#emc-devel | Logs for 2006-03-31

[01:23:33] <steves_logging> steves_logging is now known as steve_stallings
[01:24:14] <steve_stallings> Yes, there are free tools is you can stomach Windows. SWP knows details I think. We have commercial as does he.
[01:24:39] <jmkasunich> what kind of windows is needed? I have a W95 box ;-)
[01:25:03] <cradek> jmkasunich: I bet you could install win95 nicely in qemu
[01:25:06] <jepler> jmkasunich: Remember we were talking about bdi2 and whether it could build emc2 with my makefiles? The make 3.80 that I built is very small (64kb gzipped) and can be installed with just a few shell commands as root (gunzip, mv, chmod).
[01:25:44] <steve_stallings> Not sure but I would bet against W95, possibly even W98.
[01:26:19] <jmkasunich> cradek: what did you say between "qemu" and "3.80"? its blank here
[01:27:04] <cradek> jepler> jmkasunich: Remember we were talking about bdi2 and whether it could build emc2 with my makefiles? The make 3.80 that I built is very small (64kb gzipped) and can be installed with just a few shell commands as root (gunzip, mv, chmod).
[01:27:38] <jmkasunich> ah
[01:28:15] <jmkasunich> so if somebody mailed me that gz and told me the commands I could put it on the 2.xx farm slot
[01:28:36] <jepler> http://emergent.unpy.net/files/sandbox/make.gz
[01:28:40] <jmkasunich> how'd we get from w95 to that anyway?
[01:28:53] <jepler> gunzip make.gz; chmod +x make
[01:29:03] <jepler> as root: mv /usr/bin/make /usr/bin/make.old; mv make /usr/bin/make
[01:29:11] <cradek> even if it builds on bdi2, it's going to be a crippled emc2 installation without gtk etc.
[01:29:20] <cradek> or bwidget
[01:29:35] <jmkasunich> there is gtk on that box
[01:29:42] <jmkasunich> I think?
[01:29:47] <jepler> cradek: tkemc should still run
[01:29:50] <cradek> huh, I'm surprised
[01:29:50] <jepler> but the config picker won't
[01:30:00] <jepler> no idea about halscope, which is the thing in gtk
[01:30:20] <jmkasunich> I never actually run emc on the farm, only compile
[01:30:36] <jmkasunich> those things slow to a crawl if you start X
[01:30:42] <jmkasunich> they are text only installs
[01:31:00] <jmkasunich> shitty video card and very limited CPU/memory
[01:31:03] <cradek> are these /compile_farm pages up to date?
[01:31:15] <cradek> I thought TNG built last time I looked
[01:31:16] <jepler> cradek: i figured they broke with the change to dreamhost
[01:31:23] <jmkasunich> it did
[01:31:25] <cradek> yeah, maybe so.
[01:31:27] <cradek> ok
[01:31:39] <cradek> checking for GTK version... none
[01:31:40] <cradek> configure: WARNING: GTK not found, some utilities (halmeter, halscope) will not be built
[01:31:40] <jmkasunich> did the dns change go thru?
[01:31:47] <cradek> bdi2 does not seem to have gtk
[01:31:58] <jmkasunich> oh well...
[01:32:03] <jmkasunich> did the dns change go thru?
[01:32:12] <cradek> yes I think the dns is all stable now
[01:32:21] <jmkasunich> today's the 30
[01:32:28] <jepler> it's based on redhat 6.2?
[01:32:35] <cradek> I think so
[01:32:35] <jmkasunich> 30th, and the farm hasn't updated since the 26th....
[01:32:38] <jmkasunich> yes
[01:33:04] <jepler> lftp archive.download.redhat.com:/pub/redhat/linux/6.2/en/os/i386/RedHat/RPMS> ls gtk+-*
[01:33:08] <jepler> -rw-rw-r-- 6 ftp ftp 847121 Mar 08 2000 gtk+-1.2.6-7.i386.rpm
[01:33:11] <jepler> -rw-rw-r-- 6 ftp ftp 1049427 Mar 08 2000 gtk+-devel-1.2.6-7.i386.rpm
[01:33:14] <jepler> lftp archive.download.redhat.com:/pub/redhat/linux/6.2/en/os/i386/RedHat/RPMS> ls gtk+-*
[01:33:16] <jepler> -rw-rw-r-- 6 ftp ftp 847121 Mar 08 2000 gtk+-1.2.6-7.i386.rpm
[01:33:19] <jepler> -rw-rw-r-- 6 ftp ftp 1049427 Mar 08 2000 gtk+-devel-1.2.6-7.i386.rpm
[01:33:22] <jepler> oops
[01:33:25] <jepler> one paste would have sufficed
[01:33:29] <cradek> cool
[01:34:16] <cradek> why can I never find the site news for sf? I know it's somewhere
[01:34:34] <jepler> cradek: it's in one of those losenges on the front page, upper-right
[01:34:40] <jepler> cradek: it helps that they don't put any text on them
[01:34:53] <cradek> oh an expanding green lozenge, of course
[01:34:54] <jepler> site status -- green ball
[01:34:58] <jepler> duh!
