IRC Logs for #crux Friday, 2019-03-08

*** timcowchip has quit IRC00:02
TimB_jaeger: https://gist.github.com/TimB87/699a06edb7466f4034c59ea8fb47688200:24
TimB_it includes a little readme with what changed00:25
ryuodlcusa: not really, no. i don't remember that really.00:32
TimB_I have updated it once more, I added rm -fr $PKG/usr/share/doc for gvfs00:34
TimB_ryuo: now I'll have a look at the openjdk8 patch00:34
TimB_so, in openjdk8, I have hotspot as a tar.xz in $SRC, but it is not unpacked?00:44
TimB_ok, so I'll manipulate the tar file00:45
ryuoTimB_: Oh. I thought the script unpacked it... i guess it needs to be hard-patched.00:46
TimB_now the makefile checks the checksum for the hotspot tar :D00:47
ryuomd5sum -um00:47
TimB_I'll give that a look myself, I should be able to do that00:47
ryuoerr00:47
ryuopkgmk -um00:47
TimB_no, the makefile for openjdk00:47
ryuoOh.00:48
ryuoI see.00:48
TimB_ERROR: No up-to-date OpenJDK HotSpot zip available o:) I wonder why00:48
ryuoACTION facepalms.00:48
ryuowell, the alternative is to tell it to skip version checks.00:49
TimB_oh, if there is a switch for that, sure00:49
ryuoi noticed one when i was poking around.00:50
ryuo# Disable this check by setting DISABLE_HOTSPOT_OS_VERSION_CHECK=ok.00:51
TimB_indeed --with-hotspot-checksum00:52
TimB_seems to work, thx!00:53
TimB_jaeger: I just noticed that the patch is not complete...00:55
dlcusa@ryuo, you can see the doc at http://dlc.casita.net/~dlc/CRUX_3.4_Commonwealth_Manual_CCIA_0.0.pdf (still WIP but some of the detail may be interesting to you).00:57
TimB_gist is updated00:58
*** john_cephalopoda has joined #crux01:01
ryuodlcusa: i see.01:08
ryuodlcusa: interesting. i'm trying to redesign pkgutils and preserve most if not all existing behavior.01:09
ryuodlcusa: though i might find it doable to expand some to do more.01:09
ryuopkginfo isn't pragmatic to extend, but pkgrm/pkgadd could be extended to operate on multiple packages at once.01:10
ryuothat could have some performance gains for batch operations.01:10
cruxbot[contrib.git/3.4]: ntp: 4.2.8p11 -> 4.2.8p1301:15
dlcusaryuo: I perceived you want it to merely work alike for the first release.01:16
ryuodlcusa: uh, yea. i doubt i'd have much success with convincing people to switch if it didn't.01:17
ryuoit's not like when Liden was designing CRUX. we don't have the luxury of discarding backwards compatibility all at once.01:17
dlcusaRoger that.01:17
ryuoi just wanted to open the tools up to other kinds of frontends perhaps.01:18
ryuoreduced duplication of work.01:18
ryuomost C programs have to reinvent the database reading code at the very least.01:18
dlcusaNot to mention having maintainable toolsets.01:18
ryuoso, may as well reduce how much is in the frontends except the bits that pretty much only useful to it.01:19
ryuoplus, a shared library means that the lock is also shared.01:19
ryuocurrently only pkgutils is even aware of the lock.01:19
ryuomy revdep didn't even know it existed. it's a read only tool though so it didn't harm much.01:20
dlcusaThe lock only serializes within the local host, not across platforms.01:20
ryuothat too could be an issue, but isn't so much for how CRUX currently works.01:20
ryuoit's only really an issue for networked filesystems.01:21
ryuohm.01:21
ryuoi'll have to see how apt does locks.01:21
dlcusaThe lock is to keep pkgutils from stepping on itself.  Most CRUX boxes only have one admin.01:21
ryuoindeed, but it's still important to have locks.01:22
ryuowe could perhaps change the locking mechanic.01:22
dlcusaEspecially for what I'm attempting.01:22
Anselmoalso even if there is only one admin,that's no reason you cant step on your self :P01:22
ryuoi may have to resort to a lock file like shells normally do.01:23
ryuothis would signify the existence of a write lock.. though i have no way to support read locks with such a system.01:24
