| ScottL | troy_s, i liked the second post as well :) | 00:08 |
|---|---|---|
| troy_s | ScottL: That's good. Hopefully it shows the issue. It is a direct response to the top ranking comment on Reddit which is, in no uncertain terms, completely inaccurate and incorrect regarding when bit depths apply. | 00:09 |
| troy_s | ScottL: I _hope_ that is apparent now. | 00:09 |
| troy_s | ScottL: Clear or confusing for you? | 00:09 |
| ScottL | troy_s, the bit depth issue is clear to me, but it is analogous to audio which i am fundamentally familiar | 00:25 |
| troy_s | ScottL: It is sad that some people, even those with solid backgrounds, can't understand that dividing an odd number forces you to round data. | 00:26 |
| troy_s | ScottL: And that means a pretty noticeable image quality issue. | 00:26 |
| ScottL | troy_s, what blows my shorts off, is that when reading the comments in the first post it is clear that GIMP's fundamental flaws (i.e. bit depth) are obscured by it's superficial flaws (i.e. UI) | 00:26 |
| troy_s | ScottL: Yes. But to be fair, a commenter over at Hacker News basically summed up that one can negotiate UI issues but cannot negotiate capability issues. | 00:27 |
| troy_s | ScottL: And the real issue is priorities. Who is GIMP for? If it is for this new "High end imaging" thing that Peter Sikking got them to commit to, why in the _world_ is single window mode a funded change over say, deeper bit depth? | 00:28 |
| ScottL | troy_s, absolutely, and while those from windows or mac are probably not going to negotiate the UI, those within Linux _have_ to...and then find the bit depth lacking :( | 00:28 |
| troy_s | ScottL: Another key point missed. | 00:28 |
| troy_s | ScottL: I was simply chased around with pitchforks by people that assumed I wasn't committed to working 100% in a Libre environment. | 00:29 |
| ScottL | troy_s, oh, i agree about audience! if 8 bit depth is all you need then why are you using the GIMP instead of something with less under the hood? | 00:29 |
| troy_s | ScottL: And the reverse, all the blah blah about not being PS (sidetrack issue really) doesn't cover "OK, so 8 bit is a happy place, are we serving _that_ audience as efficiently as we could?" | 00:30 |
| troy_s | ScottL: Because it feels like it doesn't quite serve any audience as well as it could... sort of stuck with an identity crisis. | 00:31 |
| ScottL | troy_s, yeah, serves all audiences mediocrely but none well | 00:33 |
| troy_s | ScottL: Funny thing... You read David Airey? | 00:35 |
| ScottL | troy_s, no i haven't | 00:48 |
| troy_s | ScottL: http://www.davidairey.com/business-basics/ | 00:48 |
| ScottL | troy_s, wow, penguin pete needs some drugs, psychoanalysis, or anger management :/ | 00:49 |
| troy_s | ScottL: Yeah ... that one almost tipped me. Not because I was impacted but more because of the sheer... ignorance? of the whole... thing. | 00:49 |
| ScottL | troy_s, and i had just read to the point where one of the gimp devs posted (death, i think) | 00:50 |
| troy_s | ScottL: There's a great comment in there. "I see steps at 40 levels. I don't see any at 255. The author is an idiot." | 00:50 |
| troy_s | ScottL: Yep. | 00:50 |
| troy_s | ScottL: Bigger issue I saw, having been around libre a while, is why is the flagship imaging app down to 2 developers? | 00:50 |
| troy_s | ScottL: To me, personally, that speaks of something gone terribly wrong. | 00:50 |
| troy_s | ScottL: Terribly. | 00:51 |
| troy_s | Wholly COW | 00:59 |
| troy_s | that last comment ... as with about ... 60 above it ... clearly didn't read the post. | 00:59 |
| troy_s | Ugh. | 00:59 |
| ScottL | troy_s, LOL, but sadly | 01:08 |
| troy_s | ScottL: I originally asked the question because Blender likely has even more a niche audience than say GIMP, and yet it seems to be _extremely_ healthy. Now the list of whys is probably long and rather spread, but I'm more interested in knowing why GIMP _isn't_ there. | 01:10 |
| troy_s | ScottL: A good chunk is manpower. But why can't it attract the minds then? | 01:10 |
| troy_s | ScottL: I can only think Blender is at least as complex as GIMP. | 01:10 |
| ScottL | troy_s, aye, i would expect blender to be as complex as GIMP, but i think it serves a distinct audience, doesn't it? | 01:13 |
| troy_s | ScottL: Audience aside, I just worry about what Alexandre posted there being 2.5 developers or what not. | 01:14 |
| ScottL | troy_s, well, look at ubuntu studio, i think it's analogous | 01:14 |
| ScottL | troy_s, who does it serve? everyone? no one? | 01:14 |
| troy_s | ScottL: I understood that GIMP's core was in fact Alexa and Mitch. | 01:14 |
| ScottL | troy_s, who should get involved to help? | 01:14 |
| troy_s | ScottL: I think GIMP has gone through that vision phase. Peter Sikking, the fellow involved with Krita (and a good read on both counts) is active with GIMP. | 01:15 |
| troy_s | ScottL: So Krita chose "high end digital painting" (and has made huge strides in that regard including workshopping with Revoy) and GIMP settled on high end image manipulation. | 01:16 |