[01:35:00] <jmkasunich> the new sf webpage layout sucks
[01:35:02] <cradek> idiots
[01:35:16] <cradek> CVS Service: "Partial Outage In-Progress"
[01:35:26] <jepler> I'd call it .. the opposite of progress
[01:35:34] <cradek> There is not a current estimate for the duration of this outage, but when we get one, it will be posted on the site status page
[01:35:59] <jmkasunich> something died
[01:36:23] <cradek> 21 hours ago (assuming their dates are UTC)
[01:36:54] <cradek> I guess sf is worth every penny we pay for it
[01:37:08] <jepler> cradek: you do have the option of throwing away good money to get some kind of subscription
[01:37:22] <jmkasunich> according to the compile farm logs, slot 3 (TNG) successfully compiled emc2head on mar 30, at 05:43 UTC
[01:37:23] <cradek> it's frustrating that our project is so crippled every time they're screwed up
[01:37:48] <jmkasunich> but the results obviously didn't go where we are looking for them
[01:38:26] <jmkasunich> figured it out
[01:39:09] <jmkasunich> steve stallings created a special account on the old server, "farm_results", and thats the ID the farm is attempting to use on the new server :-(
[01:39:19] <cradek> aha
[01:41:20] <steve_stallings> Yea, but on the new server access is all or nothing, not by directory like the old one. I did mail him a list of accounts. Not sure what was done with it.
[01:44:00] <jmkasunich> I'm surprized he isn't here (or at least SWP_away)
[01:44:04] <jmkasunich> is he traveling?
[01:47:05] <steve_stallings> Left shortly before you got on....8-)
[01:47:39] <jmkasunich> thats right, he's heading out to some show in calif, right?
[01:48:05] <steve_stallings> Embedded show?
[01:57:57] <jmkasunich> hmm, doing "ftp -p linuxcnc.org" and logging in as me gets me into my own directory, not the one that the compile farm page is being served from
[01:59:20] <cradek> jmkasunich: use cd?
[02:00:17] <jmkasunich> can't cd out of what appears to be my own home directory
[02:00:26] <cradek> darn
[02:00:38] <jmkasunich> well, I can cd deeper down, but that doesn't help
[02:00:54] <jmkasunich> there is a compile_farm dir there, but I created it for testing back when SWP gave me the account
[02:01:07] <jmkasunich> it has a 3/12 version of index.html in it, and nothing else
[02:01:31] <jmkasunich> the version on the website is the one that he copied over (along with the logs) on 3.26
[02:01:35] <jmkasunich> 3/26
[02:01:44] <steve_stallings> OK, so he can set users to directories. Compile_farm was on the list that I sent him.
[02:02:21] <steve_stallings> But in this case "him" was Alex
[02:07:06] <jmkasunich> ftp> pwd
[02:07:06] <jmkasunich> 257 "/" is current directory.
[02:07:20] <jmkasunich> does that mean I'm chrooted into my own little space?
[02:11:05] <cradek> probably
[02:34:56] <steve_stallings> steve_stallings is now known as steves_logging
[02:41:23] <jmkasunich2> hello
[02:42:14] <cradek> oh no, he's teaming up on us
[02:42:27] <jmkasunich2> jepler: what was the url for that make.gz again? and can I ftp it? (no browser on this box)
[02:42:41] <cradek> http://emergent.unpy.net/files/sandbox/make.gz
[02:42:44] <cradek> do you have wget?
[02:42:51] <cradek> elinks?
[02:43:00] <jmkasunich2> no
[02:43:04] <cradek> lftp?
[02:43:25] <cradek> dcc?
[02:43:35] <jmkasunich2> I have ftp
[02:43:52] <jmkasunich2> dcc has never worked for me,probably because of the nat router
[02:44:21] <jepler> lynx?
[02:44:34] <jmkasunich2> nope
[02:44:40] <cradek> gah
[02:44:56] <jepler> python -c 'import urllib'
[02:45:00] <cradek> I'll engrave it on a stone tablet and send it to you on my donkey
[02:45:14] <jmkasunich2> I have ftp
[02:45:19] <jepler> I don't have an ftp server
[02:45:22] <jmkasunich2> oh
[02:45:26] <cradek> nobody else uses ftp
[02:45:36] <cradek> do you have an ftp server on one of your other boxes?
