Whither Krita?

[enki]

Boudewijn Rempt wrote:
> On Monday 14 September 2009, LukasT.dev at gmail.com wrote:
>   
>> On Monday 14 September 2009 19:33:07 Boudewijn Rempt wrote:
>> Let's share my ideas.
>>
>>     
>>> MyPaint is fast, uses 16 bits/channel, supports ora, has a nice brush
>>> engine (although it's not too hard to make a paintop for krita that uses
>>> MyPaint's brush lib, it's definitely going to be _slow_).
>>>       
>> Let me know about any brush engine you like from MyPaint and I will try to
>> make it native for Krita so it will be fast enough.
>>     
>
> There are some very interesting things about mypaints brush engine. One of 
> them is that MyPaint used to be really stroke based, i.e., it would store 
> strokes, and strokes were intended to be editable at a later stage. It's still 
> possible to change the brush parameters and see the last stroke change 
> accordingly. I have got a version of the mypaint brush engine integrated in a 
> krita paintop, but never got it finished enough to allow painting. The other 
> thing about mypaints brush engine is that it works in 15 bit pre-multiplied 
> rgb.
>
>   
What makes MyPaint powerful is that all brush options are combined in 
one paintops. So for example, you can create a smudge + airbrush + 
jittering+ pixel pencil preset. Most of mypaint's features are available 
in Krita, but scattered between different paintops. I think that Brush 
options should be shared more between paintops. Sensors are a step in 
the good direction, but more paintops need to use them. That would be a 
quadruple win :
- No duplicated code.
- UI easier to use.
- Infinite possibilities of creation for the user.
- Better tablet support for all paintops.
>>> There are a
>>> couple of issues with its user interface, especially for tablet users
>>> since it relies on keyboard shortcuts a lot, but nothing really
>>> problematical. I actually use MyPaint myself more often than Krita
>>> nowadays!
>>>       
Keyboard shortcuts is what I miss the most when using Krita. Tablet 
being an absolute device, you have to make big movements to use the UI. 
Plus small widgets are hard to click. When you work ten hours on an 
image, it's frustrating that most of your movements are done for 
changing settings instead of painting.
Also, keyboard shortcuts permit to adapt the workflow to specific cases 
even when the UI is not adapted : every actions is at a distance of 1.

>> Keyboard shortcuts is something I miss in Krita a lot. I consider it a big
>>  bug when I can't configure shortcuts for Tool Docker. I investigated the
>>  issue a little, but not much...
>>     
>
> Yeah, that's one of those things that broke during our integration with 
> KOffice. It worked quite well in 1.6, but is broken right now. And 
> investigating the fix takes so much time.
>
>   
Here shortcut seems to works well, but the tools I need are not in the list.
>> I see some groups of users who would like to use Krita:
>> 1.) GIMP has ugly interface for them -- they don't like it and they like
>>  Krita 
>>  2.) KDE users who wants to replace GIMP(or GNOME based app) with
>>  KIMP, ehm...Krita (mainly to have clean KDE desktop solutions)
>> 3.) People who think we have very nice tablet support, even better then
>>  other open-source apps
>> 4.) Artists (who does not care about what libs or language is Krita written
>>  in :), digital painters, matte painters, concept art painters
>> 5.) People hungry for various colorspaces
>>
>> Every group has different use cases. Maybe would be cool to write down
>>  these use-cases?
>>     
6.) Beginners and casual users will also prefer Krita because they can 
paint polygons, squares and circles without having to master selections 
tools, and because the UI is easier to understand.
>
> Well, only 4 and 5 can give us real use cases. But we've tried to write down 
> usecases for Krita before. That didn't help a lot -- but then, I was sick at 
> the time and not involved, which didn't help either. But right now, I am most 
> interested in finding out what we need and what we have in common.
>
>   
I started to write something, but it's far from being finished. I think 
it should aim at several angle of view.
who, what media, what formats, normal and extreme uses, variety of 
techniques used, abolutely needed features vs. "just better" features.
>>> It also makes sense to identify the really big problems in Krita, both in
>>> relation to that vision (so that will have to wait) and problems we know
>>> about right now, like:
>>>       
>> As I don't know about the vision, 
>>     
>
> That's because there isn't one, really.
>
>   
>> I will share ideas later in discussion.
>>
>> Anyway so far we are trying to be [Gimp+MyPaint] and I think it is good way
>> but we need more developers :( I like the way to be [Gimp+MyPaint] for KDE.
>> Developers will not come in the first hardest part of the development when
>> things crashes etc..
>>
>> We need users so that we can motivate the developers to come. If you notice
>> last year just two developers came in -- me and dimitry and we were
>> motivated by GSoC. Also Vera seems to be promising these days.
>>     
>
> We're integrating her stuff in Krita now. It needs a lot of work, really, 
> since it's very much a learning process for Vera, but the ui ideas she brings 
> are very sound.
>
>   
>> This imply that we need better support for community. E.g. Krita forum
>>  would help a lot I think.
>>     
>
> For that, it would be good to be more independent. But then, we need someone 
> to manage all that. Attempts at getting krita.org started up have faltered 
> because, for instance, all that web content management stuff is too hard for 
> me. Heck, I cannot really do stuff with the krita pages at koffice.org 
> anymore! Maybe we should look at how mypaint does this.
>
>   
Yeah in others apps, it's the community of users that do that. But I 
think it's too soon.Things will happens naturally when Krita will be 
more stable and more polished.
>> So far Krita is development platform for me as I can test many algorithms
>>  in it.
>>     
>
> That is something you and Cyrille have in common at least :-)
>
>   
>>  C++ and Qt are my favourite as I learnt them through the time I
>>  contribute to Krita. And Krita allows me to use them as wide as I want.
>>     
I think that the best point. Even if others programs were 10 times 
better than Krita, Krita would still be needed, because there will 
always be C++/Qt/KDE programmers.