IDEA! Forks and Sporks:Taming the specter of busted blenders

General discussion about the development of the open source Blender

Moderators: jesterKing, stiv

MrMunkily
Posts: 79
Joined: Mon Oct 14, 2002 5:24 am

IDEA! Forks and Sporks:Taming the specter of busted blenders

Postby MrMunkily » Wed Oct 16, 2002 3:08 am

WARNING!!!! MAMMOTH POST AHEAD!!!!!

I've been reading up on elysiun a particular thread regarding the next generation of "material" interface. Many have objected to this because it seems to be even more complicated than before. I think we need to address this sort of argumentation before it splits blender into incompatible forks.

Main Points:

1. Many People want different things from blender. I'm not just talking about the render/game difference, but more importantly the interface arguments.
People seem to either think that:

A) The interface is too hard/too nonstandard and needs to be rearranged (like more menus and such)

B) The interface is great and should be cleaned up and extended
(fix inconsisencies and make better hotkey references and add features in paralell to old ones)

C) The interface is good for now, but won't allow for the kind of power needed later. It has to be rearranged to be perhaps somewhat more complicated (the disputed material 'tree' is a good example of what will probaly be proposed in this area)

2. People may become discouraged or find that the new blender's not going to cut it for their needs or preferences. This could cause (GASP!) forks. This is not a bad thing at all, but _will_ cause unpleasantness and will hurt coherent porting, bugfixing, and feature improvement efforts across the board. If we have 3 different blenders, (for sake of example call em EasyBlender, ClassicBlender, and NextGenBlender for the 3 main groups) then we will have file format difficulties, feature disparities, and above all, incompatability!

3. I think we can agree that incompatability would be an über-crappy way to develop and use blender, so I propose an über-tenative plan O' action: How about we put the happies on hold (sorry, I seriously know many of you REALLY want new features NOW, I do too) and focus on:

A:\ Compiling Blender

B:\ Understanding Blender's Guts and Cleaning up Blender

C:\ Making sure that Blender's file format can be understood in the future, or failing that (though I don't want to), ditching it entirely and making the change to something better

D:\ Making Blender's core seperate from the myriad things people want to change. (Interface, Renderer, Monkey-Mesh...)

E:\ Extending, plugging in, mutilating intefaces, forking intefaces, eatng cheese, throwing pies, chucking CVS serves from cliffs and having fun hacking blender.

of course, that's jut my opinion, I could be wrong.

-Jon

engalicorn
Posts: 27
Joined: Sun Oct 13, 2002 7:37 pm
Location: Edmonton, Alberta, Canada

Postby engalicorn » Wed Oct 16, 2002 3:43 am

:!: :!: :!: (ANOTHER) MONSTER POST :!: :!: :!:


These all all good points that must be adressed and i would like to give my side of it all.

People may become discouraged or find that the new blender's not going to cut it for their needs or preferences. This could cause (GASP!) forks. This is not a bad thing at all, but _will_ cause unpleasantness and will hurt coherent porting, bugfixing, and feature improvement efforts across the board. If we have 3 different blenders, (for sake of example call em EasyBlender, ClassicBlender, and NextGenBlender for the 3 main groups) then we will have file format difficulties, feature disparities, and above all, incompatability!


I completly agree 3 different blenders would mean that some would be worked on more then others causing the most popular one to leap ahead. But some would say that this would be a test to see what people want in the real world, however if we were to drop the other 2 blenders once one had proven itself then the persons working on the dropped blenders would lose a substainal amount of time causing resentment.

The interface is great and should be cleaned up and extended
(fix inconsisencies and make better hotkey references and add features in paralell to old ones)


Totally agree!! The interface is what makes blender uniqe. Sure i think that there needs some tweaks (such as a save on quit option so you dont blow your work when you accidently hit the X :oops: done that!! :oops: ) and blender has always been known as a program that is difficult however extreamly rewarding.



I think we can agree that incompatability would be an über-crappy way to develop and use blender, so I propose an über-tenative plan O' action: How about we put the happies on hold (sorry, I seriously know many of you REALLY want new features NOW, I do too)


This is a good point from what ive heard on the chatrooms the code is a mess, now i have little expreience with C++ but i know what they are talking about. Putting features on hold now will allow us to be more productive in the future.


Well seems my side of the story is almost exactly like MrMunkily's guess i didnt have to put all that down after all.

Well Ciao for now :twisted:

MrMunkily
Posts: 79
Joined: Mon Oct 14, 2002 5:24 am

Postby MrMunkily » Wed Oct 16, 2002 3:56 am

Good to know - I'm just thinking that blender could be the next mozilla literally - an expansible 3D core to be built on. Perhaps there could be a core BlenderRender scanline renderer as well.

senorTUDBZD
Posts: 3
Joined: Wed Oct 16, 2002 1:59 am

Postby senorTUDBZD » Wed Oct 16, 2002 4:47 am

Tottaly agree! Having blender fork into several variations would be devastating...im not a programmer or anything, but i think all of you programmers out there should focus on just organizing, and compiling blender for now...you should as quickly as possible make the new blender usuable for everyone, that includes non-programmers...we dont need to focus on new features yet, they will come, we just need to get it working...and ASAP! We dont wanna loose some of blenders usurs, especially the newbies saying, "what the hell?!" Compile blender...then make new features for it, like a more advanced game engine for example...I also think ton, and the other webmasters should put up a section explaining what blender is, and offering downloads of blender 2.23, or links to downloads of it, because, just imagine when us experts, and advanced usurs started out...well...atleast for me, it was kinda confusing, and adding this compiling stuff to it may confuse them so much, they just say forget it.....but, the interface? i think its great! its weird at first, but i LOVE blenders interface! we shouldnt really change it a whole lot, just put the new features in with the old, hand and hand...

jeotero
Posts: 107
Joined: Wed Oct 16, 2002 5:31 am

Postby jeotero » Wed Oct 16, 2002 5:59 am

yeap the core should be compatible between all blender versions, on top of that, create your own interface, renderer, etc.

Goofster
Posts: 137
Joined: Mon Oct 14, 2002 12:26 pm

Postby Goofster » Wed Oct 16, 2002 12:26 pm

seperated core, customizable interface...addon.... have you read Eskil's post on the foundation forum on elysiun?

NExtGen blender

Roel

gorgan_almighty
Posts: 51
Joined: Wed Oct 16, 2002 12:26 pm
Location: Kent, UK

Postby gorgan_almighty » Wed Oct 16, 2002 12:38 pm

MrMunkily wrote:Good to know - I'm just thinking that blender could be the next mozilla literally - an expansible 3D core to be built on. Perhaps there could be a core BlenderRender scanline renderer as well.


Yeah just wait till Microsoft get their hands on it!!!!! :!: :!: :!: :!: :!: :shock:

Charlls
Posts: 10
Joined: Wed Oct 16, 2002 10:33 am

Postby Charlls » Wed Oct 16, 2002 3:48 pm

this all point to modularization of Blender


ie: splitting functionality in dynamically loadable libraries (plugins)

Im not aware of the current plugin infrastructure in Blender (maybe Ton could extend on this)

anyway, for this we must arrange general interfaces which each plugin implements

but for this we need to create C++ abstract classes headers (the interfaces) and plugins objects which inherit from those abstract classes, seems that Blender is extensively wrote in C, so im not sure what are the implications for a C++ subsystem


Return to “News & Chat”

Who is online

Users browsing this forum: No registered users and 1 guest