On Friday 23 October 2009 21:03:36 ext Gena wrote:
On Tuesday 20 October 2009 19:12:34 André Pönitz wrote:
There have been some changes in the area today, and by now I
can't reproduce the issue anymore. Could you please check
(after the 12 hour synchronization delay) whether it works for
On Oct 20, 2009, at 7:24 PM, ext Bravo.Alex wrote:
Why is it dead:
a) it’s obsolete, we now have a new XYZ plugin supporting
QTestLib in Qt Creator
b) the author is working on something else and doesn’t have
time/interest to work on it
c) it didn’t do much, so it’s not
ext Damien Fagnou wrote:
Hi ,
we use a lot of Lua in our company , we have some syntax highlight
definitions that include lua keywork and some of our own extension. and
the editor we current use for that is using the highlight definition to
do the auto complete [just dictionary style ]
On Monday 26 October 2009 11:34:11 André Pönitz wrote:
...
I just tried to reproduce it but failed. Can you please post a recipe
to reproduce the problem?
Andre'
Emmm... There were no special actions. Just rebuilt, set breakpoint and started
the debugger. I've just tried today's snapshot
On Monday 26 October 2009 12:29:21 ext GenaCid wrote:
On Monday 26 October 2009 11:34:11 André Pönitz wrote:
...
I just tried to reproduce it but failed. Can you please post a recipe
to reproduce the problem?
Andre'
Emmm... There were no special actions. Just rebuilt, set breakpoint
Hi,
I can see that it is a smart option to have one shortcut for
comment/uncomment. But I would prefare the option to set two different
shortcuts, one for comment and one for uncomment. E.g. when I wand to
uncomment a block, then I have to mark the exact block, if I mark a line
too much (e.g.
On Monday 26 October 2009 13:18:20 Poenitz Andre (Nokia-D-Qt/Berlin) wrote:
On Monday 26 October 2009 12:29:21 ext GenaCid wrote:
On Monday 26 October 2009 11:34:11 André Pönitz wrote:
...
I just tried to reproduce it but failed. Can you please post a recipe
to reproduce the problem?
Hi,
1) There is a bug (in current snapshots too) with parsing code like the
following:
c.f()(x)(y)(z);
But, this is legal c++.
2) C++0x patterns like y = {1,2}; (using initializer lists) are also parsed
as syntax errors.
Is C++0x support planned in the short term? I heard that Qt is also
On Monday 26 October 2009 15:18:20 André Pönitz wrote:
Refering to Message-ID: 200910261429.21544.gena...@inbox.ru:
There is something strange in the log: Your project does not seem
to be linked to libdl, so the dumpers can not be loaded using dlopen.
Could you please try to disable custom
This isn't down to Creator.
This all depends on the toolchain that Creator has been setup to use - GCC,
MinGW or MSVC's compiler (nmake?).
Check that the toolchain you are using supports C++0x.
On Sat, Oct 24, 2009 at 11:03 AM, Onay Urfalioglu onay.urfalio...@gmail.com
wrote:
Hi,
1) There
Out of interest, have you managed to reproduce my breakpoint issues?
(where breakpoints cannot always be set unless gdb is actually
running.) This issue got a little better with the latest nightly
Windows build but is still giving me problems.
2009/10/26 André Pönitz andre.poen...@nokia.com:
On
On Monday 26 October 2009 14:33:36 ext Robert Caldecott wrote:
Out of interest, have you managed to reproduce my breakpoint issues?
I was not able to reproduce it after the first change I mentioned
last week.
(where breakpoints cannot always be set unless gdb is actually
running.) This
OK, re-creating this is easy for me. I am running Qt Creator 1.2.93
rev 97e7b7bbf9 (Oct 21).
1. Start Qt Creator.
2. Create a new Qt4 GUI project (MainWindow based).
3. Open main.cpp and use F9 to add a breakpoint on line 8.
4. Hit F5. Program builds, gdb starts and the breakpoint is triggered.
On Mon, Oct 26, 2009 at 03:46:21PM +, Robert Caldecott wrote:
OK, re-creating this is easy for me. I am running Qt Creator 1.2.93
rev 97e7b7bbf9 (Oct 21).
that's stone age. please try something *current*.
___
Qt-creator mailing list
Eh? That's the latest Windows version of Qt Creator I can find on
your FTP snapshot server.
2009/10/26 Oswald Buddenhagen oswald.buddenha...@trolltech.com:
On Mon, Oct 26, 2009 at 03:46:21PM +, Robert Caldecott wrote:
OK, re-creating this is easy for me. I am running Qt Creator 1.2.93
Oha, good point. There will be new ones up in a few hours (unless
compilation breaks :).
ext Robert Caldecott schrieb:
Eh? That's the latest Windows version of Qt Creator I can find on
your FTP snapshot server.
2009/10/26 Oswald Buddenhagen oswald.buddenha...@trolltech.com:
On Mon, Oct
On Mon, Oct 26, 2009 at 04:09:26PM +, Robert Caldecott wrote:
Eh? That's the latest Windows version of Qt Creator I can find on
your FTP snapshot server.
well, that means that our nightly build system is kinda lagging. from
the developers' pov it is still *wy* outdated when it comes
I'll look for an update first thing in the morning...
