Change a4fea05a43b8fe57b9f43afeb6ed6bd88601c053, Reduce verbosity while reloading

[Milian Wolff]

On Friday 17 August 2012 15:33:40 Aleix Pol wrote:
> On Wed, Aug 15, 2012 at 1:17 AM, Aleix Pol <aleixpol at kde.org> wrote:
> > On Tue, Aug 14, 2012 at 7:42 PM, Andreas Pakulat <apaku at gmx.de> wrote:
> >> Hi,
> >> 
> >> On Tue, Aug 14, 2012 at 7:32 PM, Andreas Pakulat <apaku at gmx.de> wrote:
> >>> Hi,
> >>> 
> >>> On Tue, Aug 14, 2012 at 1:31 PM, Aleix Pol <aleixpol at kde.org> wrote:
> >>>> On Tue, Aug 14, 2012 at 8:24 AM, Andreas Pakulat <apaku at gmx.de> wrote:
> >>>>> Hi,
> >>>>> 
> >>>>> just read my irc backlog and noticed that commit (I'm not reading
> >>>>> commit mails anymore). I'm wondering about two things:
> >>>>> 
> >>>>> a) currently it looks like the reload-questions is not shown anymore
> >>>>> when the document is dirty. That sounds wrong, since it would mean the
> >>>>> user as unsaved changes and the document is reloaded, so he can loose
> >>>>> data
> >>>> 
> >>>> It's dirty in the file-system, the flag modified in the IDE is
> >>>> Modified.
> >>>> Also IDocument::reload is triggered when you call the reload action.
> >>>> 
> >>>>> b) Is this really the right way? This function is also called when
> >>>>> switching between documents, which means that with the change one
> >>>>> could loose saved changes now by simply switching through tabs or so.
> >>>> 
> >>>> Switching tabs doesn't and shouldn't call IDocument::reload()
> >>>> I made this patch because it doesn't make sense to me that when we
> >>>> explicitly reload a file to see what changed in the file system, we
> >>>> get a dialog asking if we are sure we want that.
> >>>> Yes, that's why I'm reloading in the first place.
> >>> 
> >>> After double-checking the actual code I take back my comments and
> >>> agree with you. The patch is completely ok. Thanks for working on
> >>> that.
> >> 
> >> Only thing lacking  (after some more digging through the code) is that
> >> File->Reload to reload just the current document will still show the
> >> warning. But thats not fixable by us I guess since the action is
> >> contributed by katepart.
> >> 
> >> Andreas
> >> 
> >> --
> >> KDevelop-devel mailing list
> >> KDevelop-devel at kdevelop.org
> >> https://barney.cs.uni-potsdam.de/mailman/listinfo/kdevelop-devel
> > 
> > That's correct.
> > TL;DR: Patches welcome
> > 
> > Also when you change a bunch of files (for example when you git
> > checkount <branch>) you'll get all files complaining that they have to
> > be reloaded (which is why I wanted to improve "Reload All"). Now we
> > have an alternative but I'm not sure if we still want those dialogs
> > there, still. Maybe we could add a "Reload All" option to that dialog?
> > (which belongs to kate, AFAIR).
> > 
> > I'm unsure what's the best way to solve this, but I'm sure we want to
> > improve the control we have there. It has a big impact in our user
> > experience.
> > In any case, why does KatePart even throw dialogs at us? Shouldn't be
> > the host application responsible to handle those cases by providing
> > all the needed information?
> > 
> > Aleix
> 
> What about reloading automatically files modified in disk that aren't
> modified in the IDE and that are not the current document? (I think
> the dialog is fine for the current document)

No! (see Andreas' comment on accidental git checkout -f which sometimes can be 
rescued by having the files still open)

> This would solve most of our problems there and also would make us not
> to need the "Reload All" action.
> It's not that we open documents to preserve their state...
> 
> Aleix
-- 
Milian Wolff
mail at milianw.de
http://milianw.de
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 198 bytes
Desc: This is a digitally signed message part.
URL: <http://mail.kde.org/pipermail/kdevelop-devel/attachments/20120817/f4e134e5/attachment.sig>