| *** povbot has joined #maemo-meeting | 15:24 | |
| *** mgedmin has joined #maemo-meeting | 15:26 | |
| mgedmin | povbot, are you here too? | 15:26 |
|---|---|---|
| povbot | mgedmin: Error: "are" is not a valid command. | 15:26 |
| mgedmin | okay, I'm learning new things about the channel list in /whois: they're incomplete | 15:26 |
| frals | if the channel is +s it wont show in whois unless you share the channel with the person | 15:29 |
| zaheerm | :) | 15:32 |
| mgedmin | irc is hard, let's go shopping | 15:33 |
| GeneralAntilles | Shopping! | 15:34 |
| *** Lorthirk has joined #maemo-meeting | 15:35 | |
| *** Lorthirk has quit IRC | 15:41 | |
| *** Lorthirk has joined #maemo-meeting | 15:56 | |
| *** jeremiah has joined #maemo-meeting | 15:57 | |
| jeremiah | hello :) | 15:57 |
| *** GeneralAntilles has left #maemo-meeting | 16:11 | |
| *** GeneralAntilles has joined #maemo-meeting | 16:11 | |
| *** lma has joined #maemo-meeting | 16:12 | |
| lma | Hi all | 16:12 |
| Stskeeps | 'lo | 16:12 |
| *** rmoravcik has left #maemo-meeting | 16:19 | |
| *** qwerty12 has joined #maemo-meeting | 16:24 | |
| *** rmoravcik has joined #maemo-meeting | 16:24 | |
| jeremiah | hiya | 16:29 |
| jeremiah | Gonna make a sandwhich | 16:29 |
| zaheerm | same | 16:32 |
| *** kurtan has joined #maemo-meeting | 16:33 | |
| andre__ | sudo make sandwich? | 16:33 |
| jeremiah | indeed | 16:34 |
| * VDVsx steals jeremiah's sandwich , muuhuhuh | 16:34 | |
| jeremiah | hey! | 16:34 |
| * mgedmin lurks | 16:34 | |
| *** RST38h has joined #maemo-meeting | 16:34 | |
| VDVsx | any midgard hacker around ? | 16:35 |
| *** Jaffa has joined #maemo-meeting | 16:35 | |
| Jaffa | Watching, not participating :-( | 16:35 |
| jeremiah | Jaffa! | 16:35 |
| jeremiah | You made it! | 16:35 |
| *** denver has joined #maemo-meeting | 16:35 | |
| VDVsx | need some help enabling the sponsorship fields in the Maemo BCN event ^ | 16:35 |
| jeremiah | So | 16:37 |
| jeremiah | I guess we'll get started | 16:37 |
| jeremiah | http://wiki.maemo.org/QAMeeting | 16:37 |
| jeremiah | The agenda is above. | 16:37 |
| jeremiah | I have tried to bring in all the things that have been discussed on the lists. | 16:37 |
| *** VDVsx changes topic to "Maemo meeting channel | Q&A and Extras-testing discussion - November 10, 14:30 UTC | Agenda: http://wiki.maemo.org/QAMeeting" | 16:38 | |
| VDVsx | :) | 16:38 |
| jeremiah | But I may have forgotten something so feel free to add. | 16:38 |
| jeremiah | Hopefully, we all agree that the overall mission, i.e. Quality Assurance, is good | 16:38 |
| *** suihkulokki has joined #maemo-meeting | 16:39 | |
| *** viq has joined #maemo-meeting | 16:39 | |
| jeremiah | But maybe there are some things we can do to fix the process? | 16:39 |
| jeremiah | If we can agree on what those things are, we can have a plan of action | 16:39 |
| jeremiah | Firstly, I think everyone agrees that keeping karma is good? | 16:39 |
| qwerty12 | Aye. | 16:40 |
| *** danielwilms has joined #maemo-meeting | 16:40 | |
| VDVsx | jeremiah, for testers ? | 16:40 |
| VDVsx | or apps ? | 16:40 |
| RST38h | jeremiah: Should I repeat the same suggestions I made to you in our previous discussion? | 16:40 |
| jeremiah | I was thinking more for devs and apps | 16:40 |
| lma | Package karma across versions? | 16:40 |
| jeremiah | RST38h: Yes please. | 16:40 |
| RST38h | So that everybody can discuss them | 16:40 |
| RST38h | Ok, here they go: | 16:40 |
| Jaffa | lma: yes; package karma across versions | 16:40 |
| jeremiah | lma: Yeah, package karma across minor changes and/or version | 16:40 |
| RST38h | 1. Temporarily lower the vote threshold from 10 to 5 votes | 16:41 |
| jeremiah | Yeah, I think this is a good idea too. | 16:41 |
| RST38h | 2. Possibly set threshold for existing Extras apps updates to 2-3 votes | 16:41 |
| jeremiah | I have heard others mentions that. | 16:41 |
| RST38h | (note on 1: I consider it a temporary measure, let us lower and watch how it goes) | 16:42 |
| jeremiah | RST38h: Well, I think keeping karma will solve 2. | 16:42 |
| *** fiferboy has joined #maemo-meeting | 16:42 | |
| lma | I have mixed feelings about package karma... More specifically, about negative karma | 16:42 |
| *** sjgadsby has joined #maemo-meeting | 16:42 | |
| RST38h | jeremiah: yea | 16:42 |
| RST38h | 3. Change requirements for testers from the strict list Nokia uses for its internal testing to "Application wortks for me, does not waste battery, does not reboot the system, takes less than nMB of . space etc) | 16:43 |
| zaheerm | ok back | 16:43 |
| jeremiah | lma: Well, how about we say that the negative karma gets deleted while positive gets kept? | 16:43 |
| RST38h | In other words, make testing requirements sane | 16:43 |
| lma | I think negative karma at least should be reset for new versions, but also counting 30 "not optified" thumbs down is a broken way to do it | 16:43 |
| RST38h | 4. Fix the bug that allows people to vote multiple times | 16:43 |
| VDVsx | RST38h, jeremiah , I'm more for a checklist process, if for example no one checks if the package is optified and it has 5 votes, it can be promote, something along these lines :) | 16:43 |
| Jaffa | RST38h: The list isn't a strict *Nokia* list at the moment. | 16:43 |
| VDVsx | *can't be promoted | 16:44 |
| RST38h | Jaffa: I know, but it is still pretty scary for volunteers | 16:44 |
| RST38h | 5. Streamline voting UI and make it faster | 16:44 |
| jeremiah | So a community checklist which means that testers just run through the list and give thumbs up if they all pass? | 16:44 |
| jeremiah | RST38h: 5. will be taken care of soon. | 16:44 |
| RST38h | If you want to take my PackRat code that shows repository contents, by all means, take it | 16:44 |
| RST38h | It has been written for a different purpose but it will work for a simple list of repo packages | 16:45 |
| RST38h | With votes | 16:45 |
| jeremiah | I'll look at it - its called PackRat? | 16:45 |
| RST38h | jeremiah; I will msg you the link | 16:45 |
| jeremiah | cool | 16:45 |
| RST38h | Ok, I am done with the list of suggestions | 16:45 |
| jeremiah | Great, I've added them to the agenda and we'll zip through the agenda now. | 16:46 |
| jeremiah | We agree that there needs to be user interface streamlining so that the whole process is faster | 16:46 |
| jeremiah | Shall I write that as an actionable item? | 16:47 |
| Jaffa | jeremiah: Needs an actual /design/ though. "Make it better" isn't actionable ;-) | 16:47 |
| jeremiah | We also think that a tester should follow the checklist closely, and not be too subjective. | 16:47 |
| jeremiah | Jaffa: True. :) | 16:47 |
| jeremiah | 'Faster' is kinda actionable. | 16:47 |
