IRC Logs for #crux Tuesday, 2015-01-06

*** doomicide has quit IRC00:35
*** miniml has quit IRC00:41
*** tilman has quit IRC01:04
*** tilman has joined #crux01:05
*** JSchilli1 has joined #crux01:16
*** JSchilli1 has quit IRC01:39
*** rmull has quit IRC02:49
*** rmull has joined #crux02:59
*** BitPuffin has quit IRC03:16
*** rmull has quit IRC03:22
*** rmull has joined #crux03:23
*** mavrick61 has quit IRC03:33
*** mavrick61 has joined #crux03:34
*** BitPuffin has joined #crux04:12
*** Feksclaus has quit IRC04:16
*** BitPuffin has quit IRC04:17
*** miniml has joined #crux04:28
rmullDoes anyone here use pulseaudio?05:35
*** nogagplz__ has joined #crux05:35
*** nogagplz_ has quit IRC05:37
rmullI'm having trouble getting dbus and pulseaudio to be connectable05:37
rmullIf anyone has any tips for configuring dbus to accept pulseaudio connections I would appreciate it...05:40
rmullUnable to contact D-Bus: org.freedesktop.DBus.Error.NoServer: Failed to connect to socket /tmp/dbus-rPvAjPH3X0: Connection refused05:42
rmullGet this when starting pulseuadio05:42
rmullThat socket doesn't appear to exist05:44
*** BitPuffin has joined #crux06:01
diverseI don't think most people here would ever want to touch pulseaudio, but maybe there is somebody here06:03
rmullI need it for bluez06:06
diverseoh really?06:06
*** BitPuffin has quit IRC06:06
rmullYes, for the stuff specifically concerning my device's profile06:06
rmullit's horrible06:06
diversejaeger do you need pulseaudio for using bluez on your steam box?06:06
rmullIt's not for all bluez stuff06:08
rmulljust for this a2dp stuff06:09
diverseso it's pretty niche06:09
rmullI got it working07:10
rmullI feel dirty07:11
rmullBut it works07:11
Romsterrmull, copy compat-32/apulse-32 to apulse and see if using that fixes your issue. see skype for a example usage of apulse-3207:40
*** asie has joined #crux07:46
*** BitPuffin has joined #crux07:50
*** BitPuffin has quit IRC07:54
*** asie has quit IRC08:02
*** onodera has joined #crux08:13
diverseGetting sleepy, I'll ttyl everyone08:20
*** c0x has quit IRC08:46
prologicjust got my new kbd and mouse08:49
prologicwireless :)08:49
prologicso far it feels quite good to type on and use the mouse pointer08:50
prologicposition of CTRL and SUPER are a bit getting used to though08:50
*** leo-unglaub has joined #crux08:57
*** leo-unglaub has quit IRC09:22
*** BitPuffin has joined #crux09:39
*** BitPuffin has quit IRC09:44
*** xeirrr has joined #crux09:46
*** c0x has joined #crux09:47
*** hhhhhhhh has joined #crux10:06
*** xeirrr has left #crux ("Leaving")10:15
*** linXea has quit IRC10:16
*** linXea has joined #crux10:16
nwetjeanre linXea !10:23
linXeaen Uppsalabo =)10:23
nweyupp =)10:23
linXeajovars, sitter och går igenom exjobbet och gör finjusteringar10:23
linXeapresentation on en vecka drygt10:23
nwe�fan :) g�r det bra d�?10:23
nweah :D10:23
linXeajodå, det går bra. Alltid lite nervöst att lämna ifrån sig typ 65 sidor text för rättning10:24
nwehehe f�rst�r jag :)10:25
linXeaHur står det till med dig då?10:25
nwejod� det �r bra h�r, sitter och slackar hemma.. (�r barnvakt �t tjejens dotter)..10:25
nwesatte henne framf�r tv:n :P10:26
linXeaheh, dumburken är ett säkert kort10:26
linXeaJag skall pilla lite till på detta sen borde jag söka lite jobb.10:26
linXeaKänns lite overkligt att vara klar med studierna nu.10:27
*** crash_ has joined #crux10:27
nwemjo, finns inte s� mkt att g�ra h�r hos mig :P.. ja det kan jag f�rst� :)10:28
*** c0x has quit IRC10:46
*** c0x has joined #crux10:57
*** mhe has quit IRC11:15
*** crash_ has quit IRC11:18
*** BitPuffin has joined #crux11:24
*** BitPuffin has quit IRC11:30
*** c0x has quit IRC11:40
*** onodera has quit IRC11:58
*** onodera has joined #crux11:58
*** BitPuffin has joined #crux12:07
*** BitPuffin has quit IRC12:41
*** c0x has joined #crux12:53
*** kori has quit IRC13:23
*** Camille_ has joined #crux13:27
*** onodera has quit IRC13:28
*** kori has joined #crux13:28
*** kori has quit IRC13:28
*** kori has joined #crux13:28
*** hhhhhhhh has quit IRC13:30
*** Lukc has quit IRC13:31
*** kori has quit IRC13:32
jaegerdiverse: nothing on it requires pulse except for some steam stuff, but I actually use pulse on most of my machines these days13:39
