**** BEGIN LOGGING AT Sat May 19 02:59:59 2012
May 19 03:00:44 <psycho_oreos> yeah I can see that but still that image doesn't really suit the message that the author wants to convey if one has even just a hint on the things the author uses it in their picture/photo. It kinda really ruins the whole thing if you know what I mean
May 19 03:01:02 <k273> I can see your point, psycho_oreos 
May 19 03:10:13 <Hurrian> http://goput.it/7io.jpg should be demotivatored.
May 19 03:11:06 <Hurrian> if these prehistoric hard drives weren't so rare, I'd bet that people would try to use them as electric car motors
May 19 03:11:35 <k273> o_O
May 19 05:44:19 <r00t|home> Hurrian: while they sure have stronger motors than today's HDDs, i doubt that they can power a car...
May 19 06:23:29 <Estel_> freemangordon,  if everyone assumes that "thumb2 can't work on N900" yet isn't able to give any details - beside reffering to erratas...
May 19 06:24:10 <Estel_> that You've already made working, + another single errata that is irrelevant due to way first errata works
May 19 06:24:39 <Estel_> + without it system crashes every few seconds, and with it, it runs fine for ages + no indication of errors whatsoever
May 19 06:25:09 <Estel_> ...You can safely conclude that other people are spreading FUD, because they haven't did research You've performed
May 19 06:25:55 <Estel_> just use erratas, and if someone, anytime, notice anything that might be related to regression, You will think about it again
May 19 06:26:09 <Estel_> Personally, I doubt that will ever happen, it just "works"
May 19 06:26:17 <Estel_> my 2 cents
May 19 06:29:32 <Estel_> I totally appreciate Your will to trace every single information that might lead to finding any regressions, but, it just seems everyone is saying "it wont work" without any merit. Especially, that You actually "experienced" that it works via Ubuntu 12.04
May 19 06:30:16 <Estel_> on anither topic - freemangordon, do You remember, that we tried to chase ancient reboot bug?
May 19 06:30:24 <Estel_> I got idea what may be causing it
May 19 06:30:47 <Estel_> Could You edit /etc/default/kernel-boot and enable "early ssh"?
May 19 06:30:54 <Estel_> then, try rebooting few times
May 19 06:31:05 <Estel_> if You get hit by bug, it is it
May 19 06:34:00 <Estel_> ping Pali
May 19 07:43:30 <dm8tbr> Estel_: so ARM employees are spreading fud, huh? did you even bother to read the relevant IRC log?
May 19 07:45:46 <dm8tbr> Estel_: the most elegant way to prove beyond doubt would be if there would be explicit test-cases for the errata involved and it could be tried out in a before/after fashion. So far I don't think those exist.
May 19 07:47:30 <dm8tbr> the fact that 'it just runs' is not proof. I've seen such bugs strike under regular operation once every 24/48/72h or even less often (statistically).
May 19 07:49:41 <ShadowJK> I recall mru tends to spend weeks on working around hw and compiler bugs whenever a new arm chip arrives
May 19 07:49:58 <ShadowJK> so if he gave up on it, I'd take his word on it
May 19 07:51:26 <Hurrian> the technical docs say that the chip /should/ be able to run thumb/thumb-2 binaries. but there's a problem in the chip, so we can't run them properly.
May 19 07:51:39 <Hurrian> a software fix is out, but it doesn't work (???)
May 19 07:52:16 <dm8tbr> ShadowJK: also he specifically said 'tricky' and that it might not be worth it. he didn't rule out the possibility that it could be fixed.
May 19 07:52:45 <Hurrian> other than ubuntu blatantly failing to work properly in a chroot, occasionally rebooting the phone or crashing apt-get upgrades, we're not getting any indication that the bug is actually doing it's buggy thing
May 19 07:53:36 <Hurrian> since this bug happens randomly (but it should happen, given TI acknowledged the fault), it's difficult to make tester for it
May 19 08:00:27 <Hurrian> let's chalk it up to the list of things Nokia didn't do to optimize Linux for the N900
May 19 09:10:00 <Estel_> dm8tbr,  of course I've read logs. Basically, this "arm employe" fail to provide rationale on "how exactly" freemangordon's implemented errata couldn't work.
May 19 09:10:11 <Estel_> it's just like "trust me on that (tm)"
May 19 09:10:40 * dm8tbr headdesks
May 19 09:10:43 <dm8tbr> *plonk*
May 19 09:10:43 <Estel_> trust may work when some noob ask if kp developers could make more ram in N900 ;) but in this case, it's too low for a reason
May 19 09:11:31 <Estel_> thats why freemangordon concluded, how he hates that everyone is saying "it can't work" but fail to provide technical reason
May 19 09:12:21 <Estel_> if YOU think that freemangordon's proposed implementation of two erratas WON'T work, it's YOUR duty to provide reason, not his to prove that he is not a camel (tm)
May 19 09:13:27 <Estel_> up to now, he provided good arguments why kernel and bin errata should work, and why, in case of using those two, 3th errata is irrelevant (flushing whole range by 1st errata, when 3th errata is about flushing part of range)
May 19 09:14:09 <Estel_> now, I don't care if it was said by random arm employe or not - if he fails to provide technical counter-arguments, it's not valid his reasoning, IMO
May 19 09:14:55 <Estel_> any counter-arguments appreciated, instead of plumping sounds and hitting desk ;)
May 19 09:19:55 <Estel_> as I see there is no proper (or any) counter-reasons, I will say something more
May 19 09:20:06 <Estel_> for me, this conversation on *beagle was miserably pathetic
May 19 09:20:19 <Estel_> mru just said "it won't work, stick to arm code"
May 19 09:20:37 <Estel_> while freemangordon's questions about any other erratas etc meet only silence
May 19 09:21:05 <Estel_> "even if You can set secure bit, it's hard or impossible to do it properly"
May 19 09:21:08 <Estel_> "why?"
May 19 09:21:16 <Estel_> "just stick to arm code or get other chip"
May 19 09:21:20 <Estel_> cmon, it's pathetic
May 19 09:21:45 <Estel_> that guy just don't have idea, and instead of saying that stright "sorry, I don't know", hes playing "wise".
May 19 09:22:24 <Estel_> just my 2 cents, as non-developer, following logic, common sense, and overall sticking with meritocratic approach.
May 19 09:22:59 <Estel_> see ya later, I'm very interested how all of this is going to conclude.
May 19 10:28:16 <teotwaki_> Python can be so redundant sometimes...
May 19 10:28:23 <teotwaki_> except Exception as exception:
May 19 12:32:13 <kerio> teotwaki_: you'd use e or something
May 19 12:32:26 <kerio> and expecting Exception is probably stupid
May 19 13:56:49 <DocScrutinizer> I'm all with Estel_ and freemanG about those who claim to know should share their wisdom with freemanG who has proven to be willing to really learn and wasn't like "HLP PLZZ!!". I'm all with dm8tbr regarding we need a proper testbed showing efficiency of erratum bug fix in a 100% reproducible and analyzable way, so dudes like mru can look into it and explain what's still missing
May 19 13:59:07 <Estel_> I'm all with DocScrutinizer about those two sentences ;)
May 19 14:00:30 <Estel_> but, I think You'll agree, that - in case of testbed being extremely hard to prepare (not feasible) due to bug nature, AND and the same time, arguments about invalidity if erratas (that freemangordon used) wrong, we may conclude that observation in "real life" is enough?
May 19 14:01:05 <DocScrutinizer> no
May 19 14:01:30 <Estel_> Well, it seems to me that such practice isn't used against other bugs too
May 19 14:01:32 <DocScrutinizer> there's no such thing like an impossible to prepare testbed
May 19 14:01:40 <Estel_> I hope so.
May 19 14:01:48 <Estel_> and I said "not feasible"
May 19 14:01:58 <Estel_> what I understand from those bits of info
May 19 14:02:35 <Estel_> nature of this bug is so low-level random
May 19 14:02:53 <Estel_> that we would need expensive hardware debuggers to prepare "proper" testbeds
May 19 14:03:24 <Estel_> yet, result of this bug = crashes every few second, that we experience in thumb2 compiled ubuntu on N900 *without* errata
May 19 14:03:36 <DocScrutinizer> if you understand the nature of the bug, you're able to prepare a testbed that will provoke the bug sooner or later, and it will provoke the bug in a way you're closely monitoring and able to read the gdb trace of it later on and spot the one instruction that failed and didn't behave like expected
May 19 14:03:44 <Estel_> ...and, dissapearing (crashes) totally, *with* errata enabled as per freemangordon
May 19 14:04:22 <Estel_> generally, true, and I would love to see this
May 19 14:04:48 <Estel_> but just *in case* of proper testbed requiring ARM hardware cpu tools ;)
May 19 14:05:06 <Estel_> I think it's not feasible to still chase it, *if* it doesn't give any bad results
May 19 14:05:22 <Estel_> thinking like that, we would never have hostmode on N900
May 19 14:06:02 <Estel_> many things there are not "tested" for 100%, and some of them even obviously wrong (idle power consumption), yet, we use it hapilly
May 19 14:06:50 <Estel_> if freemangordon would need to spend 10k $ on equipment to hardware'ish test those low-level things, that TI already did...
May 19 14:07:00 <DocScrutinizer> that's absolute nonsense
May 19 14:07:01 <Estel_> just to confirm that *their* errata is working like they said
May 19 14:07:26 <Estel_> ... just to be 100% sure that we're not hit by bug that doesn't produce any real-life visible errors...
May 19 14:07:34 <Estel_> It would be not feasible, IMO ;)
May 19 14:07:48 <DocScrutinizer> as I understand it this errata are pretty clearly defined and not at all occuring at random
May 19 14:08:36 <Estel_> well, those who say that errata can't work told "us" that freemangordon's result (dissapearing of crashes in thumb2 ubuntu) are not valid
May 19 14:08:40 <Estel_> due to random nature
May 19 14:08:59 <Estel_> what I understand from errata is also *"ot*random nature
May 19 14:09:34 <Estel_> s/"ot*random/*not*random
May 19 14:09:55 <Estel_> but, this "not random" nature was hammered out, as it obviously stopped crashing
May 19 14:10:25 <Estel_> now, we're hunting invisible random bug that doesn't give any problems (!) and it's even unknown how this bug works and why it's random
May 19 14:10:37 <Estel_> cause people claiming it's existence don't want to share it
May 19 14:10:51 <Estel_> with all proportions, it's like abil's "mysterious" fix for N900
May 19 14:11:22 <Estel_> sometimes, reading logs from #beagle and so on, I think that everyone just "heard" about this bug
May 19 14:11:54 <Estel_> I bet that this is just big bullsh|t, due to fact, that previous erratas were *"ot* working, while people using them though that they've worked
May 19 14:11:57 <DocScrutinizer> no, we're testing a deterministic silicon erratum with a basically random testbed
May 19 14:12:05 <Estel_> true
May 19 14:12:13 <Estel_> and You know what? I may be wrong...
May 19 14:12:45 <Estel_> But I think that no one before freemangordon tried to enable this errata for N900 using proper secure monitor or how it's called
May 19 14:12:50 <DocScrutinizer> and the check for efficiency of any patch is "I didn't see anything bad happen" :-S
May 19 14:13:05 <Estel_> = those people though that errata is enabled, yet still encountered random crashes
May 19 14:13:09 <Estel_> -> gave up.
May 19 14:13:23 <DocScrutinizer> quite possible
May 19 14:13:31 <Estel_> then, FUD started to live, that it doesn't work even with errata enabled
May 19 14:14:13 <DocScrutinizer> Estel_: on such delicate topic please check your wording - an erratum can't get enabled
May 19 14:14:22 <Estel_> true.
May 19 14:14:29 <Estel_> thanks for correcting me.
May 19 14:14:40 <Estel_> but I'm sure You get what I mean, i.e. merit
May 19 14:14:45 <DocScrutinizer> sure
May 19 14:15:04 <Estel_> no irony that I appreciate it - others would like to catch me on such details.
May 19 14:15:21 <Estel_> ... and sophism'ish try to use it as counter-reasoning ;)
May 19 14:15:25 <DocScrutinizer> just you won't get atention from thiose we're trying to convince we need their expertise (mru et al) if we sound like noobs
May 19 14:15:34 <Estel_> yep.
May 19 14:15:58 <Estel_> erm, proper sentence would be "errata applied" ?
May 19 14:16:25 <DocScrutinizer> patch / workarround applied
May 19 14:16:31 <Estel_> yep.
May 19 14:16:46 <Estel_> btw, it seems to me that "mru et all" just doesn't care ;)
May 19 14:16:56 <DocScrutinizer> indeed
May 19 14:17:30 <DocScrutinizer> it a set of annoying silicon errata against a chip revision they all never again will touch in whole their life
May 19 14:18:02 <Estel_> ... which bring me to conclusion that if does work (and it obviously does, as thumb2 compiled ubuntu stopped to crash) we could just movde forward...
May 19 14:18:24 <Estel_> and go back to this only if someone encounter bug apperance, OR idea for proper testbed.
