**** BEGIN LOGGING AT Mon Mar 06 03:00:00 2017 Mar 06 07:55:37 well, old (2012) behavior was that https://maemo.org* was redirected to https://bugs.maemo.org (???!). Now actually, midgard forces you to login once you use https Mar 06 07:56:05 since that old redirect has been removed Mar 06 10:09:21 funny stuff is HTTPS works as supposed once you logged in Mar 06 10:12:03 with the recent brainfuck mandatory https in browsers, I wonder how many disgruntled users think they are not allowed to browse maemo.org anymore, without authenticating Mar 06 10:15:54 * DocScrutinizer05 also idly wonders what's the correct time spam for black ribbons and condolences Mar 06 10:16:46 I'm absolutely not the right person to judge on such stiff, as you cna see easily Mar 06 10:17:06 "stuff", and "can" Mar 06 10:18:30 thanks to missing publication date I can't even tell how long that is there already Mar 06 10:25:46 seems to be since around 2016-11-16 or sth Mar 06 10:26:53 I _think_ 4 months of condolences is enough Mar 06 10:27:44 @council: ^^^ ? Mar 06 10:29:16 ohmy sorry! time SPAN, not time spam Mar 06 10:31:57 * DocScrutinizer05 idly ponders if midgard actually could roll back to page revision 1.29 (from 1.30) Mar 06 10:37:29 pretty sure that not: http://wstaw.org/m/2017/03/06/plasma-desktopf17764.png but http://wstaw.org/m/2017/03/06/plasma-desktopo17764.png Mar 06 11:04:58 midgard, my ass. You never know what's fucked up by design, what by implementation, what by maemo addons, and what by database corruption and bitrot Mar 06 11:06:03 which boils down to a simple constant W*T*F??!?? Mar 06 11:08:31 but of course http://atomictoasters.com/wp-content/uploads/2011/09/world_as_seen_by.png Mar 06 15:49:49 freemangordon, Wizzup: what are the open questions regarding porting Maemo? Mar 06 15:50:25 I just read the "getting organized" thread again Mar 06 15:50:31 it's all still quite confusing Mar 06 15:51:18 I don't see the ultimate goal Mar 06 15:54:02 For me the goal would be like: Mar 06 15:54:20 using "standard linux" (debian, devuan) Mar 06 15:54:48 port the best applications that are easy to port from Maemo to standard linux Mar 06 15:55:07 reimplement pieces of software that are too difficult to port Mar 06 15:55:29 -> this should give a nice mobile OS in the long run Mar 06 16:01:36 Maybe it would also be nice to compare with other mobile Linux - Distros Mar 06 16:03:16 I found the following: QtMoko, SHR, QuantumSTEP, OpenSUSE (according to http://projects.goldelico.com/p/gta04-main/page/Software/), maybe there are more? Mar 06 16:03:46 please comment... Mar 06 16:18:26 bbl Mar 06 17:02:29 ravelo: the "ultimate goal" is to have as much as possible functionality from fremantle running on top of some recent distro, I don;t think such a goal needs anything written to be clear :) Mar 06 17:03:55 and no, porting pieces from here and there won;t give us the expected result - we need everything possible ported - h-d - hh, mce, dsme, etc, etc Mar 06 17:29:16 freemangordon: how's the progress now? ;-) Mar 06 17:30:06 MoeIcenowy: I am on hildon-home and its dependencies, right now trying to port libhildonfm Mar 06 17:30:32 this is gtk2 port ofc, for gtk3 see the relevant thread on TMO :) Mar 06 17:31:21 unfortunately I am very short on spare time recently so the progress is not that fast as one would like to be, but :) Mar 06 17:34:51 MoeIcenowy: and ofc, it can;t be any faster with guys and girls only asking for progress, without actually helping on the development part :p Mar 06 17:36:20 sorry about that, btw ;) Mar 06 17:37:08 freemangordon: how did it go with the osso-bookmark-engine debian package? did it work properly? Mar 06 17:37:32 * freemangordon is sorry for his