| troy_s | ScottL: The problem is really that it seems development is crawling... and I'm certain there are reasons. If the whole 'toxic developers' thing is a myth or has passed, then there is room for others to maybe help extend that in a productive pattern. | 01:16 |
| ScottL | holstein, finished with the -generic kernel test and now i'm downloading abogani's -lowlatency kernel | 01:57 |
| ScottL | holstein, performance with the -generic kernel was pretty poor in natty contrasted against performance in maverick or lucid :| | 01:58 |
| ScottL | troy_s, reading through some of the comments on lwn and hacker it seems like there is a LOT of confusion, misinformation, misunderstanding | 02:03 |
| troy_s | ScottL: There is. But LWN and HackerNews pretty much nails it. | 02:03 |
| troy_s | ScottL: Reddit had a top comment that said "It isn't an issue" and was utterly incorrect. | 02:03 |
| ScottL | troy_s, and I am certainly not knowledgeable about the GIMP's development | 02:03 |
| troy_s | ScottL: I just don't really know about the development. Very odd. | 02:03 |
| ScottL | troy_s, i'm not trying to feign insight, but for an outsider a few things come to mind: | 02:40 |
| ScottL | troy_s, the developers might be coders but perhaps they really aren't project leads or managers | 02:41 |
| ScottL | troy_s, i say that because if they truly have been working on GEGL for many years and don't have it implemented yet then perhaps they might change their approach or developer | 02:41 |
| ScottL | troy_s, also, if the 'toxic developer' is a myth or even if it is true, perhaps a little PR might be in order | 02:42 |
| troy_s | ScottL: Agree on that front. | 02:42 |
| ScottL | troy_s, and from one of the articles it said they code in C, have high standards for formatting, and don't want "islands of code", this doesn't seem conducive for patching | 02:43 |
| ScottL | troy_s, unless they were to employ someone dedicated or semi-dedicated to integrate the patches (of course i don't know that they don't do that) | 02:44 |
| ScottL | troy_s, but i think that an honest and open (admitting a few foibles if possible) appeal to the community would probably go along ways for them to garner community understanding if not support | 02:46 |
| ScottL | troy_s, of course i could be completely wrong as well :P | 02:47 |
| ScottL | about everything | 02:47 |
| ScottL | ailo, holstein, abogani : i did some testing (very rudimentary) with the -generic kernel and the -lowlatency kernel | 03:02 |
| ScottL | firstly, i could not start with -rt privileges on the -generic kernel as predicted by ailo and holstein | 03:03 |
| ScottL | note that i was using on board audio in a dell 4700 (no firewire interface or pci card) | 03:04 |
| holstein | -lowlatency is BANGin | 03:04 |
| holstein | i was totally skeptical for the firewire | 03:04 |
| holstein | it works great abogani | 03:04 |
| ScottL | and my "testing" was browsing the internet on firefox and seeing if xruns appeared in qjackctl | 03:04 |
| holstein | yup, i had JACK running at 1.2ms | 03:04 |
| holstein | fired up FF | 03:04 |
| holstein | nothing | 03:04 |
| ScottL | with the -generic i could run at 256 frames/11.6 msecs and receive incidental xruns throughout my "testing", i.e. reading about GIMP | 03:05 |
| ScottL | if i lowered that to 512 frames/23.2 msecs i received no xruns | 03:05 |
| ScottL | BUT | 03:05 |
| ScottL | with -lowlatency i could run with rt privileges and got as low as 64 frames/2.9 ms without any xruns | 03:06 |
| ScottL | i stopped there without "testing" further because it's getting late and the family needs some attention | 03:06 |
| ScottL | ;) | 03:06 |
| ScottL | even though that's hardly conducive how it would handle real audio work, i'm still highly encouraged about the delta i saw | 03:08 |
| holstein | ScottL: what to you think? | 03:08 |
| holstein | we can add abogani 's PPA ? | 03:08 |
| holstein | ScottL: you can see the new found sense of urgency that i have right? | 03:09 |
| ScottL | holstein, i would like absolutely like to add abogani's ppa, but i would have wanted to even if the -generic worked better ;) | 03:13 |
| holstein | :) | 03:14 |
| ScottL | holstein, and i say that because i'm guessing that we will not get the -lowlatency into the repositories | 03:14 |
| holstein | hey, at least it works at all for non-FW | 03:14 |
| ScottL | holstein, but you are right, this becomes much more critical due to the non-realtime privileges _and_ the firewire issues | 03:16 |
| holstein | ScottL: i feel like have the PPA is ok though | 03:17 |
| holstein | as long as its installable | 03:17 |
| holstein | out of the box | 03:17 |
| holstein | ScottL: you think so too? | 03:19 |
| holstein | i mean, ideally, shipping with -lowlatency | 03:19 |
| ScottL | i think shipping with the -lowlatency kernel so that everything (rt, firewire, etc) works out of the box without installing anything extra | 03:21 |
| ScottL | but we can't control that currently, that's up to the ubuntu kernel team (UKT) | 03:21 |
| ScottL | however, the next best thing would be to enable abogani's ppa but have the user actively install the -lowlatency kernel | 03:22 |
| ScottL | s/enable/enable by default | 03:22 |
| holstein | something in the menu even right? | 03:22 |