May 19 14:20:03 <DocScrutinizer> well, I'll not support any such enterprise as long as no decent story about how the bugfix works is delivered. And hosnestly I'm not interested that much in all this thumb thing either - I mean it's not like it will magically turn N900 into N990
May 19 14:20:27 <DocScrutinizer> there's not a single ticket in tracker that gets closed by changing anything to thumb
May 19 14:20:38 <Estel_> sure thing, but decent story how it works *is* given already
May 19 14:21:13 <Estel_> but we can use thumb2 compiled things with, lets say, Mer kernel (nemo kernel)
May 19 14:21:30 <DocScrutinizer> errr wut?
May 19 14:21:55 <Estel_> let's say, for example, about this overrated ubuntu 12.04 thing
May 19 14:22:11 <DocScrutinizer> you're aware we need to compile any kernel for N900 anyway, and thumb vs arm is a compiler flag, not a property of the kernel c source
May 19 14:22:20 <Estel_> re-comiling this whole sh...t to non-thumb2 is non feasible
May 19 14:22:23 <Estel_> of course
May 19 14:22:34 <Estel_> I mean that with little help from U-boot
May 19 14:22:47 <Estel_> we can set errata properly and use (properly compiled) nemo kernel
May 19 14:22:59 <Estel_> with, for example, Lubuntu 12.04
May 19 14:23:05 <Estel_> without recompiling the latter
May 19 14:23:24 <Estel_> proper nemo kernels for N900  are already available
May 19 14:23:27 <Estel_> and used ;)
May 19 14:23:49 <DocScrutinizer> so how the heck are they related to thumb vs arm then?
May 19 14:23:50 <Estel_> havent tested it myself, but reportedly, without unity bloat, ubuntu works blazingly fast
May 19 14:24:10 <Estel_> ubuntu 12.04 is thumb2 compiled
May 19 14:24:26 <DocScrutinizer> and that's related to mer kernel how?
May 19 14:25:17 <DocScrutinizer> s/mer/nemo/
May 19 14:25:17 <infobot> DocScrutinizer meant: and that's related to nemo kernel how?
May 19 14:26:32 <DocScrutinizer> honestly, pushing THUMB silicon erratum patches into CSSU to enable nemo kernel to enable ubuntu, that's paradox
May 19 14:28:43 <Estel_> Who said to cssu?
May 19 14:28:58 <DocScrutinizer> aiui freemanG already accepted my suggestion to have a sysnode for enabling his thumb patches, in powerkernel. So everybody is free to enable it. I'm not convinced moving *fremantle* stuff to thumb will bring any benefit fo the general masses
May 19 14:29:18 <Estel_> mer kernel is newer version than default Maemo one + it contain some things (modules?) for hardware
May 19 14:29:31 <Estel_> that doesn't work with vanilla kernel of same version number
May 19 14:30:05 <Estel_> btw, errata isn't enabled in kernel-power, as for non thumb compiled code it's only slowdown (yet, not noticeable in real time)
May 19 14:30:41 <DocScrutinizer> Estel_: thanks for explaining that to me - after all I suggested it
May 19 14:30:48 <Estel_> :P
May 19 14:31:05 <Estel_> CSSU doesn't have anything to do with it
May 19 14:31:14 <Estel_> cause only usage for errata is outside Maemo
May 19 14:31:17 <DocScrutinizer> that's one of my points
May 19 14:31:41 <Estel_> erm, haven't You noticed that whole errata saga is related to N900as hardware, not Maemo?;)
May 19 14:32:16 <DocScrutinizer> yet freemanG can't forget his dream to have Qt and GTK and whatnot else turned to thumb, in CSSU
May 19 14:32:50 <DocScrutinizer> that's where whole argument started
May 19 14:34:13 <Estel_> hm
May 19 14:34:30 <Estel_> if it would work faster than without thumb, why not?
May 19 14:34:58 <Estel_> after all, we have cssu -devel for purpose ;)
May 19 14:37:07 <ShadowJK> I'd like to see benchmarks, because iirc at the time omap3 came out and the people in #beagle tried thumb2, once all errata was applied the speed benefit was lost, and they still had random crashes weekly
May 19 14:38:00 <ShadowJK> (and mru wasn't an arm employee at that time, they probably hired him because he was filling up their bugtrackers)
May 19 14:39:20 <DocScrutinizer> :-D
May 19 14:41:31 <ShadowJK> ivwe also seen this convo in #mer / nemo / etc
May 19 14:41:48 <DocScrutinizer> (why not) simply because I bought a device with an OS that's supposed to work error free, and that property somehow been warranted by Nokia at that time. CSSU is meant to continue on that proposition, not introducing a "FMG said 'WFM'" type of scary huge modification, just for... what really now? 
May 19 14:42:45 <DocScrutinizer> and no, cssu-devel also is meant to host projects *meant* for cssu-T and eventually cssu-S
May 19 14:43:22 <DocScrutinizer> hosting thumb Qt there is as nonsensical in my book, as it was to host mer there
May 19 14:46:01 <DocScrutinizer> somebody once decided it's not worth the effort or not rewarding or simply not feasible to have fixed thumb in fremantle. This decission been done at Nokia once, and unless we get a *scientific* approach to prove why we could change that now, and also why we *should* change that now, I don't see the benefit of including any such stuff to CSSU
May 19 14:46:26 <Estel_> well, someone decided at nokia that we wont have hostmode
May 19 14:46:45 <Estel_> yet we have it, despite not being totally *scientific* proper usb compliant approach
May 19 14:46:54 <DocScrutinizer> yes, and I took *scientific* approach to show it can be done nevertheless
May 19 14:47:25 <Estel_> if it will result in like what ShadowJK described - speed benefit lost and random problems, or even just speed benefit lost, of course
May 19 14:47:29 <Estel_> no way to put it
May 19 14:47:34 <DocScrutinizer> it took me ~9 months to collect all the pieces
May 19 14:48:04 <Estel_> if, on the other hand, it would turn out working great, "scientific approach" would be needed to reason why *not* put it there
May 19 14:48:18 <DocScrutinizer> and it's been reviewed and explained to the sincle opcode how it works and that it actually works like that
May 19 14:48:37 <Estel_> of course, but still things like low transfer speed (4 mbps) are strange
May 19 14:48:46 <Estel_> yet, we have hostmode available already
May 19 14:48:57 <Estel_> it would be stupid to wait until we tune all clock values etc
May 19 14:49:14 <DocScrutinizer> Estel_: I'm not going to argue on that basis anymore, it's basically OT
May 19 14:49:52 <Estel_> btw, I agree that no way of putting it into CSSU if it don't work. If it work and gain benefits, just put it. simple as that ;) after all, it's better replacement (deven compile-wise) of current things
May 19 14:49:56 <Estel_> no problem
May 19 14:50:18 <Estel_> btw my personal expectation is that we're not going to see any benefits from thumb2 on Maemo ;)
May 19 14:51:00 <Estel_> it's just strange to me, that this errata and thumb thing rise so much controversy, while other things with more impact on Maemo doesn't.
May 19 14:51:06 <DocScrutinizer> after all hostmode is a dedicated app everybody is free to enable and disable or just don't use it. Thumb-Qt in CSSU is forvefeeding stuff to a targeted custumer base that has no choice at all, except that nasty "you're free to not use CSSU"
May 19 14:51:16 <Jaffa> Estel_: It's ironic that similar people are pushing for thumb in CSSU as also bemoaned the effort into coming up with a good packaging solution for rewrites in CSSU
May 19 14:51:48 <DocScrutinizer> which I promise I'll kill somebody eventually if this argument "your're free not to use..." will get used again
May 19 14:52:00 <Estel_> Jaffa, no idea what You're talking about - I'm not pushing for thumb in CSSU and I don't know anyone who is doing so.
May 19 14:52:36 <Jaffa> Estel_: So no-one's pushing for it, no-one's doing the work and this is an entirely pointless discussion? Goodo.
May 19 14:52:54 <Estel_> well, consensus seems to look like "CSSU is for both things not-applicable via extras AND FOSS rewrites/bugfixesof closed bits"
May 19 14:53:20 <Estel_> Jaffa,  no one is pushing for it, yet, one talented guy is working to check if it gains any benefit for Maemo
May 19 14:53:40 <Estel_> got problem with that? ;) last time I've checked everyone was free to work on whatever she/he want
May 19 14:53:48 <DocScrutinizer> that's exactly NOT the precise thing he does
May 19 14:54:17 <DocScrutinizer> he *assumes* it is beneficial for maemo, and pushing to get Qt-thumb et al into CSSU
May 19 14:54:31 <Estel_> honestly? it is. Have You ever heard that he is planning to put thumb2 and enable errata for CSSU without proving that it's faster/doesn't result in random crashes?
May 19 14:54:36 <merlin1991> gotta love the reoccuring theme on here
May 19 14:54:43 <Estel_> how he is pushing it into CSSU?
May 19 14:55:13 <DocScrutinizer> well, even you did, some posts above
May 19 14:55:18 <Estel_> nope.
May 19 14:55:31 <Estel_> I just said that if it turns out to be working good, why not.
May 19 14:55:45 <Estel_> If it doesn't provide benefits/maked regressions, a no-go.
May 19 14:55:50 <Estel_> simple as that.
May 19 14:55:59 <DocScrutinizer> and "turns out to be working" been up to your discretion to define that
May 19 14:56:06 <DocScrutinizer> not with me
May 19 14:56:07 <Estel_> honestly, I see more problem with replacing kp enter and enter in terminal ;)
May 19 14:56:18 <DocScrutinizer> cheers, I need a coffee now
May 19 14:56:26 <Estel_> thumb2 is just different argument for compilation, yeo?
May 19 14:56:35 <Estel_> yep*what do You want to define here
May 19 14:56:47 <Estel_> working faster/doesn't crash randomly = OK
May 19 14:57:11 <Estel_> no speed advantage OR worse speed OR random bugs - if anything check
May 19 14:57:15 <Estel_> then TRASH
May 19 14:57:30 <Estel_> seriously, CSSU got  more controversial things already
May 19 14:57:50 <Estel_> I just don't get why thumb2 heats up discussion so much
May 19 14:58:40 <Estel_> take Your time, I'll be off soon
May 19 14:59:54 <Estel_> + I'm absolutely sure freeman is reasonable guy and he won't want to include thumb2 into stable CSSU even on smallest signals that it may pose any regressions OR don't provide benefits.
May 19 15:00:23 <Estel_> btw, Jaffa, are You working on something for CSSU, that collides with thumb idea?
May 19 15:02:01 <Jaffa> Estel_: Not that I know of. Just experience with ARM coding and some of the problems switching instruction modes can have.
May 19 15:02:33 <Estel_> I see. All meritocratic reasons are worth to consider
May 19 15:02:49 <Jaffa> Given CSSU doesn't have much organised regression testing the more lower level changes, the greater the risk of a large, unnoticed, regression
May 19 15:03:09 <Estel_> so if You already know about anything that could break or will find it in future, everyone will be glad to hear it
May 19 15:03:19 <Jaffa> Estel_: Of course. But low-level changes are risky; which means the balance of evidence has to be dramatic.
May 19 15:03:19 <Estel_> well, with guys like You and DocScrutinizer testing it religiously
May 19 15:03:27 <Estel_> I doubt anything will get unnoticed ;)
May 19 15:03:32 <Jaffa> Estel_: BTW, English tip - "you" isn't capitalised.
May 19 15:04:13 <Estel_> heh, sure thing, I'm using it as a way of showing respect to someone I'm talking with
May 19 15:04:35 <Estel_> if it's totally wrong in english, I'll drop it, thanks for tip
May 19 15:04:55 <Estel_> I agree about balance of evidence
May 19 15:05:23 <DocScrutinizer> I'm not doing any regression tests
May 19 15:05:32 <Estel_> DocScrutinizer,  kinda joking
May 19 15:06:36 <Estel_> btw what about this "You" versus "you", are, erm, you sure that it is't allowed in english at all (grammar)
May 19 15:06:37 <Estel_> ?
May 19 15:06:52 <Estel_> I always though it's form used in letters etc
May 19 15:07:03 <Estel_> You, He, She, etc
May 19 15:07:04 <DocScrutinizer> and freemanG for sure is no unresponsible fool, it just seems he - just like you - has not much experience with risk management in SW projects
May 19 15:07:05 <Estel_> His
May 19 15:08:04 <Estel_> I agree that I don't have experience in risk management in SW, can't say for freeman. Although, I wouldn't worry unless we have "physical" reason for it (something collapsing)
May 19 15:08:29 <Estel_> DocScrutinizer,  are You naive english speaker?
May 19 15:08:34 <Estel_> what, bullshit question
May 19 15:08:35 <DocScrutinizer> yes ;-P
May 19 15:08:52 <DocScrutinizer> I'm a very naive english speaker
May 19 15:09:00 <Estel_> epic typo
May 19 15:09:30 <Estel_> almost like one I made when I wanted to write to someone "at least You got moral satisfaction"
May 19 15:09:43 <Estel_> and it resulted in "at least You get oral satisfaction"
May 19 15:09:54 <DocScrutinizer> lol
May 19 15:13:37 * ShadowJK imagines hardfloat abi is bigger boost than thumn2+errata anyway
May 19 15:16:02 * DocScrutinizer frowns at Qt using floats at all, figues a performance boost of factor 100 for Qt rewrite using proper fixedpoint arithmetics
May 19 15:16:39 <DocScrutinizer> still I never would suggest to include such stuff into CSSU
May 19 15:17:11 <DocScrutinizer> as *nobody* can figure what's going to happen
May 19 15:19:39 <DocScrutinizer> actually I think Qt used float/real just for two reasons: a) coders always are lazy and generally have no idea about proper fixedpoint, b) on contemporary X86, float is basically free of penalty due to powerful coprocessor
May 19 15:20:01 <DocScrutinizer> b) doesn't hold true on ARM / embedded though
May 19 15:22:43 <Estel_> Intel inside ;)
May 19 15:26:08 <Estel_> DocScrutinizer,  do whole qt rewrite, You'll get both N950 and N9 in both categories ;)
May 19 15:26:38 <Estel_> considering Your calculation of N950 value (R&D / number of N950produced) You'll be a millionare
May 19 15:27:11 <Estel_> erm, IF Jaffa is right about You being totally denied in proper english, it will be hard like hell to get accustomed...
