Ok, probably an easy sugar-build bug, will take a better look as soon as I'm back at a keyboard :)<br><br>On Wednesday, 19 March 2014, Walter Bender <<a href="mailto:walter.bender@gmail.com">walter.bender@gmail.com</a>> wrote:<br>
<blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">On Wed, Mar 19, 2014 at 12:00 PM, Daniel Narvaez <<a href="javascript:;" onclick="_e(event, 'cvml', 'dwnarvaez@gmail.com')">dwnarvaez@gmail.com</a>> wrote:<br>
> And would<br>
><br>
> 2.5) osbuild run<br>
<br>
Yes. Repeated osbuild runs don't adversly impact the profile. Just<br>
check sugar makes a mess of things.<br>
<br>
><br>
> work properly? If so, it sounds most likely like a sugar-build bug, check<br>
> should use a completely separated profile.<br>
><br>
><br>
> On Wednesday, 19 March 2014, Walter Bender <<a>walter.bender@gmail.com</a>> wrote:<br>
>><br>
>> On Wed, Mar 19, 2014 at 11:43 AM, Daniel Narvaez <<a>dwnarvaez@gmail.com</a>><br>
>> wrote:<br>
>> > Went back and reread your email... It's actually intentional that check<br>
>> > starts from a clean state every time, why is that a problem? Also why do<br>
>> > you<br>
>> > think gsettings are not working properly in run?<br>
>><br>
>> Maybe the problem is only in the cpsection code. But try the following:<br>
>><br>
>> (1) osbuild run<br>
>> (2) set nick in cpsection/aboutme<br>
>> (3) osbuild check sugar<br>
>> (4) osbuild run<br>
>> (5) check user name<br>
>><br>
>> The user name will be Walter Bender, which is the last name in the<br>
>> series of set_nick calls in check.<br>
>><br>
>> -walter<br>
>> ><br>
>> ><br>
>> > On Wednesday, 19 March 2014, Walter Bender <<a>walter.bender@gmail.com</a>><br>
>> > wrote:<br>
>> >><br>
>> >> On Wed, Mar 19, 2014 at 11:30 AM, Daniel Narvaez <<a>dwnarvaez@gmail.com</a>><br>
>> >> wrote:<br>
>> >> > On a phone so I can't look very deeply into this right now... Maybe<br>
>> >> > we<br>
>> >> > are<br>
>> >> > using the in memory gsettings backend... Can you check if dconf is<br>
>> >> > installed<br>
>> >> > inside the broot?<br>
>> >> ><br>
>> >> > ./osbuild shell<br>
>> >> > rpm -qa | grep dconf<br>
>> >><br>
>> >> dconf-0.18.0-2<br>
>> >><br>
>> >> ><br>
>> >> ><br>
>> >> > On Wednesday, 19 March 2014, Walter Bender <<a>walter.bender@gmail.com</a>><br>
>> >> > wrote:<br>
>> >> >><br>
>> >> >> In trying to fix a problem with sugar/tests/test-user-profile.py I<br>
>> >> >> have bumped into a strange situation regarding gsettings. Perhaps<br>
>> >> >> you<br>
>> >> >> can comment?<br>
>> >> >><br>
>> >> >> The bug is that the test does not preserve/restore the original<br>
>> >> >> settings, so I added a setup and teardown to do just that.<br>
>> >> >><br>
>> >> >> But what I am finding is that reading from gsettings (or the model<br>
>> >> >> in<br>
>> >> >> cpsection/aboutme) doesn't give me the same values when using check<br>
>> >> >> vs<br>
>> >> >> run.<br>
>> >> >><br>
>> >> >> Well, not exactly. I can set values in the control panel and they<br>
>> >> >> are<br>
>> >> >> preserved between sessions. But these values are not seen by check.<br>
>> >> >> So<br>
>> >> >> new default values are set. I suspect that run is getting its data<br>
>> >> >> from gconf and that the gsettings are not actually being set<br>
>> >> >> properly<br>
>> >> >> in either check or run.<br>
>> >> >><br>
>> >> >> Where are the gsettings stored by sugar-build? I cannot find them.<br>
>> >> >><br>
>> >> >> thx.<br>
>> >> >><br>
>> >> >> -walter<br>
>> >> >><br>
>> >> >> --<br>
>> >> >> Walter Bender<br>
>> >> >> Sugar Labs<br>
>> >> >> <a href="http://www.sugarlabs.org" target="_blank">http://www.sugarlabs.org</a><br>
>> >> ><br>
>> >> ><br>
>> >> ><br>
>> >> > --<br>
>> >> > Daniel Narvaez<br>
>> >> ><br>
>> >><br>
>> >><br>
>> >><br>
>> >> --<br>
>> >> Walter Bender<br>
>> >> Sugar Labs<br>
>> >> <a href="http://www.sugarlabs.org" target="_blank">http://www</a></blockquote><br><br>-- <br>Daniel Narvaez<br><br>