| ScottL | the next choice would be to have either the user enable the ppa by hand or have it in ubuntustudio-controls as a check box, then the user can download the kernel | 03:23 |
| ScottL | holstein, possibly as a menu item, but we need to think through it both from a user perspective and from a developer perspective and make sure it's right for both | 03:24 |
| holstein | well, its a band-aid either way | 03:30 |
| holstein | SO i think something obvious and easy to remove later | 03:30 |
| holstein | ScottL: maybe dont mess with sources | 03:33 |
| holstein | and just a menu item that does it all | 03:33 |
| holstein | adds the PPA and installs | 03:33 |
| holstein | thats east to make and put in, and take out | 03:34 |
| ScottL | holstein, i don't think we can install the kernel without adding the ppa to the sources.list | 03:50 |
| ScottL | btw, i managed to get down to 32 frames and 1.45 msec _without_ xruns :) | 03:50 |
| holstein | i know right? | 03:50 |
| holstein | i think its better :) | 03:50 |
| holstein | ScottL: can the menu entry do the whole thing? | 03:51 |
| holstein | add the PPA too? | 03:51 |
| ScottL | i suppose it could, it would be starting a new app that we would need to make | 03:53 |
| ScottL | but ailo's suggestion of a script that runs directly after installation and asks the user (or suggests to the user) to enable the ppa, install the kernel, and rebbot might work well also | 03:53 |
| ScottL | again, we should probably identify how we want it to interface with the user and find the preferred method for that and then focus on how to implement it, maybe iterate and make concessions between user experience and our abilty to implement it | 03:56 |
| holstein | i like the script even better | 03:59 |
| holstein | so everyone sees it | 03:59 |
| holstein | and gets to decide | 03:59 |
| holstein | and that could be take out too | 03:59 |
| holstein | easily | 03:59 |
| ScottL | now that i've played more with the -lowlatency kernel i'm anxious to learn more about creating it from abogani | 04:46 |
| paultag | ScottL: I'm stoked to see forward progress on it :) | 05:02 |
| ailo | hey, tanders12 | 07:55 |
| * abogani waves all | 09:10 | |
| ailo | Hey, AutoStatic. | 10:29 |
| ailo | I'm wondering about the irq thing. How could we find out if irq conflicts still cause a lot of headache? | 10:30 |
| AutoStatic | Hello ailo | 10:31 |
| ailo | I've read a little about the juju, but I know so little to begin with | 10:31 |
| AutoStatic | Running ffado-diag could already reveal if IRQ conflicts will cause headaches ;) | 10:34 |
| ailo | I don't have a device myself, so I can't do any tests. Seems like you are perfect for the job, since you have a clear conflict. holstein has had no problems with -lowlatency, but I could of course ask him to check. | 10:36 |
| abogani | What does that utility intend for "IRQ conflicts"? Irqs shared? | 10:37 |
| AutoStatic | I'm using several FireWire devices with Ubuntu, both at home and at work. And they run better and more stable with a realtime kernel and rtirq | 10:37 |
| AutoStatic | abogani: you mean ffado-diag? | 10:38 |
| ailo | AutoStatic: Well, according to holstein he runs stable at less than 2ms with -lowlatency and firewire. | 10:39 |
| abogani | AutoStatic: Yeah. | 10:39 |
| AutoStatic | That's perfectly possible | 10:39 |
| ailo | But I don't know about holstein's hardware.. | 10:39 |
| AutoStatic | abogani: ffado-diag also outputs something like cat /proc/interrupts | 10:40 |
| AutoStatic | That's all actually | 10:40 |
| AutoStatic | And you can run stable at less than 2ms but I wonder what will happen when you open up a DAW with a dozen plugins, some external softsynths and a sequencer | 10:41 |
| ailo | AutoStatic: Sure. I have pci, and cannot use less than 64 frames/period without xruns. But it often depends on the apps too, I guess. | 10:43 |
| ailo | Still, it seems very acceptable, and very stable. | 10:44 |
| ailo | at least for the average user | 10:44 |
| ailo | But, if there is a big problem with irq, I suppose it won't seem so for those who have it. | 10:45 |
| AutoStatic | If that average user doesn't have a FireWire controller that shares its IRQ with other devices | 10:45 |
| ailo | AutoStatic: Do you think you could try Natty on your laptop to find out? | 10:46 |
| AutoStatic | Unfortunately no | 10:47 |
| AutoStatic | If I would've more time I would love to do so | 10:47 |
| AutoStatic | But unfortunately I have other priorities | 10:48 |
| ailo | AutoStatic: No problem. Well find out sooner or later. I will check with the ffado guys, if they know something about that. | 10:48 |
| ailo | It's too bad we don't have a live cd | 10:49 |
| AutoStatic | About what? IRQ conflicts? | 10:49 |
| ailo | Yes. | 10:49 |
| AutoStatic | They have a Wiki page up about that | 10:49 |
| AutoStatic | http://subversion.ffado.org/wiki/IrqPriorities | 10:49 |
| ailo | I've seen it, I think, but I don't know what has changed with the juju stack | 10:50 |
| AutoStatic | Not much really | 10:50 |
| AutoStatic | I think | 10:50 |
| ailo | So, you think it can be dealt with without rtirq, then? | 10:50 |