May 19 15:27:29 <Estel_> to not use it. I love those letter'ish forms.
May 19 15:28:11 <DocScrutinizer> while you're at it, here's another one: was / is it you who constantly uses "ho ever" instead of "however"
May 19 15:28:13 <DocScrutinizer> ?
May 19 15:31:15 * DocScrutinizer breaks an old kbd into 104 pieces and rewrites each keycap with a nice "W". Throws the 104 "W" into the channel - "to whom it may concern"
May 19 15:32:41 * DocScrutinizer slaps <self> for constantly usong "dunno" instead of "don't know"
May 19 15:32:51 <DocScrutinizer> using*
May 19 15:39:08 <Estel_> ;)
May 19 15:39:12 <Estel_> yea, it's me
May 19 15:39:23 <Estel_> it's funny, as I though that hoever is proper
May 19 15:39:25 <Estel_> it's funny, as I though that hoever is proper
May 19 15:39:43 <Estel_> yet, spell checker corrected me to ancient form of "ho ever"
May 19 15:39:55 <Estel_> well, need to trust spellcheckers less ;)
May 19 15:40:10 <Estel_> unless planning to play in original shakespare
May 19 15:40:14 * ShadowJK suddenly realizes what he's currently working on is kinda silly
May 19 15:40:32 <Estel_> compared to our englighted discussion?:P
May 19 15:40:49 <ShadowJK> heh
May 19 15:41:42 <Estel_> ShadowJK,  why in Your reswap script You're using 600MB as value for "respawn needed" ?
May 19 15:41:55 * DocScrutinizer HO HO HOOO
May 19 15:42:04 <ShadowJK> because my primary swap was somewhere between 600 and 700 in size
May 19 15:42:10 <ShadowJK> now actually 2 gigs
May 19 15:42:39 <Estel_> ;) just though that using 768 mb, 600 is when fragmentation got chances of starting to bite
May 19 15:43:02 <DocScrutinizer> sounds about right
May 19 15:43:31 <ShadowJK> fragmentation starts right after size of swap actually
May 19 15:43:32 <Estel_> it's kinda funny, as using 768, I - most of the time - "feel" when UI become sluggish, then, using reswap, it's 50-100 MB after 600 treeshold
May 19 15:43:43 <Estel_> only after? can't start before?
May 19 15:43:51 <ShadowJK> i dont think so
May 19 15:44:16 <Estel_> I feel it ~700 MB of 768 swap used, sometime, little less, and fopr sure it isn't placeb
May 19 15:44:19 <Estel_> strange
May 19 15:44:23 <Estel_> placebo*
May 19 15:44:34 <DocScrutinizer> AIUI swap writes are strictly sequential until whole swap space got "used up"
May 19 15:45:15 <Estel_> i.e. written at least once during swap "life"?
May 19 15:45:24 <DocScrutinizer> where it starts to re-use formerly used and then freed swap chunks -> fragmentation
May 19 15:45:36 <Estel_> I see
May 19 15:45:38 <ShadowJK> Currently I'm rewiring a "fixed with chewing gum, duct tape and metal wire" system of 30kW in power, only I found no proper cable so I'm using some scavenged cable I found that viosually looks like it *might* handle the power. It feels silly replacing a half-arsed system with slightly less half-arsed
May 19 15:46:33 <Estel_> heh. What is using those 30kW?
May 19 15:46:34 <DocScrutinizer> ShadowJK: sounds like you're at work ;-P
May 19 15:47:19 <ShadowJK> yep
May 19 15:47:47 <DocScrutinizer> going to have another nice fireworks?
May 19 15:47:54 <DocScrutinizer> ;-D
May 19 15:48:23 * DocScrutinizer foggily recalls some talk about IBGTs
May 19 15:48:36 <ShadowJK> SSRs
May 19 15:48:54 <ShadowJK> actually I just threw out 6 40 Amp omron SSRs
May 19 15:49:00 <DocScrutinizer> motor? or heating?
May 19 15:49:05 <ShadowJK> heating
May 19 15:49:40 <ShadowJK> going to be 8 Siemens contactors driven with 20second "pwm" duty cycle now 
May 19 15:51:05 <DocScrutinizer> well, depending on physical I parameter of that PID system, a 20s PWN is/might_be absolutely sane - reducing number of cycles that may wear the circuitry
May 19 15:51:16 <ShadowJK> These are one of the few SSRs that actually survived here
May 19 15:51:26 <DocScrutinizer> hehe
May 19 15:51:42 <ShadowJK> But that's because they were protected by some "hidden" fast-blow mcbs that were forgotten at some point
May 19 15:52:03 <ShadowJK> because there was a second set of slow blow mcbs after those...
May 19 15:52:17 <DocScrutinizer> wow, strange design
May 19 15:52:19 <DocScrutinizer> :-)
May 19 15:53:04 <ShadowJK> When old regulation system died a new one was just hurriedly slapped ontop of old thing
May 19 15:53:23 <DocScrutinizer> usual practice
May 19 15:53:59 <DocScrutinizer> idling machinery costs more than running machinery
May 19 15:54:05 <ShadowJK> ya
May 19 15:54:18 <ShadowJK> why im doing it on saturday 7pm
May 19 15:56:30 <DocScrutinizer> Estel_: that's the type of thinking that teaches how to maintain huge systems. When your little patch has any potential to stop a factory with 3000 employees for a whole day, you're not suggesting "let's try what happens when we replace Qt by a thumb compiled version"
May 19 15:57:02 <DocScrutinizer> same for a sw system that's used by 1000s of users
May 19 15:57:31 <DocScrutinizer> some of them maybe doctors, firemen, wallstreet brokers, whatever
May 19 15:57:54 <DocScrutinizer> 10% speed gain simply doesn't matter
May 19 15:58:19 <DocScrutinizer> if such huge unpredictable risk is the downside of the equation
May 19 15:59:36 <DocScrutinizer> you not even start thinking about such move, unless you got proper testbed that - on a scientific basis - proves your patch is OK and rock solid
May 19 15:59:49 <DocScrutinizer> all this rationale doesn'T apply to hostmode at all
May 19 16:00:52 <DocScrutinizer> still we (I avid the "I" here) invested any reasonable effort to understand and explain and trace on machine level what's going on
May 19 16:00:58 <DocScrutinizer> avoid*
May 19 16:01:38 <DocScrutinizer> and it's no surprise ShadowJK been one of the valuable contributors in that effort
May 19 16:03:35 <ShadowJK> you can call me Mr Luddite ;-)
May 19 16:09:42 <DocScrutinizer> Estel_: I was able to *guarantee* there's no possible hw impact with using hostmode *beta*, as well as no possible incorrect usage that would cause any harm to either N900 or attached device (besides the inevitable always existing threats involved with unplugging mounted storage)
May 19 16:10:21 <DocScrutinizer> for any ARM->THUMB adventure, we can actually guarantee nuttin
May 19 16:11:01 <DocScrutinizer> except huge headache in future bug triage, as _all_ future bugs have one more possible cause: this very thumb migration
May 19 16:13:18 <DocScrutinizer> while benefit is what exactly? Allegedly improved speed? More free space in root for what exactly?
May 19 16:14:38 <kerio> what's thumb?
May 19 16:14:51 <DocScrutinizer> hell, even improved speed theoretically can introduce new bugs showing up, that previously simply were covered by the system sluggishness
May 19 16:14:52 <kerio> also more space in root to remove optification
May 19 16:15:12 <DocScrutinizer> kerio: you can't remove optification
May 19 16:15:14 <kerio> :c
May 19 16:15:18 <kerio> but i really want to!
May 19 16:15:37 <DocScrutinizer> you can tweak your system to what I suggested in
May 19 16:15:41 <DocScrutinizer> ~optification
May 19 16:15:41 <infobot> optification is a inventive duct tape workaround to reclaim space in fs root, done due to the fact the systeminit *and* partitioning is FUBAR,  http://wiki.maemo.org/Documentation/Maemo_5_Developer_Guide/Packaging,_Deploying_and_Distributing/Installing_under_opt_and_MyDocs, or ""OMG - I wish they looked into FHS and moved /usr to eMMC"", http://www.pathname.com/fhs/pub/fhs-2.3.html#PURPOSE2 bullet1,2 and fhs-2.3.html#PURPOSE16 dot3"
May 19 16:15:47 <DocScrutinizer> some time ago
May 19 16:15:54 <DocScrutinizer> ~factinfo optification
May 19 16:15:55 <infobot> optification -- created by DocScrutinizer <~jr@openmoko/engineers/joerg> at Mon Mar 15 10:29:39 2010 (796 days); last modified at Wed Nov 30 04:17:02 2011 by DocScrutinizer!~halley@openmoko/engineers/joerg; it has been requested 98 times, last by DocScrutinizer, 13s ago; it has been locked by DocScrutinizer.
May 19 16:16:53 <DocScrutinizer> you can do that at the expense of making your system incompatible to any NOKIA OTA update
May 19 16:17:29 <DocScrutinizer> as all the official repos and fiasco images rely on optification being in place
May 19 16:18:22 <DocScrutinizer> I hoped for Nokia doing the sane cut on PR1.2 and again on PR1.3 to replace optification by what I suggested up there
May 19 16:18:25 <DocScrutinizer> they didn't
May 19 16:18:29 <DocScrutinizer> risk management
May 19 16:20:08 <kerio> :/
May 19 16:22:27 <DocScrutinizer> now we're stuck with binaries living in /usr/[s]bin which SHALL live in /[s]bin, and thus /usr has to live in rootfs rather than own partition, where optification kicks in and symlinks each binary NOT needed at early boot time form /usr/[s]bin/* to /opt/*/[s]bin/*
May 19 16:23:29 <DocScrutinizer> definition of /usr -> all stuff NOT needed during early boot. Nokia ignored that
May 19 16:24:23 <Jaffa> Estel_: FWIW, "You" in English is only ever capitalised at the start of sentence. "He" tends to be capitalised when being religious and talking about God. It might also be capitalised in contracts if it's defined at the start. But never in conversational English, or even business letters.
May 19 16:24:29 <kerio> yeah, i've read the FHS :9
May 19 16:24:30 <kerio> :)
May 19 16:24:53 <DocScrutinizer> it's up to *maintainer* of distro to move appropriate stuff from /usr/sbin to /sbin
May 19 16:25:29 <DocScrutinizer> it's not like devels of XY binary decided for all times it has to live in /usr/sbin
May 19 16:25:44 <DocScrutinizer> Nokia never got the catch on that
May 19 16:26:18 <kerio> DocScrutinizer: the thing is, most stuff that installs by default in /usr/ should probably not be needed during boot
May 19 16:26:32 <kerio> i mean, we need to load pulseaudio just for the fucking mr hands video
May 19 16:27:03 <DocScrutinizer> so the *real* optification whould've moved stuff from /usr/sbin to /sbin, and then moved /usr to eMMC and mount it
May 19 16:27:13 <DocScrutinizer> kerio: exactly
May 19 16:27:34 <kerio> fucking nokia
May 19 16:27:41 <DocScrutinizer> the devel introducing that hands-video whould've been responsible to make PA move from /usr to /
May 19 16:27:47 <kerio> oh well, nokia is dead anyway
May 19 16:28:09 <kerio> no, the devel introducing that hands video should've been shot and buried in unholy ground
May 19 16:28:20 <DocScrutinizer> ack ;-P
May 19 16:28:21 <kerio> but yeah, as a second option he should've moved pulseaudio to /[s]bin
May 19 16:30:15 <kerio> hell, why didn't nokia realize that only linux nerds would've bought the n900 willingly?
May 19 16:30:53 <DocScrutinizer> Jaffa: (re proper English) what's the (etymological) rationale for capital "I" then?