mood, he is a little sick and not in the best condition Mar 06 17:37:37 bencoh: yes Mar 06 17:37:42 oh, cool then Mar 06 17:37:47 with some tweaks for scratchbox Mar 06 17:37:58 oh? strange, since I built it on sb Mar 06 17:38:13 bencoh: https://github.com/community-ssu/osso-bookmark-engine/commits/master Mar 06 17:38:20 no idea, it failed here Mar 06 17:38:32 different sb version I suppose Mar 06 17:38:38 or different automake/autoconf version maybe Mar 06 17:40:01 could be Mar 06 17:40:07 well I'll try and keep those changes in mind next time :) Mar 06 17:40:27 bencoh: though, I've never seen libtool used in makefile :) Mar 06 17:40:46 wait, automake/autoconf? Mar 06 17:40:58 err nevermind, there is no autotools there Mar 06 17:41:00 autofoo is not used by your packaging Mar 06 17:41:05 :nod: Mar 06 17:41:10 I was referring to libtool actually :) Mar 06 17:41:46 and, exactly ... autotools usually abstract all those libtool calls, meaning we never have to bother with it Mar 06 17:42:29 mhm Mar 06 17:42:50 and other makefile-based maemo packages just don't use libtool, meaning you have to hardcode lib names and stuff like that, which I didn't want to do Mar 06 17:42:59 but anyway, the one on github compiles in both SB and devuan, so it should be ok Mar 06 17:43:04 nice :) Mar 06 17:43:17 could serve as a base for other osso-* packages then I guess Mar 06 17:43:25 mhm Mar 06 17:43:25 unless someone at devuan says it's utterly broken :] Mar 06 17:43:38 (actually, it is utterly broken per modern devuan standards) Mar 06 17:43:49 (s/devuan/de**an/) Mar 06 17:43:58 though I still prefer using autofoo to wrap libtool Mar 06 17:44:02 me too Mar 06 17:44:15 just didn't want to add it since it wasn't there in the first place Mar 06 17:44:20 maybe I should've :] Mar 06 17:44:38 how did you decide on whether it was there? Mar 06 17:45:11 there was no auto* (configure.ac/makefile.am) files in repository, so I just went with the flow Mar 06 17:45:22 s/files/file/ Mar 06 17:45:23 bencoh meant: there was no auto* (configure.ac/makefile.am) file in repository, so I just went with the flow Mar 06 17:45:31 ~botsnack Mar 06 17:45:32 bencoh: :) Mar 06 17:45:42 but ofc, this is RE job, not porting :) Mar 06 17:46:07 no file can appear without we putting it there :) Mar 06 17:46:12 yeah I was referring to your repository, not to original nokia work :) Mar 06 17:46:26 well I'll just go for autotools next time then ;) Mar 06 17:46:30 mhm Mar 06 17:46:37 meanwhile I learned some more about libtool that night, so all is fine :) Mar 06 17:46:44 :) Mar 06 17:47:07 (I had forgotten how broken and ugly it can get, then :D) Mar 06 17:54:43 Test Mar 06 17:55:06 Please somebody let me know if you can read me - I need to test my client. Mar 06 17:55:25 Enrico___: I can Mar 06 17:55:34 Thank you. Mar 06 17:56:31 Ok, my question: I have a semi-bricked N900. Thanks to Pali and NIN101, I have booted it with rescue OS from sd card. Mar 06 17:56:42 I am able to access it from my Mac by telnet. Mar 06 17:57:06 And also by graphic interface if I enable mass storage mode. Mar 06 17:57:24 In the latter mode, I can see the MyDocs directory. Mar 06 17:57:30 But by telnet I can't. Mar 06 17:57:36 So where is it? Mar 06 17:58:47 Enrico___: /dev/mmcblkNp1 Mar 06 17:58:59 where N is either 0 or 1 Mar 06 17:59:16 Ok let me try. Thank you. Mar 06 17:59:19 you should mount it for it to be visible Mar 06 17:59:30 check with sfdisk -l Mar 06 17:59:49 I have mounted it. Mar 06 18:00:03 cool Mar 06 18:00:24 I find the following directory: /mnt/maemo/ Mar 06 18:00:37 Where all maemo seems to be. But no trace of MyDocs. Mar 06 18:01:00 do sfdisk -l Mar 06 18:01:11 mydocs is usually ~28gigabytes big Mar 06 