maemo.org - Talk

maemo.org - Talk (https://talk.maemo.org/index.php)
-   Applications (https://talk.maemo.org/forumdisplay.php?f=41)
-   -   [M5] [Announce] Qalendar -- a replacement for the default calendar (https://talk.maemo.org/showthread.php?t=90820)

sixwheeledbeast 2013-09-04 13:00

Re: [Announce][Fremantle] Qalendar 0.6.1 -- a clone of the default calendar
 
Quote:

Originally Posted by deryo (Post 1371944)
Hi gidzzz.

I found a minor bug in the GUI: When you leave the calendar open showing week view and wait for the next day, the column of that day is being highlighted correctly but not its head. See attached image.

Hmm, yes I confirm this too. Seems it relates to my bug report above
http://talk.maemo.org/showpost.php?p...4&postcount=58

OVK 2013-09-04 13:12

Re: [Announce][Fremantle] Qalendar 0.6.1 -- a clone of the default calendar
 
Sorry if this has been already mentioned/asked but would it be possible to add a possibility to mark a calender entry "private" (for Exchange calender)?

int_ua 2013-09-08 09:25

Re: [Announce][Fremantle] Qalendar 0.6.1 -- a clone of the default calendar
 
Quote:

Originally Posted by OVK (Post 1372013)
Sorry if this has been already mentioned/asked but would it be possible to add a possibility to mark a calender entry "private" (for Exchange calender)?

AFAIU you can add private calendar and just use it for your private events.

iceskateclog 2013-09-08 11:58

Re: [Announce][Fremantle] Qalendar 0.6.1 -- a clone of the default calendar
 
Got segmentation fault with Qalendar.

Reproduced with removing qalendar from dbus file, killling qalendar process, installing qalendar-dbg.

I just launch qalendar from terminal, it opens in week mode and just swipe to three next weeks, windows closes with segmentation fault.

Here's (maybe useful) output:

Code:

$ dpkg -p qalendar   
Package: qalendar
Priority: extra
Section: user/office
Installed-Size: 676
Maintainer: Grzegorz Gidel <grg.gdl@gmail.com>
Architecture: armel
Version: 0.6.1-1
Depends: calendar-backend (>= 0.6-20+0m5), libc6 (>= 2.5.0-1), libgcc1 (>= 1:4.2.1), libqt4-core (>= 4.7.0~git20100909-0maemo1+0m5), libqt4-dbus (>= 4.5.3~git20090723-0maemo4+0m5), libqt4-gui (>= 4.5.3~git20090723-0maemo6+0m5), libqt4-maemo5 (>= 4.6.2~git20100224-0maemo1+0m5), libqt4-xml (>= 4.5.3~git20090723-0maemo4+0m5), libstdc++6 (>= 4.2.1-4maemo9+0m5)
Size: 222042
Description: An open source clone of the default Maemo calendar application
Homepage: http://talk.maemo.org/showthread.php?t=90820
Maemo-Display-Name: Qalendar


Code:

$ gdb qalendar
GNU gdb (GDB) 6.8.50.20090417-debian
Copyright (C) 2009 Free Software Foundation, Inc.
License GPLv3+: GNU GPL version 3 or later <http://gnu.org/licenses/gpl.html>
This is free software: you are free to change and redistribute it.
There is NO WARRANTY, to the extent permitted by law.  Type "show copying"
and "show warranty" for details.
This GDB was configured as "arm-linux-gnueabi".
For bug reporting instructions, please see:
<http://www.gnu.org/software/gdb/bugs/>...
(gdb) run
Starting program: /usr/bin/qalendar
Detected language: "en_US"

