| 00:05.54 | *** join/#edev ico2 (n=ico2@modem-3973.chimpanzee.dialup.pol.co.uk) |
| 00:08.28 | *** join/#edev redblue (i=star@ppp158.108-253-207.mtl.mt.videotron.ca) |
| 00:24.43 | MonMotha | the issue is getting the part placed close enough and then hitting the reflow profile well enough to actually reflow everything without also baking the part |
| 00:24.54 | MonMotha | you can xray it all you want, but that just shows whether or not it worked |
| 00:24.57 | MonMotha | doesn't help you actually do it |
| 00:39.23 | ds2 | It seem that chip makers would benefit if they sample premounted parts on a breakout board of sorts |
| 00:46.53 | *** join/#edev chouimat (n=dieu@kde/developer/chouinard) |
| 00:47.03 | *** join/#edev catastrop (n=catastr0@d14-69-178-65.try.wideopenwest.com) |
| 01:17.13 | *** join/#edev linac (n=lin@25.127.61.59.broad.fz.fj.dynamic.163data.com.cn) |
| 01:25.03 | *** join/#edev joe_bleau (n=joe_blea@adsl-76-213-137-175.dsl.fyvlar.sbcglobal.net) |
| 01:25.53 | *** join/#edev chouimat|tv (n=dieu@r2351064.cidc.net) |
| 01:30.43 | MonMotha | ds2: the issue is how do you break out some of these high density BGAs? |
| 01:31.03 | MonMotha | 324 pads is a lot |
| 01:31.08 | MonMotha | 560 is even more |
| 01:31.22 | MonMotha | the chip would get huge |
| 01:33.31 | *** join/#edev silverpower (n=shrike@atheme/bofh/silverpower) |
| 02:00.45 | *** join/#edev chouimat (n=dieu@kde/developer/chouinard) |
| 02:13.30 | jbevren_ | MonMotha: I have -no- idea how they break out the 939 pin amd's |
| 02:13.33 | jbevren_ | it boggles me |
| 02:16.40 | MonMotha | jbevren: those are PGAs to begin with, though |
| 02:16.48 | MonMotha | so you don't end up with the offset via grid thing |
| 02:16.51 | jbevren_ | shrug |
| 02:17.00 | jbevren_ | you have a point :) |
| 02:17.04 | MonMotha | I'm not sure it would make routing that much easier, but it makes soldering easier if you have a through-hole socket |
| 02:17.06 | jbevren_ | the vias are already there |
| 02:17.15 | MonMotha | that said, I'm not sure the sockets are actually through-hole |
| 02:17.25 | jbevren_ | dI'm trying to remember |
| 02:17.29 | jbevren_ | but I"m pretty sure mine is |
| 02:17.43 | jbevren_ | and I cant be bothered to disassemble my PC to find otu |
| 02:17.44 | jbevren_ | lol |
| 02:25.36 | MonMotha | my PC is too old for that socket |
| 02:27.31 | *** join/#edev prpplague (n=dave@ppp-70-249-148-31.dsl.rcsntx.swbell.net) |
| 02:27.39 | prpplague | ho ho ho |
| 02:27.50 | prpplague | jbevren_ whats cookin? |
| 02:29.39 | MonMotha | you aren't normally on this late |
| 02:30.58 | prpplague | yea just been busy alot in the last few months |
| 02:31.39 | prpplague | finally got around to wiring up ethernet to my desktop area |
| 02:33.25 | prpplague | MonMotha: been real busy getting feedback on the hammerboard |
| 03:12.41 | sunrunner20 | sockest |
| 03:12.53 | sunrunner20 | are soldered to the surface of the pcb |
| 03:13.13 | prpplague | sunrunner20: ??? |
| 03:13.36 | sunrunner20 | [21:18:54] <MonMotha> that said, I'm not sure the sockets are actually through-hole |
| 03:14.15 | sunrunner20 | there simply would not be enough structurel integrity to put any measure of force on the chip if it went through the pcb |
| 03:14.31 | prpplague | what item was he refering to? |
| 03:15.12 | sunrunner20 | pga 939 |
| 03:25.31 | MonMotha | yeah, I was under the impression that the sockets were not actually through-hole |
| 03:25.47 | MonMotha | they take leaded PGA type parts, but the socket is more of a BGA type thing |
| 03:25.55 | MonMotha | as far as how the socket is mounted to the PCB, that is |
| 03:26.12 | MonMotha | prpplague: what was the place you had do your boards for the 0.8mm BGA? |
| 03:26.35 | MonMotha | Advanced Circuits can't do it, and my production alternative (Calumet) may or may not (and they don't specialize in quick-turn anyway) |
| 03:27.50 | prpplague | MonMotha: a place called W-Tech in the dfw area |
| 03:27.59 | MonMotha | prpplague: website? |
| 03:28.24 | prpplague | sorry don;t know it, remind me in the morning and i'll get info from the EE |
| 03:29.06 | MonMotha | ok |
| 03:29.16 | MonMotha | I've got a completed design ready for proto |
| 03:29.22 | MonMotha | 8 layers, 8/18 vias, 4/4 trace/space |
| 03:29.45 | MonMotha | I can trade off via size for 5/5 trace/space if needed |
| 03:29.53 | *** join/#edev chouimat (n=dieu@r2351064.cidc.net) |
| 03:30.08 | MonMotha | the rest of the board is routed at 5/5 (though not strictly necessary, it was very, very nice in some areas), so that wouldn't be easy to compromise on |
| 03:46.32 | *** join/#edev DrZimmerman (n=theo@2001:5c0:8fff:fffe:0:0:0:6af1) |
| 03:48.10 | Exiles | Exiles hammer board has he most Free space |
| 03:51.54 | *** join/#edev FireEgl (i=FireEgl@FireEgl.CJB.Net) |
| 05:38.08 | *** join/#edev FireEgl (i=FireEgl@FireEgl.CJB.Net) |
| 06:32.34 | *** join/#edev redblue (i=star@ppp027.108-253-207.mtl.mt.videotron.ca) |
| 07:11.44 | *** join/#edev steliosk (n=Stelios@ipa226.211.tellas.gr) |
| 07:40.32 | *** join/#edev hw (n=hw@p578b3905.dip0.t-ipconnect.de) |
| 08:38.45 | *** join/#edev silverpower (n=shrike@atheme/bofh/silverpower) |
| 10:47.05 | *** part/#edev SpaceMonkey (n=andrewmo@ovation1.gotadsl.co.uk) |
| 11:04.23 | *** join/#edev steliosk (n=Stelios@ipa226.211.tellas.gr) |
| 12:08.31 | *** join/#edev prpplague (n=dave@ppp-70-249-148-31.dsl.rcsntx.swbell.net) |
| 12:09.00 | prpplague | jbevren_: you alive? |
| 12:14.58 | Exiles | no |
