[01:01] <persia> micahg: zequence: http://bugs.debian.org/335568 is the end of the saga: I don't find the middle anywhere convenient, but I suspect it's somewhere in b.d.o archives for linuxsampler (along with rationale why non-free wouldn't work)
[01:01] <persia> err, debian bug 336568
[01:02] <persia> micahg: Everyone who runs flash runs binaries installed by untrusted sources: that's the point of -installer packages.
[01:05] <persia> Ah, there it is: debian bug 328121 - with rationale why it can't be non-free, and reports of discussions with upstream.
[01:06] <persia> (hint: when searching for very old things, search for archived bugs)
[02:19] <micahg> persia: I don't consider Adobe an inherently untrusted source
[03:11] <persia> micahg: You can't know it was compiled at Adobe: there isn't a crytographically verifiable record from notice of build to delivered source.  I don't disagree that most folk are willing to trust Adobe, but I don't see a philosophical difference between trusting one upstream (Adobe) and another (LinuxSampler).
[03:12] <persia> For extra fun, flashplugin-installer doesn't download from Adobe anymore: now it downloads from Canonical, and while most users trust the path by which either the source or binary gets from Adobe to Canonical, it's not inherently trustworthy, it just happens that users coincidentally trust the vendors involved.
[03:21] <len-dt> In the same way that most of us don't trust a down load from MS?
[03:26] <persia> len-dt: Perhaps, although when the MS font package was hosted by MS, most folk did trust MS.  Now they trust some random guy who claimed to have downloaded it from MS before MS stopped hosting it.  Trust is a funny thing.
[03:28] <persia> But I didn't intend to get into a diatribe about trust: the point was only that *-installer packages were intended to install stuff that the distribution couldn't trust, so users had to make their own trust decisions.
[03:28] <len-dt> Ya, I thought about that as I was writing it. There is a difference between font and executable.
[03:29] <len-dt> It also lets the user know what the vendor expect in it's use.
[03:29] <persia> I don't know enough about the environment in which font hinting code runs to agree or disagree with you.
[03:30] <persia> Yes, indeed, it does integrate the vendor experience with the distribution.  When done properly, it further allows the user to manage the installed content as they might manage any other package on the system.
[03:30] <micahg> -installer packages are less about trust and more about distribution rights
[03:31] <persia> micahg: Then I don't understand "I don't think we'd want an installer installing binaries built by untrusted  sources
[03:31] <persia> "
[03:31] <persia> If it's not about trust, then why do we care?
[03:31] <micahg> well, I guess I took the trust for granted in those cases, but the reason why they're needed is distribution rights
[03:32] <persia> Sure, I can agree with that :)
[03:32]  * micahg steps away for a bit...bbl
[03:36] <persia> So, setting trust aside, I'm not convinced that users actually receive a license that allows them to run linuxsampler after downloading it, but there are lots of things on the internet like that ...
[03:36] <persia> (and there are only a few places that have laws that prohibit copying without license (e.g. DCMA))
[03:51] <len-dt> persia, the user doesn't receive a licence to run LS, They already have that. What they get is something that lets them know the licence is not like the rest of the deb/ubuntu licences and allows them to decide before installing if they wish to do so.
[03:53] <len-dt> All the SW we offer has restrictions on how it can be used. LS is no odd that way.
[03:53] <len-dt> It is rather what those restrictions are that is odd
[03:59] <persia> What license to use do they receive?  My understanding is that the additional restriction invalidates the no additional restrictions clause of the GPL.
[04:00] <persia> And the additional bit doesn't say anything about the right to copy or use it otherwise...
[04:01] <len-dt> Or not to.
[04:01] <persia> Right, but in places not Hondouras or Nicaragua, the Berne convention applies, which means there isn't any license to copy or use by default.
[04:01] <persia> I don't happen to be in one of those places, and suspect this is also true for most of our users.
[04:01] <len-dt> I'm not a lawyer for sure.... I would guess a lawyer would have a harder time with it than I would though.
[04:02]  * persia is also unlicensed to dispense legal advice
[04:02] <len-dt> Good point :)
[04:03] <persia> That said, when we're just passing on a license that appears to allow arbitrary use, it seems sane to me.  When something more complicated happens, I seek counsel.  In this case, counsel I retain wouldn't be able to give a statement that would help Ubuntu Studio distribute linuxsampler, so I haven't bothered to do so.