| AutoStatic | I've never tried the JuJu stack | 10:50 |
| AutoStatic | No, I don't think so | 10:51 |
| ailo | But, maybe it has to be done manually then? | 10:51 |
| AutoStatic | An IRQ conflict is an IRQ conflict | 10:51 |
| AutoStatic | If a datastream from a Wifi controller interferes with a datastream of a FireWire controller than FFADO will nag | 10:51 |
| AutoStatic | What do you mean by 'manually'? | 10:52 |
| ailo | I don't know :). I was thinking maybe by tuning the system, either by shutting something down, or adjusting something else. | 10:52 |
| ailo | Anyway, I'll check with that ffado guys, to see what they think. | 10:53 |
| AutoStatic | You could disable devices by modprobe -r'ing them | 10:54 |
| AutoStatic | Or unbind drivers (which is a bit more complicated) | 10:54 |
| AutoStatic | But if your FireWire controller shares its IRQ with your GPU then you have a problem | 10:55 |
| ailo | AutoStatic: Would you get the same kind of problems on Windows, or Mac? | 10:56 |
| AutoStatic | Most probably yes | 10:57 |
| AutoStatic | But Windows or Mac drivers are probably more forgiving | 10:57 |
| AutoStatic | And more optimised for use with a specific FireWire soundcard | 10:57 |
| ailo | That's why I'm thinking perhaps the new stack works a little better. | 10:57 |
| AutoStatic | But then I have 0 experience with using FireWire soundcards with Windows or Mac | 10:58 |
| abogani | You can always avoid Irq sharing change a single line in every drivers of interest. | 11:03 |
| AutoStatic | abogani: You mean by fixing the IRQ's of the concerned devices? | 11:06 |
| abogani | AutoStatic: Yes. it should be done with cautions but it a viable option for expert. | 11:07 |
| AutoStatic | Never thought about that actually | 11:11 |
| AutoStatic | Quite some work though ;) | 11:14 |
| ailo | I just finished a script that checks some conditions: Is the user in audio group, Is -lowlatency installed, is the system running on a realtime capable kernel (-lowlatency/-realtime/-rt) | 11:16 |
| ailo | I'm trying it with qjackctl. The script is run before startup. | 11:16 |
| ailo | The idea is that it will notify the user if something is wrong. | 11:16 |
| AutoStatic | So basically the same as this script: http://code.google.com/p/realtimeconfigquickscan/ | 11:17 |
| ailo | Not really. Mine is very simple and adapted to UbuntuStudio | 11:19 |
| ailo | If the system is not running on realtime capable kernel, a notification can appear when starting jack from qjackctl | 11:19 |
| abogani | ailo: Could you put it somewhere? | 11:20 |
| ailo | Actually, it's not doing anything yet. Only checks the conditions. How do I do pastebin? | 11:21 |
| ailo | http://paste.ubuntu.com/553561/ | 11:22 |
| ailo | I'm not a scripter :). So, propably there are a lot of mistakes | 11:23 |
| AutoStatic | "If not, notify the user that realtime is impossible, and that firewire will most likely not work" | 11:24 |
| AutoStatic | What do you mean by realtime? | 11:24 |
| AutoStatic | Running JACK in realtime mode? | 11:24 |
| ailo | Yes | 11:24 |
| AutoStatic | You can run JACK in realtime mode with a generic kernel too | 11:24 |
| ailo | No | 11:25 |
| ailo | Not on Natty | 11:25 |
| AutoStatic | Oh yeah the cgroups stuff | 11:25 |
| ailo | The text should be replaced. I'm more worried that the functions are just hacks | 11:26 |
| AutoStatic | Don't see any functions ;) Just a lot of if then clauses ;) | 11:27 |
| AutoStatic | But looks good to me, but I'm not a Bash expert either ;) | 11:27 |
| ailo | This kind of script could be tool to make sure the user know he's missing something. | 11:28 |
| ailo | In qjackctl there's options to run scripts. It would be perfect to have it there | 11:29 |
| AutoStatic | You could try working with exit statuses | 11:36 |
| AutoStatic | That will drastically simply your script | 11:36 |
| AutoStatic | simplify | 11:36 |
| ailo | What's that? I noticed I should have exit after checking if the user was root.. | 11:37 |
| ailo | The script should initiate one of two things: 1. If basic conditions are not met, then open Ubuntustudio-controls to fix them 2. If user is running on generic kernel, make a notification | 11:39 |
| AutoStatic | http://tldp.org/LDP/abs/html/exit-status.html | 11:39 |
| AutoStatic | It saves you from comparing the output on stdout to a a variable | 11:39 |
| AutoStatic | If the exit status of a certain command is 0 than the command was succesful | 11:40 |
| ailo | ok | 11:40 |
| ailo | It's c thinking, I quess :) | 11:41 |
| ailo | I can barely write hello world. I've spent too little time understanding those things. Seldom have time... | 11:42 |
| AutoStatic | Same here | 11:44 |
| ailo | lunchtime! | 11:45 |
| AutoStatic | Same here ;) | 11:45 |
| AutoStatic | Enjoy your lunch | 11:45 |
| ScottL | abogani, did you see my comments earlier about "testing" -generic vs. -lowlatency? | 13:27 |
| ScottL | abogani, from my rudimentary (and not audio centric) test -lowlatency performs *extremely* well | 13:27 |
| ScottL | ailo, holstein and i were talking about this script last night, i'm wondering if a script run after installation of the OS that says, | 13:28 |
