libera/#devuan/ Thursday, 2019-10-17

tom_workgnarface, E: The value 'unstable' is invalid for APT::Default-Release as such a release is not available in the sources00:52
tom_workgnarf_08, E: The value 'ceres' is invalid for APT::Default-Release as such a release is not available in the sources00:52
gnarfacetom_work: uh... hmm.   are you able to get any other src packages from there?01:04
tom_workgnarface, have not tried01:05
tom_workany package you have in mind?01:05
gnarfacenothing in particular.  literally any of them should work for this test.01:06
gnarfacei am able to get mysql-workbench from ceres right now using deb.devuan.org01:06
gnarfacemaybe you've got a simple mistake in your sources.list or the repo you're pointing to is being updated at the moment.01:07
gnarfaceyou have the deb-src lines too, right?01:08
gnarfacetry just for science01:09
gnarfaceand use "ceres" instead of "unstable"01:09
gnarfacedeb-src unstable main01:09
gnarfacehere on this line, change this to ceres just to eliminate that variable01:10
tom_workdeb-src ceres main01:10
gnarfacei'm in the US and it was not my experience that pkgmaster was the fastest01:10
tom_work$ apt source -t ceres mysql-workbench01:10
tom_workReading package lists... Done01:10
tom_workE: The value 'ceres' is invalid for APT::Default-Release as such a release is not available in the sources01:10
gnarfaceyou re-ran "apt-get update" after that sources.list change, right?01:11
tom_workwell, apt update01:11
gnarfacestrange issue.  i cannot reproduce it here.01:11
gnarfacewas your ascii install already up-to-date?01:12
tom_workCalculating upgrade... Done01:12
tom_work0 upgraded, 0 newly installed, 0 to remove and 0 not upgraded.01:12
OksanaI know that when hibernate happens, it happens for me with s2disk. According to hibernate scriptlets/conf files, that means that disk.conf is in effect. According to hibernate.conf, that means that tuxonice.conf fails, don't know why (perhaps, tuxonice not installed?). Therefore, I can comment out tuxonice, warning will go away, hibernate will happen one second faster, due to not trying tuxonice, everybody wins - except, I would have l01:14
gnarfaceOksana: your message was cut off at ", I would have"01:17
Oksana- except, I would have liked to install tuxonice instead. But then, I don't know whether tuxonice is better than s2disk, or not.01:18
gnarfaceOksana: i can't help you with the opinions, but you can check what packages are installed with "dpkg -l"01:18
gnarfaceit should be easy to install if you prefer it01:19
gnarfaceif it conflicts with packages already installed, you should be warned that they will be removed first01:19
gnarfacepay attention to what it wants to remove if that is the case.  don't just blindly say "Y" to everything01:19
tom_workgnarface, was there any issue with ASCII 2.0 Beta in 2018 that could effect a modern up-to-date stable ASCII 2.0 install?01:20
tom_workregarding APT01:20
tom_workor, I'm on comcast right now. perhaps comcast is hijacking the connection and having their own cache?01:20
gnarfacetom_work: nothing comes to mind01:20
OksanaI see but it mentions "A kernel patch for TuxOnIce is available separately. " How do I know whether my kernel has a patch for TuxOnIce?01:21
tom_workI wouldn't put it past them because I have personally experience UDP/53 hijacking on comcast's network before01:21
gnarfacetom_work: comcast has been known to do worse, but you should be able to test that theory by just using wget on the https link to the package itself01:21
* Oksana uses standard Devuan Ascii, I think; no idea whether it's relevant01:21
tom_workgnarface, perhaps I should try installing apt-https-transport and switch the repo to https?01:21
tom_workI'm not worried about packages downloaded and install in the past because ounce downloaded they are verified for tampering internally via GnuPG right?01:22
gnarfacetom_work: eh, there's some repos in the round-robin that don't handle it gracefully.  i wouldn't quite do that just yet, but you should at least pick one that is working and try to get the package directly01:22
gnarfacetom_work: yes, the packages should all still be signed01:22
tom_workgnarface, my theory is comcast is hijacking http to devuan repos and serving a stale cache01:23