Program received signal SIGSEGV, Segmentation fault.
0x4023e356 in QWidget::frameGeometry() const ()
  from /usr/lib/libQtGui.so.4
0x4023e356 <_ZNK7QWidget13frameGeometryEv+6>:        ldr        r2, [r3, #8]
(gdb) bt
#0  0x4023e356 in QWidget::frameGeometry() const ()
  from /usr/lib/libQtGui.so.4
#1  0x00033520 in WeekHintProfile::widgetDnComparator (
    w1=<value optimized out>, w2=0x258950)
    at ../../src/week/WeekHintProfile.h:103
#2  0x00033de4 in std::__insertion_sort<__gnu_cxx::__normal_iterator<ComponentWidget**, std::vector<ComponentWidget*, std::allocator<ComponentWidget*> > >, bool (*)(ComponentWidget*, ComponentWidget*)> (__first=..., __last=...,
    __comp=0x334f4 <WeekHintProfile::widgetDnComparator(ComponentWidget*, ComponentWidget*)>)
    at /scratchbox/compilers/cs2007q3-glibc2.5-arm7/bin/../l---Type <return> to continue, or q <return> to quit---
ib/gcc/arm-none-linux-gnueabi/4.2.1/../../../../arm-none-linux-gnueabi/include/c++/4.2.1/bits/stl_algo.h:2382
#3  0x00033ee0 in std::__final_insertion_sort<__gnu_cxx::__normal_iterator<ComponentWidget**, std::vector<ComponentWidget*, std::allocator<ComponentWidget*> > >, bool (*)(ComponentWidget*, ComponentWidget*)> (__first=..., __last=...,
    __comp=0x334f4 <WeekHintProfile::widgetDnComparator(ComponentWidget*, ComponentWidget*)>)
    at /scratchbox/compilers/cs2007q3-glibc2.5-arm7/bin/../lib/gcc/arm-none-linux-gnueabi/4.2.1/../../../../arm-none-linux-gnueabi/include/c++/4.2.1/bits/stl_algo.h:2462
#4  0x0002fda0 in WeekWidget::populate (this=0x24a3d8)
---Type <return> to continue, or q <return> to quit---
    at /scratchbox/compilers/cs2007q3-glibc2.5-arm7/bin/../lib/gcc/arm-none-linux-gnueabi/4.2.1/../../../../arm-none-linux-gnueabi/include/c++/4.2.1/bits/stl_algo.h:2868
#5  0x00032850 in WeekWidget::reload (this=0x24a3d8)
    at ../../src/week/WeekWidget.cpp:141
#6  0x00033070 in WeekWidget::setDate (this=0xbeedf460,
    date=...) at ../../src/week/WeekWidget.cpp:54
#7  0x0002c0a8 in WeekPlug::setDate (this=0x248188,
    date=...) at ../../src/week/WeekPlug.cpp:91
#8  0x0002c28c in WeekPlug::gotoNextWeek (this=0x248188)
    at ../../src/week/WeekPlug.cpp:118
#9  0x0006d090 in WeekPlug::qt_metacall (this=0x248188,
---Type <return> to continue, or q <return> to quit---
    _c=QMetaObject::InvokeMetaMethod, _id=2,
    _a=<value optimized out>) at moc_WeekPlug.cpp:79
#10 0x40902148 in QMetaObject::metacall(QObject*, QMetaObject::Call, int, void**) () from /usr/lib/libQtCore.so.4
#11 0x4090be4e in QMetaObject::activate(QObject*, QMetaObject const*, int, void**) () from /usr/lib/libQtCore.so.4
#12 0x4090be4e in QMetaObject::activate(QObject*, QMetaObject const*, int, void**) () from /usr/lib/libQtCore.so.4
Backtrace stopped: previous frame identical to this frame (corrupt stack?)
(gdb)

Anything more I can do?

gidzzz 2013-09-08 18:39

Re: [Announce][Fremantle] Qalendar 0.6.1 -- a clone of the default calendar
 
Quote:

Originally Posted by ade (Post 1370836)
The last measurement is after bringing Qalendar to be foreground the next day. As I see qalendar partly beeing moved to swap (which explains less ram used), the total amount of memory used seem to have shrunk. Any viable explanations (some memory reclaim by OS?)? Well, it definitely took away some worries I had about it :)

I do not know what to think of it. I did a similar test with smem, but the results were somewhat different. It took about 4 hours and USS was generally stable at around 7.5 MB, but I have noted one sudden dive to 4 MB, and back to 7.5 MB 15 minutes later. Swap was 0 all that time.

Your 15 MB also seems a bit high compared to my results, especially considering that the views are populated on demand (so no events should be loaded if the window was never shown after restarting). I've got a schedule of some movie festival, which contains 453 entries, most of which are in the same week, and with all views (month, week, agenda) open on that week plus 60 tasks and 2 notes, USS was 12 MB.

