mozilla :: #rust-community

18 Apr 2017
11:42skadetheJPster: are you coming to rustfest kyiv?
12:21theJPsterskade: I'm afraid not.
12:21skadetheJPster: hm, anyone else with some embedded experience?
12:22theJPsterI'll go ask
12:22skadeI have these two Hails, but I really can't do much more with them then flash them :)
12:22skade(guess what the weather was when I was driving home with them in my backpack)
12:23theJPsterOh, the boards for TOck
12:23skadeI haggled and got two preview devices
12:24skadewe're also thinking about buying a couple for europe
12:26theJPsterI need to take some time to look at Tock. I worry about fragmentation, and having little bits of embedded support here and there that aren't really joined up.
12:27skadein which sense?
12:27theJPsterThey have a bunch of drivers for the SAM4L, but the drivers aren't standalone - they require Tock.
12:27theJPsterI have a bunch of drivers for the LM4F120, but I need to look at the work required to use them with Tock.
12:28theJPsterIt just feels like we should be able to agree something, and write all the drivers once for everyone to use.
12:28theJPsterMaybe that's just not possible.
12:30skadeah, I see
12:30skadehm, well, not far enough into that, but it feels like some abstraction might solve that?
12:31skadelike, define a set of traits for common tasks that can be OS-specific, but not really of interest for the driver
12:32theJPsterIt feels like Rust has the bits needed to construct a solution (which C totally doesn't have). I just don't think anyone's got that far yet.
12:32skadeI'm quite happy with trait-only crates in some of my projects
12:32skadeit is java-esque and slow, though, and hard to test
12:33skadebecause a test basically needs to be done by having two implementations
12:33theJPsterI have, but it's lacking traction
12:33skadethe thing I hate is that the only thing where I haven't grokked yet how traction works, it's these efforts
12:34skadeeveryone wants them, no one wants to work on them
12:34skadealthough, I do have a hunch
12:34theJPsterI'd love to work on them. But that's a hard sell at the office.
12:34* skade looks left and right, then silently whispers: "standards commitees"
12:35theJPsterThere was a blog post by a colleague which basically said, Rust is like C was in the mid-80s.
12:36theJPsterIt'll take ISO (and maybe a parallel implementation) before it really takes off.
12:37skadeI don't really see that
12:37skadelike, languages do just fine without ISO nowadays
12:38skadebut working groups for common implementations to agree upon, not so much
12:38skadeC in the 80s was also C without hyper-collaboration (internet)
12:38skadethat's what a lot of people tend to forget
12:38theJPsterThat is true. And the primary implementations were not open source.
12:39skadeit's no surprise languages like ruby and python only _really_ took off since they tapped into open communities
12:39skadeand they couldn't before around 2000ish
12:39skadeand ruby is ISO-standardised, but the standard isn't worth anything :)
12:47skadetheJPster: an interesting data-point, by the way, is that rust is surprisingly popular at universities.
12:47theJPsterThat's comforting to know.
12:48skadeWell, it means I have to prepare a lecture :(
12:48skadeNever did that
12:49theJPsterI learned Java, Modula-2 and Standard ML. None of which I've used since leaving.
12:49skadeI learned SML and Java and I like both :)
12:49theJPsterActually, Warwick also did a half course in common UNIX tools and Bash. That was useful.
12:49skadecursing puppet on my day-job though :)
12:50theJPsterI sometimes idly think I should start a University course that teaches things I actually use day to day.
12:50skadeon the other hand, i must say that my programming languages prof was A+ and was very good at giving context to the stuff he was teaching
12:50skadeI don't think that's the point of uni :)
12:50theJPsterMores the pity!
12:51skadeWell, the more I program, the more I dig out my old theory courses and comments of my professors in passing.
12:52skadelike, I had a great prof in data retrieval. she was very good at providing a mental framework for assessing data tasks. I'm working with databases, so that's worth 3 proglang courses.
12:55theJPsterMaybe the problem is I spent more time in Engineering than CS.
12:58jdmI regret how much I tried to ignore theory in school now
13:02skadei pretty much failed in electrical engineering
13:02skadepretty much = i failed and had to leave for another uni
13:02skadebest thing that ever happened :)
13:04theJPsterThat's course 101 at my new University. Failing is good.
13:04theJPsterOr at least, not as bad as you think.
13:05theJPsterAlong with CS120: Branching and Merging in Subversion without losing your mind.
13:05theJPsterCS215: The art of coaxing users into generating decent bug reports
13:06theJPsterCS307: Delivering releases even when you're Internet is down.
13:06theJPsterugh. s/you're/your/
13:08carols10centsCS401: You've Inherited A Pile of Poo, Now What?
13:08theJPsterFourth year course - tough one that.
13:09theJPsterCS402: Dealing with a PHB.
13:11skadecarols10cents: that's about where I dropped out :D
13:17skadewell, in the end, I think it was a good thing to have a hard break in the "kindergarten -> school -> civil service -> uni -> career" path, consciously deciding to study the same thing another time definitely influences the way you view the subject
15:00meetingbotThis week's agenda: The meeting is this Weds @ 4PM and 10PM UTC. No minutes found for last week. :( *mournful bleating*
17:08edunhambaa to you too, meetingbot
17:13theJPster4PM *and* 10PM?
17:13theJPsterCan you watch a six hour delayed relay somehow?
17:13edunhamtheJPster: I think that's why we have IRC logs
17:23theJPsterbooyaa[, any luck finding a venue for Rust London?
20:36booyaa[theJPster: not for april, looks like next meetup is 2nd of may
21:00meetingbotThis week's agenda: The meeting is this Weds @ 4PM and 10PM UTC. No minutes found for last week. :( *mournful bleating*
19 Apr 2017
No messages
Last message: 10 days and 5 hours ago