**** BEGIN LOGGING AT Thu Jun 18 02:59:57 2009
Jun 18 13:46:22 <prpplague> man openocd is really gone to hell in a hand basket
Jun 18 13:47:43 <Kebianizao|work> prpplague: what do you mean?
Jun 18 13:50:17 <prpplague> Kebianizao|work: the current openocd is completely bloated, over engineering with useless features, and just about completely unusable
Jun 18 13:50:42 <Kebianizao|work> prpplague: heh! I find it useful
Jun 18 13:50:52 <Kebianizao|work> what useless feature you can think about?
Jun 18 13:51:30 <prpplague> Kebianizao|work: it would take me an entire week to go over it all
Jun 18 13:51:54 <Kebianizao|work> I have the time :) (just kidding)
Jun 18 13:52:10 <prpplague> yea but i don't, atleast not this week
Jun 18 13:52:43 <Kebianizao|work> prpplague: have you expressed your opinion in the ML, I think opinions like yours are welcome
Jun 18 13:53:06 <Kebianizao|work> and you don't need to give the full list, you can start with what you find worst
Jun 18 13:53:33 <Kebianizao|work> it's clear that they won't change it upside down, but still some comments may be useful
Jun 18 13:53:50 <prpplague> Kebianizao|work: yes i have, i've given up on that
Jun 18 13:54:39 <Kebianizao|work> oh, could you point me to that? I would like knowing. I guess approx date is enough
Jun 18 13:55:11 <prpplague> august/september of 2008
Jun 18 13:55:32 <Kebianizao|work> I see, thanks
Jun 18 13:55:54 <prpplague> drath been in the channel lately?
Jun 18 13:56:10 <Kebianizao|work> quite new here, sorry I couldn't tell
Jun 18 13:56:17 <prpplague> ahh
Jun 18 14:01:45 <Kebianizao|work> a pity but I've bee unable to find your e-mail in the ML archive :S
Jun 18 14:03:57 <prpplague> oh well doesn't really matter anyway, we have a fork of one of the older versions
Jun 18 14:04:35 <prpplague> just that we get inquires every once in awhile about using the yagarto openocd version
Jun 18 14:08:11 <Kebianizao|work> prpplague: is that fork available?
Jun 18 14:09:17 <prpplague> http://www.elinux.org/images/8/80/Openocd.tar.gz
Jun 18 14:14:08 <Kebianizao|work> do you maintain that as open source? I mean do you have a tracker, repository? do you snapshot openocd or also add some changes?
Jun 18 14:16:35 <Kebianizao|work> at first sight I miss tcl which I guess is part of you define as bloat
Jun 18 14:17:43 <prpplague> Kebianizao|work: yes we have a repo, and yes tcl is a major part of the bloat
Jun 18 14:18:59 <Kebianizao|work> I don't want to abuse your time, but if you consider your fork world (re)viewable I'd like having a link. I'm quite curious about it
Jun 18 14:20:33 <prpplague> Kebianizao|work: not as of yet, thats the reason i joined the channel today, i used to hang out here with drath in the good old days, and i need to double check with him a  few things before making it available
Jun 18 14:20:54 <Kebianizao|work> ah, that's ok
Jun 18 16:46:35 <prpplague> dirk2: hey hey bud
Jun 18 16:46:49 <dirk2> hey
Jun 18 18:11:14 <prpplague> drath: greetings
Jun 18 18:11:17 <drath> hey prpplague 
Jun 18 18:11:59 <drath> "completely sick" is probably among the nicer things i could say about the current openocd state ;)
Jun 18 18:12:05 <prpplague> drath: what cooking over in your neck of the woods?
Jun 18 18:12:15 <prpplague> drath: hehe
Jun 18 18:15:04 <drath> prpplague: is it ok if we discuss this topic here? or would you prefer an email?
Jun 18 18:15:21 <prpplague> drath: here is fine for me
Jun 18 18:16:33 <prpplague> drath: basically we are looking at rolling back to around the end of 2007 for version
Jun 18 18:16:59 <prpplague> drath: fix all the know issues with that rev so it is stable
Jun 18 18:17:22 <prpplague> drath: do a release, and then use that as a basis for improvements/changes
Jun 18 18:17:59 <drath> prpplague: yeah... it's been far too much "reworking" for questionable reasons, and dozens of new features/technologies without caring about improving the existing codebase...
Jun 18 18:18:42 <drath> prpplague: did you or the other folks you mentioned try to express your concerns on the list?