Maybe those tools are just not completely reliable? I have no idea, never used them before.



Quote:

Originally Posted by deryo (Post 1371944)
You'd be my true hero if you implemented the possibility to delete single occurrences of repeating events as well.

The recurrence editor is not finished yet. One of the things that are waiting to be added is support for exceptions. It will allow to "delete" single occurrences in the editor, but currently I do not know if there will be something like a quick shortcut in the context menu.

Quote:

Originally Posted by deryo (Post 1371944)

The calendar database seems to have a place to store categories, but from what I have read in that bug report, they will not be synchronized, if that's what you're also after (Qalendar is not the thing which does the sync). Anyway, I added it to my todo.

Quote:

Originally Posted by deryo (Post 1371944)
- When an alarm rings give the user not only the option to remind him 5 minutes later, but let him specify the amount of time for that reminder. (I don't know if this is even possible, as I fear that alarm dialog is built into maemo. Is it?)

Yes, the ringing alarm dialog is not a part of Qalendar.

Quote:

Originally Posted by deryo (Post 1371944)

I don't get it. I thought an alarm for a task can be set in both Calendar and Qalendar.

Quote:

Originally Posted by deryo (Post 1371944)
And I want to participate! It seems there is no german translation available. I don't need it, but I would love to provide and maintain it. Please tell me how to do that!

The translations are managed at https://www.transifex.com/projects/p/qalendar/, so the easiest way is to create an account there, request a language and translate using the online editor or Qt Linguist.



Quote:

Originally Posted by sixwheeledbeast (Post 1372012)
Hmm, yes I confirm this too. Seems it relates to my bug report above
http://talk.maemo.org/showpost.php?p...4&postcount=58

The relation is that they are the outcome of nonexistent time-based autorefreshing. I have not done anything about it for today's update, because I feel like I need to give the mechanics behind it some more thought and on the other hand I did not want to wait any longer to publish the fix for the crash.



Quote:

Originally Posted by OVK (Post 1372013)
Sorry if this has been already mentioned/asked but would it be possible to add a possibility to mark a calender entry "private" (for Exchange calender)?

I'd do it as int_ua says. I don't think it is possible to mark a single event as private without changing more than Qalendar.



Quote:

Originally Posted by iceskateclog (Post 1372864)
Anything more I can do?

Nope, it is already fixed. Thanks for the backtrace.



Qalendar 0.6.2 has just arrived. Nothing big, just more polish and less bugs.
  • Show calendar type in event and task details.
  • Scroll to and highlight the selected calendar when opening the calendar picker.
  • Treat event end dates as non-inclusive, which fixes misinterpretation of some events as multi-day.
  • Event time icons display only the starting time for zero-length events.
  • Added a note that the time in the alarm picker is the amount of time before the event.
  • Pop up a notice when trying to go to previous/next event and it is not possible.
  • Fixed unexpected looping when trying to go the next event from the last existing event for the second time.
  • Fixed outdated scroll hints and possible crash when no non-all-day events are present in the week view.
  • Fixed zero-length events at midnight not appearing in the week view.
  • Updated translations.

iceskateclog 2013-09-09 05:37

Re: [Announce][Fremantle] Qalendar 0.6.2 -- a clone of the default calendar
 
Yay \o/ Qalendar 0.6.2 doesn't crash.

Pizza funded :>

OVK 2013-09-09 07:59

Re: [Announce][Fremantle] Qalendar 0.6.1 -- a clone of the default calendar
 
Quote:

Originally Posted by int_ua (Post 1372842)
AFAIU you can add private calendar and just use it for your private events.

Unfortunately this does not cover my use cases.

