| freemangordon | Wizzup: actually, it can do 'autoremove' | 07:16 |
|---|---|---|
| freemangordon | but, that does not help much | 07:16 |
| freemangordon | anyway, I don't see hows was initially generated the list of signed packages in /var/lib/hildon-application-manager | 07:19 |
| freemangordon | maybe they just run some script when building the image | 07:29 |
| freemangordon | hmm... | 07:57 |
| freemangordon | ./var/lib/dpkg/status:Conflicts: hildon-application-manager-settings, nokia-repository, nokia-system-domain, nokia-certified-domain, nokia-selected-domain, maemo-extras-domain | 07:57 |
| Wizzup | It might not matter so much | 08:30 |
| Wizzup | and yes I believe those are just the old pkg names | 08:30 |
| freemangordon | but those names imply that initial domain files contents were not created by HAM | 08:34 |
| freemangordon | I grepped the whole fremantle root fs, and the only one that touches domain files is apt-worker | 08:35 |
| freemangordon | but, it does so only on package install/upgrade | 08:35 |
| freemangordon | I don;t believe all fremantle packages were installed through HAM | 08:36 |
| freemangordon | which means something else created domain files when rootfs was created | 08:37 |
| freemangordon | the problem is that now a random package coming from extras can replace libc6, for example | 08:37 |
| Wizzup | I am certain most packages were not ham installed | 08:41 |
| Wizzup | I think we should either have ham act the same way as apt, read the same repos, the same dbs, or have it only handle extras and nothing else | 08:41 |
| Wizzup | So why would ham need a special file with installed pkgs? | 08:41 |
| Wizzup | So know what domain it came from? | 08:41 |
| Wizzup | to know* | 08:42 |
| freemangordon | ye | 10:25 |
| freemangordon | yes | 10:25 |
| freemangordon | so a random package from extras to not be able to upgrade a system package | 10:26 |
| freemangordon | read https://github.com/maemo-leste/hildon-application-manager/blob/master/src/apt-worker.cc#L4936 | 10:26 |
| Wizzup | ok, but hildon-meta is not from extras | 10:37 |
| freemangordon | umm... yes, but somebody can put one in extras, no? | 10:43 |
| Wizzup | freemangordon: sure | 10:58 |
| Wizzup | but then our current setup wrt extras is ok, right? | 10:58 |
| Wizzup | will read the src soon | 10:59 |
| freemangordon | current == the one in -experimental? | 11:01 |
| Wizzup | I mean in the sense that extras can't install/upgrade pkgs from devuan | 11:05 |
| Wizzup | or maybe it is not ideal.. | 11:05 |
| Wizzup | let me get to my laptop (on d4 atm) | 11:06 |
| sicelo | freemangordon: upower issue opened ... hope i worded it correctly, https://gitlab.freedesktop.org/upower/upower/-/issues/301 | 12:11 |
| sicelo | freemangordon: or anyone else, just asking for very basic style review before i submit to ML, https://paste.debian.net/1348543/ | 15:31 |
| sicelo | it's meant to add voltage_max property for N900's fuel gauge | 15:33 |
| sicelo | the longer term goal is to eventually report other capacity_levels on this driver. currently it only reports Full (100%), Normal (everything else), Low (6%), and Critical (0%). | 15:36 |
| sicelo | i plan to do something like (voltage_max - voltage_min) * 0.75. values higher than this will report capacity_level High, and lower will report Normal. the other levels are unchangeable as they hard depend on flags set by hardware | 15:38 |
| freemangordon | sicelo: why not use checkpatch.pl? | 16:24 |
| sicelo | ah, yes, dumb me | 16:33 |
Generated by irclog2html.py 2.17.0 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!