jaegerIt's much less terrible than it used to be13:39
tilmani think even steam no longer requires pulse to be installed separately13:45
tilmanmight be they added pulse to the steam runtime. or something.13:46
jaegerMight be, honestly I haven't looked too closely13:47
jaegeramusingly I use pulse with the alsa-plugins package so that software still uses alsa, pulse just does the work behind the scenes13:49
jaegerI like being able to individually mute applications and mix audio streams without a lot of fuckery13:50
Camille_yeah pulse is in steam runtime, I used steam without pulse on arch13:55
*** doomicide has joined #crux13:58
*** jdolan has quit IRC14:01
jaegermy steam box is off right now so I can't check, but I probably don't even have the 32-bit pulse lib installed outside of the steam runtime anymore14:03
jaegerI built that box pretty recently14:03
dxlr8rbah, I like esound :P14:04
jaegeresound did mix really well but I can't recall if it had the ability to control individual streams' volumes14:06
tilmandont think so14:06
dxlr8rthe last time I messed around with those things I actually used esound... since I have not used Linux as desktop :/14:08
dxlr8rso no need for audio :P14:08
jaegerromster and I are probably the "desktop-iest" users who hang around here regularly, I imagine14:09
jaegerAnd I don't actually know his usage habits but I guess that based on the large number of ports he has14:09
dxlr8resound was great. I had an PC with a faulty on board soundcard, it made all kinds of noises. so I set up my old Power Mac 9600 with Debian and esound and sent the sound to it and from there to my headphones14:09
dxlr8rgreat times :P14:09
dxlr8rI sometimes visit Linux desktop land. but usually with live CD's, Ubuntu etc. I don't get my hands dirty :)14:10
dxlr8rit just "works", or not :P and if not I really don't care :p14:11
jaegerI use crux full-time at work as my desktop and part-time at home, and I like a bit more featureful DE than the severe basic stuff14:11
jaegerin my case it's MATE14:11
dxlr8rI have tried... so... many times to use Linux for desktop :/14:12
*** jdolan has joined #crux14:12
jaegerIt's far from perfect, in my opinion, but it works well enough that I've done it for years14:12
dxlr8rit's not the lack of technical competence14:12
dxlr8rbut I don't like what I get14:12
dxlr8rnothing is perfect jaeger :)14:13
jaegerperfect is also subjective14:13
dxlr8rI don't need much from a DE/wm, but those few things are not there on linux14:15
jaegerwhat are those few things?14:16
dxlr8rI don't give a crap about it's integration with social media etc. :P14:16
dxlr8rI'll try to explain14:16
dxlr8r(english not my first language)14:16
*** jdolan has quit IRC14:16
jaegerdon't feel obligated to, I'm just curious14:16
dxlr8rsay I have a folder open in my file browser. and I want to open something inside that folder from an application. in Mac OS and OS X you could simply drag the folder from finder and into the "open" dialog, or even take the file inside and drag it into it's icon in the dock14:18
teK_I can execute any file with a predefined suffix in my shell :}14:19
teK_like $ ./sample_doc.odt14:19
dxlr8rsecond, this is actually possible now in gnome 3 to some degree. I want ctrl/cmd-tab to only cycle applications and not windows. I want a second modifier to take care of the windows inside that app14:19
dxlr8rteK_: what database does that use? self maintained?14:19
tilmanthere is xdg-open14:20
tilmanusing shared-mime-info iirc14:20
dxlr8rand I use this approach mostly to open it in not the default app14:20
teK_dxlr8r: self maintained, yes :)14:20
dxlr8ror else I could just double clock it14:20
