libera/#devuan/ Friday, 2019-01-18

* Xenguy notes an update for libsystemd0 ...00:35
Jjp137yep just installed it an hour or so ago...00:37
cdI don't know if this is a Devuan issue, but I typically use Debian in a QEMU VM with QEMU's port forwarding which enables me to use SSH to the VM from my host system. But for some reason despite using identical /etc/ssh config files and ~/.ssh/authorized_keys with the same checksums between both the Debian and Devuan VM, I can't get an SSH connection to the Devuan system.00:40
cdIs there something particular with Devuan that could cause a connection to not be established?00:40
debdogshort answer: no00:44
debdogdoes ssh spit out an error message?00:45
debdogwhat about firewall?00:45
cdsftp will quickly return a connection refused error, but00:45
debdogcan youping it?00:45
cdnormal ssh just nothing happens00:45
cdnmap on 0.0.0.0 with the port I have forwarded via QEMU for ssh 22 says it's open00:48
debdogssh in as root or ordinary user?00:49
cdordinary user00:49
debdoghmm00:49
cdI'm using the same entry in .ssh/config on my host system that I use on both the Debian and Devuan VM.00:50
cdThey're both configured the same, using the same key00:50
debdogthere might be issues with the certificate thingies.00:52
cdOn the VM?00:52
debdogon devuan. not certain whether debian has theses probs, too00:53
debdoggnutls is the keyword here00:55
debdogcan you try with just a password? if it isn't too much a hassle00:55
cdI'll try00:56
* debdog has issues with git clone https:... because of this. while git clone git:... works well00:57
cddebdog: I don't think it's working with password auth either.01:05
cdCould it be some issue where it's wanting IPv6 support as well even though the host only has IPv4?01:06
debdoghmm, ok then.01:06
cdIt's listening for SSH on IPv6 as well on the Devuan VM01:06
debdogsorry, that's the only thing I can think of which might be different to debian01:07
g4570n!ping03:26
infobot~pong03:26
systemdletegnarface:  Thanks; forgot about that usb legacy suppport.  Just checked and it is enabled.03:39
systemdletefinally heard back from MSI support.  They could not explain why there is a mismatch of version numbers between installed BIOS's and those available for upgrade on ther site.  The respondent only said to match by date.  So, after hunting around the BIOS (man, they do make this more difficult than necessary!), I found the date, which was 042813.  That matches the release date for what they call version "H.F" available for03:42
systemdleteupdating the BIOS.   There are 2 newer versions, but the description of the changes made don't sound like they would help me.  I might upgrade the BIOS anyway just to avoid future issues.03:42
systemdleteI've never seen a MB where version numbers were so disjoint and confusing.  OK, matching by date does tie them together, but other mfrs manage to provide plain old version numbers across the installed and support site versions.03:43
systemdleteI've done BIOS updates on MBs and other devices without any confusion or trouble, aside from the usual nonsense back in the day when an update didn't complete or hung...  But understanding which version to update to was never confusing to me.03:44
gnarfaceseems suspicious to me03:45
systemdleteThis surprises me.  The board itself is not bad at all -- at least aside from this USB issue, and I'm not really 100% certain it is a USB hardware problem on the board.   I've been trying to find a similar but different distro to install in another partition and compare behavior.03:45
gnarfacewell the people who make the website aren't the same people who make the motherboard03:46
systemdlete(I agree suspicious maybe, but I'd like to try comparing distro behaviors on same hardware first)03:46
systemdletedo you think they ever talk to each other?  Or maybe meet, say, once a year?03:46
gnarfacei'd be surprised if they're even in the same country03:46
gnarfaceor the same parts of the company03:47
systemdletemaybe on one of those corporate-sponsored retreats where they catch their coworkers while falling backwards?03:47
