juxt

imre 2017-04-12T10:20:52.977522Z

question to you arch-ers: do you use desktop/window managers (or ecosystems) and which one do you find most pleasant to work with daily?

ul 2017-04-12T10:29:53.107811Z

either i3wm or kde depending on the mood 😃

imre 2017-04-12T10:30:42.121395Z

how's kde these days? I know it used to be bloated as hell a good few years ago

ul 2017-04-12T10:32:33.147866Z

quite nice as of my taste: performant and stable enough, though you have all the stuff like automount/network/energy saving etc. without need for installing and tuning too much by hand

ul 2017-04-12T10:32:51.152378Z

from the UX endpoint, still doesn't catch up with MacOS

ul 2017-04-12T10:33:21.159195Z

KDE5 is much better than 4th was

imre 2017-04-12T10:33:32.161762Z

sounds nice

imre 2017-04-12T10:33:42.164010Z

I'm getting a little fed up with gnome

imre 2017-04-12T10:33:59.167991Z

might give it a shot

dominicm 2017-04-12T10:35:19.187303Z

I use bspwm

jonpither 2017-04-12T20:05:40.781561Z

@dominicm not sure I get https://github.com/juxt/mach/issues/24

jonpither 2017-04-12T20:05:57.788136Z

you are running mach main plan

jonpither 2017-04-12T20:06:26.800283Z

hmm

dominicm 2017-04-12T20:06:40.805830Z

@jonpither plan depends on main. main gets called twice.

dominicm 2017-04-12T20:06:44.807521Z

This might be OK behaviour.

jonpither 2017-04-12T20:06:47.808673Z

you would expect Mach to not run the latter target

jonpither 2017-04-12T20:06:56.811909Z

yeah, unsure what should exactly happen

jonpither 2017-04-12T20:07:00.813454Z

I'll spin up a Make

dominicm 2017-04-12T20:07:11.817895Z

Make doesn't have this problem, because it's based on file freshness.

dominicm 2017-04-12T20:07:17.820581Z

If you have novelty set, this isn't a problem.

jonpither 2017-04-12T20:07:43.830324Z

funnily enough tho, in trying to reproduce, I find a related bug with the new mach/props, the props get evalled per rule invocation (which may again.... be OK

dominicm 2017-04-12T20:07:50.833410Z

Not sure if mach should assume non-freshness in the case of it already being run in the same process.

dominicm 2017-04-12T20:08:00.837160Z

Yeah, you get it I think ¯\(ツ)/¯

dominicm 2017-04-12T20:08:02.838003Z

Might be okay.

dominicm 2017-04-12T20:08:03.838654Z

Might not.

jonpither 2017-04-12T20:08:39.852497Z

yeah. If you get my new bug also, then we have 2 bugs of similar class

dominicm 2017-04-12T20:10:47.903313Z

not got brain power to figure out what mach/props are right now. but it wouldn't surpirse me.

dominicm 2017-04-12T20:11:03.909320Z

I can think of a solution of kind (surrounding putting everything in a delay)

jonpither 2017-04-12T20:11:32.920523Z

yeah

jonpither 2017-04-12T20:11:42.924652Z

working up the courage to tackle your async streaming output thing

jonpither 2017-04-12T20:11:44.925574Z

[jon@juxtxps07 git]$ make foo echo "baz" baz echo "bar" bar [jon@juxtxps07 git]$ make foo baz echo "baz" baz echo "bar" bar make: 'baz' is up to date.

jonpither 2017-04-12T20:12:05.933506Z

foo depends on baz

dominicm 2017-04-12T20:12:10.935419Z

make has inherent freshness.

jonpither 2017-04-12T20:12:12.936182Z

that's quite cool isn't it

jonpither 2017-04-12T20:12:41.947987Z

what do you mean? inherent freshness

dominicm 2017-04-12T20:12:55.953827Z

everything has a file it produces

dominicm 2017-04-12T20:13:03.956654Z

Oh wait

dominicm 2017-04-12T20:13:10.959438Z

you used phony targets, not file writing

dominicm 2017-04-12T20:13:12.960097Z

or did you?

jonpither 2017-04-12T20:13:17.961910Z

phony yeah

jonpither 2017-04-12T20:13:36.969867Z

just echoing "baz" and "bar"

dominicm 2017-04-12T20:13:51.975434Z

so make knows it already ran it

jonpither 2017-04-12T20:14:15.984786Z

yes

jonpither 2017-04-12T20:14:21.987152Z

it keeps track

dominicm 2017-04-12T20:15:03.003956Z

very clever

jonpither 2017-04-12T20:15:19.010189Z

easy enough for us to do tho

jonpither 2017-04-12T20:18:58.096309Z

the bug is slightly worse also, mach foo bar (and foo depends on tar depends on bar, and foo depends on bar), then bar prints out 3 times - no de-duping