#emc | Logs for 2005-04-07

[09:57:18] <alex_joni> yo paul_c
[09:57:32] <anonimasu> alex_joni: hey what's up?
[09:57:49] <paul_c> Morning Alex
[09:57:57] <alex_joni> I'm freaking pissed
[09:58:07] <alex_joni> on a ethernet connection
[09:58:10] <alex_joni> darn cable
[09:58:17] <alex_joni> got 10-40% loss on it
[09:58:19] <anonimasu> :/
[09:58:25] <anonimasu> bad hub/switch?
[09:58:35] <alex_joni> nah.. bad cable
[09:58:41] <alex_joni> I changed the switch
[09:58:44] <anonimasu> ah ok
[09:58:50] <alex_joni> changed the other parts receiver
[09:59:33] <alex_joni> I'll change it with an STP
[10:09:54] <anonimasu> ok I am working on my code still
[10:10:03] <anonimasu> :)
[10:10:08] <alex_joni> nice
[10:12:20] <anonimasu> it's going way too slow :/
[10:37:12] <alex_joni> an0n: you did some php.. right?
[10:59:35] <anonimasu> yes
[10:59:55] <alex_joni> never mind.. found it
[11:00:04] <anonimasu> ok
[11:00:05] <anonimasu> :)
[11:00:08] <alex_joni> was trying to access a variable from php ($var)
[11:00:18] <alex_joni> but it seems on newer php's that doesn't work anymore
[11:00:25] <alex_joni> you need $_GET['var']
[11:00:54] <anonimasu> ah yeah
[11:01:05] <anonimasu> they deprecated it at version 3.somthing
[11:01:06] <alex_joni> is there a way to turn this off?
[11:01:16] <anonimasu> yeah
[11:01:21] <anonimasu> you can enable globals in the config
[11:01:21] <robin_z> umm!
[11:01:25] <anonimasu> but that's a bad way
[11:01:27] <alex_joni> meep?
[11:01:27] <robin_z> ffs dont do that
[11:01:46] <alex_joni> how come?
[11:01:46] <anonimasu> robin_z: he asked if it was possible not if it was any good ;)
[11:01:50] <robin_z> its a HUGE security hole
[11:01:54] <robin_z> true
[11:01:54] <alex_joni> right
[11:01:54] <anonimasu> yep
[11:01:59] <anonimasu> very very insecure
[11:02:04] <robin_z> whatever, just dont do it.
[11:02:05] <alex_joni> ok. I'll update the pages :)
[11:02:08] <alex_joni> lol
[11:02:26] <robin_z> alex_joni: do you know the names of your vars?
[11:02:35] <anonimasu> $var=$_GET[thing];
[11:02:36] <alex_joni> yes
[11:02:40] <anonimasu> ''
[11:02:57] <anonimasu> just do it before you use them and it should work out fine
[11:03:12] <alex_joni> right
[11:03:22] <alex_joni> $thing=$_GET['thing']
[11:03:33] <anonimasu> yeah
[11:03:38] <robin_z> perl -pi -e"s/\$var/\$_GET['var']/g" *.php
[11:04:41] <alex_joni> yeah that too.. but there aren't that many
[11:04:45] <robin_z> or do what anonimasu said and explicitly import them
[11:04:59] <alex_joni> so adding that stuff ontop would be ok
[11:05:04] <robin_z> yeah
[11:05:09] <alex_joni> errr. at the beginning of the page
[11:05:24] <robin_z> you can easily imagine the security risk im sure.
[11:05:30] <alex_joni> yup
[11:05:35] <alex_joni> I was wondering
[11:05:41] <alex_joni> how big is 21U ?
[11:05:43] <robin_z> 3.5"
[11:05:43] <anonimasu> um
[11:05:48] <robin_z> oh 21
[11:05:54] <anonimasu> 21*19"
[11:05:55] <robin_z> 21 x 1.75"
[11:05:56] <anonimasu> err
[11:05:59] <anonimasu> 1.75
[11:06:00] <anonimasu> yeah
[11:06:02] <alex_joni> right
[11:06:26] <alex_joni> about a meter
[11:06:31] <alex_joni> 93 cm
[11:06:46] <robin_z> you'd be amzed how many PHP scripts cough up "interesting" data with blah.php?debug=1 ;)
[11:06:47] <alex_joni> 21 * 1,75 * 2,54 .. right?
[11:06:56] <alex_joni> lol
[11:06:57] <robin_z> yep
[11:06:58] <anonimasu> robin_z: mine does ;)
[11:07:59] <alex_joni> I think 21U should be enough for 2 PC's, one KVM, one Monitor & Keyb, one switch, maybe some telephone switch (1U)
[11:08:18] <anonimasu> 2yeah
[11:08:24] <robin_z> oh, half a rack :)
[11:08:33] <alex_joni> yeah
[11:08:39] <robin_z> in a colo?
[11:08:41] <alex_joni> a full size rack would be overkill
[11:08:44] <alex_joni> colo?
[11:08:53] <robin_z> colocation facility .. data centre
[11:09:02] <alex_joni> nah.. next to my desk
[11:09:06] <robin_z> right
[11:09:24] <alex_joni> probably won't fit under the desk :D
[11:09:36] <anonimasu> 21u is large
[11:09:44] <robin_z> yeah
[11:09:53] <alex_joni> I need to have some spare space
[11:10:06] <alex_joni> and I don't have 19'' rackmountable cases for the pc's
[11:10:22] <robin_z> I just stick the machines "somewhere else" and use a laptop on the desk
[11:10:28] <robin_z> ssh is my friend
[11:10:33] <alex_joni> mine too :D
[11:10:47] <alex_joni> my main server doesn't even have a VGA card :D
[11:10:54] <robin_z> heh
[11:11:12] <alex_joni> actually now it has (I had to stick a few more HDD inside)
[11:11:18] <robin_z> its sort of useful... for the odd log in an find out whats broken sort of thing
[11:11:18] <alex_joni> needed a card to get to BIOS ;)
[11:11:49] <robin_z> this is what I find weird about 'doze
[11:12:03] <robin_z> linux boxen have ususally no X installed on servers
[11:12:12] <alex_joni> heh.. no /var/log/messages ?
[11:12:17] <robin_z> Doze still has all that grapicla shit
[11:12:39] <alex_joni> yeah
[11:12:43] <robin_z> no /var/log/messages??
[11:12:52] <alex_joni> on doze ;)
[11:12:52] <robin_z> heh
[11:12:52] <alex_joni> I miss that
[11:13:02] <alex_joni> and the crappy logs windoze has just annoy me
[11:13:24] <robin_z> some things need the console still
[11:13:35] <robin_z> like kernel panics :)
[11:13:49] <robin_z> and cards dropping into promiscuous mode always flag on the console
[11:14:50] <alex_joni> yeah
[11:15:14] <robin_z> if youve been hacked, thats usually the first warning
[11:15:22] <anonimasu> yep
[11:15:30] <robin_z> that or tripwire getting excited
[11:15:47] <alex_joni> tripwire?