| 12:15.26 | Exiles | but I have 18 gigs free on my hammer board |
| 12:16.10 | *** join/#edev brad_mssw (n=brad@shop.monetra.com) |
| 12:31.38 | prpplague | 18 gigs? |
| 12:31.54 | prpplague | Exiles: usb hd? |
| 12:32.19 | Exiles | yeah |
| 12:40.45 | *** join/#edev redb1ue (i=star@ppp139.109-253-207.mtl.mt.videotron.ca) |
| 13:01.19 | *** join/#edev chouimat|busy (n=dieu@r2351064.cidc.net) |
| 13:16.28 | *** join/#edev prpplague (n=DaveAnde@mail.americanmicrosystems.com) |
| 13:17.20 | *** join/#edev ico2 (n=ico2@modem-3105.chameleon.dialup.pol.co.uk) |
| 13:34.39 | chouimat | morning |
| 13:38.43 | prpplague | chouimat: greetings my fine french canadian coder |
| 13:40.31 | CP|too | somebody new came in? |
| 13:41.12 | prpplague | CP|too: hehe |
| 13:42.23 | chouimat | CP|too: ha ha ha ha ha |
| 13:48.27 | prpplague | CP|too: ping |
| 13:50.35 | chouimat | prpplague: I have a few interesting numbers for my gp ... seems the Volts are lower after each try to charge it ... |
| 13:50.50 | prpplague | greenphone? |
| 13:51.09 | prpplague | chouimat: remind me why you purchase a greenphone? |
| 13:51.25 | *** join/#edev koen (n=koen@dominion.kabel.utwente.nl) |
| 13:51.31 | koen | prpplague: ping |
| 13:51.32 | chouimat | prpplague: got it free |
| 13:51.38 | prpplague | koen: greeeeeings!!!!! |
| 13:51.52 | prpplague | koen: what is cooking? |
| 13:52.07 | koen | waiting for a delivery :) |
| 13:52.12 | prpplague | chouimat: ahh, free hardware |
| 13:52.37 | prpplague | koen: i didn't know your girlfriend was pregnant!!! hehehe |
| 13:52.46 | koen | heh |
| 13:52.49 | koen | not that :) |
| 13:53.04 | prpplague | koen: neo1973 stuff? |
| 13:53.21 | koen | no, a scanner to scan film negatives |
| 13:53.38 | prpplague | lovely |
| 13:53.55 | prpplague | koen: what brings you to #edev? gonna start hanging here? |
| 13:54.06 | *** join/#edev FireEgl (i=FireEgl@Sebastian.Atlantica.US.TO) |
| 13:54.18 | koen | I missed you in #ep93xx and ibot told me you were here :) |
| 13:54.45 | prpplague | ahh |
| 13:54.56 | chouimat | prpplague: I'm 100% for free hardware and it worked well when I was able to test it ... now I just need to get this "charging" issue fixed ... aka another one I guess |
| 13:55.09 | prpplague | koen: just was saying hello, didn't really need anything |
| 13:55.33 | prpplague | koen: we just started beta testing one of our development boards based on the s3c2410 |
| 13:55.37 | koen | I saw the hammer board was public knowledge now |
| 13:55.42 | prpplague | koen: http://www.elinux.org/Hammer_Board |
| 13:55.45 | prpplague | koen: yea |
| 13:55.56 | prpplague | koen: goes on sale in a couple weeks |
| 13:56.04 | prpplague | koen: shipping with 2.6.22 |
| 13:56.12 | prpplague | koen: cleaning up the source cd now |
| 13:56.46 | prpplague | koen: the PHB's are going over all the RoHS stuff now as well |
| 13:56.55 | chouimat | bbl need to do an HD transplant from dead computer to working one ... later |
| 13:57.01 | prpplague | chouimat: okie dokie |
| 13:57.38 | chouimat | prpplague: power outage last night fried a power supply (I hope it just this) ... |
| 13:57.57 | prpplague | chouimat: i'm thinking it's more likely you spilled beer on it |
| 13:58.26 | chouimat | no beer here |
| 13:58.29 | koen | prpplague: I wondered if you were open to trade a hammer in exchange for having OE support it |
| 13:58.48 | brad_mssw | prpplague: any idea on cost? ;) |
| 13:59.09 | prpplague | koen: yea, i have you on the list for eval kit, but we hadn't finished the RoHS compliancy by the time we were shipping |
| 13:59.28 | prpplague | brad_mssw: complete kit will be $229, just the cpu board in qty 1 will be $159 |
| 13:59.50 | prpplague | koen: should be able to hook you up with a kit in about 2-3 weeks |
| 13:59.55 | brad_mssw | prpplague: did you say there were or weren't pinouts for a battery-backedup RTC? |
| 14:00.12 | koen | prpplague: cool, thanks |
| 14:00.34 | koen | we got some influx from gumstix people, which reminded me of your mini boards |
| 14:01.17 | prpplague | brad_mssw: there are no pinouts for the battery backed RTC |
| 14:01.47 | prpplague | koen: yea, we are hoping to get alot of the gumstix business since you don't need special prototype equipment to work with the hammer |
| 14:03.31 | chouimat | seems I won't have to do anything .... it's back from the dead |
| 14:04.42 | prpplague | koen: have you heard of anyone else talking about the hammer board? |
| 14:05.09 | koen | yes, rwithby in #oe |
| 14:05.32 | koen | did you already send a message to linuxdevices.com? |
| 14:05.51 | prpplague | koen: no not yet, we are finalizing the retail website and finishing the beta testing |
| 14:05.57 | prpplague | koen: we were waiting for that |
| 14:06.27 | prpplague | koen: what kind of comments are you hearing about the hammer? good? bad? ugly? |
| 14:25.46 | koen | comments like "prpplague made a board" |
| 14:26.13 | CosmicPenguin | wow koen |
| 14:26.27 | koen | hey Jordan |
| 14:27.54 | chouimat | hi koen |
| 14:28.02 | koen | hey chouimat |
| 14:28.11 | CosmicPenguin | prpplague: sounds like you've generated some great buzz |
| 14:28.13 | CosmicPenguin | good for you |
| 14:28.13 | koen | prpplague: hop over to #oe to answer some hammer questions :) |
| 14:32.39 | *** join/#edev prpplague^2 (n=dave@ppp-70-249-148-31.dsl.rcsntx.swbell.net) |
| 14:36.09 | prpplague | CosmicPenguin: working on it |
| 14:43.47 | *** join/#edev g1powermac (n=g1powerm@adsl-18-80-154.sdf.bellsouth.net) |