[04:03] <len-dt> It sounds like you are saying though that with an addon the GPL part becomes invalid and so the only licence is the addon.
[04:04] <persia> That's my understanding, yes.
[04:04] <persia> Under GPL section 10 (see /usr/share/common-licenses/GPL)
[04:04] <len-dt> I can wrap my head around that.
[04:05] <len-dt> Even without rereading it.
[04:05] <persia> I totally sympathise with what upstream is trying to do, but I believe they should review that with counsel, and come up with something that has broader consensus of compliance.
[04:06] <persia> Until then, I certainly wouldn't be comfortable having a copy on any system I took to the US or the EU (and I ought to refresh my understanding of affected regions)
[04:08] <len-dt> section 10? Part of the no warranty part?
[04:08] <len-dt> Or am I reading the wrong one?
[04:08] <persia> 10. Automatic Licensing of Downstream Recipients : third paragraph.
[04:09] <len-dt> I 3was looking at version1 not 3
[04:09] <persia> Section 6 of GPLv2
[04:09] <persia> I don't think that many folk still use GPLv1, but it is section 6 there as well.
[04:10] <len-dt> It's ok I am reading 3 now.
[04:11] <persia> Mind you, the only real way to get an answer is for someone to have a dispute in court, in a sufficient set of jurisdictions that a consensus develops, but I don't think anyone wants that to happen: frightful waste of resources.
[04:12] <len-dt> generally personal resources.
[04:13] <persia> Depends.  A couple weeks ago I went to an event where 90% of attendees were paid by their employers to review open source licenses.  All volunteer and contributory resources, but not all personal.
[04:21] <len-dt> So basically, it sounds like someone (not me :) needs to write either a straight GPL or commercial offering that works the same with a clear licence. (we would prefer GPL, but what is there is worse than paying for the same thing)
[04:23] <persia> Except that if it isn't GPL, it may not be able to link against all the GPL libraries it uses, making this very complicated.
[04:23] <persia> Given the history, I suspect some copyright holders thought they were contributing under the GPL, and others under the non-commercial bit, making any change potentially difficult.
[04:24] <persia> The easiest way (although perhaps the most expensive) would be to purchase a license to redistribute the material under arbitrary terms from the authors, then distribute that as GPL, removing the restriction.
[04:25] <len-dt> Ya, but what I am saying is that what is there is really not legal for anyone to use right now... it just has the understanding that only under certain uses will legal action be taken.
[04:25] <persia> That matches my understanding.
[04:25] <persia> Which is annoying, because lots of folk like it.
[04:26] <len-dt> Yet the FLOSS world has gone to a lot of work to make this kind of SW legal and easy to distribute and distributing something without that protection of knowing it is right doesn't make sense.
[04:27] <len-dt> Even without the threat of law suit
[04:27] <persia> Indeed, and there are other models that folk use that are accepted: GPL+CLA+commercial license options, or MIT+trademark, etc.
[04:28] <persia> With either of these, it is easier to pay upstream to use something commercially than navigate the potential exceptions available.
[04:29] <persia> (This is why Debian has iceweasel and Ubuntu has Firefox: Canonical pays Mozilla to work around the trademark restriction)
[04:29] <len-dt> I didn't know that
[04:30] <len-dt> I did find out that it is best to leave the home page as is in the distro ...
[04:30] <persia> Why?
[04:31] <persia> Did all the browser-branding patches drop out over the past couple years?
[04:31] <len-dt> the mozilla licence says some interesting things.
[04:31]  * persia hasn't looked at rebranding the browser experience since 2009
[04:31] <persia> Heh, yes it does :)
[04:31] <micahg> persia: yeah, dropped in 10.10 and in 10.04 after the move to rapid release
[04:32] <micahg> (browser branding)
[04:32] <len-dt> It is ok to add book marks. But it does say somethings about the home page.
[04:32] <persia> micahg: Ah, thanks for the update.
[04:32] <len-dt> I wasn't confident I could interpret it well enough...
[04:33] <persia> micahg: Is there anything else that partially replaced it, or are we back to where we were in 2006?
[04:33] <micahg> persia: nothing replaced it, lack of maintainer for the branding (and rapid release made it not worthwhile)
[04:35] <persia> Makes sense.  It might be nice to have a cleaner API for changing default bookmarks and homepage contents for different flavours, but that needs highly motivated folk to do.
[04:35] <persia> Homepage might be just an alternatives implementation, but bookmarks are trickier.