[11:16:03] <anonimasu> signs binaries and stuff..
[11:16:08] <robin_z> sorta
[11:16:20] <robin_z> more to the point it watches certain directories
[11:16:38] <robin_z> if stuff happens you didnt expect it "does stuff"
[11:16:59] <robin_z> ranging from mailing you, to dropping the shutters on the firewall
[11:18:02] <alex_joni> how about unplugging eth?
[11:18:03] <alex_joni> :D
[11:18:50] <robin_z> thats effectively what i do, put a DROP rule in each of the iptables
[11:19:06] <alex_joni> right
[11:19:15] <alex_joni> iptables -P DROP
[11:19:18] <alex_joni> iptables -F
[11:20:17] <anonimasu> I like "drop in on rl0 from any to probability 95%"
[11:20:33] <alex_joni> heh
[11:21:12] <anonimasu> probability is a funnny function
[11:21:14] <anonimasu> ^_^
[11:24:36] <robin_z> whatever, tripwire works well
[11:24:56] <robin_z> it saved a friend of mine just recently ...
[11:25:00] <anonimasu> it's mean as hell on lans
[11:25:19] <robin_z> client didn't pay ...
[11:25:19] <alex_joni> is it a standard thing?
[11:25:29] <robin_z> box was moved to new location ...
[11:25:40] <robin_z> tripwire kept mailing him ... with its new IP
[11:26:03] <alex_joni> nice
[11:26:04] <alex_joni> ;)
[11:26:10] <robin_z> apparently, for some unknown reason, the disk died ... weird huh?
[11:26:24] <anonimasu> alex_joni: in openbsd's pf
[11:26:28] <alex_joni> is it available on debian-sarge?
[11:26:37] <robin_z> alex_joni: apt-get install tripwire
[11:26:37] <anonimasu> no
[11:26:41] <anonimasu> :)
[11:26:44] <anonimasu> although tripwire is
[11:26:47] <robin_z> tripwire --init
[11:28:42] <robin_z> of course, putting your tripwire binaries and the keys on read-only media is not the workds worst plan ...
[11:28:55] <alex_joni> heh
[11:28:56] <alex_joni> yeah
[11:29:08] <alex_joni> so tripwire --init should be enough?
[11:29:58] <robin_z> well, review the policy file in /etc/tripwire/twpol.txt
[11:31:51] <robin_z> and run tripwire --check from time to time, from cron
[11:32:57] <robin_z> the truly paranoid keep their tripwire database on a seperate machine
[11:32:58] <alex_joni> okie
[11:33:46] <alex_joni> robin: where do I specify actions to perform?
[11:33:49] <alex_joni> in twpol.txt?
[11:33:56] <robin_z> yep
[11:34:03] <robin_z> that says what will be checked
[11:36:09] <alex_joni> I've seen that
[11:36:20] <alex_joni> where do I specify what email it should send when it fails
[11:36:37] <fenn> good morning
[11:36:53] <robin_z> * robin_z tries to remember
[11:36:58] <alex_joni> fenn: afternoon
[11:37:28] <fenn> i think it's almost my bedtime but i can't seem to go to bed
[11:37:41] <fenn> what's the current topic of discussion?
[11:38:08] <anonimasu> tripwire
[11:38:20] <fenn> blegh
[11:38:50] <alex_joni> heh
[11:38:56] <robin_z> or 50 pole brushless AC servos
[11:38:58] <alex_joni> fenn: any topic you'd wanna discuss?
[11:40:00] <alex_joni> try unmounting the resolver from that AC drive, then put it back on :D
[11:40:00] <fenn> hmmm
[11:40:17] <fenn> well i'm going to go get a van norman 12 h/v milling machine
[11:40:31] <fenn> never done anything like that before, worried it might fall and kill me
[11:40:41] <alex_joni> heh
[11:40:44] <robin_z> alex_joni: auto intitialisation is the plan :)
[11:40:45] <fenn> or break my friend's trailer
[11:41:06] <robin_z> fenn: weight?
[11:41:08] <alex_joni> robin: maybe a script is needed (check from cron, if tripwire --check fails, do some action like: mail ...)
[11:41:09] <fenn> 1800 lbs
[11:42:02] <robin_z> alex_joni: hmm, nah .. tripwire --test --email sends a mail automagically .. jobs run from cron mail their owner anyway ...
[11:42:58] <robin_z> fenn: 1800lbs since you say lbs, i assume you are in the US ...
[11:43:06] <robin_z> check your towing laws :)
[11:43:06] <alex_joni> robin: thx
[11:43:16] <fenn> robin: yep indiana
[11:43:52] <anonimasu> bleh I need to go back to work
[11:44:00] <anonimasu> need to test this stuff on a plc..
[11:44:01] <robin_z> in the UK, you would struggle to tow it legally
[11:44:07] <anonimasu> the code's growing like cancer.
[11:44:12] <robin_z> ick
[11:44:36] <anonimasu> I have 10 objects doing diff stuff, and I need to get some finished for other ones..
[11:44:38] <anonimasu> :)
[11:45:50] <robin_z> fenn: and you have some sort of hoist or crane at both ends of the journey?
[11:45:59] <fenn> eh hehehehe
[11:46:15] <fenn> i have a winch with which i hope to pull it up a ramp onto a trailer with
[11:46:20] <robin_z> ahh.
[11:46:29] <fenn> egyptian roller style
[11:46:34] <robin_z> yessss ...
[11:46:37] <robin_z> well,
[11:46:54] <robin_z> 1800 lbs is a lot ...
[11:47:02] <anonimasu> keep your friends with you
[11:47:06] <fenn> yeah it is a big winch, but that's not what i'm worried about
[11:47:14] <robin_z> the trailer is the worry
[11:47:28] <fenn> we just pulled a 35 foot sailboat onto the trailer a couple weeks ago with the same winch
[11:47:35] <robin_z> yeah?
[11:47:45] <robin_z> well, sounds like the trailer will take the load
[11:48:02] <robin_z> but ...
[11:48:07] <fenn> i gotta get it stood upright too, since it's lying on its side in the mud
[11:48:15] <anonimasu> :/
[11:48:24] <robin_z> if you could just drop it onto the ara above the axle, Id sya yes
[11:48:35] <fenn> handles busted off and belt covers smashed :( but for $200 its a steal
[11:48:40] <robin_z> but pulling it aver than long ramp ...
[11:48:46] <robin_z> over
[11:48:51] <robin_z> that long ramp
[11:49:05] <fenn> i might be able to get the guys at the scrap yard to pick it up with one of their monster forklifts
[11:49:13] <robin_z> yeah
[11:49:18] <robin_z> much better plan
[11:49:37] <fenn> you think rolling it down a ramp is a good idea?
[11:49:43] <fenn> with the winch
[11:49:44] <robin_z> mmmm ...
