#linuxcnc-devel Logs

Mar 19 2018

#linuxcnc-devel Calendar

11:21 AM jepler: forum bump on the way, php security updates, ugh
11:34 AM seb_kuzminsky: thanks for maintaining the forum jepler
11:40 AM jepler: yw
11:49 AM seb_kuzminsky: linuxcnc is trademarked, right? how's that handled?
11:50 AM seb_kuzminsky: i vaguely recall cradek saying something about how we're piggybacking on somebody elses' trademark-fu... the linux foundation maybe?
11:53 AM jepler: seb_kuzminsky: yes, roughly
11:55 AM jepler: seb_kuzminsky: we have a Linux Sublicense Agreement from the Linux foundation (linuxfoundation.org / linuxmark.org)
11:57 AM jepler: there is no registration or pending registration of a "linuxcnc.org" or "linuxcnc" trademark. Wikipedia's unaccurate and not jurisdiction-specific summary of trademark law says that "Proprietary rights in relation to a trademark may be established through actual use in the marketplace, or through registration of the mark with the trademarks office"
11:58 AM seb_kuzminsky: cool
11:58 AM seb_kuzminsky: i'll add that info to the "Services and Responsibilities" document i'm working on
11:58 AM jepler: the emc-board archive has some communications with linuxfoundation on it, circa 2011
12:09 PM seb_kuzminsky: there's no emc-board archive on sf any more
12:24 PM seb_kuzminsky: i added trademark info here: http://highlab.com/~seb/linuxcnc/services-and-responsibilities.html
12:24 PM seb_kuzminsky: what project legal stuff or infrastructure am i missing in that table?
12:48 PM skunkworks: the wiki i assume is on dreamhost too?
01:25 PM jepler: skunkworks: yes
01:28 PM jepler: it wouldn't hurt for it to be a separate item
01:35 PM jepler: seb_kuzminsky: https://emergent.unpythonic.net/files/sandbox/linux-sublicense-grant-orig.txt
01:35 PM jepler: seb_kuzminsky: we should put that somewhere!
01:38 PM jepler: https://emergent.unpythonic.net/files/sandbox/linux-sublicense-request-orig.txt might also be worth preserving
02:12 PM jepler: .. no new crashes or hangs from afl-fuzz on rs274 since I (locally) fixed 3 different identified hangs.
02:22 PM cradek: that's pretty amazing for something written in C a long time ago
02:22 PM jepler: it may still be the case that afl-fuzz isn't finding "interesting" Cases
02:23 PM jepler: #O<sub>=3#<sub>=3#g20sub>=3#b
02:25 PM jepler: t0 m0g49
02:25 PM jepler: hm I'm not sure about this error message
02:25 PM jepler: READ => #O<sub>=3#<sub>=3#g20sub>=3#b
02:25 PM jepler: Named parameter not terminated
02:27 PM cradek: the error is #g ?
02:28 PM cradek: yeah
02:28 PM jepler: READ => #O<sub>=3
02:28 PM jepler: READ =>
02:28 PM jepler: wait this is OK syntax?
02:28 PM cradek: I think that's right enough
02:29 PM cradek: I thought < had to follow #
02:30 PM cradek: I think #a<b>=1 is an uncaught error
02:30 PM cradek: just #a<b> says "Equal sign missing in paramter setting"
02:30 PM cradek: so it thinks #a<b> is a parameter
02:31 PM jepler: #b=3 is bad, #b>=3 is good
02:31 PM jepler: G[#b>] is an error as it should be
02:37 PM cradek: if((line[*counter] == '<') || isalpha(line[*(counter)]))
02:38 PM cradek: I think this is just a bug
02:38 PM cradek: a mistake from before he decided on requiring <...>
02:40 PM cradek: the character after # can be any isalpha and it still works as if that character is <
02:40 PM cradek: #Qabc>=1
02:40 PM cradek: #<abc>=1
02:40 PM cradek: these are equivalent
02:41 PM cradek: so #O<sub>=3 defined the named param "<sub" to be 3 (yes < are allowed in param names)
02:51 PM seb_kuzminsky: the wiki us UseModWiki 1.0.5 (aka 1.05?), released August 28, 2009
02:51 PM seb_kuzminsky: UseModWiki appears to still be actively maintained, 1.21 was released December 1, 2017
02:55 PM seb_kuzminsky: i added the wiki to that table
03:06 PM Chris_Morley: andypugh :that widget was made by Norbert but I can look at it, when I get a chance no problem.
03:15 PM hazzy-dev: Chris_Morley: I think I fixed it. I just replaced 48 with gtk.ICON_SIZE_LARGE_TOOLBAR, which is also 48 px
