The Clash of Doom

If you have been reading the news pages or following the Acorn Games Newsgroup recently, you will be aware of the controversial 'free' Doom port for RISC OS: Doom-It-Yourself by Andreas Dehmel. Unlike previous ports, which R-Comp (the publishers of the commercial release) have been able to quickly silence, DIY is apparently immune to this, as it is distributed in source form, and needs to be compiled. This is in fact easier than it may sound, and we've managed it to bring you the first comparison of it with the commercial release.
A key argument for those who are suspicious of DIY is that it is inferior to Acorn Doom, so we've run it through a few arbitrary speed tests (ie. how fast it appears to run) and we'll have actual FPS figures soon to corroborate them. Also, we look at general support, bugs and other related issues.
The article is now updated to take account of the many major improvements recently to DIY.

-------------------

Speed
Both are, in general, quite playable on a RiscPC 700 in standard (320x256) resolution, and jerky to a generally unplayable extent at 480x352. All tests were carried out with a Risc PC 700, 8+1MB RAM, and *RmFaster SharedCLibrary.
In Ultimate Doom (Doom I) there is a difference in small levels; Acorn Doom shows a definite speed advantage. However, on very large and complex levels, such as in the Thy Flesh Consumed episode, this speed difference is less obvious, and DIY does occasionally show a small advantage.
In Doom II, on the early levels, Acorn Doom has a noticeable edge in all resolutions. There is little to tell between the two on larger levels; DIY often has a slight edge there but this is rarely noticable - in general Acorn Doom is the faster.
There seems to be some crucial difference between the plotters; Acorn Doom offers a generally more stable refresh, but this means that a very large level produces a more pronounced slowdown than DIY which seems to disregard the complexity of the level. The base plotting speed of Acorn Doom appears to be faster despite this, so a simple level is markedly more responsive.
The speed differences are unlikely to be noticable on a StrongARM where the frame rates are very smooth on Acorn Doom already; one would expect a similar increase with DIY, but the current release is not optimised for it, so Acorn Doom is likely to have whatever edge there is. DIY is under constant development however, and we have heard of some updates that markedly increase SA speed and offer a few extra FPS on slower machines.
On DIY, the 16-bit mode offers a faster display for simpler scenes than the 8-bit, but the big slowdown (on slower RiscPC's) kicks in for more complex ones.
The latest version of DIY, due for release in the week of the 25th May, features a 24-bit mode now very similar in speed to Acorn Doom, even on slow Risc PC's, and with very little slowdown compared to its own 256 colour mode.
Additionally, while taking the screenshots below I noticed something very peculiar - Acorn Doom only uses 320x240 in a 320x256 mode, so this could well be the origin of much of its small speed advantage. DIY takes up all the available screen area, so has to plot more of the scene anyway.

-------------------

Some screenshots
These demonstrate some of the visible differences. The Acorn DOOM screenshots are taken in standard (320x240) resolution and the DIY screenshots in 320x256. Both are 24-bit and at gamma correction level 1, with DIY's hardware gamma correction enabled. Click on them to view them in more detail:

[R-Comp DOOM]
Acorn DOOM's 'cut down' view of the weapon
[DIY DOOM]
DIY Doom showing the whole weapon
[R-Comp DOOM]
The 'cut down' plasma gun of Acorn Doom
[DIY DOOM]
DIY DOOM showing the plasma gun in its full glory
Recommend that you look at these two shots at full size to appreciate the difference:
[R-Comp DOOM]
Acorn DOOM's reduced 320x256 mode and distorted invisibility effect
[DIY DOOM]
DIY Doom's full transparency in action

-------------------

Other issues

DIY Cons

Acorn Doom Pros
DIY Pros
Acorn Doom Cons

------------------- Back to miscellaneous index