08:00:49 #startmeeting 08:00:49 Meeting started Thu May 31 08:00:49 2018 UTC. The chair is Jaymzz. Information about MeetBot at http://wiki.merproject.org/wiki/Meetings. 08:00:49 Useful Commands: #action #agreed #help #info #idea #link #topic. 08:01:08 #info Meeting information and agenda can be found here:  https://lists.sailfishos.org/pipermail/devel/2018-May/008391.html 08:01:10 morn Jaymzz 08:01:31 r0kk3rz: morning :D 08:01:31 I am the meeting’s chairperson today and will be doing my best to keep time and order. Please behave, respect the timings and be gentle. 08:01:38 \o jaymzz 08:01:39 #topic Brief introduction (5 min). Please prefix your name/handle with # info 08:01:46 Piggz_: o/ 08:01:55 #info lpr, community 08:01:57 Hi there! 08:02:00 #info James Noori, Sailor @ Jolla 08:02:06 elfio: hello :) 08:02:12 #info Matt Austin, community 08:02:22 ha, i have no hash key ...fail 08:02:25 #info elfio/chipiguay/pablo, community member and developer 08:02:36 #info Martin Kolman, community & modRana developer 08:02:44 #info Lewis Rockliffe, community hacker extraordinaire 08:02:46 info, piggz community porter and developer 08:02:55 #info Raine Mäkeläinen, developer @ Jolla 08:03:10 Piggz_: Just copy it from here 08:03:16 otherwise it won't be registered 08:03:36 #info Matti Kosola, Sailor @ Jolla 08:03:44 im on a jolla tablet with a bt keyboard...that isnt easy :) 08:03:46 #info John888, community 08:03:58 dont worry about it 08:04:12 #info nrde, community lurker, readonly 08:05:51 Not too many here today 08:05:58 1st topic coming up 08:06:06 #topic patches for CVE-2014-4323 and CVE-2016-5696 still not in kernel-adaptation-sbj (15 min – asked by lpr) 08:06:15 #link https://together.jolla.com/question/163935/validate-input-arguments-from-user-space-in-kernel-msm-mdp-cve-2014-4323-remote/ 08:06:25 #link https://together.jolla.com/question/142161/vulnerability-tcp-rfc-5961-of-linux-kernel-since-v32/ 08:06:33 #info CVEs mentioned above seem to be forgotten/ignored for jolla1 08:06:41 I think these patches should come to jolla1 too 08:06:55 Yes we will provide fix for above CVE's for Jolla 1 in 2.2.1 release 08:07:03 Actually let me make that # info 08:07:07 #info We will provide fix for above CVE's for Jolla 1 in 2.2.1 release 08:07:33 lpr I guess that answers the topic? :) 08:07:39 :) 08:07:41 ok, fine. just got confused by jovirkkus statement: @lpr: According to our tool the fix for CVE-2014-4323 is in Jolla1, JollaC, Jolla Tablet and Xperia. 08:08:03 #Jaymzz : yes 08:08:05 Yes but we decided to make it for J1 as well :) 08:08:19 Jolla 1 deserves to be loved <3 08:08:20 :D 08:08:35 yes, :) 08:08:43 moving on to your next topic 08:09:00 #topic upgrade to Linaro 4.8 2015.06 (MER#1899) (15 min – asked by lpr) 08:09:07 #link https://releases.linaro.org/archive/15.06/components/toolchain/gcc-linaro/4.8/ 08:09:14 #link https://bugs.merproject.org/show_bug.cgi?id=1899 08:09:15 Mer bug 1899 in gcc "upgrade to Linaro 4.8 2015.06" [Task,New] 08:09:28 #info upgrade to the latest and last Linaro 4.8 toolchain release (2014.04 -> 2015.06) , so package-build-process will not change but we'll see performance improvements and bug fixes 08:09:33 I think we should use latest 4.8 version 08:09:51 #info We have been discussing to upgrade gcc to latest 4.9, with no ETA to be announced at the moment. 08:10:02 Why not a version higher? ;) 08:10:23 why not a few more higher ;) 08:10:26 ooi what concrete work needs to be done to make such an upgrade happen? 08:10:38 yeah, GCC is really outdated at this point 08:10:44 stop being so precious about sb2? 08:10:53 because easiest think would be to upgrade 4.8 first 08:11:14 Piggz_: Good question, I think the amount of work involved plays a big roll here 08:11:21 I am still on 4.8.5 on opensuse leap 42.3 08:11:22 r0kk3rz: how is sb2 a stumbling block? don't we just qemu whatever compiler binary? 