**** BEGIN LOGGING AT Tue Apr 27 02:59:57 2010 Apr 27 03:00:09 to beat a dead horse further... new Date(0) = (java.lang.String) 1 Jan 1970 00:00:00 GMT Apr 27 03:00:48 we see (in this half of the world...) Dec 31st because of our neg tz offset Apr 27 03:01:04 i always thought that was odd before, glad I tried it :D Apr 27 03:01:21 Thank you Sun for the toGMTString() method :P Apr 27 03:03:02 Grabber50: I think you're right, I think I'm UTC-4 at the moment and the phone starts up thinking it's some period of time off from when it actually is until it can sync with the local carrier Apr 27 03:10:33 anyone have instruction on how to install the termainl app on the palm pre running webos 1.4 Apr 27 03:11:33 bsdbandit: install Preware, and then install Terminal from Preware Apr 27 03:11:43 rwhitby: is Terminal working again? Apr 27 03:11:44 http://install.preware.org/ Apr 27 03:11:46 bsdbandit; you need to install Preware (instructions at http://preware.org) .. then use it to install Terminal Apr 27 03:11:58 Tman: apart from the menus, yes. Apr 27 03:12:09 kk Apr 27 03:16:25 thanks guys Apr 27 03:33:59 has anyone recompiled the kernel Apr 27 03:34:05 on the palm pre Apr 27 03:35:47 yeah Apr 27 03:36:00 it's a bit painful. Apr 27 03:36:31 I wouldn't recommend trying it overclocked either Apr 27 03:36:54 best bet is to set up a cross compiler toolchain on a PC tbh Apr 27 04:00:59 sbromwich: is there any activity inthe Qt front? Apr 27 04:04:09 rehi Apr 27 04:08:35 sethron: don't know, sorry Apr 27 04:09:04 anyone know where to go to get stellarium ported? Apr 27 04:09:28 does it use SDL? Apr 27 04:09:43 dtzWill: In case you're still around, it's working! Except I'm still having the same screen update issue I was having with pure SDL, where it only updates when I touch the screen (while the splash animation is playing) Apr 27 04:13:10 Tman: oh was that the driving force behind all this? x.x Apr 27 04:13:47 Tman: okay well either use SDL_PollEvents (as part of your loop to get input, etc) or call SDL_PumpEvents in your app's main loop Apr 27 04:13:48 dtzWill: lol, well partly, but some faster rendering was something I think I was going to need Apr 27 04:13:53 ah Apr 27 04:14:03 Tman: lemme know if that doesn't fix it, it was something like that. Apr 27 04:14:25 Tman: random thing: if you put it into card mode does it update properly? :) (slightly different issue, but curious) Apr 27 04:14:34 that does it :D Apr 27 04:14:42 umm it updated when going into card mode, hold on Apr 27 04:15:06 Tman: errr if it's fixed it's fixed. card mode is a curiosity of mine, if you already fixed it no worries. Apr 27 04:15:17 no, it doesn't update in card mode, only while it's moving Apr 27 04:15:42 Tman: kk. good to know, I'm curious how these things manifest themselves. but pumpevents fixed it you say? Apr 27 04:15:49 yes it did Apr 27 04:15:54 \o/ great Apr 27 04:15:56 ok there is a random line going from the center of the screen to the first point of the first line (I presume) Apr 27 04:16:25 Not sure what exactly could be causing it... Apr 27 04:16:36 well center is 0,0, might wanna play with your vertices and indices your passing, make sure you're doing that right? Apr 27 04:16:57 I don't remember if GL_LINES draws consecutive lines or disjoint pairs, I'm leaning towards disjoint but idk Apr 27 04:17:16 I think disjoint too Apr 27 04:17:28 vertices/indices seem to be right Apr 27 04:23:16 ah well, I've made good progress today Apr 27 04:23:44 I'm gonna have to shelf this project for awhile in favor of a couple others.. will definitely come back to it though Apr 27 04:42:38 new kernel for anyone who wants to play: www.fop.ns.ca/pre/uImage.141-10.1_domperignon3 Apr 27 04:42:47 this may have new and amusing power management side effects Apr 27 04:43:44 how do i do a git fetch