[02:45:43] <jmkasunich2> I _thought_ there was lynx or something on this boc
[02:45:46] <jmkasunich2> box even
[02:45:54] <cradek> elinks, links, lynx
[02:45:57] <jmkasunich2> lemme see if I can start x
[02:46:02] <jepler> python -c 'import urllib; open("make.gz", "w").write(urllib.urlopen("http://emergent.unpy.net/files/sandbox/make.gz").read())'
[02:46:13] <cradek> haha
[02:46:31] <jepler> but who knows how old a python that'll actually work on
[02:46:39] <jmkasunich2> X works
[02:46:41] <jepler> or if there's even a python on bdi2 I suppose
[02:46:45] <jmkasunich2> slowly, but it works
[02:46:53] <jmkasunich2> so I can use konqueror
[02:46:53] <jepler> scp?
[02:47:01] <jepler> you have kde, but not wget?
[02:47:20] <cradek> wow
[02:48:02] <jmkasunich2> I have the default stuff that BDI-2.xx installs
[02:48:25] <jmkasunich2> don't seem to have konqueror (I guess thats on TNG) but there is netscape
[02:50:20] <jepler> I agree there doesn't seem to be any sane commandline file grabber or text-mode web browser
[02:51:15] <jepler> perhaps we could provide make.gz as a hex listing with check bytes
[02:51:24] <jepler> it's only 64k; I'm sure I typed that many listings in my life
[02:51:24] <jmkasunich2> ;-P
[02:52:01] <jepler> but these hypothetical bdi2 users will all be able to fire up netscape, so there's no problem
[02:52:41] <jmkasunich2> ok, file is unzipped and chmod'ed +x
[02:52:43] <jmkasunich2> now what
[02:54:15] <jepler> as root: mv /usr/bin/make /usr/bin/make.old; mv make /usr/bin/make
[02:55:24] <jmkasunich2> how nice - something you can install by copying a single file....
[02:55:34] <jmkasunich2> I thought that wasn't permitted these days
[02:55:44] <jepler> make 3.80 is a good 3 or 4 years old, iirc
[02:56:06] <jepler> whee .. I hope this works for you. it works in my bdi2 chroot, but that's not quite as good a test.
[02:57:00] <jepler> on the other hand, the chroot is nice and fast
[02:57:00] <jepler> You now need to run 'sudo make setuid' in order to run in place.
[02:57:00] <jepler> real0m41.733s
[02:57:38] <jmkasunich2> heh
[02:57:47] <jmkasunich2> got thru the configure,its depending now
[02:58:00] <cradek> then you're probably home free
[02:58:41] <jmkasunich2> on this box it will probably take 15 mins
[02:58:57] <jmkasunich2> done with depends, compiling now
[03:12:28] <jmkasunich2> woot! it compiles on BDI-2.xx
[03:13:04] <jepler> does that mean all the lights are green?
[03:14:14] <jmkasunich> yes
[03:14:34] <jmkasunich> or will be once I make it upload to the new server
[03:14:55] <jmkasunich> darn
[03:15:24] <jmkasunich> the new farm box doesn't like its new memory
[03:15:37] <jepler> how so?
[03:16:01] <jmkasunich> hangs up during bootup memory test
[03:16:06] <jepler> that's no good
[03:16:47] <jmkasunich> I don't think its the memory, its the slot
[03:16:51] <jmkasunich> it has 4 slots
[03:16:59] <jmkasunich> and it doesn't like ram in the odd numbered ones
[03:17:07] <jepler> how odd
[03:17:22] <jepler> do they have to be identical modules or something?
[03:17:28] <jmkasunich> unfortunately I have 3 x 256M dimms, so I need to use at least one odd slot
[03:17:48] <jmkasunich> allegedly they can be used in any mix
[03:17:57] <jmkasunich> all three are HP, same part number
[03:18:17] <jmkasunich> one is original to this box, the other 2 were removed from similar boxes
[03:18:33] <jmkasunich> it bitches about non-HP ram, must be some ID code in there
[03:18:45] <jmkasunich> doesn't bitch about these, just gets iffy
[03:18:58] <jmkasunich> hesitates and eventually stops during the ram test
[03:19:21] <jmkasunich> I have another mobo, maybe I should swap
[04:28:11] <jmkasunich> I'm afraid we might need to stick with the old compile farm...
[04:28:52] <cradek> still ram problems?
[04:28:59] <jmkasunich> worse
[04:30:07] <jmkasunich> after changing back to the original mobo, the DIMM lights came on as expected (2 on, 2 off since 2 dimms installed)
[04:30:21] <jmkasunich> it was going thru bootup, then all of a sudden the screen went blank
[04:30:35] <jmkasunich> now the dimm lights don't come on and I get no video at all
[04:30:43] <cradek> crap
[04:31:37] <jmkasunich> the disks spin up and do some seeking, but I think thats their on-board brains
[04:31:47] <cradek> yeah
[04:32:26] <cradek> heck, I better get to bed
[04:32:30] <jmkasunich> I wonder if I could have static zapped something....