systemdleteor speak the same language?03:47
gnarfacemore like speak parts of some of the same languages, badly03:48
gnarfacebut you're right that this seems to not be a problem for other vendors so i suspect schadenfreude03:48
systemdleteI doubt it could be related to software engineers' general lack of ability to communicate with other members of the same species.  That never happens.03:50
gnarfaceheh03:51
systemdletewhat a dumb thought!03:51
systemdleteI mean, in the years i worked in I.T. I never experienced it even one time.03:52
systemdleteEveryone got along perfectly.03:52
systemdleteThere was never any competition or misunderstandings.03:52
systemdleteAnd they never called a customer, even one with a legitinmate problem with our products, a complete *****03:53
systemdleteWe just told them to RTFM and all was well throughout the land.03:53
systemdleteanyway.  I might upgrade the BIOS anyway just to head off other issues not related to this.03:53
gnarfacesystemdlete: has google really returned nothing at all about this showing up for anyone else before on any linux distro?04:00
gnarfaceand did you ever check the logs for any error messages when the mouse goes out?04:02
gnarfaceyou really seemed to omit quite a lot of the tests i asked about04:02
gnarfacethat one specifically might be important04:03
gnarfacecheck /var/log/messages, i guess or dmesg... maybe /var/log/daemon.log too... check all the logs04:03
gnarfacecheck for anything that looks like it changes when the mouse goes out04:04
gnarfaceso far i seriously can't be sure whether it's a hardware or software issue04:04
xrogaanKatolaZ: Took 2-3 days, but debian finally solved their issues with libsystemd.12:36
buZzby removal?12:37
amesserlol12:42
djphbuZz: you're today's hero13:07
HumanG33kMy debian testing as never been as slow as now14:10
HumanG33kbut i think it’s because i use FF and thunderbird …14:11
HumanG33kQuestion related to devuan stable, why when my users use shutdown button it only close the session ?14:12
HumanG33kdefault install from last iso14:12
nemothat's kinda broad. but I assume that's XFCE14:12
nemoHumanG33k: the options in the XFCE logout menu are configurable14:13
nemoHumanG33k: they should have more than one choice unless the others are disabled14:13
nemoright click on it14:14
nemobut I haven't switched back to XFCE yet, so don't know more14:14
KatolaZHumanG33k: you should not have that problem, TBH14:14
KatolaZjust be sure you have consolekit installed14:14
HumanG33kis it not provide by default install ?14:16
HumanG33kor maybe it is provide by the option "command line tools options" (or similar) in the installer14:18
HumanG33ki install 2 devuan and have the trouble on both14:19
KatolaZno HumanG33k14:19
KatolaZit should be on the default install14:19
HumanG33kand the default dhcp discover is loooooooonnnnnnnng at boot14:20
HumanG33kit’s just an idea but it’s can be cool to have somewhere a webpage with pckge with systemd dependency in debian in order to easely report issues on their bug tracker, a least for testing on sid ?14:25
nemoanyone have any hardware suggestions? building a new system17:08
nemowas thinking AMD just to cut down on meltdown impact a bit, and just 'cause they seem just slightly less skeezy than intel17:09
r3bootnemo: meltdown-like issues are valid for all CPU's that do pipelining, including AMD17:15
nemospectre17:16
r3bootyep, same story17:16
r3bootif you want to protect yourself for that, right now, at this moment, it's best to use a CPU that doesnt do that, like a VAX, or maybe a 68k17:16
nemoheh17:16
r3boot(which is ofcourse not realistic)17:17
nemor3boot: point is. *apparently* AMD was a bit more cautious on its verification, so supposedly had fewer meltdown (not spectre) vulnerabilities17:17
nemotherefore takes less of a processor hit17:17
r3bootyou know what they say about assumptions right? :)17:17
nemoheh17:17
r3bootIig, spectre/meltdown are *very* hard to solve, unless the hardware architecture of the CPU's change17:18