to just a subdirectory Apr 27 05:40:09 running palm-emulator is just supposed to "work" Apr 27 05:40:31 lies Apr 27 05:40:53 enlighten me oh great halfhalo Apr 27 05:41:06 whats not working or what is Apr 27 05:41:09 i can fix both Apr 27 05:41:12 anyone know how to upload/dl a video on the emulator (for testing) Apr 27 05:41:24 i ran palm-emulator and nada... Apr 27 05:41:30 must be my virtualbox install... Apr 27 05:41:53 do palm-emulator from command line and see what it says Apr 27 05:42:22 finally a message - virtualbox not installed... Apr 27 05:42:28 great,, --thanks. Apr 27 06:16:02 g'night Apr 27 06:31:12 need help....none of my installed patches seem available since my phone updated. what is the problem? Apr 27 08:12:31 hape: I had to change the cross-compile pdl Makefile, cause it was putting the PalmPDK in the downloads directory, and the autobuilder would automatically redistribute that - I moved it to the doctors directory (which is where all the stuff we download but do not redistribute goes) Apr 27 08:14:29 rwhitby: by default, does the meta-doctor bypass activation? Apr 27 08:15:28 jhojho: by default, unless you uncomment any lines, the meta-doctor does nothing Apr 27 08:17:37 so at a minimum, I need BYPASS_FIRST_USE_APP? Apr 27 08:18:02 if I want to have a palm profile? Apr 27 08:23:17 yep Apr 27 08:23:30 jhojho: if you can activate that is (e.g. GSM) Apr 27 08:23:59 jhojho: for CDMA, you will need to bypass activation to be able to run the first use app to log into a profile without having a working cdma radio Apr 27 08:24:22 hape: I have also update navit/Makefile to not require /media/cryptofs/apps on the build host Apr 27 08:25:34 so for activating my new pre plus. BYPASS_ACTIVATION = 1 and no BYPASS_FIRST_USE_APP then? Apr 27 08:31:21 hi all guys i have a problem on my palm pre when i install from preware the patch "advanced phone menu" my phone menu (restarting luna) remains the same. Apr 27 08:31:34 could anyone help me? Apr 27 08:46:34 rwhitby: ok, did I get it right that the org.webosinternals.navit folder should now be created in the build dir? Apr 27 08:58:17 rwhitby: Could you please du the same for the espeak packgae Apr 27 09:39:46 hi all guys i have a problem on my palm pre when i install from preware the patch "advanced phone menu" my phone menu (restarting luna) remains the same. Apr 27 09:39:58 could anyone help me? Apr 27 10:39:16 ? Apr 27 10:40:33 i having an issue im trying to change the bootlogo on webos 1.4 when i copy my pic over the bootlogo the changes never take effect Apr 27 10:40:37 what am i missing Apr 27 11:36:59 hape: will do Apr 27 11:38:30 hape: cannot download http://espeak.sourceforge.net/test/espeak-1.43.20.zip Apr 27 11:38:37 hape: is there a mirror for that file? Apr 27 11:40:29 hape: changing to 1.43.21 Apr 27 11:41:22 hape: espeak updated Apr 27 12:21:07 rwhitby: tey do this updates verry often. Apr 27 12:22:03 rwhitby: As you see we tried to switch to a cvs checkout (as we do it for navit) but faild. We are not sure why. Apr 27 12:46:48 hallo Apr 27 12:47:00 jemand da Apr 27 12:51:31 so halbwegs Apr 27 12:56:53 does preware alpha need the package manager service? Apr 27 12:57:16 or is it internal to the program? Apr 27 12:57:35 it doesn't need the service Apr 27 12:57:50 okay, thought so, just wanted to make sure Apr 27 12:57:52 thanks Apr 27 13:33:42 hey does anyone know where the file for the little tap effect is? Apr 27 13:49:37 (the ripple, that is) Apr 27 13:50:58 nevermind, found it :) Apr 27 13:55:18 morning Apr 27 13:55:53 morning...i think Apr 27 13:56:05 * uNiXpSyChO needs more coffee Apr 27 13:56:05 yeah it's still a bit early to tell Apr 27 13:57:30 and... I know I should be but I'm slightly boggled by some of the PC posts; "if I install the compcache kernel does that mean I'm