May 19 16:31:16 <kerio> DocScrutinizer: a lowercase "i" is the imaginary unit
May 19 16:31:25 <kerio> and/or the index of a loop
May 19 16:31:39 <kerio> clearly we need to avoid ambiguity :P
May 19 16:31:53 <Jaffa> DocScrutinizer: Anyone who claims English is sensible or logical doesn't know it ;-)
May 19 16:32:04 <Jaffa> DocScrutinizer: IOW, no idea. Just "is". I'm not a linguist, unfortunately
May 19 16:32:10 <DocScrutinizer> Jaffa: that's for sure :-D
May 19 16:32:31 <kerio> DocScrutinizer: anyway, we've complained about optification many times
May 19 16:32:39 <kerio> to nobody in particular, but still
May 19 16:32:44 <DocScrutinizer> yep
May 19 16:32:50 <DocScrutinizer> this ship is gone
May 19 16:33:12 <DocScrutinizer> unless we bring it back in CSSU next month ;-P
May 19 16:33:16 <kerio> the real thing i'm mad about is that there's no good modern linux pocket computer
May 19 16:33:43 <kerio> even if i bought another n900 (mine is about to die) it would still have too little ram and too little cpu
May 19 16:34:54 <DocScrutinizer> kerio: you *might* consider it a challenge to get nice stuff running despite (or even due to) low memory and little CPU grunt
May 19 16:35:29 <DocScrutinizer> as bloatware rarely ever is the solution to demand for decently working apps
May 19 16:35:58 <kerio> yeah, but it still sucks :(
May 19 16:36:17 <kerio> openpandora + dumbphone is ridiculous
May 19 16:36:32 <kerio> although it could actually be sensible
May 19 16:36:40 <DocScrutinizer> kerio: without sarcasm, could you think again and then explain what exactly sucks, in terms of UX
May 19 16:37:45 <DocScrutinizer> ok, HAM is slow as hell
May 19 16:37:52 <kerio> i suppose i can't actually talk though
May 19 16:37:56 <kerio> i'm stuck on PR1.2
May 19 16:38:01 <DocScrutinizer> phone-ui sometimes laggy
May 19 16:38:29 <kerio> and i actually love the UX
May 19 16:38:45 <kerio> (apart from HAM, yeah)
May 19 16:39:06 <kerio> (but then again, who uses HAM?)
May 19 16:39:22 <DocScrutinizer> though I found a LOT of phone-ui lag is introduced by rotation and lately (on my system) by trackerd starting indexing as soon as recaller is writing the call .wav
May 19 16:39:36 <kerio> ooooh right
May 19 16:39:38 <kerio> forgot about that
May 19 16:39:40 <Sicelo> w/in 25
May 19 16:39:45 <kerio> tracker is a fucking stupid idea
May 19 16:39:45 <Sicelo> sorry.
May 19 16:40:16 <DocScrutinizer> indeed
May 19 16:41:13 <DocScrutinizer> a fsckng stupid idea nd concept they perfected in HARM
May 19 16:41:38 <kerio> DocScrutinizer: it would be awesome on a quadcore with 8gb of ram
May 19 16:41:45 <DocScrutinizer> no
May 19 16:42:25 <DocScrutinizer> I hate the whole concept of replacing classical pathnames by a flat relational database that's not even integrated into filesystem
May 19 16:43:22 <DocScrutinizer> HECk - WHAT FOR we _got_ a filesystem when trackerd is set up to eliminate it
May 19 16:44:00 <kerio> but iphone!
May 19 16:44:06 <kerio> and ipad!
May 19 16:44:26 <kerio> i suppose that tracker is ok if you only make it index a single directory
May 19 16:44:38 <DocScrutinizer> that's a concept done by somebody trying to mimic apple while not even understanding the basics about it - completely ignoring this whole concept never panned out on iPhone as well
May 19 16:44:52 <kerio> well, it pans out quite nicely for music
May 19 16:45:03 <DocScrutinizer> not even there
May 19 16:45:05 <kerio> less so for videos and pictures, i guess
May 19 16:45:13 <kerio> and OH GOD WHY so for generic files
May 19 16:45:30 <DocScrutinizer> yup
May 19 17:09:54 <Estel_> Jaffa,  thanks a lot for explanation - will need to get used to it ;)
May 19 17:10:30 <Estel_> I can't recall why I always though You is a polite way in letters
May 19 17:11:15 <Estel_> DocScrutinizer,  historically, bot You and I (+ He She) was used by capital letters, thats why I've mentioned shakespare
May 19 17:11:30 <Estel_> it still live when talking religiously about god as jaffa pointed
May 19 17:11:43 <Estel_> well, religy doesn't like changes
May 19 17:12:09 <Estel_> it was a concept of great respect for someone
May 19 17:12:17 <Estel_> in many languages it still live
May 19 17:12:37 <Estel_> i.e in polish you is "ty", but, in letters, we still use "Ty"
May 19 17:12:46 <Estel_> not in common writing of course
May 19 17:13:19 <Estel_> DocScrutinizer,  as for testbeds erc, fortunatelly, we're not nokia and instead of risk management, we may use common sense :P
May 19 17:13:56 <Estel_> any other drawback from doing optification as You suggested, besides nokia OTI breakage?
May 19 17:14:03 <Estel_> OTA*
May 19 17:15:23 <Estel_> BTW, as for tracker and recaller, it may be good idea to exclude recording's catalog from tracker via tracker cfg
May 19 17:15:49 <Estel_> my recaller recordings are written to truecrypt partition, that is excluded as a whole
May 19 17:15:53 <DocScrutinizer> yup
May 19 17:16:03 <Estel_> but even simple directory filter should do the job
May 19 17:16:08 <DocScrutinizer> my trackerd.cfg is fsckdup
May 19 17:16:09 <Estel_> ok, now little request
May 19 17:16:12 <Estel_> ;)
May 19 17:16:20 <Estel_> Could You please take look at:
May 19 17:17:03 <Estel_> goddamn, something emptied my clipboard
May 19 17:17:28 <Estel_> fuck that, I need to refresh zerobin page @ 10 kbps connection
May 19 17:17:51 <Estel_> thanks godness it's not bloated
May 19 17:17:53 <Estel_> http://sebsauvage.net/paste/?a2f700981cc87f88#WyeBRNt5PAn+VUg32YDuWeTpzVmX+q6pR+OB5tL6r3g=
May 19 17:18:03 <Estel_> ...and tell me if it's ash or bas compliant?
May 19 17:18:17 <DocScrutinizer> wrt "common sense" - "we" need to take care of the majority of devels that aren't even aware of the term and concept. This mindset of those usually results in statements like "we need to put this into CSSU, as it's the right thing and it's bleeding edge and it would go unnoticed when we don't"
May 19 17:19:21 <DocScrutinizer> Estel_: I'm not a living bash, so I have a hard time to evaluate if a script is bash or ash compliant
May 19 17:20:06 <Estel_> for me, if initial testings show no regressions AND decent gain, cssu-devel is good testbed, and cssu-testing later. BUT we don't have any candidate now, so we're sharing wild boar's skin :P
May 19 17:20:09 <Estel_> ok, sorry
May 19 17:20:24 <Estel_> it's just that this simple script doesn't work for me and I can't find reason
May 19 17:20:34 <Estel_> and by "doesn't work" I really mean it
May 19 17:20:38 <Estel_> no error messages
May 19 17:20:44 <Estel_> no exit codes
May 19 17:20:46 <Estel_> nothing
May 19 17:20:56 <Estel_> it's a wrapper for "eject" program
May 19 17:21:51 <Estel_> my skill are low so I could miss somethimg, but I can't any place where it should fail
May 19 17:22:05 <Estel_> now, executing it just starts new line in terminal
May 19 17:22:16 <DocScrutinizer> Estel_: looks like ash to me
May 19 17:23:20 <DocScrutinizer> consider >>set -vx<<
May 19 17:25:59 <DocScrutinizer> Estel_: also check true content of all your ENV vars you use there
May 19 17:26:22 <DocScrutinizer> some of them generated by grep etc might not look like you expect them to do
May 19 17:28:23 <Estel_> yea, checked that
May 19 17:28:35 <Sicelo> i'm sure you can also re-write this line to only use 'grep' instand of also using cat .. (cat /proc/driver/musb_hdrc | grep Mode=) 
May 19 17:28:38 <Estel_> also ensuredd that all commands parsed to "eject" works when done by hand
May 19 17:28:49 <Estel_> Sicelo, thanks
May 19 17:29:04 <Estel_> I just dont get one thing
May 19 17:29:16 <Sicelo> btw, this is N900? 
May 19 17:29:25 <Estel_> IF grep etc would give unexpected or non-existent results...
May 19 17:29:28 <Estel_> yes
May 19 17:29:46 <Estel_> then, eject would be called with wron/lack of values and complain
May 19 17:29:54 <Estel_> instead of this, I just get new line
May 19 17:30:09 <Estel_> in terminal. No errors no "fuck You" no anything
May 19 17:30:15 <ShadowJK> grr. I have phase to phase continuity of 40R where there's supposed to be none. I spotted a plastic junction box deformed from heat.. Good thing I measured stuff before trying to flip on power
May 19 17:30:42 <DocScrutinizer> Estel_: add >set -vx< after shebang
May 19 17:31:09 <DocScrutinizer> ShadowJK: OUCH
May 19 17:31:13 <Estel_> Sicelo, of course it's not my script, unfortunately. Blue_led wrote that, but hes absent. I asked about it a week ago without reply
May 19 17:31:30 <DocScrutinizer> ~220** / 40
May 19 17:31:31 <Estel_> DocScrutinizer,  erm
May 19 17:31:46 <DocScrutinizer> ~220**2 / 40
May 19 17:31:46 <infobot> 1210
May 19 17:32:11 <Estel_> set what? remember, it's not my script, I'm trying to fix it
May 19 17:32:20 <DocScrutinizer> set -vx
May 19 17:32:29 <DocScrutinizer> as 2nd line
May 19 17:32:33 <Estel_> please write in simple sentences what to replace with what, and, prefferably, why, so I can learn
May 19 17:32:36 <DocScrutinizer> after shebang line
May 19 17:32:40 <Estel_> and do it myself next time :P
May 19 17:32:51 <Estel_> ouh, literally
May 19 17:33:01 <Estel_> I though You're using mental shortcut
May 19 17:33:17 <DocScrutinizer> #!/bin/sh
May 19 17:33:19 <DocScrutinizer> set -vx
May 19 17:33:20 <DocScrutinizer> banner(){
May 19 17:34:10 <Estel_> Yea, did that, What it's supposed to do?
May 19 17:34:25 <DocScrutinizer>  -v  Print shell input lines as they are read.
May 19 17:34:32 <DocScrutinizer>  -x  Print commands and their arguments as they are executed.
May 19 17:35:16 <Sicelo> Estel_: so where can i also get 'eject' ?
May 19 17:35:20 <DocScrutinizer> help set
May 19 17:35:38 <Estel_> DocScrutinizer,  thanks a lot!
May 19 17:35:41 <DocScrutinizer> yw
May 19 17:35:43 <Estel_> Sicelo, wait a second
May 19 17:36:57 <Estel_> Sicelo, normally, I would give You a direct link, but I'm on ~10kbps connection, so wouldn't You mind if I give Your "verbal" instructions?
May 19 17:37:10 <Estel_> go to tmo and write "cdrom eject" in search box
May 19 17:37:16 <Estel_> first result should be it
May 19 17:39:38 <Estel_> Sicelo,  few days ago I've tested optical media reader and hostmode, of course it worked great, and eject program is nice addition, helps in automating things
May 19 17:39:43 <Sicelo> ol. ok. thanks :p
May 19 17:39:52 <Sicelo> *lol
May 19 17:40:03 <Estel_> blue_led script should give it icing on top
May 19 17:40:33 <Estel_> I planned to upgrade this script to provide more functions of eject, but first, I need to make it working at all:P
May 19 17:41:27 <Estel_> now I noticed that while in peripheral mode, it gives me banner "peripheral mode", but in hostmode AND optical media attached, it just ended with nothing (while executing commands by hand worked)
May 19 17:41:36 * Sicelo wanted 'eject' to eject the fake cd-rom in a zeroconf usb modem
May 19 17:41:57 <DocScrutinizer> Estel_: 
May 19 17:42:00 <DocScrutinizer> t900:~# [ $e ] && echo 1
May 19 17:42:02 <DocScrutinizer> 1
May 19 17:42:03 <DocScrutinizer> t900:~# e=0
May 19 17:42:05 <DocScrutinizer> t900:~# [ $e ] && echo 1
May 19 17:42:06 <DocScrutinizer> 1
May 19 17:42:40 <DocScrutinizer> I.E >[ $e ]< is wrong
May 19 17:42:59 <Estel_> I see...
May 19 17:43:38 <DocScrutinizer> use >[ "1" == $e ]
May 19 17:44:21 <Estel_> replace
May 19 17:44:22 <Estel_> [ $e ] && exit
May 19 17:44:25 <Estel_> with
May 19 17:44:45 <Estel_> [ "1" == $e ] && exit
May 19 17:44:45 <Estel_> ?
May 19 17:44:57 <DocScrutinizer> t900:~# e=0
May 19 17:44:58 <DocScrutinizer> t900:~# [ "1" == $e ] && echo true
May 19 17:45:00 <DocScrutinizer> t900:~# e=1
May 19 17:45:01 <DocScrutinizer> t900:~# [ "1" == $e ] && echo true
May 19 17:45:03 <DocScrutinizer> true
May 19 17:45:18 <DocScrutinizer> so yep
May 19 17:46:09 <Estel_> allright, testdriving...
