [16:50] repete_: time check, 10 mins remaining, should we get to the Q+A from questions on #ubuntu-classroom-chat [16:50] Sure. [16:50] QUESTION: could UNR be used to revive old computers , and will there be iso images for virtualbox so people can have an look at it this way? [16:50] knome QUESTION: So Asus EeePC is a netbook? [16:51] oops sorry repete_ , was starting from the top [16:51] np [16:51] go ahead. [16:51] ok [16:51] knome: yes EeePC is a netbook [16:52] QUESTION: what is the difference between ubuntu-mid and unr [16:52] QUESTION: what is the difference between ubuntu-mid and unr [16:52] * njpatel let's bfiller post the questions from now on [16:52] * bfiller smiles [16:52] ubuntu-mid is targetted at devices which have touchscreens and normally small displays with high dpis [16:53] unr (although capable of use with touch-screens) is targetted more at slightly larger screens and point-and-click ui [16:53] Are there any plans for Ubuntu on the Nokia n810? [16:53] billybigrigger: QUESTION: what is running on the homescreen? looks like a gnome panel with kde4 menu??? [16:54] The home screen consists of gnome-panel and a clutter-based launcher [16:54] the launcher replaces the traditional desktop [16:54] no kdelibs in sight ;-) [16:55] billybigrigger: QUESTION: what is the default install size for UNR? [16:55] billybigrigger: it's about 2.5 GB installed [16:55] meant to fix on 4 GB SSD and greater [16:55] s/fix/fit === artir__ is now known as artir [16:56] QUESTION: will UNR will have USB Install Option [16:56] has anyone heard of retail point of sale software? [16:57] I am at work so I have to make this legit [16:57] xander21c: the UNR image will be installable via bootable USB image, so yes [16:57] QUESTION: There's a rumour around that Asus doesn't like the xandros distribution and is currently testing ubuntu for the eee pc. Do you know anything about that? [16:57] repete_: care to comment on this question? [16:58] Even if I did I couldn't comment :-) [16:58] I know many people imediatly put ubuntu on there eee when they get them [16:58] I don't know anything about it either [16:58] I will say that *many* people have install UNR on a eeePC [16:58] next question .. [16:58] QUESTION: could UNR be used to revive old computers , and will there be iso images for virtualbox so people can have an look at it this way? [16:58] as the launcher is clutter-based this poses a bit of a problem [16:59] You have to be able to boot via usb for it to work [16:59] clutter uses opengl for displaying, so if the old computer can run opengl decently, everything should work fine [16:59] I tried back in the day with pentium 2s and small dog [16:59] currently we don't have plans to provide iso image, only .img [16:59] whta is the diference between ubuntu-eepc and UNR [17:00] for the same reason, it won't work inside virtualbox (it may do in software rendering, but it'll suck) [17:00] (last question) [17:00] hhlp: not sure what you mean [17:00] ok we are out of time, what irc channel can people follow up on? [17:00] hhlp, eeePC is a device and UNR is the software that (could) run on that device. [17:01] #ubuntu-mobile [17:01] people can follow up on #ubuntu-mobile [17:01] thx all! [17:01] excellent, thanks guys! [17:01] thank you === jcastro changed the topic of #ubuntu-classroom to: Current Session: Community Q+A: https://wiki.ubuntu.com/MeetingLogs/openweekintrepid/NetbookRemix | Welcome to Openweek, questions in #ubuntu-classroom-chat please || Session details here: https://wiki.ubuntu.com/UbuntuOpenWeek [17:02] are you on facebook I found this via jono on facebook, add me paul gysler [17:02] whats our topic [17:02] ok next session is ubuntu Q+A [17:02] please ask questions in #ubuntu-classroom-chat [17:02] jcastro, you might want to remove the netbookremix -url [17:03] knome: duh, thanks [17:03] np === jcastro changed the topic of #ubuntu-classroom to: Current Session: Community Q+A | Welcome to Openweek, questions in #ubuntu-classroom-chat please || Session details here: https://wiki.ubuntu.com/UbuntuOpenWeek [17:03] jono: take it away, man of metal! [17:04] hey everyone! [17:04] OK I still am at work so I will ask the question, has anyone heard of a point of sale software for ubuntu in a retail store like ace hardware [17:04] yea [17:04] hammer__, please, speak in #ubuntu-classroom-chat [17:04] hammer__: please keep this channel ontopic, we're having a session [17:05] ok folks, welcome to the Community Q+A session :) [17:05] help me out I am noob on IRC [17:05] my name is Jono and I am the Ubuntu Community Manager working at Canonical [17:05] my job is to help the Ubuntu community to be a strong community [17:06] And a musician, "Linux outlaw" [17:06] hammer__, please....#ubuntu-classroom-chat [17:07] oh [17:07] at every open week session I have a Q+A where you can ask any questions you like about the ubuntu community, me, my team or related topics [17:07] so, head over to #ubuntu-classroom-chat and ask your questions there, prefixing them with QUESTION, eg: [17:07] FreelanceJazz will be pasting in the questions [17:07] QUESTION: Why is metal the greatest artform in the work? [17:07] world [17:07] oopshehe [17:07] hehe [17:07] ... lol [17:08] hehe [17:08] thanks FreelanceJazz [17:08] No problem Mr Bacon [17:08] so let me give you folks a few mins to ask some questions, and then we will begin [17:08] QUESTION: Any support for Ubuntu on the Nokia n810? [17:09] from michlinuxguy [17:10] Ubuntu does not run on the n810 right now, but a bunch of related Free Software technologies do [17:11] we have seen the Nokia Internet Tablets be a great breeding ground for free software [17:11] but Ubuntu is not available for it [17:11] unimatrix9: QUESTION: is the ubuntu community structure becoming to complex for mortal soul to understand, or do you think its very transparant? [17:12] good question [17:12] FreelanceJazz: question on ubuntu-clasroom-chat please [17:12] nxvl, he is pasting them for me [17:12] oh [17:12] :D [17:12] then ignore me [17:12] No worries [17:12] nxvl, will do :) [17:12] * jono chuckles [17:12] I think the community always faces the risk of becoming too complex, but have worked hard to keep it as simple as possible [17:13] much of this is about communication - helping people to understand how to slide their brick in the wall [17:13] with so many teams, structures, processes and governance, that can be difficult at times - but in the next cycle we are going to continue to improve this [17:14] I am confident that we can always make the community a simple place to get involved, and initiatives like Ubuntu Open Week, Ubuntu Developer Week, 5-A-Day, Jams and more, really help this along [17:14] but, one important point... [17:14] if you think it is too complex - let me know [17:14] jono AT ubuntu DOT com [17:14] xander21c: QUESTION: Is there cases of Ubuntu LoCO currently working us companies giving support? [17:15] not sure I understand the question, can you ask again [17:15] DoruHush: QUESTION: What others subjects should we choose to discuss? [17:16] in which context, DoruHush ? [17:16] I think they mean 'what can we talk about in this session?' [17:16] no [17:16] read the lines above that [17:16] ubuntu.ro open nights [17:16] 19:10 DoruHush: with only 3 days, called ,,Ubuntu.ro Open Nights" with this sessions: [17:16] 19:10 DoruHush: ,,Ubuntu.ro Team", ,,If you used windows - about open alternatives", ,,How to become involved in translatoins", [17:17] 19:10 DoruHush: ,,Compiling a program - workshop", ,,Using subtitles when payng video files in vlc - workshop". [17:17] 19:10 DoruHush: The ,,Ubuntu.ro Open Nights" will take place next week on #ubuntu-ro IRC channel. [17:17] 19:10 DoruHush: https://wiki.ubuntu.com/RomanianTeam/Proiecte/Suport/SerileDeschiseNov2008 [17:17] ^ context [17:17] I am not in ubuntu-ro, so I can't really answer [17:17] LjL: QUESTION: Can teams (not limited to LoCo teams) obtain an "official" status? If so, how, and how is a team's status checked? [17:18] we don't really have the concept of official teams, but we do have established teams [17:18] some teams have been around since the birth of ubuntu, so many see them as official [17:18] but the definition of official and approved in our community is anything approved by the Community Council [17:18] if the CC deem something right for Ubuntu, you can consider it as official [17:19] RoAkSoAx: QUESTION: xander21c asks if there are cases where Ubuntu LoCo Teams are currently working with companies, by giving support [17:19] sure, there are many cases all over the world where companies look to LoCo teams for advice and support [17:20] many companies typically hear about Ubuntu, and they look to the community to find out more [17:20] and in many cases they find their LoCo who also help them [17:20] this is an excellent, and important function that LoCo teams provide for Ubuntu [17:20] :) [17:20] cyphermox: QUESTION: what do you think is the most interesting achievement that we've made as a community? Aside from releasing Intrepid, of course ;) [17:20] ooh great question [17:21] aside from delivering Ubuntu, every six months, which is in itself an incredibly achievement... [17:21] I think the most stunning thing that we have achieved is our sense of adventure in our community [17:21] we really do have a strong sense of changing things, and having an impact === neverendingo_ is now known as neverendingo [17:22] but importantly, this sense of drive is underlined with an even stronger sense of togetherness in how we move forward [17:22] I am proud that our community actually "feels" like a community [17:23] we work together, we manage problems together, we celebrate together and we console each other together - Ubuntu is not just about kick-arse distribution development, but about friends, and that is the what I think is our crowning achievement [17:23] Xoke: QUESTION: Do you think Nvidia will ever sort their drivers out properly on linux. Not neccersarily opening them up but getting the darn things to work properly (nvidia-96 and 71 still have issues on intrepid). I'm not blaming ubuntu on the nvidia screwups mind, just wondering if they will ever open up to ubuntu [17:24] proprietary software is always riddled with these kinds of issues, and its frustrating for a distribution of course because we generally have a tapestry of free software, and when we see bugs, we fix them [17:25] we can't do that with closed sourced code, and that makes us feel uncomfortable about compromising the quality and stability of Ubuntu [17:25] but I think one day, NVidia *will* get it [17:26] but the driving force is going to be economics - NVidia need to know it is not commercially sensible to have a closed source driver [17:26] and market changes will make that