| Jaffa | Yup. Stock answer would be "server move is coming" :-/ | 16:48 |
| jeremiah | But it will get faster once we have a dedicated machine for it. | 16:48 |
| lma | Yes, the downloads section is available for user ratings. This is QA only. | 16:48 |
| jeremiah | yeah. | 16:48 |
| RST38h | Jaffa,jeremiah: I can come up with the list of exact things that need to be improved | 16:48 |
| jeremiah | But I also think that there are too many packages on the package page | 16:48 |
| jeremiah | Just a whole list of extraneous info | 16:48 |
| RST38h | Jaffa,jeremiah: Please, consider simply basing work off PackRat UI, as it has been specifically optimized for browsing package lists | 16:48 |
| Jaffa | RST38h: Again, those are bugs; there needs to be a list of suggested changes | 16:48 |
| Jaffa | RST38h: Personally, I found the PackRat UI not *much* better than the existing packages one. | 16:49 |
| RST38h | Jaffa: Yes, I can do such a list | 16:49 |
| RST38h | Jaffa: The existing packages one is not really a UI, just a list of package names with vote counts | 16:49 |
| RST38h | Jaffa: In order to see package details you have to click on the name and wait for it to load | 16:49 |
| jeremiah | RST38h: Don't forget that much of the interface code has hooks into the build system too | 16:49 |
| jeremiah | So that there is a lot more going on than just listing repos | 16:50 |
| RST38h | jeremiah: You only need to plug into the package list taken from the repo | 16:50 |
| jeremiah | It has to move packages and dependencies too | 16:50 |
| RST38h | jeremiah: The rest is optional | 16:50 |
| Jaffa | RST38h: I don't disagree, but pushing your own code as a panacea is disingenuous. | 16:50 |
| RST38h | Jaffa: Yea, I know :( | 16:50 |
| RST38h | Jaffa: In fact, not insisting on using the code verbatim, just use the ideas at least | 16:51 |
| jeremiah | Yeah, we have to respect the wishes of the code author somewhat. | 16:51 |
| RST38h | jeremiah: Actually you do not have to, even if you just take the general ideas, I will be happy already :) | 16:51 |
| jeremiah | One of the reasons the interface is slow is that it calculates dependencies. | 16:51 |
| RST38h | jeremiah: But dependencies are already in the package list | 16:52 |
| jeremiah | But you need to know those dependencies to promote. | 16:52 |
| Jaffa | I think it could also use the hand of a designer to make sure that any useful, but possibly extranseous, information is hidden and the most important is emphasised. | 16:52 |
| jeremiah | Jaffa: Yes. | 16:52 |
| qwerty12 | I would suggest writing comments - when thumbing down an item - are made compulsory. Thumbing down is all well and good, but if the maintainer has no idea what to fix, it does not help. | 16:52 |
| jeremiah | So can we say that it needs a designer's UX love? | 16:52 |
| RST38h | qwerty: +1 | 16:52 |
| jeremiah | qwerty12: Good suggesstion | 16:53 |
| RST38h | jeremiah: It needs a design with common sense in mind | 16:53 |
| jeremiah | "Thumbs down needs a comment." | 16:53 |
| lma | qwerty12: I would change that to thumbs down needs a bug report | 16:53 |
| RST38h | jeremiah: I.e. the top list should present the maximal amount of useful information you can present without making it unreadable | 16:53 |
| Jaffa | lma: How do you enforce that? | 16:53 |
| RST38h | jeremiah: All the other information goes into collapsible sections, Google Reader style | 16:53 |
| jeremiah | lma: Seems unneccessary if we are saying that a tester has to follow the checklist | 16:54 |
| * Jaffa really doesn't want to mandate bugs.maemo.org for every tiny little application | 16:54 | |
| jeremiah | RST38h: Collapseable sections - good idea. | 16:54 |
| lma | jaffa: for bugzilla/garage it could be automated, for external bugtrackers maybe human moderation? | 16:54 |
| jeremiah | lma: Complicated | 16:54 |
| RST38h | jeremiah: And the list is displayed solely based on the Packages.gz file | 16:54 |
| RST38h | jeremiah: All the dependencies are listed there already | 16:54 |
| * suihkulokki wants to be able thumb up/down from application manager | 16:54 | |
| RST38h | suihkulokki: This requires changes to HAM = not gonna happen soon =( | 16:55 |
| jeremiah | suihkulokki: I'll add that as a 'wishlist' item since that might be really useful. | 16:55 |
| suihkulokki | I do not want to navigate the site to find the right package/version combo | 16:55 |
| RST38h | suihkulokki; But we can do it the opposite way: provide easy way to install packages straight from the package list | 16:55 |
| RST38h | suihkulokki: With autogenerated .install files, just like PackRat does | 16:55 |
| qwerty12 | I doubt that's going to happen anytime soon (we are, after all, talking about Hildon Application Manager) but maybe danielwilms can be bribed to add that functionality to his app? | 16:55 |
| suihkulokki | RST38h: otoh ham is free sw so someone can implement it | 16:55 |
| RST38h | suihkulokki: correct. but nokia has to include your changes into fw updates | 16:56 |
| suihkulokki | I can promise to kick people to make it happen =) | 16:56 |
| RST38h | suihkulokki: There is a guy implementing alternative HAM in QT. Talk to him! | 16:57 |
| RST38h | suihkulokki: Actually the guy is indeed danielwilms :) | 16:57 |
| jeremiah | heh | 16:57 |
| Jaffa | RST38h: The danger with install files in maemo.org/packages/ is that it makes access to apps in -testing perhaps too easy for a user who doesn't know the difference between Maemo Select, maemo.org/downloads and "Maemo Extras" | 16:58 |
| qwerty12 | RST38h: danielwilms is not writing an alternative HAM; just an app to let you browse the apps on maemo.org. Choosing to install something will invoke Hildon Application Manager. | 16:58 |
| jeremiah | Jaffa: Oh, that is a good point. | 16:58 |
| danielwilms | qwerty12 exactly ;) | 16:58 |
| RST38h | qwerty: A bit different implementation, same purpose :) | 16:59 |
| Jaffa | RST38h: There's also an argument that installing through HAM 9checking the section, the icon, the version and the description) are all part of the QA process! | 16:59 |
| RST38h | Jaffa: Well, even current packages interface lets you install stuff I think | 16:59 |
| Stskeeps | some status area icon that asks you once in a while how your experience with X package was (after running it)? :P | 16:59 |
| RST38h | Jaffa: As for installing from the top list, we can provide a JS warning dialog box | 16:59 |
| suihkulokki | hrmh, the app manager has no proper "installed apps" list, so my idea doesnt work | 16:59 |
| Jaffa | RST38h: The current packages interface does not provide .install files for -testing | 17:00 |
| RST38h | Jaffa: If you really like to scare off the chickens, we can also change the browserbackground to red while showing dialog box | 17:00 |
