| norayr | BlagovestPetrov[: you may find this useful: https://leste.maemo.org/Development/Porting_Packages | 01:11 |
|---|---|---|
| BlagovestPetrov[ | thanks :) | 01:12 |
| BlagovestPetrov[ | Where is the source of libclutter? | 03:12 |
| BlagovestPetrov[ | it's not in Leste's GitHub and there is no debian directory in the official source | 03:12 |
| BlagovestPetrov[ | it library is completely deprecated since July 2022. Should we fork it? | 03:13 |
| tmlind | sicelo: sorry care to clarify.. do you mean that things work ok with all four musb patches applied or not? note that first fix can't be completely left out, the call for device_set_of_node_from_dev() is needed with the rest of the series | 06:25 |
| sicelo | with all four applied, things work, yes | 06:29 |
| sicelo | with only the 3-patch series, doesn't work. | 06:29 |
| sicelo | however, if 239071064732bc4a is reverted, then the 3-patch series is enough ... it works | 06:31 |
| sicelo | maybe should get HNS to test with this last configuration. hope it works for him too | 06:31 |
| tmlind | sicelo: ok thanks for confirming, yup makes sense | 06:32 |
| tmlind | it should work also for hns, i think he tried with gta04 which is another omap3 phone | 06:32 |
| tmlind | that whole trickery with moving the of node to a different device is super confusing though.. | 06:33 |
| tmlind | i can see why we want to avoid trying to populate irq resources from devicetree though, there may be dependencies that are unresolved because of deferred probe | 06:34 |
| sicelo | yes, makes sense | 06:38 |
| tmlind | uvos: hmm so is motoactv also using cpcap for pmic? | 07:04 |
| tmlind | freemangordon: looks like your modem .shutdown callback link expired, got that patch in a git tree somewhere? | 07:30 |
| uvos | tmlind: no its not related to the mapphones at all and dosent use cpcap | 11:13 |
| uvos | its related to the droid 1 | 11:13 |
| uvos | and is built from the same kernel tree | 11:13 |
| uvos | presumably this means it also uses twl5030 | 11:13 |
| uvos | ie like droid 1 it should more or less like the omap3 ti referance implementation | 11:14 |
| tmlind | uvos: ok | 11:30 |
| uvos | anyhow i just find it ammusing that they shoved a android phone into a watch, i dont think it will be terribly usefull as a device | 11:31 |
| tmlind | yeah well in theory battery life should be decent | 11:32 |
| uvos | not really, compeard to a proper gps watch like a garmin device that uses more like a microcontroller base | 11:33 |
| uvos | depends on what you expect by decent i gues ;) | 11:34 |
| tmlind | should be less than 10mW while idle, but the battery is probably tiny too :) | 11:43 |
| Wizzup | BlagovestPetrov[: there is https://github.com/maemo-leste-upstream-forks for clutter | 11:49 |
| BlagovestPetrov[ | thanks! | 11:49 |
| Wizzup | BlagovestPetrov[: how far did you get currently? | 11:49 |
| BlagovestPetrov[ | I didn't do a lot. iphbd works ok for now | 11:50 |
| BlagovestPetrov[ | I was mostly stuck because of clutter | 11:51 |
| uvos | tmlind: 230mAh so yeah :P | 11:51 |
| Wizzup | enough for ~1 day activity or so? | 11:51 |
| uvos | Wizzup: at best | 11:51 |
| Wizzup | well I think if there is no modem it ought to idle at lower | 11:52 |
| uvos | they also have a SGX530driving a 220x176 pannel | 11:52 |
| uvos | hillarious overkill :P | 11:52 |
| Wizzup | kinda fun though | 11:52 |
| uvos | can probubly push more tirangles than the pannel has pixels | 11:52 |
| uvos | easly | 11:52 |
| BlagovestPetrov[ | Wizzup: summarized, it's ~2 hours :) libclutter looks ok now | 11:54 |
| Wizzup | cool | 11:55 |
| Wizzup | uvos: https://packages.debian.org/bullseye/armhf/gconf2 | 12:28 |
| Wizzup | lol | 12:28 |
| Wizzup | it's just there | 12:28 |
| Wizzup | sicelo: let me try the rndis stuff now | 12:33 |
| Wizzup | sicelo: yeah so I don't see a network device, I probably lack a kernel option | 12:45 |
| Wizzup | oh, maybe it did work | 12:45 |
| Wizzup | yeah it seems ok | 12:46 |
| Wizzup | freemangordon: I took a stab at gtk2 on chimaera, I didn't have to touch any of our patches | 13:55 |
