Re: [Libreoffice-ux-advise] FilePicker reworked

Hi Cedric!

Am Donnerstag, den 10.11.2011, 09:33 +0100 schrieb Cedric Bosdonnat:
> Hi Christoph,
>
> On Wed, 2011-11-09 at 23:00 +0100, Christoph Noack wrote:
> > Sounds very good to me ... just a small question: What do you mean by
> > places. Is it more the "kind of data source" (smb vs. cmis vs. file
> > system), or something like hierarchical paths within a data source. How
> > many of such places might exist (if its the former). I'm currently think
> > about what the "root" element in such a file dialog should be ... 
>
> I was thinking it more in the GNOME way with a list of items like:
>
> Fixed items:
>   * Recently used (could replace the Recent documents menu item)
>   * Local File System
>   * Home
>
> Server items:
>   * FTP Server
>   * WebDav Server
>   * CMIS Repo
>   * SMB share

Okay, so "our" file picker will (might) replace the "platform specific"
file pickers, if server stuff is used, right? So we need to provide at
least the same interaction quality, I guess. Until today, the file
pickers were rather a "well, better than nothing" fallback.

> > By the way, I tried to find the specification for the OOo file dialog
> > rework ... without success. But I've noticed that Caolan helped to bring
> > the Gnome file dialogs "into" OOo:
> > http://people.redhat.com/caolanm/file-chooser-spec.sxw
>
> Sure and that one is great.... but only for GNOME users :)

Hehe, I rather thought that Caolan might have some good input. Since it
is now known that he worked on the file picker, he can't escape ;-)))

> > And (for maybe further use), I've found and interesting blog posting
> > showing a number of different file dialogs:
> > http://martin.iturbide.com/?page_id=935
>
> Great link, I think it could help. The only file-pickers providing
> almost all the features I was mentioning are the GNOME and Windows 7
> ones.

Mmh, although I really like to have a really consistent solution ...
combining e.g. local file handling and other sources. If people use the
CIFS functionality, how likely is it that they also require the other
sources (likeliness = importance, usage rate). I'd like to know whether
a separate solution might make sense (I'm sure you've thought about
that.)

Thank you Cedric!

Cheers,
Christoph