Darkradiant 1.0 Roadmap

[OrbWeaver]

These are currently what I believe to be critical features for the 1.0 release. The 1.0 release is considered to be the point at which, if the mod were completed, the editor would be ready to advertise to users as the official Dark Mod editor.

 

Proper texture alignment on patches - ability to copy and paste textures between brushes and patches maintaining the correct alignment.

Full light support - all missing functionality regarding lights needs to be identified, and will include support for projected lights.

Robust support for all TDM maps - no features must cause the editor to choke or otherwise misbehave

More complete skin selection - selection for arbitrary skins on models (e.g. "gen" skins) rather than just those associated with the model

Objectives editor - all mappers will need to set objectives in their maps, so a GUI is needed for this

Stim/response editor - many users will want to set up custom stim/response on entities.

Prefab insertion - the currently disabled "Insert prefab here" option needs to be enabled, so that prefabs can be inserted at a given location; also processing of "bind", "target" and other entity connection keys will need to be verified.

Sound editor - at the very least a simple GUI for browsing and selecting sounds is necessary, since at the moment the mapper has to manually inspect the files and type in the exact name of the sound shader.

Support for shader expressions - addnormals etc.

 

After 1.0, there are other features which will be important in the long term.

 

Particle editor - since the aim is for TDM to be editable without using DoomEdit (esp. on Linux), some sort of particle editor is required. Since particles rarely use bump maps or other shading, it is acceptable for the particle editor to use OpenGL lighting rather than per-pixel lighting at this stage.

Improved renderer - including support for dynamic shadows and older graphics cards, and hopefully a performance boost

Other features - as may be requested from time to time

[New Horizon]

Looks good Orb. The only part I'm hesitant about is leaving the Stim/ Response editor as 'post' 1.0. Having a stim/response GUI is a pretty big part of the Thief mission building experience as mappers have used them in dromed and T3Ed to string together some previously un-thought of tricks. Even basic stim/response editing would be welcome, but it's definitely something I would almost be inclined to list this ahead of a particle editor.

[greebo]

I tend to agree with New Horizon, that the Stim/Response editor is somewhat more important than the particle editor.

 

And I would add one more thing to the list: better handling of copy & paste of Textures from Brushes onto Patches. This is one thing DoomEdit does better than DarkRadiant. Currently, if you copy the texture from a brush onto a adjacent patch (an inverted bevel, for example), only the texture name (and perhaps the scale) is copied, but the thing is misaligned. In DoomEdit, the textures get perfectly aligned. I needed this a lot in my bonehoard section with all those bevels and I always switched to DoomEdit, because it is so much more convenient than aligning it by hand with the patch surface inspector.

 

Apart from that, the roadmap seems good. Should we go into more detail which of those parts should be tackled first / priorities / what the versions till 1.0.0 should cover?

[OrbWeaver]

I have switched S/R Editor and particle editor, added the patch texture alignment thing and ordered the tasks approximately as I think they should be prioritised. Feel free to suggest alternative priorities if there is something I have missed.

[New Horizon]

I have switched S/R Editor and particle editor, added the patch texture alignment thing and ordered the tasks approximately as I think they should be prioritised. Feel free to suggest alternative priorities if there is something I have missed.

 

Looks great!

[OrbWeaver]

Skin selection is now done, but we will need to implement the "Insert prefab" support which is currently only a placeholder menu option, so I have added this to the list.

[OrbWeaver]

Inserting prefabs is now implemented I believe, but I have added the Sound Editor as a pre-1.0 priority because currently there is no adequate way to select sounds for speakers.

[greebo]

I have a general question: Are there any plans for intermediate releases between 0.9.0 and 1.0.0? For me, the time span between 0.8.1 and 0.9.0 appeared a bit long, maybe we should decrease the interval a bit?

 

(Not that I'm itching to release anything, just curious.)

[OrbWeaver]

There's no reason why not, that's what the minor version numbers are for.

[greebo]

Perfect, maybe in a few weeks there is enough in the portfolio to justify another release.

[OrbWeaver]

By the way, if you want to have a look at the Objectives Editor feel free -- I've done a certain amount of it, but I'm not sure when (or if) I will actually get round to deciphering the cryptic objective component specifier system, and implement the required custom panel for each component type. This may well be similar to the Stim/Response editor so you might be in a better position to figure out how to do this.

[greebo]

Didn't you intend to change the Objectives system in the SDK? It's not yet finished, and Ishtvan seems to want to work on the Melee system, so this task may be free.

 

I probably can't do much in the all-to-near future anyway - currently, I'm trying to figure out the last few unknown mechanisms of the scenegraph. I learned a lot the last few days about it, and there is a lot of weirdness going on, I say.

[OrbWeaver]

Didn't you intend to change the Objectives system in the SDK? It's not yet finished, and Ishtvan seems to want to work on the Melee system, so this task may be free.

 

I could have a look at that I suppose. Unfortunately at the moment, Eclipse has refused to work properly on my system and no amount of re-installing different versions seems to fix it, which means that I have only gvim to use for development. Much as I like it as an editor, it is not a particularly good IDE.

[greebo]

My Eclipse crashed regularly on both of my Linux installations until I deactivated the C++ indexer, but I don't know what type of problem you're experiencing though.

[OrbWeaver]

My Eclipse crashed regularly on both of my Linux installations until I deactivated the C++ indexer, but I don't know what type of problem you're experiencing though.

 

3.2.2 was working fine, then the "Create class" option magically stopped working, complaining that the classname was blank even when something was written there. 3.3 RC1 seemed promising, but the "Replace with local history" function didn't work, displaying a completely blank list every time. I think Eclipse has reached the stage of complexity where it is just impossible for everything to work properly.

 

I also tried KDevelop, which has some really nice features (indexing that actually works, unlike Eclipse's indexer which takes up all your CPU and still allows no search functionality), but kept going into a 100% CPU lockup so I gave up with it too.

[greebo]

mohij suggested Code::Blocks a while ago, maybe this proves to be useful?

[greebo]

Can somebody unstickify this thread, please? It has been functionally replaced by the bugtracker by now.

[greebo]

Thread resurrection:

 

@OrbWeaver: is it too pretentious if we just switch the DarkRadiant version to 1.0.0 along with the first TDM 1.00 release? Or should we rather wait until we have (almost) complete support for the D3 engine (like particles and animations) in-place? The latter might not happen until it goes open source sometime in 2010 (maybe).

[Komag]

I like the idea of 1.0

 

Next year you can make 2.0 or whatever

[OrbWeaver]

@OrbWeaver: is it too pretentious if we just switch the DarkRadiant version to 1.0.0 along with the first TDM 1.00 release?

 

No, that was always my intention. 1.0 of DarkRadiant corresponds with 1.0 of the main mod.

[greebo]

So be it then - I'll try to be prepared for a DarkRadiant 1.0.0 release around the time the main mod is going out for the testers.

[Fidcal]

Is V1 available internally before then? Maybe I missed a post. Or is this just a renumbering of what we have?

[greebo]

I can provide a test build over the next days, but it will be a close call.

[Fidcal]

I'm not trying to rush it or anything but just wondered. If it is to be released at the same time as Release 1 then presumably you would be doing it a day or two early anyway so that would be fine. And if it is not released until a few days after that is OK too by me anyway.

[greebo]

Ok, here are the two release candidates for Windows:

 

32 Bit: http://www.bloodgate...diant-1.0.0.exe

64 Bit: http://www.bloodgate...t-1.0.0.x64.exe