[04:32:37] <cradek> talk to you this weekend if not sooner
[04:32:48] <jmkasunich> night
[04:33:03] <jmkasunich> I should go to be too
[04:33:06] <jmkasunich> * jmkasunich is depressed
[04:58:22] <jmk_cf> * jmk_cf is happy again!
[04:58:48] <jmk_cf> the original mobo must be frotzed, but the 2nd one works fine when set to 550MHz
[04:59:06] <jmk_cf> and all 768M of memory works too
[05:04:53] <jmk_cf> heh, according to dmesg they're actually running at 600MHz anyway!
[13:23:31] <alex_joni> morning ray
[13:24:33] <rayh> Hi alex.
[13:24:53] <rayh> Have you tried to access the cvs at sf. I've not been able to yesterday or today.
[13:26:04] <alex_joni> * alex_joni tries it now
[13:26:50] <alex_joni> nope, doesn't work :(
[13:27:29] <alex_joni> Project CVS Service: Partial Outage In-Progress Last updated: 2006-03-30 Pacific
[13:27:33] <rayh> Ah at least it isn't just me they don't like.
[13:27:46] <alex_joni> ( 2006-03-30 04:37:30 - Project CVS Service ) On 2006-03-30 the developer CVS server had a hardware issue that required us to take the service offline. We are actively working on this problem and hope to have it back up soon. There is not a current estimate for the duration of this outage, but when we get one, it will be posted on the site status page (this page).
[13:27:51] <alex_joni> https://sourceforge.net/docman/display_doc.php?group_id=1&docid=2352
[13:28:07] <rayh> Any way to know what time it might come back on line?
[13:28:39] <rayh> * rayh looking
[13:29:59] <alex_joni> nope, they don't know either :(
[13:32:00] <rayh> Darn. Perhaps it's time to take dreamhost up on cvs.
[13:32:57] <alex_joni> I'd rather wonder about a backup service
[13:33:03] <alex_joni> and a way to synchronize the two
[13:35:30] <rayh> Sure if such a thing were possible.
[13:44:47] <jepler> rayh: given that his own damned packages won't rebuild on his own damn distribution, I don't think there's a lot of point in asking paul for help building debian packages
[13:45:58] <rayh> I love it. Thanks, jepler.
[13:46:39] <jepler> but, like you, I keep trying to get concrete help from him, which just enables him to spew more bullshit about something being wrong with emc2
[13:47:56] <jepler> it's more frustrating than productive
[13:48:20] <jepler> oops -- off to work! talk later.
[13:48:25] <rayh> Sure. And I appreciate your testing and keeping at him.
[13:48:31] <rayh> See you.
[13:52:10] <anonimasu> hm
[13:52:12] <anonimasu> that's a mess
[13:56:26] <anonimasu> I cant see why paul dislikes emc2
[14:04:40] <cradek> I think he doesn't disklike emc2 as much as he dislikes me
[14:04:43] <cradek> dislike even
[14:05:31] <cradek> but that's ok, it's his problem, not mine
[14:06:18] <cradek> brb
[14:08:00] <rayh> cradek, I think that the disaffection with emc2 began long before your emc time.
[14:08:23] <rayh> You are just a high profile target.
[14:23:12] <skunkworks> anonimasu: we must have spent an hour last night running through the lathe videos :)
[14:23:34] <skunkworks> very cool.
[14:30:54] <rayh> Great. Wish I could watch here.
[14:35:52] <skunkworks> still working on your wifi link ray?
[14:36:28] <skunkworks> I will have one set up for a relitive maybe this weekend.
[14:39:26] <rayh> Really. I saw some 6k links but they were a bit much for me.
[14:40:01] <skunkworks> this is using a linksys access point with a dlink wireless router
[14:40:11] <rayh> But like me, most of my friends nearby also live in gulleys.
[14:40:37] <rayh> I've got a WRT54G
[14:41:35] <rayh> and of course the sacred cnc-workshop link WET11
[14:41:40] <anonimasu> skunkworks: I got so sick.. wanting a lathe ;)
[14:41:43] <anonimasu> a fast cnc lathe :D
[14:41:47] <anonimasu> last night after watching
[14:42:22] <anonimasu> skunkworks: pretty insane, I think that shows why you need synch points when doing multispindle work :)
[14:43:36] <skunkworks> I thought when I first started looking into this that I could just buy 2 routers and use them as a wireless link. after a bunch of research I found out that they need the option to do that. So I took one of the routers back and bought a access point. (linksys says it will only work with other linksys access points or routers for the link but seems to work just fine with my dlink)
[14:44:06] <skunkworks> anonimousu: very cool - fast. muti spindle and everything.
[14:45:58] <rayh> I've got a line on 3 slant bed Hardinge chuckers that had fanuc controls if you are interested.