03:20 PM Chris_Morley: excellent, that was my guess too.
03:23 PM hazzy-dev: I can create a PR if you think what I did is sufficient.
03:23 PM hazzy-dev: This is my commit: https://github.com/KurtJacobson/linuxcnc/commit/ba2b827
04:25 PM Chris_Morley: hazzy-dev :i'm not very good with git-hub so just pushed the fix. thanks though. another bug squashed!
04:27 PM hazzy-dev: I think the fix should go in 2.7 and then be merged up to master??
04:28 PM Chris_Morley: ha yes of course - damn
04:28 PM hazzy-dev: :)
04:31 PM hazzy-dev: I guess I should have just created a PR, then you could have just hit the big green button :D
04:31 PM hazzy-dev: I still can, it will take like 3s if it is easier
04:33 PM Chris_Morley: im cherry picking it now
04:34 PM hazzy-dev: oops, OK
04:39 PM Chris_Morley: ok - I caused enough trouble for today - time to get ready to work for real...ttyl
04:42 PM hazzy-dev: Sorry, I was trying to make things easier for you, not harder!
04:45 PM jepler: it's hard to know what another person will prefer. When it comes to me, I'd rather get a PR than an issue, if the reporter thinks they have a clear idea what needs to be changed. Sometimes that can go in right away, and if not it's a much better starting point for discussion!
04:57 PM hazzy-dev: jepler: I agree, fairnorth had everything he needed for a good PR, which would have saved several people some time, and he would have had the pleasure of contributing, but I guess he may not have been comfortable doing that. He posted on the forum to begin with after all.
04:58 PM jepler: "not everyone has X years professional and Y hears hobby experience with git" is something I can't consistently remember
04:59 PM jepler: wow 2009ish, it'll be 10 years soon
05:14 PM seb_kuzminsky: jepler: i added the license number and the original request & reply emails that you linked
05:14 PM seb_kuzminsky: i didn't fix the horrible html-email formatting
05:18 PM Chris_Morley: hazzy-dev : yes - I just didn't see your work until i had already done the same lol - i'm working nights so probably shoulda just stayed out of it :)
06:52 PM jepler: no more interesting findings from afl-fuzz http://media.unpythonic.net/emergent-files/sandbox/rs274-findings3/graphs/index.html
07:02 PM jepler: I didn't look through all 2000+ files that afl-fuzz created for itself, mind you
07:03 PM jepler: but no more hangs, and never any crashes
07:30 PM andypugh: How did you even find that bug?
07:31 PM cradek: Jeff was fuzzing the interpreter and happened to notice that weird input
07:31 PM cradek: http://searchsecurity.techtarget.com/definition/fuzz-testing
07:32 PM cradek: did you know you could have a variable named <<<? like this: #<<<<>=1
07:32 PM cradek: that's probably not a feature
07:33 PM jepler: READ => #<<>=7
07:33 PM jepler: READ => G0Z#<<>
07:33 PM jepler: 6 N..... STRAIGHT_TRAVERSE(0.0000, 0.0000, 7.0000, 0.0000, 0.0000, 0.0000)
07:33 PM jepler: you can ever successfully refer to that one!
07:34 PM cradek: I think you can refer to all of them once you understand the bug (first character is assumed to be <)
07:34 PM jepler: READ => #a>=9
07:34 PM cradek: #abc>=1
07:34 PM jepler: READ => G0Z#<>
07:34 PM jepler: 7 N..... STRAIGHT_TRAVERSE(0.0000, 0.0000, 9.0000, 0.0000, 0.0000, 0.0000)
07:34 PM andypugh: Well, that’s not a bug then. It’s an undocumented feature for when you run out of all possible alphanumeric variables.
07:34 PM jepler: oh so you can
07:35 PM cradek: oof the empty variable name!?
07:35 PM cradek: wild
07:35 PM cradek: ... g0 #<bc> (goes to 1)
08:12 PM hazzy-dev: jepler: https://imgs.xkcd.com/comics/git.png
08:20 PM jepler: hazzy-dev: you can totally get day to day work done with git if you care to learn it to that level and no further.
08:27 PM hazzy-dev: Absolutely! Git is one of those things that is very easy to use at a level that is fine maybe 95% of the time, but then it starts to get a little mysterious, to me at least :)