happen - but I think Intel and ATI will drive them that way [17:26] unimatrix9: QUESTION: If Canonical was to stop supporting Ubuntu, because of the financial crisis ( lets hope it does not happen ) what effect would this have on the community? (NB: edited to make it make real sense in the form of an official question) [17:26] ok, so first and foremost, Canonical is *not* going to stop supporting Ubuntu [17:27] and not stop developing Ubuntu [17:27] I know we are in worrying times right now, but we are confident we can ride the storm out [17:28] and Mark is 100% behind that [17:28] so firstly, don't worry [17:28] but now onto how sustainable Ubuntu is without Ubuntu [17:28] oops [17:28] without Canonical [17:29] if Canonical went away, it would naturally affect Ubuntu - we hire a lot of developers and other staff who work on training, support, marketing, certification, testing and more [17:29] all of these functions revolve around the Ubuntu platform [17:29] but at the heart of Ubuntu is community, and open and transparent community, and with this central governance in place, it means that if Canonical went away, Ubuntu could carry on [17:29] LjL: QUESTION: Wouldn't it be beneficial, for various coordination tasks, for teams or projects that have been approved by the Community Council to bear a mark in Launchpad (by being themselvse a member of Ubuntu Members, or another such trick)? === santiago-pgsql is now known as Guest70218 [17:30] LjL, possibly - I am sure that just outlining "approved by the CC" status would change the social economics of the project that much, but possibly - will note it down === Guest70218 is now known as santiago-ve [17:31] r000n: QUESTION: Why Ubuntu includes Mono software even in LiveCD version? Problems with MS patents impossible? [17:31] ahhh the great Mono debate :) [17:31] I am not going to delve into this too far as Mono has been debated often enough across the net [17:31] but its important to remember a few things: [17:32] the Mono project very specifically separate out the core library and framework (which is all legally above board) from the bits which could worry a few lawyers [17:32] as such, the Mono project themselves have insulated the vast majority of potential legal threats [17:34] secondly, a lot of people rag on Mono (the clearly legal bits), but forget that Microsoft have targeted many other parts of the Linux stack, so lets not see Mono as the only problem [17:34] DoruHush: QUESTION: Ubuntu will include in the future multimedia made by the users, that will be placed in „Example” folder. What is the procedure, if one whant to contribute. [17:35] this is the Ubuntu Free Culture Showcase - it is a competition on each release cycle where we like to showcase quality Free Culture media by including it in the Examples/ folder in Ubuntu [17:35] we will be launching the Ubuntu Free Culture Showcase later in the cycle - just stay tuned for more details [17:35] I will announce it on my blog, which will appear on Planet Ubuntu and The Fridge [17:35] Picklesworth: QUESTION: What should new teams and development projects around Ubuntu do so that people are better able and more welcome to contribute? [17:36] the best gift to a new contributor is an outstretched hand === lemma is now known as lemmAway [17:37] every new contributor wants to feel welcomes, and like there is someone there who can help [17:38] as such I recommend that when you see new people, introduce yourself, ask them how you can help them, and get them up and running with things to do [17:38] I also recommend that all teams make it really easy for someone to find out: [17:38] * how to join the team [17:38] * what they need to know to contribute [17:39] * what the team hopes to achieve [17:39] this will help people get up and running quickly :) [17:39] sloopy: QUESTION: Finding info (HOW TO's, ubuntu specific FAQ's, etcs) on the help pages (help.ubuntu.com, wiki.ubuntu.com,etc) is hard to browse often, so i have to resort to using google to find specific info within those pages, any suggestions on how to make help more accessable/easier to get answers to questions and issues? [17:40] I think there are lots of opportunities for us to improve how we deal with help [17:40] I think it would be great to have better methods of searching and indexing help [17:40] and better processes for updating, refining, and knowing what people need help on [17:40] I would recommend working with the docs team on this - this is valuable work === ntinos__ is now known as ntinos_ [17:41] xander21c: QUESTION: Do LoCos can be registered as some kind of organization to obtain profit by giving ubuntu support === lemmAway is now known as lemma [17:41] no, we don't offer direct funding right now through a programme - Canonical does sponsor some LoCos, but that is at our discretion [17:42] cyphermox: QUESTION: Would there be a benefit is growing your team to overlook different parts of the world, e.g. someone for NA, somebody else for Europe, or Asia, etc.? [17:43] I don't think it makes that much sense right now to focus on geographical regions, but my team could benefit from focus on additional areas of the community [17:43] in fact, right now I am hiring a new horseman or horsewoman to focus on translations, and I would like to expand the team into other areas in the future [17:43] unimatrix9: QUESTION: will ubuntu have legal open source DVD playback any time soon? [17:44] I am not really the best person to answer that question - I don't work with ISVs [17:44] ok so we are done with all the questions so fatr [17:44] any more? [17:45] RoAkSoAx: QUESTION: can we register our LoCo team as some kind of organization in our own country? which kind? [17:46] some team have done that with LoCos to make them legal entities, mainly for tax and financial reasons [17:46] I would only recommend that you do this if you absolutely have to [17:46] its a paperwork nightmare [17:46] unimatrix9: QUESTION: are most of the community based in europe , how does it look from the other side of the world? [17:46] the community really is worldwide [17:47] as an example, we have 170+ LoCo teams, in pretty much every country in the world [17:47] we have a huge worldwide spread in all teams and in all kinds of contribution [17:47] any more questions? [17:47] samgee: QUESTION: when Gobuntu died both you and Mark Shuttleworth expressed interessed in working closer together with gNewSense. How is that coming along? [17:48] I have not had a chance to check in with the development team to if we have made any more changes in that direction [17:48] I think Gobuntu helped us in one very specific way though [17:49] it framed a new question in Ubuntu, a question that helped us assess freedom in different ways [17:49] any more q's? [17:51] ok I think we are done [17:51] thanks for the questions folks! [17:51] Wait! [17:51] yusuf_: Question: Ubuntu without the internet is not very intresting. is there improvment being made on this aspect? [17:52] I will finish on this one [17:53] we want Ubuntu to be valuable in the unconnected world, and that is why we care so much about getting as much on the disc as possible - we want to be able to ship a complete system for offline use [17:53] but with the entire world becoming so connection, naturally, a lot of our focus will be there [17:53] thanks folks! [17:53] jono: thanks! \o/ [17:53] and thanks to FreelanceJazz for helping with pasting the questons :) [17:53] No worries [17:53] Xoke> jono, you have to end by saying 'communitising the community, using community tools' :P [17:53] lol [17:53] communitising the community, using community tools [17:54] thanks jono :) [17:54] :) [17:54] for the whole session, not just cutting and pasting :P [17:54] jono: one word... belonging :) [17:54] ok next up is a session on Kubuntyu, stay tuned folks :) [17:54] kenvandine, :) [17:54] Kubuntu, rather [17:55] About ~5 minutes until Kubuntu, now's a good chance for everyone to take a quick break [17:56] I'd rather keep my front row place [17:58] heh === jcastro changed the topic of #ubuntu-classroom to: Current Session: Kubuntu - reinventing QA the resource-limited way | Welcome to Openweek, questions in #ubuntu-classroom-chat please || Session details here: https://wiki.ubuntu.com/UbuntuOpenWeek [17:58] ok JontheEchidna, it's all yours! [17:59] Ok! Let's rock this. [17:59] * knome sheds dudus and steals the best place [17:59] My name is Jonathan, you can call me Jon === santiago-pgsql is now known as Guest17735 [17:59] I assume you are either idling here because you were in the last session [17:59] or you came here to listen to my type about Qa [17:59] which in this case is Quality assurance === Guest17735 is now known as santiago-ve [18:00] Ever been bitten by a bug? Have you ever seen people complain about bugs in Kubuntu on the web? [18:00] I'm guessing most people have, but fear not! You can help [18:01] Starting in the middle of Intrepid we have started to try to make a process to help ensure that Kubuntu is good [18:01] and as a result it is easier than ever to jump in [18:01] To answer the question "How can I help improve the quality of Kubuntu" I will show what we at Kubuntu (should) do to handle bugs and make sure they get resolved [18:02] -Bug reports are the most direct means of feedback [18:02] Without them, we cannot know about bugs [18:02] Something that absolutely anybody can do is report bugs [18:03] you can't know how irritating it is to hear people say "omg kubuntu is so buggy" but then never say what was wrong :/ [18:03] So bug reporting is a great way of helping to make sure issues get resolved [18:03] But, as anybody who has taken a look at the Ubuntu bug tracker may already know, we get a *lot* of bug reports [18:04] and bug reports do pile up over time.... [18:04] so to be able to efficiently handle them all, we must triage them and otherwise get the number of reports down [18:05] Kubuntu is lucky that it generally only has to worry about bugs for KDE or KDE related packages and Kubuntu tools such as the restricted drivers manager [18:05] but we still have ~3000 bugs in launchpad that contain either the words "kde" or "kubuntu" [18:05] and at the moment we have 4 or 5 people triaging the whole thing [18:06] Here is what we do to handle bug [18:06] s [18:06] First, we look for duplicate bugs [18:07] Second, we see if the bug has proper information to be handled [18:07] We want to see at least: Kubuntu/KDE version [18:07] -good backtraces if the bug is a crasher === lemma is now known as lemmAway [18:07] and instructions on how to reproduce are vital for non-crash bugs, and very nice even for bugs with a backtrace [18:08] If you see a bug that lacks any of the above, please request the information from the reporter and set