[14:46:53] <skunkworks> ray: bought a bunch of pigtails and connectors here. they also have some pretty cheap antannas. (I made a cantinna as the distance isn't far) http://sharperconcepts.zoovy.com/
[14:49:54] <anonimasu> too bad I'm in sweden :)
[14:50:56] <skunkworks> we have a big 2 turet old cincinati with an acromatic controller on it. (another emc retrofit in the making)
[14:52:26] <rayh> Thanks for the link skunkworks.
[14:53:30] <rayh> Oh. That sounds fantastic anonimasu. Are these the sort of spindles that hand off work from one to the other for second end operations?
[14:53:56] <rayh> oops. wrong guy.
[14:54:03] <skunkworks> right - they were also cutting while bar feeding thought the spindle.
[14:54:37] <rayh> Sure. That sounds like it will cause a demand for syncro between spindles.
[14:54:41] <anonimasu> rayh: yeah
[14:54:53] <rayh> But we'll have to create multiple spindles first.
[14:55:02] <anonimasu> rayh: it includes having multiple controllers..
[14:55:11] <anonimasu> it were explained at some forum at cnczone.. about programming them..
[14:55:21] <rayh> I service a couple of Mazak 10 lathes that do this and live tooling.
[14:55:29] <anonimasu> they had 3 programs running at the same time
[14:56:19] <anonimasu> http://cgi.ebay.com/New-3hp-Servo-Motor-With-Brake-Dual-Shaft_W0QQitemZ2512861692QQcategoryZ25290QQcmdZViewItem
[14:56:47] <rayh> I've worked on a couple of old Mazak T4 controls that had two independent tools.
[14:56:53] <anonimasu> I am going to quote my plc dealer for a vfd with positining support..
[14:57:07] <anonimasu> and a 4 hp motor.
[14:57:13] <anonimasu> offtopic though :/
[14:57:34] <rayh> I don't think so. Just future topic.
[14:57:43] <anonimasu> yeah probably
[14:57:54] <anonimasu> I am amazed at the turn/mill/rotate stuff
[14:58:28] <rayh> Watching one mill a square or triangle is something.
[14:58:57] <skunkworks> nothing diffent than 3 axis plus rotory.
[14:59:12] <anonimasu> yeah true
[14:59:16] <skunkworks> in my mind anyways ;) (that doesn't say much)
[14:59:17] <rayh> Except that it is only 2 axes and a rotary.
[14:59:38] <rayh> and one of them is not alligned.
[14:59:53] <rayh> requires rho theta calculations.
[15:00:14] <rayh> where x moves in and out as the rotary moves.
[15:00:42] <rayh> to make a flat cut.
[15:00:54] <anonimasu> atleast at thoose lathes they had x/y/z on the mill spindle..
[15:01:07] <anonimasu> I saw some other video with xyzr
[15:01:08] <anonimasu> :)
[15:01:10] <skunkworks> I was just going to say that.
[15:01:17] <rayh> The ones I've delt with did not have any y
[15:01:21] <anonimasu> ah ok
[15:01:50] <rayh> Traditional lathe with the live tooling on the turret.
[15:02:05] <skunkworks> ray - so you would use a ball mill to make a flat cut on a rund stock?
[15:02:07] <skunkworks> round
[15:02:09] <rayh> What you are describing sounds more like a 2 in 1
[15:02:33] <anonimasu> gang tooling lathe.. or something..
[15:02:39] <rayh> You can use an ordinary end mill
[15:03:05] <rayh> It's rotation could either be around x or z
[15:04:06] <rayh> Draw a box on a cdrom and rotate slowly and imagine a tool that only moves toward the center or away and you'll get the idea.
[15:04:41] <rayh> Starts a flat cut away from the center and moves in as the cd rotates slowly.
[15:04:50] <anonimasu> yeah I know what you mean
[15:04:51] <anonimasu> :)
[15:05:09] <rayh> odd sort of kinematics.
[15:05:22] <anonimasu> but with xyz you can machine threads internally..
[15:05:23] <anonimasu> :D
[15:06:22] <rayh> off from the center of spindle rotation?
[15:06:51] <anonimasu> like the T huydralic connector they machine
[15:07:08] <anonimasu> yeah with that lathe you could..
[15:07:24] <rayh> Interesting.
[15:13:34] <rayh> Hey skunkworks. I suppose I'd need 2 YSC-HG2424G-5PK and pretty much line of site.
[15:14:32] <rayh> sight.
[15:28:30] <rayh> Morning Dan.
[15:28:42] <dan_falck> hi ray
[15:28:59] <dan_falck> I see threading is moving along
[15:29:35] <rayh> Yes it is. Say thanks for the info on DNS. We got it changed over.
[15:29:58] <dan_falck> Yes. The web page looks great. I like it.
[15:30:30] <rayh> Alex added search the other day and blew me away.
