On Jan 27, 2008 8:52 AM, William Skaggs wrote:

One of the points that emerged from the ongoing discussion is that it is time to start making a tentative roadmap for 2.8. I hope this time that Peter and his group can be brought into the discussion as early as possible, and will play a role in shaping the plan.

As I wrote earlier, Peter in his LGM talk listed the things he considered the top priorities at the UI level:

10. better printing support 9. painting with brushes
8. improve the text tool
7. save for web
6. organize brushes, palettes, gradients in categories 5. avoid pop-up dialogs
4. better painting tools
3. layer trees
2. adjustment layers
1. single window interface

Of these items, #1 is controversial, and #2 and #3 depend on Gegl developments. Each of the remaining things are potential targets for 2.8 -- in fact, I don't see any reason why it wouldn't be possible to do several of them. In particular, I think it should be possible, and very cool, to implement the "blobs of paint" idea that is part of #4.

I thought that #4 was dependent on (the use of, but not the overall adoption of) Gegl. Perhaps I've misunderstood you. At the moment, we do have some useful GEGL ops that could be used to back the 'blobs of paint' idea.

At a casual glance, most of the approachable ops would work best for mask adjustment:
* range adjustment (gimp-levels
* halo adjustment (gimp-curves)
* hardening (gimp-threshold)
* blurring
* inversion (good for stenciling, ie. layer masks) * noise

There are some that work with colors: * colorbalance
* colorify
* white balance

Can we use the transform ops (rotate/flip/etc)? That would be very nice. Initially, I figure we will only make the brush data available to ops, so it could
be done with a linear stack of Ops. After a while, you might want to combine brushes, so that might develop into a normal tree.

Actually, technically it would start as a tree, looking like this:

brush-output (sink) |
+- add_alpha
|
++--- last_mask_op
| |-- some_mask_op
| +-brush-mask (source)
+--- last_color_op
|-- some_color_op
+- brush_color (source)

If we wanted to combine blobs of paint, that would amount to re-separating the mask and color, and connecting the output to the newly added blob.

More advanced things, such as spatially-dependent paint (think of the clone tool in aligned mode), would have to wait until GIMP was using GEGL for images, so that a brush could know what region was requested in terms of the image coords.

I think 'blobs of paint' would be a good place to experiment with GEGL graphing UI, that could later be refactored to be more generally usable for GIMP.

Hi,

On Sat, 2008-01-26 at 22:22 +0000, William Skaggs wrote:

As I wrote earlier, Peter in his LGM talk listed the things he considered the top priorities at the UI level:

10. better printing support 9. painting with brushes
8. improve the text tool
7. save for web
6. organize brushes, palettes, gradients in categories 5. avoid pop-up dialogs
4. better painting tools
3. layer trees
2. adjustment layers
1. single window interface

I think you misunderstood Peter's talk. These are the top 10 user requests, not the top priorities. It is very important to analysize user requests but it would be wrong to use this list as the basis for a discussion about the future roadmap. This should happen based on the product vision.

Sven

Sven wrote:

I think you misunderstood Peter's talk. These are the top 10 user requests, not the top priorities. It is very important to analysize user
requests but it would be wrong to use this list as the basis for a discussion about the future roadmap. This should happen based on the product vision.

right on.

top-10 (as felt by the developers) user requests of the last years.

and at the lgm we addressed them as well as we could at that time (printing? what can we say about printing? OK, a couple of things...)

--ps

founder + principal interaction architect man + machine interface works

http://mmiworks.net/blog : on interaction architecture

Sven Neumann gimp.org> writes:

I think you misunderstood Peter's talk. These are the top 10 user requests, not the top priorities. It is very important to analysize user requests but it would be wrong to use this list as the basis for a discussion about the future roadmap. This should happen based on the product vision.

Sven

So, what is the product vision?

One more question regarding the list (I know it's not a roadmap but still): Will a move to pure SDI help to get all the various dialogs to float properly above the image windows, never be lost behind them, and not have their own taskbar buttons?

Hi,

On Wed, 2008-01-30 at 12:29 +0000, Michael Grosberg wrote:

One more question regarding the list (I know it's not a roadmap but still): Will a move to pure SDI help to get all the various dialogs to float properly above the image windows, never be lost behind them, and not have their own taskbar buttons?

There will be no move to a pure SDI. And you can already have all dialogs float properly above the image window without their own taskbar buttons. Assuming of course that your window manager implements the respective window hints properly.

Sven

Hi,

On Wed, 2008-01-30 at 12:29 +0000, Michael Grosberg wrote:

So, what is the product vision?

http://gui.gimp.org/index.php/GIMP_UI_Redesign#product_vision

Sven

Prduct vision:

- GIMP is Free Software; - No problems about that;
- GIMP is a high-end photo manipulation application and supports creating original art from images; - Critical points:
- Make a native support to up to 32 bit image (CMYK, HDRI...);
- Improve GFig and paths to be a a minimal consistent vector solution to help a drawing (Not asking about a Inkscape inside Gimp.
Just the minimal support);
- GIMP is a high-end application for producing icons, graphical elements of web pages and art for user interface elements; - Critical point:
- The GFig again, it help so much in this vision topic; - GIMP is a platform for programming cutting-edge image processing algorithms, by scientists and artists; - I think python-fu and script-fu is going well, same as compiled plugins, since we have seen in the last months after 2.4 release great plugins with some good features; - GIMP is user-configurable to automate repetitive tasks; - Well, here is a oppose of above. python-fu and script-fu is not so user-configurable. If the case is the automate repetitive tasks, what
about make a UI for the python-fu or script-fu and put some of buttons and
lists just to user choice? I'll make a example of this and put in UI discussion in i'ts blog;
- GIMP is easily user-extendable, by 'one-click' installation of plug-ins.
- Again same as above. Make a extension menu like Firefox is so great idea, again I'll look at UI's blog.

Well, why dont start to work with the visions? I see it's as a great idea.

On Feb 1, 2008 1:41 PM, Isaque Profeta wrote:

Improve GFig and paths to be a a minimal consistent vector solution to help a drawing

Improving GFig probably doesn't make sense at all. Beginning with the fact that drawing of that kind should not be done in a separate window. I believe bgo has a number of shapes drawing tools related feature requests.

Alexandre

About Gfig, my idea is it, make it a tool, not a filter. Agree to you.

On Feb 1, 2008 9:03 AM, Alexandre Prokoudine wrote:

On Feb 1, 2008 1:41 PM, Isaque Profeta wrote:

Improve GFig and paths to be a a minimal consistent vector solution to

help

a drawing

Improving GFig probably doesn't make sense at all. Beginning with the fact that drawing of that kind should not be done in a separate window. I believe bgo has a number of shapes drawing tools related feature requests.

Alexandre
_______________________________________________ Gimp-developer mailing list
Gimp-developer@lists.XCF.Berkeley.EDU https://lists.XCF.Berkeley.EDU/mailman/listinfo/gimp-developer

Hi,

Well, why dont start to work with the visions? I see it's as a great idea.

That's what we are doing for one and a half year now.

Your mail pointing out where the vision is not yet implemented was really completely unneeded. We are well aware that it's still a vision and that a lot needs to be done to get there.

Sven