Jun 18 18:19:16 <drath> prpplague: the situation has been rather problematic for me, as i don't have time to contribute myself, so it's hard to stop those who are contributing
Jun 18 18:19:50 <prpplague> drath: yea back in august/september of 2008 when i was doing some of the work on the cortex-a8 and icepick dev
Jun 18 18:20:51 <prpplague> drath: i had a couple conversations with the current maintainer
Jun 18 18:21:40 <prpplague> drath: it's just that i think all the changes, new features and re-works have gotten out of control
Jun 18 18:26:39 <drath> prpplague: yeah, definitely... i don't think you'll be able to convince oyvind to change the current development, so a fork might really be necessary... it would be nice if you could keep me posted, maybe we can eventually bring both lines of development back together one day
Jun 18 18:27:27 <prpplague> drath: right, i was considering using either the linuxtogo.org site or getting a repo up and running on the elinux.org site (CELF sponsored)
Jun 18 18:27:37 <prpplague> drath: you have a preference?
Jun 18 18:28:24 <drath> prpplague: whatever suits you best
Jun 18 18:28:59 <prpplague> drath: probably go with the linuxtogo.org site if it isn't too difficult to get setup
Jun 18 18:29:37 <drath> prpplague: i suppose a branch on the original openocd site would cause too much confrontation, otherwise i'd offer you admin access there
Jun 18 18:30:48 <prpplague> drath: yea i think that would be problematic, and based on the incidents with the GPE stuff, it might make more sense to be seperate
Jun 18 18:43:55 <drath> prpplague: allright, gotta go... i'll try to check by every once in a while
Jun 18 18:44:01 <drath> prpplague: bye
Jun 18 18:44:14 <dirk2> sorry, but any hint about which target architecture/board you are talking here
Jun 18 18:44:22 <dirk2> ?
Jun 18 18:45:37 <prpplague> dirk2: hehe, openocd in general
Jun 18 18:46:58 <dirk2> yes, ok, I got this ;) Any special target in mind? Seems that Dave is helping a little with TI stuff, btw
Jun 18 18:48:36 <prpplague> dirk2: dave who?
Jun 18 18:49:10 <dirk2> David Brownell
Jun 18 18:55:58 <prpplague> ahh
Jun 18 18:56:11 <prpplague> dirk2: no special target in mind
Jun 18 18:56:36 <prpplague> dirk2: just a number of us are very unhappy with the overall direction that openocd is headed, including drath
Jun 18 19:02:56 <prpplague> dirk2: when the TCL stuff was added and they started changing the config file structures, thats when everything started going down hill
Jun 18 19:07:08 <dirk2> ok. my hope is that it will calm down some time and that somebody with the time and knowledge will help 
Jun 18 19:09:20 <prpplague> dirk2: for which item? the cortex-a8/icepick stuff?
Jun 18 19:10:50 <dirk2> prpplague: Yes, OMAP3. Icepick seems to be basically solved. Cortex A8 is missing, though. There are some scripts that can do already a lot with Cortex A8, but converting them to some legacy OpenOCD (C-) code and adding the missing stuff (cache, MMU) is still missing
Jun 18 19:11:10 <prpplague> dirk2: ahh
Jun 18 19:11:30 <prpplague> dirk2: i've lost all interest in the beagle
Jun 18 19:11:43 <prpplague> dirk2: (not that i had that much to start with)
Jun 18 19:12:11 <dirk2> prpplague: and, you are right, as soon as we have something working, it is broken by some new changes again, and the comment is then "your previously working code was wrong"
Jun 18 19:12:24 <prpplague> dirk2: yep
Jun 18 19:16:27 <dirk2> unfortunately, none of the core developers seem to be here often. Instead, you get sometimes ~100 mails from mailing list per day, not able to track them
Jun 18 19:16:37 <prpplague> yea
Jun 18 19:17:08 <prpplague> dirk2: mailing list has it's usage, but for something that is moving as fast as openocd has been, interactive conversation tends to be better(IMHO)
Jun 18 19:18:48 <dirk2> anyway, good luck with your fork, bye
Jun 18 19:20:45 <prpplague> dirk2: later bud
Jun 18 19:20:52 * prpplague waves bye to dirk2 
Jun 18 20:57:33 <Zoiah> Is there any JTAG tool that supports FT2322 (Olimex ARM-USB-OCD), Motorola MPC8245 and 2 Intel Strataflashes 8x wired in 16x, using the memory mapped flash instead of having to figure out which pins the flash is attached to exactly?