[15:30:48] <dan_falck> I'll have to try it out.
[15:31:15] <rayh> I searched for ini and was really impressed with what it found.
[15:31:54] <dan_falck> In the post about threading last night, how does one 'loop through G33'? to do multi pass threading
[15:32:41] <cradek> dan_falck: I used an O loop. I can commit that file if you like (but can't right now)
[15:32:52] <dan_falck> ok. thanks.
[15:33:02] <jepler> cradek: didn't you post the file to the list or something?
[15:33:07] <cradek> did I?
[15:33:31] <cradek> (that machine is off, and cvs isn't working anyway)
[15:33:33] <jepler> I could have sworn I remembered seeing the O-codes
[15:34:02] <cradek> maybe I committed a version with a loop
[15:34:16] <jepler> http://cvs.sourceforge.net/viewcvs.py/emc/emc2/nc_files/threading.ngc?rev=HEAD&view=auto
[15:34:29] <cradek> aha
[15:34:32] <anonimasu> hm, is that command?
[15:34:48] <dan_falck> thanks I'll take a look
[15:34:51] <cradek> that's not exactly the program I used to cut the 3/8-16 last night, but it's close enough
[15:35:30] <rayh> Right the commit message showed it but you cant get the committed file.
[15:35:44] <cradek> actually I think that program cuts a lefthanded thread...
[15:48:40] <skunkworks> ray- pretty much anything is going to be line of sight for long distance at 2.4ghz
[15:51:07] <rayh> Okay. I'll have to take a few pictures off the roof and see what's out there.
[15:53:31] <cradek> rayh: don't rule out a pole tower - they're simple to put up if you have the space
[15:54:05] <cradek> hmm, aiming might be a challenge because you have to do it before you pull up the tower
[15:54:50] <anonimasu> cradek: you can climb thoose towers too :)
[15:54:58] <cradek> ha, maybe you can
[15:55:09] <anonimasu> but it's kind of scary :)
[15:55:13] <cradek> http://www.absak.com/catalog/product_info.php/cPath/32_54_169/products_id/761
[15:55:23] <cradek> I helped put up something like this
[15:55:28] <anonimasu> bring hooks and stuff
[15:55:33] <cradek> very light and simple - obviously not meant for climbing
[15:55:38] <anonimasu> ah that kind
[15:55:45] <anonimasu> thoose arent nicely climbable ;)
[15:55:49] <cradek> no.
[15:56:27] <cradek> but cheap and simple, and we pulled it up with just a lawn tractor
[15:57:21] <anonimasu> :)
[16:46:33] <rayh> cradek, sorry away for a bit. Yes a tower might work well. The top of my roof is about 27 feet up and I am on a knoll but I may have to go a bit higher to see any useful place.
[16:49:35] <cradek> hmm, I just drew a picture, and it seems like doubling the height doesn't double your line of sight - it does much less than that
[16:49:58] <cradek> if I knew some math, maybe I could figure it out.
[16:51:52] <skunkworks> cradek - have you set up any long distance wireless? I have heard you want the length of coax between the router/accesspoint/nic to be as short as posible. Would you mount the router on the antanna?
[16:52:52] <cradek> no, I haven't, but running CAT5 up the tower to the router/antenna sounds like a good idea
[16:54:07] <skunkworks> that would be my thought. (I have powered devices also through the network cable)
[16:54:23] <rayh> Right. Depending upon distance to the barriers
[16:55:07] <rayh> Power up the tower isn't a problem here.
[16:55:34] <rayh> I can put a box there for the electronics.
[16:55:54] <cradek> rayh: do you have cell phone service out there?
[16:56:06] <cradek> digital service I mean
[16:56:17] <cradek> sometimes you can get decent internet speed over that
[16:56:37] <cradek> (it's probably expensive though)
[17:01:28] <rayh> no.
[17:02:03] <rayh> I've got one bar, with glasses and bottles. When folk come with cell phones, they get less than that.
[17:05:02] <cradek> if you don't want to surf much, you can still do a lot of useful things with a dialup, I guess. you get a lot done working with us.
[17:05:49] <rayh> Boy the uploads and downloads are deadly slow. As the pool of data gets larger the lag gets more noticable.
[17:07:07] <rayh> Thanks for the links and help guys. I really don't think satellite is the answer though.
[17:07:32] <rayh> It's good enough for surfing but long downloads and uploads rarely succeed.
[17:07:57] <cradek> rayh: do you get >33k connections?
[17:08:16] <rayh> nope 1.2k on a good run.
[17:08:25] <cradek> iirc 33.6 is the fastest you can get with normal phone lines
[17:08:50] <cradek> oh I mean kbit
[17:08:58] <cradek> (the modem speed)
[17:09:05] <rayh> Right but I've got 16 miles of copper and a whole bunch of junctions.