| RST38h | Jaffa: Sorry, my mistake | 17:00 |
| RST38h | Jaffa: Only links to .debs | 17:00 |
| lma | has anyone tried packagerate? | 17:00 |
| Jaffa | lma: it feeds into maemo.org/downloads. Tried it once, it hadn't been Fremantlised | 17:01 |
| RST38h | Jaffa: Anyway my point is that there is plenty of ways to make the user understand the risks when he clicks on that link | 17:01 |
| Jaffa | RST38h: Indeed. Perhaps even -testing install files should be limited to logged in users. | 17:01 |
| RST38h | Jaffa: The ones who can vote, indeed | 17:01 |
| jeremiah | That might be a good procedure | 17:01 |
| RST38h | Can't vote => no .install for you | 17:02 |
| Jaffa | RST38h: with a "...don't forget to rate against the QA checklist after testing" JS popup in-page. | 17:02 |
| RST38h | Jaffa: True | 17:02 |
| jeremiah | So, the tester logs in, installs via .install, and then votes? | 17:02 |
| jeremiah | Otherwise there is no .install file visible? | 17:02 |
| jeremiah | (If they don't log in?) | 17:03 |
| qwerty12 | With a reminder to disable extras-devel (if enabled) | 17:03 |
| Jaffa | jeremiah: Well: tester logs in, reviews app metadata on web, installs via .install, tests and then votes ;-) | 17:03 |
| *** kurtan has left #maemo-meeting | 17:03 | |
| Jaffa | jeremiah: indeed | 17:03 |
| jeremiah | excellent. | 17:03 |
| RST38h | Here is the approximate warning text: "Please keep in mind that you are downloading and installing beta software that has not yet been fully tested. Press Cancel if you are afraid of the risks. If you do install this package, make sure you vote afor it at <...>" | 17:03 |
| jeremiah | I'll add that to our wiki. | 17:03 |
| jeremiah | so, we have a clear idea of what we want from testers: | 17:04 |
| jeremiah | log in, follow the checklist, install, test, vote and comment | 17:04 |
| jeremiah | That is roughly the algorithm they should follow | 17:05 |
| jeremiah | With little leeway to deviate from the checklist | 17:05 |
| jeremiah | So that developers know what to expect from testers. | 17:05 |
| RST38h | Well you do not want to make it look like a bootcamp | 17:05 |
| Jaffa | RST38h: That's where a friendly UX guy comes in ;-) | 17:05 |
| RST38h | So, setting some strict procedure with time limits is wrong | 17:05 |
| jeremiah | Well, we do want to strengthen the criteria no? | 17:06 |
| Jaffa | Where did timelimits come from? | 17:06 |
| RST38h | Jaffa: Judging from Maemo5, UX guys are anything but friedly :) | 17:06 |
| jeremiah | No timelimits. | 17:06 |
| RST38h | Ok | 17:06 |
| jeremiah | And the UX guys are good. Plus we have some good maemo.org UX guys | 17:06 |
| RST38h | Jaffa: I would say if we can figure out how to make procedure friendly without UX guy we should do it :) | 17:06 |
| VDVsx | well, it should be also easy to vote/proceed, if you've the app already installed, please don't forget that | 17:07 |
| RST38h | Another thing: make testing a privilege so that people WANT to be testing stuff | 17:07 |
| * Jaffa nods | 17:07 | |
| jeremiah | yes | 17:07 |
| jeremiah | No - that is too hairy | 17:07 |
| lcuk | in an ideal world voting should be running down the checklist actively checking the elements which pass qc | 17:07 |
| RST38h | Well...true | 17:07 |
| jeremiah | Let whoever wants to participate join in | 17:07 |
| qwerty12 | Oh, out of curiosity, what's the stance on voting up your own apps? ATM, I see people voting their stuff down as a hint not to vote for it, but, really, they should be allowed to remove it from the queue | 17:07 |
| RST38h | lcuk: not gonna happen with volunteers | 17:08 |
| jeremiah | You should be able to vote up or down as you like on your own app | 17:08 |
| lcuk | "In an ideal world" | 17:08 |
| RST38h | qwerty: Personally I always vote my apps up | 17:08 |
| RST38h | qwerty: And see no problem with that: I tested it, it worked, I voted it up | 17:08 |
| jeremiah | I think people _should_ vote for their own apps | 17:08 |
| jeremiah | Yup | 17:08 |
| VDVsx | and you should be also able to remove your own app from the queue | 17:08 |
| lma | Hold on, why can't volunteers follow a simple checklist? | 17:08 |
| Jaffa | I also think the maintainer should only be able to vote their app *down* to remove it | 17:08 |
| RST38h | Yes, removal should be possible | 17:08 |
| Jaffa | jeremiah: I strongly disagree | 17:08 |
| qwerty12 | Jaffa: +1 | 17:08 |
| RST38h | In case someone fucks up | 17:09 |
| jeremiah | Really? Why? | 17:09 |
| Jaffa | I thought my app was suitable for testing by promiting to -testing. That's an implicit +! | 17:09 |
| Jaffa | err, +1 | 17:09 |
| jeremiah | okay | 17:09 |
| RST38h | Jaffa: Not necessarily | 17:09 |
| VDVsx | agreed | 17:09 |
| jeremiah | that is kinda true | 17:09 |
| RST38h | Jaffa: Want a scenario? | 17:09 |
| jeremiah | No corner cases! | 17:09 |
| Jaffa | If we let some maintainers vote their own apps up, we've reduced the threshold to one less | 17:09 |
| Jaffa | RST38h: Go on | 17:09 |
| RST38h | There are corner cases | 17:09 |
| qwerty12 | It's your own app, so voting up screams of bias | 17:09 |
| *** yerga has joined #maemo-meeting | 17:10 | |
| jeremiah | Well you should possess that bias | 17:10 |
| Jaffa | qwerty12: +1 | 17:10 |
| RST38h | Let us say I have an app that I think is ready for consumptions by power users but not yet ready for the new Extras | 17:10 |
| jeremiah | But maybe uploading it is enough to say "I like this!" | 17:10 |
| RST38h | Keeping it in -devel is wrong because -devel is seriously dangerous | 17:10 |
| RST38h | So I move it to -testing for the power users to install (new -testing is like the old extras, right?) | 17:10 |
| RST38h | But I am not voting for it because I do not think it is ready for the general populace that needs to know what button to press | 17:11 |
| jeremiah | I think if we lower the threshold to 5 - then it is reasonable to assume that you need five votes from someone _other_ than you! | 17:11 |
| jeremiah | So, let's go with Jaffas proposal not to allow devs voting on their own apps | 17:11 |
| Jaffa | RST38h: No, new -testing is a staging post for new Extras. It's only similarity to old Extras is the fact that you can upload pretty much anything | 17:11 |
| RST38h | jeremiah: Maybe, but I really do not think we should add some special checks for voter being the author | 17:11 |
| Jaffa | RST38h: Then you can choose not to promote it. | 17:12 |
| RST38h | Jaffa: Functionally, -testing is like the old extras really | 17:12 |
| jeremiah | Well, we can at least ask them not to | 17:12 |
| Jaffa | RST38h: If you *don't* thumbs up your own app, are you saying that no-one else should bother testing it as you only want it in a repo | 17:12 |