teK_works for urls, too :p14:20
dxlr8rif I can have a DE/wm that does that then maybe there is hope :)14:21
dxlr8rand it would hurt it to be scriptable without having to restart it either14:21
dxlr8rtrying to build gcc on my new kernel :)14:32
rmulljaeger: Among a number of other things, one of the things that got my pulse/bluez5/dbus stuff working correctly was running EVERYTHING that needed to work as root. Do you happen to have any insight on the best way to avoid this?14:34
jaegerrmull: I let my X session start pulse itself, so it's running as my user. bluetoothd needs to run as root as far as I can tell14:38
jaegersame with bluetoothctl14:38
jaegerI haven't looked for any way to change that, it's not a large concern for me14:39
rmullYes - I think the issue is connecting pulse to dbus14:39
rmullThis was at 2AM last night so I don't recall the details I guess14:39
jaegerwell, dbus is usually running two sessions14:39
jaegerone system-wide and one user14:39
rmullThe system-wide one is started by the rc script - for the user one, I wrapped my call to my window manager in .xinitrc in a dbus-launch14:40
rmullDoes that sound right?14:40
jaegerI assume that setup handles whatever pulse needs, I haven't had problems with it (and thus haven't really investigated)14:40
rmullAlright, thanks14:40
jaegeras far as I know that should be ok, yeah14:40
rmullI'll bring it to the mailing list if I can't figure it out14:41
jaegersince I'm using MATE it handles launching some of that stuff for me, but dbus-launch should work as well14:41
*** jdolan has joined #crux14:41
rmullI haven't yet managed to hook pjsua up correctly so that it can use my headset as its input/output device, so I still need to sort that out. I was able to play music over it with mpv and paplay though.14:42
rmullI'm surprised pjsua doesn't have more traction with crux users. It's a voip softphone.14:42
rmulluses portaudio for its audio stuff14:43
rmullNot really sure what the deal is with portaudio14:43
*** doomicide has quit IRC14:48
jaegerI have no need for voip softphones so that's why it has no traction with me :)14:52
*** leo-unglaub has joined #crux15:22
*** hhhhhhhh has joined #crux15:23
rmullIs anybody here well versed in programming with TCP sockets?15:26
jaegernot I15:28
rmullI wrote a small server using epoll and it's buggy15:29
rmullNot really sure where to turn for help15:29
jaegerstackoverflow or similar site, perhaps?15:31
rmullYeah, that's probably resaonable15:32
rmullWas hoping for something a little more chatty15:32
diversea little more casual huh?15:33
rmullYeah, I don't want to have to get all dressed up15:33
diverseno problem with that ;P15:33
diversewell it's one thing to be versed in TCP sockets but it's another depending on the language you are using, because you know, APIs happen ;)15:35
rmullIt's C. socket(2)15:35
tilmanrmull: show ze codes!15:36
tilmanalso, what exactly is the problem?15:36
tilmanyou're not mixing up edge-triggered vs. level-triggered wakeups, are you?15:36
tilman(random idea #1)15:36
rmullI write data into a socket, then the client disconnects before the socket can be drained. I then close the socket, but when the client reconnects, it's getting all of the queued up data. Should I be using shutdown(2) instead of close(2)?15:36
tilmanso you are the server?15:37
rmullYes, and I am managing my clients by their fd number, which is probably a bad idea15:37
rmullI was thinking if I use shutdown, then the pending tx data will be dropped, and the client won't get the queued up stuff when it reconnects15:38
rmullI was about to try it15:38
tilmani don't know that one15:39
tilmanhow would data from socket #1 end up in socket #2 for the client? o_O15:40
rmullI think it's got something to do with using the socket's file descriptor number to manage the connections15:40
rmullI call close on the socket, but it then goes into TIME_WAIT state or something, so it's not actually closed even though the fd number gets decremented15:40
