From kde-kimageshop Wed Jan 05 23:24:50 2011 From: JL VT Date: Wed, 05 Jan 2011 23:24:50 +0000 To: kde-kimageshop Subject: Re: Report of Meeting: plans 2011 Message-Id: X-MARC-Message: https://marc.info/?l=kde-kimageshop&m=129426994826772 MIME-Version: 1 Content-Type: multipart/mixed; boundary="--===============1731561621==" --===============1731561621== Content-Type: multipart/alternative; boundary=20cf3054a529196efd049921b1af --20cf3054a529196efd049921b1af Content-Type: text/plain; charset=ISO-8859-1 (Most of this message was written right before Silvio added his answer =P I will complete it in the same spirit as if I hadn't read Silvio's answer) On Wed, Jan 5, 2011 at 5:49 PM, LukasT.dev@gmail.com wrote: > > > 3. > > As regards Pentalis, I would like to having him working on a new mixing > > brush engine. > > In the past, I have read he wished to work on it. > > As for me, I am confident he is the right guy to finish this task ;-) > > Needless to say, also for Pentalis there is really NO rush! > > Regarding the fact, that you resources are limited, I think maybe > somehow more wiser solution would be to support Pentalis in the Google > Summer > Of Code program (if there will be one) with this idea and if that fails, > let's > use your resources. > Actually, I am not going for GSoC this year. I won't be a student anymore when applications open. I could enroll to something to be able to participate (like I did in 2010) but that is not my focus this year. > > Also other idea: if the community feels like this is important feature (I > remember David Revoy said he used it) to be done, I have probably the > biggest > experiences with the brush engines, so I suppose _maybe_ I would be more > effective regarding the time needed to implement it (port it from GIMP > Painter) > and we can put it in my Action Plan. > I think you would be more efficient at porting it from GIMP Painter. But David, and myself, and probably Silvio too, want a more powerful brush. In the case of David Revoy, he wanted something that felt like mixing watercolors; in my case, I want something capable of giving a wide range of textures to the brush on top of mixing colors, a texture regulated by parameters (or in nerdspeak: parametrical texturing), in addition to the ability to mix colors not simply by painting dabs on the canvas like the smudge brush (or GIMP Painter's) but by other methods too. > I don't want to take him this sponsored project, I just wanted to share I > might be more effective, like Dmitry is more effective in the backend stuff > like > canvas, tiles, threading, I'm more confident regarding the brush engines. > > Pentalis can be sponsored to work on some different project -- to finish > Impasto? or to restore Emanuelle's work? Or still can be discussed..But if > he > really wants to work on this, then no...I will be here for you to help you > ;) > My point was only about being efficient with the Silvio's money in this > case. > Actually I really want to work on this (I was already working on it before I was offered sponsorship). Polishing Phong Bumpmap, Impasto, and the Halftone brush are things that will come with time (Halftone and Bumpmap first, Impasto much later), since I feel I owe those, however I don't think any of those is too important for Krita at the moment compared to a good mixing brush. As for restoring Emanuelle's work... now that's something really really interesting but I'm not sure if it's worth the effort. The Painterly Framework was made with the very idea of simulating subtractive color mixing in mind (if it was made thinking of something different then please correct me). If a brush engine is made that can mix colors like that (or in other fashions), then there is no need for the painterly framework anymore, at least not with any urgency. Knowing myself, It's only a matter of time before I try to solve the problem theoretically, but whether it turns into code that goes into Krita is something I don't know; it /will/ turn into a lengthy wall of text describing my findings somewhere though. Now comes what would've gone into a separate e-mail after Silvio's answer ---> Silvio wrote: > > > Pentalis can be sponsored to work on some different projects -- to finish Impasto? or to restore Emanuelle's work? > Wow. What a conundrum! :-) > It goes without saying that having a mixing brush engine is vital for Krita :-) Whether it is developed by Pentalis, thanks to a Google summer of code, or through my sponsorship it is up to you to decide (Pentalis plus Krita's developers: Boud, Cyrille, Sven and so on). > Let's think about the better options to try to satisfy everyone ;-) > Naturally, now it is also up to Pentalis to speak his mind ;-) Well, my mind says that porting GIMP Painter's mixing brush is not a long term solution. I envision the mixing brush engine as something made with object oriented programming in mind, modular, thorough and clear; Something that covers many needs and where the theory behind the code (the "model") is evident (such that others can extend it without trying to figure out what was I thinking). Of course I also want it to do more than just mixing. I don't think GIMP Painter's mixbrush was made with that vision in mind since the author himself described it as a learning exercise. I had a large e-mail describing all the thought I've given to the topic and it seems this is a good moment to send it to the mailing list. This should open up the discussion for others to give feedback on what they expect from the brush. I've been painting lately, so at the very least I know what I need myself (part of that already described earlier in the e-mail). I think I am the right person to do this because I'm passionate about it, and have given MUCH thought to possible solutions (aka theoretical work). Code is where things get less fun but writing it myself saves all the time needed to explain the idea point by point to someone else (I think I've heard people say that they took longer explaining something than coding it themselves). After the code is done, programmers, who are more experienced at reading someone else's code, will probably understand what I was thinking after reading the code I write. Then it all will make more sense than with just a theoretical overview. Anyway, I've spoken my mind. Wall of text incoming in another thread!. --20cf3054a529196efd049921b1af Content-Type: text/html; charset=ISO-8859-1 Content-Transfer-Encoding: quoted-printable
(Most of this message was written right before Silvio added his answer= =A0=3DP =A0 I will complete it in the same spirit as if I hadn't read = Silvio's answer)

On Wed, Jan 5, 2011 at 5:49 PM, Luk= asT.dev@gmail.com <lukast.dev@gmail.com> wrote:

> 3.
> As regards Pentalis, I would like to having him working on a new mixin= g
> brush engine.
> In the past, I have read he wished to work on it.
> As for me, I am confident he is the right guy to finish this task ;-)<= br> > Needless to say, also for Pentalis there is really NO rush!