| jeremiah | And their vote appears in the interface. | 17:12 |
| RST38h | Jaffa: Do not forget that I can also thumb it down | 17:12 |
| Jaffa | RST38h: Yes, and I think an author thumbing their app down should remove it from the QA queue and/or -testing | 17:12 |
| jeremiah | This might be a good time to bring in PPAs | 17:13 |
| qwerty12 | I see no problem with thumbing down your own app. It's more of a hint that you wish to have the program removed from the queue | 17:13 |
| RST38h | Dunno, it looks like useful feature to me, people use both up and down | 17:13 |
| jeremiah | Do we want or need personal repos? | 17:13 |
| Jaffa | Indeed | 17:13 |
| Jaffa | jeremiah: I don't see any need (yet) | 17:13 |
| jeremiah | Nor do I. | 17:13 |
| RST38h | jeremiah: we do not I think | 17:13 |
| VDVsx | me neither | 17:13 |
| jeremiah | But it is something we should keep an eye one | 17:13 |
| jeremiah | on | 17:13 |
| zaheerm | i think ppas are a bad idea | 17:13 |
| *** Lorthirk has quit IRC | 17:13 | |
| jeremiah | Okay good: Nix to PPAs | 17:14 |
| zaheerm | we have a playground for stuff, extras-devel | 17:14 |
| Jaffa | I'd argue (strongly) on: app authors *prevented* from thumbing up own app (I don't think RST38h will listen to a rule which says you shouldn't ;-)) and that a maintainer thumbing it down removes it from the QA list | 17:14 |
| zaheerm | ppas will cause lots of conflicts | 17:14 |
| RST38h | Jaffa: Naah, the rule without enforcement is useless | 17:14 |
| qwerty12 | Jaffa: Agreed | 17:14 |
| VDVsx | +1 | 17:14 |
| RST38h | Jaffa: But I would still vote against any kind of enforcement or special rules on this | 17:15 |
| jeremiah | No I think the rule without enforcement works because you see if the dev has voted or not | 17:15 |
| RST38h | Keep it simple! | 17:15 |
| qwerty12 | Which brings me to another point: Is it possible to switch votes? I see people hitting the wrong thumb... | 17:15 |
| Jaffa | RST38h: Why? | 17:15 |
| VDVsx | qwerty12, yes | 17:15 |
| Jaffa | qwerty12: It is an obvious missing feature which should be implemented | 17:15 |
| VDVsx | we need that feature | 17:15 |
| RST38h | Jaffa: Makes things too complicated. You probably know how I feel about extra policies | 17:15 |
| Jaffa | RST38h: Makes it complicated for whom? | 17:15 |
| RST38h | Jaffa: For everybody | 17:15 |
| Jaffa | RST38h: Eh? I don't understand. It's an extra line of PHP and a sentence in a doc: "you cannot thumb up your own app" | 17:16 |
| RST38h | The more policies you add, the more things devs have to keep in mind | 17:16 |
| RST38h | Yes, it is an extra line in PHP, but it is also an extra policy for people to keep in mind | 17:16 |
| Jaffa | jeremiah: But what's the sanction against a developer who ignores the suggestion? | 17:16 |
| jeremiah | So people should be able to switch their vote - that should be too hard to do. | 17:16 |
| RST38h | jeremiah: +1 on that | 17:16 |
| *** vre has joined #maemo-meeting | 17:17 | |
| jeremiah | Jaffa: Public humiliation | 17:17 |
| jeremiah | :) | 17:17 |
| lcuk | yikes | 17:17 |
| lcuk | switching votes brings complications | 17:17 |
| RST38h | if you are not enforcing something, do not require it | 17:17 |
| Jaffa | RST38h: Who is keeping it in mind? Maintainers are prevented from clicking a button by the fact the button is not present or disabled (possibly with a message) | 17:17 |
| lcuk | "i didnt mean to vote down my own app, but i cant vote it up" | 17:17 |
| lcuk | because its removed/disabled | 17:17 |
| VDVsx | lcuk, only for the maintainer ;) | 17:17 |
| Jaffa | lcuk: If voting down your own app => removal, then that's an argument for not having voting up (no mistaken button presses) | 17:17 |
| lcuk | yes, but still | 17:17 |
| VDVsx | mistakes happens | 17:18 |
| jeremiah | So voting should be disabled for the deb/maintainer | 17:18 |
| jeremiah | Thus not allowing bias or mistakes. :P | 17:18 |
| RST38h | more mistakes happen than anyone would like to acknowledge, even | 17:18 |
| lcuk | any kind of voting for maintained | 17:18 |
| VDVsx | IMO, yes | 17:18 |
| lcuk | maintainer | 17:18 |
| RST38h | so yes, there should be a way to change one's vote | 17:18 |
| Jaffa | jeremiah: Maybe all voting disabled, and a separate buytton for "Demote this package" (like "Promote") | 17:18 |
| lcuk | it should be an explicit removal option | 17:18 |
| lcuk | +1 | 17:18 |
| RST38h | jeremiah: non-maintainers may also make mistakes | 17:18 |
| VDVsx | Jaffa, +1 | 17:18 |
| jeremiah | I much prefer terms like 'promote' and 'demote' | 17:18 |
| qwerty12 | Jaffa & lcuk: Agreed | 17:18 |
| RST38h | jeremiah: Or maybe the tester has not understood sometihing, voted uapp down, then developer explained it in the comments | 17:19 |
| jeremiah | And the buttons should be separated so you cannot slip. | 17:19 |
| RST38h | jeremiah: how will this tester change his vote? | 17:19 |
| jeremiah | Maybe the tester has to check each check box in a list and then it gets karma automatically? | 17:19 |
| VDVsx | click in the unselected thumb ? | 17:20 |
| jeremiah | then you're just checking checkbox entries | 17:20 |
| RST38h | jeremiah: not really | 17:20 |
| RST38h | jeremiah: you are complicating it | 17:20 |
| jeremiah | Easier actually | 17:20 |
| jeremiah | but whatever | 17:20 |
| RST38h | Even the current model is not being widely used | 17:20 |
| jeremiah | We can skip that | 17:20 |
| RST38h | Add more chckboxes - and it gets more complicated | 17:20 |
| jeremiah | I dunno - it is pretty well used. | 17:20 |
| *** lma has quit IRC | 17:20 | |
| jeremiah | in any case, we'll move on | 17:21 |
| RST38h | not judging from the vote counts :( | 17:21 |
| jeremiah | What should we do about cli apps? | 17:21 |
| jeremiah | Or libs? | 17:21 |
| *** lma has joined #maemo-meeting | 17:21 | |
| jeremiah | Are they a separate class? | 17:21 |
| zaheerm | libs should probably get promoted when an app that deps on them gets promoted | 17:21 |
| RST38h | separate category? | 17:21 |
| qwerty12 | libs shouldn't really be a question. They shouldn't even be in user/ | 17:21 |
| Jaffa | jeremiah: I'm very strongly against libs being promoted without an app using them | 17:21 |
| RST38h | libs should not be shown to the user | 17:21 |
| jeremiah | zaheerm: What if you are just uploading a lib | 17:21 |
| RST38h | cmd line apps should have an extra category | 17:22 |
| jeremiah | Okay, if you upload a lib, you need to upload an app that uses taht too? | 17:22 |
| qwerty12 | Let libs be in another section with an app in user/ depending on them | 17:22 |
| Jaffa | So any argument about randomg bindings, or new libraries are entirely unpersuasive to me | 17:22 |