not running SuperKernel"... ummmmmmmmmm... Apr 27 13:57:40 s/should/shouldn't/ Apr 27 13:57:40 sbromwich meant: and... I know I shouldn't be but I'm slightly boggled by some of the PC posts; "if I install the compcache kernel does that mean I'm not running SuperKernel"... ummmmmmmmmm... Apr 27 13:58:15 i saw that. lusers Apr 27 13:58:26 *nod* Apr 27 13:58:58 but at least now they can't complain someone hasn't done the work Apr 27 13:59:18 now BFS will kick my ass Apr 27 13:59:27 not sure i want to take on that quagmire Apr 27 13:59:30 I'm not convinced myself... it's like, what was it, stacker? for disks back in the drdos days, nice idea but... Apr 27 13:59:49 I don't think BFS will go anywhere on such an old kernel Apr 27 14:00:17 maybe -rt would be better? ;-) *nudge* Apr 27 14:00:18 I'm still fiddling with the power stuff trying to ameliorate the constant hammering from powerd Apr 27 14:00:33 it will, yeah, but I'm distracted by the prcm stuff Apr 27 14:00:59 TI have published a known errata with... CLK1? where if you try ti put it to sleep it hangs the device coming out of wakeup Apr 27 14:01:03 sounds familiar? Apr 27 14:01:14 oh? Apr 27 14:01:43 so now I'm trying to map that back to how palm named the clocked and I *think* it's DOM_PER Apr 27 14:02:12 so I went through last night after I chucked that code over to you and for a laugh disabled all the DOM_PER stuff I could find Apr 27 14:02:39 tres, tres smooth but the battery life dropped like a stone... drained down to ~25% in 9 hours of doing nada Apr 27 14:02:50 should i apply that patch to the compucache sources i have? Apr 27 14:03:02 ouch Apr 27 14:03:09 I'm still of the opinion that powerd is causing too many wakesups per second Apr 27 14:03:17 if you're feeling particularly nasty and vicious... sure ;-) Apr 27 14:03:29 i'm also thinking powerd isnt letting the phone fully sleep Apr 27 14:04:03 looking at the code I'm not too sure which files I acutally changed any more, but I'm happy to throw the lot at you if you want Apr 27 14:04:05 ditto Apr 27 14:05:14 I think if the phone's gone into suspend after 15 minutes with no activity (voice call in progress, etc) it can be taken as read that the user's done with the device and *everything* can be put to sleep bar the DSP listening for phone calls Apr 27 14:06:00 and conversely, if the device is plugged into power, powerd shouldn't even bother trying to turn anything off Apr 27 14:08:09 have you tried your patch 1st? ;-) Apr 27 14:08:30 for prcm? Apr 27 14:08:48 ya Apr 27 14:09:02 I left it running overnight off charge, normally I'd expect battery to be around 85%, I was down around 25% Apr 27 14:09:42 www.fop.ns.ca/pre/prcmdrain.tar.gz if you want to play Apr 27 14:10:16 ah. i thought I was your 1st hamster :) Apr 27 14:10:23 very smooth in the UI though Apr 27 14:10:36 no, I try everything on my own device first Apr 27 14:10:46 is this diff from the kernel you had me load last night? Apr 27 14:10:51 compiled test kernel is... jeez I shouldn't try coding so late... sec Apr 27 14:10:52 yeah Apr 27 14:11:25 domperignon3, I think Apr 27 14:12:09 uname -a timestamps 20100427043353 Apr 27 14:12:30 i still had that one running when i left this morning. i'll patch this one in to see later tonight. Apr 27 14:13:04 be fun to see how fast the battery drains with compcache as well Apr 27 14:13:10 all I had running was battery monitor Apr 27 14:13:17 haven't even installed cron on it yet Apr 27 14:13:42 compucache seems to be a battery drain already Apr 27 14:14:23 *nod* Apr 27 14:14:42 spods on the forum don't seem to understand the concept of TANSTAAFL Apr 27 14:15:56 flame them Apr 27 14:16:11 ignoring them is better for my blood pressure :-) Apr 27 14:16:29 correcting one muppet on the interweebs just means there's a squillion more to go. Apr 27 14:16:42 