dlcusaThe lock prevents that (but prt-get doesn't use the lock to the best of my knowledge).01:24
ryuoindeed it doesn't, but it uses pkgadd to install and such so it does indirectly.01:24
Anselmowell, its no problem to be building 10 things at once. you just dont want to be logging installed files 10 times at once01:24
dlcusaThe point is prt-get's resources are not serialized, so you don't want to run more than one at a time.01:25
ryuodlcusa: indeed. i'm working on the first layer of that, pkgutils.01:26
ryuois the advisory lock too primitive?01:26
ryuoor should we use something else?01:26
ryuohm.01:26
dlcusaHardly, as long as everybody obeys the traffic light.01:26
dlcusaCRUX software maintenance is a fairly closed system.01:27
ryuodlcusa: but... serialized? i was told that's just about data conversion to a data format. what you're describing is properly synchronrized resource accesses.01:27
ryuohm.01:28
ryuodlcusa: question. can you write actual C code?01:28
dlcusaMy model uses several layers of locking but there's still advisory locks--all done in shell scripts.01:28
ryuoyes, that's a kernel limitation.01:29
ryuothe kernel has mandatory locks, but it's broken, so it's not advised to use it.01:29
ryuocooperating locks is about the only solution i know of that currently work.01:30
ryuo(advisory as they call them)01:30
dlcusaThere's no need to lock individual files, only abstractions that represent a well-defined group of resources.01:30
ryuolike locking the whole pkgutils directory?01:30
dlcusaPrecisely.01:30
ryuoi can see why Liden chose that. there's various things you need to do under there but it's hard to say what in advance.01:30
ryuoso it makes more sense just to lock the whole thing during writes.01:31
ryuodlcusa: i'm also working on fixing some implementation inefficiencies i discovered.01:32
ryuo2 really.01:32
dlcusaYes, and maximize concurency for non-modifying accesses (so the world doesn't change on you enexpectedly).01:32
ryuopkgadd/pkgrm don't support batch operations. you have to do a fork+exec for each one.01:32
dlcusaThat's more prt-get's cup of tea.01:32
ryuoindeed... but i see no harm in expanding them to support 1 or more files at once.01:33
ryuoit's still compatible with the old approach.01:33
dlcusaAs long as they do it safely--hence the layers of locking.  Read the serialization section of my doc.01:33
ryuoi'll look at it more in awhile.01:34
dlcusaNo real rush.01:34
ryuoeventually i'll see what i can do about prt-get.01:34
ryuodlcusa: also, the other inefficiency i discovered is that pkgadd recompiles the REs defined in /etc/pkgadd.conf every time it does a check of them.01:35
ryuowhich makes RE execution more expensive.01:36
ryuocaching the compiled engine would reduce that cost.01:36
dlcusaDetecting still-valid compiles could be a significant win.01:37
dlcusaI'm not sure they're used that much but I haven't analyzed it.01:38
ryuodlcusa: actually, they're all valid. comparisons don't modify the engine.01:38
RomsterTimB_, pkgmk -kw ; then look for the extracted source on where to patch?01:38
ryuonot sure why it was designed that way.01:38
TimB_Romster: it's sorted out already, but the tar gets extracted after the check01:44
TimB_so I had to patch the tar :)01:45
*** xor29ah has joined #crux01:45
TimB_built fine too, so thx ryuo01:45
dlcusaryuo, I learned C in 1989.01:45
ryuoi see.01:45
ryuook.. i'll assume that means you know more than I do.01:46
dlcusaMaybe.  There's also pockets of ignorance.01:46
dlcusas/also/always/01:46
ryuosome features i still consider useless... namely digraphs01:47
dlcusaDitto.01:47
ryuoC11 added very little of interest, but _Generic and _Static_assert have promise.01:47
ryuoC99 was a lot more interesting.01:48
ryuocompound literals, loop initializers...01:48