| zaheerm | jeremiah, i did so when uploading gstreamer plugins | 17:22 |
| Jaffa | RST38h: "extra category" => i18n work and shipping an update in a firmware | 17:22 |
| zaheerm | ogg-support for example is not an app but a meta package that deped on others and did a post-install hook | 17:22 |
| RST38h | Jaffa: Ok. Maybe we have a category already that can fit semantically? | 17:22 |
| jeremiah | zaheerm: Aha - what do we do then? A new lib for an existing app? | 17:22 |
| Jaffa | zaheerm: Then ytou should upload a pretty meta-package with an icon and a human-understandable description as to why they should install them. | 17:22 |
| zaheerm | Jaffa, exactly | 17:22 |
| jeremiah | Seems a bit much to ask. | 17:22 |
| zaheerm | not really | 17:23 |
| jeremiah | Well, if you're okay with that, then so am I. :) | 17:23 |
| zaheerm | if you want something promoted that is really useful and will help apps currently installed etc. you should do the meta package | 17:23 |
| Jaffa | jeremiah: If the library is a plugin, then it doesn't need to be a separate package if the user can access the functionality | 17:23 |
| VDVsx | note that auto-promotion of libs can break existent apps | 17:23 |
| zaheerm | VDVsx, good point... | 17:23 |
| jeremiah | that is really problematic. | 17:23 |
| Jaffa | jeremiah: However, that's different from me wanting to put libmycoollibrary (or, libhildonextras) in Extras without an app using it | 17:23 |
| VDVsx | Jaffa, for what do you need the lib there ? | 17:24 |
| Jaffa | I think that's outside the scope of this discussion, isn't it? QA can't help there. Not much can without full blown central, standardised library version | 17:24 |
| zaheerm | VDVsx, i think lib upgrades need to be taken on a case by case basis.... | 17:24 |
| jeremiah | okay. | 17:24 |
| fiferboy | I think Qt is a good example of Jaffa's point too | 17:24 |
| zaheerm | VDVsx, maintainers should know what apps to test etc. | 17:24 |
| Jaffa | VDVsx: That's my point. Without an app, libraries (not plugins or enabling meta-packages) should not be pushed | 17:24 |
| jeremiah | I think we have consense that libs are not something we put through the QA interface that exists now. | 17:25 |
| VDVsx | Jaffa, ok, 100% agreed | 17:25 |
| jeremiah | Or rather, consensus. | 17:25 |
| RST38h | jeremiah; we put them through implicitely | 17:25 |
| jeremiah | RST38h: If they are connected to an app | 17:25 |
| RST38h | jeremiah: When you test an app you test its dependencies as well | 17:25 |
| RST38h | jeremiah: Correct. But there is no reason to push a standalone lib nobody uses | 17:25 |
| jeremiah | Currently that doesn't happen apparently. | 17:25 |
| Jaffa | And we don't want it to. | 17:26 |
| VDVsx | true | 17:26 |
| Jaffa | Perfect :) | 17:26 |
| RST38h | jeremiah: The only problem is you cannot enforce this | 17:26 |
| zaheerm | you shoudl also test other apps who have same deps | 17:26 |
| jeremiah | This is something that hhedberg complained about though. | 17:26 |
| RST38h | jeremiah: because plugins are also libraries | 17:26 |
| zaheerm | to make sure upgrading a dep doesn't break those apps | 17:26 |
| RST38h | jeremiah: and stuff like font packs is just data | 17:26 |
| jeremiah | That people are blaming mauku for microfeed's problems | 17:26 |
| Jaffa | jeremiah: That's because he wants lots of other people to use microfeed, and for its updates to be visible to people. Given microfeed installs a Control Panel plugin, it is visible | 17:26 |
| jeremiah | okay, so it falls into the 'app' category. | 17:27 |
| zaheerm | facebook widget also had a problem....due to a lib not being updated automatically... | 17:27 |
| zaheerm | not that that is a community app | 17:27 |
| jeremiah | So our rule still stands - without an app a lib does not get promoted | 17:27 |
| RST38h | yes | 17:27 |
| zaheerm | where an app is a real app or a meta package | 17:27 |
| VDVsx | yup | 17:27 |
| jeremiah | cool | 17:27 |
| VDVsx | now CLI :P | 17:28 |
| jeremiah | Have we decided that a bugtracker is a blocker? | 17:28 |
| lizardo | jeremiah: I just remembered some one asked on the list : "what about security and bug fixes?" | 17:28 |
| RST38h | VDVsx: I would say we choose a category that semantically fits cli apps and put them there | 17:28 |
| zaheerm | bugtracker or maintainer email address or just: bugtracker? | 17:28 |
| RST38h | jeremiahl No | 17:28 |
| lma | Yes please, you can't do any QA meaningfully without a bugtracker | 17:28 |
| RST38h | jeremiah: As long as there is a way to contact maintainer, that is it | 17:28 |
| Jaffa | jeremiah: I think trivial apps can treat /packages/ as a simple bugtracker (whilst it allows separate comments) | 17:28 |
| jeremiah | lizardo: Yeah. That is a big question. | 17:28 |
| lma | Private email won't cut it, it needs to be publically visible | 17:29 |
| VDVsx | what about background packages ? for e.g | 17:29 |
| Jaffa | lma: Right, so a comment on the package page is visible. | 17:29 |
| * RST38h has Google Group for reporting and discussing bugs | 17:29 | |
| vre | Do all the project need to have maemo bugtracker for automation then? | 17:29 |
| * RST38h is not going to create a separate Bugzilla for this thing: he does not have time to keep track of all these sites | 17:29 | |
| lma | jaffa: yes, but it'll be a nightmare to discuss, track across versions and so on | 17:30 |
| Jaffa | lizardo: jeremiah: Security update policy for libs could be to contact the debmaster directly | 17:30 |
| RST38h | So, please, define the "bugtracker" liberally | 17:30 |
| Jaffa | lma: If it's a blocker, it doesn't need to be tracked across versions. | 17:30 |
| Jaffa | lma: Not every app is complex enough to require such big infrastructure (and all bug trackers suck; some just suck more than others) | 17:30 |
| zaheerm | maybe infra should create some kind of "bugtracker" functionality for apps that don't want to maintain own bug tracker and want to go into extras-testing ? | 17:30 |
| RST38h | zaheerm: The problem is maintainer willing to use it | 17:31 |
| jeremiah | Jaffa: I think we need something more involved, like a security team. | 17:31 |
| jeremiah | Or some more formal policy | 17:31 |
| RST38h | zaheerm: You can create a bugtracker but if I can't be othered to keep track of it, what use is it? | 17:31 |
| jeremiah | Something that co-ordinates with Nokia too | 17:31 |
| Jaffa | jeremiah: Up to the debmaster then ;-) | 17:31 |
| qwerty12 | maemo.org's Bugzilla is open to Extras' products | 17:31 |
| jeremiah | Jaffa: Okay, I'll try and wrestle with that. | 17:31 |
| zaheerm | RST38h, yah so in promoter interface, it can tell you to either specify a bug tracker or use this one created on maemo.org and lack of response will result in package being demoted? | 17:32 |