tom_workas in a non-modified index and debs, but not up to date01:23
tom_workhmm nope. using https and apt-updating still doesn't resolve it01:24
tom_workgnarface, are you completely out of ideas or just not interested in helping me debug this further01:25
tom_workI'd understand if you were tired out, but at this point I'm completely out of further troubleshooting ideas too01:25
gnarfacetom_work: well i am out of ideas, but you've refused the one idea i had01:26
tom_workrefused which one?01:26
gnarfacedownload it with wget or a browser01:26
gnarfacei'm attempting to work deductively here.  eliminate the question of whether the issue is with apt or not01:26
gnarfacei'm no fan of comcast and i agree they could be at fault here01:27
gnarfacebut we don't know yet01:27
gnarfacewhat we do know is you're having a problem nobody else has had yet01:27
gnarfaceand people HAVE had problems with comcast01:27
gnarfacejust not this one, yet01:27
gnarfacebut comcast has been known to do excessively weird things in isolated regions01:28
gnarfacebut i'm wondering if something weird happened to your install instead01:28
gnarfacelike maybe you accidentally got a backported version of apt back in beta and it hasn't been updated since due to a version conflict01:29
gnarfaceor something weird01:29
gnarfacealso, did you ever test trying to get the source of some other package?  any other package?  try like, rxvt-unicode01:29
tom_workok, I'm trying to figure out the right link for it, but mysql-workbench source requires a lot of other helper files so this is hardly an ideal situation01:30
gnarfacethis is just to test the repo itself01:30
gnarfacemake sure the file is there01:30
gnarfacemake sure you're really connecting to the repo01:30
gnarfacewhen you have problems nobody else has, the debugging process has to start by throwing out stale assumptions01:31
tom_workgnarface, test the unstable sources or can i do the regular?01:31
gnarfacetry both01:32
tom_worksorry i mean01:38
tom_workyeah I can download this with wget01:38
gnarfacethe first one 404's on me, the second one works01:38
tom_workfirst link is a typo01:39
gnarfacethe second one works in http or https even01:39
tom_workyes for me too01:39
rrqall pool/DEBIAN will be redirected01:40
gnarfacehang on, lemme test one thing...01:40
gnarfacetom_work: just checked something i should have sanity-checked for you earlier, but it's not your fault here.  the apt in ascii must be too old for even getting ceres src packages.  i was testing from ceres, sorry.01:43
tom_workis this fixable?01:44
gnarfacetom_work: well, you can work around it01:44
tom_workhopefully not by resolving all the dependencies by hand and downloading by hand?01:45
gnarfacenah just put a ceres install in a chroot.  you have a spare 500MB - 1GB or so on disk, right?01:45
gnarfaceyou can debootstrap into a directory01:45
gnarfacethat should still work from ascii to ceres01:46
tom_workI've never debootsrapped ceres before01:46
gnarfaceat least well enough to get src packages01:46
gnarfaceok stand by i can dig up an example command...01:46
gnarfacetom_work: ok try this:  mkdir temp_chroot; debootstrap ceres ./temp_chroot
gnarfaceif it bails out you might have to exclude some buggy packages, it has happened before01:49
tom_workgnarface, do I need to run debootsrap as root yes?01:50
OksanaWhat's the difference between pm-hibernate and hibernate?01:51
gnarfacetom_work: no, but you do need write access to that directory01:51
tom_workE: debootstrap can only run as root01:52
tom_workyou sure?01:52
gnarfaceno i guess not, lemme test01:53
gnarfacei guess it might need root to set permissions right01:53
gnarfacei'm getting it confused with dh tools01:53
gnarface(which can use the "fakeroot" binary)01:54
tom_workI have AppArmor installed so perhaps01:54
tom_workwhich is HMAC not ACL01:54
tom_workso maybe01:54
tom_workit's fine though I have root01:54
tom_workDon't worry about it01:54
gnarfaceyea my bad, it probably needs root just to set permissions right01:55
gnarfacetom_work: lemme know when the debootstrap command finishes.01:59
tom_workit's finished, I'm trying to download the sources now in the chroot02:00
