XnView Software

What is the most important features to have a good version (better than current linux&macosx version)? It's to know which code to write
For me, it's batch operation (processing, rename, timestamp, iptc) + slideshow

Hi Pierre

I know this isn't what you are asking, but the most important thing for me would be modularization of features: that is, if a GUI feature or functionality or user interface is available in one context in XnView MP, it should be available everywhere that context applies.

For example:

• The ability to display the grid anywhere an image can be displayed, including adjustment previews.
  
  The ability to invoke the multi-rename dialog from any filenaming dialog, including the simple file-renamer all the way to the batch dialog's folder controls.
  
  The ability to display full-featured thumbnails anywhere a file-list is displayed (such as batch whatever, slideshow design, etc).
  
  The full set of navigation tools available from any context where navigation might take place, including navigating through folders, categories, favorites, rankings, etc.
  
  Full viewer-like image zoom, zoom lock, and navigation anywhere a preview appears (such as any adjustment dialog, or the compare panel, or...).
  
  The ability to name and save settings from any dialog where settings might differ from time to time (such resize, canvas size, rename, DPI, and so-on)
  
  Etc.

I realize this would be initially an ambitious undertaking as functional and feature modules are isolated, identified, and implemented. But by building self-contained modularity of functions and features (with parameterized interfaces), a relatively "lightweight' implementation of XnView could result, but with a level of increased functionality and convenience no other similar software has. XnView would become an "empty" framework into which essentially any feature desired could be added.

I guess one way of looking at this is XnView as just a bundle (but a highly granular bundle) of plug-ins integrated by a generic application framework.

Thanks for listening and not laughing too hard at this!

JohnFredC wrote:I know this isn't what you are asking

I realize this would be initially an ambitious undertaking as functional and feature modules are isolated, identified, and implemented. But by building self-contained modularity of functions and features (with parameterized interfaces), a relatively "lightweight' implementation of XnView could result, but with a level of increased functionality and convenience no other similar software has. XnView would become an "empty" framework into which essentially any feature desired could be added.

Oh I try to write new code like that, for example, we can have more than 1 folder tree view already in the browser, or use the category view in view mode or ...

But my first goal is to provide a "better" version for linux & macosx users...

For me, its batch operations + Copy/Move To Folder & Crop (both from fullscreen perspective). Whatever makes fans use emulator?!

Hello,
For me the best way to do is to have a clone of the most ergonomic and advanced XnView product (the Windows version).
As the XnWiew Windows version should look like the XnViewMP win (as you said at the begining of this MP project),
The XnViewMP version for Mac or Lin*x should look like the XnViewMP for Windows.
By this way, maybe than the development (the standardizing, the tracking changes for products improvement, and so on ) will be easier for you too, this will be also easier for users with this "interoperability".
That's it for the moment, about my personal ideas for XnViewMP Un*x or Mac Versions ... !

JohnFredC wrote:...I guess one way of looking at this is XnView as just a bundle (but a highly granular bundle) of plug-ins integrated by a generic application framework....

I like this idea too.

My wishlist for XnView MP:

1.) Same user interface concept as XnView (Windows)
2.) Stability and reliability
3.) Many formats supported
4.) Unicode support
5.) Language in text files (for easy translation)
6.) Handling improvements (e.g. dashed selection, hiding/showing grid in rotation dialog)

From what I have read and heard so far, 3-5 are on their way. So only few wishes left...

XnView 1.93.6 is well designed, superior to other tools of a similar nature. Because of this, it is best not to "reinvent the horse", but instead, preferable to implement all the current fine features and GUI into the MultiOS version.

Priorities for implementation scheduling should reflect the hierarchy of tasks addressed by XnView's software category:

1. Image viewing
2. File management/image conversion
3. Basic image manipulation (resizing, contrast, gamma, etc.)
4. Basic slideshow
5. Automation (batch operations)
6. The sky's the limit

(A more appropriate response to Pierre's request than my previous one, n'est-ce pas?)

My mains Priorities for implementation scheduling:
0.) lightness & Stability
1.) Same user interface concept as XnView (Windows)
2.) Image viewing, many formats supported
3.) Slideshow, quality and fast zoom functions.
4.) Basic image manipulation (conversion, resizing, auto-contrast/level, etc.)
5.) Automation (batch operations)
6.) Unicode support and language in text files (for easy translation)
n.) ....

For me we have already:

* Same user interface concept as XnView (Windows)
* Image viewing (currently fullscreen, next, prev)
* Many formats supported
* Unicode support
* Language in text files (XML)

One thing, i've added in this version a menu style, with QT we can customize all widget, i'll try to make a good dark theme.

You can perhaps check the 'read ahead' & 'cache behind', better than xnview win

Well, this is perfect, then you have just to try to do the same things for the XnViewMP Lin*x & Mac versions !

oops66 wrote:Well, this is perfect, then you have just to try to do the same things for the XnViewMP Lin*x & Mac versions !

It's already the same, same source for all plateformes

Exactly the same source for all plateformes...
Ah, ok cool, I did not know, it was a quiproco !
Then I should have the MP unix version very soon

oops66 wrote:Exactly the same source for all plateformes...
Ah, ok cool, I did not know, it was a quiproco !
Then I should have the MP unix version very soon

Ok ok, i build it

1. Lightness, speed & stability

2. Extensive format support.
(Length of the supported formats list is simply amazing. Quality and integration of the decoders could really be improved though) isn't this the core of any viewer ?

3. Custom input device mapping. (am I missing something ? Is there really any native way to configure how you interact with the program?)

4. Slightly more polished interface (ie:thumbnails style) Perhaps making it perfect with the help of a talented ui designer instead of adding too many clumsy visual personalization options. Right now it feels like the user is given some visual customization options to compensate the lack of a proper ui design.

When I think of it, these are about the same suggestion I'd already have submitted 5 years ago.
I'm really sorry, obviously theses aren't the most exiting task to add to your todo list pierre.

instead of adding too many clumsy visual personalization options

One of the features I like best about XnView is its "visual" configurability. If everyone wanted XnView to look as much like LightRoom as possible, then I'd agree (because that would be all right with me).

But I doubt everyone does.

So I'll continue to hope that all of the existing customization (and more!!) will eventually be present in the MultiOS version, too.