| 00:48.05 | *** join/#picogui CIA (CIA@cia.utility.freenode) |
| 00:57.52 | *** join/#picogui turran (~jl@192.Red-80-37-186.pooles.rima-tde.net) |
| 01:07.10 | *** join/#picogui lurgyman (~lurgyman@home-dhcp6-8.Colorado.EDU) |
| 01:20.16 | *** join/#picogui CIA (CIA@cia.utility.freenode) |
| 01:42.05 | *** join/#picogui Leno (LLAMA@12-255-38-19.client.attbi.com) |
| 02:37.43 | *** join/#picogui captain_pistachi (~blardebla@home-dhcp6-158.colorado.edu) |
| 02:41.46 | *** join/#picogui purple_c1w (~jupiter@home-dhcp6-158.Colorado.EDU) |
| 04:35.47 | njs | ~seen scanline |
| 04:35.48 | | scanline is currently on #picogui (2d 8h 25m 5s) #bzflag (2d 8h 25m 5s). Has said a total of 21 messages. Is idling for 6h 26m 44s |
| 04:41.33 | scanline | hm? |
| 04:42.16 | njs | scanline: totally random question, but I figure you're more likely to know than most: what's the latency on a USB connection? |
| 04:45.11 | *** join/#picogui chaoscon (ph33r@h24-71-217-78.ek.shawcable.net) |
| 04:45.59 | chaoscon | anyone that works with CIA around? |
| 04:46.01 | scanline | njs: depends on the device's polling interval |
| 04:46.20 | scanline | njs: no less than 1ms or so |
| 04:46.29 | scanline | chaoscon: yep |
| 04:47.05 | chaoscon | ah scanline you are alive hehe |
| 04:47.10 | njs | hmm, >1ms? how high does it go? |
| 04:47.12 | scanline | I think so |
| 04:47.30 | chaoscon | does CIA hate being called from a .qmail file? |
| 04:47.37 | scanline | njs: the polling interval is an 8-bit value in milliseconds, so between 1 and 255ms, plus whatever latency the host controller and kernel cause |
| 04:47.44 | scanline | chaoscon: I don't know |
| 04:47.48 | chaoscon | and what exactly do I pass to testClient to see if he really is respoding |
| 04:47.52 | chaoscon | responding* |
| 04:47.53 | scanline | chaoscon: are you setting up your own bot? |
| 04:48.21 | chaoscon | yea, I can get him online and all |
| 04:48.21 | njs | (don't need any really specific numbers, I'm trying to figure out how annoyed I should be that my new box has no game port :-)) |
| 04:48.38 | chaoscon | but I cant find why he isn't responding correctly do anything |
| 04:48.53 | chaoscon | do = to |
| 04:49.03 | njs | he probably doesn't like you. overthrow a few small countries to show him you care. |
| 04:49.25 | scanline | njs: ah :) |
| 04:49.47 | scanline | njs: USB's polling-onlyness is rather annoying, but I managed to stretch low-speed USB pretty far with the mi6k |
| 04:50.21 | njs | is the polling interval hardwired into the device, or settable on host? |
| 04:50.33 | scanline | yes :) |
| 04:50.43 | *** join/#picogui captain_pistachi (~blardebla@home-dhcp6-158.Colorado.EDU) |
| 04:50.51 | njs | = depends on the device? |
| 04:51.01 | scanline | the device's endpoint descriptors specify polling intervals, but the kernel module doesn't have to follow it |
| 04:55.04 | njs | wow, they still make serial port gamepads? wacky. |
| 04:56.19 | *** join/#picogui lurgyman (~lurgyman@home-dhcp6-8.Colorado.EDU) |
| 04:57.04 | njs | hmm, let me revise that: you can still buy them old remaindered ones "new" :-) |
| 04:57.08 | njs | s/them// |
| 05:18.41 | *** join/#picogui lurgyman_ (~lurgyman@home-dhcp6-8.Colorado.EDU) |
| 07:39.45 | captain_proton | ~whaleslap scanline |
| 07:39.46 | | ibot slaps scanline upside and over the head with one freakishly huge killer whale named hugh |
| 07:44.14 | scanline | oh no! |
| 08:06.44 | *** join/#picogui scanline (~micah@home-dhcp5-110.Colorado.EDU) |
| 10:58.58 | *** join/#picogui TD (~mike@m34-mp1.cvx5-c.man.dial.ntli.net) |
| 11:24.31 | TD | scanline: navis host key changed? |
| 12:58.42 | *** join/#picogui TD_ (~mike@m14-mp1.cvx5-c.man.dial.ntli.net) |
| 13:20.14 | *** join/#picogui lalo (~lalo@200-102-210-082.paemt7012.e.brasiltelecom.net.br) |
| 14:05.16 | *** join/#picogui TD (~mike@m14-mp1.cvx5-c.man.dial.ntli.net) |
| 14:05.39 | *** join/#picogui lurgyman (~lurgyman@home-dhcp6-8.colorado.edu) |
| 14:23.53 | *** join/#picogui TD_ (~mike@m25-mp1.cvx5-c.man.dial.ntli.net) |
| 15:27.53 | *** join/#picogui prpplague (~billybob@12.148.134.14) |
| 15:53.59 | *** join/#picogui turran (~jl@192.Red-80-37-186.pooles.rima-tde.net) |
| 16:08.35 | *** join/#picogui TD (~mike@m32-mp1.cvx5-c.man.dial.ntli.net) |
| 17:01.31 | *** join/#picogui TD (~mike@m32-mp1.cvx5-c.man.dial.ntli.net) |