| Wizzup | freemangordon: it built, so maybe we're ok | 13:55 |
| Wizzup | freemangordon: G_INLINE_FUNC and G_CAN_INLINE are deprecated, will remove from clutter 0.8 | 14:08 |
| BlagovestPetrov[ | https://hastebin.com/uniqofosuj.bash | 16:10 |
| BlagovestPetrov[ | this is for libhildonmime | 16:10 |
| BlagovestPetrov[ | the gtkdoc xml is not properly parsed | 16:11 |
| BlagovestPetrov[ | have you experienced this before? | 16:11 |
| Wizzup | https://gitlab.gnome.org/GNOME/librsvg/-/issues/457 | 16:16 |
| Wizzup | maybe check how they fixed it | 16:16 |
| BlagovestPetrov[ | thanks:) | 16:20 |
| Wizzup | there is some patch, but I don't know if it was applied | 16:20 |
| Wizzup | librsvg apparently now uses someting else | 16:20 |
| Wizzup | so you might have to search the git hostory | 16:20 |
| Wizzup | history | 16:20 |
| Wizzup | https://gitlab.gnome.org/GNOME/librsvg/uploads/002a29454303fe74f36b90a569ebcd3d/Add-fallback-for-when-tree_index.sgml-is-not-available.patch this is the patch that I mean | 16:21 |
| Wizzup | but it's not clear if it is the 'right' solution | 16:21 |
| Pali | Hello! U-Boot for N900 should be fixed again. I sent patches to ML. I have tested just on one my N900. Could somebody test it also on other N900 device? | 18:00 |
| BlagovestPetrov[ | guys, do you know what is "outo" I don't find any packages with the same name. libosso-test depends on it | 18:54 |
| BlagovestPetrov[ | testing framework? | 18:54 |
| BlagovestPetrov[ | * guys, do you know what is "outo" ? :) | 18:55 |
| sicelo | Wizzup: ty for the u-boot/serial stuff. seems to have led to a fix :-) | 19:06 |
| sicelo | and thanks too for the hildon-usb-gadgets | 19:06 |
| sicelo | i wonder if there's no bug/regression in whoever sets IP address for gprs on at least d4 - today internet wouldn't work, and it was due to wrong ip address and gateway getting set. i ended up manually setting it to the value shown in `list-contexts` ofono script, and that worked | 19:08 |
| sicelo | anyway, i'll re-test it and report properly (if it happens again) | 19:08 |
| Wizzup | sicelo: hm, what was the 'wrong' ip? | 19:12 |
| Wizzup | sicelo: it's in -devel btw | 19:12 |
| Wizzup | sicelo: happy to help, sorry it took so long @ serial | 19:20 |
| Wizzup | Pali: thanks for figuring it out | 19:20 |
| Wizzup | BlagovestPetrov[: never heard of it, let's see | 19:20 |
| BlagovestPetrov[ | it may be a typo in the Control file | 19:20 |
| BlagovestPetrov[ | it's just for the tests package | 19:21 |
| Wizzup | does not seem like a typo | 19:21 |
| Wizzup | I don't think we ever tested or built this | 19:21 |
| Wizzup | does dpkg-buildpackage complain? | 19:21 |
| BlagovestPetrov[ | https://github.com/maemo-leste/libosso/blob/master/debian/control#L60 | 19:21 |
| Wizzup | yes, but does dpkg-buildpackage complain? | 19:22 |
| Wizzup | it was a maemo pkg btw | 19:22 |
| Wizzup | ~10 years ago | 19:22 |
| Wizzup | http://repository.maemo.org/stable/4.0/3.2_vs_4.0_content_changes.html search for outo | 19:22 |
| Wizzup | it wasn't ever in fremantle (maemo 5) | 19:22 |
| Wizzup | BlagovestPetrov[: note that it's not a build-depends, so we probably just never ran into it | 19:23 |
| BlagovestPetrov[ | sorry, the food came | 19:29 |
| BlagovestPetrov[ | build-depends is ok but it's not installable | 19:29 |
| Wizzup | yes, that's why we never ran into the problem | 19:30 |
| Wizzup | I doubt anyone ever ran those tests | 19:30 |
| Wizzup | :) | 19:30 |
| BlagovestPetrov[ | ok, I'll leave this packagew | 19:31 |
| Wizzup | BlagovestPetrov[: what do you mean exactly? | 19:33 |
| Wizzup | BlagovestPetrov[: my point is that if libosso builds, then it's in the same state (i.e. likely working) as it is *right now* in maemo leste | 19:34 |
| Wizzup | so the tests package not working because outo is not there doesn't mean we shouldn't add the package to the repo now | 19:34 |
| Wizzup | if that makes sense | 19:34 |
| BlagovestPetrov[ | right. I'm talking about the tests package | 19:55 |
| Wizzup | :) | 19:58 |
| BlagovestPetrov[ | https://github.com/eniac111/libhildonmime/tree/maemo/chimaera | 20:27 |