gnarfaceoh, you knew how to chroot already?  good02:00
gnarfacei didn't know if you'd need the bind mounts on /sys, /proc, /dev and /dev/pts or not, and whether you knew how or not02:00
gnarfacejust getting src packages probably doesn't need them02:01
tom_workthat's only necessary for systemd02:01
gnarfacewell, not just systemd.  Steam needs it too :-p02:01
gnarfaceprobably a few other things.  the nvidia binary driver updates, for some reason...02:02
gnarfacedon't get me wrong, i'm not sure everything that needs it deserves it02:02
tom_workbasically only shit software needs that02:02
tom_worksteam is such a mess. you know steam for linux is only i386?02:03
tom_workthe complaining Valve did to Canonical was not over "legacy games". It was mostly because Steam itself is only 32bit02:04
gnarfacethat's false.  but i do know it's an atrocious mishmash of i386 and amd64 components, just like Wine and most of the rest of Windows software these days.  and i don't approve.  but if it was entirely i386, the situation would actually be better.02:04
gnarfaceif it was all one of ANY architecture, the situation would be better.  but the intractable problems it has are largely to do with the fact they couldn't just pick one and stick with it02:05
gnarfacebut that's how Windows does it so that's why02:05
tom_workgnarface, kind of unrelated, but couldn't they just use qemu TCG to run any i386 code instead of requiring all those libs? sorta like how you'd run AMD64 Linux binaries on PPC64 or ARM64?02:06
gnarfaceso you have a Steam client with both i386 and amd64 requirements, running launcher scripts that may be in either i386 or amd64 or a mix of both, in turn launching programs that may be i386 or amd64 and tend to change between updates02:06
tom_work(not talking about qemu-system)02:06
tom_workwhich emulates an entire machine02:06
tom_work>so you have a Steam client with both i386 and amd64 requirements02:07
tom_workgnarface, thank you. I think I can take it come here02:07
gnarfacei'm really not sure about TCG but pretty much across the board the situation with the Steam linux client is apparently due to the fact they're figuring this out as they go along, and only bothering to ask the community for guidance after there's a large outcry of anger02:07
tom_workit's something02:08
tom_workat least they gave us an updated DirectX to Vulkan translation layer for Wine-staging02:08
gnarfaceand even then, sometimes their hands are tied simply because they're publishing binary-only software from 3rd parties.  so they can't just force everything to amd64 at package time like a regular linux distro would.02:08
gnarfacesome of that i386 stuff they're stuck with02:09
gnarfacebelieve me, they would demand their vendors all use amd64 only if they could02:09
tom_workI don't know why they don't just bring the libs they need with them and do a LD_PRELOAD at launch if the host system dosn't provide some obscure lib02:09
gnarfaceas far as i can tell, they tried that but they fucked it up by not understanding the fundamentals of library loading or dependencies.  *more than once* i've had to fix persistent ongoing bugs by using LD_PRELOAD *myself*02:10
gnarfaceand *always* to override the use of their bundled libraries in favor of using the native ones from my distro02:10
gnarfaceso they're trying to do what you suggest, just incompetently02:11
gnarfaceand again, most of the blame seems to lie with 3rd party publishers.  all Valve's own in-house published games run great these days.02:11
tom_workexcept for the client itself02:12
gnarfaceheh, yea the client is a train wreck, but it's getting better :)02:12
gnarfacethey seem to have problems tracking dependencies in a cross-distro compatible way, accurately02:12
gnarfacea lot of the LD_PRELOAD nightmares i have are due to the fact they are bundling their own libraries to get around bugs in OTHER distros.  so the Debian steam package deletes a handful by wildcard at every start-up :-p02:13
gnarfacetom_work: (fyi use the steam package from non-free in the repos.  don't get the one from their website; it's for ubuntu and will give you much more headaches)02:13
tom_workI only use steam on my Gentoo workstation rn02:14