08:11:33 * chriadam_ doesn't know much about the sdk 08:11:34 and it's working 08:11:39 many projects (such as rinigus Valhalla package) have to resort to using custom new gcc builds 08:11:48 chriadam_: yeah but theres a performance hack that uses the platform gcc because its heaps faster 08:11:49 jaymzz:but really, its just delaying the work, has to happpen some time 08:12:07 chriadam_: so if theres a mismatch, BadThingsTM happen 08:12:14 mismatch between what and what? 08:12:27 abis? 08:12:29 the sb2 target and the host afaik 08:12:49 thats the gist as i understand it anyway 08:13:02 (I'm here, but late and lurking. Hi.:) 08:13:27 lbt: abranson: can either of you shed some light on this? I'm curious :-) 08:13:37 * M4rtinK_phone_ mentions Fedora is already at GCC 8... 08:13:54 M4rtinK_phone_: yes yes we know fedora is amazing 08:14:15 r0kk3rz: it is ;-) 08:14:15 on gcc, ive been using this with no major issues https://build.merproject.org/package/show/home:rinigus:toolbox/opt-gcc 08:14:25 Jaymzz: no ETA is a bit too far in the future :) I saw gcc-packages to 4.8.3-1.9.5 recently 08:14:36 its set to 4.8 ABI mode 08:14:52 r0kk3rz: yep, that's what I meant 08:14:58 what version is that? 08:15:29 Piggz_: We have to allocate our time properly based on priorities and our paying customer's needs. We can't allocate too much time to this. 4.9 seems to be a version that should not be so hard to upgrade to. 08:16:09 lpr: It should not be too far. I'm not giving an ETA yet as I don't want to risk it in case it gets delayed. But it should not be too far away based on what I've heard :) 08:16:35 reasonable enough...nice to see 6.4 version with 4.8 compat available tho 08:16:52 gcc4.9 would enable glibc2.25+ :) 08:16:54 the thing is that paying customers also benefit from upgrades and stuff 08:17:30 it just doesnt sound cool on paper 08:18:23 ApBBB: Indeed. but then again it is about priorities. If we are getting paid for X amount of hours to do some work, and we allocate it to these upgrades, we aren't acting very smart, business-wise 08:19:05 also unless we want to get pidgeonholed to using and outdated unsupported version of all libs/frameworks, we need to upgrade the GCC 08:19:29 upstreams are not going to support building with 4.8/4.9 forever 08:19:49 latest gcc4.9 is https://releases.linaro.org/components/toolchain/binaries/4.9-2017.01/ 08:20:03 some already don't, even just due to c++ 11 support 08:20:32 c++11 is fine, its c++14 thats the trouble 08:20:44 r0kkerz, after installing 6.4 in a target, what is needed in builds to use it? 08:21:16 Piggz_: tell your toolchain to use it, and its best to static link libc 08:21:49 Piggz_: so for a qt project, something like this - https://github.com/r0kk3rz/harbour-matrix/blob/master/harbour-matrix.pro#L17-L19 08:22:26 #link https://releases.linaro.org/components/toolchain/binaries/4.9-2017.01/ 08:22:32 good to know 08:22:34 #link https://github.com/r0kk3rz/harbour-matrix/blob/master/harbour-matrix.pro#L17-L19 08:22:59 * M4rtinK_phone_ things there should ideally be some general process for regularly updating core components so we don't end up in fire fighting mode so often & wit missmatched stone age/new versions of stuff 08:23:42 M4rtinK_phone_: jolla doesnt want to regularly update core components, they've said that quite bluntly on many occasions 08:23:44 2 minutes left on this topic, lpr do you think you have your answer already? :) 08:24:17 r0kk3rz: well, then thats a problem 08:24:21 Jaymzz: yes, thanks 08:24:41 having a relatively stable base to work from is a feature 08:25:01 Cool, I think we can move on to the general discussion and continue this discussion there. Will do so in a minute or so. 08:25:05 for the record, you don't upgrade core components in RHEL for the same reason 08:25:57 but you have major RHEL releases with updated core componens every few years, while there is a single outdated core release of Sailfish OS 08:27:01 just because its not updated in a big monolithic jump doesnt mean it hasnt been updated 08:27:28 Alright I'm chnaging the topic here, you guys can continue the same discussion now :) 08:27:29 #topic General discussion (20 min) 08:27:49 it didn't change the topic :D 08:27:55 #topic General discussion (20 min) 08:28:03 space? 08:28:14 Jaymzz: oh, I forgot to say that for cve-2016-5696 patch it is important to use the right one for kernel3.4: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux-stable.git/commit/?h=linux-3.4.y&id=d91a2aa46cbc95c9854d4a444fc6acee444ca655 08:28:25 is 2.2 readyish?? 