| ScottL | "1. 'ello, we've noticed you don't have the awesome -lowlatnecy kernel, which is highly recommended, please click the 'make me awesome' button below to enable to ppa" | 13:29 |
| ScottL | "2. next you will need to install (via synaptic or apt-get install) the -lowlatency kernel" | 13:30 |
| ScottL | < or possible automate step #2 when the 'make me awesome' button is selected > | 13:30 |
| ScottL | "3. reboot to experience awesomeness" | 13:30 |
| ScottL | < or possible automate step #3 as well > | 13:30 |
| abogani | ScottL: Happy to hear it. | 13:32 |
| ScottL | "4. without this kernel most audio intefaces will suffer performance degredation due to inability to run with -rt privileges and firewire devices will most likely not work at all" | 13:32 |
| ScottL | "5. these settings can also be adjusted at any time by using ubuntustudio-controls (System -> Administration -> Ubuntu Studio Controls )" | 13:32 |
| ScottL | "6. Rock on" | 13:33 |
| ScottL | ailo, this way it doesn't pester someone, but it pops up a dialogue right off the bat after installation | 13:33 |
| ScottL | ailo, this can be part of one of the ubuntustudio-* packages rather than having to adjust a jack package | 13:33 |
| ScottL | OR | 13:34 |
| ScottL | perhaps we can add a tab to the ubuntustudio-controls for a system-tweak page | 13:34 |
| ScottL | right after installation ubuntustudio-controls can be started and opened to this page with all the warning bells and whistles and flashing lights | 13:35 |
| ScottL | abogani, i would also like to start learning how to build the -lowlatency kernel, how difficult would that be? | 13:35 |
| ScottL | abogani, and how would that be effected? | 13:36 |
| ScottL | oh, ailo, you might also talk to paultag, he likes to code :) he could probably be a big help with your script, et al | 13:36 |
| ScottL | good morning kubotu | 13:42 |
| ScottL | oh, i forgot , you are muted ;) | 13:43 |
| ScottL | why we have three (possibly 4) bots in here confuses (and concerns) me :/ | 13:43 |
| abogani | ScottL: Could you explain better what do you meant with previous two questions, please? | 13:44 |
| * abogani doesn't feel well today and also his ability to understand English ... | 13:46 | |
| ScottL | abogani, i am sorry you are not feeling well, would you prefer to discuss this when you feel better? | 13:56 |
| paultag | ScottL: what's up? :) | 13:56 |
| ScottL | hi paultag, how are you this morning? ailo is working on some scripts to help users with understanding they don't have the -lowlatency kernel (which is highly recommended or even necessary) and then install it | 13:57 |
| ScottL | paultag, i thought you might give him some oversight or help in the coding department :) | 13:58 |
| paultag | ScottL: I'm well, thanks! | 13:58 |
| paultag | ScottL: sure, that's not hard at all :) | 13:58 |
| paultag | ScottL: you can check in the /boot/ folder for a kernel with -lowlatency in the name, and with an up-to-date version string | 13:59 |
| ScottL | paultag, to be honest though, we haven't really decided on a particular vector to help the use at this point though | 13:59 |
| paultag | ScottL: aye, but if we check in /boot or on the grub menu, you can detect hand-rolled kernels | 14:00 |
| paultag | as well as any installed kernels :) | 14:00 |
| ScottL | paultag, ailo seems to prefer working with a script to start qjackctl to do this, i am advocating maybe a script that runs once at first boot after installation and pops open a dialogue box | 14:00 |
| paultag | ScottL: I think I agree with you | 14:01 |
| paultag | ScottL: It's easy enough to whip up something for python + GTK+ to inform people | 14:01 |
| ScottL | paultag, here is what ailo has so far (i haven't looked at it though): http://paste.ubuntu.com/553561/ | 14:01 |
| ScottL | paultag, you may not know, but when would someone use glade? | 14:01 |
| * ScottL realizes that's kinda a cold question from nowhere | 14:02 | |
| paultag | ScottL: it's a good start, but it uses dpkg to find the kernel, and outdated kernels will pass that | 14:02 |
| paultag | ScottL: Glade rules, it's for creating GTK+ XML files that you use in the app | 14:02 |
| paultag | ScottL: with GUIs, you don't want to code it into bytecode, that's a waste. You create XML files that are read at runtime | 14:02 |
| paultag | ScottL: so Glade is a really nice frontend to create these files | 14:03 |
| ailo | Hi | 14:04 |
| paultag | howdy ailo | 14:04 |
| ailo | My script was just an idea, but a way to perhaps make it easier for new users to know that they are missing something | 14:04 |
| paultag | ailo: for sure | 14:04 |
| ailo | It would only do something, if some condition was not met, that we would consider critical | 14:05 |
| ailo | But, already using the generic kernel with jack, there is a critical problem in a way | 14:05 |
| ScottL | paultag, so the XML are like the lv2 plugins sort of, you have a framework that reads the XML files and does things based on the XML file | 14:06 |
| paultag | ScottL: exactly, at least for the GUI -- the actualy "meat" ( callbacks and stuff ) are all handled in "real" code | 14:06 |
| paultag | the actual * | 14:07 |