the status to incomplete [18:09] If the backtrace for a crasher contains lots of (no debugging symbols found) lines, the user will need to install the approriate debugging symbol package [18:09] plasma and kwin have debugging symbols in kdebase-workspace-dbg [18:09] most other apps are easy to figure out since they belong to a KDE module [18:10] QUESTION: what are backtraces? and how to get such a thing? [18:10] good question! [18:10] backtraces are logs that give information about where in a program's source code the crash happened [18:10] this can help pinpoint buggy code [18:11] and they help to find duplicates [18:11] continuing [18:12] Sadly, many people are what can be called "hit-and-run reporters" [18:12] They report a bug and assume that the developers are magical fairies that can fix anything [18:13] if nobody responds to a request for information within a month, we generally have to close the bug since it can't be investigated [18:13] and would be clogging the bugtracker otherwise [18:13] But what to do when we do get enough info to investigate the bug? [18:14] If the bug is not something we at Kubuntu have caused, the best course of action is to let KDE know about they bug [18:14] Once we get enough info, reports should be filed upstream and linked to our bugtracker [18:14] This is essential to getting bugs fixed [18:15] Because just like us, KDE can't fix bugs they don't know aobut [18:15] The first step when upstreaming a bug is to search the KDE bug tracker! [18:15] Duplicate bugs aren't nice to the bug triagers upstream [18:16] If you can't find the bug upstream then we can report it [18:16] KDE has a nice feature where you can report a bug to KDE from the About menu of the application. If you have the same version of KDE as the user reporting the bug, you can use this to save time filling out info about your software setup :D [18:17] Once you have posted the bug upstream, you will then need to link the bug back to launchpad === Milyardo_ is now known as Milyardo [18:17] Above the bug description in Launchpad there should be a button that says "Also affects project". From there you paste in the url to the KDE bug report [18:18] A more in-depth look at upstreaming bugs will be given by jcastro in the next session "Upstreaming bugs" [18:18] It is also important to keep an eye on the upstream bug [18:18] subscribe yourself to either the upstream bug or the launchpad bug with the upstream watch [18:19] As bugs get fixed upstream, the fixes do have to make their way back down to us [18:20] If an upstream bug is marked as Fixed, we can set the launchpad bug to "fix committed", because the fix has been committed upstream [18:20] If we are early on in the development cycle, the bugs will generally be taken care of when we update the KDE packages to the latest KDE release [18:21] But if the last KDE release before our Final release has been made, then we should patch the bugs with the upstream fix to make sure they aren't in the final release [18:22] To do this we grab the patch from KDE. They generally tell us what svn revision we can find the patch in [18:22] And then we add that patch to our packages, following this method: https://wiki.ubuntu.com/PackagingGuide/Complete#quilt%20%28example%20package:%20xterm%29 [18:22] Once the patch is sponsored and uploaded, we can mark the bug fixed :) [18:23] Of course, if you have C++ skills you are emplored to attack any of the bugs either in Launchpad or the KDE bug tracker [18:23] This benefits both us and KDE [18:24] = Kubuntu bugs = [18:24] actually, before I get in to non-KDE bugs, are there any questions? [18:25] Continuing on... [18:25] oh, wait [18:25] QUESTION: what about bugs in old kde3 apps that won't get updates from upstream? (like kaffeine) [18:26] Unfortunately we can't do much to get upstream to fix bugs in unmaintained apps [18:26] If volunteers want to fix bugs, then by all means we do accept patches :) [18:27] But, as the topic suggests, we don't really have the resources to fix all the bugs ourselves [18:27] any more? [18:28] I would like to note that if a KDE3 app such as Kaffiene *does* turn out to have a security bug or such, that we will provide patches for that for supported distros [18:28] up until the support for that distro ends [18:29] Hardy should have another year of security updates to ensure that Linux stays secure [18:29] So, that's basically what we do to handle bugflow [18:30] We always would like help, and not only with bugs [18:30] If you are interested in helping everybody's favorite Debian-based KDE distro, we'd be glad to include you in our community [18:30] you can find us in the #kubuntu-devel channel [18:31] If you're interested in actually fixing bugs, that can be very simple too, once you gain experience [18:32] In fact, Friday at 15:00 UTC there will be a session by Daniel Holbach about bugfixing [18:32] "the thing with kaffeine is that the maintainer uploaded some patches to the svn but said that he won't release another old qt3 version, cause the qt4 port will get ready in time. and i was wondering if i might get one of the atsc scanning fixes in the current ubuntu package" [18:32] This actually raises a very good point [18:33] If you know patches that have been included in svn, please file bugs at Kubuntu about this! [18:33] If somebody finds the time we'll gladly include patches coming from upstream [18:33] although, I think we're still one minor version behind the latest kaffiene release... eheh [18:34] If you'd like to help participate in updating patches, we're always looking for new packaging people too! [18:34] We will train you in the secret art of BatPackaging! [18:35] and maybe one day any one of your could be preparing packages for the newest KDE release [18:35] thanks for the info. i thought that i need to get the patch in debian first [18:36] This is handled by us actually [18:36] At the beginning of each Ubuntu/Kubuntu cycle we merge our packaging with Debian [18:36] we take changes that debian has made to their packages since their last merge [18:37] and we also tell them what our changes have been, in case they want to include them too [18:37] Anyway, if you'd like to participate in packaging or merging, you can always join in the fun at #kubuntu-devel [18:38] This is also handy: https://wiki.ubuntu.com/PackagingGuide/Complete [18:38] Oh! I almost forgot to give out resources for forwarding bugs upstream! [18:39] Unfortunately, due to limited manpower, we still have several dozen bug reports lying around that need to be forwarded upstream [18:39] This is a handy tool to check which packages are in need of bugs being forwarded: https://launchpad.net/ubuntu/+upstreamreport [18:39] and this is a list of most kde-specific bugs that need forwarded [18:39] http://tinyurl.com/5b4nwo [18:40] QUESTION: whom would I have to talk to in #kubuntu-devel? [18:40] Anybody really! [18:40] Riddell, apachelogger (who is unsuspiciously absent), vorian, myself [18:40] the great thing about kubuntu is the community in my opinion [18:41] in fact, it was Nightrose who got me in to Kubuntu, by suggesting that I notify people about the problems I was seeing in the KDE4 edition of Hardy [18:41] things sorta snowballed from there... [18:42] QUESTION: Is apachelogger affiliated with the Apache project? [18:42] I have no clue actually. If he wasn't so unsuspiciously missing you could ask him [18:42] he's more of an Amarok guy [18:43] Amaroklogger doesnt' quite roll off the tounge... [18:43] :P [18:44] QUESTION: do a lot of you also work on KDE/Amarok/others? [18:44] Several of us do, in fact. [18:44] Smarter works upstream with KDE and hacks on accessibility apps [18:45] Nightrose and apachelogger do a lot of work with amarok [18:45] making packages for Kubuntu [18:45] and Nightrose is actually an amarok team member [18:45] Riddell contributes patches upstream when he's not too busy being the almighty Kubuntu robot [18:46] I have started working with mornfall of the Adept project [18:46] and we also have a few gnomies on our team :P [18:47] Kubuntu really is a mixture of some of the best people in the Free-software world [18:47] QUESTION: Do you cooperate with the Ubuntu/Xubuntu teams [18:47] Good question too [18:47] Recently NCommander has been contributing to Kubuntu. He's an Xubuntu dev too [18:48] I haven't heard him complaining about what we do, so we must be doing something right :P [18:48] I think our relations with Ubuntu could be better, but both sides will have to work at that [18:49] QUESTION: What do you think about many people being unsatisfied with the new KDE4 desktop compared to the 3.x series? [18:49] Now that is the million-$CURRENCY-question, now isn't it... [18:50] Personally I think anybody who doesn't like KDE4 is a hater :P [18:50] * knome coughs loudly [18:50] nah, some people have valid reasons for not likign KDE4 [18:50] People are creatures of habit [18:51] and even though the KDE team has done a great accomplishment rewriting as much of KDE as they have done in 8 months (they have done most of it) I recognize that it's not all there yet [18:51] but at the same time I recognize that things will get better [18:51] in addition, we'll also see things in KDE4 that we never coudl have seen in KDE3 [18:52] QUESTION: what was the overall reaction to Kubuntu 8.10? [18:52] I'd like to think that the overall reaction is good [18:52] You do have a very vocal minority that frequents such places as dot.kde.org [18:53] where the announcement thread turned into a 300-post complaint about KDE4 not being ready [18:53] but I've also heard very positive things about Kubuntu 8.10 even from people who thought they wouldn't like it [18:54] QUESTION: Are you going to offer something shiny with the next Kubuntu release and try to make new users or to make KDE4 more stable and maybe suitable for the people who stopped using Kubuntu again? [18:54] I'd say... both! [18:54] The next KDE4 release itself is going to come with many shiny things [18:54] Desktop cube effect [18:54] panel autohiding [18:54] and lots of neat polished features and enhancements [18:55] but at the same time we'd like to emphasis on stability [18:55] which is what this whole session is about, really [18:55] we're running out of time, so the next question is the last [18:55] QUESTION: what do you plan to realize in 9.04? [18:56] This is a rought draft of things we're going to discuss as possible features for 9.04: https://wiki.kubuntu.org/KubuntuUDSJauntySpecs [18:56] I must emphasize that not all of those features will be pursued [18:56] but this is a wishlist the devs have put together for discussion at the developers summit [18:58] Next up we have Jorge Castro with his session "Upstreaming bugs" [18:58] thanks JontheEchidna! [18:58] I'll