All,
Working further with kmymoney, I have run across a build failure due to a
missing icon: kmymoneytitlelabel.png. The failure occurs at the end of the build
during the packaging stage:
/usr/bin/install -c -p -m 644
/build/tde-kmymoney/src/kmymoney/kmymoney2/widgets/$file
/build/tde-kmymoney/pkg/tde-kmymoney/opt/trinity/share/apps/kmymoney2/pics; \
done
/usr/bin/install: cannot stat
'/build/tde-kmymoney/src/kmymoney/kmymoney2/widgets/kmymoneytitlelabel.png': No
such file or directory
The icons is not in the tde git tree:
15:23 phoinix:/dat_e/tde/tde/main/applications/kmymoney> find . -name
kmymoneytitlelabel.png
<nothing>
The build system looks for the icon here:
kmymoney2/widgets/Makefile:689:WIDGET_PNGS = kmymoneytitlelabel.png
kmymoney2/widgets/Makefile.am:54:WIDGET_PNGS = kmymoneytitlelabel.png
kmymoney2/widgets/Makefile.in:689:WIDGET_PNGS = kmymoneytitlelabel.png
Does anyone have a patch/substitute for this icon? It is apparently triggered
when building with 'libofx' features enabled:
./configure \
--prefix=${TDEDIR} \
--with-qmake=${QTDIR}/bin/tqmake \
--with-qt-dir=${QTDIR} \
--with-qt-includes=${QTDIR}/include \
--with-qt-libraries=${QTDIR}/lib \
--with-extra-includes="/usr/include/tqt:/usr/include/tqt/Qt" \
--with-extra-libs="${TDEDIR}/lib:${TDEDIR}/lib/trinity" \
--sysconfdir=${TDEDIR}/etc \
--localstatedir=/var \
--disable-dependency-tracking \
--disable-gcc-hidden-visibility \
$enable_pdfdocs \
--enable-new-ldflags \
--enable-rpath \
--enable-ofxplugin \
--enable-ofxbanking \
--enable-qtdesigner \
--disable-sqlite3 \
--enable-qtdesigner \
--disable-debug \
--disable-final \
--enable-closure
I am tinkering with enabling more of the options after Francios confirmed he
was able to build with those options enabled. The build completed fine, the
packaging of the icon is the problem.
--
David C. Rankin, J.D.,P.E.
Here is a link to a 13 MB patch that finishes renaming tdevelop components.
http://humanreadable.nfshost.com/misc/tdevelop-kdev-tdev-rename.diff
There are no build failures or errors or warnings. I don't know how to use tdevelop but my modest tests indicates the patch is stable. You folks who use tdevelop can test more thoroughly and update the patch as needed.
Darrell
All,
from today the version number of Kate syntax highlighting files has been realigned to the version number of the KDE repository, to simplify comparison for users who do not keep older copies of those files.
If a file has been edited to adapt it to TDE/fix errors/add extensions, the version number is followed by a dash and a progressive number, as suggested by Slavek a few weeks ago.
Ex:
- KDE repo file: version=1.2.3
- TDE repo file without any change: version 1.2.3
- TDE repo file with changes: version 1.2.3-1
In future, if more local changes are done, the number after the dash will be increased. If the changes come from the KDE repo, the version number will be update accordingly and the number after the dash will restart from 1.
For the same reasons, I have also reverted kateversion="2.5" to be the same as in the KDE repo.
NOTE:
Most files have a version number that is either 0.01 or 0.1 less than the version number previously available in the GIT tree.
If you previously downloaded any file to your local ~/.trinity folder, please consider either deleting or updating it accordingly to prevent old version numbers to mask new files with lower version number.
The new files have just been pushed, so they should be available from tomorrow for update using the Kate GUI update dialog. Alternatively just update from the GIT repo directly.
Cheers
Michele
All,
This is strange. After going through initial setup of kmymoney, I get a
segfault whenever I click in the account "ledger" window. I ran it in gdb and it
disclosed the following error:
17:32 valhalla:~/bld/bld/html2ps> gdb kmymoney
GNU gdb (GDB) 7.7
<snip>
Using host libthread_db library "/usr/lib/libthread_db.so.1".
reading file
start parsing file
startDocument
reading accounts
reading transactions
reading securities
reading prices
reading reports
endDocument
Program received signal SIGSEGV, Segmentation fault.
