IRC Logs for #crux Monday, 2019-11-04

*** xor29ah has joined #crux00:02
*** zimmer__Dl has joined #crux00:02
*** zimmer_Dl has quit IRC00:05
*** bkdwt has joined #crux00:09
*** nthwyatt has joined #crux00:10
*** darfo has quit IRC00:11
joacimfinally. took only two hours to reach the point where i get prompted to reboot the machine00:13
bkdwttwo hours? hahaha00:13
bkdwtAHA, I know it!00:26
bkdwt=======> ERROR: Md5sum mismatch found:MISSING   9d6bd09066b26c1a3b43e14ab37a67c1  harfbuzz-2.6.2.tar.xzNEW       1551bb7ebe970d3466787cd26cfa7f76  harfbuzz-2.6.2.tar.xz00:26
bkdwtharfbuzz downloaded directly from freesoftware site has a md5 and downloaded by prt-get has another md500:27
bkdwtThat's why some packages are failing for compile00:28
bkdwtAnd -is flag is not working in this case00:32
bkdwtThe only thing to do is wait until tomorrow to see if everything is back to normal.00:33
TimB_why does it check md5sums for you? its supposed to check the signature00:36
bkdwtTo be honest I don't know what's going on. I need to download some packages manually, edit Pkgfile and install them.00:46
*** john_cephalopoda has quit IRC00:47
TimB_try and check if you provide a valid in /etc/ports and haven't disabled signatures in prt-get.conf00:53
TimB_sorry, last one is pkgmk.conf00:53
TimB_time to sleep, good luck00:53
bkdwtgood night00:53
*** bkdwt has quit IRC00:56
*** bkdwt has joined #crux00:57
*** john_cephalopoda has joined #crux01:00
*** bkdwt has quit IRC01:44
*** samsep10l has quit IRC02:35
*** BurnZeZ has quit IRC02:39
*** tilman_ has joined #crux02:39
*** BurnZeZ has joined #crux02:40
*** tilman has quit IRC02:43
*** samsep10l has joined #crux03:07
*** abenz has joined #crux03:36
ryuodlcusa: so anyway, we can provide you with either file hosting or just a web redirect. our server is already configured for LE certs so it would be trivial to setup for that too.03:59
*** abenz_ has joined #crux03:59
*** abenz has quit IRC04:02
*** SiFuh has joined #crux04:07
*** SiFuh has quit IRC04:12
*** SiFuh has joined #crux04:15
Rudolphwhat do you use for LE?04:20
ryuoRudolph: web server's own php script.04:27
ryuothough now it's bundled with the web server.04:29
Rudolphah, nice04:29
ryuoit parses the web server configuration to avoid a lot of repetition.04:29
*** SiFuh has quit IRC04:33
ryuoRudolph: though i learned something new about mysql's views today. they absolutely suck for performance if you try to do anything advanced with them like aggregate functions because mysql apparently can't merge these with the user's select query so it uses a slower fallback.04:40
ryuoit's not the final query that's slow... trying it fully expanded works just fine. i guess it's something they felt was too hard to support.04:41
ryuobecause even the most recent version has this limitation.04:42
ryuothough not sure what benefit views really offer... you can do the same with a proper select query in the application.04:42
ryuoat best it helps abstract some of the repetitive, static parts.04:43
ryuoi guess it's a stored select query, though it's called a view.04:43
Rudolphyeah, i've hard they're terrible. is that one of the things mariadb is supposed to fix?04:44
Worksterwhy can't they just map the views to the real cells04:45
ryuoWorkster: to some extent they can. but most views are read only in principle. updating them is an extension.04:46
ryuoWorkster: but it's a lot more complicated for some kinds of queries. mysql only supports in a limited manner.04:46
Worksteroh a update would make many many regular commands for each field. probably.04:46
ryuoWorkster: but really mapping back to the originals only makes sense if you're wanting to modify data.04:47
Workstermost views are for read only i would imagine04:47
ryuoWorkster: example, mysql's limited to updating only one of the underlying tables of a view. if you try to update more than one in a single statement it'll error out.04:48
ryuothat's regardless of how the mapping works.04:48
ryuoplus it's a select query which can discard columns. mapping it back to the original table is non-trivial.04:48