| 14:55.01 | *** join/#edev tech404 (n=evan@southernae.com) |
| 15:17.15 | *** join/#edev chouimat (n=dieu@kde/developer/chouinard) |
| 16:02.31 | *** part/#edev tech404 (n=evan@southernae.com) |
| 16:02.31 | prpplague | CosmicPenguin: ping |
| 16:08.15 | CosmicPenguin | hi |
| 16:18.03 | prpplague | CosmicPenguin: hey, BUG() does return |
| 16:18.12 | prpplague | CosmicPenguin: or atleast all my tests indicate that it does |
| 16:18.16 | CosmicPenguin | its not supposed to |
| 16:19.06 | prpplague | CosmicPenguin: i did a small module with nothing more than a BUG() and a printk |
| 16:19.18 | prpplague | CosmicPenguin: it sure returned for me |
| 16:20.10 | CosmicPenguin | is CONFIG_BUG enabled? |
| 16:20.14 | prpplague | CosmicPenguin: do you have an example of some x86 code where they have BUG() as a none return |
| 16:20.52 | CosmicPenguin | prpplague: I have i386 bug code that very clearly goes into dead loops |
| 16:21.35 | prpplague | CosmicPenguin: can you point me to it on gitweb ? |
| 16:21.49 | CosmicPenguin | include/asm-i386/bug.h |
| 16:23.33 | prpplague | ahh |
| 16:25.14 | prpplague | CosmicPenguin: ahh ok, it looks like the define for the arm bug.h stuff is indeed none returning |
| 16:28.57 | *** join/#edev silverpo1er (n=shrike@adsl-234-207-62.cha.bellsouth.net) |
| 16:44.19 | *** join/#edev prpplague^2 (n=dave@ppp-70-249-148-31.dsl.rcsntx.swbell.net) |
| 17:18.30 | *** join/#edev lyakh (n=lyakh@dynamic-unidsl-85-197-26-24.westend.de) |
| 17:32.42 | prpplague | CosmicPenguin: so the question i have is, should BUG() always be a no-return function across all platforms? |
| 17:33.13 | CosmicPenguin | I think thats a matter of opinion |
| 17:34.33 | CosmicPenguin | which why i think its currently ambigious |
| 17:34.36 | prpplague | CosmicPenguin: ppc and i386 seem to be no-return |
| 17:35.01 | prpplague | mips is no-return |
| 17:35.32 | prpplague | sh is no-return |
| 17:36.01 | prpplague | s390 is no-return |
| 17:36.09 | *** join/#edev emil|k (n=emil|k@e179239245.adsl.alicedsl.de) |
| 17:36.21 | prpplague | even m68k is no return |
| 17:37.26 | prpplague | wait |
| 17:37.38 | prpplague | hmm |
| 17:39.54 | prpplague | that is incorrect |
| 17:40.05 | prpplague | CosmicPenguin: i read that wrong |
| 17:40.11 | prpplague | CosmicPenguin: its a while(0) |
| 17:41.15 | prpplague | CosmicPenguin: yea its the for(;;); that causes the no return |
| 17:45.26 | prpplague | CosmicPenguin: i think its going to boil down to the question of whether or not the arm kernel people thing BUG() should be a non-return function |
| 17:49.13 | CosmicPenguin | That sounds like the sort of thing Russell probably has an opinion on |
| 17:49.36 | CosmicPenguin | To me, a BUG() that doesn't crash the system is easy to ignore |
| 17:49.36 | prpplague | CosmicPenguin: yea, problem is rmk doesn't hang on any open irc channels |
| 17:49.46 | CosmicPenguin | probably for good reason |
| 17:51.48 | prpplague | CosmicPenguin: well, my problem is that A) the code that uses the BUG() in a non-return fashion they don't want to changed, and B) the code that defines BUG() as a returnable function they don't want to change |
| 17:55.03 | ds2 | have a header that secretly redefines bug in an obscure way? ;) |
| 17:56.45 | prpplague | ds2: its not about hacking it, i've already corrected the code so that it compiles cleanly, its about making the correction in a proper way that it gets accepted upstream |
| 18:09.11 | ds2 | prpplague: *nod* |
| 18:09.29 | ds2 | I do find that not all upstream reviewers are diligent :) |
| 18:10.13 | prpplague | ds2: indeed |
| 18:10.17 | prpplague | CosmicPenguin: i found the problem |
| 18:10.39 | prpplague | CosmicPenguin: indeed the BUG() in the asm-arm/bug.h is intended as a no-return |
| 18:10.48 | prpplague | CosmicPenguin: but the problem is in how it is defined |
| 18:12.55 | Exiles | cool |
| 18:13.02 | prpplague | Exiles: ?? |
| 18:13.09 | Exiles | I'll have my pin headers tommorow it looks like |
| 18:13.19 | prpplague | lovely |
| 18:14.04 | prpplague | wmat: ping |
| 18:17.09 | CosmicPenguin | prpplague: how it is defined? |
| 18:17.15 | prpplague | http://git.kernel.org/?p=linux/kernel/git/torvalds/linux-2.6.git;a=blob;f=include/asm-arm/bug.h;h=7b62351f097d2f49ff6fcfb7ea0e0c43e11d012f;hb=HEAD |
| 18:17.19 | prpplague | CosmicPenguin: line 15 |
| 18:19.01 | prpplague | CosmicPenguin: changed it to http://www.pastebin.ca/676013 |
| 18:19.47 | prpplague | CosmicPenguin: that code syntax is consistant with other bug.h defines |
| 18:20.36 | *** join/#edev MothaWork (n=monmotha@149.164.193.61) |
| 18:20.58 | MothaWork | ok, so I'm playing around with buildroot |
| 18:21.01 | MothaWork | this is proving...frustrating |
| 18:21.10 | prpplague | MothaWork: hehe |
| 18:21.17 | prpplague | MonMotha: more than OE? |
| 18:21.31 | MothaWork | I can't say I've had the "pleasure" of playing with OE |
| 18:21.35 | MothaWork | from what I've been told, it's worse |
| 18:21.48 | prpplague | MothaWork: hehe |
| 18:21.51 | prpplague | MothaWork: funny |
| 18:21.51 | MothaWork | this is doing dumb stuff though |
| 18:22.03 | prpplague | MothaWork: like what? |
| 18:22.04 | MothaWork | like trying to build cairo before it builds x11r7 |
| 18:22.13 | MothaWork | as you can imagine, that doesn't work too well |
| 18:22.49 | prpplague | MothaWork: yea thats a dependency issue, xwindows builds haven't really been maintained alot in buildroot |
| 18:22.57 | MothaWork | I see |