| abogani | ailo: Why don't you use zenity? | 14:07 |
| ailo | To notify? | 14:07 |
| ailo | I see. Zenity does more than that.. | 14:08 |
| ailo | I was meaning to either use zenity, or notify-send. Using an Ubuntustudio logo. | 14:09 |
| paultag | oooh, I like notify-send | 14:10 |
| paultag | on boot, bitch | 14:10 |
| paultag | I dig it | 14:10 |
| ailo | Also, my script would not be an all-in-one solution, just a small helper, just in case | 14:10 |
| ailo | There's also the problem with which kernel is the default kernel | 14:11 |
| paultag | ailo: screw it, there's an easy out | 14:11 |
| paultag | ailo: on gnome login, run your script. check uname for the kernel name, and check if it has lowlatency in it | 14:11 |
| paultag | ailo: if not, then send an angry notify message at the screen | 14:11 |
| paultag | otherwise shut up ;) | 14:11 |
| ScottL | ailo, did you see my suggestion about having it open to a 'system tweak' tab within ubuntustudio-controls? | 14:11 |
| ailo | Yes, that was my plan too | 14:12 |
| ScottL | oh good :) | 14:12 |
| ailo | If user is not in audio group, or -lowlatency is not installed - run Ubuntustudio controls | 14:12 |
| ailo | But there would be needed some explanation to why it pops up as well, perhaps | 14:13 |
| paultag | ailo: I like having the notify send, so that if they're just writing a word doc, they're not bugged | 14:13 |
| paultag | ailo: so just let them know every login that their kernel is crap | 14:13 |
| ailo | hihi | 14:13 |
| paultag | but in the background ;) | 14:13 |
| paultag | it's also much easier to code :) | 14:14 |
| paultag | ( and will handle updates, if they upgrade stock, but not lowlag ) | 14:14 |
| ailo | I suppose the best solution when installing Ubuntustudio audio stuff, is to include those things in the install script (like jackd has for setting realtime). | 14:16 |
| ailo | Well, gotta go. Have fun. | 14:17 |
| paultag | cheerio! | 14:17 |
| ScottL | ailo, ubuntustudio-audio meta package depends on ubuntustudio-controls | 14:17 |
| ScottL | ailo, paultag , holstein : also try to remember that we need to consider both the user that installed via ubuntustudio ISO and one that "upgraded" from vanilla ubuntu | 14:18 |
| ailo | ScottL: Yes. Quite important. | 14:18 |
| paultag | ScottL: aye, if we take the "login check" route, it should do the same thing | 14:18 |
| holstein | hey guys :) | 14:52 |
| holstein | on my natty install, i "upgraded" from vanilla | 14:52 |
| holstein | if that helps this process | 14:52 |
| ScottL | paultag, we had also talked about fixing the editing of memory and rtprio with ubuntustudio-controls | 15:30 |
| ScottL | paultag, you might look here for ubuntustudio-controls: https://wiki.ubuntu.com/UbuntuStudio/ReleasePlanning#Proposed%20Improvements | 15:31 |
| ScottL | i finally started working on the gnome-classic default xsession issue :) it feels good to be making (hopefully) positive progress on it | 16:11 |
| ScottL | holstein, yeah, i see the email again from colin watson about the ubuntustudio-video package :( i'll look into it probably some tomorrow to see what's going on | 16:28 |
| tanders12 | abogani: are there any limitations to running your natty kernels on maverick? | 18:17 |
| falktx | hey there | 20:41 |
| falktx | ScottL: do you have a minute? | 20:41 |
| holstein | tanders12: im using the -realtime one | 20:49 |
| holstein | in maverick | 20:49 |
| holstein | my wifi adapter seems to work occasionally with it | 20:50 |
| holstein | other than that, its fine *and* dandy :) | 20:53 |
| === detrate` is now known as detrate | ||
| tanders12 | holstein: yeah I was just wondering if there were any known issues other than wifi | 21:25 |
| ScottL | falktx, yeah, i'll be in and about with the kids and dogs, but i'm here | 21:46 |
| holstein | ScottL: im running out too | 21:46 |
| falktx | ScottL: too late now | 21:46 |
| holstein | but as always, if i can help with that issue, PM me or something | 21:47 |
| holstein | im up late :) | 21:47 |
| tanders12 | holstein: if your laptop has your graphics and fw on the same IRQ is there anything you can do? | 21:54 |
| holstein | tanders12: i was told there is | 21:54 |
| holstein | the bios *when possible | 21:54 |
| holstein | other than that | 21:54 |
| holstein | i think there is a software solution | 21:54 |
| holstein | but ive never needed to try it | 21:54 |
| tanders12 | hm | 21:54 |
| tanders12 | ok | 21:54 |
| holstein | tanders12: someone in #opensourcemusicians was either doing it | 21:54 |
| holstein | or trying it | 21:55 |
| holstein | didnt keep track of the results | 21:55 |
| tanders12 | i think ill start with the ffado irq info and go from there | 21:55 |
| tanders12 | the thing is that windows doesn't seem to have any glitches so i wonder whats up | 21:55 |
| holstein | hmmm | 21:56 |
| holstein | tanders12: you said TI chip right? | 21:56 |
| tanders12 | yeah | 21:56 |
| tanders12 | shared with i915 and yenta | 21:56 |
| tanders12 | but i think yenta is the controller | 21:57 |
| holstein | tanders12: what is the issue? | 21:57 |
| holstein | lots of xruns? | 21:57 |
| tanders12 | xruns | 21:57 |
| tanders12 | ^^ | 21:57 |