tilmananyway, AFAIK close'ing the socket in the server process on error is fine15:40
rmullBut is it sufficient if there is data queued up on the socket?15:41
rmullIt's a little weird15:41
rmullThanks for listening :)15:41
tilmanthink so15:41
tilmanafaik shutdown is only(?) useful if you want to keep data flowing in only _one_ direction15:42
rmullWell, the docs say that close(2) will try to complete the transmission, whereas shutdown(2) will drop it15:42
*** BitPuffin has joined #crux15:43
tilmanstill, the connection you are establishing after the error isn't the same that got broken earlier, so why would data arrive there? :o15:44
rmullYes, good question15:45
rmullMaybe I'm not clearing the rest of the client's state well enough15:46
tilmanare you restarting the client program after the error?15:46
rmullIt's an embedded device15:46
rmullWhich can be expected to do all sorts of broken annoying things15:46
tilmandunno then15:47
jaegerthat's inconvenient, heh15:49
*** thetornainbow has joined #crux15:51
rmullI wasn't resetting my client's state cleanly enough16:13
tilmanyou mean in the server process?16:14
rmullYeah. I buffer the packet separately before sending it into the socket16:14
rmullSo that was just hanging around, and getting more and more packet appended to it16:14
tilmanoccam's razor? ;)16:14
rmullMakes me wonder if I should just jam everything into the socket without buffering it separately16:14
rmulland let the OS worry about it16:14
tilmanyou really don't want to call into the kernel for very few bytes16:15
rmullSometimes it's a few bytes, but occasionally it's ~150kB16:16
rmullI'm not going to change anything because it Just Works at the moment16:17
rmullThis is for work, not for fun :P16:17
*** Camille_ has quit IRC16:33
leo-unglaubgood evening :)17:02
leo-unglaubi just build crux and replaced openssl with libressl17:02
leo-unglaubactually it works quit fine :)17:02
leo-unglaubbut still, personally i prefer polarssl over libressl/openssl ... much better api design17:04
teK_leo-unglaub: did you get my mail wrt the CSR?17:22
leo-unglaubteK_: ah, yes!!! i did ... but sadly i cannot decrypt it17:23
teK_haha :D17:23
jaegerbit ironic17:23
teK_yes but the screwup was on gpg's side ;)17:23
teK_still, probably, by leo-unglaub ;)17:23
leo-unglaublol, sure ... blame me for GPG ...17:24
teK_I think you lost the key?17:25
leo-unglaubno, of course not ... but ... it rejects my passphrase ..17:25
teK_oh boy17:25
teK_so you lost the key to the key :}17:25
teK_you better have your recocation certificate handy17:26
teK_also: query!!17:26
*** lnds has joined #crux17:33
rmullWill crux be switching to libressl?17:39
leo-unglaubrmull: it's a drop in replacement17:42
leo-unglaubso basically you can swap it out at any time17:42
diversewell the important thing is, is it mature enough to be useable?17:45
diverseare all the quirks taken out?17:46
rmullIs there any hardware acceleration?17:47
joacimfor 3D? I have a spare voodoo2 if you need one17:50
leo-unglaubdiverse: mature is relative with libressl17:50
joacimI have some spare vortex 2 cards if you need audio acceleration17:50
leo-unglaublibressl works, but not very well ... the problem is openssl works, but not very well17:51
diverseokay, so that means it's practical to use as a replacement17:52
diverseby your terms ;)17:52
leo-unglaubyes, they made sure that it stays binary compatible17:52
diverseyou mean API compatible?17:52
leo-unglaubäh, i meant api compatible17:53
leo-unglaubnot abi compatible, sorry17:53
leo-unglauba little typo on my part17:53
joacimor abi?17:53
leo-unglaubno, they are not abi compatible17:53
diversewhy would they need to be anyway? :P17:54
joacimIt would be convenient17:54
joacimespecially for binary distros17:54
leo-unglaubdiverse: well, the reason for that is that you can to do a large company and simply replace the files, rebuild and ship17:54
rmullAnybody want to weigh in on this?