Worksterit's underlying structure can not cross table boundaries with a single statement, but what about table joins04:49
Worksteryou can write to a joined table04:49
ryuosqlite doesn't support it at all though it has hooks to implement your own insert/update/delete logic for views if you use triggers.04:49
ryuoWorkster: you don't join tables when you do insert/update/delete.04:49
ryuotable joins only happen on select queries.04:49
ryuoaka read operations.04:49
Worksterno but mysql could use a joined table to write to more than one table at the same time04:50
Worksterfrom the view04:50
ryuoin theory, but they don't support it. that only really works though if there's a 1 to 1 mapping.04:50
Worksterguess they haven't figured that out04:50
ryuowhat do you do when your mapping the same raw in a junction table join?04:50
ryuoin my case I was trying to make a view that consolidated the information from the auxiliary table into a more usable form for the web application04:51
ryuowhich is why i was using aggregates04:51
ryuoi only needed the summary.04:52
ryuoi have a many to many between a users and roles setup and the permissions overlap heavily so i need to either check for a specific permission in all the roles a user has or consolidate them with a sort of logical OR.04:52
ryuoincidently mysql has BIT_OR aggregate function.04:53
ryuoi only care if there's an assigned role that gives the user a specific permission.04:54
ryuobut how to implement this is what i've been working on.04:54
ryuohow much to put in the DB and how much to put in the application.04:54
Worksterwhy not a permissions table and then use the id index from permissions table for the users table? no duplication.05:41
Worksterbut i haven't seen the data so don't know if that is viable.05:42
Worksteryou'd only need a row for each permission combination05:42
ryuoWorkster: because there's multiple types of users and users may have roles with overlapping permissions.05:46
ryuoWorkster: your idea is a flat model. it wouldn't support reusable permissions so to speak.05:46
ryuoif that existed it would only work if the application had some complex translation logic.05:47
ryuoWorkster: the whole point of roles is to be able to reuse a set of permissions for multiple users.05:48
ryuoplus it also makes it easier for the admin users to manage it.05:48
ryuoplus this also allows users to be assigned to multiple roles if needed.05:49
ryuoit's akin to unix user/group model but for the web.05:49
ryuothough fair point, i guess i could always just use SQL to search for it via a joined query.05:56
UNIXn3rdwhat does unsupported scheme file mean?06:34
ryuoUNIXn3rd: context?06:34
UNIXn3rdjre from alan's repo06:35
ryuowhen does it occur?06:36
UNIXn3rd$ prt-get depinst jre06:38
UNIXn3rdprt-get: installing /usr/ports/alan/jre06:38
UNIXn3rd=======> Downloading 'file:///jre-8u181-linux-x64.tar.gz'.06:38
UNIXn3rdfile:///jre-8u181-linux-x64.tar.gz: Unsupported scheme 'file'.06:38
UNIXn3rd=======> ERROR: Downloading 'file:///jre-8u181-linux-x64.tar.gz' failed.06:38
UNIXn3rdprt-get: error while install06:38
UNIXn3rdI noticed the URL looked local06:38
ryuooh, this kind of port.06:38
ryuoit probably expects you to get the tarball from somewhere else.06:38
ryuoyou see these a lot in the arch AUR06:39
ryuowhere they can't provide a download URL.06:39
ryuodo you really need the proprietary jre? is openjre not good enough?06:39
ryuoif you really need it, i'd suggest reading the pkgfile for instructions or so06:39
UNIXn3rdI never heard of openjre06:40
ryuothe runtime of openjdk06:41
ryuono idea if CRUX provides a different package for the jdk and jre.06:41
UNIXn3rdopenjdk is in crux, its for that libreoffice bin06:42
ryuoi'd say just install openjdk then.06:42
ryuoif it runs on the open source one, that's infinitely preferable.06:43
ryuowhatever it is you need java for anyway06:43
ryuoon our server we only use it for minecraft06:43
UNIXn3rdI don't do java programming its more for running libreoffice for college stuff06:46