dlcusaHaven't looked at those nuances.  These days I mostly script.01:48
ryuoah.01:48
RomsterTimB_, ryuo be sure to add a flyspray bug report on openjdk8 with the solution01:48
dlcusaI can likely QA your code.01:49
ryuowell, what i have is a bit messy still. i'm still figuring out how to package pkgutils into a library.01:49
ryuohttps://bitbucket.org/ryuo/pkgutils/src01:50
ryuothe git commits are kinda uninformative, but it's a WIP. i generally let loose on experiments.01:50
Romsterpedja, actually not sure why i did the CFLAGS CXXFLAGS to cmake. might of been from another port i looked at or a cmake bug back ages ago.01:50
*** xor29ah has quit IRC01:52
ryuoi've largely finished the frontend stub though.01:53
ryuoi thought it would help me figure out how better to design the library.01:53
TimB_Romster: I don't have a flyspray account yet, but I can do that tomorrow01:54
dlcusaryuo, why do you think a library is indicated?01:54
ryuodlcusa: ... why do i think a library is necessary?01:54
*** xor29ah has joined #crux01:54
TimB_I'll try and get some more sleep today, so I'm off for today. laters01:54
Romsternini TimB_01:55
dlcusaCan you say what's the driver for that?01:55
ryuodlcusa: Well, code reuse mostly. otherwise there's not much point.01:56
dlcusaWhat needs to reuse it?01:56
ryuorevdep and prt-get both need to use the database.01:56
ryuofor some reason or another.01:56
ryuoproviding it as a library provides another option than just relyng on the CLI.01:56
ryuoefficiency gains to some degree. shell is ridiculiously inefficient.01:57
dlcusaYou've found prt-pkg/revdep doing that or believe they need to?01:57
ryuodlcusa: I know revdep needs database read access. I had to reinvent a reader because no library exists.01:58
dlcusaLet me remind you of the CRUX prime directive--there must be a strong reason to make something more complicated.01:58
ryuodlcusa: well, it would also mean that anything using the library will also end up using the same locking mechanism as pkgutils.01:59
ryuohm.02:00
dlcusaIt might.  The lock is to serialize data accesses, not neccessarily the code accessing the data.02:01
ryuowell, the lock is useless if nothing else is aware of it...02:01
ryuohm.02:02
dlcusaI agree itwould be more convienient if non-pkgutils software needs to invoke those functions.  I'm just saying show the need.02:02
ryuowell, that's part of it. it also opens up the possibility of making the library code accessible to languages other than C or C++.02:03
dlcusarevdep currently reads /var/lib/pkg/db?02:04
ryuoyes. it needs to know about packages to do its  job.02:04
*** tilman_ has joined #crux02:04
dlcusaThat's good enough, then.02:04
cruxbot[contrib.git/3.4]: obs-studio: remove cflags/cxxflags to cmake02:04
ryuoit sounds like you're saying i shouldn't bother.02:05
ryuohm.02:06
dlcusaNot at all--I'm just saying show why it needs to make things more complex.02:06
ryuowell... i've spoken to users that wanted better ways to integrate with pkgutils, but ended up doing giant custom hackjobs because there was no library for them to interface with..02:09
ryuoi forget who i spoke to about it.02:09
ryuoit seemed like something people would want to use, so i thought that was enough.02:10
dlcusaI agree--that's good enough reason to build and document the library.02:10
ryuoand i figured i might as well reexamine how pkgutils does things while i'm at it.02:11
ryuosee if there's a better way to do things without breaking behavior.02:11
dlcusaAlso agreed--you might find undetected bugs.02:11
ryuoi'll see if i can integrate it with Perl as the first language to bind it to.02:14
ryuothat's the only real interpreted language that ships with Core.02:14
dlcusaGood.  Gotta step away for awhile.  I'll be around.02:19