| 18:23.07 | MothaWork | I'm not familiar enough with it to know how to fix the dependencies, though |
| 18:23.13 | MothaWork | any resources I should consult? |
| 18:23.18 | prpplague | MothaWork: most of buildroot is designed around headless or custom display devices |
| 18:23.34 | MothaWork | well, this is going to be running a custom display app, but it will be drawing to X using WxWidgets |
| 18:23.39 | MothaWork | since WxWidgets doesn't have a framebuffer backend |
| 18:23.42 | prpplague | MonMotha: should be able to look at the makefiles |
| 18:23.51 | prpplague | MonMotha: it follows basic makefile dependencies |
| 18:23.53 | MothaWork | which makefiles? the <package>.mk? |
| 18:24.00 | prpplague | yea |
| 18:25.26 | ds2 | wouldn't directfb be lighter weight? |
| 18:26.04 | MothaWork | ds2: the issue is that WxWidgets pretty much only has X11, Qt, or GTK as a target |
| 18:26.08 | prpplague | ds2: indeed for fixed application usage, but many people want the advantage of having the ability to use pre-existing x apps |
| 18:26.23 | MothaWork | size isn't an issue, really |
| 18:26.27 | MothaWork | we've got 256MB of flash |
| 18:27.13 | ds2 | DOH |
| 18:27.46 | MothaWork | prpplague: in Makefile.in, there's some stuff to handle X |
| 18:27.51 | MothaWork | in package/Makefile.in, that is |
| 18:28.09 | MothaWork | it supports tinyx, xorg, an xggi, but no x11r7. I don't really know what that stuff does, though |
| 18:28.59 | prpplague | MothaWork: those are other variants of xwindows builds |
| 18:29.12 | MothaWork | (damn this lack of sloppyfocus) |
| 18:29.19 | MothaWork | yeah, I'd like to use x11r7 if possible |
| 18:29.29 | MothaWork | there are options for the others in the config stuff for buildroot |
| 18:31.11 | prpplague | x11r7 is the base code |
| 18:31.32 | MothaWork | I mean the 7.2 modularized stuff |
| 18:32.23 | prpplague | what part of x11r7 needs the cairo? |
| 18:32.39 | MothaWork | it needs Xlib |
| 18:33.10 | prpplague | oh, its trying to build cairo before x11r7? |
| 18:33.17 | MothaWork | yeah |
| 18:33.22 | MothaWork | cairo lists $(XSERVER) as a dep |
| 18:33.28 | MothaWork | but $(XSERVER) I don't think is being set |
| 18:33.32 | MothaWork | so I guess that is the problem |
| 18:34.45 | MothaWork | it also needs Xft |
| 18:38.40 | MothaWork | god this OS sucks |
| 18:41.02 | prpplague | MothaWork: which on is that? hehe |
| 18:41.02 | *** join/#edev redb1ue (i=star@ppp202.108-253-207.mtl.mt.videotron.ca) |
| 18:41.48 | MothaWork | I'm running Vista Business |
| 18:41.55 | MothaWork | even with all the eye candy turned off, it's still slow as hell |
| 18:41.56 | chouimat | ewww |
| 18:42.06 | MothaWork | it was what was recommended by Altium for designer |
| 18:42.11 | MothaWork | obviously they wanted some testing done |
| 18:42.20 | MothaWork | the answer to said testing is "it doesn't work very well" |
| 18:42.36 | MothaWork | turning off Aero (which is pretty much the only new feature of Vista) makes it quite a bit faster |
| 18:42.40 | MothaWork | also makes it look like Windows 2000... |
| 18:43.33 | MothaWork | and my X servers still run horribly |
| 18:43.40 | MothaWork | Xming runs better in WINE than it does on Vista |
| 18:43.47 | chouimat | ouch |
| 18:44.09 | MothaWork | cygwin X and Netsarang X are almost unusuably slow (Xming is, too) |
| 18:44.12 | MothaWork | I have not tried X-Win32 |
| 18:44.39 | chouimat | MothaWork: conclusion: stay away from vista |
| 18:45.40 | prpplague | Exiles: any more news on hammer dev? |
| 18:45.55 | prpplague | Exiles: i've not had time to repost the new source yet |
| 18:47.28 | MothaWork | chouimat: pretty much |
| 18:47.36 | MothaWork | this thing is going to get downgraded when we're done with this project |
| 18:47.40 | MothaWork | we'll just run XP on it |
| 18:47.50 | MothaWork | unfortunately, the EDA software, while pretty good, requires Windows |
| 18:47.58 | MothaWork | most commercial EDA software does |
| 18:48.07 | MothaWork | and I really don't feel like doing some of this design in Eagle |
| 18:49.02 | chouimat | MonMotha: a friend did some comparaison between Vista and OS X and OS X feel and was faster on the same hardware |
| 18:49.32 | MothaWork | I'm sure |
| 18:49.48 | MothaWork | Vista is an absolute pig of an OS |
| 18:54.16 | Exiles | do you have new new kernel by chance? |
| 18:54.32 | MothaWork | also, I'd really like to know why all of this stuff requires gfortran |
| 18:54.35 | MothaWork | or even bothers to look for it |
| 18:57.05 | MothaWork | also, I'd really like to know why all of this stuff requires gfortran |
| 18:57.08 | MothaWork | oops |
| 19:07.37 | MothaWork | yeah, I don't think anybody has ever actually tried to build using the modularized x11r7 thingie |
| 19:07.43 | MothaWork | the dependencies are very incomplete |
| 19:17.40 | prpplague | CosmicPenguin: ok, got to the root of the problemS |
| 19:17.45 | prpplague | CosmicPenguin: i.e. more than one |
| 19:20.00 | Exiles | I agree osx does feel and act alot faster on equal hardware |
| 19:24.33 | prpplague | CosmicPenguin: ping |
| 19:28.38 | CosmicPenguin | here |
| 19:31.45 | prpplague | CosmicPenguin: ok so here is the scoup, if BUG is disablled it is defined as nothing, i.e. "#define BUG() " |
| 19:32.26 | prpplague | CosmicPenguin: which means that BUG() should NEVER be used in a condition where a BUG() defined as null be used |
| 19:32.53 | prpplague | CosmicPenguin: does that make sense the way i explained it? |