| vre | For external bugtrackers there should be a link for it, and for a blocker a link to the report. | 17:32 |
| Jaffa | jeremiah: My point being that security issues can be considered outside of the normal QA process | 17:32 |
| Jaffa | zaheerm: Eugh | 17:32 |
| zaheerm | security issues should be considered outside of normal QA process | 17:32 |
| jeremiah | Jaffa: Yeah, that sort of dawned on me as I started writing. :) | 17:32 |
| RST38h | zaheerm: I would say bugtracker should not make a difference | 17:32 |
| RST38h | zaheerm: Lack of it should not block the package | 17:32 |
| jeremiah | Should security issues be consider outside the QA process? | 17:32 |
| zaheerm | ok so maintainers email should be enough? | 17:33 |
| RST38h | Lack of maintainer contact info should block the promotion though | 17:33 |
| RST38h | zaheerm: yes, email should be NECESSARY | 17:33 |
| lcuk | about the lib thing, if a bug is identified there, but the main app doesnt need to change, how does it get promoted? | 17:33 |
| Jaffa | Package doesn't need a bugtracker to be end-user suitable. Package needs a way to publicly higlight issues with the app which prevent it from being end user accessible. | 17:33 |
| zaheerm | lcuk, app package version update with dep on newer lib? | 17:34 |
| Jaffa | At the moment, that can be a bug tracker, a Google Group or a comment on the package page. If the package page consisted of checkboxes, perhaps each one gets a small "notes" field for the tester to complete; however it becomes more self-documenting. | 17:34 |
| Jaffa | zaheerm: +1 | 17:34 |
| jeremiah | RST38h: I think we already capture contact data at upload now. | 17:34 |
| jeremiah | In fact, I know we do. | 17:35 |
| lcuk | zaheerm, anything, how does qt get updated if its a lib, shopper and birdwatcher might be user apps, but qt isnt | 17:35 |
| jeremiah | So we always have a valid email to contact dev with. | 17:35 |
| lcuk | same goes for libliqbase etc | 17:35 |
| jeremiah | Yeah see! The checkbox list is a good idea! | 17:35 |
| jeremiah | :P | 17:35 |
| * RST38h is against the checkbox list | 17:35 | |
| jeremiah | :) | 17:35 |
| zaheerm | lcuk, if there is an important bug in the lib, it would be user visible in one of the apps..that app should do the update | 17:36 |
| jeremiah | Okay, I think we have gone over the whole agenda pretty much. | 17:36 |
| RST38h | As dumb as it sounds, each extra click you force the voter to make reduces the number of potential voters by 50% | 17:36 |
| Jaffa | RST38h: It theoretically improves the documentation (no need to go on a hunt for the QA guidelines) and speeds up the process: "yes, yes,. yes, no, yes, yes, yes" | 17:36 |
| lcuk | but the app where the bug was identified wasnt pushed yet | 17:36 |
| jeremiah | So we can continue with unstructured conversation. | 17:36 |
| fiferboy | A Qt app doesn't necessarily pull all the Qt libs, though | 17:36 |
| lma | CLI! | 17:36 |
| lcuk | it can and will happen | 17:36 |
| VDVsx | CLI ^ | 17:36 |
| Jaffa | RST38h: I think highlighting how easy it is to rate an app will *increase* participation. | 17:36 |
| RST38h | Jaffa: People are dumb. Very very dumb. | 17:36 |
| VDVsx | :P | 17:36 |
| jeremiah | Jaffa: Exactly - and the dev knows exactly what to fix! | 17:36 |
| Jaffa | RST38h: So, checkboxes easy :-p | 17:36 |
| Jaffa | lcuk: fiferboy: As zaheerm says, if there's a bug which affects nobody, it's not a bug. | 17:37 |
| Jaffa | jeremiah: indeed. | 17:37 |
| jeremiah | With checkboxes we have a way to limit the tester's own bias | 17:37 |
| lcuk | it does though, its just not been identified by user app yet. the -dev testing code spotted it | 17:38 |
| Jaffa | Right, let's wrap up on libraries (and maybe checklists) and then CLI? | 17:38 |
| lcuk | zero day so to speak | 17:38 |
| jeremiah | So they are not testing 'background colo' | 17:38 |
| jeremiah | Yeah, let's wrap up | 17:38 |
| vre | Íf the checklist is made so, that not all the points need to be done by one person - it's great idea. Creates participation too. | 17:38 |
| jeremiah | http://wiki.maemo.org/QAMeeting | 17:38 |
| VDVsx | With checkboxes anyone can test ;) | 17:38 |
| Jaffa | lcuk: And when an app using that feature gets promoted, it pulls the fixed libs up with it | 17:38 |
| jeremiah | I pust actionable items there. | 17:38 |
| jeremiah | VDVsx: Another great point! :) | 17:38 |
| lcuk | jaffa, the testing code is in the lib source itself | 17:38 |
| lma | vre: +1 | 17:38 |
| RST38h | Jaffa: No, they are even more dumb than you think | 17:39 |
| Jaffa | lcuk: then why do you need to push the fix? | 17:39 |
| Jaffa | RST38h: Don't patronise me. | 17:39 |
| jeremiah | We have logged this chat so I will try and go over it and condense things. | 17:39 |
| Jaffa | RST38h: (please) | 17:39 |
| RST38h | Jaffa: Sorry | 17:39 |
| jeremiah | Thank you everyone for you help! | 17:39 |
| jeremiah | It has been very useful! | 17:39 |
| VDVsx | CLI :´( | 17:39 |
| Jaffa | CLI includes background? | 17:39 |
| jeremiah | I will summerize and send email to the lists and post to talk, etc. | 17:39 |
| Jaffa | e.g. openssh-server? | 17:39 |
| RST38h | Jaffa: It is the sunny optimism tat people will be clicking through all this stuff that makes me do that | 17:39 |
| lcuk | jaffa, cos it frobulates the hackerability of stuff and causes battery drain | 17:39 |
| lma | CLI! | 17:40 |
| Jaffa | jeremiah: I think we need to tackle CLI ;) | 17:40 |
| Jaffa | RST38h: It's not sunny optimism, it's trying to streamline the process for devs & testers. We disagree on the effect it'll have. | 17:40 |
| VDVsx | do we want CLI apps available for regular users ? | 17:40 |
| lcuk | jaffa, but the end user app still doesnt need to change cos the bug was found in the lib | 17:40 |
| Jaffa | lcuk: if it's that bad, contact debmaster | 17:40 |
| lma | define "regular"? | 17:41 |
| VDVsx | JOE - the user :P | 17:41 |
| Jaffa | VDVsx: Extras should be full of software. The user should know what they are getting. | 17:41 |
| * Jaffa starts from there. | 17:41 | |
| lma | We want CLI apps to be available via H-A-M, so that the user sees updates, can restore them after reflash and so on | 17:41 |
| jeremiah | cli | 17:41 |
| jeremiah | right | 17:42 |
| Jaffa | lma: s/CLI/useful/ | 17:42 |
| VDVsx | without icons, GUI, ... totally against that | 17:42 |
| VDVsx | also will confuse users | 17:42 |
| jeremiah | Yeah, it will. | 17:42 |
| Jaffa | VDVsx: So you don't think openssh-server (or client) should be in Extras? | 17:42 |
| lma | I'm totally against "social" apps, so what, I just don't install them | 17:42 |
| suihkulokki | lma: remove your irc client NOW | 17:43 |