[11:49:46] <robin_z> not really
[11:49:51] <robin_z> depends on the trailer
[11:50:00] <alex_joni> giant fork lift is a great ideea
[11:50:03] <robin_z> the bit over the axle is a good place to place it
[11:50:07] <fenn> what if i put blocks under the end so it didnt tilt down
[11:50:20] <robin_z> but .. the long extended ramp ... I dunno
[11:50:32] <robin_z> might just buckle and twist :(
[11:51:11] <robin_z> maybe if you can tilt it and put blocks under at a couple of places
[11:51:24] <robin_z> if you just block the end ..
[11:51:39] <robin_z> well .. its like a bridge isnt it .. between the wheels and the end
[11:51:46] <robin_z> thats what, 10 foot?
[11:51:55] <fenn> yeah i was intending to put blocks under the end of the trailer and under the middle of the ramp
[11:52:03] <fenn> i think its about 10 feet
[11:52:20] <fenn> the trailer, not the ramp (ramp's 6 ft horizontal)
[11:52:28] <robin_z> yeah, I think
[11:53:13] <robin_z> basically, with that weight
[11:53:13] <robin_z> I wouldnt want it to go over any section of more than the base of the machine without a block under
[11:53:15] <robin_z> so block it every 5 foot maybe
[11:53:31] <fenn> ok thanks for the advice
[11:53:36] <robin_z> depends how big and butch your trialer is :)
[11:53:57] <fenn> hrm well i don't want to break it cause it's a not-really-friend of a friend's trailer
[11:54:08] <robin_z> ahh :)
[11:54:33] <robin_z> the best way with these thing is a fork truck
[11:54:40] <robin_z> and gantrys work well
[11:54:50] <fenn> gantry means what?
[11:54:56] <robin_z> umm
[11:55:04] <robin_z> big beam on legs
[11:55:13] <robin_z> or steel beam in a ceiling
[11:55:18] <fenn> with thing that slides on i beam on bearings?
[11:55:33] <robin_z> hoist it up .. drive the trailer out from underneath, lower it down
[11:55:40] <fenn> oh
[11:55:46] <robin_z> doesnt need to be that complex ...
[11:56:37] <fenn> don't know if i'm really ready for this thing or not
[11:56:48] <fenn> only prior experience is my gingery lathe :)
[11:56:57] <robin_z> this thing is in a scrap yard?
[11:57:00] <fenn> yea
[11:57:08] <anonimasu> hiring somone to tow it would probably be easy/cheap
[11:57:12] <anonimasu> or to lift it up the trailer..
[11:57:15] <robin_z> for $50 they might even drop it off for you
[11:57:27] <anonimasu> dont risk your life if you dont absolutely have to
[11:57:35] <robin_z> yeah
[11:57:44] <robin_z> a mill for $200 is cheap
[11:57:55] <robin_z> but $200 + 1 right foot is expensive
[11:58:04] <anonimasu> exactly..
[11:58:28] <anonimasu> when a mill's on the roll it's on the roll.
[11:58:33] <robin_z> the press I just moved cost $900 in crane fees ..
[11:58:41] <robin_z> but I still have all my fingers and toes
[11:58:43] <fenn> i read an article about a guy who was trying to move a mill he bought at an auction, then it fell on him and nearly bit his leg off
[11:58:52] <robin_z> yeah
[11:58:54] <fenn> he kept taking pictures :)
[11:59:05] <alex_joni> robin: ever changed your policy on tripwire?
[11:59:12] <robin_z> alex_joni: yeah
[11:59:15] <anonimasu> if it starts to move, get away :D
[11:59:26] <robin_z> I think I just did an init afterwards, to re-sign the db
[11:59:32] <anonimasu> brb, going to work now
[11:59:42] <alex_joni> I ran into some problems, when I do tripwire --update-policy twpol.txt
[11:59:58] <alex_joni> it complains about elements missing from the database
[12:00:14] <alex_joni> if I do a new --init it still uses the old policy
[12:00:20] <robin_z> oh ..
[12:00:27] <robin_z> I think I have forgotten
[12:02:12] <robin_z> yeah
[12:02:23] <fenn> can you make a useful tripwire database that will fit on a floppy?
[12:02:41] <alex_joni> ### Error: Policy Update Changed Object.
[12:02:41] <alex_joni> ### An object has been changed since the database was last updated.
[12:02:41] <alex_joni> ### Object name: Conflicting properties for object /etc/tripwire
[12:02:41] <alex_joni> ### > Modify Time
[12:02:46] <robin_z> fenn: sure
[12:02:51] <robin_z> mine is 1.2m
[12:04:26] <alex_joni> if I do a tripwire --update it just says Error: File could not be opened.
[12:04:33] <alex_joni> and it fails to find the database
[12:04:59] <robin_z> * robin_z tries a update-policy
[12:10:17] <fenn> well, if anyone's interested, I started a wiki on building a machine shop from the ground up
[12:10:30] <fenn> http://www1.atwiki.com/gingery_machines/
[12:10:34] <alex_joni> nice
[12:10:53] <alex_joni> fenn: remember to add your machine to the EMC-wiki once working ;)
[12:10:56] <fenn> will cover everything from foundry and whitworth surface plates to hexapod machining platforms
[12:11:31] <fenn> alex: add my machine? the van norman or the hexapod or the lathe or my computer or what?
[12:11:50] <fenn> the wiki is on a wiki host
[12:11:56] <fenn> somewhere in japan
[12:12:11] <alex_joni> there's an emc-wiki (on redpoint.co.uk)
[12:12:17] <alex_joni> and there's a list with machines running emc
[12:12:26] <robin_z> www.linux.cnc.org
[12:12:30] <alex_joni> once you have an machine running emc, you might want to add it there
[12:12:31] <robin_z> follow the wiki link
[12:12:37] <fenn> ah
[12:12:37] <alex_joni> www.linuxcnc.org
[12:13:36] <robin_z> would someone care to add wiki.linuxcnc.org to the DNS and point it to www2.redpoint.org.uk
[12:13:43] <robin_z> as a CNAME
[12:13:51] <alex_joni> bug Steve about that
[12:13:56] <alex_joni> as it's his hosting
[12:14:19] <rayh_> Sounds good to me.
[12:14:41] <rayh_> How does one go about that?
[12:14:41] <alex_joni> * alex_joni is starting to get annoyed by tripwire
[12:14:47] <robin_z> what?
[12:14:56] <robin_z> modifying the DNS??
[12:15:04] <rayh_> add the wiki.linuxcnc
[12:15:15] <robin_z> just change the DNS
[12:16:24] <alex_joni> mcedit /etc/bind/zones.conf ;)
[12:16:24] <alex_joni> or whatever the file is called
[12:16:24] <alex_joni> probably some click & pray for doze (I think Steve is running it on doze)
[12:16:24] <rayh_> And we'd do that on Steve's server?