nemor3boot: point is the patches impact Intel disproportionately for now17:18
r3bootand pipelining is needed to make cpu's fast17:18
nemor3boot: simply because (speaking of assumptions) AMD made fewer17:18
r3bootIf amd would apply similar patches, they would also face a slowdown, it's as simple as that17:18
r3bootbut, expect a slowdown of 10~30% depending on your workload17:19
r3boot(if you apply all patches)17:19
nemohttps://en.wikipedia.org/wiki/Meltdown_(security_vulnerability)#Affected_hardware17:19
r3bootNote, no independent researchers confirmed that ;) This could just as well be AMD protecting their business / seeing an opportunity to seize on the intel problems17:20
nemo"Researchers have indicated that the Meltdown vulnerability is exclusive to Intel processors"17:20
nemor3boot: again.. it's not certain, nothing is17:20
nemobut right now, with deployed patches, AMD is an advantage17:21
nemoany costs they had in their architecture was already paid up front17:21
nemotherefore the performance is more "realistic"17:21
r3bootI'm just saying it's not clear cut, since you need to believe the vendor on their blue eyes that they patched17:21
nemosure. that's obvious17:21
KatolaZnemo: it looks like AMD still need a but more power17:21
KatolaZbut less than 15 year ago17:21
KatolaZ(w.r.t. Intel equivalent)17:21
nemobut with what the distros are pushing out right now, there is no AMD impact17:21
nemoKatolaZ: yeah, so you take AMD vs Intel perf, and then subtract the meltdown patch mitigation slowdown17:22
nemowhich AT PRESENT is intel only 😉17:22
r3boot(and alpha, and itanium, but you dont hear a lot about that :P)17:22
KatolaZnemo: I buy hardware very rarely17:22
r3bootbut, w/e17:22
nemor3boot: yeah no intention of buying that so who cares ☺17:23
KatolaZactually, arm is not bad at all as an arch17:23
nemowhich is why most people ignore it17:23
golinuxHuman_G33k: A list like this? https://pkgmaster.devuan.org/libsystemd.txt17:23
nemoKatolaZ: trouble w/ arm unfortunately is as a desktop it is a real PITA17:23
nemoKatolaZ: Wine for example. or virtualisation17:23
KatolaZwell, depends on the desktop ;)17:23
nemoKatolaZ: I do have a number of ARM linux machines tho17:23
KatolaZthere are arm processosr which support virtualization17:24
r3bootnemo: you'd be surprised as to how many mission-critical applications are still running on those cpu's ;)17:24
KatolaZnot the 30$ ones, though17:24
nemor3boot: no impact on *my* buying decisions anyway ☺17:24
r3bootahja17:24
r3bootnobody except you cares about that as well ;+17:24
nemor3boot: for those, I can basically, at present say, for any machine I'm going to buy, meltdown is intel only patch performance impact17:24
nemor3boot: I was addressing your " but you dont hear a lot about that :P"17:25
nemo'cause a lot of others are in 'sact same boat17:25
nemoi.e. consumers17:25
nemor3boot: for most of us, meltdown is intel-exclusive17:25
KatolaZnemo: actually it seems like meltdown/spectre-like mitigations will probably affect larger sections of the kernel anyway17:25
KatolaZmeaning that they will result in rewrites over extensive parts of the kernel17:25
nemoKatolaZ: sure sure.  I was just drawing distinction between meltdown and spectre, since they really are different beasts17:26
KatolaZand that's for all the archs17:26
nemoKatolaZ: It almost seems the push to hyphenate them is a heavily intel driven exercise, given they cut (a few) more corners17:26
nemobut anyway17:26
nemoKatolaZ: https://www.amazon.com/AMD-Processor-Wraith-Stealth-Cooler/dp/B07B41WS48/  waddaya think?17:26
KatolaZnemo: I told you, I rarely buy hardware at all, so I am the worst person to ask an opinion on new hw17:27
KatolaZ:\17:27
nemo'k ☺17:27
Human_G33kif i try to migrate a debian testing to a devuan i will broke it ? smb/amule/transmission/openvpn are the main soft on it21:47
gnarfaceHuman_G33k: maybe, maybe not.  check the forums, it has been discussed.22:01
golinuxHuman_G33k: Our "testing" is beowulf.  Always use the release name in your sources.list.22:13