*** stoffepojken has quit IRC06:46
ryuodoesn't matter; either should work06:46
*** stoffepojken has joined #crux07:01
dlcusaRomster, ryuo, it looks like dynadot doesn't offer redirecting, but their .net transfer/renewal rate is US$10.99, so I guess I'll take you up on your hosting offer for the collection tree at least.  Are you offering a vhost with my choice of OS or just a webserver account?07:26
ryuodlcusa: 3 things I can offer.07:27
ryuodlcusa: vhost where it directs people to files on the server, that would come with a regular user account.07:28
ryuodlcusa: redirect which would rewrite all requests to your real web server.07:28
ryuodlcusa: or, if romster agrees, we can setup a container to provide your "choice of OS"?07:29
ryuowe already have containers on this server setup.07:29
dlcusaIt would be easiest to redirect* URIs to* at this time.07:33
ryuoi can see about that.07:35
ryuoi believe the url rewriter can do that.07:35
dlcusaSo I'll transfer the zone from GoDaddy to dynadot that will point all traffic to the A record Romster posted the other day.07:35
ryuoso uh07:35
ryuothe server supports dual stack.07:36
ryuoup to you if you wish to add IPv6.07:36
ryuodlcusa: ok. once the dns propagates i'll see about configuring the host.07:37
ryuoi've never done this before, but it should be fairly easy.07:37
dlcusaMight as well since the casita server supports IPv6, too (on a IBM z15 box).07:38
ryuodlcusa: does your web server support https?07:38
ryuodlcusa: check the records for to get the addresses for the server.07:38
ryuoif you'd like i can have the redirect upgrade http to https.07:39
dlcusaI think not--I can ask the sysadmin if need be.07:39
ryuoi'll test it.07:39
ryuoconnection refused.07:40
ryuook, so no https for this.07:40
ryuoi won't bother with setting up LE then.07:40
dlcusaI'll email my benefactor and let you know what he says.07:40
dlcusaSo now dlcusa's A-host ( should be forwarding to and is currently timing out.07:56
ryuodlcusa: it needs
ryuothat's not our server address.08:06
ryuo2607:5300:120:3ac::1 for AAAA08:06
ryuoif you care to set that up.08:06
ryuoanywho i think i'm done configuring the server.08:06
ryuonow to reload.08:08
ryuoand it's ready.08:08
dlcusaBingo--even works again!08:08
ryuothis is how i configured your portion of our host.08:09
ryuodlcusa: not sure what you did? i assumed your dns would be using our server IPs.08:11
ryuois your dns using the correct addresses now?08:12
dlcusaNo, I changed the forwarding to the dotted decimal you just posted.08:12
Romsterthat comes up as
Romsterand 404 no pages08:13
Romsterso it's hitting our site now08:13
ryuoi'm not sure what's going on? i was expecting him to forward to our actual ip address.08:13
dlcusaI'mn getting the pages (?).08:13
ryuooh. no wonder.08:13
Romsterthen shounld't that make show and not the ip08:14
ryuothe way i set it up expects...08:14
ryuoto receive08:14
ryuothe HOSTNAME from being the first in the chain08:14
ryuowhatever is using right now is directing it to the ip address, which uses the fallback host.08:14
dlcusaPaleMoon is showing the casita address08:14
ryuo--2019-11-04 02:14:14--
ryuoResolving (
ryuoConnecting to (||:80... connected.08:15
ryuoHTTP request sent, awaiting response... 301 Moved Permanently08:15
ryuoLocation: [following]08:15
ryuothe A record should be for
ryuois that what you did?08:15
ryuoif so we're using an old copy of it.08:15
ryuoour web server is confused by you set it up.08:16
ryuoit was expecting to be the first in the chain, not the second.08:16
dlcusaYes, I updated it in response to your 08:06:31 post.08:16
*** kalamaki has quit IRC08:16
ryuothen it just needs to propagate i guess08:16
ryuothe joys of slow dns updates08:17
dlcusaI expect.08:17
ryuoi disabled logging of the redirects.08:18
ryuoi don't want to flood our logs with useless stuff.08:18
dlcusaWell, at least it's not chitchat, somewhat on topic.08:19
ryuono idea why they'd break these redirects. they're trivial to setup.08:20
dlcusaI'm positive they expected to sell more services thereby.08:21