be happy to answer further questions in #kubuntu [18:58] ...to a degree though :P [18:58] * Oxyhydrogen applauds [18:58] ^_^ [18:58] * rgreening cheers === jcastro changed the topic of #ubuntu-classroom to: Current Session: Upstreaming Bugs | Welcome to Openweek, questions in #ubuntu-classroom-chat please || Session details here: https://wiki.ubuntu.com/UbuntuOpenWeek [18:58] ok I will start in about ~2 minutes [19:01] ok, thanks everyone for coming, my name is Jorge Castro and for the next hour I will be talking about upstreaming bugs [19:01] Hopefully if you're at this session you've attended bdmurray and pedro's talks about bugs [19:01] but if not hopefully you can get something out of this session anyway. :D [19:01] So what exactly do I mean by upstreaming bugs? [19:02] First off, what we ship as "Ubuntu" contains a bunch of software written by other open source projects [19:02] this can be large projects like linux, GNOME, KDE, X.org, and others [19:02] but also includes smaller projects that you might not have heard of before [19:03] Ubuntu fits into the picture by bundling this software and shipping it as a distribution every 6 months [19:03] As such it is up to us to make sure that when our users report bugs that the bug report makes its way to the right developer so that they can address it. [19:04] We do this in our bug tracker, launchpad, which has the ability to link Ubuntu bugs with upstream bug trackers. [19:04] So in this session I will go over some workflows on how to do this efficiently, and give you some examples of ideal bugs that we've forwarded. [19:04] Any questions so far? [19:04] (Please ask in #ubuntu-classroom-chat) [19:05] Ok, moving on ... [19:05] For what I call the "top100 projects" in ubuntu, we generate a chart that shows us how well we're linking bugs [19:05] in this instance it's the "top100 projects sorted by open bugs" [19:06] we make this chart available here: https://launchpad.net/ubuntu/+upstreamreport [19:06] We call this the "upstream report" [19:06] what this report allows us to do is to glance at a large cross section of the distro to see how well we're doing overall, and more importantly, to find out where we need to improve [19:07] So everyone if you would, click that link so you can follow along [19:07] So let's look at an example [19:07] I will just pick openoffice.org, which is the third row down [19:07] As you can see, right now there are 523 open bugs in openoffice.org in Ubuntu [19:08] that means that right now there are 523 possible defects that need to be addressed. [19:08] Of these, 364 are in a triaged state, which means that a developer has looked at it and has determined that it is indeed a bug. [19:08] The upstream column is where it gets interesting [19:09] That number is 338. [19:09] That means that out of those 523 bugs reported, a developer has determined that 338 of them are bugs in OpenOffice itself [19:09] http://img530.imageshack.us/img530/3820/pascii6.png [19:10] of those 338, 315 have a bug link in launchpad to the openoffice bug tracker [19:10] this means that we have a 93.20% linkage rate [19:10] which is very good (we always want to be over 90%) [19:11] so basically, when a developer determines that a bug is upstream, they open an upstream task in ubuntu, which then shows up in the upstream column [19:11] then it's up to bug people like us to make sure that these bugs get linked upstream, which I will show you in a minute [19:12] So, what really you need to be concerned about is the last column [19:12] these are bugs that a developer has deemed to be "upstreamable" [19:12] but has NOT been linked to an upstream bug report [19:12] as an example, if you scroll down to "kdebase" [19:12] you'll see that the last column is a 12. [19:12] click on the 12 [19:13] and it takes you to this list: https://bugs.edge.launchpad.net/ubuntu/+source/kdebase/+bugs?search=Search&field.status_upstream=pending_bugwatch [19:13] So these 12 bugs are something that kubuntu developers and QA people have determined to be an upstream issue [19:13] so let's take a look at one [19:14] please click on bug 172682 [19:14] which takes you here: https://bugs.edge.launchpad.net/ubuntu/+source/kdebase/+bug/172682 [19:14] Launchpad bug 172682 in kdebase "Dolphin cannot open a tar file on a smb share" [Wishlist,Confirmed] [19:14] So, as you read the bug report you will find where Jonathan says: "The best bet for getting this feature is to file a wishlist report at http://bugs.kde.org. If you (or anybody) files a wishlist it would also be handy to link the report here or at least post the link to the wishlist in the comments." [19:15] So ideally, you would go to the KDE bugtracker [19:15] and then first search to see if a bug is already filed. [19:16] If it is, you would get the url to that bug, click "Also affects project" and paste the bug url in there [19:16] at that point launchpad will make a bug watch to monitor the bug [19:16] if in the future the bug is fixed the bug watch will reflect that [19:16] And at that point the bug is considered fixed upstream, but NOT yet in ubuntu [19:16] at that point the bug would show up in the harvest tool so a developer can look at it [19:17] I have an example of this just from today [19:17] https://bugs.launchpad.net/ubuntu/+source/brasero/+bug/292201 [19:17] Launchpad bug 292201 in brasero "brasero don't add files to data project by drag and drop" [Low,Fix committed] [19:17] So this bug was reported on november 1st. [19:18] at one point Pedro decided that this was a bug in brasero itself, not in anything that we've done in Ubuntu [19:18] so he filed a bug in the upstream GNOME bugzilla [19:18] then he pasted the link in a comment, clicked "Also affects project" and pasted the GNOME bugzilla URL in there [19:18] if you look at the top of the bug report you will see that it lists "gnome-bugs #559173" [19:19] If you click on that you can go see the report in the GNOME bugzilla [19:19] If you see Pedro's report here: http://bugzilla.gnome.org/show_bug.cgi?id=559173 [19:19] Gnome bug 559173 in general "brasero don't add files to data project by drag and drop" [Normal,Resolved: fixed] [19:19] you will see that he also put a link to the launchpad bug in that report [19:20] so if you check both pages you'll see this workflow [19:20] reporter -> pedro -> brasero developer [19:20] Basically pedro is being a link between an ubuntu user and a GNOME developer [19:20] the great thing is, anyone can be this link! [19:21] so what I tell people is to pick a small project to start off with [19:21] something you want to care about [19:21] and then go through it's bugs in ubuntu, and search for those bugs in the upstream bug tracker [19:21] and then make the link in launchpad and in the upstream bug tracker [19:21] This helps get information to upstream developers faster [19:22] and ultimately will help get fixes out to the user faster, which is what it's all about [19:22] QUESTION: when clicking on Also affects project, the form asks for a project name, do we just paste the bug url anyway? [19:22] Sometimes some information for a certain package is not filled out properly in launchpad [19:22] It won't let you paste a url into that field [19:23] what I do is fill in the project name, hit search, then click on the title, and then it will fix that project and allow you to put a URL in [19:23] QUESTION: will the status for the gnome-bugs link in launchpad automatically update? [19:23] thanks rick_h_ [19:23] Yes [19:23] in that brasero example it has not updated yet, which is why it's unknown [19:23] QUESTION: How do you find out if the bug really also affects upstream and is not ubuntu-related? [19:23] We use a tool called harvest as a "radar" that shows us when something is fixed upstream [19:24] That is a tough question, probably the crux of the matter [19:24] If you use the last column in the report and look at those bugs [19:24] those will already be determined to be upstream, so you're safe to look at those and make linkages [19:24] with some bugs it's easy [19:25] usually if it's a feature request or a wishlist it's very much an upstream bug [19:25] so "I wish firefox did this and that" is probably an upstream bug [19:25] "firefox crashing" would be a harder one to solve [19:25] which is why you should consult people in #ubuntu-bugs or the corresponding team for that package [19:26] still, there are plenty of bugs in the last column that you can link that are already determined to be upstream [19:26] so there's no shortage there [19:26] The way I do it is just work off the last columns, and let smarter people make the decision on wether it's upstream or not. :D [19:28] < homy> sorry, I don't understand that "last column" jcastro talks about. [19:28] ok no problem [19:28] jcastro: can you post the list to the page with the "last column" again? [19:28] https://edge.launchpad.net/ubuntu/+upstreamreport [19:28] that last column on the upstream report [19:28] That is the list of bugs I look at [19:28] Ok, so how does this all tie together? [19:29] When a bug watch shows that a bug is fixed upstream it shows up in Harvest [19:29] the harvest page for brasero looks like this: http://daniel.holba.ch/harvest/handler.py?pkg=brasero [19:29] so basically, in the past 2 months there have been 6 watches that have been shown as fixed [19:30] the reason we have this harvest page [19:30] is that an ubuntu developer can check on each of those bugs, find the fix, and fix the package in ubuntu [19:30] so basically, I see it this way [19:31] if you make links to upstream trackers, as those get fixed, they start to show up on harvest [19:31] we basically are what I call "feeding the harvest machine" [19:31] from there ubuntu developers have a nice little list of possible fixes that they can look at to fix the packages in ubuntu [19:32] so really, if you know how to search for bugs in launchpad and bugzilla [19:32] just making the link between the two can really help developers find the right fixes [19:32] while I am at it, I did a blog post on how this all works here: [19:33] http://stompbox.typepad.com/blog/2008/09/introducing-the.html [19:33] that has screenshots and examples as well [19:33] So if you're lost right now you can refer to that document and the logs from this session [19:33] < homy> ah, I think I understood: when I click on the "40" in the firefox-3.0 column for example, I get all upstreamable firefox-3.0 bugs that don't have an upstream bug report yet? [19:33] yes! correct! [19:34] Now [19:34] the upstream report is a great tool for finding possible upstream targets [19:34] but like all tools, you can use it to cause havok as well [19:34] so just because you see 40 bugs doesn't mean you should go open up 40 bugs in firefox's bug tracker and paste in