what are you running on compache? Apr 27 14:16:54 webOS? Apr 27 14:16:57 i mean.. Apr 27 14:17:00 how are you using it Apr 27 14:17:01 I just put straight facts out there and let the numpties do their own googling Apr 27 14:17:46 i compiled it in the enabled the swap Apr 27 14:18:26 does it give any perceptible performance improvement that you've noticed? Apr 27 14:18:31 uNiXpSyChO: sorry i am mixing up 2 similar tricks in my head Apr 27 14:18:59 PuffTheMagic: which other trick? Apr 27 14:20:00 sbromwich: hard to tell. i cant run more apps, but the lots of apps i got running seem to not jerk and delay reloading to foreground. Apr 27 14:20:59 uNiXpSyChO: using a fs supporint compression in a loopback file on tmpfs Apr 27 14:21:01 uNiXpSyChO: with compcache do you still get the scrolling jerkiness in preware alpha? Apr 27 14:22:04 sbromwich: thats one thing i didnt try. i should have a better list of tests :). i just test for basic things like can i make a call, camera, SMS. Apr 27 14:22:21 yeah, my test cases aren't as extensive as they should be. Apr 27 14:22:44 maybe if we had someone looking for a coding project who could code up some sort of benchmarking app in mojo? Apr 27 14:22:57 gosh, who could do such a thing? Apr 27 14:23:11 help us, obi oil kenobi, you're one of our only hopes! Apr 27 14:23:16 sbromwich: actually Apr 27 14:23:21 me and oil could do that :D Apr 27 14:23:34 well volunteereded :-) Apr 27 14:23:37 sbromwich: mayby port unixbench or some Apr 27 14:23:45 is there any sdl based benchmark tools? Apr 27 14:24:03 PuffTheMagic: the method i put in was a compressed swap. not sure i've seent eh other trink Apr 27 14:24:28 uNiXpSyChO: this was something we did before compache cameout Apr 27 14:24:32 gentoo ricing Apr 27 14:25:14 ah Apr 27 14:27:01 prcm changes reduced wakeups per second by around 20 though Apr 27 14:27:05 according to powertop Apr 27 14:27:20 but it kills the battery quicker? hmmmmm ;) Apr 27 14:28:32 well, it's not putting stuff to sleep, it's just idling the hardware Apr 27 14:29:47 does that IVA2 in suspend work? i tried it and it caused an Oops for me. Apr 27 14:30:06 yeah, it's tied in with the PREEMPT stuff Apr 27 14:30:30 it's a matter of fiddling with the timing for bitbanging the hardware Apr 27 14:31:02 some of the less finicky hardware works well with doing a cond_resched instead of udelay, too Apr 27 14:31:45 i have heard of some finicky hardware out there that cant even do 250MHz cleanly. but mine runs good. Apr 27 14:34:04 the issue seems to be changing the clock while messing around with suspending/waking hardware... because it's doing udelay it's doing timing loops based on cpu speed, my guess is the busywait is making ondemand up the clock frequency so the 100ms wait becomes (say) 20ms and the hardware goes wonky Apr 27 14:36:40 like i always have said "ondemand" is bad. but users love it. Apr 27 14:37:03 *nod* Apr 27 14:37:17 I just locked it to performance with my naive governor and it works great Apr 27 14:38:28 I suspect playing with min/max frequency in the prcm stuff would also fix the hang from suspend, but that would basically just be bandaiding the problem Apr 27 14:39:02 uNiXpSyChO: on demand isnt bad, its just that the pre doesnt really have a usable low end Apr 27 14:39:11 so the useful range is sorta small to switch between Apr 27 14:40:21 did I talk about the harmonics ondemand causes? Apr 27 14:40:25 on a PC it isnt bad, but on an embedded device where ticks of the CPU clock cause battery is bad. the ondemand code is bloat-torious. Apr 27 14:41:05 I thought I mentioned it at one point... it's bad enough that it's audible if you plug in external speakers and up the speaker volume, it's possible to hear the cpu changing frequencies Apr 27 14:41:09 something like "performance" or userspace