*** SiFuh has joined #crux03:08
cruxbot[contrib.git/3.4]: python3-urllib3: fix path for new python versions03:15
Romsteri for one would like a daemon for probably called pkgdl that can download all the sources in the background as prt-get is building. and it could do stuff like download new versions of installed ports after it sees new files after ports -u03:20
*** Stealth has joined #crux03:25
*** _________mavric6 has joined #crux03:44
ryuoRomster: one thing at a time, Romster. I think it's best to redesign the foundation first before going off on new ideas for pkgmk or ports.03:52
ryuonot many here are fluent with C, so it's not going to happen quickly.03:53
tilmanryuo: https://crux.nu/gitweb/?p=tools/pkgutils.git;a=shortlog;h=refs/heads/pkgutils606:04
tilmanryuo: that's my ancient attempt at lib-ifying pkgutils and rewriting the whole thing in c. doesn't fully work though06:04
ryuoa BST implementation... interesting. if i needed that, i'd use the POSIX one though...06:13
ryuotilman: thanks. this may be useful for architecture design.06:15
ryuotilman: i was working my way backwards through the whole pkgutils as it stands right now.06:15
*** Abdullah7 has joined #crux06:35
*** Abdullah has joined #crux06:50
*** Abdullah has quit IRC06:51
*** Abdullah has joined #crux06:51
ryuowow. this has some good ideas.07:07
ryuoi'd forgotten how i could use openat(). that's a better idea than storing the string form of the root path.07:08
*** guido_rokepo has joined #crux08:37
*** abenz has joined #crux08:42
*** Abdullah has quit IRC09:11
*** admincomplex has joined #crux09:16
TimB_currently about to submit a flypsray task, do I assign it to j_v directly or is somebody else doing that?09:56
TimB_regarding openjdk8 on linux 5.009:59
*** joacim has joined #crux10:03
*** onodera has joined #crux10:28
*** onodera has quit IRC10:28
*** blueness has joined #crux11:13
*** abenz has quit IRC11:24
*** jue has joined #crux11:35
cruxbot[opt.git/3.4]: python: update to 2.7.1611:39
cruxbot[opt.git/3.4]: scite: update to 4.1.411:39
cruxbot[opt.git/3.4]: dovecot: update to 2.3.511:39
*** xor29ah has quit IRC11:41
*** xor29ah has joined #crux11:44
cruxbot[opt.git/3.4]: gvim: update to 8.1.099911:47
cruxbot[core.git/3.4]: vim: update to to 8.1.099911:47
cruxbot[contrib.git/3.4]: monit: update to 5.25.311:49
*** xor29ah has quit IRC11:51
*** xor29ah has joined #crux11:52
*** abenz has joined #crux12:24
cruxbot[opt.git/3.4]: [php] update to 7.2.16"12:52
john_cephalopodaNew Laptop, New CRUX!13:12
ryuojohn_cephalopoda: lol.13:22
john_cephalopodaryuo: I got 4 CRUX devices now. My desktop, my old laptop [screen broken rn], my new laptop and my RasPi server.13:36
ryuoi see.13:36
john_cephalopodaIt's just the best OS :D13:40
ryuohm.13:43
Anselmocrux must take over all the machines13:51
Anselmoone day we will rule the earth13:51
dlcusaGood.  Gotta step away for awhile.  I'll be around.14:01
*** SiFuh has quit IRC14:03
*** joacim has quit IRC14:04
*** SiFuh has joined #crux14:05
*** joacim has joined #crux15:30
john_cephalopodaOoooh *A*16:15
john_cephalopodaIt compiles so fast, I can't even read the compiler messages any more!16:15
john_cephalopodaSSD and a new Intel... I think my laptop is faster than my desktop :D16:15
ryuojohn's Desktop: I thought he had something special! :( Good bye. -runs blkdiscard on its internal SSD-16:19
ryuowe*16:19
ryuolol16:19
jaegerWhat did you get?16:27
john_cephalopodaMy computer is i7-4770 with 12GB RAM and a 1TB HDD.16:31
ryuoquad core?16:31
john_cephalopodaMy new laptop is i7-8550U with 32GB RAM, a 2TB HDD and a 1TB SDD.16:32
ryuooh.16:32
john_cephalopodaBoth quadcore with 8 threads.16:32
Anselmoo16:32
jaegerThe i7-8550u and i5-8250u are beastly for laptop CPUs. I have one of the latter16:32
jaegerAnd it's nice to have a 4c/8t i516:33
ryuofrom what i've heard, intel's mobile i7s only recently caught up to their ivy bridge counter parts.16:33
