Review Request 120794: Implement Unique Mode Properly in Rekonq

[David Narváez]

> On Oct. 28, 2014, 6:09 p.m., Andrea Diamantini wrote:
> > src/application.cpp, line 731
> > <https://git.reviewboard.kde.org/r/120794/diff/3/?file=322475#file322475line731>
> >
> >     It's not clear to me why you chose parse() instead of process() here.
> >     Can you please argument this?
> 
> David Narváez wrote:
>     It would be equivalent to process() because the --help and --version arguments were already handled by the original application if they were present (i.e., the application would have died before requesting our activation).

Moreover, if you were to use process() here, and external entity could "shoot your application dead" if it sends --help or --version parameters through the right D-Bus call. I guess this is an important point to mention in a porting guide.


- David


-----------------------------------------------------------
This is an automatically generated e-mail. To reply, visit:
https://git.reviewboard.kde.org/r/120794/#review69318
-----------------------------------------------------------


On Oct. 29, 2014, 1:57 p.m., David Narváez wrote:
> 
> -----------------------------------------------------------
> This is an automatically generated e-mail. To reply, visit:
> https://git.reviewboard.kde.org/r/120794/
> -----------------------------------------------------------
> 
> (Updated Oct. 29, 2014, 1:57 p.m.)
> 
> 
> Review request for KDE Frameworks and rekonq.
> 
> 
> Repository: rekonq
> 
> 
> Description
> -------
> 
> This is my humble attempt to implement the Unique Mode properly. I have been trying to do this for the longest time in a way that avoids code duplication but I can't find a way to jump over all the hurdles these API impose. I tried learning from other ports from KUniqueApplication but a quick look at LXR shows there are plenty of applications that blindly ported to Unique Mode but didn't bother implementing activateRequested and the one I found that did was plasmawindowedcorona.cpp which does not need a QCommandLineParser, so the code duplication is less evident. At this point, I would like someone who knows about the QCommandLineParser + KDBusAddons dance to look at this and tell if it is reasonable or not.
> 
> The current patch just makes it possible to open several Rekonq applications. It does not do the right thing when a Rekonq window is already open in the current activity and a user clicks a link elsewhere (step 4 in the Testing Done section) because it starts a brand new Rekonq window, but that's a different patch. It also does some funky thing asking you if you want to restore the previous session when nothing has crashed, I have to check that.
> 
> 
> Diffs
> -----
> 
>   src/application.h 7ccd60d 
>   src/application.cpp c7c297d 
>   src/main.cpp 7592f7a 
> 
> Diff: https://git.reviewboard.kde.org/r/120794/diff/
> 
> 
> Testing
> -------
> 
> 1. Open one Rekonq window
> 2. Try opening Rekonq again
> 3. Try opeining Rekonq from a command line with some URLs
> 4. Assuming Reknoq is your default browser (why wouldn't it be?) click on a link somewhere (I click on the links at the title of the Konversation channels I am in)
> 5. Open rekonq from the console using rekonq --incognito
> 6. Open rekonq from the console using reknoq --webapp twitter.com
> 7. Open rekonq from the console pointing it to some local HTML file
> 
> Before this patch, nothing happens in steps 2 - 7. After a first version of this patch that does not avoid the QCommandLine parser if the argument list is not empty, the window opened at 1 crashes because the activateRequested signal passes an empty list of arguments - not even the binary name - so QCommandLine parser dies. With this patch, every step opens a new window properly, step 5 opens an incognito window and step 6 opens a webapp window (simple window). Step 7 opens the current working directory because local URL handling is broken.
> 
> 
> Thanks,
> 
> David Narváez
> 
>

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.kde.org/pipermail/kde-frameworks-devel/attachments/20141029/4d93bb75/attachment.html>