| jeremiah | The experience a ordinary user would have with the bash shell from extras would be bad | 17:43 |
| Jaffa | VDVsx: As long as the description makes clear that this runs-in-background/in-X-Terminal/... | 17:43 |
| VDVsx | Jaffa, yes, if in a appropriate category | 17:43 |
| jeremiah | They wouldn't understand how to use it. | 17:43 |
| lma | what? | 17:43 |
| suihkulokki | irc client is a social app too | 17:43 |
| jeremiah | The would click on it and this terminal would pop up | 17:43 |
| jeremiah | and they would be sad. | 17:43 |
| VDVsx | or event don't find a icon | 17:43 |
| lma | tsk... | 17:43 |
| jeremiah | cli apps are for devs and people who can handle extras-devel | 17:43 |
| qwerty12 | suihkulokki: Everyone knows that the people on IRC don't have a real life... | 17:44 |
| lcuk | does HAM store anywhere which categories users are in when they install apps? | 17:44 |
| VDVsx | and call nokia care "my device is brokes" | 17:44 |
| VDVsx | *broken | 17:44 |
| lcuk | ie do most people click "All" | 17:44 |
| Jaffa | jeremiah: There's a big marketing win for having something like SSH server visible in Extras though | 17:44 |
| jeremiah | Jaffa: Then it needs another avenue to get in | 17:44 |
| Jaffa | lcuk: Since our patches to improve the view have been integrated, I don't use "All" that much | 17:44 |
| vre | I think this will be maemo-users QA process creation is iterative process, so I'll just get better over time - learn by doing. Can't see all the problems yet.. | 17:44 |
| Jaffa | vre: indeed, but that doesn't mean we shouldn't address the problems we've already encountered | 17:45 |
| lma | people are already arguing for merging the tools repo, I don't think we should creating more | 17:45 |
| vre | Jaffa: sometimes I feel that there should be users, powerusers (ssh) and guru-users (vi) :) | 17:45 |
| VDVsx | ok, probably you should ask Nokia about it ? they do the marketing :P | 17:45 |
| Jaffa | vre: :) | 17:45 |
| VDVsx | S/you/we | 17:45 |
| Jaffa | VDVsx: The don't market to the /. crowd :-p | 17:46 |
| vre | Jaffa: definitely, fix what we can and know | 17:46 |
| * Jaffa actually doesn't care. As long as users know what they're getting, I don't mind. | 17:46 | |
| Jaffa | Happy to go with a consensus that applications which don't auto-start/plugin and have no non-X Terminal way of starting do not get into Extras. | 17:47 |
| lma | Then where do they go? | 17:47 |
| *** danielwilms has quit IRC | 17:47 | |
| VDVsx | note that can be a bit difficult to explain to some people, what is a command line application | 17:47 |
| Jaffa | So, you could include ssh-server (but not client), vim (cos it has an icon and starts X Terminal automatically) and excludes socat or findutils | 17:47 |
| lma | Suppose tomorrow a security hole is found in openssh, if it's not visible in h-a-m how will users know about it? | 17:47 |
| Jaffa | lma: Hmm, I was implicitly assuming promotion of non-user/ stuff. | 17:47 |
| qwerty12 | It's also worth noting tz1 (the maintainer of socat) does not know how HAM works | 17:48 |
| Jaffa | ...but that conflicts with not promotign libs | 17:48 |
| Jaffa | lma: Again, that's the security argument. What if it was found in libssl. | 17:48 |
| lma | libssl is Nokia's problem :-) | 17:48 |
| lma | so it will be updated via SSU eventually | 17:49 |
| Jaffa | Maybe a meta-package of "command line utils" isn't the worst idea, then | 17:49 |
| mgedmin | um, that's like a meta-package "applications" | 17:50 |
| lma | hm it's a very bad idea, why install everything and the kitchen sink if you only want vim? | 17:50 |
| Jaffa | vim is special and separate anyway | 17:50 |
| Jaffa | mgedmin: True | 17:50 |
| RST38h | even worse, why install vim if I want everything else | 17:50 |
| Jaffa | OK, bad idea. | 17:50 |
| Jaffa | So, options are: | 17:50 |
| Jaffa | 1) CLI apps which need to be run from X Terminal linger in -devel/-testing | 17:51 |
| Jaffa | 2) New user/cli (or, with a subsection-aware HAM, user/network/cli and user/desktop/cli) for them | 17:51 |
| * mgedmin has always been rooting for a separate app-manager category for CLI apps | 17:51 | |
| RST38h | Jaffa: I have got another suggestion | 17:52 |
| qwerty12 | Jaffa: CLI apps having a section of non-user/ and promotion of them would be more preferable to 1) | 17:52 |
| Jaffa | qwerty12: A convention of cli/... for stuff which is promotable separately without allowing promotion of everything outside user/? | 17:52 |
| RST38h | Jaffa: For some CLI apps (mc,pico,nano,htop,etc) you can actually create a desktop file that launches app inside an xterm. Can we generate such files effectively making apps desktop-based? | 17:52 |
| VDVsx | qwerty12, true, that's ubuntu/fedora/suse way | 17:52 |
| lma | 2 gets my vote | 17:53 |
| zaheerm | RST38h, +1 | 17:53 |
| Jaffa | RST38h: I'd say developers should do that for those apps, yes. (as I did for vim). However, soemthing like socat needs params... | 17:53 |
| zaheerm | except someone needs to improve the flat icon menu | 17:53 |
| lma | heh, so a script that launches sshd in osso-xterm with a nice action is okay then? | 17:53 |
| Jaffa | lma: yes. | 17:54 |
| lma | s/action/icon/ | 17:54 |
| qwerty12 | socat is only in Extras with a user/ section because the maintainer thinks that only apps with a user/ section gets installed as a dependency by HAM. | 17:54 |
| lma | that's totally superficial IMHO | 17:54 |
| zaheerm | actually openssh i think should have a status bar icon | 17:54 |
| Jaffa | lma: Well, I think sshd doesn't count as a CLI app - it's a background process/plugin type argument. | 17:54 |
| zaheerm | like openvpn-applet does vpns | 17:54 |
| lma | how is x11vnc different? | 17:55 |
| mgedmin | ugh, no inflation of statusbar icons! | 17:55 |
| Jaffa | But a zenity front-end to ssh(-client) would be sufficient. | 17:55 |
| Jaffa | lma: I don't know. I've not used it recently. | 17:55 |
| mgedmin | I wouldn't mind a small GUI app that indicates whether sshd is running or not, and buttons to start/stop it | 17:55 |
| mgedmin | maybe control panel applet? | 17:55 |
| mgedmin | overkill imho | 17:55 |
| zaheerm | mgedmin, yah control panel applet would work too... | 17:56 |
| lma | both are background processes that give remote access, they shouldn't be treated differently | 17:56 |
| mgedmin | install opensshd -> it's always running -> everybody's happy | 17:56 |
| RST38h | Jaffa: To be more specific, anything that makes use of Curses should have a desktop file | 17:56 |
| mgedmin | ssh is encrypted and authenticated | 17:56 |
| Jaffa | lma: Does X11vnc auto-start? | 17:56 |
| Jaffa | RST38h: Yeah, probably. | 17:56 |
| mgedmin | x11vnc is non-encrypted passwordless backdoor | 17:56 |