ryuothey were slower for a few generations.16:34
jaegerMobile CPUs are always slower than their desktop counterparts to some extent16:34
jaegerTo be expected, given the cooling constraints16:34
john_cephalopodaAnd the power constraints.16:34
john_cephalopodaDon't want to have my battery empty in 10 minutes :D16:35
jaegerindeed16:35
AnselmoI want mine emptied in five ~16:36
jaegerfor you there are dual GPU laptops :D16:41
TimB_batteries? https://image.made-in-china.com/202f0j00zYMfNdtUHPkj/2018-Promotion-40000mAh-Power-Bank-Car-Jump-Starter-Laptop-Battery.jpg16:41
kyaaaACTION  surviving with  ahtlon II x3 435 / via c7 6300 / turion II M500 / j190016:43
jaegerIf they do what you want them to, that's enough :)16:48
ryuoC7!?16:52
ryuoI recall it makes molasses seem fast.16:53
john_cephalopodakyaaa: Are you running CRUX on it?17:07
*** guido_rokepo has quit IRC17:08
jaegerTimB_: starting to look at the MATE updates now. I'd prefer to leave the mate metaport alone but I've no objection to having other metaports that install subsets or whatever17:08
TimB_jaeger: I think I'm fine with either, I just like the idea of giving users a choice about how much they want to install so I made this split :) then again, it's easy enough to achieve individually17:15
TimB_and to be honest, I was looking at the archwiki when I first got the idea.. so yeah :D17:17
jaegerAll good. I used to do something similar with the gnome ports years ago. had 'gnome' and 'gnome-light' I think17:17
TimB_I'm somewhat nervous over here, updated to the early 3.5 branch in hopes to get rid of my odd compiler error with i3, and I manually bumped glibc. If all fails, this is gonna end in a fresh install.. which would be fair, this is maybe 8 years old and survived different mainboards and harddrives :D17:25
ryuojaeger: before GNOME went over to the dark side?17:26
jaegerbefore 3, if that's what you mean17:27
jaegerTimB_: you can always recover from an install ISO without doing a complete reinstall, if need be17:27
TimB_jaeger: yeah, but at some point I already installed everything from your updated iso and it didn't help, so yeah, this is mostly an act of desperation17:29
TimB_maybe it works, which would be cool17:29
TimB_maybe it won't.. idc anymore. a clean slate can be nice, and it's not like I'd abandon all config files17:29
john_cephalopodaWhat are you using for reading RSS/Atom feeds?17:37
TimB_over here, I use the integrated function in nextcloud17:38
TimB_I hoped there was a good cli alternative, but newsbeuter (I think?) needs rust, so thx but not thx17:38
TimB_are you looking for a gui or cli reader?17:39
john_cephalopodaThere's "newsboat", which doesn't seem to depend on rust.17:39
john_cephalopodaBut I'd like to have something graphical.17:39
TimB_newsboat isn't updated since 2013 or something?17:39
TimB_I dunno, rust seems like a overkill but 2013 seems pretty dead17:40
john_cephalopodaLast release is December 2018, according to https://newsboat.org/17:40
john_cephalopodaOh, wait, actually Feb 2019.17:40
TimB_oh, it's even feb17:40
ryuoDHMO is dangerous. It kills people every year.17:40
TimB_haven't seen that17:40
ryuo>.>17:40
TimB_I have used rssowl in the past, which was okay17:41
TimB_comes with a thunderbird'ish gui17:42
juenewboat 2.14 depends on rust, that's the reason why our opt port is still at 2.1317:42
ryuoseems newboat finally rusted.17:43
ryuoACTION coughs.17:43
TimB_old iron already?17:43
TimB_newsbeuter was last released in feb 201517:44
TimB_well then, I'll grab opt/newsboat ^^17:45
cruxbot[opt.git/3.4]: mailx: update to 14.9.1317:45
jaegergo away, rust!17:46
john_cephalopoda"Tiny Tiny RSS is a free open source web-based RSS/Atom reader and aggregator, powered by AJAX. To host it, you need to setup a LEMP or LAMP stack on your system."17:47
