IRC log for #maemo-ssu on 20151029

06:17.47*** join/#maemo-ssu _rd (~rd@p5B2C7F05.dip0.t-ipconnect.de)
07:02.22*** join/#maemo-ssu amiconn (~amiconn@rockbox/developer/amiconn)
08:14.27*** join/#maemo-ssu freemangordon_ (~ivo@213.222.56.174)
08:50.42*** join/#maemo-ssu LauRoman (~LauRoman@5-14-1-85.residential.rdsnet.ro)
10:09.11*** join/#maemo-ssu jonwil (~jonwil@27-33-80-219.tpgi.com.au)
10:16.57jonwilhi
11:38.12*** join/#maemo-ssu NishanthMenon (nmenon@nat/ti/x-ljjhcnqizgiaoahh)
13:42.49*** join/#maemo-ssu sparetire_ (~sparetire@unaffiliated/sparetire)
13:44.42*** join/#maemo-ssu ZetaR (~user@2602:47:3fb:fc00:d996:1aa1:88fc:42a8)
16:23.18*** join/#maemo-ssu Pali (~pali@Maemo/community/contributor/Pali)
16:49.58*** join/#maemo-ssu RedW (~redw@89-76-164-87.dynamic.chello.pl)
16:59.39*** join/#maemo-ssu Pali (~pali@Maemo/community/contributor/Pali)
18:26.17*** join/#maemo-ssu futpib (~futpib@176.104.194.78)
20:07.56Sicelobencoh: seems after you talked about video playback issues, my N900 got jinxed :p
20:07.59SiceloOct 29 22:06:57 fremantle mafw-dbus-wrapper[7898]: dspvdec2: eos timed out after 10258 ms
20:08.03SiceloOct 29 22:06:57 fremantle mafw-dbus-wrapper[7898]: dspvdec2: eos took 10275 ms
20:08.15Sicelonow playing videos on my N900 is quite uphill task
20:09.35bencohoO
20:09.56bencohdid you try something funny?
20:09.59keriolol
20:11.36Sicelomaybe it just doesn't like some of the videos. simple ones play fine
20:12.06*** join/#maemo-ssu M4rtinK (~M4rtinK@ip-89-177-124-111.net.upcbroadband.cz)
20:42.30*** join/#maemo-ssu M4rtinK2 (~M4rtinK@ip-89-177-124-111.net.upcbroadband.cz)
20:54.22*** join/#maemo-ssu jonwil (~jonwil@27-33-80-219.tpgi.com.au)
20:55.23jonwilhi
21:11.03freemangordonjonwil: HI!
21:11.06freemangordonops
21:11.11freemangordonsorry for the caps
21:11.17jonwil:)
21:11.50freemangordonjonwil: doesn't look good https://github.com/community-ssu/osso-systemui-modechange/blob/master/osso-systemui-modechange.c#L76
21:12.22jonwilThats just what hexrays showed me...
21:12.36freemangordonsore, but still doesn't look good
21:12.40freemangordon*sure
21:12.44jonwilok
21:12.53jonwilif you want to fix it, be my guest :)
21:13.06freemangordonthat condition doesn't make much of a sense
21:13.10freemangordonI will
21:13.24freemangordoncould you share the DB?
21:13.24jonwilanyhow,
21:15.37jonwilhttps://drive.google.com/file/d/0B9idqO9KygGsZ1YzeTY0Umo5dTg/view?usp=sharing
21:16.03jonwilOther than fixing that, do you have a particular target in mind to reverse engineer/clone next?
21:16.39freemangordonsystemui daemon?
21:16.54freemangordonI think that's the last man standing
21:17.06freemangordonsystemui stuff that is
21:17.40jonwilnope, we still need osso-systemui-actdead, osso-systemui-devlock, osso-systemui-emergency and osso-systemui-splashscreen
21:17.46jonwilalthough we have -dev packages for some of those
21:18.13jonwil-emergency makes calls to some stuff in librtcom-call-ui that we have no documentation of whatsoever
21:18.29freemangordonhmm, right
21:19.06freemangordonjonwil: BTW, why did you remove "fprintf(stderr, "Special key listener %i\n", event->keyval);" from modechange_key_press_event_handler?