gnarfacethat was another boneheaded move they made.  they based their client and steamos on Debian, but then for some stupid reason decided to baseline all their bundled libraries on some ancient version of Ubuntu that was already obsolete by then.02:14
tom_workI'm planning to migrate it to Devuan as well like the rest of my servers and laptops in the future. probably sometime after beowulf02:15
tom_work>steamos on Debian, but then for some stupid reason decided to baseline all their bundled libraries on some ancient version of Ubuntu02:15
golinuxOksana . . . you should not have be using debian repos ie buster on devuan.  It could put you in a world of hurt02:15
gnarfacetom_work: yea, it seems to me they were making decision by committees that didn't have any prior Linux development experience, and fundamentally didn't understand the nature of library versioning as it pertains to cross-platform, cross-distro compatibility.  now they are just in a long slow clean-up cycle02:16
gnarfacetom_work: they could have stood to take a week off to study the Debian release cycle and just move in tandem with that.  it would have helped them a lot.02:17
gnarfacetom_work: instead they made a bigger mess by trying an approach that was not supposed to need anyone to understand that02:18
buZzi run steam from devuan repo on my buster02:18
gnarface(whoever made that decision was not qualified to make it)02:18
buZzworks fine?02:18
buZzwell, its probably just grabbed from debian02:18
gnarfacebuZz: it should work mostly fine these days.  like i was saying the remaining issues are largely to do with specific games02:18
buZzyeah the whole binaryonly 32bit for life stuff is annoying02:19
tom_workI'd like to think Valve employees are able to make sane long-term business decisions, as they have been around for so long and do a lot against DRM proliferation, price fixing etc; however this going head first with no plan into Linux porting makes me skeptical02:21
tom_workI can't tell if they just wanted to get Linux games fast or wither they just don't care02:21
gnarfacethey wanted to stab Microsoft in the back, and that worked well.02:22
tom_workdidn't M$ stab them first with UWP and Microsoft App Store?02:22
gnarfacei forget exactly, but yes it was all a disagreement over something Microsoft was planning to do with the win10 release.02:23
gnarfaceall the publishers were up in arms about it02:23
gnarfaceValve was the only one big enough and brave enough to step in the line of fire though02:23
buZzmaybe ?02:24
buZzheh > They’ll never completely break it, but will continue to break it until, in five years, people are so fed up that Steam is buggy that the Windows Store seem like an ideal alternative.02:25
gnarfacebuZz: yea, that looks relevant, but i recall the discussions starting a couple years before that02:26
gnarfacebuZz: there was also something about breaking compatibility with old 32-bit games that would not be able to be re-built for 64-bit due to missing ownership or missing sources02:27
gnarfaceand since that is directly the cause of a lot of their support issues, i can see why Microsoft would have wanted do do that, but Microsoft forgot something key; gamers built their empire.02:27
gnarfaceanyway, this is drifting well off-topic, we should continue this in #debianfork02:28
* gnarface admits blame for that02:28
* buZz goes to check if devuan buster's scummvm supports 7th guest02:29
buZzoh, should be, it has 2.0 and its supported since 0.13 :D02:31
tom_workwhat were you saying gnarface?02:42
gnarface(that's all you missed)02:45
tom_workHave you or anyone else here installed and used OpenZFS .8 release yet on Devuan?02:48
* gnarface has not yet tried zfs 03:16
tom_workat all or on devuan?03:36
gnarfaceat all on Linux03:42
specinggnarface: haha, people will never move off windows en-masse03:45
specingtom_work: play FOSS games03:45
furrywolfsure they will.  the heat death of the universe is believed to be a certain thing.03:46
gnarfacespecing: they won't, but the only thing that has brought more new users to Linux than that Valve/Microsoft spat was the release of Ubuntu03:47
golinuxOff topic03:47
gnarfacesorry, sorry03:47
golinuxAgain . . .03:47
gnarfacei took the bait and fed back too, i know.03:47
