void IBasicVersionControl::showXXXX()

I think majority of us agreed that showXXXX()  should be
void, since
it is totally async. When using this method, the caller
doesn't need
to take care of the operation -- call and forget operation.

Is there any objection that I change this into void ??

_______________________________________________
KDevelop-devel mailing list
KDevelop-develkdevelop.org
https://barney.cs.uni-potsdam.de/mailman/listinf
o/kdevelop-devel

dukju ahn wrote:
> I think majority of us agreed that showXXXX()  should
be void, since
> it is totally async. When using this method, the caller
doesn't need
> to take care of the operation -- call and forget
operation.
> 
> Is there any objection that I change this into void ??

I don't recall that consensus . My
argument for returning a job is 
you can wait on the user to finish using the GUI, e.g. if a
script needs 
to invoke a GUI method. But maybe others feel
differently...

-- 
Matthew
ESIG: .sig file not available


_______________________________________________
KDevelop-devel mailing list
KDevelop-develkdevelop.org
https://barney.cs.uni-potsdam.de/mailman/listinf
o/kdevelop-devel

On 05.06.07 12:38:50, Matthew Woehlke wrote:
> dukju ahn wrote:
> > I think majority of us agreed that showXXXX() 
should be void, since
> > it is totally async. When using this method, the
caller doesn't need
> > to take care of the operation -- call and forget
operation.
> > 
> > Is there any objection that I change this into
void ??
> 
> I don't recall that consensus . My
argument for returning a job is 
> you can wait on the user to finish using the GUI, e.g.
if a script needs 
> to invoke a GUI method. But maybe others feel
differently...

Well, maybe this should be done on a case-by-case basis, for
example
I don't see a reason to wait for showDiff(), but for
example
showCommit() might make sense to wait for. (think of doing
something
after the commit is done).

Andreas

-- 
Break into jail and claim police brutality.

_______________________________________________
KDevelop-devel mailing list
KDevelop-develkdevelop.org
https://barney.cs.uni-potsdam.de/mailman/listinf
o/kdevelop-devel

On 05.06.07 14:03:26, Matthew Woehlke wrote:
> Andreas Pakulat wrote:
> > On 05.06.07 12:38:50, Matthew Woehlke wrote:
> >> dukju ahn wrote:
> >>> I think majority of us agreed that
showXXXX()  should be void, since
> >>> it is totally async. When using this
method, the caller doesn't need
> >>> to take care of the operation -- call and
forget operation.
> >>>
> >>> Is there any objection that I change this
into void ??
> >> I don't recall that consensus . My
argument for returning a job is 
> >> you can wait on the user to finish using the
GUI, e.g. if a script needs 
> >> to invoke a GUI method. But maybe others feel
differently...
> > 
> > Well, maybe this should be done on a case-by-case
basis, for example
> > I don't see a reason to wait for showDiff(), but
for example
> > showCommit() might make sense to wait for. (think
of doing something
> > after the commit is done).
> 
> Yes, I can agree with that. Also is there a detach()
method of VcsJob 
> (i.e. something that says 'run, and I don't care about
the result, clean 
> yourself up')?

No, and how should it delete itself? deleteLater in a
private slot
connected to finish()? I don't like that. After all the user
explicitly
requests a VcsJob from the version control interface so he's
responsible
for properly deleting it.

Andreas

-- 
Your talents will be recognized and suitably rewarded.

_______________________________________________
KDevelop-devel mailing list
KDevelop-develkdevelop.org
https://barney.cs.uni-potsdam.de/mailman/listinf
o/kdevelop-devel