is better as the code is simple... "performance" is the simplest and smallest codebase. Apr 27 14:41:16 *nod* Apr 27 14:42:11 no shit u are going to hear the cpu change freqs, the lower freqs are too slow Apr 27 14:42:39 yeah, the problem is that can also spill over to the mic and screw up voice conversations Apr 27 14:44:32 what freqs do you run ondemand at for min/max, PuffTheMagic? Apr 27 14:45:10 sbromwich: i dont Apr 27 14:45:24 well Apr 27 14:45:30 you don't run ondemand? Apr 27 14:45:48 nope Apr 27 14:45:56 run stock 500 Apr 27 14:46:28 do you have it doing any underclocking at all? Apr 27 14:46:44 i tried ondemand a few times Apr 27 14:46:49 i didnt tune it Apr 27 14:46:59 i know it got a little laggy Apr 27 14:47:11 but that goes down to what 150 something 250 somewthing right? Apr 27 14:47:28 up_threshold defaults to 95, which is why it's laggy on an embedded device Apr 27 14:47:44 yeah i i know about that tunable Apr 27 14:47:48 it will clock down to whatever scaling_min_frequency is set to Apr 27 14:47:52 like i said i didnt really play with it much Apr 27 14:48:16 but im sure there is a nice little range in there Apr 27 14:48:26 that lets it idle a little slower Apr 27 14:48:33 125, 250, 500, 550, 600, 720 and 800MHz Apr 27 14:48:37 :/ Apr 27 14:48:49 nothing between 250 and 500? Apr 27 14:48:50 idling is not the issue at point though Apr 27 14:48:51 that sucks Apr 27 14:48:55 sbromwich: I'll run your kernel if you let me run my Pre at 4.77Mhz Apr 27 14:49:03 sbromwich: people have different prioities Apr 27 14:49:11 nt4cats: you'll have to talk to TI for that one ;-) Apr 27 14:49:24 nt4cats: reminds me of a 6502 CPU Apr 27 14:49:57 sbromwich: can a freq between 250 and 500 be added like u added 720 and 800? Apr 27 14:49:59 PuffTheMagic: we are discussing ondemand wakeups per second and side-effects from it, not necessarily the effectiveness of it as a scheduler per se Apr 27 14:50:18 PuffTheMagic: only what's available on TI dies Apr 27 14:50:41 i thought u were playing with the voltages Apr 27 14:50:43 uNiXpSyChO would know for sure but I think those are the ranges, above that I think there's 1GHz and 1.2GHz Apr 27 14:50:49 sorta like how the undervolting patch works Apr 27 14:51:03 no, I'm playing with current draw reduction not voltage reduction Apr 27 14:52:03 as a side-effect that may allow smartreflex to start working, but it's not something I'm explicitly working towards Apr 27 14:52:27 did smartreflex ever work in previous kernels without lockups, anyway? Apr 27 14:53:16 some users swore by it. i doubt it ever worked correctly since the default stock kernel was configured for somethnig else. Apr 27 14:53:20 if palm is not going to work on the pre kernel any more Apr 27 14:53:30 maybe we should try to use the upstream omap3 code Apr 27 14:53:44 now that we have a group for kernel hackers Apr 27 14:54:08 sbromwich: you're making the kernel patches ? are these public ? Apr 27 14:54:16 the reason smartreflex failed was because (if I read the TI docco aright) CLK1 doesn't like being made to sleep and strops out when smartreflex comes into play Apr 27 14:54:47 xaiki: they're test at the moment, but I've got a tarball up somewhere based off the webos git tree... sec Apr 27 14:54:48 isnt that what Palm puts to sleep manually? Apr 27 14:55:09 xaiki: www.fop.ns.ca/pre/prcmdrain.tar.gz Apr 27 14:55:54 uNiXpSyChO: yeah, and I think it's also the cause of the "hang from wakeup from suspend", but tracking it down to be sure is a bit of a bear Apr 27 14:56:45 xaiki: I also have a test kernel if you know how to make novacom boot mem and want to give it a whirl? Apr 27 14:57:54 sbromwich: I didn't follow all of the conversation, I'm mostly interested in seeing what's beeing patched =) Apr 27 14:58:38 xaiki: see also