[04:36] <micahg> yeah, would be nice to have that again
[04:36] <len-dt> persia, I gave up after spending a week I could have been doing something better.
[04:36] <persia> Are the default bookmarks hardcode and changing, or are they something we can dig into?
[04:37] <micahg> hardcoded ATM
[04:37] <persia> len-dt: I totally understand.  I spent about a week trying to use the branding infrastructure when it was there for a contract arrangement, and we were never quite satisfied, so don't think my comments are "it was wonderful in the old days".
[04:37] <len-dt> There is a mozilla dir in /etc where the mozilla site tells how to use it, but the ubuntu version uses something else
[04:37] <persia> micahg: And they change per upstream release (potentially)?
[04:38] <len-dt> Yup.
[04:39]  * persia has a richer understanding of why 40% of the desktop flavours default to alternate browsers
[04:39] <len-dt> persia, I ended up making *.desktop files to add to a submenu on the main menu.
[04:39] <len-dt> This works for studio with xfce, but I would guess not unity
[04:39] <persia> len-dt: That's slightly more scalable, because folk like me who don't use Firefox can still access them.
[04:40] <persia> Could work in unity fine, if one sets up the right lenses or installs other menu providers (e.g. classicmenu-indicator)
[04:40] <persia> But it's decidedly less obvious :)
[04:41] <len-dt> Yes it uses whatever the default browser is.
[04:41] <len-dt> I ended up doing our IRC channel that way too
[04:41] <persia> Err, well, default URI MIME handler, which might not be the browser the user typically launches, but yeah.
[04:42] <persia> A .desktop file for the IRC channel?  Do all the IRC clients even declare support for that?
[04:42] <len-dt> Uses exo-open --launch WebBrowser http://ubuntustudio.org/ for exec
[04:43] <len-dt> IRC is python /usr/lib/ubuntustudio/irc_auto_starter.py   oops zequence put a script in there
[04:43] <persia> Ah, so not the XDG MIME stuff at all.  You might want to trigger sensible-browser, rather than WebBrowser, but that's an implementation detail.
[04:43] <len-dt> it was xchat. Yes it does support that
[04:44] <len-dt> I set xchat up to open in the xchat client already open or open a new one.
[04:45] <persia> Cool
[04:45] <len-dt> zequence, added the dialog so tha the user knows they are going to a live chat.
[04:46] <len-dt> Whats the difference with sensible? I just used what xfce suggested.
[04:47] <micahg> persia: exo-open is the handler similar to sensible-browser
[04:47] <persia> sensible-* are provided by Debian, and Ubuntu had adopted them.  They are scripts that try to get something that does what you want from all the packages that provide it, rather than depending on interrelationships not documented in the package metadata
[04:48] <micahg> exo-open uses the system settings (like default system browser)
[04:48] <persia> micahg: By "system settings", you mean the XFCE configuration?
[04:49] <micahg> well, I'm not sure which implementation it's using on the back end
[04:49] <len-dt> The xfce settings manager has a preferred applications setting
[04:50] <persia> That means we're probably using the Xfce Preferred Applications Framework
[04:50] <len-dt> I am not sure if they are stored in the env though.
[04:51] <persia> The settings GUI only exports a subset.  I don't know if it uses upstream defaults, or has a maintainer script to determine the installed state of the system and make best guesses.
[04:52] <len-dt> The settings GUI is what our user sees, so it needs to work as that does.
[04:52] <persia> I'm a fan of xdg-mime and xdg-open, but since we're using Xfce, these may not be the best choice currently.
[04:52] <persia> Indeed.
[04:53] <len-dt> As the default browser I can choose "debian sensible broser" :)
[04:53] <len-dt> *browser
[04:53] <persia> Which probably ought be the default default, but that's another conversation :)
[04:54]  * persia runs errands
[04:54] <len-dt> This is my raw 12.10 disk so I don't have another browser installed to try different settings to see what works.
[06:07] <len-dt> micahg, I created the bug for jack... when you said regression-release and release name does that mean quantal, or a package name? I am not sure if it is pulse or jack that is the problem.
[06:09] <len-dt> Bug #1075044
[06:35] <zequence> falktx: pong
[06:37] <zequence> len-dt: About the bug. Do you ever get jackdbus to grab the audio card?
[06:43] <zequence> micahg: persia: I assume that the logic around allowing flash is that it's on a trusted url, and that it has been tested. 
