Old News <Jan 2004 to Dec 2004>


   12/28/2004

I am now working on a low level game. All of the blitters 'n sh**. It should be able to work on any platform with 68020 compatible software (Macintosh, Amiga... er... emulated). The most difficult part of it is bit depths lower than eight bits (I have to use bitfields to carry around non-byte aligned pixel data). At least this is more reliable (and possibly much faster) than some of the QuickDraw software on my Mac right now!

As far as Old Durandal... I am dealing with 64 bit math -- all programs should use "Microseconds" instead of "Delay" for timing. I will try to implement a feature to convert a film into a compressed movie file (30fps!).
I will be putting up a 68K build that is months old. It will be slightly modified to try and fix a small problem with the black color when using shapes with crappy color tables. You will have to E-mail me to get the encryption password for the StuffIt archive to be useful.

   12/20/2004

Sunday shtuff. Not as great as I'd hope I would have written it.

   11/29/2004

Days old... (busy)
Running playing M1 maps...

Tame thy fists
"Suicide II"

Test map
Night vision (the first image in invincibility and invisibility)... it got boring. So I added some friends (Assimilated friends) with a new cheatcode.


Can't... move...
Before alien textures were detoured...

Much better
And after.

   11/29/2004

Did you know they have Grand Theft Auto for gameboy? Pimpin' and killin' in pocket sized package -- pretty risky and stupid for Nintendo.

New write-amonn-a-junk.

   11/10/2004

I take back what I said about M1A1. They have probably put more work into their project than I have in my projects. I still can't play M1A1, since I only have a 68LC040 Mac.
source.bungie.org still seems to be down.

Here is something I wrote on the weekend regarding the marathon engine, gaming, and politics. Be advised that it is not censored. I don't time myself, so I don't know how long it took for me to write. I know for sure, it took less than three hours. If you don't believe that I wrote it, then test me.

Okay...
Halo 2 is the one to look for. It may not have the best graphics, but... it's from Bungie! Or Microsoft- er, something. I don't know. Halo 2 is supposed to have more of the real sludge we deal with today. Cause and effect also includes political bullshit.

Grand Theft Auto: San Andreas lacks no violence, and paves the way to low expectations in games. Watch the black guy run into walls and shoot people - stereotypes are fun!

   11/10/2004

After accidentally taking a risk... the full MaraMap Ed source code is out. I'm done with that sh**.

Loooookkkk......

Night vision with Marathon One textures
(lava set - lava looks like metal grating)

This was taken just a few hours ago (about six-something P.M. EST). Not only is the image in Night Vision mode, Old Durandal (the program name) is using an M1 map. I did this in effort to not have to redo the level, or anything. As far as the shapes: just put the M1 panel textures in M2 order (switches, rechargers, comm.), then general textures, then the last of the M1 panels (the last at index 37). The program does the rest. It reorders everything, including the monsters, items, and scenery. Nothing was done to the map.
M1A1 now looks like crap.


source.bungie.org appears to be still down. I wonder what is going on over there (or if anything is going on at all).

   11/7/2004

The previous post was actually from the 26th of October, although the information completely reflects the 25th. The number of node aliases was actually at 256 (and is still at that number).
I've got 8 bit color to work completely, although there are problems with the handling of a few colors - especially with find_or_add_color, after implementing...

NIGHTVISION!

You'd think that the majority of Americans would die for their freedom. Heh, I guess we're too fat, now. Bush got reelected!! Moral values my ass.

   10/25/2004

I finally solved that major bug that corrupted memory. It involved the paths trying to allocate a pointer before the size was set in the dynamic limits (during startup).

After setting the maximum number of node aliases to 128, I finally got zero render problems (although some of it doesn't show up because of bugs in the map). Look:
Desla
You can see the frame rate, number of objects, and number of nodes (those three now come standard in the debug version). Over 2,600 nodes.
And even at the lowest resolution, there is still a low frame rate (more nodes being sorted). The second is the limit to the frame rate.

I played Missed Island, and it was pretty short. Look at Waterloo Waterpark with it's texture set!

You can see the transparent texture for the window has the "IGNITER" thing on it.

And...
Here's something you couldn't do in Marathon Infinity (I got there by using a cheatcode):

And here's the overhead map:

   9/25/2004

Long awaited (yeah right) images are up! Check out the gamesbeat page. DOOM II and Oh no! More Lemmings.

   9/9/2004

I am completely awake now. Ask me anything in the universe. I can give an answer that may suprise you.

   8/26/2004

I am back after nearly a month... just to show you I'm not dead. It may be wiser to start with a board game - er, whatever. I came up with a new floating point format that's optimized for the m020 - Motorola 68020 and later (like i386 represents the Intel 80386 and later op-codes). 32 bit significant digits plus 31 bits of exponent (and 1 bit for the sign). It's way faster than the S.A.N.E. FPU substitution code. The square root function I wrote for it performs usually at most 68 instructions! And the funny thing is, it takes less time for the code to compute the square root of two than it is the square root of one. The 'multiply' operation is usually the fastest of them all. After all of this, I wrote a high level sine function using the arithmetic function code for all of that. The accuracy of the sine function goes to less than a fraction or error! All simple tests have come to there most closest equivalent values. pi->-0, pi/2->1.0, pi/4->0.B504F334, pi/6->0.5... even on my old 68k Mac, a number of these operations can be done in less than 2 microseconds.