Jun 18 20:59:09 <prpplague> Zoiah: i'm not familiar with the MPC8245, what is that a MIPS core?
Jun 18 20:59:40 <Zoiah> It's a PPC core.
Jun 18 21:00:11 <Zoiah> I often found it in networking equipment and some small storage things.
Jun 18 21:00:35 <prpplague> Zoiah: ahh, you will probably have trouble getting that to work via jtag without some serious docs
Jun 18 21:00:57 <Zoiah> UrJTAG supports MPC824x just fine, but not the 2 x8 in x16 setup.
Jun 18 21:02:13 <prpplague> Zoiah: ahh, was unaware of that
Jun 18 21:02:25 * prpplague makes a note to get the latest UrJTAG
Jun 18 21:03:00 <Zoiah> I got pointed at TopJTAG which is a closed-source Win32 application, but that uses pintoggling directly to the flashchip.
Jun 18 21:03:17 <prpplague> ahh
Jun 18 21:03:23 <Zoiah> And I do not have schematics of my devices (3com 3870, edgecore es4549, dell powerconnect 3324), so this is probably not feasible for me.
Jun 18 21:04:19 <Zoiah> All these devices use the same OEM and basic structure (MPC8245 behind JTAG COP pinuot and 2 PLDs behind PLD JTAG pinout).
Jun 18 21:05:31 <roxfan> urjtag only does boundary scan, but i guess you don't need debugging here
Jun 18 21:05:56 <Zoiah> UrJTAG does flashing through memory i/o in the cpu core.
Jun 18 21:06:11 <Zoiah> So if you know where your flash is mapped into memory, you can flash that way.
Jun 18 21:06:32 <roxfan> hmm it does? that's new
Jun 18 21:07:21 <Zoiah> Openwince also did this.
Jun 18 21:07:27 <Zoiah> http://buffalo.nas-central.org/wiki/JTAG_Software/Openwince
Jun 18 21:08:41 <roxfan> this looks to me like boundary scan...
Jun 18 21:10:39 <Zoiah> Yes, it uses boundary scan to do memory i/o to the CPU.
Jun 18 21:12:42 <Zoiah> I am new to these kinds of things, so I might be using the wrong terms. :)
Jun 18 21:12:51 <roxfan> i think it just toggles pins connected to the flash
Jun 18 21:13:16 <roxfan> at least that's what was happening when i tried it a year or so ago
Jun 18 21:13:22 <Zoiah> How would it know to which pins the flash is connected? And why would it require the location in the memory of the cpu?
Jun 18 21:14:12 <Zoiah> Because if UrJTAG knows to which pins it's connected, I could feed that information to TopJTAG maybe.
Jun 18 21:15:43 <roxfan> well, the pin descriptions are specified in the BSDL file
Jun 18 21:16:34 <roxfan> so you basically need to mimic the signals on them as if the cpu is reading or writing data
Jun 18 21:16:44 <Zoiah> The pins of the cpu, but i can imagine there's more than 1 way to wire a flashchip to a chip.
Jun 18 21:17:20 <roxfan> yeah, often it gets tricky
Jun 18 21:18:36 <Zoiah> So, because the location of the flash chip in the memory is defined, I assumed it was doing memory i/o on the cpu.
Jun 18 21:18:48 <Zoiah> That should be possible, right?
Jun 18 21:19:04 <Zoiah> http://www.youtube.com/watch?v=JhojoT1o1M4
Jun 18 21:19:09 <Zoiah> Sorry, did not mean to paste that.
Jun 18 22:11:32 <prpplague> man o man i wish i could find an old copy of openocd for winblows
Jun 18 22:19:04 <roxfan> http://www.argussoft.ru/webroot/delivery/files/library/atmel/at91/Eclipse/openocd-r247-20071230.exe
Jun 18 22:19:24 <prpplague> oh dandy, thanks
Jun 18 22:19:35 <prpplague> roxfan: you have anything closer to R734?
Jun 18 22:20:08 <roxfan> not at the monent
Jun 18 22:20:12 <roxfan> but i'll look around
Jun 18 22:20:58 <prpplague> thanks
Jun 18 22:23:42 <roxfan> http://www.siwawi.arubi.uni-kl.de/avr_projects/arm_projects/winarmtests/
Jun 18 22:26:16 <prpplague> thanks
Jun 18 22:26:19 * prpplague heads home
**** ENDING LOGGING AT Fri Jun 19 02:59:58 2009