[17:09:26] <cradek> man, you're lucky it works at all if you're 16 miles from anywhere
[17:09:56] <rayh> There is a u turn in the middle of that. Some sort of disagreement between phone systems.
[17:10:58] <cradek> in the prime of dialup I used to get connections at 46,666 but never the promised 56k
[17:14:42] <jepler> dialup .. how happy I am to be free of it
[17:32:12] <alex_joni> hi jeff
[17:32:15] <alex_joni> still around?
[17:35:48] <rayh> Hi alex. Ready for the weekend?
[17:37:25] <jepler> alex_joni: oh kinda
[17:37:27] <jepler> alex_joni: what's on your mind?
[17:39:15] <alex_joni> was wondering about http://wiki.debian.org/KernelModulesPackaging
[17:39:33] <alex_joni> but I guess you know more about that than me..
[17:40:05] <jepler> I don't feel like I know debian that well
[17:40:43] <alex_joni> I found various snippets of shell code that seem to do the promised thing
[17:40:48] <alex_joni> but I couldn't say really
[17:41:14] <alex_joni> something like this:
[17:41:14] <alex_joni> . # sh -c 'KVERS=`uname -r`; KSRC=/usr/src/kernel-headers-$KVERS; \
[17:41:14] <alex_joni> . KDREV=`dpkg -s kernel-headers-$KVERS \
[17:41:14] <alex_joni> . | awk '"'/^Version:/ {print \$2}'"'`;\
[17:41:14] <alex_joni> . export KSRC KVERS KDREV; \
[17:41:17] <alex_joni> . make -f debian/rules KSRC=$KSRC KVERS=$KVERS \
[17:41:20] <alex_joni> . KDREV=$KDREV kdist_image'
[17:41:59] <alex_joni> brb
[17:42:16] <jepler> the thing is, I know of one guy right now who cares about building debian packages against a different kernel, and he won't contribute a patch.
[17:44:49] <cradek> no, but he'll snipe at us about it
[17:45:55] <jepler> so I say, wait for the contributed patch
[17:46:01] <jepler> instead of going to the trouble just to serve one ungrateful bastard
[17:48:05] <jepler> who'll find something else to complain about anyway
[17:48:15] <jepler> in an effort to keep his linux distribution relevant
[17:48:33] <jepler> s/an effort/a pathetic effort/
[17:54:13] <skunkworks> I had directway for about 2 years. worked well (2 way satallite.) up to 2000kbs download and around 30kbs upload
[17:56:13] <alex_joni> jepler: ok ;)
[17:59:09] <skunkworks> heavy rain and snow took it out. it also has fair access policy (fap) in short it is about 169Mb in about 4 hour period (it is more complicated than that)
[18:07:44] <skunkworks> you would not be able do download iso without a download manager of some kind. (break it into 169mb chunks)
[18:10:46] <skunkworks> but all and all - tons better than dialup but no replacement for dsl/cable or wireless.
[18:12:08] <skunkworks> (your comunicating to satellites that are geosyncronous orbit. Ping times are around 1 second +/- depending on the time of day (gets worse the more people are on)
[18:13:22] <skunkworks> <forest gump>And thats all I have to say about that</forest gump>
[18:50:59] <anonimasu> hm
[19:28:36] <skunkworks> logger_devel: bookmark
[19:28:36] <skunkworks> See http://solaris.cs.utt.ro/irc/irc.freenode.net:6667/emcdevel/2006-03-31#T19-28-36
[19:29:53] <jepler> skunkworks: sounds like it would not be bad for e-mail, but suck for casual web browsing
[19:30:47] <cradek> with multi-second latency, you'd want streaming uucp
[19:32:57] <jepler> modern http does have some streaming capability
[19:33:01] <jepler> they call it pipelining
[19:58:12] <skunkworks> in and out
[19:58:40] <skunkworks> it wasn't too bad - it did some funky stuff as far as an "accellerator"
[19:58:40] <skunkworks> <skunkworks> it had a "proxy server" setup that used a differnt protocol for the comuniction between the enduser and the base station.
[19:59:09] <alex_joni> * alex_joni is back
[20:08:01] <rayh> Hi alex
[20:08:17] <alex_joni> hi ray
[21:07:08] <jepler> Specifically, regarding CVS,
[21:07:09] <jepler> we're planning to upgrade our service based upon improvements that we've
[21:07:09] <jepler> successfully deployed as part of our Subversion offering.
[21:07:13] <jepler> -- sourceforge newsletter I just got
[21:07:42] <alex_joni> that shouldn't be hard (to improve it)
[21:09:55] <jepler> They've updated the status page now. "We currently expect this outage to last 48 hours, at minimum."
[21:10:10] <jepler> does that mean "we're going to take the weekend off, and then see if we can't fix it on monday"?