TQObject::controlElementDataObject (this=this@entry=0x0) at kernel/qobject.cpp:120
120 kernel/qobject.cpp: No such file or directory.
Here is a screenshot of the window immediately before the crash:
http://www.3111skyline.com/dl/dt/trinity/ss/kmymoney-ledger.jpg
Clicking on any of the tabs [Deposit] [Transfer] [Withdrawal] will cause the
segfault. What does the kernel/qobject.cpp error tell us? (anything? or do we
need to rebuild with debugging enabled?) The reason for the question is I don't
know if this clearly discloses a naming issue, or if it is just the start of the
debug chain that would tell us more with debugging enabled in the build. Let me
know. Thanks.
--
David C. Rankin, J.D.,P.E.
Hi Guys,
On Tuesday 18 February 2014 22:38:05 Darrell Anderson wrote:
> The primary reason I got frutrated yesterday is not my lack of C++
> skills. My frutration comes from the fact that asking for coaching
> help in this list is futile.
>
> Darrell
I belive, and I'm sure someone will correct me if I am mistaken, the
problem is that it can be very difficult to read and understand code
that has been written by a third party without documentation and/or
explanation of what a particular piece of code is supposed to do. C++
along with code resusability tends to distort perception of what is/was
intended originally. Lack of documentation within the code seems to be
one of those things that either gets ignored or overlooked by the
programmer for what ever reason.
Just my 2p's worth...
--
Best Regards:
Baron.
I want to add a couple of more points to my previous email below.
1) internationalization/translation (tde-i18n) is another area where users without any programming knowlegde can help a lot
2) the contribution a person can give to TDE does not depend on his/her skills but on the amount of time he/she is willing to put it TDE. You could be the best C++ programmer of the world, but if you only contribute 5 minutes a month your contribution would be very small indeed (but not futile though). Or you could be just a normal user without any programming experience who is willing to spend a couple of hours each week on updating documentation: in such case your contribution would definitely be bigger than the best C++ programmer of the world.
So it is not about skills, it is about time.
Feel free to pass this email to the user mailing list too (I know, I should subscribe... :) )
Cheers
Michele
>> The primary reason I got frutrated yesterday is not my lack of C++
>> skills. My frutration comes from the fact that asking for coaching
>> help in this list is futile.
> Yep - amen. I'd love to learn what is required to fix the systemd
> user session tracking issue, but for me, >at this point, to sit down
> and try to follow the c++ inheritance and polymophism through
> tdelibs/tdebase to >arrive at the points in the code that need
> attention -- is -- (looking for word stronger than >futile...checking
> kthesaurus...got it!) simply absurd, cockeyed, derisory, idiotic,
> laughable, ludicrous, > nonsensical, preposterous, ridiculous --
> to the point of making "a contribution so small as to belaughable."
> You c++ guys shell out some advise and guidance and we will be more than happy to help :)
As usual, different people have different opinions, so I will give mine too.
1) Number of bugs in bugszilla
As a developer, I prefer to have "10 bugs filed and one fixed" than "1 bug filed and 1 fixed". It means we are doing a good job with testing (and that we have too little active developers :( ).
Having a list of bugs that are almost all resolved, does not mean the project is better. It means we are not testing well at all. Luckily, we have people like Darrell, David, Francois, Alex and others that constantly report any issue they find. Maybe the issue will be resolved 2 years later, but at least we know it exists. So please keep up with filing bugs :)
2) Applications in the tree
The more we have, the better. Not because we want to show to the world how many beautiful applications we have (the first comment against that would be that most of them are 3-4 years old at least), but because when we need *that* particular application for *that* particular job in *that* particular day, the application is already available in our DE. If we don't have *that* application, we need to start looking around for an application doing *that* particular job, and Murphy's laws dictates that when we need *that* particular application, we are usually already short in time :)
3) Contribution to TDE
While I understand very well the frustration of Darrell, I strongly disagree with the comments marked as ">>" and ">" at the beginning of this email. No contribution to TDE, however small it could be, is ever going to be considered "futile" or "looking for word stronger than futile...checking kthesaurus...got it!) simply absurd, cockeyed, derisory, idiotic, laughable, ludicrous, nonsensical, preposterous, ridiculous -- to the point of making a contribution so small as to be laughable."