http://git.webos-internals.org/?p=kernels/patches.git;a=summary Apr 27 14:59:15 sbromwich: ok, *that*'s what I was looking for =) Apr 27 14:59:29 xaiki: I'm seeing issues with wake from suspend and my rather caveman approach is just to keep beating code and removing bits to see how far I can go before it breaks Apr 27 15:00:07 xaiki: most of my changes get dumped before they get there, only stuff I know is safe gets committed Apr 27 15:02:48 hi all could anyone help me installing an app to my pal pre? Apr 27 15:03:34 i've installed the patch advanced phone menu on preware but it just does anything Apr 27 15:04:34 /proc/power/clock_bail_stat has some interesting stats... Apr 27 15:05:26 8 + printk(KERN_INFO "Help me Obi Pre Kenobi... Timeout in check_device_status(), deviceid %d, " Apr 27 15:05:30 ok =) Apr 27 15:06:34 xaiki: some of the errors are named identically (or almost so) so I put in some markers to make it easier to see what was going on Apr 27 15:07:44 yeah, not criticizing, just made me smile, I'm not understanding what these patch do at all. Apr 27 15:07:55 and the overclocking one is even more obscure, I should look more deeply Apr 27 15:08:42 xaiki: what it's doing is catching some of the udelay stuff and instead (or as well) doing a cond_resched so other threads get to run while waiting on the hardware Apr 27 15:09:31 sbromwich: where does that patch come from: http://git.webos-internals.org/?p=kernels/patches.git;a=blob;f=overclocking/opp6-720-opp7-800-min-opp5.patch;h=79f454bbb781df07c10dd63633f254f512463b83;hb=628585e9fe73b60912ee4e9f9d93ecf9db83820c ? Apr 27 15:10:03 you ? where did you get the values for core_l3_vdd2_volts ? Apr 27 15:10:47 xaiki: I think that's uNiXpSyChO's patch Apr 27 15:11:13 I think the settings were backported from a later processor but he can say for sure Apr 27 15:12:09 uNiXpSyChO: can you remember if it was an immediate lockup with smartreflex, if it was random, or if there was something specific that made it lock up? Apr 27 15:14:01 ahhh... random lockups Apr 27 15:21:11 it was a random lockup Apr 27 15:21:25 * uNiXpSyChO needs to stop doing work while at work Apr 27 15:27:28 heh Apr 27 15:53:41 uNiXpSyChO: http://linux.derkeiler.com/Mailing-Lists/Kernel/2010-04/msg04972.html Apr 27 15:53:50 I think that might explain the whole DOM_PER / wakeup thing Apr 27 15:55:09 ohhhhhh Apr 27 15:55:24 gonna apply that soon as i can get home to play Apr 27 15:56:51 well that's for omap2... I'm trying to figure out the equivalent for omap3 Apr 27 15:59:34 prcm_set_wkup_dependency seems to be the equivalent Apr 27 15:59:36 alot of the code we use is in omap2 files tho. i dont remember and i can be completely all wet. Apr 27 16:00:34 this bit is different Apr 27 16:00:59 sbromwich: power draw eh? Apr 27 16:01:25 I've found where it sets a dependency on PRCM_WKDEP_EN_MPU so now I need to figure out what that is ;-) Apr 27 16:01:36 oilsworkn: well if you *will* let any idiot have edit access to the wiki... ;-) Apr 27 16:02:14 * oilsworkn fixes that Apr 27 16:03:40 heh Apr 27 16:03:57 * uNiXpSyChO sneaks out for lunch Apr 27 16:04:14 but really, if someone add a method to the service to tell the app what the power draw is, ill add it Apr 27 16:04:52 oilsworkn: I believe there's a way to do it by querying lunasysmgr Apr 27 16:05:07 the battery monitor app does it Apr 27 16:05:16 i thought it just did % Apr 27 16:05:25 * oilsworkn has never used the battery monitor app Apr 27 16:05:37 and I'm sure I've seen a log where sawmill was muttering about coulombs Apr 27 16:05:50 just a quick question about govnah: if I change a setting in it, is it only applied while the app is running, or does it "stick"? and does it reset on reboot? Apr 27 16:06:32 VincentLaw: the app applies it Apr 27 16:06:38 and you dont need to keep it open once doing so Apr 27 16:06:41 but