| Wizzup | great, you can PR against master branch :) | 20:27 |
| BlagovestPetrov[ | ok :) | 20:29 |
| BlagovestPetrov[ | ah, I see that INSTALL is also changed in my commit. It's probably from a build patch | 20:30 |
| Wizzup | use git add -p to add specific files | 20:33 |
| BlagovestPetrov[ | I'll just revert the commit and reset | 20:36 |
| BlagovestPetrov[ | https://github.com/maemo-leste/libhildonmime/pull/2 | 20:43 |
| Wizzup | I added the pr here https://github.com/maemo-leste/bugtracker/issues/644 | 20:45 |
| Wizzup | I'll look at it, or maybe ask freemangordon to look at it (although I think he's away for a few days) | 20:45 |
| BlagovestPetrov[ | :) | 20:45 |
| Wizzup | BlagovestPetrov[: can you add a bit of context to the commit in the pr, maybe amend it with the error that you get during the build | 20:48 |
| BlagovestPetrov[ | sure | 20:49 |
| Wizzup | uvos: do we still need mce-dev ? | 21:01 |
| Wizzup | uvos: I think that is in mce now, right? | 21:02 |
| Wizzup | uvos: this might perhaps cause issues as libosso depend on mce-dev for example, and mce might depend on things that ultimately depend on libosso? | 21:04 |
| tyusha | INFO | 21:39 |
| Wizzup | hi | 21:43 |
| BlagovestPetrov[ | https://github.com/maemo-leste/clockd/pull/7 | 22:49 |
| BlagovestPetrov[ | WIP, forgot the changelog.. | 22:56 |
| uvos | Wizzup: the mce repo now builds mce-dev | 22:58 |
| uvos | this has no bearing on the packages | 22:58 |
| uvos | they dident change | 22:58 |
| uvos | ie mce-dev still exitsts, just the git repo is obsolte | 22:58 |
| uvos | regarding dependacy cycles | 23:01 |
| uvos | i gues the sollution is to make those optional in mce | 23:02 |
| uvos | since mce dosent hard depend on anything anymore | 23:02 |
| uvos | besides glib | 23:02 |
| Wizzup | uvos: yes, this is a problem | 23:08 |
| Wizzup | uvos: hmmmm | 23:08 |
| uvos | the other soltion is to evict the debain packaging files form mce | 23:08 |
| uvos | and maintain them elsewhere like debian packages are genneraly maintained | 23:09 |
| uvos | then you can have mce and mce-dev in seperate control files again | 23:09 |
| Wizzup | I can see a few other solutions, but yeah | 23:09 |
| uvos | i gues its not possible to make it build just one pacakge from a given debian dir? | 23:09 |
| Wizzup | I mean there's many ways to do this, we could even just import mce-dev from beowulf | 23:11 |
| Wizzup | and then once we build mce the problem is solved | 23:11 |
| uvos | antother hacky sultion for now to achive the bootstrap would be to just remove alll the depends from the control file | 23:11 |
| Wizzup | in that case I'd rather import | 23:11 |
| uvos | build mce (it will suceed with all the mameo modules not built) | 23:11 |
| uvos | and use the resulting packages for the bootstrap | 23:11 |
| uvos | and then re add them | 23:12 |
| uvos | whatever works best | 23:12 |
| uvos | really i dont like that the package hard depends on all the stuff really | 23:12 |
| uvos | but i dont know how to make the debian packaging stuff behave like i want | 23:12 |
| uvos | since those hard depens are fake | 23:13 |
| uvos | if you build mce iwth all dependancies installed | 23:13 |
| uvos | and then remose sme | 23:13 |
| uvos | *some | 23:13 |
| uvos | mce will just stop loading the corrisponding module | 23:13 |
| uvos | and continue on otherwise | 23:13 |
| Wizzup | it should list what -we- need for mce to build the way we want it | 23:15 |
| Wizzup | so in that sense it's fine I think | 23:15 |
| uvos | building is not the thing | 23:15 |
| Wizzup | but currently it will not build because of the missing depends, the ci will just fail | 23:15 |
| uvos | the thing is that even after the build | 23:15 |
| uvos | with everything enabled | 23:15 |
| Wizzup | what about just importing mce-dev? | 23:15 |
| uvos | the binary package dosent hard depend on anything | 23:15 |
| uvos | sure that would work | 23:15 |
| uvos | whatever you like | 23:16 |
Generated by irclog2html.py 2.17.0 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!