gnarfacei'll be good now03:48
golinuxAt least that wasn't in this channel03:48
golinuxImpulse control is a "good thing" [tm]  :)03:49
Oksanaspecing: I don't know how many people use Windows now... But I think that a century later I will hear 'haha, people will never move off linux en-masse'04:41
EHeMA challenge for gnarface/tom_work/others with Steam accounts:  Seems one or more of their developers don't understand e-mail either, as a result rather a lot of legal e-mail addresses get rejected during account creation; such as any address with '+' in the user portion.  If someone (or perhaps several people) are up for it, file a bug in their reportting system that [[:alpha:]][[:alnum:]]*@<anything DNS, including IDN> rejects many legal e-mail addresses.05:16
tom_workdh binary --max-parallel=3 --with python205:30
tom_workdh: Sorry, but 11 is the highest compatibility level supported by this debhelper.05:30
tom_workdebian/rules:24: recipe for target 'binary' failed05:30
tom_workmake: *** [binary] Error 2505:30
tom_workIs it not possible to build ascii backports for Ceres packages?05:31
gnarfacebackports don't always work without patches05:32
gnarfacethough, the first thing that comes to mind looking at that command is, could the python version be too old for ceres?05:32
gnarfaceit's just a guess, i don't know05:33
tom_workpython2 isn't supported in ceres?05:33
furrywolfpython is by far the biggest problem I run into trying to backport things.  it, and its packaging, is broken in every way.05:34
furrywolfand they fucking did it again!  3.8 is not compatible with 3.7.  they once again changed a bunch of stuff, INCLUDING THE GOD DAMN ABI, with a minor version bump.  again.05:36
tom_work>did it again!  3.8 is not compatible with 3.705:36
tom_workWHAT THE FUCK05:36
tom_workthat's it fuck porting stuff to python3, i'm just using perl05:37
furrywolfif you just run .py source, 3.7 files should work on 3.8, usually.  but anything that gets into the abi will break.  and, of course, 3.8 adds a bunch of new stuff, including a new operator, that makes code that works in 3.8 liable to not work on 3.7.05:37
furrywolf3.8 adds := as an operator, which does the same thing as = (yes, seriously), to make sure people's code stops working in older versions, again.05:38
tom_workwait, so only pyc files will need to be regenerated not source code modifications?05:38
furrywolfyou can distribute python binaries.  they'll stop working.05:39
EHeMI note that Python 3.8 is alpha according to the first mention I ran across.05:40
furrywolfbreaking things in minor versions is something python constantly does, and it causes constant problems.  for example, when backporting.  some tools needed to build ceres packages need python 3.7, but you can't install 3.7 on ascii without major breakage, since it breaks things that were written for 3.5.05:42
furrywolfwhen backporting ceres stuff, I have to force-depends install 3.7 and then re-install 3.5 afterwards to un-break everything else.05:42
furrywolfbecause of their constantly breaking stuff with minor version changes, debian packages of python stuff tend to depend on exact versions, like Depends: python3 (<< 3.6), python3 (>= 3.5~)   for borgbackup...05:43
furrywolfso you can't upgrade python without also using a ceres version of everything else, or forcing depends and letting it all break for the short time you need the new version.05:44
tom_workStill, I don't think I'm going to do any new deployments in python05:45
tom_workperl instead05:45
systemdleteis there a way to coerce browsers like ff and chrom* to open javascript ( type calls I guess) in a new tab, similar to the way middle click (used to) work?  There's an ancient bug on bugzilla for mozilla and friends, 19 years old.  Still not fixed apparently.05:46
systemdleteI tried to google, but I am worried that my google-fingers or google-magic or google-karma just isn't working for some reason.05:47
systemdletesome solutions I found involve greasemonkey scripts and the like, but the concern is those may create problems for other, more natural webpages.05:48
EHeMPython is an interesting beast, but I'm mostly writing Python when I need to extend an existing program, or have something which exclusively provides Python interfaces.05:48