May 19 17:46:32 <DocScrutinizer> set -vx would have shown you immediately
May 19 17:46:35 <DocScrutinizer> ;-)
May 19 17:48:54 <Estel_> used itp and original version ended with
May 19 17:48:59 <Estel_> e=1
May 19 17:49:16 <Estel_> [ $e ] && exit
May 19 17:49:27 <Estel_> [ 1 ]
May 19 17:49:29 <Estel_> exit
May 19 17:49:44 <Estel_> modified one ends with
May 19 17:49:56 <Estel_> e=1
May 19 17:50:06 <Estel_> [ "1" == $e ] && exit
May 19 17:50:17 <Estel_> [ 1 == 1 ]
May 19 17:50:19 <Estel_> exit
May 19 17:50:35 <Estel_> I mean output when executing with -vx
May 19 17:51:14 <Estel_> whats the difference between 1 and 1== 1? :P my non-coder lack of skill is showing
May 19 17:52:16 <Estel_> well, I'll test it later with usb optical drive - I know, it's dumb that I've asked aboout it without optical with me, but I though script will contain obvious errors (which seems to be the case)
May 19 17:52:24 <DocScrutinizer> Estel_: obviously it sets e=1, so it's meant to exist on [ 1 == $e ]&&exot
May 19 17:52:31 <DocScrutinizer> exit* :-P
May 19 17:53:52 <Estel_> I trust You on that (tm) :P
May 19 17:54:41 <DocScrutinizer> Estel_: so just one of those lines executes e=1
May 19 17:54:44 <DocScrutinizer> *Peripheral*)banner "Peripheral mode";e=1;;
May 19 17:54:45 <DocScrutinizer> *Host*)[ ! -e /dev/sr* ] && ( banner "Host mode detected but no optical drive attached";e=1);;
May 19 17:55:20 <Estel_> ...and causes exit. got it
May 19 17:57:21 <Estel_> btw, why script allowed him to define "e" as 0 or 1 inside first part, and later, as /usr/sbin/eject"?
May 19 17:57:21 <DocScrutinizer> in any case function banner() should have printed a notification
May 19 17:57:48 <DocScrutinizer> unless it failed, e.g due to missing run-standalone.sh
May 19 17:57:52 <Estel_> hm, without host mode it prints notification, but when I've tried it with real disk attached, it did nothing
May 19 17:57:59 <Estel_> ouh
May 19 17:58:03 <Estel_> quite possible
May 19 17:58:23 <DocScrutinizer> Estel_: (why did it allow) shell env are usually not strict types
May 19 17:59:07 <DocScrutinizer> Estel_: this script needs to be run under user
May 19 17:59:20 <Estel_> with sudo
May 19 17:59:42 <DocScrutinizer> still I guess a >run-standalone.sh dbus-send --system --type=method_call \< won't hurt
May 19 18:00:02 <Estel_> I see
May 19 18:00:11 <DocScrutinizer> no sudo
May 19 18:00:16 <DocScrutinizer> run it as user
May 19 18:00:16 <Estel_> ?!?
May 19 18:00:26 <Estel_> original package provided sudoers.d file
May 19 18:00:35 <Estel_> and tried to run it as sudo
May 19 18:00:37 <DocScrutinizer> sure, for eject
May 19 18:00:48 <DocScrutinizer> that's idiotic
May 19 18:01:13 <Estel_> user ALL = NOPASSWD: /usr/sbin/eject.sh
May 19 18:01:28 <DocScrutinizer> third and second last line use sudo - the script itself needs to run as user
May 19 18:01:38 <Estel_> blame blue_led, I'm just stating how it looked
May 19 18:01:43 <Estel_> hm
May 19 18:01:59 <Estel_> in fact sudoers file is quite verbose
May 19 18:03:43 <DocScrutinizer> I also frown at >dbus-send --system< - afaik you should contact org.freedesktop.Notifications on session dbus
May 19 18:04:11 <DocScrutinizer> not all sure about it though
May 19 18:04:26 <DocScrutinizer> you might wnat to check back at phone control wikipage
May 19 18:05:09 * Estel_ is uploading something to zerobin, which take ages on current connection
May 19 18:05:48 <Sicelo> btw, what do you want 'eject' for, when a 'umount' would do just as well, afaict?
May 19 18:05:49 * DocScrutinizer is afk, for some RL
May 19 18:05:52 <DocScrutinizer> sorry
May 19 18:06:19 <DocScrutinizer> eject does more than umount
May 19 18:06:28 <Estel_> Sicelo,  usb drives without button (relying on software eject function)
May 19 18:06:34 <Estel_> the slim, lightweight type
May 19 18:06:35 <DocScrutinizer> it also does "safely remove"
May 19 18:06:39 <Estel_> DocScrutinizer,  np
May 19 18:06:43 <Estel_> thanks a lot
May 19 18:06:49 <Estel_> + safely remove
May 19 18:06:51 <Estel_> yea
May 19 18:07:19 <Sicelo> ah, i forgot you were talking about optical drives :D
May 19 18:07:52 <Estel_> + eject got in fact many other nice functions ;) also for other media types (ancient floppies :P)
May 19 18:08:37 <Estel_> http://sebsauvage.net/paste/?8e7e886a078a7a24#vdqYUEeNJJBMYT7eHOzHj71aK5zp9qlTa6AFQNkkxD0=
May 19 18:08:39 <Estel_> sudoers file
May 19 18:08:50 <Estel_> unnecessary elevated privileges for eject.sh
May 19 18:08:50 <Estel_> ?
May 19 18:09:01 <Estel_> if only eject needs it and have granted?
May 19 18:13:13 <DocScrutinizer> yes
May 19 18:15:05 <Estel_> Will test it, re-write as per Your instructions, then publish fixed version... Of course with proper credits to real author of fixed lines ;)
May 19 18:15:15 <Estel_> I mean "real author of fix"
May 19 18:15:24 <Estel_> for broken lines
May 19 18:15:34 <Estel_> thanks a lot
May 19 18:15:56 <Estel_> I wonder if blue_led never used his own script, or mysteriously it worked for him, lol
May 19 18:16:00 <Sicelo> lol.. i had even thanked that blue_led post with eject, but now i didn't even know it existed :P
May 19 18:16:11 <Estel_> heh
May 19 18:16:23 <Estel_> what is Your tmo nick?
May 19 18:17:06 <Estel_> I wonder if kp contain modules for writing optical media
May 19 18:17:13 <Sicelo> qhubekela ;)
May 19 18:17:26 <Estel_> ancient versions contained it, but later, all cdrom related stuff got mysteriously removed
May 19 18:17:32 <Estel_> thanks, lol
May 19 18:17:38 <Sicelo> used to be my nick here as well, unti started to use my real name
May 19 18:17:52 <Estel_> better choice, IMO
May 19 18:18:40 <Estel_> would just love to write DVD with some data from N900
May 19 18:18:46 <Estel_> just love it :P
May 19 18:19:03 <Estel_> and later, blueray :P
May 19 18:19:29 <Sicelo> writing was supported before? :O
May 19 18:19:31 <Estel_> seriously, not only for sake of it
May 19 18:19:33 <DocScrutinizer> KP50 should have all iso9660 et al kernel modules again
May 19 18:19:37 <DocScrutinizer> afaik
May 19 18:19:41 <Estel_> yea
May 19 18:19:44 <DocScrutinizer> if not, spank pali
May 19 18:20:01 <Estel_> The problem - I've no idea how linux writing via command line goes
May 19 18:20:05 <Sicelo> yay! /me kisses his N900 :D
May 19 18:20:28 <Estel_> would need to search for informations how to even start :P
May 19 18:20:39 <Estel_> good to share files when pendrives are not around
May 19 18:21:36 <Estel_> many times You just throw things on dvd disc -as I'm using N900 as real computer with me, it would be great
May 19 18:22:17 <Estel_> now I'm only waiting for Rpi to arrive, and dream of connecting N900 via hdmi out to monitors and TVs will become true :D
May 19 18:22:28 <Estel_> with decent resolution
May 19 18:23:12 <Estel_> Sicelo,  do You remember how to test writing from command line? (optical)
May 19 18:23:32 <Sicelo> i'm worse than you, lol
May 19 18:23:43 <Estel_> btw I wonder if writing tools (also GUI) from ed would have proper access to modules... They should, yep Doc?
May 19 18:24:04 <Estel_> seriously, Sicelo?I haven't noticed and though otherwise
May 19 18:24:14 <Estel_> (no irony)
May 19 18:25:02 <Sicelo> hmm, not so sure i understand you there. otherwise, what i meant is that i'm noob, in the real sense of the word
May 19 18:25:24 <Estel_> ...and I've written that I though I'm more noob that You ;)
May 19 18:25:25 <DocScrutinizer> Estel_: hm? writing tools from ed?
May 19 18:25:37 <Estel_> tools for writing things on optical disc
May 19 18:25:42 <Estel_> in Easy Debian
May 19 18:25:46 <DocScrutinizer> aah
May 19 18:25:51 <DocScrutinizer> easydeb
May 19 18:25:57 <Estel_> or to backup&write them like EAC or laserdisk
May 19 18:26:12 <DocScrutinizer> I'm not sure about ED
May 19 18:26:39 <Estel_> well, generally cups works with USB printers
May 19 18:26:45 <Estel_> easy debian cups
May 19 18:26:55 <DocScrutinizer> but generally you need root access for all those tools - unless your physical device in /dev/* got user with full privileges
May 19 18:26:55 <Estel_> so why optical media should be different?
May 19 18:27:25 <Estel_> yea
May 19 18:27:34 <DocScrutinizer> simple as that: ls -l /dev/sr*
May 19 18:27:50 <Estel_> in easy debian, of course?
May 19 18:27:54 <DocScrutinizer> even stat /dev/sr*
May 19 18:28:00 <Estel_> if it sees it it can use it
May 19 18:28:03 <Estel_> understood
May 19 18:28:15 <DocScrutinizer> from wherever you want to use those writing-tools
May 19 18:28:33 <Estel_> ssh :P
May 19 18:28:41 <Estel_> joking
May 19 18:29:36 <Estel_> I just wonders why things that suffer from low memory (thus swapping) when run as easy debian chrot doesn't suffer so much when run as native ubuntu on N900
May 19 18:29:49 <Estel_> as long as You get rid of unity
May 19 18:29:55 <Estel_> i.e. browsers
May 19 18:30:23 <Estel_> reportedly, chromium and iceweasel works with speed of microB on (thumb2 :P) ubuntu @ N900
May 19 18:30:38 <Estel_> (joke with thumb2, I'm sure it's not the reason)
May 19 18:31:04 <Estel_> ubuntu isn't less memory hungry than maemo
May 19 18:31:21 <Estel_> I would say the opposite
May 19 18:31:24 <Estel_> so wtf?
May 19 18:31:27 <Sicelo> Estel_: if someone cold convince DocScrutinizer and everyone else that worked on h-e-n to also bring it on to Ubuntu/Debian, whatever ..
May 19 18:31:35 <Sicelo> *could
May 19 18:31:36 <Estel_> chroot debian is quite native
May 19 18:32:11 <Estel_> erm, Sicelo, I may be wrong, but what's the problem with running hostmode on ubuntu/debian?
May 19 18:32:29 <Estel_> modules and scripts for hostmode are same
May 19 18:32:39 <Estel_> h-e-n itself is just gui running scripts
May 19 18:32:50 <DocScrutinizer> Sicelo: h-e-n basically is hostmode patches in PK
May 19 18:32:59 <Estel_> same scripts should work under, lets say, ubuntu 12.04
May 19 18:32:59 <DocScrutinizer> the gui is a nobrainer
May 19 18:33:09 <Estel_> exactly
May 19 18:33:26 <Estel_> using KP in ubuntu OR Nemo kernel with hostmode patches added...
May 19 18:33:29 <Estel_> and Y
May 19 18:33:44 <DocScrutinizer> lol
May 19 18:33:44 <Estel_> + scripts and You have hostmode
May 19 18:34:13 <DocScrutinizer> thought you've hit return instead backspace as you tried to delete the Y
May 19 18:34:17 <DocScrutinizer> ;-P
May 19 18:34:33 <Estel_> true, almost
May 19 18:34:34 <Sicelo> does Nemo's kernel include those patches yet?
May 19 18:34:48 <Estel_> no idea.
May 19 18:35:05 <DocScrutinizer> I'd be really surprised
May 19 18:35:11 <Estel_> but no big deal to include them for someone who can compile kernel already
May 19 18:35:30 <DocScrutinizer> hostmode hacks are really really nasty botch, not upstream-able
May 19 18:35:32 <Estel_> btw DocScrutinizer, what's the status of hostmode on N950?
May 19 18:35:37 <Estel_> vanilla have it?
May 19 18:35:45 <Estel_> or they screwed it again?
May 19 18:35:50 <Sicelo> as using KP for ubuntu sounds like a not so good idea .. at least for debian i had serious issues with udev .. nearly every /dev node had to be mknod'ed by hand
May 19 18:36:02 <DocScrutinizer> limbo state, as I rarely ever touch HARM
May 19 18:36:26 <Estel_> DocScrutinizer,  no surprise here
May 19 18:37:30 <DocScrutinizer> javispedro and me had a private session trying to test a "stock" openmode kernel with just some USB OTG related stuff enabled in devconfig, some weeks ago. We had to stop the session for both getting tired, and we haven't resumed yet
May 19 18:38:15 <DocScrutinizer> as for academic considerations it should work OOTB on N9
May 19 18:38:23 <DocScrutinizer> nfc about N950
May 19 18:38:41 <Estel_> DocScrutinizer,  You got N950 via last year program, or N9?
