https://bugs.kde.org/show_bug.cgi?id=374798
Jan Kundrát <j...@kde.org> changed:
What |Removed |Added
----------------------------------------------------------------------------
Resolution|--- |WONTFIX
Status|UNCONFIRMED |RESOLVED
--- Comment #6 from Jan Kundrát <j...@kde.org> ---
(In reply to Robert Kratky from comment #5)
> The problem with the round trip through the browser is that I have my
> desktop environment configured in a way that only a certain limited subset
> of MIME types is opened through the browser when I click on them. For
> example, I use an image viewer to open images by default. So, when I click
> on a URL that points to an image in Trojita, the file is opened by the image
> viewer.
>
> Now, that particular helper app that's used to open a specific type of file
> may or may not have the functionality to save that file to a disk (for
> example, my GIF viewer isn't capable of doing that). So, I'm stuck with two
> options:
>
> * copy the URL from Trojita and use wget or similar
> * copy the URL from Trojita and open in a browser
>
> Both are possible but both are annoying and the 'round trip' has gotten a
> lot more complicated than just clicking on a link.
The DE that I use (Plasma) has a switch which controls whether an URL is always
opened in a web browser, or whether the MIME type is guessed/checked (I don't
actually know, maybe it uses its generic network stack to download or HTTP HEAD
it first) and then an appropriate local application is used for its handling.
Perhaps you can use a similar feature as well?
I believe that it's the job of the "rest of the desktop environment" to provide
a wrapper which gets triggered when a user "just clicked a web link" to ask
what they intend to do. I don't want to manage filetype associations within
Trojita, either; these are bits which we simply delegate to the DE.
> I don't see that many examples of this in my workflow, but wouldn't it be
> possible for Trojita to time out in such cases and report something like
> 'Unable to access'? I don't know how Trojita handles HTML content in
> messages, but the fact that it's capable of displaying it properly rendered
> means that it must be capable of downloading arbitrary stuff from the
> Internet, or not?
Yes, the libraries which we use can be asked to download stuff from the
Internet. However, this is limited to e-mail display, there's no extra UI, no
download resuming and what not. There's no user-facing download manager tied to
the HTTP stack.
> That's a good point, but couldn't this be hidden, for example, behind the
> systray icon? And I don't mean progress bars or something like that. Just a
> note that Trojita is trying to access so and so. Just an idea.
It can be done for sure. But I do not think that this /should/ be done within a
MUA.
--
You are receiving this mail because:
You are watching all bug changes.