[12:16:24] <rayh_> NT
[12:16:24] <robin_z> no
[12:16:31] <robin_z> on the name server
[12:16:44] <alex_joni> depends.. maybe Steve's server is the DNS server ;)
[12:17:11] <robin_z> ;linuxcnc.org.INNS
[12:17:11] <robin_z> ;; ANSWER SECTION:
[12:17:11] <robin_z> linuxcnc.org.26956INNSns1.coile.com.
[12:17:11] <robin_z> linuxcnc.org.26956INNSns1.pmdinc.com.
[12:17:25] <alex_joni> well.. not ;)
[12:17:48] <rayh_> pmdinc is his.
[12:18:04] <robin_z> well,
[12:18:05] <robin_z> debian:/var/lib/tripwire# host ns1.pmdinc.com
[12:18:06] <robin_z> ns1.pmdinc.com has address
[12:18:06] <robin_z> debian:/var/lib/tripwire# host www.linuxcnc.org
[12:18:06] <robin_z> www.linuxcnc.org is an alias for linuxcnc.org.
[12:18:06] <robin_z> linuxcnc.org has address
[12:18:15] <robin_z> so smae rack I gues
[12:18:16] <robin_z> s
[12:19:08] <rayh_> * rayh_ is humbled once again by his limited web ability.
[12:19:26] <rayh_> rayh_ is now known as rayh
[12:20:04] <rayh> I'll post a note to steve.
[12:21:18] <alex_joni> robin: dpkg-reconfigure tripwire seemed to do the trick
[12:24:14] <alex_joni> almost anyways...
[12:24:35] <alex_joni> doing a tripwire --check outputs one error /etc config file changed
[12:26:17] <fenn> does the creative commons license actually provide you with any liability protection?
[12:26:29] <fenn> or is it just smokescreen
[12:27:22] <robin_z> liability protection, hell no
[12:27:34] <robin_z> its exactly what it says on the tin
[12:28:04] <fenn> is there even any such thing as liability protection?
[12:28:12] <robin_z> if you want to copy it and use it feel free
[12:28:21] <robin_z> so long as you offer the samew terms
[12:28:30] <fenn> i mean, if i tell someone to go stick their head in a furnace, what do i have to do so they can't sue me?
[12:28:30] <robin_z> yes, I have liablity insurance
[12:31:01] <fenn> afraid i only speak english, japanese, spanish, and italian
[12:32:33] <alex_joni> robin: when I run tripwire --check I get a violation on /etc/tripwire
[12:32:44] <alex_joni> and I did run an tripwire --update
[12:32:55] <alex_joni> but it still does the same, any ideas?
[12:33:43] <robin_z> http://www.tripwire.org/qanda/index.php
[12:33:45] <rayh> There does seem to be some discussion of REQUIRED BY APPLICABLE LAW.
[12:34:08] <fenn> so somone in poland could sue me but not in argentina... hmmm
[12:34:18] <fenn> * fenn cackles
[12:35:58] <rayh> "Linux Today - InfoWorld: GPL May Be Unenforceable Under German Law"
[12:36:47] <fenn> apparently it is illegal to grow garlic in your garden in germany
[12:37:12] <rayh> I can understand that!
[12:37:22] <fenn> it is a medicine since it has "some beneficial health value" so yu are practicing medicine without a license ;)
[12:37:29] <robin_z> well, the way it works
[12:37:39] <robin_z> if you accept the GPL, you can use the software
[12:37:44] <robin_z> if you dont, you cant
[12:38:03] <robin_z> so if its not useable under German law .. then germans cant use Linux ...
[12:43:00] <SWP_Away> SWP_Away is now known as SWPadnos
[12:44:08] <robin_z> hey steve
[12:44:46] <robin_z> SWPadnos: how about adding wiki.linuxcnc.org pointing as a CNAME to www2.redpoint.org.uk
[12:45:23] <alex_joni> robin: that's stephen
[12:45:31] <alex_joni> * alex_joni was talking about SteveStallings
[12:45:36] <robin_z> ahh
[12:45:40] <robin_z> worng steve
[12:46:16] <SWPadnos> sure :)
[12:46:47] <SWPadnos> (sorry - should have marked myself !away after getting coffee :) )
[12:46:54] <rayh> I sent stallings a note a bit ago.
[13:02:05] <alex_joni> robin: does tripwire --init && tripwire --check work properly for you?
[13:11:55] <alex_joni> hey les
[13:13:59] <les> hello alex
[13:26:33] <nevyn_> nevyn_ is now known as nevyn
[13:33:04] <alex_joni> les: what's new?
[13:33:30] <les> oops was away
[13:33:36] <alex_joni> yeah .. me too
[13:33:51] <les> I was just checking about the birmingham cnc show
[13:34:09] <alex_joni> right
[13:34:22] <alex_joni> got some nice pics of your emc machine?
[13:34:34] <alex_joni> besides the one on lmwatts.com ?
[13:34:43] <les> hmm let me check
[13:37:56] <les> only a few...
[13:38:10] <les> I have one pic here of when it was being built
[13:38:23] <alex_joni> good quality?
[13:38:32] <les> yes want it?
[13:38:39] <alex_joni> yup
[13:38:44] <alex_joni> can you mail it to me?
[13:39:06] <les> sure...oh I also have autocad drawings
[13:39:45] <les> I can give you the drawings for the machine design if you want
[13:40:30] <les> email ad?
[13:40:30] <alex_joni> thx.. that would be great too
[13:55:32] <alex_joni> hey steve
[13:55:42] <alex_joni> got the message from rayh?
[13:56:26] <stevestallings> hi Alex, yes, trying to chat with Robin now, but no response yet
[13:56:40] <alex_joni> * alex_joni prods robin_z
[13:57:19] <stevestallings> DNS is on my server, but there are issues.....
[13:57:27] <alex_joni> like?
[13:58:27] <stevestallings> as best I can tell (I am not a guru), the CNAME record must point to another server in the same domain, e.g. mywikiserver.linuxcnc.org
[13:58:27] <robin_z> issues?
[13:58:27] <robin_z> nah
[13:58:27] <robin_z> cnames can point to ANYTHING
[13:58:31] <robin_z> dont forget the dot
[13:58:31] <stevestallings> I think what is needed is a redirect from a web server
[13:58:51] <robin_z> no
[13:58:51] <robin_z> its a CNAME thats required
[13:58:52] <robin_z> but dont forget the dot at the end
[13:59:30] <robin_z> wiki CNAME www2.repoint.org.uk.
[13:59:38] <robin_z> if you do :
[13:59:39] <robin_z> wiki CNAME www2.repoint.org.uk
[13:59:56] <robin_z> then it becomes www2.repoint.org.uk.linuxcnc.org
[14:00:10] <robin_z> the dot at the end is important
[14:00:33] <alex_joni> repoint?
[14:00:39] <robin_z> redpoint. sorry
[14:02:05] <stevestallings> I know about the dot at the end. My DNS admin will even add it automatically. What it will not do is feed me coffee until I quit reversing the ALAIS and HOST fields.
[14:02:17] <les> robin quick qusetion: what is the name of that birmingham show?
[14:03:30] <alex_joni> robin: does www2.redpoint.co.uk even work?