everything [19:35] Instead you should use it as a guide to get started with [19:35] You don't want to be in a vacuum forwarding these bugs without working with someone [19:35] so for example for these bugs I would look at them [19:35] determine which ones would be easy low hanging fruit [19:36] then I would post on the ubuntu-mozilla mailing list or ask someone in the irc channel [19:36] something like "I am interested in helping triage bugs upstream, I am looking at this list here, any thoughts?" [19:36] or something like that [19:37] at which point someone on the team would give you some advice or tell you what to do or point you to more information [19:37] It is important to tell someone in a team that you are looking at those bugs [19:37] for example the desktop team does a great job of doing linkages, they have their little way to handle upstreaming bugs [19:38] so I try not to touch them unless it is obvious that I can do it [19:38] even then, sometimes I make a mistake and link the wrong bug [19:38] at which case someone will tell you it's a mistake and then fix it [19:38] Any more questions so far? [19:39] ok so more tips [19:39] So it's obvious that bugs filed under "linux" are far and away the most open bugs. [19:39] These are tough because in a lot of cases they are hardware specific [19:39] and not very easy to track down [19:40] so in the case of kernel bugs you'll definately want to work with the kernel team on that [19:40] Sometimes you look at an ubuntu bug [19:40] and then you find a bug in the upstream project that might be similar [19:40] but you might not know [19:41] there is absolutely nothing wrong with posting a comment asking "Is this the same bug as this one?" [19:41] I'm sure many of you have seen these kinds of comments [19:41] it's basically people who /think/ they might have found the same bug, but aren't sure, so they're asking for help from someone who might know [19:41] there's nothing wrong with that of course! [19:42] Ok, so the last thing before I open it up to more questions [19:42] https://launchpad.net/ubuntu/+upstreamreport [19:42] If you scroll to the very bottom [19:42] you'll see a total for the top100 projects === syslogd_ is now known as syslogd [19:43] the two that I look at is the percentage of bugs that are determined to be upstreamed, and the amount of linkages [19:43] those numbers are 20.12% and 93.68% [19:43] this tells us that out of all the bugs we have reported [19:43] 20% are determined to be upstreamed [19:43] and of those, 93.68% are linked [19:44] The 93% tells us that when we do find that a bug is upstreamable, that we do a great job at making that link [19:44] that's people out there finding the bug in an upstream tracker and linking it in launchpad. [19:44] I am interested in growing the first number. [19:44] The more we determine that bugs are upstream the larger our targets for linkages become [19:45] One last example [19:45] look at virt-manager in this chart [19:45] it's red for a reason. :D [19:45] Out of 21 possible bugs that can be upstreamed, 0 have been reported to the virt-manager developers [19:45] that's where we can help out to get those bugs to the right people. [19:46] More questions? [19:46] Ok well, I realized I chatted a bunch and it's a lot to swallow [19:46] but hopefully you will be able to use the report to find bugs that can be linked [19:46] if anyone has any feedback or questions on forwarding reports, please let me know [19:47] I am always on irc and can be reached via jorge@ubuntu.com [19:47] So uh ... smoke if you got 'em! [19:47] ty, Great job on this! [19:48] jcastro: Yes, thank you! [19:48] Next up will be nand with Brainstorm Q and A! (I suspect this session will be popular!) [19:48] in about ~12 minutes [19:50] QUESTION: Are you planning to extend the "upstreamed" list for more packages? [19:50] yes [19:51] currently we went for the top100 by open bugs because that covers a huge area [19:51] eventually you'll be able to just ask for a package [19:51] and get all that info [19:52] in fact I think that the top 100 packages constitute about half of all open bugs in Launchpad [19:52] yes [19:52] I got another question [19:52] let me paste it [19:52] "Hi I got a question. You said that one should not touch all bugs for a package in one go for example. I did not get the reason behind this and my other question is what can be done to increase the reporting in upstream trackers both automagicaly and\or with human labour do you think?" [19:53] ok, so the reason we don't do this automatically is that we depend on your human brain to do a common sense thing [19:53] sometimes the bug might be incorrectly marked as upstream [19:53] or sometimes a developer won't mark a bug as upstreamed until he/she gets more information [19:53] what we don't want is to flood upstream bug trackers with bad bugs [19:54] ideally when an upstream gets a link from us in their bug tracker they should (hopefully) think "Ah, another one of those ubuntu studs with a rocking bug report" [19:54] This is why it's important to ask someone in a team working on that package for guidance [19:55] Because as it turns out, if you're forwarding bad bugs upstream that shouldn't be, you end up doubling the work for the ubuntu developer, the upstream developer, and probably everyone subscribed to the bug [19:55] they especially don't like it when it turns out to be an ubuntu bug :P [19:55] So it's important that you try to do the right thing [19:56] instead of fast-fooding a bunch of bad bugs [19:56] that being said, if you make a mistake, just fix it and move on [19:56] I mislinked a bug this morning and found out right away that I had made a mistake. :D [19:56] Question: So what can we do to make it more effective_ [19:56] So really alot of this is experience [19:57] I recommened participating in hug days [19:57] where you can join a channel and during that entire day you are focused on one package [19:57] our bugmaster and QA people are around for these [19:57] so those are the days to ask questions and pick their brains [19:57] after a while you will just start to get good at it, like any skill [19:57] ok so I am out of time, thanks everyone [19:58] * c00l2sv help [19:58] nand: you're up in 2 minutes! [19:58] * c00l2sv nu are nimeni chef să ajute un pic la traducerea WordPress-ului (300 de stringuri au mai rămas!) === jcastro changed the topic of #ubuntu-classroom to: Current Session: Ubuntu Brainstorm Q+A | Welcome to Openweek, questions in #ubuntu-classroom-chat please || Session details here: https://wiki.ubuntu.com/UbuntuOpenWeek [19:59] ok! [19:59] * c00l2sv http://tradu.softwareliber.ro [20:00] ok ok, so now's Ubuntu Brainstorm time [20:00] hmm [20:00] ok let's start with what is Ubuntu Brainstorm [20:00] or rather [20:00] what is its use [20:01] I guess you all have ideas on how Ubuntu could be improved [20:01] even lots lots lots of ideas [20:01] but the thing is, your ideas are subjective to your relation to your computer, of what you're doing [20:02] charla de kubuntu arranca en dos horas [20:02] your requests may be not as important for others [20:03] how do we start on this. [20:03] so basically it is not easy to grasp what are the main requests out there. Some people can be very vocal about something, but they may represent a tiny portion of the userbase [20:04] so Ubuntu Brainstorm is a place where you can post your ideas, and vote on the ones you agree on [20:04] This week cut and past between programs such as vector graphics from inkscape to openoffice. [20:04] and the voting system makes it easy to assess what is the most asked [20:04] mttr: please discuss on #ubuntu-classroom-chat [20:05] ok thanks [20:05] ok so we can see what the most asked, and that will be taken into account during the next Ubuntu Developer Summit, where the planning is laid out [20:06] so that we can maximize the awesomeness factor of Ubuntu [20:06] for people who didn't quite follow what's going on on Brainstorm, you can fetch a summary I wrote : http://www.ndeschildre.net/downloads/UbuntuBrainstorm810Report.html [20:06] Ok [20:07] Now, Ubuntu Brainstorm is far from perfect [20:07] it is an experiment : besides Dell's Ideastorm, this has never been used on this scale [20:08] so things are not perfect . Often ideas are duplicates of others, ideas are not well formed [20:08] we are open to critics [20:09] A new version of Ubuntu Brainstorm will come Soon (c) which will try to fix some of these issues [20:09] You can see a preview here: http://devel.ideatorrent.org/ [20:09] :) [20:10] The main improvments are Idea and Solution separation, to force users to structure their ideas === Rafik_ is now known as Rafik [20:10] and more moderator tools, including a brand new moderator team (23 people so far!) [20:11] ok... any question so far? [20:12] QUESTION: what happens to our old ideas with the new system? [20:13] they will be of course transfered to the new system. But as there was no idea/solution separation, there will be a "fake" solution [20:13] See for example : http://devel.ideatorrent.org/idea/14433/ [20:13] From now on, you will be able to post several solution to one idea [20:13] and the voting will occur on a per-solution basis [20:14] Part of the job of the new moderators will be to separate the rationale from the solution to make some clean ideas [20:14] ok, so let's continue with moderators [20:15] Being a moderator on Ubuntu Brainstorm is a hard work [20:15] because until recently we were a very few [20:15] and we gets dozens of duplicate reports every day [20:16] (So far, almost 4000 ideas have been marked as dup!!) [20:16] and marking as duplicate is not an easy task === santiago-ve is now known as Guest34739 [20:16] as it looks [20:16] you have to read fully and understand both ideas, and make a decision, sometimes hard [20:17] that's why, this plus the limited number of moderators, there was unfortunately some ideas incorrectly marked as dup [20:17] but behooold! [20:17] things will change soon [20:18] once the new version is out [20:18] anyone will be able to join the Ubuntu Brainstorm moderator team to give an hand! [20:18] the team: https://launchpad.net/~brainstorm-moderators [20:18] and https://wiki.ubuntu.com/Brainstorm/BrainstormModerators [20:19] there will be two level of moderators [20:19] 1) Idea Reviewer : anyone can become one, and your job will be to review incoming ideas [20:19] Ooooops [20:20] I happen to have forgot another new fanstastic feature :) [20:20] http://devel.ideatorrent.org/ideas_in_preparation/ [20:20] You may have noticed this "Idea sandbox" [20:20] Now, when an idea will be submitted, it will have to be reviewed by 2 moderators, for some very basic guidelines [20:21] See more explanations here: http://devel.ideatorrent.org/faq/#question14 === Guest34739 is now known as santiago-ve [20:22] In summary, moderators will check you are not posting a dup, a bug report, and check that your idea is precise (no more "Make Ubuntu Faster"), and with some kind of realistic solution [20:22] so [20:22] 1) Idea Reviewer : anyone can become one, see explanation on https://launchpad.net/~brainstorm-moderators [20:22] the job will be to review incoming ideas, discuss with the author to add more details, and so one [20:23] and once you did a great job for some time, you can become a: [20:23] 2) Moderator : who can virtually do anything [20:24] To summarize, being a moderator, you can: [20:24] - Review incoming ideas [20:24] - Mark ideas as duplicates [20:25] - Separate rationale and solution from imported ideas from Brainstorm v1 [20:25] The main goal being [20:25] *Making Ubuntu Brainstorm clean and exploitable* [20:26] because anyone do not expect to spend hours through bad ideas [20:27] ok, some more questions? [20:27] QUESTION: as an idea reviewer, is your task only approving/declining idea's, or does one have to write a reason? [20:28] as an idea reviewer, you can just approve, or discuss with the author to help him improve his idea [20:29] ultimately, if the idea is not one, or the author do not correct it, you can mark it as "not an idea", or "already implemented" (another kind of bad idea) [20:29] Ah, and if you want to see what to expect [20:30] you can just log in using the test accounts here : http://demo.ideatorrent.org/ [20:30] you will see what are the moderators tools for each kind of roles (moderator, admin, idea reviewer ,....) [20:30] I recommend you to check that! [20:30] QUESTION: have any other projects (e.g. gnome, firefox, etc) expressed an interest in using ideatorrent? [20:31] I have seen some interests in mailing lists/forums, but so far, no project have actually started using it (besides blender using an old customized version of it) [20:32] the weird thing is, I get more requests from corporate companies (Stockholm airport is going to launch its ideatorrent soon, and another *BIG* name too) than from open source projects! [20:33] QUESTION: what *big* company in particular? [20:33] sssshh, that's a secret! [20:33] I guess you'll know quite soon === root_ is now known as papapep [20:33] Go Sweden ! sorry could not help myself:-) [20:33] QUESTION: The current models for determining what's "popular" seem really really bad - an idea with 100 up votes and 0 down votes is considered less popular than an idea with 1101 up votes and 1000 down votes, simply because more people have voted on the latter. The current hack to this is to count ideas as less popular as more time elapses in a weird hyperbolic fashion (divide by number of days). Why not just move to [20:34] * nand is waiting for the end of the question :) [20:34] what a question [20:34] anyway, starting with the start of the question [20:34] I often get this remark [20:35] but if you look closely at http://devel.ideatorrent.org/, you'll see that you can now see the proportion of +1, -1, +0 votes [20:36] and that, when an idea has > 100 votes, the proportion is more or less always identical : 80 to 95 percent of +1 [20:36] See for examaple : http://devel.ideatorrent.org/idea/14433/ [20:37] The green and red bar just under the number of votes [20:37] 419 promotions / 31 demotions [20:37] (in the tooltip) [20:37] so in conclusion, [20:38] and that answers: QUESTION: Why not just move to a "percentage upvote" model after some 100-vote grace period? [20:38] looking at stats I have, this would not be a good model [20:39] but I'm still open to discussion on better improved models to order ideas [20:39] QUESTION: Even if it's often "90% up votes" surely there's meaning in lowering one that's only 85% or the occasional one that's worse than that? [20:39] I quite don't get the question... [20:40] QUESTION: I'm especially worried about overlooked ideas. An idea might be very very good, but few people understand it (and thus don't vote on it), meaning you won't see it in that "popular" listing even when it gets >90% approval [20:42] hmm, I think you misunderstand the current "popular" algorithm [20:42] the current "popular" idea algorithm is quite simple : it is ordering based on (number of votes) divided by (time since creation) [20:42] or more precisely, by the number of votes per unit of time [20:43] so you could even see negative ideas on the popular listings, if you go at the end of the list [20:44] ok, still a little time... [20:45] so if you're interested about Brainstorm and moderation, you can go to #ubuntu-brainstorm, or discuss on the mailing list here: https://launchpad.net/~brainstorm-moderators [20:46] QUESTION: when will brainstorm.ubuntu.com actually *use* ideatorrent? [20:46] well [20:46] hrm [20:46] hmm [20:46] hopefully within the month. But don't take that as a promise! [20:47] (and note that brainstorm.ubuntu.com IS using ideatorrent. It just got some major changes!) [20:47] QUESTION: so if an idea from the sandbox gets approved by 2 reviewers it will get to the popular ideas? [20:47] yes. [20:48] An approved idea doesn't neccesarily have to be popular.. [20:48] "Popular ideas" is kind of not the real good name, I agree, but it's better than "approved ideas" [20:48] let's say I put that in lack of a better name :) [20:48] QUESTION: ideatorrent is just a new drupal theme? or something more? [20:49] Ideatorrent is a Drupal module (open source of course) [20:49] you can download it at http://www.ideatorrent.org [20:49] :) [20:49] 10 mn left... [20:49] go go go questions! [20:50] * nand just realize http://www.blenderstorm.org is actually quite used, 700+ ideas so far [20:51] QUESTION: how will you make approved, but not-popular idea's visible? [20:51] well, first new non-popular yet ideas will appear on the "latest ideas", "most popular these 24H" lists, where most of the initial voting goes [20:52] then, if it is quite popular, it will go on "most popular this week" and get more votes [20:52] then, if it is quite popular, it will go on "most popular this month" and get more votes [20:52] then, if it is quite popular, it will go on "most popular this 6 months" and get more votes [20:52] and so on :) [20:52] that's the current classic behavior [20:52] plus some cheaters are using digg.com to help themselves :) [20:52] Quite a lot of ideas are digged [20:52] and some of them reached the digg popular state [20:53] QUESTION: how much ideas implements from Brainstorm? [20:53] Concerning Intrepid only, that's 18 : http://brainstorm.ubuntu.com/implemented_ideas/ [20:54] but note that quite a few of them, we missed and are still in the "New" state, because of lack of moderators! [20:54] QUESTION: Why are you calling drawing attention to an idea with Digg "cheating" -- it seems like it would only give something more votes total, not necessarily more upvotes [20:54] it was just a joke :) [20:55] I encourage all possible external websites to add more votes to the current 1.3 million of them! [20:55] and you can promote your ideas using the images links on your blog e.g. http://brainstorm.ubuntu.com/idea/10701/promote/ [20:55] Question: Is there an widget\applet\gadget to subscribe to new apporved ideas with an voting mechanism directly from the desktop? [20:56] That could be an interesting thing to look at. You will be able to suscribe to a RSS feed, but I haven't looked at the voting question yet... [20:57] ok ok, 3 minutes before some Virtualization talk! [20:59] ok then thanks all for your interest, I'm hoping to see you soon to discuss moderation and join our awesome big brand new moderation team! [21:00] thank you nand! [21:00] o/ [21:00] Hi, everyone. [21:01] Welcome to the obligatory virtualisation talk. [21:01] It wouldn't be Ubuntu OpenWeek without it, would it? === jcastro changed the topic of #ubuntu-classroom to: Current Session: Virtualisation | Welcome to Openweek, questions in #ubuntu-classroom-chat please || Session details here: https://wiki.ubuntu.com/UbuntuOpenWeek [21:02] I have to say this session is as much for my sake as it is for yours. I'm hoping to get some good feedback from you about how you're using it and what your problems are. [21:02] Or perhaps even better: Some input on why you're not using virtualisation. That would be interesting. [21:02] So, to kick things off. [21:02] .. [21:02] New stuff in Intrepid: [21:03] Intrepid didn't see huge, world changing updates in the basic virtualisation stack. [21:03] The virtualisation stack in my mind consists of: [21:03] kvm [21:03] libvirt [21:03] virt-manager [21:03] virtinst [21:04] They all got updated to more recent upstream versions, which gave us a few new features, such as fully swappable kvm guests and a few other tidbits. [21:04] In intrepid, we also completely rewrote our vm builder. [21:05] If you're not familiar with vmbuilder, it's basically a tool that will create a virtual machine for you in a fraction of the time it would take you to complete a regular install. [21:06] If we have time at the end of this session, I can do a demonstration. [21:06] Just shoot questions in #ubuntu-classroom-chat when you have them, by the way. Don't wait until the end. [21:07] In case you're just tuning into the whole virtualisation buzz, let me explain a bit about what it's good for. [21:08] Actually, virtualisation comes in many shapes and varieties. We've chosen to focus on what is called full virtualisation, which is what kvm provides. [21:09] kvm is a piece of software that runs on your regular system, but looks like a completely separate computer. You can install any operating system in it you like. Windows, other Linux distros, other versions of Ubuntu. [21:09] Anything that you can do to a regular PC, you can do to one created by kvm. [21:09] This is by far the most versatile form of virtualisation. [21:10] If you're a developer, you can set up other operating systems to test interoperability between different things... [21:10] ...you can have other versions of Ubuntu running in it to see if a bug you found is a regression from an older version of Ubuntu.. [21:11] ..or even check if a bug has been fixed in a later version. [21:11] You can use it as a "clean room" for doing various tests that might hose your system. [21:11] Or you can use it in production environments to keep services separate. [21:11] The sky is the limit! [21:11] 21:08:33 < ~MadsRH> soren -> The vmbuilder is setup in the terminal? Are there any plans for a GUI? [21:12] VMBuilder is really a library, but also comes with a command line tool to use it. [21:12] ..this is what most people will use. [21:12] However, it was designed to be a library so that it could be easily embedded into other Python projects. [21:13] Perhaps