leo-unglaubif large companies had to change there code, no one would do that17:54
leo-unglaubrmull: already ahead of you ;)17:55
rmullDo its recommendations look reasonable to you?17:55
rmullI don't want to copypasta without some corroboration17:55
leo-unglaubyes, most of them are17:56
diverseleo-unglaub: eh, it's much better to rebuild it if the source is available imo17:56
leo-unglaubi am using 4096 keylength for years now without any problems17:56
leo-unglaubalso everyone should disable wreak crypto by default17:57
leo-unglaubwhat they found out now is exactly the same as the "poodle" ssl attack17:57
leo-unglaubexactly as ssl, in ssh client and sender fall back to wreaker crypto if one of the two is not able to use ECC for example17:57
leo-unglaubsame bug known for years, but now the media got it to17:58
leo-unglaubi just proposed in #crux-dev to raise the keylength to 4096 by default in the rc file17:58
leo-unglaubmy local version already does that17:58
diverseso what is the keylength currently set to?18:00
leo-unglaub4096 is a good one18:00
diverseI mean, what length is it set by default in our ports?18:00
leo-unglaubif you don't specify something, it still uses the damn 102418:01
diverseand 1024 is pretty much crackable now?18:01
leo-unglaubit was already crappy 3 years ago ;)18:01
diversewell then, what the heck are the devs doing by not increasing the defaults?18:02
diversein the upstream18:02
leo-unglaubthey argue, that you can sed over the default config to increase it *g*18:03
diversewhy do they want to keep it at 1024?18:03
leo-unglaubno idea ...18:03
leo-unglaubi would guess to still support very old hardware, ...18:04
jaegeror laziness, perhaps18:04
leo-unglaubpropobly the same reason why firefox still ships SSLv318:04
diverseheh, if you are in the security field, speed is always 2nd place to integrity18:05
leo-unglaubdiverse: sadly not ...18:06
leo-unglaubbackwards compatibility is the first important thing for most of them18:06
leo-unglaubi mean look around18:07
leo-unglaubopenssl still comes with SSLv2 !!!!!!18:07
leo-unglauband SSLv3 by default!!18:07
leo-unglaubyou have to disable it manually18:07
diverseso the existing old applications are what cripple our security... just like with openssl18:08
leo-unglaubyes, exactly18:08
leo-unglaubmost crypto software has commercial interrests, and for them breaking compat means angry customers18:08
leo-unglaubor why do you think openssl still ships support for .net 1.5 !!!!18:09
leo-unglaubeven microsoft does not support that anymore, but they are payed to keep it in18:09
leo-unglaubthats why openssl abstracted the entire memory management instead of using the os components that are proven to work correctly18:09
leo-unglaubsadly it's all about the fucking money18:09
diverseleo, if there is one thing you have already learned, its always about the fucking money, in everything ;)18:10
leo-unglaubhmm, maybe ... but i dont like that18:11
diverseyeah, because it corrupts us18:12
diverseand apparently in more ways than one18:12
leo-unglaubwell, i refuse to be ruled by money18:15
leo-unglauband thats why i always increase my keylength *g*18:15
diverseand why we use DIY systems18:15
*** JSchilli1 has joined #crux18:26
*** bifeo has joined #crux18:35
*** bifeo has left #crux ("Leaving")18:35
rmullleo-unglaub: Just to get another thing on your radar, in /etc/login.defs, we are using SHA512 for passwd hashes which is fine, but we're only using the default of 5000 rounds. If we set SHA_CRYPT_MIN_ROUNDS we can choose to use a larger number of rounds and potentially slow down login attempts.18:36
rmullIf you're bothering the crux devs about security stuff, maybe you could add that to the list18:37
rmullIf they're unwilling to do anything about md5 package checksums or package signing with pgp keys, then tehy're probably unlikely to do anything else18:38
leo-unglaubrmull: thanks for the suggestion18:40
diverseleo-unglaub: wow, so dsa was the reason the psn got hacked huh?18:40
leo-unglaubrmull: but actually 5000 is a nice value18:41
tilmani think login.defs stuff might be easier to get in than changing ports from md5 to something else18:41
leo-unglaubwhat value would you suggest?18:41
rmullleo-unglaub: I don't know, that's where someone like you steps in to suggest something18:41