18:02:16 With sfdisk -l I end up with many errors. Mar 06 18:02:40 try sfdisk -l /dev/mmcblk0 Mar 06 18:02:47 and sfdisk -l /dev/mmcblk1 Mar 06 18:03:32 unless you have badblocks which could be the reason of semibrickage Mar 06 18:03:56 Yes I see four partitions. Mar 06 18:03:57 Now? Mar 06 18:04:14 do you see one that is ~28gigs big? Mar 06 18:04:39 I think it's the first. Mar 06 18:05:02 It's exactly 28315648 blocks. Mar 06 18:05:18 if you type: mount |grep yourpartition, is it mounted? Mar 06 18:05:43 Let me try. Mar 06 18:06:12 Enrico___: what are you trying to achieve? Mar 06 18:07:00 No, I was just looking around and trying to get familiar with the system. I would like to stard Maemo. Mar 06 18:07:04 *start Mar 06 18:07:16 you can;t start it from rescueos Mar 06 18:07:18 I am trying to avoid reflashing the device. Mar 06 18:07:35 ok, but what for you need rescueos? Mar 06 18:07:37 if you cannot boot maemo in normal way, then it is broken and you have to flash it Mar 06 18:08:02 unless you know what is broken and how to fix it :) Mar 06 18:08:51 I know that I have messed something in the file structure. I was just cleaning MyDocs from a wrong expansion of some directory. Mar 06 18:09:08 broken MyDocs should not prevent maemo from booting Mar 06 18:09:52 I know I changed something like the boot.scr file. Mar 06 18:10:07 But I don't remember where it is supposed to be. Mar 06 18:10:14 freemangordon: may cause ages to complete boot properly, waiting for fsck nad stuff tho =) Mar 06 18:10:38 if maemo does not boot => reflash device Mar 06 18:10:44 Enrico___: but, as Pali said - you'd better flash it, both rootfs and eMMC Mar 06 18:10:45 boot.scr isnt in mydocs? Mar 06 18:10:50 doing all other things is vasting of time Mar 06 18:10:53 unless you have some weird bootloader installed Mar 06 18:11:52 No boot.scr in mydocs. Or better: I think I substituted it with something else, wrong, while trying to install NITDroid. Mar 06 18:12:01 ah Mar 06 18:12:12 Sorry, I know I am doing quite crazy things, I am just trying to understand. Mar 06 18:12:14 so, what you can do is to flash the kernel only Mar 06 18:12:23 it will replace u-boot Mar 06 18:12:31 but you will be able to boot to maemo Mar 06 18:12:43 where you can reinstal u-boot Mar 06 18:13:17 with flashing only the kernel you will not lose anything besides u-boot Mar 06 18:13:29 if you have already uboot, then you can boot maemo via "run attachboot" Mar 06 18:13:51 But I think Maemo's kernel is ok. If I try from u-boot the first option, boot attached OS, it tries to start. Mar 06 18:13:53 Pali: does that work if boot.scr is missing? Mar 06 18:13:59 yes Mar 06 18:14:01 ok Mar 06 18:14:05 Then it sits down and after a while it reboots. Mar 06 18:14:18 Enrico___: then you have to flash it, sorry Mar 06 18:14:46 It's not a big problem - I don't have data inside. Mar 06 18:15:02 I am just trying to avoid it if it is not necessary. Mar 06 18:15:06 then just flash it, boot rootfs and eMMC Mar 06 18:15:10 *both Mar 06 18:15:39 first eMMC and second COMBINED Mar 06 18:15:56 Enrico___: yes, I see, but it doesn;t make much of a sense if there is no data you want to keep Mar 06 18:16:13 And I have a small problem: I reflashed one time from a Windows computer (installing the flasher on my Mac is a problem). It was very easy. But now the Windows doesn't see the device. Mar 06 18:16:57 On the Mac, at a very low level (I think), it appears. Tried the equivalent of lsusb (it's a quite complicated command on a Mac). And the device appears. Mar 06 18:17:48 Now I have to understand if it is worth trying to install the flasher-3.5 on the Mac (I have some indications on how to do that, but it is complicated) and try to flash