| 19:33.07 | CosmicPenguin | nope |
| 19:33.17 | prpplague | :( |
| 19:33.19 | prpplague | great |
| 19:33.22 | prpplague | let me try again |
| 19:33.50 | prpplague | http://git.kernel.org/?p=linux/kernel/git/torvalds/linux-2.6.git;a=blob;f=arch/arm/mm/ioremap.c;h=75952779ce1991f68114dcbeef6a188c4c47d027;hb=HEAD |
| 19:33.54 | prpplague | see line 66 ? |
| 19:34.22 | prpplague | if BUG is turned off in the kernel config that becomes a blank line per the asm-generic/bug.h |
| 19:34.55 | prpplague | therefor it becomes a warning because it reaches the end of control without a return |
| 19:36.41 | prpplague | CosmicPenguin: that make sense? |
| 19:36.46 | *** join/#edev DrZimmerman (n=theo@84-72-44-194.dclient.hispeed.ch) |
| 19:36.49 | CosmicPenguin | technically its a warning either way |
| 19:36.55 | CosmicPenguin | because gcc isn't smart enough to know you won't return |
| 19:38.13 | prpplague | CosmicPenguin: ? |
| 19:38.22 | prpplague | either way? |
| 19:38.50 | CosmicPenguin | The problem isn't a return in the macro, its a return in your function, right? |
| 19:39.14 | prpplague | CosmicPenguin: no, the problem is that when BUG is turned off in the config, i get a warning |
| 19:39.24 | CosmicPenguin | you should get a warning either way |
| 19:39.41 | prpplague | CosmicPenguin: why should i get a warning either way? |
| 19:39.46 | CosmicPenguin | because the function doesn't return after a bug - and gcc isn't aware that the function won't return programattically |
| 19:40.01 | CosmicPenguin | all the parser sees is that there is no return at the end of the function |
| 19:40.44 | prpplague | CosmicPenguin: unless the BUG() define uses the noreturn attribute like in the i386 code |
| 19:40.48 | MothaWork | prpplague: another buildroot question if you have the time |
| 19:41.07 | prpplague | MothaWork: one sec, don't want to take up too much of CosmicPenguin's valuable time |
| 19:41.08 | CosmicPenguin | okay - yeah, if it uses noreturn, then the warning won't hapen |
| 19:41.15 | MothaWork | that's ok |
| 19:41.25 | prpplague | CosmicPenguin: the only time it will be an issue is if BUG() is turned off |
| 19:41.55 | prpplague | CosmicPenguin: so techincally, kernel coders shouldn't use BUG() anywhere that they can't subsitute a blank line |
| 19:42.12 | prpplague | CosmicPenguin: or atleast that's how i read that |
| 19:42.21 | CosmicPenguin | I'm not convinced thats a problem |
| 19:42.41 | prpplague | CosmicPenguin: what would you have me to test to convince you? |
| 19:42.41 | CosmicPenguin | but sure |
| 19:42.52 | CosmicPenguin | I understand that the problem might exist - I don't think anybody cares |
| 19:42.58 | prpplague | CosmicPenguin: ahh |
| 19:43.34 | CosmicPenguin | I mean, bigger picture - if developers expect BUG() to be invoked, then they're probably not cleaning up after the BUG() |
| 19:43.39 | prpplague | CosmicPenguin: well i grep'd all the i386 arch stuff, BUG() is never used in a manner that is consistant with the usage in ioremap.c for arm |
| 19:43.57 | CosmicPenguin | so return or not - your system is heading for hosedville |
| 19:44.27 | prpplague | CosmicPenguin: but isn't that bad practice when BUG() could be not used? after all its a config option |
| 19:44.51 | CosmicPenguin | lets put it this way - I didn't know it was a CONFIG option until this morning |
| 19:45.12 | prpplague | CosmicPenguin: interesting |
| 19:45.19 | CosmicPenguin | indeed |
| 19:45.21 | CosmicPenguin | lunch |
| 19:45.33 | prpplague | CosmicPenguin: thanks, i'll see if i can get some other feedback |
| 19:47.20 | MothaWork | prpplague: there's a patch file sitting in the directory that isn't getting applied by the AUTOTARGETS thingie |
| 19:47.38 | MothaWork | I can apply the patch manually, but my goal is to actually get to where "make world" works |
| 19:48.03 | MothaWork | this is libxcb in package/x11r7/libxcb |
| 19:49.00 | MothaWork | then I'm going to add support for my custom board |
| 19:49.12 | MothaWork | currently I'm just building for the dev board |
| 19:51.12 | Exiles | hmmmm |
| 19:51.20 | prpplague | MothaWork: hmm, is the patch listed as one of the ones needed in the makefile? |
| 19:51.21 | Exiles | sounds like thunder |
| 19:51.29 | prpplague | MothaWork: or is it named improperly? |
| 19:53.31 | MothaWork | prpplague: it looks named properly to me |
| 19:53.35 | Exiles | more thunder |
| 19:53.45 | MothaWork | the makefile is the bare-bones thing that just sets some vars then calls autotools |
| 19:54.12 | prpplague | MothaWork: brb |
| 19:56.13 | MothaWork | looks like it may be because libxcb is underneath x11r7, while autotools checks for package/<pkgname> directly |
| 19:56.17 | MothaWork | do you agree (when you get back)? |
| 19:56.31 | Exiles | yeah we have thunder and lighting |
| 19:56.40 | MothaWork | you need a THUNDER SHIELD |
| 19:57.08 | Exiles | wifi-max internet connection and computers are on ups |
| 19:57.45 | Exiles | so the chance of lighting killing my pc's is very slim |
| 19:58.36 | *** join/#edev T0mW (n=Tom@24.238.86.158.res-cmts.sth.ptd.net) |
| 20:06.00 | Exiles | grrrr |
| 20:10.27 | T0mW | Exiles: heh, it is Thursday |
| 20:10.32 | T0mW | ~Thursdays |
| 20:10.43 | T0mW | ~Thursday |
| 20:10.44 | purl | "This must be Thursday. I could never get the hang of Thursdays." (Arthur Dent) |
| 20:10.45 | Exiles | and I'm getting brownouts |
| 20:11.07 | T0mW | Exiles: ow, where are you located? |