| 17:40.13 | *** join/#picogui TD (~mike@m11-mp1.cvx5-c.man.dial.ntli.net) |
| 18:09.21 | scanline | TD: yes |
| 18:09.56 | TD | ok, i'll whack ssh then |
| 18:10.03 | TD | good to see it back up :) |
| 18:10.49 | TD | (he says, as he does a commit) |
| 18:11.02 | TD | scanline: i think i asked you this before, but have you looked at arch? |
| 18:11.15 | TD | i was reading the tutorial today. it seems to be gaining traction at approx the same rate as subv |
| 18:11.55 | scanline | yes, I've looked at arch |
| 18:11.59 | TD | also, you work with python a lot. how do you avoid errors caused by lack of static typing? |
| 18:12.11 | TD | the best approach i've seen is to have truckloads of unit tests, but they aren't always easy to do (for guis etc) |
| 18:12.14 | TD | thoughts? |
| 18:12.24 | scanline | the main reason I chose svn instead of arch is that you can't efficiently checkout only part of a repository with arch |
| 18:12.27 | scanline | and that's a feature I need |
| 18:12.42 | TD | hmm. ok |
| 18:13.40 | scanline | as for static typing... I've really never had a problem with that. If I make it clear what interface I'm expecting from an argument, and that interface is provided, everything Just Works. Of course, I haven't written much python code that's actually being used by others yet. |
| 18:14.17 | TD | yeah. it prompted me to think because on gnomedesktop.org, somebody was saying that considering it's written in a "safer" language (python) than C, it crashes an awful lot |
| 18:14.40 | TD | i've also seen quite a few backtraces in the RH tools that look like |
| 18:14.44 | TD | er, type errors |
| 18:16.12 | scanline | I wouldn't use python for a system that needs extreme reliability, like a kernel, window server, or firmware |
| 18:16.39 | scanline | but in the applications that I use python for, the extra power I get from dynamic types outweigh the opportunity for error |
| 18:17.55 | TD | why do you think it's harder to write reliable software with it? |
| 18:18.41 | scanline | with a statically typed language, you get the benefit of a compiler that in effect statically performs tests on your modules' interfaces |
| 18:19.00 | scanline | I'm sure you could do the same with many unit tests, but most people are too lazy :) |
| 18:19.09 | scanline | of course, if there is an error you can recover from it easier with python |
| 18:19.17 | scanline | am I doing your homework for you or something? :) |
| 18:20.23 | TD | lol, no |
| 18:20.28 | TD | no homework for me, not yet anyway |
| 18:20.31 | TD | not until october :) |
| 18:20.54 | TD | i was just thinking through the issues of moving to the "new infrastructure", ie python/.net, arch/subversion, scons/whatever |
| 18:21.16 | scanline | ah |
| 18:21.24 | TD | what impact it might have on reliability and so on |
| 18:35.15 | njs | I bet the backtraces in RH tools are real errors that just happen to look like type errors, because they didn't check for an abnormal condition when they should have and the dynamic type checking is usually the first thing to notice an inconsistency. |
| 19:08.47 | file[school] | they're magically delicious |
| 19:34.57 | *** join/#picogui CIA (CIA@cia.utility.freenode) [NETSPLIT VICTIM] |
| 19:34.57 | *** join/#picogui jojo (jojo@infidyne.com) [NETSPLIT VICTIM] |
| 19:34.57 | *** join/#picogui fragglet (~fraggle@pc2-glfd1-4-cust207.glfd.cable.ntl.com) [NETSPLIT VICTIM] |
| 19:34.57 | *** mode/#picogui [+b *!*@*.dip.t-dialin.net] by brunner.freenode.net |
| 19:38.22 | *** join/#picogui lalo (~lalo@200-102-210-082.paemt7012.e.brasiltelecom.net.br) |
| 19:40.46 | *** join/#picogui jojo (jojo@infidyne.com) |
| 20:19.05 | *** join/#picogui lalo (~lalo@200-102-210-082.paemt7012.e.brasiltelecom.net.br) |
| 20:21.39 | *** join/#picogui lalo (~lalo@200-102-210-082.paemt7012.e.brasiltelecom.net.br) |
| 20:50.42 | *** join/#picogui TD (~mike@m97-mp1.cvx5-c.man.dial.ntli.net) |
| 21:00.00 | *** join/#picogui file[wifi] (~file@mctn1-2962.nb.aliant.net) |
| 22:08.23 | *** join/#picogui lurgyman (~lurgyman@home-dhcp6-8.colorado.edu) |
| 22:44.11 | *** join/#picogui file[wifi] (~file@mctn1-2962.nb.aliant.net) |
| 23:31.26 | *** topic/#picogui by scanline -> picogui! it's small, sweet and everybody loves it |
| 23:37.59 | *** join/#picogui lurgyman (~lurgyman@home-dhcp6-8.colorado.edu) |