[06:43] <falktx> zequence: hey
[06:44] <persia> zequence: rather, it's that Canonical has a license to distribute the binary.  I think it gets tested, but I don't know.
[06:45] <persia> The key thing for distribution being permission to distribute, rather than anything related to quality of distributed material.
[06:45] <zequence> persia: Ok, so they have "permission"
[06:45] <persia> Right.  I don't know details, but I suspect a commercial contract between the two parties.
[06:46] <zequence> falktx: I'm just about to get out the door. I started remembering stuff after asking thelonious to tell you stuff
[06:46] <zequence> falktx: Your jackd2 on Quantal is the same as on Precise, no?
[06:47] <falktx> zequence: pretty much, yes
[06:47] <zequence> From the orig changelog, it seems like it's the Precise package (with the added patch we've discussed)
[06:47] <falktx> zequence: I had to adjust the multiarch a bit to fix the updates, but that's all
[06:47] <falktx> zequence: I'll soon bring jack 1.9.9 though
[06:47] <falktx> currently in testing
[06:47] <zequence> falktx: I'm trying to get jackd2 updated on all releases where we have trouble with it. 12.04-13.04
[06:48] <falktx> I use a custom jack version, so mine is different
[06:48] <falktx> my current 1.9.8 has lots of backport fixes, ladish integration, and prefer dbus over jackd
[06:49] <falktx> my 1.9.9 is the ladish version and dbus-prefered
[06:49] <zequence> Eh, why does the jack2 git source not have a 1.9.9 tag?
[06:49] <falktx> it's not released yet
[06:49] <zequence> Ah, yes
[06:50] <zequence> falktx: Ok, so it's about to be release soon. I think that will take care of things what 13.04 is concerned
[06:50] <falktx> cool
[06:51] <falktx> afaik stephan was just looking for people to test latest git and report issues
[06:51] <falktx> I've reported some, now fixed
[06:51] <falktx> so hopefully it won't take too long now
[06:51] <zequence> falktx: So, you would build jackd2 differently? What else than build options for 1.9.9?
[06:51] <falktx> zequence: I build jack2 very differently from debian/ubuntu
[06:51] <falktx> for starts, no jack1 & jack2 combo
[06:52] <falktx> that's just awful
[06:52] <falktx> jack2 is the default, ever
[06:52] <falktx> I also increase the max-client and ports
[06:52] <falktx> throw it ladish support, and a patch to prefer dbus over jackd
[06:53] <zequence> falktx: How about suggesting a new jack2 package for Debian, in the way you do it?
[06:53] <falktx> because it's not gonna be done
[06:53] <falktx> they want both jack versions
[06:53] <zequence> falktx: So, you can't start jack1 from your package?
[06:54] <falktx> in kxstudio, jack1 doesn't even exists
[06:55] <falktx> to get jack1, you need to add a special ppa that will replace current jack2 version with jack1
[06:55] <falktx> there's only 1 version of jack present in the system
[06:55] <zequence> falktx: Well, I don't see jack1 being a problem anyway. It's how we get jackd2 working, as it is that which is causing problems for us right now
[06:55] <falktx> either jack2 by default, or jack1 if the special PPA is enabled
[06:55] <falktx> zequence: what problems?
[06:56] <zequence> falktx: The jackd2 package in Debian/Ubuntu is still suffering from the bug that leaves jackdbus running in the background after a failed stop
[06:56] <zequence> On both 12.04 and 12.10
[06:56] <falktx> ah, you need to backport 2 fixes from git
[06:57] <zequence> falktx: We'd SRU it
[06:57] <falktx> https://github.com/jackaudio/jack2/commit/aa02feeacfa533a07f04e916334637b57eaac5a2
[06:57] <falktx> https://github.com/jackaudio/jack2/commit/700489b429b0edb7046b169278e3e6751e3e59fc
[06:57] <zequence> falktx: So, two patches? We won't need to change anything else than the orig source for jackd2?
[06:57] <falktx> ^those fixes the shutdown issue
[06:57] <zequence> falktx: Thanks man
[06:57] <falktx> there's also other issues though
[06:58] <falktx> but those are the most important
[06:58] <persia> zequence: For Debian, those ought be pushed as RC for wheezy, and we'll SRU for precise/quantal in Ubuntu.
[06:58] <zequence> persia: RC?
[06:59] <zequence> Opps. I'm late. bbl
[06:59] <persia> zequence: wheezy is frozen, so only release-critical bugs can be uploaded.  Once the freeze completes, it's *lots* harder to get the fixes approved as stable updates.