john_cephalopodaWTF???17:47
jaegerI've used that one in the past. It creates a webpage17:47
jaegerhence the ajax/lemp/lamp need17:47
john_cephalopodaIt's a program that is supposed to get XML files from the internet and tell me when one changed.17:47
TimB_but css?17:48
john_cephalopodaNeeding a full-blown webserver environment on a computer just for that sounds pretty overpowered.17:48
john_cephalopodasfeed can output HTML, too, but it's doing it statically.17:48
TimB_stuff just tends to get crazy nowadays17:49
ryuohttps://codezen.org/canto-ng/news/17:49
TimB_last time I was amazed was hyper, an electron based terminal17:49
ryuoanything electron based is... why?17:49
TimB_I mean, to go along a lamp stack17:50
TimB_again: css? ¯\_(ツ)_/¯17:50
ryuohttp://quiterss.org/17:50
ryuosome RSS readers that don't use rust.17:50
jaegerjohn_cephalopoda: the idea with it is that you can read it from any browser, not requiring an RSS reader software17:52
john_cephalopodajaeger: Yeah, but you could do that by generating simple html pages (maybe with some js that updates them from time to time).17:52
jaegerACTION shrugs17:53
john_cephalopodaUsing the apache server to display feeds doesn't seem like a good idea.17:53
jaegerwell, if you're going to read them from somewhere other than where the software runs, you need a web server17:53
joacimlemp is engine ex?17:59
TimB_joacim: iirc yes18:00
*** Stealth has quit IRC18:00
TimB_tiny tiny rss should be possible with nginx and mysql18:00
joacimi'm more of a wimp person myself18:01
joacimdont like the forced tab view of quiterss18:05
*** Stealth has joined #crux18:05
joacimthink overall liferea suits me better18:05
joacimexcept it has this annoying thing where it doesnt remember the size of the window since the last time i used it18:05
joacimhave to specify those preferences with fixed values in gconf18:05
john_cephalopodaI'm trying out snownews now.18:21
john_cephalopodaUhrg, of course that depends on a ton of Perl...18:23
joacimi think it is snownews i have on my crux laptop18:32
kyaaa[14:07:32] < john_cephalopoda> kyaaa: Are you running CRUX on it? <--- of course! via c7 is only 32b and very slow... full 3 days to compile everything that I need (minimal xorg and deps for palemoon-bin)19:57
TimB_jesus, as if I did nothing. same compiling error on i319:59
joacimthink a via something would be f un to have19:59
*** timcowchip has joined #crux19:59
joacimalready got a couple of nice geode boards =)19:59
*** timcowchip has quit IRC20:02
*** timcowchip has joined #crux20:03
Anselmoah. I should fix my raspi again.20:03
*** timcowchip has quit IRC20:06
*** timcowchip has joined #crux20:10
timcowchiphttps://www.youtube.com/watch?v=aEpv5XTvTMM20:10
timcowchiphttps://www.foxnews.com/entertainment/airwolf-star-jan-michael-vincent-dead-report20:11
pedja'Cons: No RGB' :)20:25
timcowchipno RGB ?20:26
pedjafrom the review of the one of the new 1660ti gpu20:27
timcowchipso "urxvt*background: rgba:0000/0000/0200/c800" won't work?20:28
jaegerapparently not on that 1660ti20:29
timcowchipTIMB: https://github.com/nivit/polybar-module-news20:36
TimB_timcowchip: I don't use polybar20:39
TimB_but looks nice!20:39
timcowchipok20:44
TimB_I mean, maybe you can make use of it for other bars, i3bar seems flexible enough20:44
TimB_I bookmarked it ;)20:44
timcowchiphttps://pasteboard.co/I4wfAoD.png20:46
TimB_does it support scrolling through news in some form?20:54
*** timcowch1 has joined #crux20:57
*** timcowchip has quit IRC20:59
timcowch1I don't think it scrolls manually21:00
*** timcowch1 has quit IRC21:00
*** timcowchip has joined #crux21:01
timcowchipit scrolls by itself21:01
timcowchipand can be made clickable, opening the feed in a browser21:02
TimB_I just installed feedparser and tested it, should be fairly easy to use with another bar, I'll give it a try later21:02