from there. Mar 06 18:19:33 What do you think about this? Mar 06 18:22:05 get some cheap old linux laptop Mar 06 18:22:14 with working usb ports Mar 06 18:23:31 It's an idea. But anyway, if the device is seen by lsusb it should be possible to flash it, right? Mar 06 18:24:06 depends on which stage it is, if it started to boot (time frame is <0.5s) it will change id Mar 06 18:24:24 that's why you start flasher first, then connect device which was fully powered off Mar 06 18:26:01 Yes, but I am able to enter update mode by holding u and starting the device. I think it enters correctly: I see the usb icon on the top right. Mar 06 18:26:42 I just don't understand why Win doesn't see it anymore and was wondering whether it be sufficient for the Mac (or for a Linux OS) to see it at low level. Mar 06 18:45:45 ~lazyflashing Mar 06 18:45:45 lazyflashing is, like, http://wiki.maemo.org/Updating_the_tablet_firmware#The_Lazy_Approach Mar 06 18:47:50 for linux. And yes, when you see device in lsusb and you have a working flasher and image, it is possible to reflash N900 Mar 06 18:48:12 usually, except for some very strange fringe cases Mar 06 19:02:08 because win and mac has slow usb enumeration which depends on couple of factors Mar 06 19:02:37 when I was testing speed of enumeration I saw that only libusb0.1 on linux is usable for userspace... Mar 06 19:02:48 pali, does holding u during boot stops the device into update mode? Mar 06 19:03:06 only if host starts enumeration Mar 06 19:03:21 and needs to do some data transfer Mar 06 19:03:49 iirc pressing U just extend delay for additional fixed internal Mar 06 19:04:11 *interval Mar 06 19:10:18 Do you mean it's not a good idea to try with the Mac? Mar 06 19:10:38 can your mac run linux? Mar 06 19:11:29 I don't know. To me it seems quite a big work to get linux running. I was thinking about trying the flasher on the Mac, just to see if it sees the device. Mar 06 19:11:45 The problem is installing the flasher. It's not that easy. Mar 06 19:12:03 Enrico___: not installing, but booting from live cd Mar 06 19:12:11 or flash drive Mar 06 19:12:47 May you suggest a Linux distribution suitable for this? I admit I never really used Linux. Mar 06 19:12:58 ubuntu Mar 06 19:13:19 * freemangordon hides :) Mar 06 19:13:34 Enrico___: every distro will do the job Mar 06 19:13:43 as long as it boots Mar 06 19:13:55 Ok, so how do I get it? Is it available from the Internet? Mar 06 19:14:02 sure Mar 06 19:14:08 Sorry for silly questions, I'm a newbie to this. Mar 06 19:14:33 yep, got it Mar 06 19:14:52 what's the problem with installing mac flasher? Mar 06 19:15:22 Enrico___: https://www.ubuntu.com/download/desktop Mar 06 19:15:29 did apple introduce friggin app signatures meanwhile? Mar 06 19:15:30 The problem is that the program, in older Macs, used to go to /usr/lib/ Mar 06 19:15:52 Enrico___: also, see the not at the lower right corner re macs Mar 06 19:16:01 Now on my El Capitan OS /usr/lib is protected by System Integrity Protection. Mar 06 19:16:03 *note Mar 06 19:16:24 aah that's how they call that crap Mar 06 19:16:27 freemangordon: thanks! I will have a look at it! Mar 06 19:16:33 DocScrutinizer05: the problem is the fruit :) Mar 06 19:16:58 Enrico___: can't you select another installation destination Mar 06 19:17:01 ? Mar 06 19:17:07 on mac? Mar 06 19:17:12 why not? Mar 06 19:17:17 or just change harddrive temporarily? Mar 06 19:17:18 I cannot from the installer. Mar 06 19:17:18 never touched it Mar 06 19:17:19 no idea Mar 06 19:17:26 The installer just does not work. Mar 06 19:17:33 But I can extract the files. Mar 06 19:17:38 can't you just extract the files? Mar 06 19:17:49 A man suggested me how to do that on