on reboot, itl be back to stock Apr 27 16:06:49 ah ok Apr 27 16:07:27 am I right in thinking in TI world the MPU is the central processor? Apr 27 16:09:06 ooh. Apr 27 16:09:08 http://www.mjmwired.net/kernel/Documentation/arm/OMAP/omap_pm Apr 27 16:12:54 http://fxr.watson.org/fxr/source/arch/arm/plat-omap/include/mach/omap-pm.h?v=linux-2.6 Apr 27 16:15:15 233 * Report the current OPP for the DSP. Since on OMAP3, the DSP and Apr 27 16:15:17 234 * MPU share a single voltage domain, the OPP ID returned back may Apr 27 16:15:19 235 * represent a higher DSP speed than the OPP requested via Apr 27 16:15:21 236 * omap_pm_dsp_set_min_opp(). Apr 27 16:15:24 ah-hah. Apr 27 16:18:53 * halfhalo_T400 is liking the new amd 6 core cpu's Apr 27 16:19:08 halfhalo_T400: how much cache? Apr 27 16:19:51 same as the phenom II quads Apr 27 16:20:03 but the most expensive one is 285 Apr 27 16:20:05 is there one with the 3 meg of L3? Apr 27 16:20:16 the... black edition or whatever? Apr 27 16:20:22 I think there may be one Apr 27 16:20:42 * sbromwich is waiting on his quad core with 3 meg of L3 Apr 27 16:20:48 the highest end one runs at 3.2GHz and is a black edition Apr 27 16:21:50 sbromwich: the 12 cores have 12MB of L3 Apr 27 16:22:11 Hell, considering I have been using a laptop which has been locked at a x3 multi and 800MHz and not noticing it a 6 core may be a little excessive Apr 27 16:22:19 ok, yeah, both models that have been reviewed have 6MB of L3 Apr 27 16:23:50 plus to build a new system around it would be cheaper than a new intel system for me Apr 27 16:24:05 bougyman: way to make me start rethinking my purchase ;-) Apr 27 16:24:22 Heh Apr 27 16:24:27 i'm a sucker for the underdog. Apr 27 16:24:33 the 6 cores start at 185 Apr 27 16:24:35 i always get one or two amd's when the new procs come out. Apr 27 16:24:52 being rich must be awesome :) Apr 27 16:24:57 I have used amd but switched to intel for the q6600 Apr 27 16:25:07 which just doesn't cut it anymore Apr 27 16:25:20 oilsworkn: all you have to do is log retarded amounts of overtime and you, too can piss away your money on expensive toys you never get to use Apr 27 16:25:30 yeah Apr 27 16:25:39 i dont even come close to overtime Apr 27 16:25:53 i don't even have a job yet Apr 27 16:25:55 hi, btw Apr 27 16:25:59 * halfhalo_T400 doesn't get overtime Apr 27 16:26:08 * halfhalo_T400 doesn't even do 20 hours a week Apr 27 16:26:18 what's a job? Apr 27 16:26:20 oilsworkn: I seem to be one of the few people capable of RTFM so I get called in aaaaaall the time to fix SEPs Apr 27 17:07:37 this channel is moving entirely too slow for me to be able to ignore work Apr 27 17:07:51 hah Apr 27 17:08:00 * oilsworkn goes to vending machine Apr 27 17:19:21 oilsworkn: You could hang out in the other one. Nothing really interesting going on there, just trying to get my login page working. Apr 27 17:19:43 the other one? Apr 27 17:22:57 #webos Apr 27 17:23:26 oh Apr 27 17:24:40 oilsworkn: I'm surprise you don't chill in #webos, lotsa mojo talk Apr 27 17:26:32 lies Apr 27 17:26:55 * zinge lies when he's sleeping Apr 27 17:27:15 I lies when I lies Apr 27 17:35:15 why would i want to do lotsa mojo talk? Apr 27 17:35:19 * oilsworkn already knows what hes doing Apr 27 17:35:21 :) Apr 27 17:35:56 and wth, the people in there are the same as in here? Apr 27 17:36:11 that accomplished a lot Apr 27 17:36:12 haha Apr 27 17:38:52 nt4cats: can you do another precorder test if you get a sec Apr 27 17:43:14 zsoc_wirc: sure Apr 27 17:43:34 zsoc_wirc: please give me very specific instructions about the sound(s) you want me to record, too. Apr 27 17:43:48 nt4cats: media captures this time Apr 27 17:44:07 should I grab a new build, I assume? Apr 27 17:44:16 no Apr 27 17:44:30 okay, firing up last night's build. readyu Apr 27 17:44:33 err ready Apr 