[14:05:28] <stevestallings> OK, it is there, but I had to go for www.redpoint... instead of www2.redpoint.
[14:06:08] <stevestallings> Goes to default page on Quacky.
[14:07:02] <stevestallings> Now what?
[14:08:50] <alex_joni> Steve: make it point to www2
[14:09:23] <alex_joni> I think robin is planning to add a CNAME from there to the actual wiki
[14:09:28] <alex_joni> or whatever
[14:10:24] <stevestallings> OK, now that I know it works I will change back to WWW2
[14:11:32] <alex_joni> coo
[14:12:01] <stevestallings> Of course all the DNS servers have the WWW address cached now. 8-(
[14:12:36] <les> robin has vanished on us
[14:12:43] <les> get mail alex?
[14:12:47] <stevestallings> you noticed
[14:13:06] <les> hi steve
[14:13:12] <stevestallings> hi les
[14:13:31] <rayh> Hey thanks for that Steve and others.
[14:13:36] <alex_joni> les: got some issues with my link :(
[14:13:47] <alex_joni> but the mail is there, just can't download it :((
[14:14:02] <les> It's big
[14:14:09] <les> I could dcc as well
[14:14:56] <les> The picture was during assembly of the machine
[14:15:19] <les> getting things square and straight to .01mm
[14:15:25] <alex_joni> it's not that big (about 1.5 Meg if I've seen it correctly)
[14:15:27] <les> which is a bit of work
[14:16:48] <stevestallings> Les, a few days ago you mentioned a GlentekCNC and I was wondering if they are related to the Glentek servo folks.
[14:17:40] <stevestallings> Hi Ray. What is the plan for wiki.linux.cnc? Should I change the links when it goes live?
[14:17:53] <les> steve...not related
[14:18:07] <les> but I am currently working with them
[14:18:55] <les> can't give much details...but I am contemplating selling a new type of cnc router
[14:19:06] <rayh> stevestallings: Don't think it matters much unless the wiki location changed.
[14:19:06] <les> not sure about the market though
[14:19:52] <rayh> a link to wiki.linuxcnc.org would be easier to maintain.
[14:20:12] <stevestallings> True. Figured that was the aim.
[14:20:21] <alex_joni> les: got the mail
[14:20:30] <les> good
[14:21:08] <robin_z> hey les,
[14:24:12] <alex_joni> * alex_joni goes home
[15:06:06] <A-L-P-H-A> item has finally came in the mail.
[15:08:00] <A-L-P-H-A> the DA300 collet holder.
[16:31:06] <les> alex take a look at the picture and drawing?
[16:31:27] <les> oops gone home
[16:31:33] <les> heh
[17:21:39] <bowika> I am sorry for disturbing you, but I am just reading the Hal_introduction.pdf.........and i read about hal_parport program, and ither hal management progs
[17:21:39] <bowika> >bowika< and I can find these on my emc2 dir
[17:21:40] <bowika> >bowika< how can i compile them or why didn't these progs compile?
[17:27:19] <Bowika> I am sorry for disturbing you, but I am just reading the Hal_introduction.pdf.........and i read about hal_parport program and other hal management progs, and I can find these on my emc2 dir.
[17:27:21] <Bowika> How can i compile them or why didn't these progs compile?
[17:36:34] <rayh> Bowika: I can't help you but there are a couple others that seem to be lurking or away right now that can.
[17:37:33] <Bowika> ok thanks, i am waiting.........
[17:59:40] <rayh> picnet: Do you work with emc2?
[18:33:58] <paul_c> There is no HAL stuff in the bdi-4 branch, and the head of emc2 won't compile on a 2.6 install
[18:34:22] <paul_c> This needs to be resolved during the codeFest.
[19:24:01] <paul_c> * paul_c pounces on alex_joni
[19:24:16] <alex_joni> * alex_joni pounces back
[19:24:55] <alex_joni> what's up ?
[19:25:33] <paul_c> waiting for something to die prematurely...
[19:25:58] <alex_joni> what's gonna die?
[19:26:25] <paul_c> dunno...
[19:26:35] <alex_joni> huh?
[19:26:54] <paul_c> Running a little app, and it spits out "we might die prematurely" once in a while.
[19:27:11] <alex_joni> lol
[19:27:16] <alex_joni> nice
[19:27:31] <alex_joni> but it won't top my favorite error message ;)
[19:27:44] <alex_joni> "you don't exist! go away!"
[19:30:44] <rayh> alex_joni: You're home.
[19:30:50] <les> look at the file and picture alex?
[19:30:55] <alex_joni> les: yes
[19:31:01] <alex_joni> nice
[19:31:16] <alex_joni> didn't get the chance to properly look at it, just a glimpse on both
[19:31:25] <alex_joni> rayh: indeed
[19:31:35] <alex_joni> http://www.plethora.net/~seebs/ops/ibm/cranky37.html
[19:39:22] <rayh> alex_joni: when you get a chance Bowika has a question about emc2 compiles.
[19:39:29] <rayh> He posted while you were out.
[19:39:47] <alex_joni> I'm trying to access the logsd
[19:39:51] <alex_joni> logs
[19:43:22] <Bowika> my questions were:
[19:43:24] <Bowika> I am sorry for disturbing you, but I am just reading the Hal_introduction.pdf.........and i read about hal_parport program and other hal management progs, and I can find these on my emc2 dir.
[19:43:25] <Bowika> How can i compile them or why didn't these progs compile?
[19:44:06] <alex_joni> Bowika: certainly you are not disturbing
[19:44:38] <alex_joni> I didn't use those myself, but if you wait a minute I'll take a look
[19:47:22] <Bowika> i had a private chat with paul_c and he said that it won't work on bdi-4
[19:47:23] <alex_joni> Bowika: as I see it
[19:47:39] <alex_joni> you should get hal_parport and hal_skeleton compiled
[19:47:39] <Bowika> ok
[19:47:54] <alex_joni> what are you trying to do?
[19:48:30] <alex_joni> the other files in hal\drivers only have a RT component
[19:48:44] <alex_joni> and thus they only get compiled as modules (filename.o)
[19:48:52] <Bowika> i just want to try how it works..
[19:49:16] <Bowika> ok
[19:49:25] <alex_joni> well.. if you really want to try you should go ahead and use the RT stuff
[19:49:25] <alex_joni> does emc2 work?
[19:49:25] <alex_joni> scripts/emc.run ?
[19:49:31] <Bowika> yes it works
[19:49:44] <alex_joni> ok.. then I assume hal is working properly
[19:49:56] <alex_joni> next you might want to try some hal components
[19:50:06] <alex_joni> do a scripts/realtime start
[19:50:23] <alex_joni> after that you can insmod the hal modules you want to play with
[19:50:50] <alex_joni> you have 'halcmd' (along with a lot of options) to set params, create signals, link those to pins
[19:51:01] <alex_joni> try to follow the steps in HAL_introduction.pdf
[19:51:14] <Bowika> it will not work 'cause as i saw there was no realtime dir in the bin dirs
[19:51:29] <alex_joni> not in the bin dir
[19:51:45] <Bowika> i only followed the commands of http://www.redpoint.org.uk/cgi-bin/emcinfo.pl?BDI-4_Install
[19:51:56] <alex_joni> errrr
[19:52:02] <Bowika> yes errrrr
[19:52:02] <alex_joni> is this bdi4 ?