leo-unglaubactually 5000 is a good value, i think PHP does 50 *g*18:42
rmullAlso, ssh doesn't appear to support the recommended key-exchange algorithm from that SSH article18:42
leo-unglaubrmull: i already have a patched package manager that works with sha515sum files :)18:42
leo-unglaubrmull: we will get there :)18:43
diverseheh Romster was trying to get me to use sha256 or 512, for my pkgutils rewrite18:43
leo-unglaubi have to migrate 20 servers from debian to crux and i want crux to be secure before i do that because i have no interreest in patching up 20 running servers manually al the time *g*18:43
rmullHow come our ssh doesn't support the kex algorithm?18:44
rmulloh, nvm, I typoed18:45
rmullworks fine.18:46
JSchilli1i just wanted to thank you guys on the mailing list for putting up with me18:49
JSchilli1i thought i was at least ok at doing this kind of thing until i installed crux, haha18:49
leo-unglaubcurrently i like my job :)19:10
leo-unglaubmy current job is do develop a small but very secure Server-Framework19:11
*** leo-unglaub has quit IRC19:18
*** leo-unglaub has joined #crux19:20
*** kori has joined #crux19:35
leo-unglaubyeah, crux ist getting more secure every day :)19:39
leo-unglaubteK_ is currently working on a nice surprise *g*19:40
teK_am I? :>19:40
dxlr8rchange the kernel of crux with the one in Windows ME and replace the init with systemd?19:44
dxlr8rI got curious, never know :P19:44
leo-unglaubhehe, i bet the windows ME TCP stack still behaves better than the linux one under high load *g*19:46
*** toriso has joined #crux19:47
tilmanteK_: what surprise?19:50
teK_you do get the concept of surprises, right? :P19:50
tilmanhate them19:51
teK_your bad19:52
diversedepends on the type of surprise, but if we knew the type, it wouldn't be a surprise now would it?20:11
leo-unglaubi am goiung to eat something20:12
leo-unglaubsee you in a few hours :)20:12
*** leo-unglaub has quit IRC20:12
dxlr8ryou need hours for that? :P20:29
cruxbot[opt.git/3.1]: libunique: fixed deps20:40
*** hhhhhhhh has quit IRC20:47
cruxbot[opt.git/3.1]: dbus: updated to 1.8.1420:52
*** lnds has quit IRC20:57
rmullParty like it's 200421:04
*** JSchilli1 has quit IRC21:06
teK_nothing about windows losing their decoration after switching virtual desktops21:07
frinnsti've not hit that21:07
teK_I do several times a day21:08
teK_have to restart fb to make them reappear21:08
frinnstgranted I dont use virtual desktops that often21:08
frinnstnever have.. old windows injury :)21:09
teK_I even have these on Windows 8.1..21:10
joacimyour cpu is too slow, and you should reset your bios21:10
teK_now I have a 27"@4k where space is not an issue but atm I'm using my 12.1" nb so I'm heavily using all four of my virtual desktops21:10
teK_joacim: ww..what? :)21:11
joacimthat's what some people tell me, when i talk about something crashing or behaving poorly21:11
frinnstI have serious screen-space issues at work. I always solve it by opening a fresh new window instead :)21:12
teK_have you tried rebooting? Or even better: you HAVE to turn it off and wait some seconds21:12
frinnstformat and reinstall21:12
teK_frinnst: yeah but they are stacking or taking all of your task bar21:12
frinnstit seems to be the crux-ml way of solving problems :)21:12
teK_{CTRL,ALT} F<1-3,5> are the most commonly used keys on my keyboards :p21:13
teK_yes, I ignore these postings entirely, tbh21:13
frinnstyeah you get sad by reading them21:14
frinnsttek btw, new ssh id thingy if you try to ssh :)21:17
cruxbot[contrib.git/3.1]: fluxbox: 1.3.5 -> 1.3.621:18
teK_please write a warning to all maintainers as they should get that error on pushing via git21:21 is down?21:24
rmulloh, no, it's fine21:24
rmullthat was weird.21:24
teK_nothing to see here, keep on moving21:24
rmullI get "Bug in Mailman version 2.1.11"21:24
teK_I started apache with strace, so mailman bitched21:25
rmullWorks now21:25
teK_won't for long :>21:26
rmullLet me know when it is working again please :P21:30
*** jdolan has quit IRC21:31
*** kori has quit IRC21:32
*** kori has joined #crux21:33
*** kori has quit IRC21:33
*** kori has joined #crux21:33
diverseah that's nice, I'll update mine21:38
diversefrinnst: just need to restart the wm, right? No need to kill X?21:39
diverseteK_: I've experienced that odd bug once actually but it didn't happan all the time21:41
teK_I switched desktops a lot21:44