May 19 18:38:58 <DocScrutinizer> I got both
May 19 18:39:08 <Estel_> ah
May 19 18:39:22 <Sicelo> OOTB on N9? 
May 19 18:39:28 <DocScrutinizer> yup
May 19 18:39:38 <DocScrutinizer> basically
May 19 18:39:55 <Estel_> at least something to make up for lack of microSD slot, eh?
May 19 18:40:15 * Sicelo must start learning to compile kernels, and try something out for N900. 
May 19 18:40:16 <DocScrutinizer> just should need proper defconfig for kernel, as N9 by far isn't as fsckdup hw as N900 regarding USB
May 19 18:40:21 <Estel_> but I would still kill a idiot who haven't connected fmtx module to antenna
May 19 18:40:49 <Estel_> hehe, it's fsckd hw n every other aspect :P
May 19 18:40:53 <DocScrutinizer> Estel_: well, that's a kinda justified notion :-D
May 19 18:41:14 <DocScrutinizer> (about antenna that is)
May 19 18:41:32 <Estel_> I wonder if it's possible to easily hw mod it to connect
May 19 18:41:38 <Estel_> some pads or whatsnot
May 19 18:41:41 <DocScrutinizer> at very least they could've spent a testpoint for the pin
May 19 18:41:54 <Estel_> something that doesn't include rebailing or drilling multilayer pcb :P
May 19 18:41:58 <Estel_> yea
May 19 18:42:08 <Estel_> they haven't made testpoint?...
May 19 18:42:21 <DocScrutinizer> I honestly doubt there is any (though that'S mere guessing)
May 19 18:42:29 <Estel_> just nothing, bga without connection to pcb?
May 19 18:42:31 <Estel_> I see
May 19 18:42:41 * ShadowJK located the phantom resistance to a previous kludge, and determined it nonissue
May 19 18:42:46 <Estel_> well, there is a chance that I'll be able to test myself
May 19 18:42:55 <ShadowJK> got sufficiently pissed off and gave up for tonight :P
May 19 18:44:27 <DocScrutinizer> Estel_: when you're a council member you're free to send my whatever hw (N9/50) for me to disassemble and take hires scans, and analyze if I can find a testpoint
May 19 18:44:39 <DocScrutinizer> s/my/me/.
May 19 18:45:09 <Sicelo> :P
May 19 18:45:16 <DocScrutinizer> I'm reluctant to do this on N950 and actually even on N9
May 19 18:45:17 <Estel_> I got strange feeling that one day, someone, just for the sake of it, will buy one of rarely available SoC like in N900 with 1GB RAM and spent 200 dollars - or just time, if she/he own infrared soldering station - and upgrade N900's RAM
May 19 18:45:33 <DocScrutinizer> I'm reluctant to do this on N950 and actually even on N9 that I got here
May 19 18:45:35 <Estel_> DocScrutinizer,  :P
May 19 18:45:54 <DocScrutinizer> Estel_: I wouldn't need a working device for that
May 19 18:45:56 <Estel_> there are chances that I'll be able to do it myself for N950
May 19 18:46:20 <Estel_> hey, MohammadAG got broken N950, doesn't he?
May 19 18:46:23 <RST38h> MEANWHILE: Inflatable Meat Balloons Are the Perfect Decorations For Your Next BBQ
May 19 18:46:34 <RST38h> WAIT. Isn't it what humans are anyway???
May 19 18:46:47 <DocScrutinizer> OTOH IIRC I got the N9 for exactly that purpose to disassemble it...? <me starts thinking>
May 19 18:47:09 <Estel_> so, problem solved
May 19 18:47:16 <Estel_> going back to previous plan :P
May 19 18:47:28 <Estel_> You said You don't touch harmattan anyway, or rarely
May 19 18:47:36 <Estel_> risk management say risk is low ;)
May 19 18:48:04 <DocScrutinizer> indeed, and it's not like N9 wasn't replaceable
May 19 18:48:13 <Estel_> btw what ribbon in N900 You'he almost torn? this one that connect screen to motherboard?
May 19 18:48:24 <DocScrutinizer> yup
May 19 18:48:42 <Estel_> for me this thing is hard to connect again during assembly
May 19 18:48:55 <Estel_> OTOH, on one device, I've torn it to half width
May 19 18:49:03 <DocScrutinizer> the ribbon itself is sturdy, it's the connector at end which is crap
May 19 18:49:03 <Estel_> and screen still worked great
May 19 18:49:09 <Estel_> hm
May 19 18:49:17 <Estel_> I got different experiences
May 19 18:49:32 <Estel_> no problem with connector, yet damaged ribbon
May 19 18:49:44 <Estel_> yet I agree that it's sturdy
May 19 18:50:11 <DocScrutinizer> this conector actually is 3 layers: real connector, then flex cable, then yet another small PCB to create 'jumpers'
May 19 18:50:16 <DocScrutinizer> all soldered together
May 19 18:50:29 <DocScrutinizer> bend it and solder joints snap open
May 19 18:50:31 <Estel_> it's just misleading, it looks like being made from many small "wires"
May 19 18:50:34 <Estel_> yea
May 19 18:51:17 <Estel_> btw this main ribbon contain much less places of real tracks that I've thought
May 19 18:51:43 <Estel_> it looks like "tear 1mm and You're screwed"
May 19 18:52:05 <Estel_> while left half is actually just ribbon, no cooper
May 19 18:53:03 <Estel_> BTW I observed strange thing. few days ago damaged N900 with secondary board (microSD slot and camera slide sensor) was torn away
May 19 18:53:17 <Estel_> was sold for almost same price as 100% working N900
May 19 18:53:28 <DocScrutinizer> ouch
May 19 18:53:35 <Estel_> I wonder if idiots bought it, or they really think it's easily repairable?
May 19 18:53:57 <Estel_> this damn non-connectable ribbon to secondary board goes inside multilayer IIRC
May 19 18:53:57 <DocScrutinizer> idiots may think it's repairable ;-P
May 19 18:54:12 <DocScrutinizer> yep
May 19 18:54:37 <DocScrutinizer> no friggin way to fix it once it's broken
May 19 18:55:12 <Estel_> ... and if it break, it does *only* on place where it goes from PCB
May 19 18:55:22 <kerio> http://i.imgur.com/GsOUl.jpg
May 19 18:55:27 <Estel_> yet I must admit it's sturdy too
May 19 18:56:14 <Estel_> I'm now hunting for good source of cheap N900's cameras - would like to open it and remove IR filter
May 19 18:56:31 <Estel_> yet I'm afraid to try it with my module without having replacement
May 19 18:57:14 <DocScrutinizer> Estel_: there are new spare camera modules for N900
May 19 18:57:38 <Estel_> not genuine ones I guess? expensive?
May 19 18:58:22 <Estel_> oh, I almost forget to ask - N950 really have 12MP camera software'ish limited to 8MP? any way/project to remove limitation?
May 19 18:59:16 <DocScrutinizer> Estel_: http://www.saremi-mobilfunk.de/product_info.php/info/p4767_Original-Nokia-N900-CAMERA-Module-5-Megapixel.html
May 19 18:59:57 <Estel_> thanks
May 19 19:00:19 <DocScrutinizer> yw
May 19 19:00:36 <Estel_> holy shit, 43 euro
May 19 19:00:57 <DocScrutinizer> you know you'll need camera module extractor tool as well. Or some nifty DIY tooling
May 19 19:01:15 <Estel_> or gilette torn into two and screwdriver :P
May 19 19:01:26 <DocScrutinizer> yup
May 19 19:01:28 <DocScrutinizer> :-D
May 19 19:01:28 <Estel_> although I agree that extractor make it easy
May 19 19:02:04 <Estel_> BTW funny fact
May 19 19:02:15 <Estel_> I lost my stylus one day - unfortunately
May 19 19:02:27 <Estel_> and styluses for N900 in my country were priced like 7 USD
May 19 19:02:42 <Estel_> then, I've bought one from china for 1 USD with shipping
May 19 19:03:01 <Estel_> it travelled half of the world, still being 7x less expensive
May 19 19:03:17 <Estel_> + styluses offered here were also from china, surely
May 19 19:03:44 <Estel_> although I must admit that original stylus is much harder, china replacements are quite elastic
May 19 19:04:13 <Estel_> lately my son bite off stylus tip, so I tried tto buy them for 1$ again
May 19 19:04:26 <Estel_> but wefre sold only in batches of 50 min :(
May 19 19:04:51 <Estel_> someone noticed that sending 1 stylus for 1 usd half a world away isn't good idea
May 19 19:05:34 <Estel_> I needed to pretend being product manager of one famous polish gsm-selling company, and that I'm interested in buying 5k or 10k styluses...
May 19 19:05:44 <Estel_> to get 4 piece sample for 4 usd shipped
May 19 19:07:15 <Estel_> ok, I' off for now
May 19 19:07:18 <Estel_> I'm
May 19 19:07:22 <Estel_> see ya!
May 19 19:24:22 <DocScrutinizer> WTF?!
May 19 19:24:33 <kerio> lol
May 19 19:24:41 <Sicelo> yay!
May 19 19:25:27 <DocScrutinizer> :-S
May 19 19:27:02 <StyXman> damn, one of my chinese batts just fell apart
May 19 19:27:09 <DocScrutinizer> there been a netspilt, or my server had problems?
May 19 19:27:32 <StyXman> DocScrutinizer: 21:23 -!- DocScrutinizer51 [~lagrange@openmoko/engineers/joerg] has quit [Ping timeout: 260 seconds]
May 19 19:27:49 <DocScrutinizer> :nod:
May 19 19:28:03 <DocScrutinizer> sary
May 19 19:28:09 <DocScrutinizer> sCary even
May 19 19:28:29 <ShadowJK> StyXman; generic one or a brand name chinese batt?
May 19 19:37:34 <StyXman> the only thing that looks like a brand name is actually a certification icon
May 19 19:37:44 <StyXman> so I guess generic is the answer
May 19 19:37:50 <StyXman> more like «cheap and nasty»
May 19 19:39:02 <kerio> cheap and nasty is good
May 19 19:39:18 <kerio> although exploding lithium batteries have a destructive potential that you shouldn't underestimate
May 19 19:41:06 * ShadowJK is currently using a japod from dx
May 19 19:41:12 <ShadowJK> electrically it's superb
May 19 20:51:49 <Estel_> hey DocScrutinizer, of course with Your suggestions eject.sh work fine
May 19 20:53:02 <Estel_> Thanks a lot, again. I've modofied it as per Your suggestions, + added something from myself (tuned it to not fail mounting filesystem of less-than-ideal cd's and in case of failure, drop messages @ user face, instead of collapsing, etc)
May 19 20:53:23 <Estel_> of course fixed permissions and deleten unnecessary elevated ones too.
May 19 20:53:51 <Estel_> I'm going to publish it with changes - could You or anyone else knowledgeable take a look if no more code cleanup suggestions should be made?
May 19 20:53:53 <Estel_> http://sebsauvage.net/paste/?d979078432f27717#H7pJu7vAsx4J9Uc6bui/YkFFuuP10nHIxJTA953NyxA=
May 19 20:55:26 <Estel_> it even have "GUI fronentd" - I mean single tray icon as per blue_led, that now opens tray, close it (like real button) and upon closing,automatically mount cd's/DVD's filesystem (before fix, it collapsed, due to lack of waiting time between closing tray and trying to mount filesystem)
May 19 20:55:43 <Estel_> It's also priceless for ultra-light trayless optical drives
May 19 20:56:42 <Estel_> if anyone sees something that should be made more clean in code, drop me a note. If not, I'll post it soon.
May 19 20:57:34 <Sicelo> m=$(cat /proc/driver/musb_hdrc | grep Mode=) .. i do feel a  grep Mode= /proc/..... would be 'neat'
May 19 20:57:38 <Sicelo> my two cents
May 19 20:57:51 <Estel_> shit, forget that, You already said about this
May 19 20:58:13 <Estel_> will it work without elevated permissions?
May 19 20:58:33 <Estel_> cat /proc/... works, but trying more irect things on it require root
May 19 20:58:50 <Estel_> well, gotta check that :P
May 19 20:58:52 <Sicelo> that line? it's practically the same as the one you have
May 19 21:07:01 <Estel_> done:
May 19 21:07:03 <Estel_> http://sebsauvage.net/paste/?ab99bbae8cde4d97#gxsaQxrg0cUaFWVEP+v/e2p5/BVqEtupw9GRDdBFyhk=
May 19 21:07:37 <Estel_> I suppose I haven't screwed anything replacing one word in line ;)
May 19 21:08:12 <Sicelo> :P
May 19 21:09:07 <Estel_> hm
May 19 21:09:21 <Estel_> I wonder why, if it detects peripheral mode, just exits...
May 19 21:09:39 <Estel_> but if detect host mode yet no optical drive, it try to mount, actually
May 19 21:09:47 <Estel_> can't get it code-wise
May 19 21:10:39 <Sicelo> peripheral mode.. i don't think it would need to do anything. i don't know, but to me, exiting is sensible..