27 17:44:46 nt4cats: I would like you to verify the following behavior: start media capture with nothing playing. then load anything that plays music and start it Apr 27 17:45:02 verify that the numbers start moving (once they catch up) Apr 27 17:45:25 then stop precorder WHILE the other app is still playing. Apr 27 17:45:32 media cap, record, starting pandora ... Apr 27 17:45:38 verify that precorder stops almost immediately Apr 27 17:46:04 pandora is playing, numbers aren't moving Apr 27 17:46:18 nt4cats: focus precorder, give it 5 seconds Apr 27 17:46:26 it has been 20 seconds Apr 27 17:46:29 no numbers Apr 27 17:46:38 nt4cats: ok you can kil it Apr 27 17:46:39 wow, now it is going Apr 27 17:46:44 oh lol Apr 27 17:46:44 numbers going fat Apr 27 17:46:47 err fast Apr 27 17:46:54 yeah it should catch up Apr 27 17:47:01 wtf Apr 27 17:47:08 phone lost irc connection Apr 27 17:47:08 position queries are /very low/ priority Apr 27 17:47:28 nt4cats: can you stop while pandora is still playin? Apr 27 17:47:39 checking ... Apr 27 17:47:59 I hit stop, time sped up fast for about 10 second and then it stopped Apr 27 17:48:07 nt4cats: please play that file and make sure there is no deadspace at the beginning (ie, it didn't start till the song started) Apr 27 17:48:25 (I mean subjective time in the universe around me, not on precorder) Apr 27 17:48:50 nt4cats: also, even tho the numbers were funky, it should not effect the actual recording Apr 27 17:49:15 I ruined it Apr 27 17:49:25 wat? Apr 27 17:49:58 there was no deadspace ... but after my phone started blaring Abba (Pandora's choice, not mine) to the amusement of my co-workers ... I hit the volume down button until the music was quiet. Apr 27 17:50:26 nt4cats: ah, yes, it records actual output volume :) Apr 27 17:50:29 the recording started but seemed to fade to zero-volume at about the time I turned down the volume on the speaker Apr 27 17:50:31 nt4cats: that is ok. Apr 27 17:50:57 I could still hear the song (quietly) while the recording was going on ... now I don't hear anything. Apr 27 17:51:06 (during playback after the point at which I dialed down the volume) Apr 27 17:51:32 nt4cats: ok. well that is a different issue lol. 1 last test Apr 27 17:51:53 nt4cats: same process, only this time, close pandora BEFORE you attempt to stop precorder Apr 27 17:52:08 nt4cats: I'm trying to verify that will cause it to hang for 1 minute Apr 27 17:52:22 zsoc_wirc: actually this might make sense Apr 27 17:53:05 nt4cats: yep. but I need to verify this behavior before I attack fixing it ;) Apr 27 17:53:27 I started with a volume of (made up number) 10, and dialed it down to 5. You recorded starting at 10, and lowered it to 5. At playback the output volume was still at 5, so the initial sound I heard was as loud as pandora ... b Apr 27 17:53:28 and then I need to figure out how to send an eos to a pipeline with no buffers Apr 27 17:54:08 okay, I'm going to start by closing everything, starting precorder, media cap, record, start pandora ... Apr 27 17:55:16 nt4cats: oh, yeah, volume will be relative by percentage. you would have to turn playback up past 50% of the way to hear somethine recordered past 50% Apr 27 17:56:20 it took about 50 seconds for the timer to start incrementing Apr 27 17:56:32 I'm waiting for it to catch up and then I'll close pandora Apr 27 17:56:40 nt4cats: you are not on 800mhz lol Apr 27 17:57:05 zsoc_wirc: nope, I'm on a stock kernel/speed here Apr 27 17:57:33 okay, closing pandora now ... Apr 27 17:57:33 nt4cats: yeah I'll fix up the details by beta for stock Apr 27 17:58:55 okay pandora closed, it the clock sped up and caught up ... and is paused/stuck at 'recording ...' Apr 27 17:58:59 (as you predicted) Apr 27 17:59:41 still seems stuck ... Apr 27 18:00:12 ... still stuck ... Apr 27 18:00:52