diverseI switch between between workspaces all the time too21:45
diverseI even drag windows across them21:45
teK_that works fine, yes21:46
diversealthough one bug that fluxbox and pekwm seem to share is that when dragging windows across workspaces, it can sometimes skip a workspace which gets annoying21:47
teK_yes? never have seen that21:48
diversedo you drag your windows across a lot?21:48
diverseit happens randomly, if you want to see it for yourself, just move a window back and forth frequently and see it jump an extra workspace when it shouldn't21:51
diverseteK_: ^21:53
teK_nope, works for me21:54
teK_(moved it ~20 times)21:54
diversekeep doing it21:54
diverseor perhaps they solved that bug in 1.3.6?21:55
teK_I did not restart fb yet21:55
tilmanteK_: that ssh config article got fefe's sign of approval. guess we're good now21:56
teK_we talked about insecure ciphers just yesterday or the day before yesterday22:00
teK_remember.. blogge oft, trainier selten22:01
teK_that feeling when a library issue you struggled for ~1.5 days with is solved by its author by adding a simple flag to fopen()22:04
teK_so the point is.. my call to fopen() was bogus :)22:05
*** leo-unglaub has joined #crux22:08
leo-unglaubhey :)22:08
leo-unglaubhas teK_ surpsized you already?22:09
teK_teK angry22:09
*** BitPuffin has quit IRC22:10
leo-unglauboh noooo22:10
leo-unglaubteK_ angry? why?22:10
*** jdolan has joined #crux22:14
diverseleo-unglaub: oohh, you don't want to make him angry or else "teK_ smash!"22:26
leo-unglaubi never make him angry ...22:26
diverseoops I forgot the ":P" emoji22:27
*** kori has quit IRC22:28
*** kori has joined #crux22:29
*** kori has quit IRC22:29
*** kori has joined #crux22:29
diverseleo-unglaub: I thought since you were into batman, I thought I make a marvel reference ;)22:30
leo-unglaubi doo22:32
leo-unglaubbut i spend today 4 hours at the dentist .. so i am not on 100% brain capacity *g+22:32
*** doomicide has joined #crux22:34
diverseyeah I know what you mean, that smell of drilled teeth particulates in the air can pass out anybody22:35
leo-unglaubthat was not the biggest problem .... my problem was that one of my tooth broke22:36
leo-unglaubyep, i had something to hard in my food22:36
leo-unglauband after a bit of food ... crack ....22:36
leo-unglaubhmmm, i am still thinking about the package signing22:38
leo-unglauband i am not happy with my solution ...22:38
diverseI think you need rest at this point22:38
leo-unglaubyeah, i should get some sleep ...22:39
leo-unglaubit's just ... i want to solve this *g*22:39
leo-unglaubplus, i am waiting for teK_ and his surprize :)22:39
diverseI think I know what his surprise is already22:40
*** kori has quit IRC22:40
leo-unglaubwell, he is not uploading nacked pictures of him ;)22:40
leo-unglaubit you think thats the surprize *g*22:40
*** kori has joined #crux22:40
*** kori has quit IRC22:40
*** kori has joined #crux22:40
diverseteK_ is not amused :P22:42
diverseleo-unglaub: no I was worried he was actually going to transform into a big green monster and smash everything in sight22:42
leo-unglaubwell, he is not Merkl *g* so i doubt that *g*22:43
leo-unglaubplease help me, i have a problem ... according to my current package sign solution the only file that i would need signed is the .md5sum file ... or the .sha512sum file22:46
leo-unglaubevery build resulting from that files must be secure ... however ... not ident22:46
leo-unglaubbecause autoconf could find some additional dependencies and build a different, but still secure package22:47
leo-unglaubam i right?22:47
teK_so the first thing to start with is the attacker model22:48
teK_which is..?22:48
leo-unglaubi have 3 different thread models22:48
leo-unglaubthe one that i cannot solve is number 322:48
leo-unglaubthread model 3 is that i install a packe and someone manipulates the already installed executable22:49
leo-unglaubi have no way to verify that all files installed are actually the ones from the package archive22:49
leo-unglaubexample: i have libopensslx installed22:49
leo-unglauband then someone gets to my laptop and replaces the .so file22:50
leo-unglaubi have no idea on how to verify the installed file against the package22:50
leo-unglauban idea would be that there is a new file called .footprint-installed ... and that file contains the checksums of all files22:50