virtinst could use VMBuilder as an alternative form of OS installation. [21:13] ...or it could be exposed directly in virt-manager. [21:13] ...or a web frontend could be used to build VM's. [21:13] Anything is possible. All we need are ideas and code monkeys :) [21:13] 21:11:22 < knome> QUESTION: How effienct kvm is compared to, let's say, VMWare? [21:14] At the risk of offending people who've worked on either side optimizing particular things in each of them, I'd say they're roughly comparable. [21:14] With a twist, though: [21:15] kvm - as you may know - only runs on newer hardware, which has virtualisation features built into the CPU. VMWare runs on any PC. [21:15] If you run VMWare on a PC without virtualisation extensions, I'm quite sure kvm will outperform it somewhat. This is due to the nature of that type of virtualisation. [21:16] VMWare can take advantage of those extensions, too, though, so on the same hardware, they'll be on par. [21:16] 21:06:36 < YokoZar> QUESTION: I'm interested in shipping a VM appliance for using an application with a graphical interface. It's very easy to do a minimal VM-based installation using the Ubuntu-server CD, but it's not at all clear to me how I get a minimal desktop suitable for a VM appliance. [21:16] Installing the Ubuntu-destop package, for instance, will install all manner of video drivers that arean't needed. [21:17] For starters, a VM appliance (or virtual appliance) is a virtual machine that you distribute to clients or customers that performs a single task. [21:18] The idea is that you preconfigure it to do a few things really well, and usually wrap it all up in a fancy wrapping. [21:19] A colleague of mine has written a HOWTO about creating such an appliance. [21:19] ...it even ended up in Linux Mag: http://www.linux-mag.com/id/4829 [21:20] It's an area where we still have a lot of work to do. it could certainly do with a better frontend for deployment.. [21:20] ..such as setting it up in the end user's network, perhaps hooking it into whatever other services it needs to interact with, etc. etc. [21:21] If anyone has the time to work on something like that, please do get in touch. [21:21] 21:07:29 < alexharrington> QUESTION: What does "fully swappable" mean in relation to a kvm guest? [21:22] Not too long ago, if you had a kvm guest running that used 2 GB of memory, those 2 GB of RAM on the host was stuck in RAM and couldn't be swapped out to disk. [21:22] This is no longer true. [21:22] This is actually the primary reason the /dev/kvm device wasn't open to everyone. [21:23] Anyone with access to that device could pin all your RAM, which was rather unfortunate. [21:24] 21:12:29 < YokoZar> QUESTION: So, is there going to be support for a "minimal desktop" sort of VM appliance sometime? Like another package seed? [21:24] I don't really think a graphical environment in an appliance is something we want to put a lot of effort into. [21:24] I'm much more inclined to go the way of your garden variety router which has a web interface for configuring everything. [21:25] Graphical environments are just a hassle :) [21:25] 21:13:13 < Zizou> Question: which is the adventaje in kernel based virtualisation comparing to software virtualisation? [21:26] I'm not entirely sure what you mean by software virtualisation.. If you mean full virtualisation, but done entirely in software, like qemu does it, the main advantage is speed. [21:26] Emulation is a slow, slow process. [21:27] If you're thinking more along the lines of jails or containers, the main advantage is versatility. [21:27] You can't run Windows in a jail or a container. You need something the pretends to be a complete PC so that Windows doesn't get all confused. :) [21:27] 21:13:37 < Israphel> QUESTION: in the setup of my mother there's a virtualization option to enable, should I enable it? [21:28] It all depends. [21:29] Perhaps your mom needs to run a few Windows apps even though she has switched to Ubuntu. Instead of going the way of dual booting, you could enable virtualisation on her machine, set her up with kvm and virt-manager and have her run those last few windows apps inside kvm. [21:29] A few years from now, I think we're going to see much more widespread use for virtualisation. In fact, by then it's likely that you'll be using it without even knowing it. [21:30] 21:14:19 < zul> QUESTION: what features does the vmbuilder rewrite has? [21:30] The new VMBuilder features an extensive plugin system. [21:30] This makes it much, much easier to add support for other distros and hypervisors, and even to add entirely new functionality. [21:31] All the libvirt integration it does is contained in a plugin by itself, for instance. [21:31] It's also a lot more friendly to watch, and it's simply much more flexible in many, many ways. [21:32] 21:14:31 < nealmcb> QUESTION: So just how fast is vm-builder, anyway :) [21:32] I forget the most recent benchmarks I've done, but under optimal conditions, I'm certain it can create a virtual machine in around a minute, probably less. [21:33] I have a few optimisations in my back pocketet that should shave off an additional 5-10 seconds from that benchmark. [21:33] 21:14:35 < ~Mauricio> QUESTION: KVM is installed for default into Ubuntu? [21:33] It's not installed by default, no. It's just a simple apt-get away, though :) [21:33] 21:21:28 < ~johnsgruber> QUESTION: Is there an easy way to tell whether my computer has virtualization extensions? [21:34] An almost sure way to tell is to run this command: [21:34] egrep '^flags.*(vmx|svm)' /proc/cpuinfo === ntinos_19 is now known as ntinos_ [21:34] If that gives you any output, your cpu has the right extensions. It might still need to be enabled in the BIOS, but installing kvm will attempt to run, and should detect if that is the case. [21:35] QUESTION: Regarding the minimal desktop, what about people who are using something like vmware on their desktop to test things? It seems like a lot of people try out Ubuntu this way, and it'd be nice to have it snappier there. [21:35] I'm not sure I understand this use case? Are we still talking about appliances? [21:35] 21:30:31 < alexharrington> QUESTION: Does/will kvm support direct access to specific host hardware - eg a TV tuner [21:36] Not right now, but it will come. [21:36] 21:31:52 < YokoZar> QUESTION: I hate to ask, but any interest in a graphical front-end to vmbuilder? [21:36] I think my money is on the web frontend. [21:37] That way, we can distribute the webified version of it as a vm, and people can run it on their windows systems or fedora systems or whatever and create even more Ubuntu based virtual machines in a snap. [21:38] Now, if you want to try out kvm, what you should do is install a metapackage we've created for it.. [21:38] * soren tries to find it.. [21:39] ubuntu-virt-mgmt [21:39] Start by installing that: sudo apt-get install ubuntu-virt-mgmt [21:40] Oh, and ubuntu-virt-server, too. [21:40] Same deal: sudo apt-get install ubuntu-virt-server [21:40] When that's done, you have virt-manager, which is the graphical frontend for (among other things) kvm. [21:41] You have libvirt, which is a handy wrapper and management library for kvm, Xen, openvz, etc. [21:41] You have kvm which is the core "hypervisor". [21:41] And you have a few other things that comes along with these packages. [21:41] First, you want to add yourself to the libvirtd group. [21:41] The short explanation is that it allows you to do stuff with your network config that you *really* don't want to be without. [21:42] As with any other new group membership, you need to log out and log back in to make it take effect. [21:43] If you don't want to do that right now, that's fine. [21:43] Just do it later. You'll be thankful :) [21:43] Fire up virt-manager, and you should see an overview screen. [21:43] Virt-manager, IIRC, is in the Applications->System menu. [21:44] 21:41:13 < maluta> QUESTION: Xen has any future in newer versions os Ubuntu? [21:44] Xen has finally started to slowly trickle into the upstream vanilla kernels. [21:44] We're obviously not going to actively keep it away from our kernels then :D [21:45] ...so in fact, the server kernel runs as a Xen domU without any modifications. [21:45] Later versions will probably work as a dom0, when that lands upstream. [21:45] I don't know when that's likely to happen, though. [21:46] We'll probably not spend a whole lot of time on Xen going forward, though. We chose to focus on kvm a while ago, and we're quite happy with that decision. [21:46] 21:42:23 < ~toobuntu> QUESTION: With ubuntu-vm-builder, we had to --addpkg foo --addpkg bar, listing every pkg separately. Does the python rewrite accept multiple pkgs in a comma separated list or something a little simpler? [21:47] I believe that's still the same. I've never considered this a problem, really. [21:47] For me, I usually only install one or two extra packages. [21:48] The way I've set up my production environment (I host a bunch of stuff for myself, some friends, etc), is with one virtual machine per service. [21:48] I have a virtual machine for mysql, one for apache, one for kannel, one for nagios, one for postfix/dovecot.. [21:48] Feel free to file a bug about this, though. I've never thought it was an issue :) [21:49] 21:43:37 < Zizou> QUESTION: Are those packages available in Debian? maybe with other names... [21:49] Those metapackages are not in Debian, no. [21:49] In Debian, just install libvirt-bin, virt-manager, and kvm, and you should be well on your way. [21:49] 21:45:25 < ~foob9> QUESTION: is there a howto available? [21:50] For getting virt-manager and all that running? Probably :) I don't know to be honest. [21:50] Ok, you have virt-manager installed and you're looking at the overview. Double click on the line that reads "localhost (system)" or "localhost (user)" depending on whether you're a member of libvirtd. [21:51] You're now connected to the hypervisor. [21:51] Click the "New" button, and you're on your way to setting up your first virtual machine. [21:52] It should be mostly self explanatory. If not, don't hesitate to bother us in the #ubuntu-virt channel. [21:52] Or in #ubuntu if you're more comfortable there. [21:53] Apart from the obvious "I don't have a computer that supports kvm" what other reasons do you have for not using virtualisation? [21:53] Is it too difficult to get started? [21:53] Do you not see the point? [21:53] I'm very interested in this, so that we can try to fix the problems. [21:54] 21:54:04 < alexharrington> soren: Initially I was mainly concerned about disk io throughput - especially in things like proxy servers and mail servers. KVM is working well though. [21:54] I/O performance has been a concern for many people. [21:55] Network performance in KVM