*** guido_rokepo has joined #crux08:23
dlcusaNow people don't have to ask how to type that squiggley hyphen anymore.08:23
Romsterthis~ thing08:24
Romsterseriously people08:24
dlcusaI mean, it's evan 7-bit ASCII--what's the prob?  Oh, modern cellphone keyboards...08:25
dlcusaNow maybe will rediscover the dlcusa collection...08:32
*** kalamaki has joined #crux08:46
Romstersymbol shift 2/2 `~\|<>{}[]08:47
Romsterliterally 2nd key on the android symbol keyboard.08:48
Romsterof it's second page.08:48
Romsterlike so hard to find :D08:48
dlcusaYou may not have any idea the percentage of eyes that glaze over just at the mention of tilde.  Of course, I'm in Amurika.09:04
*** pedja has joined #crux09:11
Romsteri guess i am aussie but i use the odd characters for sites/programming etc. so i learned them09:14
dbrookejust been reading back and I may be missing something here, but if is issuing a 301 redirect why does it not just redirect straight to rather than via to do yet another 301 redirect?10:00
Romsteryeah i am not sure why but godaddy broke redirects10:02
*** chinarulezzz has quit IRC10:08
dbrookeOK, but the current redirect at is also broken from here ¯\_(ツ)_/¯10:09
Romsterdns is not instant10:10
dbrookethough I still don't see why needs to be involved if the existing can issue 301 itself10:16
dlcusadbrooke, I gather GoDaddy is stripping the rest of the URI when it forwards.10:22
dlcusaOr maybe the webserver for casita is mishandling it.  What I know is GoDaddy changed something around November 15, 2018 and all URIs ended up displaying thr ~dlc/index.html.10:25
dbrookethe redirect at does preserve the URI but it's not redirecting to casita -
*** chinarulezzz has joined #crux10:40
dlcusaGoDaddy may have stopped forwarding because I've started a domain transfer (obtained auth code).10:45
dbrookemaybe, probably just need to wait for things to settle10:48
*** xcko has quit IRC11:24
*** obarun has quit IRC11:34
*** Anselmo has quit IRC12:04
*** ivandi_ has joined #crux12:08
*** ivandi has quit IRC12:11
*** samsep10l has quit IRC13:04
*** Kruppt has joined #crux13:26
*** Anselmo has joined #crux13:29
*** samsep10l has joined #crux13:39
*** pedja has quit IRC13:58
*** stoffepojken has quit IRC13:59
*** stoffepojken has joined #crux14:02
*** pedja has joined #crux14:05
ryuodbrooke: i did what i could on our end. the rest is up to dlcusa.14:13
dlcusaryuo, dbrooke, not me, it's up to the domain transfer process. is the GoDaddy server that handles the forwarding spec GoDaddy supports--customers cannot change the actual A record.14:31
dbrookeit was the double redirect I didn't see the need for, but I understand that's just a temporary issue until things get sorted properly14:34
dlcusaThe double redirect is the GoDaddy way, it seems.14:35
dbrookefortunately I'm able to run my own DNS and servers so can avoid the need for workarounds14:41
*** stoffepojken has quit IRC14:58
*** stoffepojken has joined #crux15:02
*** guido_rokepo has quit IRC15:11
*** guido_rokepo has joined #crux15:22
Rudolphis ping provided by any package? it doesn't seem to be in any of the core ports15:29
Rudolphah, thanks15:31
ryuoACTION plays ping pong with Rudolph's nose.15:40
RudolphACTION watches it light up15:40
Rudolphlike a christmas tree15:41
ryuogotta love apt logic. "I know I just generated an initramfs, but why don't I do it all over again!?"15:41
RudolphHey, redundancy is important, and is important15:41
pedjazypper does the same15:49
pedjaerring on the safe side, I guess :)15:49
ryuoi fail to see why it's helpful to generate it more than once per system upgrad.15:54
jaegerDo you think it's actually intentional rather than an oversight?15:56
jaegerI doubt it15:57
ryuono idea. i just know i've done more than my fair share of upgrades where that is run multiple times.15:57
ryuoyou would think they could handle initramfs and similar updates as something to do once if they don't influence the outcome of other package operations15:58
*** guido_rokepo has quit IRC16:01
ryuo"be sure that the system will always boot even when zypper is interrupted"16:02