Human_G33kyep golinux22:14
golinuxI would search the dev1galaxy forum to see if any issues have popped up for those packages.22:14
Human_G33ki think i will just try with a vm22:14
golinuxbeowulf, especially the desktop, may have issues at this point.22:14
golinuxGood idea.  If you want to install from scratch, there is a mini.iso22:15
golinuxhttps://pkgmaster.devuan.org/devuan/dists/unstable/main/installer-i386/20181125/images/netboot/22:16
golinuxThere is also a amd64 iso22:16
Human_G33ki think it’s important at least for this release to be sure migration from debian can be done22:17
golinuxYes.  But we're not there yet,.22:17
Human_G33ksome people like me may have some debian testing22:17
golinuxWe'll get there eventually22:17
Human_G33kdo you know if the kde fork is in devuan ?22:18
golinuxYes KDE is available.  Might be issues in beowulf though.22:18
golinuxRead the ASCII Release Notes so you can understand the consolekit polkit stuff22:19
Human_G33kok i will clone / upgrade my testing and stable vm and make news for fresh install22:20
golinuxPlease report your experience on the forum or DNG mailing list.22:20
Human_G33kok i will :D22:21
Human_G33kthere is somewhere schematics of devuan core package ?22:21
Human_G33k*packages22:22
gnarfacegit.devuan.org?22:23
gnarfacemost of the packages are not different from debian22:23
golinuxI'm looking for the linl22:24
Human_G33ki want to know which commom packages devuan need for DE22:24
golinuxhttps://git.devuan.org/devuan-packages22:24
Human_G33ki will start a thread tests.22:24
Human_G33kthx22:24
golinuxYou might also read the migration instructions for ascii22:26
golinuxhttps://devuan.org/os/documentation/dev1fanboy/22:26
golinuxEspecially if you're coming from systemd environment.22:27
Human_G33ki already have migrate some debian22:28
Human_G33kSome of you also knows how to custom the debian installer ?22:30
gnarfacetechnically, the devuan installer is just a custom version of the debian installer22:31
Human_G33ki know that22:31
gnarfacewell, so someone around here must know how to do it22:32
Human_G33ki just want lvl up fast for make some improvment on it22:32
gnarfaceshould also be in git.devuan.org somewhere22:32
KatolaZHuman_G33k: you can put together a preseed file, if you need to make many installs (22:33
KatolaZ(sorry if I state something obvious)22:33
gnarfacei was gonna mention that too but i wasn't sure if that's what he meant.22:33
Human_G33kno i don’t want22:33
gnarfacesince he said "some improvement" i assumed he wanted to mess with the guided partitioning, because everyone seems to hate that part22:34
Human_G33kmake it several time some time show some error22:34
gnarfaceah22:35
gnarfacefyi there should be console errors on alt+f4 or F3 or one of those...22:35
Human_G33kin fact yes i always do manual partitioning22:35
Human_G33kbecause i always do dm-crypt lvm22:35
gnarfacesometimes errors go into one of those virtual terminals that could be useful to show up in the gui22:35
gnarfaceusually just edge cases though22:36
Human_G33kif i can learn fast how to make change in the installer22:36
Human_G33ki want to make some users type/kind preset22:36
Human_G33kand hardenned options22:37
gnarfacelook into the tasksel part22:37
gnarfacei can't help with specifics, but the presets are called "tasks" or mabye "task packages"22:38
Human_G33ki know yes22:38
Human_G33ki will come back soon22:39
Human_G33k(desapear in a cyber ninja smoke bomb)22:39
KatolaZHuman_G33k: the installer is not easy to hack around23:29
KatolaZif you are just interested in package collections, then tasksel is the place to start23:30
KatolaZother components might be far less easy to customise23:31
KatolaZit's a mixture of C, Perl, and shell scripts, accumulated over about 15 years23:31
KatolaZand all interacting with each other, in obvious and less obvious ways23:31
Human_G33kKatolaZ, no trouble for me with this languages interaction may be problematic23:38
Human_G33k*,23:40

Generated by irclog2html.py 2.17.0 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!