Please remember that TDE carries on thanks to the effort of us all, not just the (few active) C++ developers.
There are tons of things that need attention.
- testing: very important, so please keep up with it :)
- documentation review: don't need to be a C++ developer to edit docs. Darrell has been doing an outstanding job with it lately. Would be good if others could help him out too.
- website review: same as for the documentation, don't need to be a C++ developer to contribute to it.
- development: yes, here you need to know some C++ :) But not all bugs are major ones, so if your skills are somehow limited, you can try to target small bugs and leave the big ones to the other developers.
TDE lives thanks to the effort of us all. Please don't give up, whether you are testing, fixing paper cut issues, reviewing docs, fixing major problems or whatever else you are doing to contribute to it.
Cheers
Michele
Hi Guys,
On Tuesday 18 February 2014 22:38:05 Darrell Anderson wrote:
> The primary reason I got frutrated yesterday is not my lack of C++
> skills. My frutration comes from the fact that asking for coaching
> help in this list is futile.
>
> Darrell
I belive, and I'm sure someone will correct me if I am mistaken, the
problem is that it can be very difficult to read and understand code
that has been written by a third party without documentation and/or
explanation of what a particular piece of code is supposed to do. C++
along with code resusability tends to distort perception of what is/was
intended originally. Lack of documentation within the code seems to be
one of those things that either gets ignored or overlooked by the
programmer for what ever reason.
Just my 2p's worth...
--
Best Regards:
Baron.
All, tdehw guys,
Checking through my logs I note a consistent error from
[tde_dbus_hardwarecontrol. The sequence of errors occurs on each login (several
times):
Mar 01 21:40:15 valhalla org.trinitydesktop.hardwarecontrol[180]:
[tde_dbus_hardwarecontrol] Name request failed with error 'Rejected send
message, 2 matched rules; type="method_call", sender=":1.18" (uid=1000 pid=1597
comm="tdepowersave [tdeinit] ")
interface="org.trinitydesktop.hardwarecontrol.Power" member="CanHibernate" error
name="(unset)" requested_reply="0"
destination="org.trinitydesktop.hardwarecontrol" (uid=0 pid=1621
comm="/opt/trinity/bin/tde_dbus_hardwarecontrol ")'
Mar 01 21:40:15 valhalla org.trinitydesktop.hardwarecontrol[180]:
[tde_dbus_hardwarecontrol] Not primary owner (-1), exiting!
Mar 01 21:40:15 valhalla dbus[180]: [system] Activated service
'org.trinitydesktop.hardwarecontrol' failed: Launch helper exited with unknown
return code 1
I'm not sure if this is an 'error' or just a normal 'test for capabilities',
not available, fail, situation?
To me it looks like dbus org.trinitydesktop.hardwarecontrol.Power checks
member "CanHibernate" and finds the name "(unset)". It checks that it is 'Not
primary owner', then throws a 'failed: Launch helper exited with unknown return
code 1'.
I want to know if this is normal behavior or if there is a problem. On all old
3.5.13-sru hal installs I have the option to suspend/hibernate as a logout
option. With R14 (same environment) I do not have that option available. Is that
expected?
--
David C. Rankin, J.D.,P.E.
All,
Just FYI, I've noticed a big drop in server responsiveness since the new tools
were installed (updated bugzilla, etc.) Not of immediate concern, but I wonder
if new features (like bugzilla automatically searching for all related bugs,
etc.) may be putting a big strain on the server. I don't know who looks at this,
or if there are any stats available, but it is something to keep an eye on in
the future.
It's working fine, just slow, so there is no immediate need for a fix, but if
anybody has access to the box, it would be worth a check to see what the load is
and where there is anything in the logs that points to a problem.
--
David C. Rankin, J.D.,P.E.