ryuodepends how that works.16:03
ryuoif it's done wrong, it'll leave you with a corrupt initrd.16:03
pedjahaven't tried it yet, but I think it'll revert to a previous working one16:06
ryuosomeone has a lot of time on their hands.16:06
*** Anselmo has quit IRC16:26
*** Anselmo has joined #crux16:29
*** samsep10l has quit IRC16:39
jaegerhrmm... dxvk still seems so much better than vkd3d17:00
*** zimmer__Dl has quit IRC17:36
*** zimmer_Dl has joined #crux17:40
*** tsaop has joined #crux18:00
Rudolphis there a script for xorg like the script for docker to make sure all of the kernel modules are installed? i've got the xfce port built but im having trouble pointing xorg to a display18:21
john_cephalopodaGraphic driver config is trivial in the kernel.18:25
john_cephalopodaWhat's your GPU?18:26
Rudolphoh, i'll be honest, this is my first time ever really dealing with x when it wasn't already installed18:26
Rudolphonboard macbook (which is intel)18:26
RudolphIntel Iris Plus Graphics 640 1536 MB18:26
john_cephalopodaTry installing xorg-xf86-video-intel.18:27
Rudolphdo i need to load the module for it afterwards?18:27
Rudolph(have it installed and rebooted already but didn't see any change)18:28
tsaopfor the record, I have circumvented the problem I had with consolekit2 and KDE 518:29
john_cephalopodaIt is just the xorg-side part, not kernel-related.18:29
Rudolphah ok18:29
tsaopsince the problem was in the inhibit signals powerdevil set, I simply disabled powerdevil18:29
john_cephalopodaRudolph: What is the exact error message?18:29
tsaoplo and behold, the system reboots from the menu18:29
TimB_tsaop: powerdevil?18:30
Rudolphinitiation: Xorg.0.log:
Rudolphoh, not sure if it matters, but this is a vm on my macbook with a vmware hypervisor18:32
tsaopTimB_: yep18:39
tsaopI checked via dbus who was inhibiting signals and there was powerdevil and kscreenlocker18:39
tsaopI disabled the screen locking and then powerdevil by removing the .desktop file from /etc/xdg/autostart18:40
TimB_oh, i didn't check, it's kf5 related?18:40
tsaopthe problem seems circumvented18:40
tsaopTimB_: indeed18:40
tsaoplet's hope CK2 gets picked up again18:40
TimB_for me, ck2 worked flawlessly except for some useless spam in syslog :)18:40
TimB_w/ mate that is18:41
tsaopyes, I was your bugreport18:41
TimB_yeah it's not merged18:41
tsaopseems abandoned18:41
TimB_hopefully it gets picked up, but what if it's the guy behind librsvg? ;)18:43
tsaopTimB_: you can expect a rust rewrite18:43
jaegerRudolph: it definitely matters since VMs don't generally have direct access to the hardware18:52
jaegerYou'd want to use intel for the host system if it were linux but since it's a VM you should look at installing the guest tools to get the proper X support18:52
Rudolphah, that makes sense18:58
stenurOn Apple i sometimes forwarded X via VirtualBox; the VM itself did not have an Xserver at all.18:59
stenurBut it was a mess.19:00
stenur(But it was SnowLeopard from 2009.)19:00
*** obarun has joined #crux19:22
Rudolphjaeger: do you know what the value should be for "What is the directory that contains the init directories (rc0.d/ to rc6.d/)?"19:32
*** replicant_ has joined #crux19:32
jaegerYou may have to make some fake ones and point it to them, or try using open-vm-tools instead19:40
*** bkdwt has joined #crux19:41
Rudolphyeah, thanks, just made some fake onesa19:42
Rudolphfound an arch linux wiki page that helped19:42
*** bkdwt has quit IRC19:49
*** Anselmo has quit IRC19:55
*** replicant_ has quit IRC20:00
*** nthwyatt has quit IRC20:24
*** darfo has joined #crux20:24
cruxbot[opt.git/3.5]: nvidia: updated to version 440.3121:35
cruxbot[compat-32.git/3.5]: nvidia-32: updated to version 440.3121:35
*** tsaop has quit IRC21:46
*** iovec has joined #crux21:57
RomsterRudolph, prt-get fsearch ping22:08
Rudolphah, handy22:18
*** Anselmo has joined #crux22:23
*** Kruppt has quit IRC23:31
*** zimmer__Dl has joined #crux23:36
*** zimmer_Dl has quit IRC23:38
*** pedja has quit IRC23:54

Generated by 2.14.0 by Marius Gedminas - find it at!