08:28:30 #topic General discussion (20 min) 08:28:47 r0kk3rz: Yeah, there were 2 of those buggers. 08:28:49 ApBBB: has it been released to EA? 08:29:04 lpr: Okay, thanks :) 08:29:07 no 08:29:16 #info for cve-2016-5696 patch it is important to use the right one for kernel3.4: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux-stable.git/commit/?h=linux-3.4.y&id=d91a2aa46cbc95c9854d4a444fc6acee444ca655 08:29:17 then no, its not readyish then is it :P 08:29:24 #link https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux-stable.git/commit/?h=linux-3.4.y&id=d91a2aa46cbc95c9854d4a444fc6acee444ca655 08:29:35 ApBBB: Yes :) 08:30:02 it is closer than you might think, unless something blows up and in that case it's still not ready :D 08:30:22 Jaymzz: cool :) 08:30:36 Jaymzz: very diplomatic :D 08:30:46 r0kk3rz: :D 08:31:31 Jaymzz: any progress on the systemd daemon thing? and on the keyboards? 08:31:47 systemd socket 08:32:02 a sec 08:32:46 sledges is on holiday right now but regarding the KB, he asked me to say this: we have started looking into how to integrate presage text engine community's effort with jolla-keyboard, and currently analysing the licences involved 08:33:29 ApBBB: Does that answer your question so far? :) 08:33:55 not much changed 08:34:31 I've recently been developing some 'Progressive Web Apps' (PWAs). I was curious if support on Sailfish OS has been considered/discussed previously? 08:34:46 I have not been involved in the previous discussion followup so unfortunatey that is all I have until Simonas is back from holiday 08:35:00 Android (Chrome and Firefox) has some quite nice integrations now. Offline support, notifications, launcher etc. A nice example is the Twitter mobile site which works very well. 08:35:11 From my understanding iOS is gradually increasing support. Chrome desktop has some support too on Windows/Linux/MacOS. 08:35:13 mattaustin: perhaps considered, but i dont think the browsers support all the things you need for PWAs 08:35:21 Follow up question: Any chance of a browser engine update? 08:35:32 iirc it is being updated 08:35:47 to quantum? 08:35:53 hah, i doubt that 08:35:56 :D 08:36:01 rainemak: care to comment? 08:36:07 would be a dream come true to have latest ff on sfos natively 08:36:39 :) Well mozilla just did release quantum firefox for enterprise with long term support! 08:37:01 so if there is a release to target, the long-term-support versions could be a good choice. 08:37:41 sure, but typically we've been an LTS version behind the latest 08:37:55 understood! 08:38:27 let's see when next browser engine update happens, we're lacking a bit behind 08:39:13 rainemak: Thanks 08:40:04 Jaymzz: Thanks for the progress update on your internal check of the text prediction for keyboard. E.g.That the licenses are currently analysed but sledges is on vacation. 08:40:51 ljo No problem :) 08:42:09 question for sailors, did PBAP get removed from the Bt stack at some point around 2.1? 08:42:18 thanks! 08:43:20 and systemd socket actiovation support (a requirement for OSM Scout Server to be in Jolla Store) - any progress on that ? 08:43:34 (or did I miss an answer ?) 08:45:34 Piggz_: it has not been removed, but due to us updating to Bluez5, it may not currently work properly 08:46:16 jaymzz: well, i confirm it doesnt work in either of my cars ;) 08:47:01 oops :D 08:47:53 which is really frustrating in 2018 ..... in 2010, it would still have been a noted absence :) 08:48:43 M4rtinK_phone_: There has been no progress on that front 08:48:50 would be a handy workaround to push whole phonebook as a vcard ... but can only do it on a contact by contqct basis 08:49:02 And we are out of time guys 08:49:17 Jaymzz: OK 08:49:27 #topic Next meeting time & date (5 min) 08:49:32 gosh.... these spaces 08:49:36 #topic Next meeting time & date (5 min) 08:50:09 June 14th 2018 at 08:00 UTC. objections? :) 08:50:19 +1 08:50:43 Alright 08:51:08 #info Next meeting will be held on June 14th 2018 at 08:00 UTC 08:51:48 Thanks for your contributions guys. See you in 2 weeks! 08:51:54 bye! 08:51:56 #endmeeting