[19:52:06] <Bowika> yes it is
[19:52:10] <alex_joni> well then.. bummer
[19:52:18] <alex_joni> bdi4 is not emc2 :)
[19:52:25] <alex_joni> and it doesn't contain hal
[19:52:37] <alex_joni> and emc2 doesn't work on bdi4 neither
[19:52:41] <Bowika> yes, it doesn't
[19:53:02] <alex_joni> that's because emc2 doesn't work on 2.6 (doesn't compile at the moment)
[19:54:24] <Bowika> i downloaded the emc2 by cvs and did a "make" and blabla .....following the commands mentioned above
[19:54:52] <alex_joni> are you sure you downloaded emc2 ?
[19:55:26] <alex_joni> did you use: cvs -d:pserver:anonymous@cvs.sourceforge.net:/cvsroot/emc checkout -Pr bdi-4 -d emc emc2 ?
[19:55:39] <alex_joni> if you did then you DIDN'T download emc2
[19:56:01] <alex_joni> you downloaded the emc from BDI4 (located in CVS under emc2, but the bdi4 branch)
[19:56:24] <Bowika> ok i run "cvs -d:pserver:anonymous@cvs.sourceforge.net:/cvsroot/emc checkout -Pr bdi-4 -d emc emc2"
[19:56:26] <alex_joni> the bdi4-emc is somewhere between emc (which runs with rcslib) and emc2 (runs with libnml)
[19:56:42] <alex_joni> so you actually have emc1 on your machine
[19:57:02] <alex_joni> and the stuff downloaded from CVS resembles more to emc1 than it does to emc2
[19:58:00] <Bowika> ok
[19:58:31] <Bowika> what do i have to do if i'd like to acquire a real emc2?
[19:59:24] <alex_joni> for now? install a 2.4 kernel
[20:00:27] <alex_joni> Bowika: if you have issues with installing / compiling kernels (and patching it with rtai) try installing BDI-Live
[20:01:31] <Bowika> i have the 2.4.18 kernel and rtlinux 3.2 installed, 'cause i usually develop in its
[20:01:44] <alex_joni> right...
[20:01:55] <alex_joni> well run that kernel, and do a cvs checkout:
[20:02:08] <Bowika> that's all?
[20:02:10] <alex_joni> cvs -d:pserver:anonymous@cvs.sourceforge.net:/cvsroot/emc checkout -dP emc emc2
[20:02:23] <alex_joni> make sure you get HEAD not the bdi-4 branch
[20:02:33] <alex_joni> after that ./configure && make
[20:02:39] <alex_joni> should be enough
[20:03:19] <alex_joni> rtlinux-free?
[20:03:26] <Bowika> yes it is
[20:04:20] <alex_joni> might work :)
[20:04:39] <alex_joni> the parts from configure were written kinda blindly (for rtlinux)
[20:04:50] <alex_joni> so I think it's great if you try it
[20:05:03] <alex_joni> if it doesn't work.. please do complain :D
[20:06:19] <Bowika> ok i have to restart my computer.....will you be here in 15 minutes?
[20:06:28] <alex_joni> think so :)
[20:07:02] <Bowika> ok i will try it
[20:07:47] <Bowika> bye for a while
[20:13:04] <rayh> alex_joni: Thanks a bunch.
[20:13:20] <alex_joni> rayh: what for?
[20:20:31] <alex_joni> alex_joni is now known as alex_joni_away
[20:41:02] <alex_joni_away> alex_joni_away is now known as alex_joni
[20:41:10] <alex_joni> * alex_joni is back
[20:43:12] <alex_joni> now where's Bowika?
[20:46:27] <SWPadnos> rebooting still?
[20:47:21] <alex_joni> starting processor: 1137/5128
[20:47:36] <SWPadnos> big cluster there ;)
[20:47:43] <alex_joni> yeah
[20:47:46] <alex_joni> =))
[20:48:09] <alex_joni> hope he'll be back soon
[20:48:18] <alex_joni> I'm preparing to crash :)
[20:48:18] <SWPadnos> getting late for you?
[20:48:30] <alex_joni> a few more mins till tomorrow
[20:48:42] <SWPadnos> heh
[20:48:52] <SWPadnos> Right - Romania
[20:49:45] <alex_joni> EEST=GMT+2
[20:51:11] <SWPadnos> We're usually GMT-5, but I'm not sure what happens with daylight savings (or standard time or whatever we're actually on in the summer)
[20:51:17] <SWPadnos> It's 17:00 here
[20:51:49] <SWPadnos> (EST5EDT)
[20:54:20] <les> hi
[20:54:27] <SWPadnos> he there
[20:54:31] <SWPadnos> hi there, even
[20:54:47] <alex_joni> hey les
[20:54:55] <les> had a long long talk with Mariss at gecko
[20:55:06] <alex_joni> cool.. bout what?
[20:55:14] <les> might try the g200x for a commercial control
[20:55:19] <alex_joni> how's the Vampire coming?
[20:55:29] <les> robin is messing with it
[20:55:39] <alex_joni> with the g200x .. yes
[20:55:45] <les> Vampire?
[20:55:53] <SWPadnos> the unkillable stepper driver
[20:56:01] <les> oh
[20:56:16] <les> new boards go on the pick and place tommorow
[20:56:23] <alex_joni> although.. people suggested it might die from a wooden stick that gets rammed through the PCB
[20:56:25] <SWPadnos> the G200X OEM?