| holstein | at what latencey? | 21:57 |
| tanders12 | aprox every minute | 21:57 |
| tanders12 | any latency | 21:57 |
| holstein | drag | 21:57 |
| holstein | you could try a couple of live CD's | 21:57 |
| holstein | that would be easy | 21:58 |
| tanders12 | easiest way to cause an xrun is to minimize/maximize ardour | 21:58 |
| holstein | the old 64studio one | 21:58 |
| holstein | AVLinux | 21:58 |
| holstein | and maybe pure-dyne | 21:58 |
| tanders12 | hm good ide4a | 21:58 |
| tanders12 | i was gonna try lucid as well | 21:58 |
| holstein | that would be a nice cross-section of different kernels | 21:58 |
| holstein | and different verions of JACK | 21:58 |
| holstein | and the firewire stack too | 21:58 |
| holstein | tanders12: i would be interested in natty | 21:59 |
| holstein | with abogani 's -lowlatency kernel | 21:59 |
| holstein | seems to work *so* well with my hardware | 21:59 |
| holstein | not that you want to run natty on a production box right now | 21:59 |
| tanders12 | well | 22:00 |
| tanders12 | my lappy has issues with unity | 22:00 |
| tanders12 | that im trying to iron out | 22:00 |
| holstein | you can choose gnome | 22:00 |
| holstein | classic startup | 22:00 |
| holstein | tanders12: IF you install natty | 22:00 |
| holstein | ping me | 22:00 |
| holstein | id like to try some tests with you | 22:01 |
| tanders12 | already installed the alpha | 22:01 |
| holstein | to confirm the new firewire settigns | 22:01 |
| holstein | settings* | 22:01 |
| holstein | tanders12: OH | 22:01 |
| holstein | how is it? | 22:01 |
| tanders12 | unity crashed in 5 min | 22:01 |
| holstein | tanders12: if you get a minute | 22:01 |
| holstein | get jackd | 22:01 |
| holstein | and the -lowlatency kernel | 22:02 |
| holstein | add yourself to the audio group | 22:02 |
| holstein | and see if jack starts | 22:02 |
| tanders12 | k | 22:02 |
| holstein | i want to confirm those are the only steps needed | 22:02 |
| tanders12 | do you have to use the alternate iso in order to use gnome? | 22:02 |
| holstein | nop | 22:02 |
| holstein | e | 22:02 |
| holstein | tanders12: at login | 22:02 |
| tanders12 | ok | 22:02 |
| holstein | down at the bottom | 22:02 |
| holstein | 'session' i think its called | 22:03 |
| holstein | you want classic session | 22:03 |
| holstein | thats good 'ol gnome | 22:03 |
| tanders12 | im in the middle of an arch install on another machine right now but it's using this one for internet sharing but I'll try it later on | 22:03 |
| holstein | tanders12: im heading out too | 22:03 |
| holstein | whenever you get a chance | 22:03 |
| tanders12 | k ill let you know how it goes | 22:03 |
| holstein | next couple days or whatever | 22:03 |
| tanders12 | k | 22:03 |
| holstein | i just wanted to catch you before you did anything else | 22:03 |
| holstein | video group or udev | 22:04 |
| tanders12 | ah makes sense | 22:04 |
| tanders12 | yeah Ill start with what you said and figure out what the minimum needed is | 22:04 |
| holstein | cool :) | 22:04 |
| holstein | im outty... laterx.. | 22:05 |
| tanders12 | cya | 22:06 |
| ScottL | anyone know what falktx wanted? | 22:22 |
| ailo | tanders12: You there? | 23:02 |
| tanders12 | ailo: yeah hows it goin? | 23:02 |
| ailo | Good. I finally remembered what I was supposed to tell you last time (no big deal) | 23:03 |
| tanders12 | lol | 23:04 |
| ailo | It was about getting dependencies for source code. I told you: sudo apt-get build jack | 23:04 |
| ailo | I mean, sudo apt-get build-dep jack | 23:04 |
| ailo | But, of course it should be, sudo apt-get build-dep jackd | 23:05 |
| ailo | Like the package | 23:05 |
| tanders12 | right | 23:05 |
| ailo | Anyway, you having problems with firewire? | 23:05 |
| tanders12 | just the same old xruns | 23:05 |
| ailo | Usually that would go away with rtirq script together with an rt kernel. | 23:06 |
| tanders12 | i still have a lot of improvements i can make though, like IRQs and getting natty going | 23:06 |
| tanders12 | dont know what rtirq is... | 23:06 |
| ailo | http://alsa.opensrc.org/Rtirq | 23:07 |
| ailo | It's a script that together with the rt patch gives higher priority to audio devices. | 23:07 |
| ailo | It usually comes natively with an rt-kernel, if a distro has that. | 23:08 |
| tanders12 | ooo nice | 23:09 |
| ailo | I'm still not sure whether this is still a problem on Natty, though. Or as big of a problem. | 23:09 |
| tanders12 | one problem i have right now is that the firewire on my laptop shares an irq bus with the graphics chip :/ | 23:09 |
| ailo | If you ever try Natty with the -lowlatency kernel, let us know of the results. | 23:10 |
| ailo | Otherwise, I can at least recommend Ubuntu Karmic vintage. Haven't messed around with Lucid so much. | 23:11 |
| tanders12 | im planning to try natty w/ lowlatency first, and if that doesn't work lucid | 23:12 |
| tanders12 | or possibly pursue maverick running the natty kernel | 23:12 |
| ailo | On Natty, all you have to do is install the -lowlatency and add yourself to the audio group | 23:13 |
| tanders12 | and ive almost got my arch system going. the base install is done i just need to add a gui :D | 23:13 |