stackexchange. Mar 06 19:17:55 I guess you doubleclick the flasher.blabla file and it starts installing Mar 06 19:18:01 ok, then, place the binay in /usr/local/lib or whatever suitable Mar 06 19:18:19 which is not crap protected Mar 06 19:18:23 Ok, I was just wondering about this solution. Mar 06 19:18:44 Do you think that /usr/local/lib would let the flasher work? Mar 06 19:18:59 freemangordon: I *guess* that security crap also will stop access to USB Mar 06 19:19:13 I see no reason why not, but keep in mind I have never ever touched mac Mar 06 19:19:42 DocScrutinizer05: makes sense Mar 06 19:20:21 Enrico___: what I would sugges is - start downloadin ubuntu iso, while it is downloading, try to extract and run flasher in /usr/local/lib Mar 06 19:20:38 or /usr/local/bin? Mar 06 19:20:42 if you fail, write that iso on dvd and boot from it Mar 06 19:21:07 There's an answer on stackexchange about SIP which mentions this location. Mar 06 19:21:23 Enrico___: can't help, sorry, you have to try Mar 06 19:21:24 http://apple.stackexchange.com/questions/193368/what-is-the-rootless-feature-in-el-capitan-really Mar 06 19:23:06 hmm, "you can turn SIP off." Mar 06 19:23:21 "Currently this requires rebooting into recovery mode and running the command csrutil disable" Mar 06 19:23:44 DocScrutinizer05: not that bad ^^^ Mar 06 19:26:13 freemangordon, how do i start porting? I know there is a long way to go... Mar 06 19:27:01 just read it Mar 06 19:27:13 so they have an 'open mode' ;-P Mar 06 19:27:15 ravelo: install devuan under a VM, then try to build and run hildon-desktop from fremantle-gtk2 repo on github Mar 06 19:27:21 otherwise: Mar 06 19:27:23 ~aegis Mar 06 19:27:23 http://www.developer.nokia.com/Community/Wiki/Harmattan:Developer_Library/Developing_for_Harmattan/Harmattan_security/Security_guide , or "The purpose of this framework is: ... to make sure that the platform meets the requirements set by third party software that requires a safe execution environment.", or http://en.wikipedia.org/wiki/Trusted_Computing#Criticism, or http://en.qi-hardware.com/w/images/1/10/ME_382_LockedUpTechnology2.gif Mar 06 19:27:27 DocScrutinizer05: but it really seems to be open Mar 06 19:27:29 0.0.5 Mar 06 19:27:40 I would prefer the other alternatives than turning SIP off... Mar 06 19:27:50 yeah, please don't Mar 06 19:28:13 Enrico___: boot ubuntu then, if installing in /usr/local/bin does not work Mar 06 19:28:16 ravelo: https://github.com/fremantle-gtk2/ Mar 06 19:28:20 aiui you could "install" the flasher to any dir you like, and run it there, via sudo Mar 06 19:28:39 freemangordon, actually sounds quite simple Mar 06 19:28:50 ravelo: keep in mind you should use "gtk2" or "upstream branches Mar 06 19:29:04 sorry, afk Mar 06 19:29:09 ravelo: it is, the problem is there are a piles of packages you have to build and install Mar 06 19:29:16 *a pile Mar 06 19:30:02 we're waiting wor Wizzup/parazyd to create a repo on devuan servers with debs, but until then... Mar 06 19:30:11 Actually there still should be a wiki page about that Mar 06 19:30:24 currently there are only those threads on tmo Mar 06 19:30:43 ravelo: agree, but that means somebody should do it :) Mar 06 19:31:08 and I can;t both port and document and test and... Mar 06 19:31:17 sure Mar 06 19:31:52 ravelo: feel free to write wiki page while you're trying to build h-d and hildon-status-menu Mar 06 19:32:33 ravelo: keep in mind hildon-thumbnail and libhildonfm are WIP ATM Mar 06 19:33:05 all other packages should be more or less ready, though I remember a guys having some problems a couple of weeks ago Mar 06 19:33:11 *a guy Mar 06 19:34:18 ravelo: also, I am using devuan x64, I'd recommend you to use the same as it increases