| 20:11.20 | Exiles | thunder and lighting storm outside |
| 20:11.26 | T0mW | Exiles: heh, Eastern Pennsylvania? |
| 20:11.26 | Exiles | and I need a Bigger ups |
| 20:11.31 | Exiles | Oregon |
| 20:11.51 | T0mW | Exiles: oh, I have 1500W unit, the largest APC supply that Staples had on the shelf |
| 20:12.29 | Exiles | I want twice that I think |
| 20:12.48 | Exiles | I have ~700 watts of equipment |
| 20:13.21 | MothaWork | prpplague: that appeared to be the problem. I solved it with a horrible, horrible hack that should never see the light of day |
| 20:13.36 | MothaWork | Exiles: build your own |
| 20:13.55 | MothaWork | I should also point out that lots of PC equipment happily runs off DC |
| 20:14.47 | Exiles | the SLA batteries are expensive though |
| 20:15.35 | ds2 | there are cheap import SLA's |
| 20:16.15 | Exiles | http://www.apcc.com/resource/include/techspec_index.cfm?base_sku=SUA3000 something like this would do |
| 20:19.06 | Exiles | saw that lighting |
| 20:25.12 | prpplague | MothaWork: lovely |
| 20:25.24 | prpplague | T0mW: whats cookin? |
| 20:25.28 | prpplague | T0mW: hammer hacking? |
| 20:26.00 | prpplague | T0mW: btw, i removed the listing you added on the wiki for the hammer, we aren't advertising it as available yet |
| 20:30.22 | MothaWork | checking for /usr/share/sgml/X11/defs.ent... configure: error: cannot check for file existence when cross compiling |
| 20:30.28 | MothaWork | I'm not sure how to fix that one... |
| 20:35.58 | *** join/#edev g1powermac_ (n=g1powerm@adsl-210-142-198.sdf.bellsouth.net) |
| 20:45.02 | MothaWork | heck, I'm not entirely sure how the check for it is ending up in aclocal |
| 20:53.50 | T0mW | prpplague: I was on the phone with customer. cup of coffee in one hand, cigar in the other, and speaker phone on the desk. The whole time trying to get my mind into gear. heh |
| 20:54.46 | prpplague | hehe |
| 20:55.14 | T0mW | prpplague: ok on the listing. I added that to try to see if that was why I couldn't put a menu on the Hammer page of the wiki. If you goto the other project pages, they have a simple menu at the top of the page which hyperlinks to the topics? |
| 20:55.14 | MothaWork | this is really silly, because it means that the X server flat out is not cross compilable simply due to documentation |
| 20:55.19 | MothaWork | that's gotta be an oversight... |
| 20:56.54 | ds2 | MothaWork, seem to recall the gumstix folks might have a working X in build root, maybe you can steal hints from that? |
| 20:57.12 | prpplague | MonMotha: like i said, the xwindows portion of buildroot has been somewhat neglected |
| 20:57.15 | MothaWork | ds2: what flavor of X are they using, thoguh? |
| 20:57.25 | MothaWork | prpplague: this would appear to be an issue with X itself, not buildroot |
| 20:57.25 | prpplague | MothaWork: tinyx more than likely |
| 20:57.50 | MothaWork | this is just the configure script flat out not working whenever it is set to cross compile |
| 20:58.32 | ds2 | MothaWork: donno |
| 20:59.27 | MothaWork | yes, it is |
| 21:00.15 | T0mW | anyone use the TSC2003 i2c touchscreen driver by Bill Gattliff? |
| 21:00.28 | prpplague | not i |
| 21:00.54 | T0mW | prpplague: I figure since the i2c is there, I would use that chip instead of and ADS thingy |
| 21:01.26 | prpplague | T0mW: yea, we have the ads stuff in stock so thats what i probably will use |
| 21:03.05 | T0mW | prpplague: I ordered two more of those EDMGRB8KJF 4K color STN displays and started doing a layout for a motherboard for the Hammer that would mount on the back of the display |
| 21:03.20 | prpplague | lovely |
| 21:03.28 | prpplague | T0mW: where did you order them from earthlcd? |
| 21:03.33 | T0mW | prpplague: yea |
| 21:03.48 | prpplague | T0mW: you gonna post schematics and such? |
| 21:03.53 | T0mW | prpplague: there is only a single HOST usb port on the S3C2410, right? |
| 21:03.58 | T0mW | prpplague: yes |
| 21:04.12 | T0mW | prpplague: heh, and sell boards. :-) |
| 21:04.30 | prpplague | T0mW: well.... their are two usb ports, one dedicated usb host, the other can be usb host or usb slave |
| 21:04.39 | ds2 | speaking of LCD, anyone got a document/page on going from datasheet numbers to values suitable for putting in #define's or passing to the framebuffer driver's commandline? |
| 21:04.42 | T0mW | prpplague: AH! great! |
| 21:05.00 | prpplague | T0mW: in addition, you can use something like the TUSB2046B as a four port hub, we have a module for that |
| 21:05.03 | T0mW | prpplague: I thought that was the case, but I wasn't sure |
| 21:05.31 | T0mW | prpplague: I was thinking about a USB port expander, but I don't know how "wild" I wanted to go. |
| 21:06.53 | prpplague | T0mW: hehe usb port expander? |
| 21:07.15 | prpplague | T0mW: you just mean a usb controller that has gpio? |
| 21:07.17 | T0mW | prpplague: Same type of O/S structure as with the Zipit: load kernel from flash into RAM, then have the kernel boot from the SD card. IOW, the card is not removable |
| 21:07.44 | T0mW | prpplague: no, I meant that four port hub expander |
| 21:08.32 | T0mW | prpplague: I will have to sneak the Hammer work in when I have time. I have a customer meeting on Wednesday and I don't have a freakin' thing done right now. |
| 21:09.08 | prpplague | T0mW: ok you mean just a standard 4 port hub then? |
| 21:09.09 | T0mW | prpplague: yes |
| 21:09.31 | prpplague | T0mW: ok yea, i'll email you the schematic for using the TUSB2046B |