Regarding the fact, that you resources are limited, I think maybe
somehow more wiser solution would be to support Pentalis in the Google Summ= er
Of Code program (if there will be one) with this idea and if that fails, le= t's
use your resources.

Actually, I am not = going for GSoC this year. I won't be a student anymore when application= s open. I could enroll to something to be able to participate (like I did i= n 2010) but that is not my focus this year.
=A0

Also other idea: if the community feels like this is important feature (I remember David Revoy said he used it) to be done, I have probably the bigge= st
experiences with the brush engines, so I suppose _maybe_ I would be more effective regarding the time needed to implement it (port it from GIMP Pain= ter)
and we can put it in my Action Plan.

I = think you would be more efficient at porting it from GIMP Painter. But Davi= d, and myself, and probably Silvio too, want a more powerful brush. In the = case of David Revoy, he wanted something that felt like mixing watercolors;= in my case, I want something capable of giving a wide range of textures to= the brush on top of mixing colors, a texture regulated by parameters (or i= n nerdspeak: parametrical texturing), in addition to the ability to mix col= ors not simply by painting dabs on the canvas like the smudge brush (or GIM= P Painter's) but by other methods too.


I don't want to take him this sponsored project, I just wanted to share= I
might be more effective, like Dmitry is more effective in the backend stuff= like
canvas, tiles, threading, I'm more confident regarding the brush engine= s.

Pentalis can be sponsored to work on some different project -- to finish Impasto? or to restore Emanuelle's work? Or still can be discussed..But= if he
really wants to work on this, then no...I will be here for you to help you = ;)
My point was only about being efficient with the Silvio's money in this= case.

Actually I really want to work o= n this (I was already working on it before I was offered sponsorship).

Polishing Phong Bumpmap, Impasto, and the Halftone brus= h are things that will come with time (Halftone and Bumpmap first, Impasto = much later), since I feel I owe those, however I don't think any of tho= se is too important for Krita at the moment compared to a good mixing brush= .

As for restoring Emanuelle's work... now that's= something really really interesting but I'm not sure if it's worth= the effort. The Painterly Framework was made with the very idea of simulat= ing=A0subtractive=A0color mixing in mind (if it was made thinking of someth= ing different then please correct me). If a brush engine is made that can m= ix colors like that (or in other fashions), then there is no need for the p= ainterly framework anymore, at least not with any urgency.
Knowing myself, It's only a matter of time before I try to solve t= he problem theoretically, but whether it turns into code that goes into Kri= ta is something I don't know; it /will/ turn into a=A0lengthy=A0wall of= text describing my findings somewhere though.


Now comes what would've gone into a = separate e-mail after Silvio's answer --->

= Silvio wrote:

> Pentalis can be sponsored to work on some different projects -- to fin= ish
Impasto? or to restore Emanuelle's work?

Wow. What a conundrum! :-)

It goes without saying that having a mixing brush engine is vital for Krita=
:-)
Whether it is developed by Pentalis, thanks to a Google summer of code, or<= /blockquote>
through my sponsorship it is up to you to decide (Pentalis plus Krita's=
developers: Boud, Cyrille, Sven and so on).

Let's think about the better options to try to satisfy everyone ;-)

Naturally, now it is also up to Pentalis to speak his mind ;-)
=

Well, my mind says that porting GIMP Painter'= s mixing brush is not a long term solution.

I envi= sion the mixing brush engine as something made with object oriented program= ming in mind, modular, thorough and clear; Something that covers many needs= and where the theory behind the code (the "model") is evident (s= uch that others can extend it without trying to figure out what was I think= ing). Of course I also want it to do more than just mixing. I don't thi= nk GIMP Painter's mixbrush was made with that vision in mind since the = author himself described it as a learning exercise.

I had a large e-mail describing all the thought I'v= e given to the topic and it seems this is a good moment to send it to the m= ailing list. This should open up the discussion for others to give feedback= on what they expect from the brush. I've been painting lately, so at t= he very least I know what I need myself (part of that already described ear= lier in the e-mail).

I think I am the right person to do this because I'= m passionate about it, and have given MUCH thought to possible solutions (a= ka theoretical work). Code is where things get less fun but writing it myse= lf saves all the time needed to explain the idea point by point to someone = else (I think I've heard people say that they took longer explaining so= mething than coding it themselves).
After the code is done, programmers, who are more experienced at readi= ng someone else's code, will probably understand what I was thinking af= ter reading the code I write. Then it all will make more sense than with ju= st a theoretical overview.

Anyway, I've spoken my mind. Wall of text incoming = in another thread!.

--20cf3054a529196efd049921b1af-- --===============1731561621== Content-Type: text/plain; charset="us-ascii" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit Content-Disposition: inline _______________________________________________ kimageshop mailing list kimageshop@kde.org https://mail.kde.org/mailman/listinfo/kimageshop --===============1731561621==--