You see, I am allowed to have "hidden" meetings during my work time. In my organisation all calendars are shared to everyone (helps in finding suitable times for internal meetings etc.). However, all of my meetings can't be seen by everyone (some of our clients require this), yet my collegues need to see that I am reserved. Hence the need for "private" appointments: people can see that I am reserved but do not see why. This is how my organisation works.

int_ua 2013-09-16 00:06

Re: [Announce][Fremantle] Qalendar 0.6.2 -- a clone of the default calendar
 
Can we have an option to add some time to the default task due date? I would love due-less tasks, but AFAIU it's not possible (checked ~/.calendar/calendardb, even -1 gives year 2106) so I thought that default due date being today +5 years would suffice :)

jackburton 2013-09-16 10:43

Re: [Announce][Fremantle] Qalendar 0.6.2 -- a clone of the default calendar
 
Quote:

Originally Posted by int_ua (Post 1374436)
Can we have an option to add some time to the default task due date? I would love due-less tasks, but AFAIU it's not possible (checked ~/.calendar/calendardb, even -1 gives year 2106) so I thought that default due date being today +5 years would suffice :)

In maecaltasks, I made it so the year 2037 indicates 'no due date'. Press the red X to toggle the year.
http://farm6.staticflickr.com/5443/9...db956c12ca.jpg
mct-no_due_date by brander snaxe, on Flickr

gidzzz 2013-09-16 14:25

Re: [Announce][Fremantle] Qalendar 0.6.1 -- a clone of the default calendar
 
Quote:

Originally Posted by OVK (Post 1373054)
You see, I am allowed to have "hidden" meetings during my work time. In my organisation all calendars are shared to everyone (helps in finding suitable times for internal meetings etc.). However, all of my meetings can't be seen by everyone (some of our clients require this), yet my collegues need to see that I am reserved. Hence the need for "private" appointments: people can see that I am reserved but do not see why. This is how my organisation works.

Now I think I see what you mean by "private", but it is not Qalendar's domain either.



Quote:

Originally Posted by int_ua (Post 1374436)
Can we have an option to add some time to the default task due date? I would love due-less tasks, but AFAIU it's not possible (checked ~/.calendar/calendardb, even -1 gives year 2106) so I thought that default due date being today +5 years would suffice :)

Actually -1 is more like 1969, because calendar-backend API uses time_t, which in our case is 32-bit signed int. Year 2106 is the effect of Qt's function to convert time_t to QDateTime having unsigned int as the argument.

But back to the point... I agree that it can be annoying to be forced to change the date each time or end up with a ton of overdue tasks in no time, but just adding some predetermined value is not what I like either. However, I am not sure what is the best way. Some ideas:


1a) Give the user an option to explicitly define the amount of time to add.

An obvious extension of your proposal, but it would require stuffing that option somewhere in the menus and I'd prefer not to clutter them up unless there's no other way.


1b) Remember the last offset used when creating a new task.

I like it for being simple and transparent, but I am not sure how (in)convenient it would be in the reality. If all your tasks are supposed to be due-less, than it acts like 1a, but otherwise I am afraid it might cause a bit of confusion when trying to add a task and seeing a default date that seemingly comes out of nowhere.

Luckily, the mentioned problem could be mitigated by something else I am planning to add anyway: a "New task" button next to the "New event" button in the day view (at least, possibly other views too). The offset for task deadline would be loaded and saved only when invoking the editor without a date context, i.e. from the tasks view.


1c) Hidden option.

Same as 1a, but hidden as a value in the config file. It would be a simple hack in about one line of code and invisible to those who do not want to see it, so I could add it in the next update if a long-term solution is not ready by then.


2a) A special distant date

But I'd rather go for 2038, most likely 2^31-1 seconds since 1970 UTC, because the whole 2037 is still completely valid. I think this would be as close to due-less as calendar-backend permits. Such tasks would not have their date displayed in Qalendar.

@jackburton: I installed MaeCalTasks to see exactly how it works, but If the screenshot is from 1.0.0-5, then I am doing something wrong or there's a bug, because I don't see the buttons with "X", "Today" and "Tomorrow".

I also have a question: was 2037 just simpler to use? Or is it to allow also the stock calendar to set such date? I guess It would be nice if Qalendar was compatible with MaeCalTasks in this matter, but I see no definitive reason to choose 2037 over 2038.

Or is that X button just a shortcut to toggle the year and 2037 tasks do not get any kind of special treating (2b)?


2b) Just a distant date

There would be a shortcut to set the date to 2037 and that's it. I think it would feel like a quick hack, so I do not really like it.


2a is quite tempting, but I am conservatively leaning towards 1b, as it does not make up any special meaning for the data. Any thoughts?


All times are GMT. The time now is 09:09.

vBulletin® Version 3.8.8