| 21:10.04 | prpplague | T0mW: as far as using the sd/mmc card, apex has the capability of reading the kernel and rootfs from a partition on the sd/mmc card |
| 21:10.17 | prpplague | T0mW: i just have to finish tweaking the spi interface for the hammer |
| 21:10.46 | T0mW | prpplague: I don't think it will appear on the board if the circuitry is too complex. I want to get into a proto quickly and the LCD + Touch + SPI ethernet would need to be done. Other stuff can wait. But, the schematic would be good to look at and maybe implement if not too much of a PITA |
| 21:11.04 | T0mW | prpplague: nod, Apex should work for that |
| 21:11.41 | T0mW | prpplague: I would like to put an OE version together using uClibc. That way, apps can also be built for the target system using buildroot. |
| 21:12.18 | prpplague | indeed |
| 21:12.23 | T0mW | buildroot for hacking apps together, OE for the basic system with the current off-the-shelf stuff. |
| 21:12.30 | prpplague | T0mW: my problem is i never can get OE to build anything |
| 21:12.53 | prpplague | T0mW: but i have to give it another try soon anyway |
| 21:13.00 | T0mW | prpplague: it takes TIME, TIME to learn the freakin' thing, and they keep changing sh!t |
| 21:14.54 | T0mW | I released the OpenZipitDistro V1.3 and forgot to compile it as 32bit for the host tools. It was compiled under an X86_64 host and... What can I say? The poor 32bit bast^H^H^H^H people are left with an app that won't program their Zipit flash. doh! |
| 21:15.38 | T0mW | prpplague: two AMD X2 systems here, one has a 32bit o/s the other is 64bit o/s. |
| 21:16.08 | T0mW | prpplague: I'm down to only two computers running 24/7 |
| 21:18.35 | MothaWork | well, I fixed it, though I'm not sure how to make a patch |
| 21:34.06 | T0mW | prpplague: I will have to do surgery on the Hammer to lift the output pin on the 3v3 regulator. That way, I can use the 3v3 output pin as a 3v3 input supply pin to power the Hammer. Otherwise, there will be h3ll to pay when using a 5volt supply to the LCD and only 3.3volts to run the Hammer (LCD_VDx lines may be overvoltaged due to the 5volt LCD). |
| 21:34.43 | T0mW | or, add some 3v3 to 5volt level shifter ICs |
| 21:34.51 | prpplague | T0mW: i folowd the first part, didn't follow the second |
| 21:36.18 | T0mW | prpplague: well, +5v to run the hammer, but the LCD I/O pins would be 3.3v due to the internal 3.3volts of the Hammmer board. I would need two voltage supplies if I don't modify the Hammer. A +5volt to the hammer and a +3.3volt to the LCD. |
| 21:37.33 | *** join/#edev FireEgl (i=FireEgl@Sebastian.Tcldrop.Com) |
| 21:40.43 | prpplague | T0mW: oh ok, it was just how you worded it |
| 21:40.53 | prpplague | T0mW: you can just remove the ldo all together if you want |
| 21:49.07 | ds2 | prpplague: might have asked before but what's the reason why you didn't bother with a switcher? space? |
| 21:49.19 | prpplague | ds2: yea |
| 21:49.32 | prpplague | ds2: and because the original design was with an ldo |
| 21:49.53 | prpplague | ds2: the customer didn't end up using the design, so we just used it as is |
| 21:49.58 | ds2 | khilman@mvista.com changed: |
| 21:49.58 | ds2 | <PROTECTED> |
| 21:50.33 | prpplague | ds2: if it sells we will release the LP version which will have 2AA battery support and suspend/resume functionality |
| 21:50.35 | ds2 | that was unexpected |
| 21:50.47 | prpplague | ds2: ? |
| 21:51.07 | ds2 | prpplague: I was wondering if a buck/boost might be better |
| 21:51.33 | prpplague | ds2: the TPS60000 series is what we will be using |
| 21:51.35 | ds2 | oh kde paste buffers...sometimes it pastes stuff months old |
| 21:52.58 | *** join/#edev redb1ue (i=star@ppp050.108-253-207.mtl.mt.videotron.ca) |
| 21:56.47 | prpplague | ds2: http://focus.ti.com/docs/prod/folders/print/tps63000.html |
| 21:57.31 | prpplague | ds2: so for 2AA's should be able to run the hammer+usb host |
| 21:58.04 | prpplague | ds2: if you go to 4AA's should be able to last awhile even running power hungry usb device such as a wireless dongle |
| 21:58.36 | ds2 | nice |
| 21:58.39 | ds2 | I like it |
| 21:59.12 | ds2 | of course usb host would still need another booster for +5, right? |
| 21:59.40 | prpplague | yea you'd need to do two boosters to be really effective |
| 22:00.11 | prpplague | 2AA's for the hammer, and 2AA's for the usb |
| 22:00.24 | ds2 | yep |
| 22:00.40 | prpplague | fully pushed the hammer carrier board is pulling about 165mA |
| 22:00.40 | ds2 | there is a maxim 2AA to 5V converter I been eyeing for a long time |
| 22:01.11 | prpplague | yea, we don't use maxim parts because there is always a problem getting them in bulk qty |
| 22:01.33 | ds2 | I am well aware of their sourcing issues ;) |
| 22:01.57 | MonMotha | prpplague: Maxim pretty much seems to target low-medium volume exclusively |
| 22:02.08 | prpplague | MonMotha: yea |
| 22:02.22 | prpplague | MonMotha: more than 1000qty and they freak out |
| 22:02.27 | ds2 | for 165mA, I'd think you can get by with a charge pump |
| 22:02.45 | ds2 | I like how maxim claims they never obsolete parts ;) |
| 22:03.13 | ds2 | <sarcasm></sarcasm> there if it isn't obvious |
| 22:03.36 | ds2 | <-- has a strong dislike for power inductors |
| 22:04.15 | MonMotha | ds2: they don't |
| 22:04.22 | MonMotha | they just jack the price up to where nobody wants them |
| 22:04.32 | MonMotha | but if you call them up and agree to pay that price, they WILL make them for you |
| 22:04.43 | ds2 | thought it was the ever growing leadtime strategy they use |