timcowchipah...cool :)21:03
TimB_currently partitioning my sd card for that raspberry pi :)21:17
TimB_this didn't went to bad21:42
TimB_timcowchip: https://nullvoid.de/crux/polybar-module-news-i3bar.png22:46
TimB_needs fixed width tho, or needs to be the very last item maybe? Or have conky paste the last 10 news? I dunno22:50
TimB_s/last/first/g22:50
TimB_let's see about making it clickable anyway, should be easy22:52
john_cephalopodaHeh... It's funny how all my computers are configured extremely similarly, so when I switch computers I don't really have any transition in how they operate.23:02
TimB_same here23:06
AnselmoI just use my laptop to reach my desktop, so when I use it it litterally _is_ the same computer ~23:07
john_cephalopodaWell, my IRC is running on my RasPi which I ssh into, so part of what I do with my computer is shared between all computers.23:11
TimB_but it can feel kind of stupid when you then ssh back to the machine you are sitting at, I can tell :D23:11
john_cephalopodaI even ssh into that server with my phone to IRC.23:11
john_cephalopodaTimB_: Lol. I only got sshd running on one machine.23:12
Anselmoah, well, I guess I also have things on different machines. . .but usually accessed through my main machine too x-x23:12
TimB_so that raspberry is updating as it seems23:12
TimB_got zsh and tmux running, nice23:12
Anselmoah, hrm23:13
TimB_is there anything I need to keep in mind that I haven't read at crux-arm.nu? Do I need to update the initramfs when I update the kernel?23:13
jaegerif you're using a kernel config with an initramfs, probably... but if not, no23:22
TimB_oh, haven't checked if there actually is one, I assumed there was one. ok23:26
jaegerTimB_: do you know if mate-icon-theme-faenza is deprecated now? Or did they just not release a new one for some reason and the 1.20 version is fine?23:41
TimB_jaeger: I checked because I thought the same, no news about it, guess there just isn't a new version for it23:48
TimB_latest commit was 02/201823:49
jaegerThe 1.20 version still seems ok so I'll leave it alone for now, just wondered23:53
TimB_mate-session-manager is already patched btw23:54
jaegerin git, yeah23:54
jaegerbut I didn't see a new release tarball yet23:54
jaegerI've patched it locally23:55
TimB_I created a standalone patch23:55
TimB_isn't it included? :^23:55
jaegerin git, yes23:55
jaegernot in the release tarball23:55
TimB_right23:55
TimB_I just thought that my diff included that patch as an extra source file, but I might have made an mistake23:56
jaegerOh. Your diff does have one but I'd already done it locally too23:56
jaegerjust grabbed it straight from the github issue23:56
jaegerer, commit23:56
TimB_ah alright23:56
jaegerif you didn't know about that, a nifty github feature is adding .patch to the end of a commit23:56
jaegerhttps://github.com/mate-desktop/mate-session-manager/commit/7f0a938467e4e3d57c68af1caa146a664458a911.patch23:56
TimB_I tried that for a minute, didn't find the button and then created a diff myself :D23:56
jaegerah, ok23:56
TimB_ah, nice!23:56
TimB_very cool23:57
TimB_I am still trying and learn more git on the go23:57
jaegerI'm in the process of verifying clean builds for all the mate package updates. Going to push those through first, then will visit the ancillary stuff23:57
jaegerlike udisks2, etc.23:57
TimB_branching and stuff hasn't been my kind of stuff up until now ^^23:57
jaegercan be very handy to know :)23:58
jaegerI use a bunch of git for work as well as crux so good practice23:58
TimB_that's what I have in the back of my mind too :)23:58
TimB_get a good knowledge of a great tool that is utilised a lot all over the world23:59
pedja'when in doubt, nuke and clone again'23:59
TimB_:D23:59

Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!