OrbWeaver

I found this gem in DarkRadiant - I think its purpose is to remove null entries from argv[], but why they bothered shifting stuff around when they could just copy the valid items into a new array I have no idea. void args_init(int argc, char* argv[]) { int i, j, k; for (i = 1; i < argc; i++) { for (k = i; k < argc; k++) if (argv[k] != 0) break; if (k > i) { k -= i; for (j = i + k; j < argc; j++) argv[j-k] = argv[j]; argc -= k; } } g_argc = argc; g_argv = argv; } I would have done it like this: void args_init(int argc, char* argv[]) { int inPtr, outPtr; char* outArgv[argc]; for (inPtr = outPtr = 0; inPtr < argc; inPtr++) { if (argv[inPtr] == 0) // null pointer break; else outArgv[outPtr++] = argv[inPtr]; } g_argc = outPtr; g_argv = outArgv; } Maybe some experienced programmer can point out why my version is crap, but it seems to me to do the same thing with more readability and fewer lines of code.

I would mention that it's mine on Monday, but I think by then we would need a separate forum just for birthdays.

I think your expectations are too high. HTML was never intended for exact, per-pixel alignment, and attempting to get this right can be a real struggle. It is also arguably not a very good way to design a web site which might be viewed on different platforms with different font sizes and resolutions.

Assuming it's a pin tumbler lock, yes.

Subversion is a source-control system, basically an improved version of CVS. GtkRadiant has only recently been updated to work with Doom 3, so there may well be missing features. Our hope is that with DarkRadiant we will be able to add customised features that are important for Dark Mod missions, such as an improved hierarchical entity browser, stim/response editor etc.

I should have mentioned I want an open source product, not a commercial one. In fact if I had the time I would be interested in developing such a tool myself, but unfortunately I don't so it will have to remain on the wish list.

You could check out the SVN and build it yourself, but there is no point as yet because no changes or enhancements have been committed. If you want to familiarise yourself with the application, you would be better off installing a recent build of GtkRadiant.

It's the polys in the scene that are important. A single object could have 50000 polys as long as it was in a room by itself without too much else going on. On the other hand, a pillar that was intended to be used in sets of 12 would not want that many polys.

If tracking is the assembler of music, then CSound is the machine code.

It's not the interface from tracking I want though - it's the interface from a MIDI sequencer (multiple channels, variable-size blocks that can be named, recorded, edited (in piano roll, notation or event view), copied and dragged around) but the internal sample handling of a tracker (load samples, perhaps edit them, have as many as you want, appy effects and volume envelope, without having to use ANY external MIDI programs). A completely self-contained music production system, in other words, that lets you do anything you want with audio without having to write lines like "F#4 00 01 32 E00" all over the place.

Ah yes, I remember you mentioning you were looking for mappers for testing. I wonder if one or two testers to would be able to get a build environment set up on Windows so that they could build and beta test from the latest SVN. Since I develop in Linux (the toolchain is just so much smoother, plus I'm in Linux for most other stuff anyway), it will be useful to have some permanent Windows testers to make sure that both platforms are functioning correctly. I'll have a look into what is required for building on Win32, I suspect that just Scons, Subversion and the relevant devel libaries should do it.

You are of course talking about Doom3Radiant. DarkRadiant is only two days old.

http://www.bbc.co.uk/youngmusician/competi...s_strings.shtml Looks like May for the final (http://www.bbc.co.uk/youngmusician/competition/final.shtml).

Like the fact that silhouetting would require a render from the point of view of every AI in the scene, rather than from a couple of fixed directions like the lightgem?

Besides, you could accomplish a hell of a lot more than a weapon with 100 hours of competent developer time.

The latter.

That sounds like it's linked to the lightgem. It wouldn't be a lot of effort to link the code that brightens or darkens the lightgem to the modify the entire hud as well.

What resolution are you using for the captured image? Presumably it can be quite small if all it is doing is calculating a light level. Perhaps a tweakable option could be provided to adjust the resolution for lower-performance users?

I remember trackers, I used to produce music with Impulse tracker on DOS. They are very much a "techi" solution, the Assembler of music production. The main advantage of them is the internal sample processing which means you don't have to use external samplers or anything. What I would like is a tracker with a sequencer-style interface, to combine the best of both worlds.

It is actually a very good design - using the game's own render code to process animated textures rather than having to re-implement everything and attempt to match what the game is doing. The drawback is in the implementation, due to an underlying limitation of the game code which only allows the rendered image to be passed back over a local socket (which is not a high-performance way of doing things). It doesn't seem to present any problems so far though, in fact for me the lightgem feels a lot more responsive in TDM than in any of the Thief games.

That's what I thought. Thief seems way too primitive to have the sort of semi-advanced image integration code necessary to calculate an overall brightness and adjust for it. Besides, if you're using hardware rendering the game code doesn't even have access to the rendered image without special tricks (*cough* renderpipe *cough*) I reckon they just have monitors with poor voltage regulation.

Generating sounds is analogous to generating image textures in many ways: 1) There is a large variety of tools available, from professional studio apps to bits of freeware that are good for one or two specific jobs. Typically a selection of tools will be employed, starting from "generation" tools and ending with a final "composition/mastering" utility (PhotoShop, Sound Forge, Cool Edit, etc). 2) A combination of "painting from scratch" or layering existing samples can be used. 3) By using high-quality original samples you can generate a very good result with relatively little work. 4) You don't need any special hardware (MIDI keyboards/graphics tablets), but more experienced artists/composers may find that their workflow is quicker with one.

Aren't they? I can't remember the last time I snogged another man or shared a bath with one.

Presumably in order to work properly it has to project an image from every side onto its opposite side. The whole surface would need to be made up of cameras and projection lenses interleaved with one another.

Many people don't understand that. I was once discussing what a load of crap Freud's pseudoreligious psychology "theories" are, and my colleague responded "but Freud invented the field of psychology". True, but it would still have been invented by somebody else if he hadn't got there first.