May 19 21:12:37 <Estel_> no, I mean something else
May 19 21:12:49 <Estel_> it should also exit when in host mode yet no optical drive
May 19 21:12:58 <Estel_> from code, it should also set e=1
May 19 21:13:08 <Estel_> resulting in different banner value, yet same exit
May 19 21:13:25 <Estel_> instead, when hostmode but no optical, "e"= remains 0
May 19 21:13:30 <MohammadAG> Estel_: No, my first N950 overheated twice so I got it replaced
May 19 21:13:40 <Estel_> of course my noobish self can't find out why
May 19 21:13:43 <MohammadAG> My second never arrived
May 19 21:13:51 <Estel_> MohammadAG,  I see
May 19 21:13:52 <MohammadAG> My third has a creaking back
May 19 21:13:59 <Estel_> wtf? they left You without replacement?
May 19 21:14:07 <Estel_> ah I see
May 19 21:14:16 <MohammadAG> ~Israel-sucks
May 19 21:14:17 <infobot> methinks israel-sucks is http://www.dhl.com/content/g0/en/express/tracking.shtml?brand=DHL&AWB=6936258346%0D%0A
May 19 21:15:02 <MohammadAG> Meh, they deleted most of the fun stuff
May 19 21:15:10 <MohammadAG> infobot: Forget Israel-sucks
May 19 21:15:11 <infobot> MohammadAG: i forgot israel-sucks
May 19 21:15:14 <Estel_> I just noticed it's empty
May 19 21:15:19 <Estel_> what was there initially?
May 19 21:15:33 <MohammadAG> Two days to reach Israel
May 19 21:15:43 <MohammadAG> Over 30 days clearance delay
May 19 21:15:47 <Estel_> ...
May 19 21:16:11 <MohammadAG> Then they asked me to make an FCC-like certificate
May 19 21:16:21 <Estel_> cute
May 19 21:16:28 <MohammadAG> So Nokia shipped me a third one, and asked for the second one back
May 19 21:16:41 <Estel_> using same shipping route?
May 19 21:16:49 <MohammadAG> Yes
May 19 21:16:52 <MohammadAG> And it worked
May 19 21:16:55 <Estel_> lovely
May 19 21:17:03 <MohammadAG> Israel's security is foolproof
May 19 21:17:14 <Estel_> and they got 2nd back, or someone from customs is enjoying N950?
May 19 21:17:19 <Estel_> yea
May 19 21:17:37 <MohammadAG> Especially since they thought my guitar amp was an explosive, and labelled it appropriately
May 19 21:17:56 <Estel_> =D not so funny in real life, I guess.
May 19 21:18:01 <MohammadAG> I didn't see Nokia sues Israel in local news so..
May 19 21:18:13 <Estel_> maybe they thought You're going to play explosive music
May 19 21:18:25 <Estel_> ;)
May 19 21:18:33 <MohammadAG> They got a robotic arm/bot to check it out
May 19 21:18:52 <Estel_> of course. Still, why explosive?
May 19 21:19:27 <Estel_> after all, it's good that they didn't concluded You're planning to use this "bomb"
May 19 21:20:14 <MohammadAG> Cause explosives look like cubes with volume, gain, delay and reverb on them
May 19 21:21:32 <Estel_> well, it's good that You've finally received Yur 3th N950...
May 19 21:21:36 <Estel_> Your*
May 19 21:23:24 <Estel_> Sicelo:  oh my, I've fixed it myself! now it exits properly on both peripheral and hostmode lacking optical drive
May 19 21:23:28 <Estel_> I feel godly ;)
May 19 21:23:33 <Sicelo> :P
May 19 21:24:12 <Sicelo> btw, i don't remember the full details, but blue_led's host-mode never worked for me..
May 19 21:24:54 <Estel_> it's probably a matter of some mistakes
May 19 21:25:11 <Estel_> hes great guy, but tends to post stuff that, apparently, he didn't test himself
May 19 21:25:20 <Estel_> containing small errors
May 19 21:25:46 <Sicelo> how do you use hostmode? blu_led's stuff, or h-e-n gui?
May 19 21:25:52 <Estel_> hopefully, final version with all bugs fixed ;) any suggestions about code cleanup etc welcome:
May 19 21:25:55 <Estel_> http://sebsauvage.net/paste/?74592b27ce099bcd#cd9vp5aOmMtxZegpCKfNNDQ31t1uOWuqkcUQuzlLkDo=
May 19 21:26:06 <Estel_> hen gui or scripts themself, it depends
May 19 21:26:19 <Estel_> scripts itself*
May 19 21:26:20 <DocScrutinizer> Estel_: (eject.sh) yw
May 19 21:27:13 <Estel_> DocScrutinizer,  thanks to "Your" -vx I was able to trace down next issue there - it's like gaining experience in good rpg ;)
May 19 21:27:14 <Sicelo> Estel_: lemme trouble you a bit.. please give me the scripts, and order of execution..
May 19 21:27:48 <Estel_> first You need to stop bme
May 19 21:27:56 <Sicelo> yeah..
May 19 21:27:57 <Estel_> damn, need to do some RL
May 19 21:28:05 <Sicelo> ow
May 19 21:28:20 <Estel_> DocScrutinizer,  could You explain him steps to use scripts without hen?
May 19 21:28:27 <Estel_> I'm must be off, sorry guys
May 19 21:28:37 <Sicelo> np. thanks :)
May 19 21:29:20 <DocScrutinizer> Sicelo: what's your problem?
May 19 21:30:03 <Sicelo> none.. but i'd love to use hostmode from terminal :-)
May 19 21:30:11 <DocScrutinizer> aah
May 19 21:30:21 <DocScrutinizer> no problem, basically
May 19 21:32:09 <DocScrutinizer> Sicelo: http://maemo.org/packages/source/view/fremantle_extras-devel_free_source/hostmode-gui/0.3/
May 19 21:32:29 <DocScrutinizer> http://repository.maemo.org/extras-devel/pool/fremantle/free/source/h/hostmode-gui/hostmode-gui_0.3.tar.gz
May 19 21:33:42 <DocScrutinizer> mainwindow.cpp
May 19 21:34:37 <DocScrutinizer> MainWindow::~MainWindow()  is about stopping hostmode:
May 19 21:34:47 <DocScrutinizer> in shell:
May 19 21:34:54 <DocScrutinizer> boostoff
May 19 21:35:03 <DocScrutinizer> echo peripheral > /sys/devices/platform/musb_hdrc/mode
May 19 21:35:12 <DocScrutinizer> ;;
May 19 21:36:18 <DocScrutinizer> MainWindow::hostmodeFS()  -> set fullspeed:
May 19 21:36:29 <DocScrutinizer> echo hostf > /sys/devices/platform/musb_hdrc/mode
May 19 21:36:37 <Sicelo> thanks. :) 
May 19 21:36:41 <Sicelo> \o/
May 19 21:36:43 <DocScrutinizer> MainWindow::hostmodeLS()
May 19 21:36:51 <DocScrutinizer> echo hostl > /sys/devices/platform/musb_hdrc/mode
May 19 21:37:01 <DocScrutinizer> MainWindow::hostmodeHS()
May 19 21:37:10 <DocScrutinizer> echo hosth > /sys/devices/platform/musb_hdrc/mode
May 19 21:37:36 <DocScrutinizer> booston for VBUS 5V
May 19 21:37:45 <DocScrutinizer> void MainWindow::enumerate()
May 19 21:37:49 <Sicelo> i had previously seen this file, but wondered what each of those commands do .. and i wasn't sure how i'd initiate enumeration -- https://garage.maemo.org/frs/download.php/8237/hostmode-working.sh
May 19 21:37:53 <DocScrutinizer> echo F > /proc/driver/musb_hdrc
May 19 21:39:51 <DocScrutinizer> so initiating a proper highspped hostmode is: #> echo hosth > /sys/devices/platform/musb_hdrc/mode; sleep 1; booston&; sleep 8; echo F > /proc/driver/musb_hdrc
May 19 21:40:15 <Sicelo> perfect 
May 19 21:40:16 <Sicelo> :)
May 19 21:40:23 <DocScrutinizer> ...and simple
May 19 21:40:32 <Sicelo> indeed ;)
May 19 21:40:46 <DocScrutinizer> the sleeps are a bit tricky
May 19 21:41:02 <DocScrutinizer> sometimes it takes even 30s after booston until ENUM will work
May 19 21:41:08 <DocScrutinizer> depends on peripheral
May 19 21:41:20 <Sicelo> i'll lengthen them .. because my device takes longer for most of the stuff i have
May 19 21:41:36 <DocScrutinizer> after ENUM succeeds it may take another 15s until mounting works
May 19 21:42:30 <DocScrutinizer> (btw same is true for any USB peripheral on my laptop as well, so nothing really h-e-n specific)
May 19 21:42:48 <Sicelo> for me it's betwen booston and enum that i have to wait longest .. sometimes i get to the point that i must restart, when suddenly everything magically comes to life :P
May 19 21:44:29 <DocScrutinizer> yeah, we operate the metorgraphics MUSB core in OMAP3430 in a special testmode that basically disables all the automatic handling of shit, so you have to walk all that speedselection and enum stuff manually, and that can sometimes be cumbersome
May 19 21:45:41 <Sicelo> well, this is the best hw/sw hack on N900, imo
May 19 21:45:43 <Sicelo> :)
May 19 21:46:26 <DocScrutinizer> on a 'normal' USB host, when things don't work, you simply unplug and replug. I have to do that every now and then on my laptop as well. Just on N900 it's a bit more than just unplugging and replugging you need to do, you basically have to redo all the manual steps
May 19 21:46:52 <DocScrutinizer> and not forget to do all the steps associated with 'unplugging'
May 19 21:46:54 <Sicelo> btw, is there a 'reset' command, for when i am done with al the hostmode things?
May 19 21:47:08 <DocScrutinizer> yep, I mentioned it first
May 19 21:47:29 <DocScrutinizer> [2012-05-19 23:34:37] <DocScrutinizer> MainWindow::~MainWindow()  is about stopping hostmode:
May 19 21:47:30 <DocScrutinizer> [2012-05-19 23:34:47] <DocScrutinizer> in shell:
May 19 21:47:32 <DocScrutinizer> [2012-05-19 23:34:54] <DocScrutinizer> boostoff
May 19 21:47:32 <Sicelo> boostoff
May 19 21:47:33 <DocScrutinizer> [2012-05-19 23:35:03] <DocScrutinizer> echo peripheral > /sys/devices/platform/musb_hdrc/mode
May 19 21:47:34 <Sicelo> :P
May 19 21:47:39 <Sicelo> ah,
May 19 21:47:46 <Sicelo> gotcha
May 19 21:48:43 <DocScrutinizer> though I always thought echo OTG > /sys/devices/platform/musb_hdrc/mode was the right thing to do, but well. Maybe this kernel needs peripheral there
May 19 21:49:29 <DocScrutinizer> and don't forget to add a sleep 30 after boostoff ;-D
May 19 21:49:58 <DocScrutinizer> possibly even a >>start bme; sleep 1; stop bme;<<
May 19 21:50:57 <DocScrutinizer> as this will most probably discharge the capacitors on VBUS line
May 19 21:52:41 <DocScrutinizer> you can't set/change hostmode speed (echo host[lfh] > /sys/devices/platform/musb_hdrc/mode) while vbus is 5V
May 19 21:53:33 <DocScrutinizer> which is the whole crux with friggin h-e-n
May 19 21:53:44 <Sicelo> ah yeah :)
May 19 21:53:57 <DocScrutinizer> as you can't detect what'S right speed for peripheral _without_ vbus 5V
May 19 21:56:30 <DocScrutinizer> another nasty detail of N900 (standard kernel) is ID pin isn't pulled up to 5V via resistor normally, so you have no chance to hw-detect plug in of a cable that grounds ID pin, to start hostmode automatically
May 19 21:57:47 <DocScrutinizer> but via shell, manually... it's straight simple as stealing a baby the lolly
May 19 21:58:45 <merlin1991> ID pin?
May 19 21:59:05 <DocScrutinizer> merlin1991: the 5th pin in mini/micro-USB
May 19 21:59:35 <DocScrutinizer> then pin that distinguishes A from B plugs for a A/B-receptacle
May 19 21:59:49 <merlin1991> okay :)
May 19 21:59:58 <DocScrutinizer> A-plugs are supposed to ground the ID
May 19 22:00:11 <merlin1991> I forgot about the extra pin on mini/micro-usb
May 19 22:00:18 <DocScrutinizer> OTG devices on seeing ID grounded are supposed to enter hostmode
May 19 22:00:19 <merlin1991> somehow my head has usb hardwired to 4 pins :D
May 19 22:02:37 <DocScrutinizer> now you may ask "why our micro-B-USB receptacle got an ID pin?". Well, simply because micro-B-*plugs* are supposed to fit into A/B-receptacles as well, and when they have 5 pins then also the micro-B receptacle has 5 pins of course
May 19 22:03:58 <DocScrutinizer> they *could* have replaced the ID pin in micro-B-receptacles by a dummy, but oh well, a dummy probably is more expensive than a true contact
May 19 22:04:53 <DocScrutinizer> OTOH meanwhile some USB-charger-specs exploit the ID pin for controlling the charger
May 19 22:05:39 <DocScrutinizer> you obviously want to do this for USB peripherals that usually have a micro-B-receptacle
May 19 22:10:54 <DocScrutinizer51> dafaq, wlan didn't notice me leaving the range of AP
May 19 22:11:16 <DocScrutinizer51> been on a AP that I left 30 min ago
May 19 22:13:06 * DocScrutinizer51 is raising big red regression-warning flag
May 19 22:13:45 <Sicelo> :O
May 19 22:14:39 <DocScrutinizer51> merlin1991: shall I open a ticket for that?