[07:01] <zequence> persia: I see. Thanks
[14:10] <len-dt> zequence, on login if PA has never run any audio stream it seems to always work. After PA uses the port then it fails. I am wondering if PA writes to disk when it gives up the port and the delay is too long for jack so jack tries to pick up the port before PA has released it.
[14:10] <len-dt> Or there is some other delay.
[14:12] <zequence> len-dt: What was the bug nomber again?
[14:14] <len-dt> Bug #1075044
[14:18] <zequence> len-dt: I'm wondering who or where would be good to ask about this.
[14:18] <zequence> falktx: Do you know anything about this?
[14:18] <zequence> ..PA not releasing card to jackdbus
[14:19] <zequence> btw, I applied the two commits to the Precise jackd2 source. Building now, to try it out
[14:20] <zequence> High time to learn about how administration of patches work
[14:22] <falktx> zequence: sorry no, I handle pulseaudio in a different way
[14:22] <zequence> falktx: Right. Is there a package for that in your repo?
[14:24] <zequence> falktx: And, your version of jack2 is probably a part of your solution?
[14:24] <falktx> no, jack2 version doesn't matter much for my solution
[14:24] <falktx> my solution is cadence
[14:25] <zequence> falktx: I see. Thanks
[14:26] <zequence> len-dt: So, we don't know what happens with the whole dbus thing. But, I would put my money on pulseaudio adding to the problem
[14:27] <zequence> len-dt: There's a #pulseaudio channel. Perhaps hang out there, and bother them for some help on finding out what the error might be?
[14:27] <zequence> len-dt: Or, in #jack
[14:27] <zequence> Or both
[14:27] <astraljava> falktx: Oh, I wasn't "giving you credit", I was just linking to the page that _has_ them, because as you put it, abogani tends to not provide them occasionally, which was at that time. :)
[14:27] <zequence> I can't since I don't have 12.10 here. And, besides, not getting the same kind of problem, I think
[14:28]  * smartboyhw just upgraded to 13.04 so...
[14:31] <zequence> len-dt: Another way is to poke David on this, all though, I don't think he's responsible for jackd being able to grab the card. But, he may be interested in helping solving the problem
[14:34] <zequence> His nick is diwic. 
[14:42] <zequence> Ok, so I've been trying to make dbus crash for a while now. Will try the same thing on 12.10 when I get home. Sweet.
[14:47] <zequence> jackdbus, not dbus :)
[15:30] <scott-work> zequence: sorry i haven't replied to any emails lately, been a very, very busy weekend
[15:30] <scott-work> zequence: although i did start creating/acquiring assets for the mode/workflow video i want to create :)
[15:30] <scott-work> ..
[15:31] <scott-work> incidentally, i have an "almost finished" mix of my first song from the album i am currently working on
[15:32] <scott-work> i'll need to talk to the guy with whom i am working and ask if he's cool with me linking it to some places for people to hear
[15:32] <scott-work> ..
[15:32] <scott-work> i hope to tidy up the PR blueprint some more this week with an email to the ML about a few points (e.g. making sure others have a voice in posting news, etc)
[15:33] <scott-work> zequence: i may also go ahead and take ownership of a few other blueprints in which i am particularly interested, unless someone objects
[16:13] <zequence> scott-work: Doesn'
[16:14] <zequence> scott-work: Doesn't seem like we have that much action on workflows just yet. Hopefully we can work on attracting some people later
[16:14] <zequence> scott-work: I'm just going to finish getting jackd2 updated, and then I'll start working on dev docs
[16:15] <zequence> I think if we get some basic stuff down, it'll be enough as basis for the general reveloper
[16:15] <zequence> developer
[16:15] <zequence> Should help, when introducing them to what we do
[16:16] <zequence> I should put up a bzr repo for the system startup script
[16:17] <zequence> i don't have US dev access yet
[16:18] <zequence> ..(not that I need to)
[16:36] <scott-work> zequence: i was planning on taking the video and maybe the graphics workflow blueprints and making sure they see progress
[16:42] <scott-work> zequence: do you want me to deactivate the "kaj ..." account and add "zequence" to us-dev team?
[18:20] <zequence> scott-work: That would be helpful, yes. There was a bug while merging the accounts. It seems to have been solved, while the accounts have yet not been merged
[19:02] <scott-work> zequence: done and done
[19:05] <zequence> scott-work: Great. Thanks!