the chances to catch some bugs that will not appear on 32bits Mar 06 19:34:51 ok Mar 06 19:35:09 NeKit: do you remember what clutter env var needs to be set so hildon-desktop to start normally? Mar 06 19:35:16 yes Mar 06 19:35:21 CLUTTER_BACKEND=x11 Mar 06 19:35:43 cool :). could you make a pull request for a modified h-d startup script? Mar 06 19:36:58 here https://github.com/fremantle-gtk2/hildon-desktop/blob/gtk2-clutter-1.x/debian/hildon-desktop.init#L26 Mar 06 19:37:26 freemangordon, sry for on emore noob question. currently maemo sits on top of some debian, tightly integrated. apt-get is used to install packages. We want to port one application after another to Devuan, so the original maemo-package manager is no more used, at least at the beginning? Mar 06 19:38:03 ravelo: what is that original maemo package manager? ham? Mar 06 19:38:18 hmm, i don't know Mar 06 19:38:42 you mean hildon-application-manager UI? Mar 06 19:38:52 ok Mar 06 19:39:07 as apt-get works ok in both maemo and devuan Mar 06 19:39:14 NeKit: thanks Mar 06 19:40:30 NeKit: hmm, wait, you don;t have a way to test that, right? you're using arch? Mar 06 19:41:41 ravelo: in short - you build packages with dpkg-buildpackage, install with dpkg -i. that works in exactly the same way both under maemo and devuan Mar 06 19:42:20 yes, I can set up Devuan chroot though so I would be more useful, but testing init scripts will be hard without LXC, for example Mar 06 19:42:43 NeKit: well, if you're interested, just install a VM Mar 06 19:43:17 it is easy and native IMO Mar 06 19:43:27 you mean KVM on ARM? Mar 06 19:43:35 or on desktop? Mar 06 19:43:39 20:30 < freemangordon> we're waiting wor Wizzup/parazyd to create a repo on devuan servers with debs, but until then... Mar 06 19:43:39 on desktop Mar 06 19:44:02 there's still progress, but I'm still somewhat block on amprolla. I can do more work on this coming week(end) Mar 06 19:44:21 Wizzup: ok, I was just explaining the situation :) Mar 06 19:44:51 NeKit: all the development I've done so far is under VMWare x64 devuan on my desktop Mar 06 19:45:08 freemangordon: ack, just elaborating as well :) Mar 06 19:45:09 no need for arm at that point, unless you want to feel it :) Mar 06 19:45:57 also, I don;t want to restrict maemo to ARM only, having intel as a target is a good option IMO Mar 06 19:47:04 ravelo: also, please, if you see something broken (not compiling, installing, dependencies are missing, does not work), please raise an issue on github and eventually make a pull request with a fix Mar 06 19:47:12 I have Intel with libhybris to try though Mar 06 19:48:00 how's clutter behaving with libhybris? you said you should tweak some stuff, right? Mar 06 19:49:25 s/should/had to/ Mar 06 19:49:25 freemangordon meant: how's clutter behaving with libhybris? you said you had to tweak some stuff, right? Mar 06 19:52:20 with latest approach the problem I had is that hildon-desktop first starts in 800x600(?) window and then resizes to fullscreen Mar 06 19:56:18 NeKit: window? thats weird Mar 06 19:56:44 that might be the default resolution for clutter_stage Mar 06 19:58:56 or maybe it is matchbox default resolution Mar 06 19:59:07 though I don;t see such effect here Mar 06 19:59:12 part of clutter/cogl initialization, right Mar 06 20:08:26 freemangordon, ok Mar 06 22:04:27 ~netsplit Mar 06 22:04:27 rumour has it, netsplit is when two IRC servers lose their link with each other, thus isolating the users on each side. A normal part of all IRC networks. An orchestra of poips and thwoops, or something which occurs frequently on OPN. Mar 07 02:21:12 Might be nice if IRC networks didn't need to be trees. **** ENDING LOGGING AT Tue Mar 07 03:00:02 2017