May 19 22:15:18 <merlin1991> nah that's just ususal fail ;)
May 19 22:15:26 <ShadowJK> stock firmware tries to stay on an AP if it can still hear it, regardless if it can get a signal through
May 19 22:16:23 <DocScrutinizer51> hear it over 250m and three blocks? I doubt...
May 19 22:17:04 <DocScrutinizer51> heck, it hardly keeps connection 20 m from AP
May 19 22:21:34 <DocScrutinizer> May 19 23:54:47 IroN900 kernel: [112213.386169] wl1251: ERROR Power save entry failed, giving up
May 19 22:21:36 <DocScrutinizer> May 19 23:54:48 IroN900 kernel: [112214.386047] wl1251: ERROR Power save entry failed, giving up
May 19 22:21:37 <DocScrutinizer> May 19 23:55:02 IroN900 kernel: [112227.643707] wl1251: ERROR Power save entry failed, giving up
May 19 22:21:39 <DocScrutinizer> May 19 23:55:02 IroN900 kernel: [112228.526794] wl1251: ERROR Power save entry failed, giving up
May 19 22:21:40 <DocScrutinizer> May 19 23:55:04 IroN900 kernel: [112229.675292] wl1251: ERROR Power save entry failed, giving up
May 19 22:21:49 <DocScrutinizer> zillions of such lines in syslog
May 19 22:23:43 <merlin1991> yep that's not wlancond
May 19 22:23:50 <merlin1991> that's actually the kernel driver
May 19 22:26:26 <DocScrutinizer> accurate log of doom impact: http://paste.debian.net/170198/
May 19 22:26:51 <DocScrutinizer> from then on loglines like above, 1/s avrg
May 19 22:31:31 <DocScrutinizer> so what's the fuckup going on there?
May 19 22:31:43 <DocScrutinizer> and what to kick ass to fix that shite?
May 19 22:32:14 <merlin1991> wl*** kernel module
May 19 22:32:43 <DocScrutinizer> (I'd even be willing to monitor syslog and trigger some modprobe -r&&modprobe in case too many of those lines spam in
May 19 22:32:52 <DocScrutinizer> )
May 19 22:33:18 <DocScrutinizer> wl1253.ko?
May 19 22:33:23 <DocScrutinizer> or firmware?
May 19 22:33:25 <merlin1991> probably
May 19 22:33:46 <DocScrutinizer> well, prolly firmware via wl1253.ko aiui
May 19 22:34:12 <MrPingu> I have them too ;)
May 19 22:35:28 <MrPingu> always ignored it... I also encountered that wlan-bug. I was 15 KM from home and yet i was conencted with home AP :P
May 19 22:38:33 <DocScrutinizer> merlin1991: recovery (manually selecting AP) http://paste.debian.net/170201/
May 19 22:38:55 <DocScrutinizer> MrPingu: CSSU?
May 19 22:41:07 <MrPingu> Yep testing flavour
May 19 22:41:17 <DocScrutinizer> merlin1991: ^^^
May 19 22:41:31 <merlin1991> DocScrutinizer: that part that claims the power save problems is stock maemo unless you have powerkernel
May 19 22:41:37 * DocScrutinizer waves with red regression-warning flag
May 19 22:41:38 <merlin1991> in any case it's not cssus fault
May 19 22:42:12 <MrPingu> I have those psm errors while I am on powerkernel and cssu...
May 19 22:42:18 <merlin1991> anything starting with "wlancond" might be cssus fault
May 19 22:42:31 <DocScrutinizer> merlin1991: I see that's a line starting with May 19 23:55:02 IroN900 kernel:
May 19 22:43:10 <merlin1991> so I wouldn't say regression, but valid bug
May 19 22:43:34 <DocScrutinizer> May 20 00:09:37 IroN900 wlancond[1695]: Deauthenticating
May 19 22:43:40 <MrPingu> Remember what I told about wlan yesterday? I wouldn't connect to AP until I rebooted
May 19 22:44:19 <DocScrutinizer> May 20 00:09:37 IroN900 kernel: [113103.380950] wlan0: deauthenticating by local choice (reason=3)
May 19 22:44:21 <DocScrutinizer> May 20 00:09:37 IroN900 EAP[21199]: EAP 2.0.39+0m5 quitting.
May 19 22:44:22 <DocScrutinizer> May 20 00:09:37 IroN900 wlancond[1695]: Setting BSSID 00:00:00:00:00:00
May 19 22:44:25 <MrPingu> and whole wlan was kinda weird as it didn't saw any AP at all (managed mode)
May 19 22:45:05 * DocScrutinizer waves and waves, like mad
May 19 22:46:25 <merlin1991> DocScrutinizer: and why?
May 19 22:46:43 <MrPingu> G2G, good night!
May 19 22:46:49 <merlin1991> ofc wlancond goes through the steps of deauthenticating and reauthenticating since you trigered it from userspace
May 19 22:47:02 <DocScrutinizer> wlancond is not tearing down association to BSSID when beacon lost
May 19 22:47:11 <DocScrutinizer> (see first pastebin)
May 19 22:48:10 <merlin1991> because the driver ist still trying to probe back
May 19 22:48:45 <DocScrutinizer> well, see line 21ff >>May 19 23:22:12 IroN900 wlancond[1695]: Setting SSID: PALAIS<< in http://paste.debian.net/170198/
May 19 22:50:35 <DocScrutinizer> line 27 >>May 19 23:22:13 IroN900 wlancond[1695]: SIOCGIWAP: bc:05:43:d2:4a:75<< is last sign of intelligent life of wlancond, before things go awry
May 19 22:51:25 <DocScrutinizer> then no more trace from wlancond until http://paste.debian.net/170201/ >60min later
May 19 22:52:27 <DocScrutinizer> well, more like 50min
May 19 22:54:04 <DocScrutinizer> I'd not claim it's definitely wlancond's fault, but at very least wlancond is *very* close to where things start to misbehave
May 19 22:55:48 <DocScrutinizer> I for one suspect a race between re-association/re-authentication and device leaving range of AP, so this process gets stuck halfway and thus blocks proper detection of loss of connection
May 19 22:56:43 <DocScrutinizer> now you can try figure if that's wl1251's fault or wlancond's
May 19 22:57:09 <DocScrutinizer> something doesn't behave in a way the other half can act on that behaviour accordingly
May 19 22:58:17 <DocScrutinizer> neither could I say if that's a regression of wlancond in CSSU, or it's all the same in stock
May 19 22:59:39 <DocScrutinizer> merlin1991: makes sense?
May 19 23:12:53 <DocScrutinizer> dnag, not intended to be THAT scary
May 19 23:18:10 <Sicelo> just a side note .. DocScrutinizer, the latest booston script on your server has a small typo ../sbin/stop bme 2>&1 | grep -q "bme (stop) killed" || _strart_bme_=$false
May 19 23:18:50 <DocScrutinizer> umm
May 19 23:19:09 <DocScrutinizer> mind to disclose where's the typo?
May 19 23:19:14 <DocScrutinizer> aah
May 19 23:19:25 <Sicelo> strart_bme
May 19 23:19:30 <DocScrutinizer> though I have to say this doesn't look like I ever typed it
May 19 23:20:09 <DocScrutinizer> lemme check that
May 19 23:20:10 <Sicelo> i was trying to underline .. didn't work
May 19 23:20:17 * Sicelo is half-awake now
May 19 23:20:19 <DocScrutinizer> aaah
May 19 23:20:42 <Sicelo> original line is   /sbin/stop bme 2>&1 | grep -q "bme (stop) killed" || strart_bme=$false
May 19 23:21:21 * DocScrutinizer scratches head and wonders if now only God knows
May 19 23:22:01 <Sicelo> hmm :\
May 19 23:23:16 <Sicelo> should be start_bme instead of strart_bme .. the varialbe is used elsewhere in the script to determine whether or not to restart bme
May 19 23:23:38 <merlin1991> hm scary
May 19 23:26:20 <DocScrutinizer> Sicelo: thanks, fixed
May 19 23:27:21 <DocScrutinizer> that's what you get when you don't use proper code-coverage tools for testing your crap
May 19 23:28:16 * DocScrutinizer ponders setting up some better editor support for shellscripts
May 19 23:28:31 <DocScrutinizer> like not allowing any implicitly defined envs
May 19 23:28:56 <DocScrutinizer> I mean, there's "define" for some reason
May 19 23:29:16 <DocScrutinizer> or, damn, is this a bashism again? :-o
May 19 23:30:29 <DocScrutinizer> ~ $ define x
May 19 23:30:30 <DocScrutinizer> -sh: define: not found
May 19 23:30:32 <DocScrutinizer> WAAAH
May 19 23:30:53 <DocScrutinizer> err
May 19 23:30:57 <DocScrutinizer> time for 12648430   ☕ 
May 19 23:31:58 <DocScrutinizer> ~ $ declare x
May 19 23:32:00 <DocScrutinizer> -sh: declare: not found
May 19 23:32:06 <DocScrutinizer> not any better
May 19 23:32:13 <DocScrutinizer> so still WAAAH
May 19 23:34:03 <Sicelo> there's still something else about booston.. that line, aiui, should result in bme only being started on exit IF it was explicitly stopped by booston.. this suggests to me that if bme was already stopped, booston should not start it on exit. but that's what it does :\
May 19 23:34:44 <DocScrutinizer> yes, that's the idea
May 19 23:35:19 <DocScrutinizer> I have to admit I added that late in the developing process and never really tested it
May 19 23:35:32 <Sicelo> bme is started on exit irregardless of it's previous state .. 
May 19 23:37:04 <Sicelo> maybe something in the 'if' block? or how the start_bme variable gets its value?
May 19 23:38:12 <DocScrutinizer> t900:~# /sbin/stop bme 2>&1 | grep -q "bme (stop) killed" || echo "bme not stopped"
May 19 23:38:13 <DocScrutinizer> t900:~# /sbin/stop bme 2>&1 | grep -q "bme (stop) killed" || echo "bme not stopped"
May 19 23:38:15 <DocScrutinizer> bme not stopped
May 19 23:38:16 <DocScrutinizer> t900:~#
May 19 23:38:29 <DocScrutinizer> second time it can't stop bme
May 19 23:43:06 <DocScrutinizer> Sicelo: seems all ok according to my tests. You actually used the fixed version (without the typo)?
May 19 23:45:34 <Sicelo> yeah. well, guess i must test further when i'm less sleepy
May 19 23:48:30 <Sicelo> lol, yeah.. my stupidity this time -- i'm assuming /usr/lib/hal/hald-addon-bme is bme :P
May 19 23:50:28 <DocScrutinizer> Sicelo: could you run it like >>debug=true booston<<?
May 19 23:53:45 <Sicelo> yeah, it works fine. :)
May 19 23:53:59 * DocScrutinizer assumes this inline env declaration been invented before shell
May 19 23:57:10 <Sicelo> g'nite
May 19 23:57:29 <DocScrutinizer> cya Sicelo 
May 19 23:57:45 <DocScrutinizer> thanks for spotting the bug
May 20 01:36:52 <LaoLang_cool> How to change the default alarm time in calendar?
May 20 01:37:23 <LaoLang_cool> Everytime I create an event or change the event date, the alarm will be set to 15min, it's very annoying
May 20 02:09:09 <DocScrutinizer> hm?
May 20 02:09:42 <DocScrutinizer> FSCK apt
May 20 02:10:50 <DocScrutinizer> why is there no --force option for apt?
May 20 02:11:00 <LaoLang_cool> hello
May 20 02:11:21 <LaoLang_cool> How to change the default alarm time in calendar? Everytime I create an event or change the event to another date, the alarm will be set or reset to 15min, it's very annoying
May 20 02:20:34 <DocScrutinizer> hm?
May 20 02:35:30 <DocScrutinizer> LaoLang_cool: sorry, seems there's no easy way to change that calendar behaviour. I checked binary and gconf, haven't seen any interesting value
May 20 02:41:29 <LaoLang_cool> DocScrutinizer, thank you, n900's calendar is worse in usage :)
May 20 02:42:06 <DocScrutinizer> yes, it has some strange usability issues
May 20 02:43:24 <DocScrutinizer> well, apt-get is worse
May 20 02:44:30 <DocScrutinizer> it's completely useless since >> The following packages have unmet dependencies:  hostmode-gui: Depends: kernel-feature-usbhost<<
May 20 02:45:08 <DocScrutinizer> NO way to fix this, it seems
May 20 02:46:05 <DocScrutinizer> so I wonder how to install wget, so I could download any package to install it via dpkg
May 20 02:48:05 <DocScrutinizer> let's see how annoying HAM can get
May 20 02:48:49 <ieatlint> mmm, ham
May 20 02:49:17 <DocScrutinizer> fsck HAM "no space on device"
May 20 02:49:30 <ieatlint> are you calling me fat?
May 20 02:50:38 <DocScrutinizer> no wget in HAM, of course
May 20 02:55:08 <DocScrutinizer> OMG http://thefuckingweather.com changed font to some awful serif
**** ENDING LOGGING AT Sun May 20 02:59:58 2012