21:19.23jonwilnot sure why I didn't put the fprintfs in there
21:19.26jonwilI cant remember
21:19.35jonwilfeel free to put them back if you think they are important :)
21:19.45jonwilnot that systemui ever runs with an actual terminal :)
21:19.55freemangordonyeah
21:19.59freemangordonI'll see
21:20.00jonwilMe, I am trying to decide whether I should work on libcodelockui or whether I should keep working on figuring out all the interfaces (dbus calls, gconf keys etc) the WiFi subsystem touches and from there which ones we would need to still support if we want to rip the whole thing out
21:20.39freemangordonI gues libcode... will be easier, ain;t?
21:21.42jonwilnot necessarily, the CluiCodeDialogPrivate struct is giving me some trouble
21:22.01freemangordonif you share the DB, I can help you with that one
21:26.11jonwilhttps://drive.google.com/file/d/0B9idqO9KygGsS0NEWFJleWxycWc/view?usp=sharing
21:26.45jonwilThe functions marked with an x have yet to be reverse engineered
21:26.58freemangordonok
21:27.00jonwiland the 2 marked with a z have some weirdness I just cant work out
21:28.08freemangordonhmm, I don;t see z-marked functions
21:28.43jonwilshould be zcodelock_password_change and zcodelock_response_signal
21:29.22freemangordongboolean __cdecl ycodelock_change_passwd(const gchar *old_passwd, const gchar *new_passwd)
21:29.36jonwilhmmm ok, maybe I sent you the wrong db
21:29.43freemangordonlooks like
21:32.51jonwilwait you are looking at codelock_change_passwd and not codelock_passwd_change
21:32.55jonwilI dont know why I named them so close
21:33.26jonwildo you see zcodelock_passwd_change now?
21:33.45freemangordonI see gboolean __cdecl xcodelock_password_change(CodeLockUI *ui, CodeLockChangeFunc *func)
21:34.10freemangordonno 'z' functions at all
21:34.20jonwilweird
21:34.39freemangordonI ordered by function name, the last one is yutils_dbus_send
21:35.14jonwiltry https://drive.google.com/open?id=0B9idqO9KygGsUl9tRWZwdVNBd1E
21:36.20freemangordonyeah, I see them now
21:36.55freemangordonhmm, I see what you mean
21:37.11freemangordonI guess those are gtk_dialog_run() result shecks
21:39.58jonwilI cant remember now why I marked zcodelock_response_signal as having wierd stuff
21:40.14jonwiloh wait yeah it wasn't picking up the _GLOBAL_OFFSET_TABLE_ stuff properly
21:40.19jonwilbut then I think I found a solution to that
21:40.27Sicelobencoh: same video plays just fine on non-HD-codec N900. hmm
21:40.30jonwili.e. how to manually point IDA at the correct variable in each case
21:41.06freemangordonjonwil: I don;t see that private struct used
21:41.47jonwil_CluiCodeDialogPrivate is referenced by the priv member of _CluiCodeDialog
21:42.04jonwiland is used by a number of the x labeled functions
21:42.25freemangordonok
21:42.43jonwilaha I see
21:42.47jonwilI created one called CluiCodeDialogPrivate and one called _CluiCodeDialogPrivate
21:42.59freemangordonhmm, why?
21:43.02jonwilcant remember
21:43.06jonwilmust have been a mistake
21:43.17jonwilI think the code may be using CluiCodeDialogPrivate and not _CluiCodeDialogPrivate
21:43.26jonwileither way the structure needs reverse engineering :)
21:43.29*** join/#maemo-ssu _rd (~rd@p5B2C7F05.dip0.t-ipconnect.de)
21:49.59jonwilhey, its not like reverse engineering is an exact science :P
21:59.06jonwilanyhow, back to fiddling with WiFi I guess for now
22:03.04freemangordonjonwil: that priv structure contains an array of (i think 9) GtkWidget *, which are code entry buttons
22:03.14freemangordonand a couple of more buttons
22:04.22jonwilthat doesn't account for all of the 21 members it has
22:04.54freemangordonor those buttons are 12, lemme check
22:05.21freemangordonnever used device lock) :)
22:05.55bencoh10 digits, Done, Clear
22:05.56freemangordonanyway, you have several gtkwidgets(buttons) one text entry and some other members
22:06.07bencohs/clear/backspace/
22:06.13freemangordonok, 12 then
22:06.52freemangordonjonwil: xclui_code_dialog_build inits most of the provate struct
22:07.00freemangordon*private
22:07.04jonwilok
22:08.47freemangordonneeds sleep, night guys
22:08.58jonwilcya
23:42.04*** join/#maemo-ssu jonwil (~jonwil@27-33-80-219.tpgi.com.au)

Generated by irclog2html.pl Modified by Tim Riker to work with infobot.