onefang | I came to this conversation late, and furrymcgee left, plus I'm still waking up. Download the package list files, compare them, create stats? apt-panopticon does a bit of that so it can check if Devuan package mirrors are up to date. Dunno what they are trying to accomplish though, but might be worth look. | 04:24 |
---|---|---|
goose1 | Have got a box here that suddenly won't boot. Removed the SSD and have now chrooted into it to reinstall the mbr. Despite the OS (ascii amd64) when I ran grub-install, I got the output "Installing for i386-pc platform. Installation finished. No error reported." That seems a bit odd doesn't it? | 14:36 |
fsmithred | no, it always says that | 14:37 |
goose1 | Ah, good ;) | 14:37 |
fsmithred | suddenly won't boot is not normal | 14:38 |
goose1 | True. My user, though has no problem doing stuff like force-powering down on a regular basis | 14:38 |
fsmithred | how far into boot did it get? | 14:39 |
GyrosGeier | is that a 32 bit PC? | 14:39 |
goose1 | Wouldn't boot at all. With no other medium present, I got a bios error message. With another one present, it goes straight to that and boots | 14:40 |
goose1 | GyrosGeier: No. amd64 | 14:40 |
goose1 | I might compare the mbr before and after grub-reinstall | 14:41 |
GyrosGeier | mh | 14:41 |
GyrosGeier | grub has more than the MBR | 14:41 |
GyrosGeier | the MBR loads additional code from the following sectors | 14:41 |
GyrosGeier | (which are unusable in a CHS based partition table, because you can only assign from cylinder 1 onwards) | 14:42 |
goose1 | RIght, yes | 14:43 |
goose1 | As it happens, before and after are identical | 14:43 |
fsmithred | did something in the bios change so it doesn't look at the ssd to boot? | 14:45 |
goose1 | That can't have happened. But when I choose it, it fails. Also, when I boot from a live the problematic SSD is not even seen, yet the bios seems to know it's present | 14:49 |
goose1 | I could maybe try it in a different sata port | 14:49 |
fsmithred | yeah, that's a good idea | 14:49 |
fsmithred | also check smart data | 14:50 |
fsmithred | smartctl -a /dev/sdX | 14:50 |
goose1 | OK | 14:50 |
goose1 | SMART overall-health self-assessment test result: PASSED | 14:51 |
* goose1 decides that it's time to put the SSD back in its box and try again with a different SATA port | 14:56 | |
fsmithred | is it a crucial BX...? | 14:57 |
goose1 | Kingston SSDNow UV400/500 | 14:58 |
fsmithred | kingston is probably ok. Crucial BX has no heat sink. Mine has crapped out under heavy load a couple times. | 15:17 |
goose1 | Right. Shall make a mental note of that. | 15:29 |
goose1 | All good now in different SATA port. Actually what was showing in the BIOS before was probably the CD drive, so neither the BIOS nor the live distro was seeing the disk | 15:30 |
goose1 | Now I have the box in my possession, it would probably be a good time to go from ascii to beowulf… | 15:31 |
critr | why not chimaera? | 15:38 |
goose1 | Ooh. I notice beowulf is oldstable already. | 15:38 |
goose1 | Is it OK to go ascii->chimaera directly? | 15:38 |
critr | i probly wouldn't just to be on the safe side. | 15:39 |
goose1 | Well i just backed up the partition, so maybe it's worth risking?. It's a stock desktop | 15:40 |
debdog | usually skipping releases doen't work out | 15:40 |
debdog | which doesn't mean you couldn't try – just to gain knowledge | 15:41 |
fsmithred | i think I tried ascii to chimaera recently and it did not work. | 15:51 |
fsmithred | goose1, here are my notes for ascii to chimaera. (Don't try this at home. Or at work. Or at your friend's house.) | 15:54 |
fsmithred | https://termbin.com/yxgi | 15:54 |
fsmithred | stock desktop might be worse because of the metapackages (which I tend to avoid) | 15:55 |
* critr alwasy just does clean installs. but i have copious install notes. | 15:55 | |
* fsmithred secretly hopes that goose1 will try it and let us know if it worked. | 15:57 | |
goose1 | Hehehe. Too late, I'm going the slow route ;) | 15:58 |
* goose1 is hoping that the ascii->beowulf problems, if they exist, will show themselves promptly | 16:00 | |
goose1 | Superficial check sees beowulf as OK. Copying that partition too before going to chimaera. Is it still wicd in chimaera? | 16:54 |
fsmithred | no, wicd is gone because it does not work with python3 | 17:03 |
fsmithred | you might want to remove it first or switch to connman or network-manager-gnome | 17:04 |
fsmithred | goose1, ^^^ | 17:04 |
goose1 | Right. So had I been relying on wifi, what would have happened on dist-upgrade beowulf->chimaera? | 17:09 |
fsmithred | I think wicd would stop working | 17:11 |
fsmithred | or might get removed. I think it depends on what/how you installed | 17:11 |
fsmithred | I think in some cases python2 stays behind and wicd does keep working | 17:11 |
goose1 | Alarming ;) Fortunately my client uses wired Ethernet | 17:14 |
samuraikid | how to fix discover plasma not show highlights on daedalus ? On Chimaera i add sid repo and it pushed the apps | 18:28 |
golinux | samuraikid: I hope that you used "ceres" instead of "sid" in your sources | 19:06 |
brocashelm | ^ | 19:07 |
goose1 | OK. First little glitch ascii->chimaera is that I had a problem with the applications menu getting into a loop of _trying_ to show itself. It failed. How do I get it back? | 19:10 |
samuraikid | golinux: ill try because when deadalus was release it bricked after update ill try now | 19:13 |
golinux | goose1: You should go ascii > beowulf > chimaera. I hope you have a good backup. :) | 19:20 |
brocashelm | the jump from ascii to chimaera is massive, so upgrade steadily from one lts to the next | 19:21 |
brocashelm | especially regarding kernel builds | 19:22 |
goose1 | I did it incrementally. Partition of each version (except chimaera) backed up | 19:22 |
goose1 | It's not the original monitor, so perhaps that doesn't help, but I haven't known menus to fail to appear at all with different geometry | 19:23 |
goose1 | Interestingly I logged in as a different user for the very first time (I think) and the applications menu showed itself as just a small empty rectangle | 19:42 |
goose1 | http://technojeeves.com/tech/xfce-panel-bug.png is what's happening. Execute is an endless loop | 19:51 |
goose1 | But now I look, I see tons of packages have been kept back | 19:56 |
goose1 | Upshot: ascii->beowulf seems OK on short exam. beowulf->chimaera miserable failure. Rolled back to beowulf | 21:53 |
critr | hmn. chimaera's working great for me. | 22:04 |
goose1 | After dist upgrade? | 22:12 |
* goose1 has standard desktop | 22:13 | |
critr | no, i always reformat and clean install. | 22:18 |
goose1 | Well that's entirely different ;) | 22:24 |
goose1 | Thanks folks | 23:34 |
Generated by irclog2html.py 2.17.0 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!