| ailo | The same udev rules file you used on Maverick comes installed by default on Natty, in /lib/udev/rules.d/* | 23:13 |
| tanders12 | nice | 23:15 |
| ailo | I should try arch sometime. Never got around to it yet. Never tried Slackware either.. | 23:15 |
| tanders12 | i love how stuff gets done in the linux world | 23:15 |
| tanders12 | on the other hand i hate how things get broken all the time... | 23:15 |
| tanders12 | but it sounds like natty's gonna be pretty sweet for audio if they get the cgroups thing under control | 23:16 |
| ailo | I don't know what will happen on the cgroups front before Natty release. They seem to be working on that. | 23:16 |
| ailo | At least the -lowlatency kernel rocks, and is not affected | 23:17 |
| ailo | Probably there will be a new way to get realtime in the future, if I've understood things correctly. | 23:17 |
| tanders12 | the official one, or the abogani? | 23:18 |
| ailo | The official is -generic, which sucks for audio right now. -lowlatency is aboganis. | 23:18 |
| tanders12 | ok thats what i thought | 23:18 |
| tanders12 | have you ever built a kernel | 23:18 |
| ailo | Or, alessio boganis :). | 23:18 |
| tanders12 | ? | 23:18 |
| ailo | Yes, I did LFS a couple of times, but I didn't spend much time with it | 23:19 |
| ailo | It's not so difficult to build the kernel itself. As long as you are using the same source as the distribution uses. | 23:19 |
| ailo | But, from what I understand, things can get a little messy when applying patches, like the rt patch. | 23:20 |
| tanders12 | i might try doing a custom one with the rt patch and that "200 line" superpatch | 23:20 |
| tanders12 | but if i do it wont be for a while | 23:21 |
| ailo | Nice initiative from Ralf to start a tut in the mail list. You saw it? | 23:22 |
| tanders12 | no? | 23:24 |
| ailo | It's on the ubuntustudio-user list. How to build a kernel. | 23:25 |
| ailo | On this thread: Fernando's comment on Ubuntu Distro's where the Tascam US-122 | 23:26 |
| ScottL | hi rlameiro :) | 23:30 |
| rlameiro | ScottL: hey! how are you ? | 23:31 |
| tanders12 | ailo: aw ok | 23:31 |
| tanders12 | ill check it out | 23:31 |
| ScottL | rlameiro, i'm doing well, been real busy with family today and didn't do much for ubuntu studio so i'm a little bummed about it, but how are you today? | 23:32 |
| ScottL | ailo, if you are interested in slackware for audio you might look into: http://www.slackermedia.info/ | 23:32 |
| ScottL | ailo, that's klaatu's guide, he's pretty hip about many things | 23:32 |
| rlameiro | ScottL: going to try some stuff on my arm board | 23:33 |
| rlameiro | I think ubuntu has a serious bug on some usb audio devices.... | 23:33 |
| rlameiro | i check it at maverick on ARM, then maverick on x86 same problem | 23:33 |
| rlameiro | so i need to check it in natty | 23:34 |
| ailo | rlameiro: What kind of problems? | 23:35 |
| tanders12 | rlameiro: yeah, I think the bug is called "I can't handle USB 2.0" ;) | 23:35 |
| rlameiro | no 2.10 | 23:36 |
| rlameiro | 2.0 | 23:37 |
| rlameiro | its a 1.1 device | 23:37 |
| rlameiro | and edirol ua4fx | 23:37 |
| tanders12 | yeah i was just kidding | 23:37 |
| tanders12 | anyone know off the top of their heads how NEC firewire chipsets typically do? | 23:38 |
| rlameiro | the system recognizes the device but can play. doesnt show up on pulseaudio and after trying to play something via alsa (via aplay) it outputs some error about audio format | 23:38 |
| rlameiro | audio -----> TI | 23:38 |
| ailo | Thanks for the info ScottL. Never heard of that until now (but didn't look for it yet). | 23:39 |
| tanders12 | yes I know, but I have an NEC laying here and I'm gonna try it and see how it does | 23:39 |
| rlameiro | tanders12: well, then share your findings with us | 23:40 |
| rlameiro | I have a ricoh... not bad, but nothing like the ti chips | 23:40 |
| ailo | rlameiro: did you try jack? | 23:41 |
| tanders12 | rlameiro: for sure | 23:41 |
| ailo | (just checking..) | 23:41 |
| rlameiro | ailo: yeap. first thing :D | 23:41 |
| ailo | rlameiro: I read about it a little. It says that you can't control a lot of stuff with Linux, right? | 23:43 |
| ailo | rlameiro: Maybe it's locked in a bad samplerate | 23:44 |
| rlameiro | it is something about the sambple and bit rate format | 23:44 |
| ailo | rlamerio: did you try it with Windows, or mac. You might be able to change the settings. (just fishing) | 23:44 |
| rlameiro | it works, I am using it in Lucid | 23:45 |
| rlameiro | it stoped working in maverick | 23:45 |
| rlameiro | my guess is something on the kernel | 23:45 |
| rlameiro | there were a lot of usb drivers changes, and i remeber that this board needed a quirck to the usb audio driver | 23:46 |
| ailo | Ok. Well, it seems Maverick is suffering quite a few problems, then :( | 23:46 |
| rlameiro | I will test it better and check what its hapening, hopefully its fixed in natty | 23:46 |
| tanders12 | yeah maverick is definitely havin audio issues | 23:49 |
| rlameiro | yeah, we had a bad release... it wasnt a perfect 10... | 23:50 |
| tanders12 | lol | 23:51 |
Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!