| lma | no | 17:56 |
| mgedmin | they should be treated differently | 17:56 |
| Jaffa | lma: right, and that's where the difference is. | 17:56 |
| Jaffa | As mgedmin says | 17:56 |
| lma | so a non-autostarting sshd is ok for extras? | 17:57 |
| Jaffa | lma: no | 17:57 |
| Jaffa | lma: Without a way for the user to start it, without going to X Terminal, no. (According to the arguments above, which I'm persuaded of) | 17:58 |
| RST38h | I am all for openssh start/stop status bar icon, that would also show current connectiong | 17:58 |
| RST38h | But who is goint to WRITE IT? | 17:58 |
| zaheerm | RST38h, suggesting it on t.m.o will get it written :) | 17:59 |
| qwerty12 | The OpenSSH argument is kinda moot, considering how it is already in Extras, and all... | 17:59 |
| zaheerm | look at quim and his chess game | 17:59 |
| lma | ironically that could /introduce/ security issues :-/ | 17:59 |
| Jaffa | RST38h: At the moment, it doesn't need to be written because sshd autostarts. If it didn't, and the maintainer (or someone else) wanted it in Extras it would need to be written. | 17:59 |
| Jaffa | zaheerm: Ooooh, brainstorm! *cough* | 17:59 |
| VDVsx | RST38h, better ask Quim to suggest it in tmo :P | 17:59 |
| RST38h | zaheerm: Yea. Right. :) | 17:59 |
| lma | qwerty12: it is now, but will the next version be able to get promoted? | 17:59 |
| zaheerm | hey give me a day and i'd write it :) | 18:00 |
| RST38h | Folks, suggesting something on tmio DOES NOT get it written, be realistic | 18:00 |
| mgedmin | sometimes it does, but that's _very_ rare | 18:00 |
| zaheerm | RST38h, suggesting a good idea has a likelihood :) | 18:00 |
| zaheerm | s/likelihood/possibility/ | 18:01 |
| VDVsx | RST38h, depends on who is suggesting ;) | 18:02 |
| Jaffa | Is there a consensus around apps which require opening X Terminal *by the user* being denied (i.e. an addition to the QA process)? And that control panel applets, autostarting (if secure and good on battery life) or .desktops which auto-open X Terminal being acceptable? | 18:02 |
| Jaffa | Is there a consensus around a cli/... prefix which would be allowed promotion (with more lax, but still some(?), QA process?) to Extras? | 18:02 |
| lma | I don't agree with that, it will just lead to third party repos popping up all over the place again | 18:02 |
| * VDVsx thinks we should create a new thread @ -devel about CLI apps | 18:03 | |
| suihkulokki | extras-unix | 18:03 |
| mgedmin | Jaffa, I'm not sure how to answer your question | 18:03 |
| zaheerm | yah i think there isn't a consensus here | 18:03 |
| Jaffa | And that cli/... prefix apps would still have .install files but not be shown in maemo.org/downloads/ (or, naturally, HAM) | 18:03 |
| mgedmin | ask people to shut up and express their votes with +1/-1 for every question separately, with pauses between questions | 18:03 |
| lma | not shown in HAM == no updates, no restore | 18:03 |
| lma | which would suck, IMHO | 18:04 |
| Jaffa | lma: write a bolton? | 18:04 |
| * mgedmin wants to be able to sudo apt-get install cmdlinetool from extras _or_ extras-devel | 18:04 | |
| RST38h | mgedmin: nooooooo | 18:04 |
| Jaffa | mgedmin: We're going round in circles here. As there's no consensus, someone needs to summarise what's been discussed so far, present concrete proposals for discussion on -devel and wind this meeting up | 18:04 |
| Jaffa | Otherwise we'll end up going round in circles on -developers as well | 18:05 |
| lma | True | 18:05 |
| mgedmin | sounds sane | 18:05 |
| Jaffa | Oh, and outline the issues (e.g. discoverability, restorability, user-friendliness, ...) | 18:05 |
| mgedmin | maybe we should elect a Maemo King | 18:05 |
| Jaffa | Meh, we've got the council for a reason | 18:06 |
| Jaffa | They take in all the feedback and co-ordinate a consensus or - failing that - a workable compromise | 18:06 |
| Jaffa | So, VDVsx and gcobb FTW :) | 18:06 |
| qwerty12 | Jaffa: Ey up, let's get penguinbait in on the action, too... | 18:07 |
| VDVsx | qwerty12, he's @ Maemo, you can ask him to join ;9 | 18:07 |
| qwerty12 | VDVsx: He's your fellow Council member... | 18:08 |
| *** penguinbait has joined #maemo-meeting | 18:09 | |
| VDVsx | Stskeeps, thanks | 18:09 |
| * Jaffa was limiting his praise (on this topic) to the council members who have contributed to the discussion of this topic. | 18:09 | |
| Stskeeps | penguinbait: http://pastebin.com/m5da645dc | 18:10 |
| VDVsx | so, discussion @ -devel then ? | 18:10 |
| VDVsx | who's wants to start the war ? (write the post :) ) | 18:12 |
| Stskeeps | collect up the proposals so it doesn't turn into a jungle? :P | 18:12 |
| Jaffa | VDVsx: Well volunteered :-p | 18:12 |
| Jaffa | And feel free to editorialise on each of them after laying out the proposal | 18:12 |
| * VDVsx don't has much spare time :´) | 18:14 | |
| jeremiah | sorry, had to go away for a bit, but I'm back. :) | 18:15 |
| jeremiah | Time to make some dinner. | 18:15 |
| VDVsx | jeremiah, will you summarize the results and send them to the ML ? | 18:16 |
| VDVsx | ok, I will deal with the CLI *war* | 18:19 |
| VDVsx | and we're done :) | 18:20 |
| VDVsx | thanks all and bye bye ;) | 18:21 |
| Jaffa | Thanks all | 18:21 |
| *** Jaffa has left #maemo-meeting | 18:21 | |
| lma | Bye! | 18:21 |
| *** lma has quit IRC | 18:21 | |
| *** frals has left #maemo-meeting | 18:22 | |
| zaheerm | thx guys bye | 18:23 |
| jeremiah | VDVsx: Yeah - I'll do taht! | 18:23 |
| jeremiah | that | 18:24 |
| jeremiah | ciao! | 18:24 |
| jeremiah | Thanks zaheerm, VDVsx! | 18:24 |
| *** strank has left #maemo-meeting | 18:26 | |
| *** zaheerm has left #maemo-meeting | 18:28 | |
| *** denver has left #maemo-meeting | 18:39 | |
| *** fiferboy has left #maemo-meeting | 18:48 | |
| *** rmoravcik has quit IRC | 19:05 | |
| *** sjgadsby has left #maemo-meeting | 19:09 | |
| *** dneary has quit IRC | 19:53 | |
| *** Lorthirk has joined #maemo-meeting | 19:56 | |
| *** qwerty12 has quit IRC | 20:34 | |
| *** vre has left #maemo-meeting | 20:44 | |
| *** andre__ has left #maemo-meeting | 20:59 | |
| *** mgedmin has quit IRC | 21:09 | |
| *** penguinbait has quit IRC | 21:20 | |
| suihkulokki | /n | 21:23 |
| suihkulokki | gah | 21:23 |
| *** suihkulokki has left #maemo-meeting | 21:23 | |
| *** lcuk has quit IRC | 21:33 | |
| *** lizardo has left #maemo-meeting | 21:35 | |
| *** jeremiah has quit IRC | 22:16 | |
| *** lcuk has joined #maemo-meeting | 22:20 | |
| *** JoeBrain has joined #maemo-meeting | 22:39 | |
| *** JoeBrain has joined #maemo-meeting | 22:41 | |
| *** Lorthirk has quit IRC | 23:49 | |
| *** penguinbait has joined #maemo-meeting | 23:57 | |
Generated by irclog2html.py 2.15.1 by Marius Gedminas - find it at mg.pov.lt!