*** JSchilli1 has joined #crux22:54
JSchilli1is it safe to visit i'm getting a certificate warning22:55
teK_which browser?22:55
leo-unglaubJSchilli1: try it without www.22:56
JSchilli1i'm not using www22:57
leo-unglaubcan you check the cert's fingerprint? sha256: 35:61:22:C1:5C:D3:5B:61:66:67:93:CD:F3:01:CC:BB:0E:A4:99:51:69:24:77:F0:04:51:8A:92:BE:72:63:3D22:57
*** mhe has joined #crux22:58
teK_I can confirm the issue22:59
leo-unglaubteK_: i am using firefox and i dont have the problem22:59
JSchilli1leo-unglaub: how do i check the fingerprint? sorry23:00
leo-unglaubif you get the warning window, select "i understand the risks" -> add Exception23:02
leo-unglaubthen you get a popup23:02
leo-unglaubthere click "view"23:02
leo-unglauband then you see the fingerprint23:02
JSchilli1seems right?23:04
teK_IE does it, too23:04
teK_gotta check the redirect rules23:04
JSchilli1it hasn't always done this, i left for a bit and came back and it was like this23:05
teK_JSchilli1: please retry23:05
teK_visit http://www.crux.nu23:05
JSchilli1no dice, it redirects to https://crux.nu23:06
teK_that's ok23:06
teK_still getting the warning?23:06
teK_it should go away with that23:06
mechaniputerI am also using firefox, and I get no warning on either my Debian or CRUX machines.23:07
teK_srsly -_-23:07
leo-unglaubit even works on my android firefox *g*23:08
leo-unglaubJSchilli1: can you restart your browser23:08
leo-unglauband please try again23:08
JSchilli1doesn't fix it23:09
teK_hit shift + ctrl + del and delete the cache for the 4 weeks or so23:09
leo-unglaubhave you removed startssl from your cert authorities?23:09
JSchilli1leo-unglaub: not that i know of23:09
JSchilli1teK_: still getting it23:10
teK_ok this is really odd23:11
teK_this is the current FF, right?23:12
jaegerany proxies? does it happen on other machines?23:13
JSchilli1teK_: indeed it is23:13
JSchilli1jaeger: no proxies, it loads just fine on my phone23:13
teK_I blame his/her ISP :p23:13
JSchilli1the phone is on wifi not cellular23:13
teK_the last thing I could think of is to move your profile and retry23:13
JSchilli1what do you mean by that?23:15
teK_you have some .mozilla/firefox folder23:16
teK_in ~23:16
teK_move that (temporaryly) just to be sure23:16
JSchilli1ahh okay23:16
teK_sorry for the inconvenience23:16
JSchilli1wow, still getting it23:16
JSchilli1guess there's not much i can do23:17
jaegerJSchilli1: does give you trouble?23:17
JSchilli1yes it does23:18
JSchilli1same warning23:18
*** irclogger______ has joined #crux23:21
*** rauz has joined #crux23:21
*** bedis has joined #crux23:21
*** pips- has joined #crux23:21
*** nrxtx has joined #crux23:21
*** novak has joined #crux23:21
*** jaeger has joined #crux23:21
*** blueness has joined #crux23:21
*** Stealth has joined #crux23:22
cruxbot[opt.git/3.1]: [notify] libunique: downgraded to 1.1.623:22
*** thetornainbow has joined #crux23:22
*** mavrick61 has joined #crux23:22
*** pejman_ has joined #crux23:22
*** z3bra has joined #crux23:22
*** jue has joined #crux23:22
*** swartulv has joined #crux23:22
*** Sleepy_Coder has joined #crux23:22
*** Papey has joined #crux23:22
*** tkln has joined #crux23:22
*** syncn has joined #crux23:22
*** fnords has joined #crux23:22
*** Wasp has joined #crux23:22
teK_JSchilli1: uhm..23:25
*** j^2 has quit IRC23:25
*** swartulv has quit IRC23:25
*** j^2 has joined #crux23:26
*** leo-unglaub has quit IRC23:26
*** swartulv has joined #crux23:27
teK_I think jaeger and I found the problem23:27
teK_trying to fix this.. hang tight ;)23:27
teK_JSchilli1: please retry23:33
JSchilli1it works! thank you23:33
teK_thanks for reporting :)23:33
teK_have fun23:33
JSchilli1hehe i will23:33
*** Agge has joined #crux23:37
teK_still strange that my FF on windows did not check the (full) path23:37
JSchilli1what was the issue?23:38
teK_startssl signed our certificate with an intermediate CA cert23:38
teK_so you have something like this:23:38
jaegerpretty common thing to do now23:39
teK_Trusted CA X --> startssl "root" CA -->  [[MISSING LINK]] --> intermediate CA --> our certificate23:39
teK_we added the MISSING LINK23:39
jaegerfirefox has the intermediate cert in it, it seems23:39
teK_FF Windows / Chromium / IE did not seem to care..23:39
*** Feksclaus has joined #crux23:49

Generated by 2.11.0 by Marius Gedminas - find it at!