systemdlete(and I looked through the extensions also, but nothing that seemed to match)05:48
furrywolfare you trying to make it open things that normally open a new window open a new tab instead?  googling says there's both a config setting for that, and an extension for that.05:49
furrywolfI really disagree with python's development model, in multiple different ways.  also, the language.  heh.05:50
systemdleteno.  this is for links that would open in the *same* window.  I'd like to cause it to open a new tab instead, leaving the current page as is.05:50
systemdlete(and thanks for responding, btw)05:50
furrywolf  does that do it?05:51
systemdleteincidentally, just to sort of join the thread here a bit, I use perl for nearly everything.  I figure whatever extra overhead it *might* add, the ability to quickly run down a problem and put in a fix is worth it.05:51
furrywolfthat's great until you try fixing someone else's perl.  :)05:52
systemdleteI acknowledge I am probably a bit sloppy there, but it does make things easier for me because its consisstent05:52
systemdleteI am *only* talking about my OWN scripts, of course.  LOL!05:52
systemdlete(and sometimes, after many months, I have to go back and remember what I did and how...)05:52
furrywolfyes, I guessed that.  few people who have tried fixing other people's perl code would describe it as quick.  :P05:52
systemdletethat extension, furrywolf, is for switching between new window vs new tab05:53
EHeMAnd Perl even does OO now too.  %-)05:53
furrywolfare you sure?  it looks like it's for links that would normally open in the same window.05:53
systemdletehow can you tell from "about this extension" -- it doesn't seem to indicate that05:54
systemdleteoh, and "the developer removed it"  -- click on the link05:55
systemdletemay I ask how you are finding these, furrywolf?   I could not05:56
furrywolfoh well.  you're on your own, then.  I tried.  :P05:56
systemdleteyes, google.05:56
systemdletegoogle... what search string, please?05:56
systemdlete(I know of google.  Believe me.  It's just that my search terms are not always too good)05:56
systemdleteEHeM, perl has done OO since 5.0.  c. 1995 or so05:57
systemdletebeen around a long time, and has been extended and improved a few times also05:57
systemdleteI do most of my own scripts in perl if they are beyond something simple, or I want to use modules05:58
furrywolfI just use right click -> open link in new tab...05:58
systemdleteThere is a perl module for nearly every purpose imaginable.05:58
systemdletefurrywolf:  Try that on a page where the devs have disabled the context menu (right click)05:58
furrywolfwhy would I use that poorly written of a site?05:59
systemdleteWell, let's say... that one is extremely... well, hmmm.05:59
systemdletehow can I say this?05:59
systemdleteLet's say that someone living in the modern age would like to meet someone else living in the modern age05:59
systemdleteso maybe they visit a dating site.05:59
furrywolf  install that first?06:00
furrywolfI'm sure there's other dating sites not written by idiots.06:00
systemdleteAnd let's say that the developers disable browsers to make it hard to keep multiple windows or tabs open simultaneously.06:00
systemdlete(nope.  You are wrong.  They are *all* written by idiots.)06:00
systemdleteI had to use ff to pay for one site because it didn't work in chrome.  OTOH, I had to use chrome to even view the site, because pics didn't show up correctly on ff06:01
systemdletein fact, I think all browsers are written by idiots too06:01
systemdleteevery one of them has their own set of standards06:01
systemdleteokcupid used to work fine.  Then they fixed it.  It has never been right since.06:02
furrywolfthe top three google results for my search,  ,  , and  , all appear to not do anything stupid with right clicks.  :P06:02
systemdleteit is nice to know that at least ONE Internet denizen is satisfied.06:02
tom_workare there any other good database clients in the Devuan repos for MariaDB? Mysql-workbench is too old to do anything but basic manual queries, building a backport is not possible, and building vanilla sources is way to complex due to pulling in an obcene amount of java dependencies06:03