| 22:04.49 | MonMotha | there's some of that, too |
| 22:04.58 | ds2 | Oh? hmm |
| 22:05.01 | MonMotha | leadtime is 1 year and it's $20/chip |
| 22:05.04 | MonMotha | but you can still get them!! |
| 22:05.27 | MonMotha | of course, the date code may be from 10 years ago, back when they actually made them (i.e. they're sourcing them from some "obsolete semiconductor components" company in China) |
| 22:05.43 | ds2 | does maxim have a fab at all? |
| 22:06.04 | MonMotha | I don't know that they do |
| 22:06.12 | MonMotha | I think they outsource their fab work and retain the mask sets |
| 22:06.49 | prpplague | yea, i think they are fabless which might explain their low volumes |
| 22:07.00 | prpplague | they probably buy "off-time" from the fab housers |
| 22:07.02 | prpplague | houses |
| 22:07.26 | MonMotha | also explains why they can afford to never EOL anything |
| 22:07.39 | MonMotha | if they keep the mask sets forever, it's not a big deal to just buy some time and do a small run |
| 22:07.41 | ds2 | actually that is not strictly true |
| 22:08.09 | ds2 | think they actually EOL'ed the MAX038 with a note about NSC not being able to make any more for them |
| 22:12.09 | jbevren | lol |
| 22:12.13 | jbevren | I showed my boss the hammer |
| 22:12.18 | jbevren | he said wtf is that (he's not a tech guy) |
| 22:12.25 | jbevren | I said its a 200mhz process controller |
| 22:12.35 | jbevren | he said woah. thats small. (he knows what a process controller is) |
| 22:12.37 | jbevren | :D |
| 22:13.04 | jbevren | then he started asking me about IOs and A/D stuff |
| 22:14.52 | jbevren | he was pretty impressed when I told him it ran linux (then explained that linux is a unix-like OS) |
| 22:15.58 | prpplague | hehe |
| 22:16.12 | prpplague | jbevren had any more time to tinker? |
| 22:16.50 | MonMotha | prpplague: well, X is still giving me trouble with buildroot |
| 22:17.08 | MonMotha | some weird issue with GLX, which I of course don't care about to begin with |
| 22:17.25 | jbevren | prp: not today no |
| 22:17.26 | prpplague | MonMotha: hehe indeed |
| 22:17.29 | jbevren | today was my last day of work |
| 22:17.31 | prpplague | jbevren: ahh ok |
| 22:17.35 | jbevren | so now I get to scramble to find income :) |
| 22:17.42 | prpplague | jbevren: what???/ |
| 22:17.47 | prpplague | jbevren: what happened? |
| 22:17.47 | MonMotha | we may be able to have WxWidgets talk GTK and have GTK talk to DirectFB |
| 22:17.49 | MonMotha | which would remove X entirely |
| 22:18.47 | prpplague | something really weird going on with the immigration support web site i use |
| 22:19.06 | *** join/#edev GPSFan (n=kenm@12.10.255.248) |
| 22:19.06 | *** mode/#edev [+o GPSFan] by ChanServ |
| 22:19.06 | prpplague | lots of "new" users and posting really troll like stuff |
| 22:19.31 | prpplague | GPSFan: greetings |
| 22:19.51 | GPSFan | prpplague: hey, back early, anything interesting hammer wise? |
| 22:20.20 | jbevren | prp: seasonal job |
| 22:20.49 | prpplague | jbevren: oh ok |
| 22:21.03 | jbevren | prp: maybe we should privmsg another channel to work on:) |
| 22:21.07 | jbevren | lol :) |
| 22:21.09 | prpplague | GPSFan: nothing major, T0mW broke a cap of the back side of the pcb when removing the pcb from the socket |
| 22:21.22 | prpplague | jbevren: for the hammer? |
| 22:21.24 | jbevren | I bent two pins :) |
| 22:21.30 | jbevren | prp: lol yeah (paranoia mode) :) |
| 22:21.34 | GPSFan | prpplague: yeah, I checked in last night from the hotel, he told me. |
| 22:21.40 | prpplague | naw, i don't mind people listening it |
| 22:21.41 | prpplague | in |
| 22:21.46 | prpplague | GPSFan: ahh ok |
| 22:21.48 | jbevren | hehe |
| 22:21.51 | prpplague | nothing else major |
| 22:21.55 | prpplague | just getting geared up |
| 22:21.56 | GPSFan | good. |
| 22:23.01 | jbevren | so where are teh troll posts? :) |
| 22:23.10 | jbevren | lol |
| 22:24.04 | ds2 | going to setup an "investment company" to help people move money out of off shore accounts like those in Nigeria? ;) |
| 22:24.14 | prpplague | http://www.visajourney.com |
| 22:24.27 | prpplague | website for helping people with US immigration issues |
| 22:24.41 | jbevren | hm nice booster |
| 22:25.15 | GPSFan | prpplague: did you ever get anywhere with the enc28j60 kernel 2.6 driver? |
| 22:26.08 | jbevren | I should get one and build a breakout board |
| 22:26.45 | prpplague | yea, the 2.6 driver is pretty close to the same as the 2.4 |
| 22:26.49 | prpplague | only minor changes |
| 22:27.06 | prpplague | however what i have working does not use the s3c2410's spi subsystem driver |
| 22:27.17 | prpplague | GPSFan: its a start |
| 22:27.21 | GPSFan | prpplague: could I get a patch? |
| 22:27.39 | prpplague | GPSFan: tuesday soon enough? |
| 22:27.50 | prpplague | GPSFan: i still have to clear a couple of items with the boss |
| 22:27.50 | GPSFan | I don't want to go back to a 2.4 kernel, Tue's is fine. |
| 22:28.08 | GPSFan | I don't want to go back to a 2.4 kernel, Tue's is fine. |
| 22:30.53 | prpplague | hehe |
| 22:30.54 | prpplague | ok |
| 22:30.59 | prpplague | no need to go to 2.4 |
| 22:31.50 | prpplague | i've gotta head out and pay a few bills, bbiaw |
| 22:32.04 | GPSFan | bye.. |
| 23:13.37 | Exiles | so, now that I've been flooded, whats next |
| 23:25.23 | *** join/#edev prpplague (n=dave@ppp-70-244-172-90.dsl.rcsntx.swbell.net) |
| 23:41.32 | prpplague | GPSFan: so how do you feel about the Hammer so far? |