2009/10/26 Oswald Buddenhagen oswald.buddenha...@trolltech.com:
On Mon, Oct 26, 2009 at 04:09:26PM +, Robert Caldecott wrote:
Eh? That's the latest Windows version of Qt Creator I can find on
your FTP snapshot server.
well, that
In article 20091026161823.ga1...@troll08.nokia.trolltech.de,
Oswald Buddenhagen oswald.buddenha...@trolltech.com wrote:
On Mon, Oct 26, 2009 at 04:09:26PM +, Robert Caldecott wrote:
Eh? That's the latest Windows version of Qt Creator I can find on
your FTP snapshot server.
well,
They tend to appear in spurts - I assume it's a manual process. There's not
been a mac version for some time now.
On Mon, Oct 26, 2009 at 5:22 PM, Stephen Chu step...@ju-ju.com wrote:
In article 20091026161823.ga1...@troll08.nokia.trolltech.de,
Oswald Buddenhagen
On Monday 26 October 2009 18:31:09 ext Danny Price wrote:
They tend to appear in spurts - I assume it's a manual process. There's not
been a mac version for some time now.
Actually it's automated. Though we had never a release where we didn't
massively change the way we create packages on at
Plus that it's not so high in the list of things to do.
If you are talking about qtcrea...@nokia team, it makes a lot of sense.
But that's why you guys opened development to outside world, right?
So that we can help you improve already great product.
So, back to qtestlibplugin:
After
1) There is a bug (in current snapshots too) with parsing code like
the following:
c.f()(x)(y)(z);
But, this is legal c++.
Curious question of the day, what does the above line of code do?
I've never seen a C++ construct like that before!
c.f() returns a function pointer or a
I like the idea of a dedicated QTest output pane - will you be able to
double-click on failures to jump directly to the code? That would be
brilliant.
2009/10/26 Bravo.Alex alex.br...@igt.com
Plus that it's not so high in the list of things to do.
If you are talking about qtcrea...@nokia
Clicking on failures is probably the main reason to integrate QTestLib with the
IDE.
Filtering different types of test results and generating a stub implementation
for a tested class would be other features of this plugin.
There's Entry 254483 in Task Tracker that mentions this feature. It
Honestly I don't know if I want to set the precedent of allowing
plugins to just arbitrarily add new tabs there.
I think Build Issues would be a valid output target, though; it's
already intended for jumping to the relevant places in the code and
one could argue that failing a unit test is a
Hi Onay,
On Oct 24, 2009, at 1:03 PM, ext Onay Urfalioglu wrote:
Hi,
1) There is a bug (in current snapshots too) with parsing code like the
following:
c.f()(x)(y)(z);
Oops, well it's parsed correctly but it looks like I forgot to handle this case
in the type checker :) I will try to fix it
Adam,
Build Issues would be a valid output target
What happens when the test itself has build issues - which is very common due
to changes in the code being tested?
Are you saying that running of test can be considered part of building?
Aren't building and running separate things according
On Mon, Oct 26, 2009 at 2:50 PM, Bravo.Alex alex.br...@igt.com wrote:
Adam,
Build Issues would be a valid output target
What happens when the test itself has build issues - which is very common due
to changes in the code being tested?
Are you saying that running of test can be considered
Hello,
I searched the documentation and the web without finding any hint or
solution to the following quest. I use the ctrl-click on methods a lot
to peek at a method declaration or definition, but I couldn't find an
easy way to get back to where I clicked on the method.
Eclipse has this
If running is part of building, then how do run the test several times (most
likely to step through it with debugger)? Do you also build several times?
That's one of the reasons building is separate from running - you build once
and then run as many times as you want.
-Original
Running the unit tests is conceptually part of building your
application. You don't have to clean and recompile your entire
application every time you make a little change; why is this any
different? Running the Test step doesn't have to recompile the whole
mess if it's already built (and,
If I have to debug my unit tests then usually something smells about the
test already - normally each test should be rather simple and be
stand-alone runnable (i.e. test B should not have to rely upon a run of
test A); so in most cases I should end up with some really small piece
of code where
Adam, you are exactly correct.
Then following your logic, if test is run, but not recompiled,
then there's no build stage happening and no running of a test.
(if running of test is part of building).
I think we are just talking about different things here.
Let's clarify the scenario we are
Thank you. It works great, thank you.
I guess it is documented, but I couldn't find it in the QtCreator
documentation. Do you have a link to the documentation page for these
features ? I would like to check and learn other shortcuts. I searched
in Qt Creator's documentation.
Brian McGillion
Hi,
I am not sure which version they appeared in, but if you use the snapshot
from
ftp://ftp.trolltech.com/qtcreator/snapshots/latest/
they will be there. As regards documentation I am not sure where the
latest is.
Br,
Brian
Thank you. It works great, thank you.
I guess it is documented,
On Mon, Oct 26, 2009 at 4:05 PM, Bravo.Alex alex.br...@igt.com wrote:
Adam, you are exactly correct.
Then following your logic, if test is run, but not recompiled,
then there's no build stage happening and no running of a test.
(if running of test is part of building).
I think you missed my
Mar wrote:
Hi,
I can see that it is a smart option to have one shortcut for
comment/uncomment. But I would prefare the option to set two different
shortcuts, one for comment and one for uncomment. E.g. when I wand to
uncomment a block, then I have to mark the exact block, if I mark a line
38 matches
Mail list logo
