Back
[10:02:43] <cradek> ... and almost certainly will be eliminated for the version 2.2 release of EMC
[10:02:46] <cradek> haha
[11:00:48] <seb_kuzminsky> nearly certainly :-)
[11:02:19] <cradek> a coworker used to say "virtually" means "not" (like when a programmer says "that's virtually done")
[11:08:36] <seb_kuzminsky> a couple of sample configs & tests use freqgen, so it's not going to be as simple as 'git rm'
[11:08:56] <seb_kuzminsky> and 'hal_demo', whatever that is
[11:10:18] <cradek> but if some intrepid adventurer would fix todo-list-#8 first...
[11:11:02] <seb_kuzminsky> ah yes
[13:55:36] <andypugh> Maybe I am too old to learn Python.
[13:58:59] <andypugh> I didn't have too much trouble learning the basics of C, but C is like a programming language. Python is like a very foreign language with wierd grammar and too much vocabulary
[14:17:46] <ju-emb> andypugh: Ack
[14:17:48] <ju-emb> I'm a C programmer for long time c, objc, c++, no problem.
[14:17:50] <ju-emb> Whenever possible I keep distance to any interpreted language.
[14:17:52] <ju-emb> Last week I started to go after python, and it's as you say, the code reads like novel and interpretation of the code is just in programmers hand.
[14:17:54] <ju-emb> On the other hand, its really powerful, with just a few lines you get a lot done.
[14:19:03] <micges> "baterries included"
[14:19:35] <micges> (libraries for most of common tasks)
[14:20:17] <andypugh> Yeah, but you can spend a lot longer looking for the library and working out how to use it than it would take to code it from scratch. Well, sometimes.
[14:20:47] <micges> yep, many times
[14:22:31] <ju-emb> tandypugh: hat's a common problem of all languages one uses for application programming.
[14:22:33] <ju-emb> the libs are packed with functions
[14:22:35] <ju-emb>
[14:22:37] <ju-emb> Problem I see is, if all the youngsters just learn java php or python, who's writing the kernels for tomorrow?
[14:23:37] <ju-emb> JREoC?
[14:25:59] <mhaberler> the youngsters which learn java or php today will have lots of time to write tomorrow's kernels once they are out of their jobs ;)
[14:26:28] <andypugh> You might find
http://coding2learn.org/blog/2013/07/29/kids-cant-use-computers/ interesting reading
[14:34:14] <micges> andypugh: here simmilar situation was already 10 years ago, now it's probably worse
[15:37:50] <cradek> andypugh: he's mostly right, and it makes me sad.
[15:38:59] <andypugh> I rememeber all the progammers I know being scared of being replaced by kids. Now they are scared that there is nobody to replace them. (Expecially the COBOL contractors)
[15:40:32] <jepler> mhaberler: I have no technical advice about what libm replacement the rtai folks should consider. My interest in this is purely from the license angle, and I'm also not excited about doing another license audit on somebody else's code.
[15:41:14] <mhaberler> ok
[15:43:29] <cradek> andypugh: there are probably about the same number as 10 years ago and as 20 years ago and maybe 30. That all the rest of the kids know how to use facebook and play angry birds is irrelevant and the only mistake is thinking it's not.
[15:44:15] <cradek> andypugh: so there's no crisis, but also: people aren't getting any smarter.
[15:45:20] <andypugh> All my friends can write software. None of the kids I know can write software.
[15:45:27] <andypugh> (Selection bias
[15:45:32] <cradek> big time
[15:45:52] <jepler> I do sometimes wonder how the comp sci coursework has changed since I was in ('93-97 or so). In a couple of 2nd/3rd year classes you got a pretty complete tour of everything from transistors to compiled languages, though it was obviously all pretty shallow.
[15:46:47] <cradek> in my day, we used to learn both typing and writing...
[15:46:50] <andypugh> You could probbaly do a masters in the history of language development by now. How many programming languages are there?
[15:46:57] <andypugh> (and why?)
[15:47:29] <andypugh> On my first computer I ran out of line numbers :-)
[15:48:12] <andypugh> (It could only handle 9999 lines, and it was BASIC, and needed line numbers)
[15:48:19] <jepler> only 9999 lines?
[15:48:50] <seb_kuzminsky> i was one one of the last classes that used a breadboard in logic circuits class - they all use fpgas now, apparently
[15:50:46] <cradek> eventually it'll be hard to find DIP ttl/cmos gates that are nicely breadboard-compatible
[15:51:36] <cradek> (and even though I'm a dinosaur, I almost always use a microcontroller where I would've previously used gates)
[15:51:46] <jepler> huh, Commodore 64 BASIC V2's maximum line number is 63999
[15:52:17] <jepler> of course you couldn't run out of line numbers, because there were only 39811 bytes for BASIC programs, and each line took a minimum of 4 bytes
[15:53:32] <micges> iirc it has also two letters variables names
[15:53:39] <jepler> yep
[15:53:48] <jepler> not letter-and-optional-digit like some cheap BASICs either
[15:54:40] <seb_kuzminsky> cradek: dont even try it, i've seen your vacuum tube circuit
[15:54:59] <seb_kuzminsky> circuits
[15:55:51] <cradek> hmm
[15:56:20] <cradek> did I show you the no-transistor digital clock that sometimes even keeps time?
[15:56:41] <cradek> hm, I haven't built a clock for a few years
[15:59:23] <andypugh> The first computer I used was a ZX81. That could have 26 variables :-)
[16:16:22] <andypugh> How strange: /home/andypugh/git/linuxcnc-dev/scripts/linuxcnc: line 716: 12540 Segmentation fault $EMCDISPLAY -ini "$INIFILE" $EMCDISPLAYARGS $EXTRA_ARGS
[16:16:58] <andypugh> (No, that isn't what line 716 looks like)
[16:24:52] <cradek> amd64 system?
[16:26:43] <cradek> perhaps
http://sourceforge.net/p/emc/bugs/342/
[16:27:49] <andypugh> Well, actually it is a Mac system.
[16:28:18] <andypugh> But it was working moments ago, as a scrren shot I posted to the forum will attest.
[16:32:00] <andypugh> Some Axis configs work, some don't. Which is very puzzling.
[16:32:09] <cradek> mine also went from working to not working for no reason I understood
[16:32:21] <cradek> get a core and see if it's the same problem
[16:32:31] <andypugh> (This is in my own development branch, so there are many possibilities)
[16:33:08] <cradek> well there's a fix in master (I think?) so maybe you should just pull it and see
[16:38:40] <andypugh> Worth a try, of course. More likely that I have caused a different problem.
[16:39:05] <cradek> why not look at the backtrace?
[16:43:20] <andypugh> Err, because I don't know how?
[16:43:29] <andypugh> I seem to have also broken YouTube.
[16:43:44] <cradek> ulimit -c unlimited; make it crash; gdb /path/to/whatver core
[16:43:54] <cradek> in this case it's probably /usr/bin/python
[16:44:02] <andypugh> 500 Internal Server Error // Sorry, something went wrong. // A team of highly trained monkeys has been dispatched to deal with this situation. // If you see them, show them this information: // ZWxthzRmjxg8AQUc9aheaEYhwMK_P3mzwodgMXsXCKqwkonOztk9JSd7sazj
[16:44:02] <andypugh> snxQ0sGYjstfsqf9so94h04xqQe95RHCsDmjU43L0imy6i1OuaFXVA2cClS7
[16:44:35] <andypugh> (That's YouTube)
[16:47:30] <andypugh> And it looks like the LinuxCNC error will remain a mystery, as a pull, make and launch seems to be working fine.