[21:10:25] <alex_joni> it already lasts more than 28 hours
[21:10:28] <alex_joni> 48
[21:11:08] <alex_joni> that's so ambiguous ("We currently expect this
[21:11:08] <alex_joni> outage to last 48 hours, at minimum."
[21:11:26] <alex_joni> from when it started? from now on?
[21:12:25] <jepler> from whenever you read the words "48 hours"?
[21:14:04] <cradek> whatever it says, it probably means we won't be able to do anything all weekend
[21:16:09] <alex_joni> buggerit
[21:17:31] <jepler> alex_joni: leaving? see ya.
[21:18:16] <alex_joni> jepler: leaving? how so?
[21:25:08] <jepler> oh, I figured "buggerit" was a prelude to you calling it an (early) night
[21:25:24] <jepler> and I saw that you restarted logger
[21:26:11] <alex_joni> jepler: that's a cron thing
[21:26:20] <alex_joni> because it dies and has no way of reviving
[21:26:29] <alex_joni> so I set it up like that in order to miss max. a day
[21:26:45] <alex_joni> I mean when I'm not paying attention if the logger is around
[21:28:49] <jepler> ah
[21:32:41] <alex_joni> jepler: I should worry about replacing it.. but I don't think it's that important
[21:32:49] <alex_joni> not motivated enough to do that
[21:38:05] <rayh> I see cvs at sf is still down.
[21:38:20] <jepler> yep
[21:38:51] <alex_joni> rayh: it'll probably stay like that till monday
[21:38:55] <alex_joni> at least
[21:39:05] <alex_joni> it doesn't look very bright right now
[21:41:06] <rayh> Yuck. That will make a mess of my desire to release sunday night.
[21:41:13] <cradek> I was able to download the cvs tarball. We could move cvs to somewhere else this weekend.
[21:41:28] <cradek> rayh: I still haven't been able to make the testing package, so we're in no shape for a release.
[21:42:01] <rayh> Right. Do you think you have a current checkout there?
[21:42:20] <cradek> no
[21:42:29] <cradek> my checkouts are all from before alex's branch
[21:43:34] <cradek> symbolic names:
[21:43:34] <cradek> v2_0_branch:
[21:43:40] <cradek> but the nightly tarball is up to date
[21:44:40] <cradek> does anyone remember the last change committed?
[21:44:49] <alex_joni> * alex_joni looks
[21:45:00] <alex_joni> Modified Files:
[21:45:00] <alex_joni> threading.ngc
[21:45:00] <alex_joni> Log Message:
[21:45:00] <alex_joni> multiple passes at 29 degrees
[21:45:22] <alex_joni> your commit on threading.ngc (1.4)
[21:45:48] <alex_joni> cradek: but that's on HEAD
[21:46:28] <cradek> revision 1.4
[21:46:28] <cradek> date: 2006/03/30 04:36:20; author: cradek; state: Exp; lines: +11 -5
[21:46:28] <cradek> multiple passes at 29 degrees
[21:46:35] <cradek> this is up to date then
[21:47:17] <alex_joni> do you see the v2_0_branch ?
[21:47:23] <cradek> yes
[21:47:28] <alex_joni> how about the TESTING tag=
[21:47:29] <alex_joni> ?
[21:47:47] <cradek> in threading.ngc:
[21:47:48] <cradek> symbolic names:
[21:47:48] <cradek> TESTING: 1.3
[21:47:48] <cradek> v2_0_branch:
[21:48:00] <cradek> so TESTING points to the version before my change
[21:48:26] <alex_joni> what's that with ?
[21:48:42] <cradek> don't worry, it's just cvs internals
[21:48:43] <alex_joni> .2 should come from people editing on the branch.. right?
[21:48:59] <cradek> that version doesn't exist yet, cvs would create it later if needed
[21:49:32] <alex_joni> ahh.. ok
[21:50:54] <rayh> I suppose that sf will blame all the problems on us and our last branch or cradek's threading code. I know thats a bug.
[21:51:19] <cradek> yep, it's usually my fault.
[21:51:23] <alex_joni> cradek: I'm bitching in ##sourceforge right now
[21:51:33] <alex_joni> but no officials around :D
[21:51:54] <cradek> whatever makes you happy
[21:52:37] <alex_joni> https://sourceforge.net/docman/display_doc.php?group_id=1&docid=2352, if you read that page carefully you'll see that either on of the admins got kicked, or he quit
[21:52:40] <alex_joni> but they are hiring
[21:54:19] <rayh> I'll see if I can fix Jason's list problems.
[21:54:56] <alex_joni> * alex_joni already looks too :D
[21:56:40] <rayh> * rayh takes a nap while sf sends a batch of adds.
[21:57:11] <alex_joni> it all looks legit to me
[21:57:18] <alex_joni> both on emc-users & emc-developers
[21:58:25] <rayh> I saw some sf stuff the other day that reset several accounts because they were bouncing.