is IMO stellar. [21:55] I've moved files from my desktop machine to a virtual machine running on it at 180 MB/s. [21:55] That's not too shabby :) [21:56] Disk I/O has been massively improved in recent kvm versions. [21:56] You'll see a dramatic performance increase in kvm in Jaunty. [21:56] 21:56:04 < alexharrington> QUESTION: any news on paravirtualised block devices - I know there is a paravirt network driver now (which we're using on a couple of guests). [21:57] The 180 MB/s was with paravirt network drivers. They're called virtio devices in kvm. [21:57] If you use virt-manager to install Ubuntu (and choose Ubuntu in the OS selection drop down), you'll have virtio networking by default. [21:58] There's also virtio block devices, but it doesn't make a lot of difference right now, actually. [21:58] (2 minutes left!) [21:58] ...because that's not where the bottleneck is. [21:58] ..but with recent kvm versions, it's likely to make a lot of difference, if you have a storage backend that can keep up. [21:58] Final questions? [22:00] Alright, let's call it a day then. Thanks for showing up, everyone. === jcastro changed the topic of #ubuntu-classroom to: Current Session: Ubuntu Training | Welcome to Openweek, questions in #ubuntu-classroom-chat please || Session details here: https://wiki.ubuntu.com/UbuntuOpenWeek [22:00] thanks soren! [22:00] And now, Ubuntu training with billycina and dinda! [22:00] take it away ladies! [22:01] hi folks [22:01] dinda is just finishing up in another community meeting so I'll kick off [22:02] so, for those of you who don't know me yet, I am Billy Cina and I look after the training programme for Canonical [22:02] in this session we plan on basically discussing: [22:02] 1. What Canonical does on the training front [22:03] 2. What work we do with the community and how [22:03] happy to take questions throughout, so stop me whenever === lemmAway is now known as lemma [22:03] so, Canonical wise [22:03] I'm here too - Belinda [22:04] if you can sum up the goal of training, it is to make migrating to Ubuntu as easy and seemless as possible for every user on every level [22:04] so that means: [22:04] a. Creating courses for different users and purposes [22:05] we currently have a system administrator course - Ubutnu Cetified professional [22:05] The Ubuntu desktop course (which we will go into more detail on during the 2nd half of the meeting) === dindatx is now known as dinda [22:06] which is really a comfort blanket course for MS users migrating to ubuntu for the first time [22:06] we also have partner enablement courses [22:06] so when we get partners on board who need to work with and customise Ubuntu, teaching them how to do that [22:07] supporting Ubuntu - teaching service providers what to do on the common trouble shooting elements etc [22:07] where to find more information - wiki, forums, escalation, reporting bugs on LP [22:08] for many of our partners, working with an OS Linux distribution is completely new playing field so we really need to get down to the basics [22:08] we are also currently evaluating the need for a server based course [22:08] So we have the courses [22:09] now in parallel, we also seek training partners who can actually teach those courses [22:09] they have to meet certain criteria [22:10] in order to ensure that there is a good match between their instruction skills, canonical expectations and student requirements [22:10] so whilst we speak to scores of partners, only a handful actually get selected to come on board the programme [22:11] so, big picture - that's what the training team at canonical does [22:11] does anyone have questions before we get onto the community element of training? [22:11] http://webapps.ubuntu.com/partners/training/ our current list of partners [22:11] going [22:12] thanks dinda [22:12] Question: Are you planning to translate de guides available on the wiki? [22:12] so, there are no plans at the moment [22:12] but [22:13] the guides and the source material are there for anyone to get the ball rolling, in any language [22:13] Question from xanders21c: which criteria do you use to select training partners? [22:13] well, a few of the basic ones: [22:14] 1. the company must be already an established linux training house [22:14] we look for partners who already have a presence and understanding of linux training requirements [22:15] 2. Must have own classrooms and preferably facilities in a number of cities [22:15] 3. we discourage too many partners in any given geographic area [22:15] 4. Goes with point 1 - quality of instructors [22:16] all must be certified Linux instructors and must become LPIC1 cetified and attend the ubuntu Train the Trainer event [22:17] 5. Types of customers - corporations, governments etc [22:17] so quite a few elements need to match [22:17] any other Qs? [22:17] not yet, continue on [22:18] ok, Belinda Lopez do you want to lead on the community work? [22:18] would you believe my doc with all that info jst went "poof"? [22:18] but I can wing it [22:18] yeah :) [22:19] so the second part: 2. What work we do with the community and how [22:19] https://wiki.ubuntu.com/Training [22:19] is the main wiki page where you can find all our information [22:20] we're a team of really enthusiastic contributors and we'd love for more folks to come join us [22:20] we just started the process of assigning tasks to assist in updating the Ubuntu Desktop course [22:21] and we hang out in #ubuntu-training. . . [22:21] the goal of the Training Team is to help develop quality training materials. . . [22:22] that are then released under to the community to use to help spread the Ubuntu love [22:22] many of the Loco teams have been using the Desktop course materials at their meetings [22:22] and several groups have begun to translate it as well [22:23] you don't need any technical expertise to help either [22:24] right now we're working on updating the course from the 7.10 version to the 8.04 LTS version [22:24] and hope to release it by December 15 [22:25] anything else I should mention, billycina? [22:25] how can people get involved? [22:26] what can they do right now? [22:26] what tools are we using to create the course? [22:26] https://wiki.ubuntu.com/Training/GettingInvolved [22:26] right now we are looking for new contirbutors to help us update the course [22:27] this entails, adopting a chapter, 5, 6 & part of 4 remain. . . [22:27] then comparing that info to the existing course and suggesting changes on the wiki [22:27] so for that step, nothing more than downloading a pdf file and having 8.04 running [22:28] but the deadline for that part is one week from today [22:28] the next step is a bit more complex but we've have some great folks willing to mentor [22:29] once the changes have been identified, then they need to be edited and be Docbook friendly [22:29] which can now be done in Open Office v. 3 [22:29] next week in #ubuntu-training, dougierichardson will lead a session on just how to do that [22:30] once a "patch" or change is made, then the patch is uploaded into Launchpad/Bzr [22:31] where we (me, other team members with commit rights) then review and merge the changes into the final courseware [22:31] so we're trying to make it as easy as possible for new users to contribute [22:31] make sense? [22:32] does to me ;) [22:32] we have a question [22:32] * dinda gives a big sigh of relief o that [22:32] shoot [22:32] probably for me to answer [22:32] FreeBuzzbee: QUESTION: References to both Training and Exams for certification. Are there fees associated with either or both? If amount too small then concerned that this might become too infrequent for the demand. [22:33] FreeBuzzbee: not sure about the 2nd half of the question but [22:33] yes, the Ubuntu Certified Professional and Desktop course are paid for [22:33] the UCP also has 3 exams - LPI 101, 102 and Ubuntu 199 [22:33] Ubuntu bug 199 in baz ""baz export" doesn't add to the revision library" [Medium,New] https://launchpad.net/bugs/199 [22:34] How much the courses cost depends on the country where they are studied [22:34] ignore the bot [22:34] the exam prices are the same globally [22:34] $155 for the LPI exams, $100 for ubuntu 199 [22:35] FreeBuzzbee: does that answer your question? [22:35] it must :) [22:35] ok, anyone have any other questions before we sign off for the day? [22:36] ok, great [22:36] so thanks everyone [22:36] it's an easy crowd this time around [22:36] :) [22:36] :) [22:37] we have another question [22:37] Belinda Lopez: hoping you know the answer to this [22:37] DoruHush: QUESTION: font encoding support for other languages translations of the Ubuntu Desktop course. How can be made. What fonts should be used and how. [22:37] DoruHush: QUESTION: font encoding support for other languages translations of the Ubuntu Desktop course. How can be made. What fonts should be used and how. [22:37] lol [22:37] yes, we're still working this one [22:38] we're looking for suggestion on what steps we might take in the development process to help [22:38] if I recall, the issue came down to non UTF-8 character sets [22:39] FreeBuzzbee: Is the course a physically attended at hotel or such? Or can most or all be subscribed online? Of the links to courses, other than seeing a time and place for it there was not clue to costs involved. Thanks for the testing costs. [22:39] FreeBuzzbee: the courses are held at training partner facilities in countries / cities where we have partners [22:40] FreeBuzzbee: please contact your nearest partner and they will quote you [22:40] there are also online courses [22:40] * billycina goes to find the link [22:40] http://www.ubuntu.com/training/e-learning [22:41] which also offers two free lessons - which is way cool [22:41] gscholz: QUESTION: Where are the translations hosted? [22:41] I saw a few on the ubuntu wiki [22:42] DoruHush: https://code.launchpad.net/ubuntu-desktop-course [22:42] thanks DoruHush [22:42] Teams have been doing translations and posting them there on the wiki [22:43] there are also the .po files available in a PPA [22:43] ack,I'd have to look for them, Nicolas Varcel's PPA has them [22:44] ok, anyone else? [22:44] I will post more info on that process on the wiki [22:44] oh [22:44] https://wiki.ubuntu.com/Training/KnowledgeBase [22:44] there is also a community training mailing list [22:44] has the info on How to Translate the course [22:45] which folks can join to receive updates on meetings and activities [22:45] and from henceforth I'll be in #ubuntu-training Wednesdays @ 20:00 UTC to answer questions [22:46] thank you Belinda Lopez [22:46] thanks all [22:46] thanks everyone! [22:47] dinda: billycina: thank you for the session [22:47] you are welcome - any time ;) === Silvy is now known as Fierelin