[20:56:34] <les> I was counting on emc
[20:56:37] <SWPadnos> no - only silver bullets (and garlic causes it to malfunction)
[20:56:38] <les> yes the oem
[20:56:44] <les> but I need a backup
[20:56:58] <SWPadnos> You can make your own :)
[20:57:05] <SWPadnos> (the schematics are published)
[20:57:25] <les> A great fear with customer support if I put emc in a commercial product
[20:57:31] <les> it works great for me
[20:57:37] <les> but I am an engineer
[20:57:42] <SWPadnos> the G200X is emc
[20:57:45] <les> customers would not be
[20:57:58] <alex_joni> SWP: partly
[20:58:02] <les> the interpreter is emc
[20:58:05] <alex_joni> not neccesarely
[20:58:06] <les> the rest is not
[20:58:13] <SWPadnos> highly modified and optimized for a small microcontroller
[20:58:17] <alex_joni> afaik there's another version aswell
[20:58:19] <les> yeah
[20:58:24] <SWPadnos> yes - the Machx version
[20:58:54] <les> the mach version is a separate thing art is doing
[20:58:58] <SWPadnos> yes
[20:59:07] <les> you can put whatever you want on the rabbit
[20:59:26] <SWPadnos> or even make a different controller board (which I may do with an AVR or something)
[20:59:29] <les> c compiler dev kit =$150
[20:59:37] <les> according to Mariss
[20:59:59] <SWPadnos> yeah - sort of. I picked up their literature at ESC, and the "real" dev kits seem like they're a bit more
[21:00:03] <alex_joni> les: c compiler dev kit for AVR = 0$
[21:00:05] <SWPadnos> I'd have to look it over again though
[21:00:17] <alex_joni> includes gcc, gdb, etc
[21:00:20] <les> I just told him I needed a controller that could do smooth 1000 ipm motions
[21:00:21] <SWPadnos> not quite - that's only if you can fit into 4k of object code
[21:00:27] <SWPadnos> the assembler for the AVR is free though
[21:00:31] <les> with .001" repeatability
[21:00:35] <SWPadnos> oh - you're talking GCC - nevermind
[21:00:42] <alex_joni> JTAG for avr = 50$
[21:00:56] <les> and not use off the shelf components for a mee too product
[21:01:35] <SWPadnos> Either the G200X or Jon Elson's board can do that
[21:01:43] <les> for emc to be considered segmentqueue must work
[21:01:50] <SWPadnos> (or a custom FPGA board)
[21:02:09] <les> and the linux must be transparent to the user
[21:02:38] <alex_joni> les: there was a suggestion a while back
[21:02:40] <les> linux is good, but I can't sell that learning curve to a signmaker or something
[21:02:40] <SWPadnos> that can only work if it's basically impossible for the user to run any other application (like CAD, word processor, email, etc.)
[21:02:42] <alex_joni> make emc run on linux
[21:02:51] <alex_joni> but the display on doze
[21:03:00] <les> exactly
[21:03:03] <SWPadnos> yes - the "ethernet CNC"
[21:03:15] <les> that is a good plan
[21:03:18] <alex_joni> make the emc-machine a SBC burried deep inside
[21:03:26] <alex_joni> maybe even with linuxbios :)
[21:03:33] <alex_joni> boots in a few seconds
[21:03:37] <SWPadnos> les - you should add segmentqueue discussions to the Fest ToDo list
[21:03:59] <les> emc's advantage is (with a working segmentqueue) is that it has sophisticated motion planning and servo loops
[21:04:05] <SWPadnos> actually - there are flash IDE hard drives (up to a couple of gig) that would significantly increase boot speed)
[21:04:33] <alex_joni> SWP: how so?
[21:05:02] <robin_sz> evening ..
[21:05:04] <SWPadnos> Well - data transfer and seek are a part of bootup (though I realize that waiting for peripherals to respond is a lot more of the time)
[21:05:34] <alex_joni> SWP: afaik flash is a lot slower than hdd's
[21:05:39] <les> anyway Mariss will send me a 200x if I buy the rabbit
[21:05:41] <SWPadnos> I should have said "flash drives that plug into an IDE connector
[21:05:51] <SWPadnos> true for writes (though marginally), not necessarily true for reads
[21:05:53] <alex_joni> yeah that too
[21:05:54] <robin_sz> les: kewl ... buy the Rabbit from mariss presubaly
[21:06:04] <alex_joni> reads ~ 10-12 MB?
[21:06:07] <SWPadnos> damn - I'll take that deal :)
[21:06:07] <les> sure why not
[21:06:09] <robin_sz> as it will need to be programmed
[21:06:20] <les> dev kit $150?
[21:06:24] <robin_sz> nah
[21:06:28] <robin_sz> oh,
[21:06:30] <robin_sz> yeah
[21:06:41] <SWPadnos> Even compactflash is in the 4-6 MB/sec write range
[21:06:47] <robin_sz> but a programming lead is $25 and there is an updater for free
[21:06:58] <robin_sz> you only need the dev to compile
[21:06:58] <les> oh and he explained the convolution scheme
[21:07:12] <robin_sz> SWPadnos: dont forget .. flash burns out in 100K r/w cycles
[21:07:15] <les> it is not convolving waypoints
[21:07:21] <robin_sz> les: makes sense now?
[21:07:29] <les> but velocity by interrupt
[21:07:35] <SWPadnos> true, but for a boot drive (non-swap) that shouldn't be an issue
[21:07:36] <robin_sz> yes, thats sounds right
[21:07:38] <SWPadnos> vibration proof as well :)
[21:07:44] <les> so constant filter poles
[21:08:01] <robin_sz> right
[21:08:06] <SWPadnos> in the rabbit software?
[21:08:11] <robin_sz> yeah
[21:08:17] <les> anyway he had good arguments
[21:08:23] <SWPadnos> he always does :)
[21:08:26] <les> I should try it
[21:08:31] <alex_joni> SWP: how about battery backed up SRAM?
[21:08:33] <robin_sz> the rabbit has the interp, tp and motion smoothers all in it
[21:08:38] <SWPadnos> nope
[21:08:41] <alex_joni> now that would rock
[21:08:45] <alex_joni> 100MB/sec+
[21:08:52] <SWPadnos> yesh - like suspend / resume
[21:09:03] <alex_joni> yup
[21:09:35] <les> although I still take issue with his using an h-bridge switching scheme that dissipates servo energy in the output fets in regenerative braking
[21:09:40] <robin_sz> les: regardless of mariss convioncing arguments, i have the results running here .. and thats the biggest convincing argument you will ever see :)
[21:10:12] <les> well it will do smooth 1000 ipm moves...
[21:10:23] <les> and .001" repeat
[21:10:24] <robin_sz> les: you should have seen the steppers run emcfin.nc after the post-processor ...
[21:10:31] <les> and is DIRT cheap
[21:10:34] <robin_sz> smooooth
[21:10:36] <robin_sz> yep
[21:10:37] <les> it works for me
[21:11:01] <SWPadnos> actually, you should be able to get higher speed and/or higher resolution with it
[21:11:06] <SWPadnos> (if you need it)
[21:11:11] <les> replaced all those lines with arcs?
[21:11:16] <robin_sz> some
[21:11:24] <robin_sz> 84% reduction :)
[21:11:27] <les> must not have been emc1
[21:11:36] <robin_sz> no was on the G2002
[21:11:56] <les> it is incapable of runnung joined arcs in anything other than exact stop mode
[21:12:04] <les> makes gouges
[21:12:15] <robin_sz> segmotqueus runs them in blended mode though??
[21:12:33] <robin_sz> I thought you had segmot?
[21:12:33] <les> robin how big is your test machine?
[21:12:43] <les> segmentqueue blends arcs yes
[21:12:51] <anonimasu> what are you discussing?