golinuxSearch engines these days suck as much as browsers all of which is pretty OT.  This is a Devuan help channel06:03
* furrywolf concludes all of the internet since 2000 or so sucks06:04
systemdletetom_work:  Evolution workbench maybe?  idk, just a thought06:04
systemdlete(thinking there might be a plugin for databases)06:05
tom_worksystemdlete, evolution?06:05
systemdletenot the email client.   It's some workbench tool for developing C, perl, etc.  I thought I remember somewhere seeing that one could do db dev with it also.  I might well be wrong.06:06
furrywolfsystemdlete:  howlr also has working right click.  in fact, everything I click on has working right click.  find better sites.  I've suggested four.  :P06:06
* systemdlete thinks, sure, furrywolf. Hit me up with furries and then run away... :P06:07
systemdletethe world according to furries, I think.06:07
tom_workwhat about furries systemdlete ?06:08
systemdleteoh nothing, tom_work, nothing06:08
* systemdlete looks for nearby place to hide06:08
tom_workno what?06:08
systemdleteyes... ?06:08
systemdleteI have NOTHING against furries, honestly.06:08
systemdleteSome of my best friends...06:09
golinuxPlease move this to #debianfork06:10
golinuxor I may have to do a refresher course on how to 'quiet'06:10
golinuxfurrywolf should no better.06:11
golinuxSays one of the worst offenders06:12
* golinux goes to my irc cheat sheet06:12
* EHeM wonders when Raspberry PI 4 revision 1.1 will show up.06:17
golinuxWe discussed that briefly at the meet today06:18
EHeMAre there rumors based on actual experience? (mostly I would like to avoid the USB-C issue)06:19
gnarfacetom_work: to be clear, i don't actually know that python is the issue.  but having to first backport dependencies is a common backporting thing06:27
gnarfacetom_work: and there is a way to get multiple versions of python installed in parallel safely, i just can't tell you how06:30
gnarfacesomeone might already have done that work for python.  you should check ascii-backports for dependencies...06:30
EHeMAre there any plans to generate netinst images for anything other than amd64 and i386?  A few non-x86 systems can be installed via netinst image (there is a chainloading UEFI image for Raspberry PI 3 and there have been experiments on the 4)08:14
gnarfacethere are some ARM images, fyi, just not using the netinst08:15
gnarfaceplatform support is only limited by available hardware and expertise08:16
gnarfaceamd64 and i386 were just what could be supported, and the ones that had the most demand by a large margin anyway08:17
EHeMMy impression was the package lists from the Debian netinst images could be used to create a netinst image by combining the listed packages into a filesystem image, I don't see a similar package list.  :-(08:31
buZz? what kind of list are you looking for08:33
buZzthe tasksel stuff?08:33
yetiread about debootstrap and friends?08:33
EHeMIn debian/dists/buster/main/installer-arm64/current/images/udeb.list is a list that looks plausible as "Create a `deboostrap` and install these packages" will result in something pretty similar to a netinst image.08:40
EHeMPerhaps not a viable approach, but one which sort of looked plausible.08:41
gnarfacewell it might be fun to try08:43
gnarfaceit probably includes systemd and needs to be altered to work08:43
gnarfacebut i don't know08:43
EHeMTime to head out...08:57
onefangI was busy, but reading the backlog...09:01
onefang"<tom_work> that's it fuck porting stuff to python3, i'm just using perl"09:02
onefangWould that be perl 5, or the incompatible so they changed the name perl 6?  B-)09:02
buZzperl6 is changing name to eh, roku?09:02
onefangSomething like that.09:02
buZzyeah raku09:02
onefangMiddle clicking on a URL always opens it in a new tab for me.09:07
onefangI use phpmyadmin as a client for dealing with MariaDB / MySQL.  There's also a fork of it in Devuan repos, though I don't recall the name, but I did try it.09:08
tom_workonefang, perl6 is a different programming language09:32
onefangI know that, hence the rename.09:32
fsmithredjust trying to figure out where I am in relation to the irclogs17:23

Generated by 2.17.0 by Marius Gedminas - find it at!