[21:12:59] <robin_sz> les this is just a test frame at home, 3 steppers in a rack
[21:13:08] <anonimasu> heh
[21:13:18] <robin_sz> les; I have a 2.5m x 1.25m plasma at the factory
[21:13:23] <les> g200x flavors of emc
[21:13:27] <les> or part emc
[21:13:29] <SWPadnos> Come on over - we can attach them to my Bridgeport :)
[21:13:38] <robin_sz> heh
[21:13:41] <les> heh
[21:14:00] <les> the emc plaque is fairly nice looking
[21:14:06] <SWPadnos> (though if I had mounts made, MY motors would already be attached)
[21:15:11] <SWPadnos> alex_joni: also, flash discs don't have to spin up - so especially if you use LinuxBIOS, you get a much faster boot
[21:15:20] <robin_sz> would a post-process version in purely lines be of interest??
[21:15:35] <alex_joni> right
[21:15:54] <SWPadnos> What does pause / resume do to that idea?
[21:16:06] <robin_sz> to which idea?
[21:16:18] <robin_sz> the G2002?
[21:16:23] <robin_sz> or segmot
[21:16:25] <SWPadnos> the lines-only post -I suppose nothing, really
[21:16:31] <les> well I have it with different granularity as far as line segments
[21:16:35] <robin_sz> should work perfectly
[21:16:40] <les> emcfine is the tough one
[21:16:57] <les> designed specifically to beat up controls
[21:17:04] <robin_sz> yeah
[21:17:12] <robin_sz> it killed the G2002 to a crawl
[21:17:30] <robin_sz> segemnts shorter than 1 interrupt period do that
[21:17:55] <robin_sz> after processing it ran as sweet as a nut
[21:18:10] <robin_sz> you could hear it making lovely sounds around the curls ...
[21:18:27] <les> that speaks well of g200x
[21:19:30] <les> Had reservations about Mariss' tp calcs as opposed to a full third order bezier
[21:19:38] <robin_sz> yeah
[21:19:39] <les> but if it works
[21:19:43] <robin_sz> its a bit of a fudge
[21:19:49] <les> yes
[21:19:50] <robin_sz> but it does seem to work
[21:20:10] <robin_sz> certainly for most "normal" tasks
[21:20:16] <les> anyway he said If I didn't like his fudge...change it
[21:20:22] <robin_sz> heh
[21:20:30] <robin_sz> thats open source for you
[21:20:37] <les> yup
[21:20:47] <robin_sz> probably best get a dev kit then :)
[21:21:23] <robin_sz> at least 60% of the code will be totally familiar
[21:21:32] <robin_sz> if you know EMC
[21:21:54] <robin_sz> the code is in the files section of the group, as binaries and source
[21:23:13] <SWPadnos> hmmm - at 1000IPM, one interrupt is 1/60 inch (0.01666) - how often do you need to do that fast a direction change (given inertia and cutter size)?
[21:23:28] <les> I guess he is using the interpreter, task handler, and such
[21:23:40] <robin_sz> roughly speaking, yes
[21:23:52] <robin_sz> without the nml ickiness
[21:24:05] <anonimasu> um..
[21:24:08] <anonimasu> 0.423164mm
[21:24:16] <les> swp: It's more of an issue of stopping at a particular point
[21:24:29] <anonimasu> that's enough to hit a clamp.
[21:24:40] <SWPadnos> I was referring to the "segments shorter than one interrupt" thing
[21:24:45] <les> could be
[21:24:56] <les> yeah I understand
[21:25:01] <les> it's an issue
[21:25:29] <les> well let have hypothetical 0.2 g accel...
[21:25:36] <les> 1000 ips....
[21:25:38] <SWPadnos> Just trying to understand why (and in what situations)
[21:25:50] <SWPadnos> 1000 IPM
[21:25:59] <les> oops right
[21:26:01] <les> so
[21:26:06] <SWPadnos> heh - that would be a problem
[21:27:31] <robin_sz> the 1khz int rate is akin to a servo update rate, its always a good idea to have it at least twice the rate of the basic system freq response .. its not hard to conceive a system with a 500hz basic freq response is it?
[21:28:01] <robin_sz> and .. you can always set it lower of you waish
[21:28:02] <les> ok .2g at that speed gives a min corner radius of 0.21"
[21:28:57] <les> so yes must slow down for sharp corners
[21:29:24] <les> ideally slowing spindle too to mantain chip load
[21:29:32] <les> but that would be problematic
[21:29:53] <les> spindles are not designed for low inertia
[21:30:22] <les> and yeas robin must not violate nyquist
[21:30:33] <les> brb
[21:30:52] <SWPadnos> me too - dinner and the wife just arrived :)
[21:34:07] <cradek_> cradek_ is now known as cradek
[21:45:47] <les> had to run out before the general store closed
[21:46:28] <les> hey chris
[21:51:18] <alex_joni> night guys
[21:51:56] <les> night alex
[21:53:25] <rayh> see you alex.
[22:02:12] <cradek> hi les
[22:39:48] <dmess> Hello all... ; )
[22:40:41] <asdfqwega> You do not exist. Go away!
[22:41:06] <A-L-P-H-A> didn't I read that earlier on slashdot?
[22:42:15] <asdfqwega> I don't know...did you?
[22:42:51] <asdfqwega> Ugh...it's been awhile since I've actually done some machining...
[22:43:19] <asdfqwega> Had to do a 2-3/8"-6 thread
[22:43:50] <asdfqwega> with no compound and no thread dial indicator
[22:44:53] <asdfqwega> But, it's like riding a bike or smashing your thumb with a hammer...you never forget how ;)
[22:46:03] <asdfqwega> Now that that is done...food!
[22:52:58] <anonimasu> goodnight everyoen
[22:58:04] <dmess> i can see my ms network... woo hoo....
[23:03:49] <paul_c> * paul_c does an nmap on dmess
[23:08:28] <dmess> whats that do?
[23:10:16] <paul_c> checks for open ports
[23:18:43] <les> better do a dmess on nmap too
[23:18:50] <les> can't be too careful.
[23:21:44] <les> Ran infamous emcfine.tap on the g200x/partly emc
[23:21:50] <les> fairly smooth
[23:21:55] <les> robin ran it
[23:25:00] <paul_c> I also had "fairly smooth" runs for the most part.
[23:25:15] <les> mess with the queue size?
[23:26:15] <les> Had a long long telphone talk with Mariss fremantis today
[23:26:45] <les> he is not using segmentqueue
[23:27:05] <les> but has changed the TP
[23:28:05] <les> emcfine.tap should really beat up kernel/user comms from what you say
[23:30:39] <paul_c> I plan on doing some tests over the weekend - But I need to get this CD finished first
[23:30:55] <les> it beats up Mariss' 22 MHz z-80 too
[23:30:57] <les> cool
[23:31:54] <les> I must get off...bad thunderstorm near
[23:31:57] <les> later
[23:32:39] <SWPadnos> see ya
[23:32:44] <les> bye
[23:32:57] <SWPadnos> paul_c: are you doing a newer BDI-4?
[23:33:31] <paul_c> No.
[23:33:48] <SWPadnos> ok
[23:41:45] <dmess> is the live CD related to BDI 4.2??
[23:44:03] <paul_c> it uses someof the packages, yes.