The term "radiosity" came from the study in dynamics of heat transfurence. When you put it into light wavelengths, you get photographic images in CG rendering. Here is an explanation.

Okay, gaming specific news. John Carmack has pushed modern video cards into a ray-tracing-like graphics engine. "photons"! Oh, my. His idea was... not an original idea. But... you do get animated light sourcing, reflection and refraction out of it - not available in most 3D games. Low polygon rates in the physical objects in the game with high detailed "normals" mapped onto their offscreen space. You also get sharp shadows. So, as you can see, the awaited graphics of DOOM III are not the most realistic, however the gameplay is better when you put the f**ker in animation. Something you didn't get with the older DOOM engine and some modern 3D modeled engines - smooth animation. The monsters seem more real in how they move. Well enough blabbing off today. I gotta go trip an' crack my head open.

   7/31/2004

No politics here. I hate politics. So.. I won't talk about Dubyah screwin' up the environment with the V.P. in *"secret meetings with polluters", or growing the deficit to immense proportion (recently came to another record!).  Hollywood does not represent any party! How could they let that happen?! And this is your junk news briefing (putting on underpants).

I might start a game called Turdis - it's like Tennis, but it deals with batting a flaming piece of sh** on a court. And around that court is a large moat of boiling oil called the "fryer ring". The Turdis ball would be ejected by a Turdis ball launcher that looks like an ass and is constructed so that the hole vibrates in such a fashion that it literally makes a fart sound when launched. The ball is on fire and the players are wearing flameable clothing... so they have a chance to catch on fire (or fall into the "fryer ring"). If the balls hit the rackets hard enough, the Turdis balls get divided up into rectangular pieces (essentially sliced) and fried into brown French fries once they land in the large "ring of fryer" (mmmmm... tasty - little burned hairs and corn and all!). All of this - of course - would be in a computer game (and not in real life). Different arenas, including a bridge like one where the movement of the players cause the bridge to rock, and you must not let the Turdis balls hit the bridge ('cause that would cause it to set fire leaving the players plunge into the large fryer below). Particle systems for the boiling water and fire; just enough for realism, but not too much, because it's got to be in real-time on PPC Macs. The only difference between a turd and a Turdis ball is a Turdis ball has an imprinted texture of something like a basketball or tennis ball, and... of course, it has a coating of lighter fluid and is on fire. You could even put in an a completely irritating audience full of characters from shows you hate so you can delibrately try to get the Turdis ball to hit them (which would set them on fire). Then at the end, if you lose all the Turdis balls you lost would be relaunched at you from the audience -- and it's a matter of dodging them that you don't get set on fire.
DO NOT TRY THIS AT HOME!! Although it would be funny seeing the Jackass guys try it (it would be expensive though).
Am I a real nutbag - or what?

* in at least one of the DNC speeches.

   7/28/2004

'Starting an expression parser -- I need to at least be able to make graphs.
I will be starting completely over on a new map editor, and it should come out a little like Vulcan/Forge. The new Marathon I've been working on is good up to the point of monster action - paths are not working right! Something I did days ago screwed up pathfinding. I keep getting that "invalid current step path", especially when a path has been taken out (the monster is using an invalid path). Monster targets may not be correct.
Due to a number of reasons, anything that involves decision making or artificial intelligence (regarding mostly monsters at this point) have been moved to a new segment called "A.I.", with "artificial_intelligence.c". Other than that, there have been major changes to monsters and projectiles (mostly definitions). Essentially, the structures are more elaborate; they contain more stuff, and more precision.
Once I get some of the problems in pathfinding fixed... I'll upload a new snapshot of the code (there is one currently up from late June - I won't give the address).

   7/8/2004

A lot of changes have been made to projectiles and monsters.

I'm trying to fix the problem with color tables under 8 bit. It should update the GWorld color table now (QuickDraw wouldn't copy directly).
If I were to make a PPC version, it would have higher per-pixel accuracy when it comes to rasterizing. Masked images, overflowing 10+ bpp channel capabilities (still fits into 32bpp) -- and possibly 2x2 dithering. I also may add light-mapping capabilities. These may not run at high performance under 68k... but they should still be optional by the preferences and/or the map and shapes files. Both of the color table and the 30bpp thing would make it wise to take QuickDraw out completely when it comes to transferring to screen.

I will leave GWorld accessors in the code for multiplatform reasons, since they should be available under QuickTime 3.0+ on non-Mac platforms.

   6/28/2004


Long distance issues? What long distance issues? A recent snapshot (less than 72 hours ago) of the more accurate rendering I've been talking about. The landscape may be a little stretched, but there is no way you can get this out of Infinity. I had to increase the number of node aliases to render at all view points in "Carnage Time Arena on Steroids". Ammunition is literally littered all over the place, but its all there in the image.

This was much easier than Desla. I can say that a million times. I'm still raising the limits, trying to get Desla to work, but the limit for maximum nodes never seems enough. The maximum number of nodes is now set to 2048 -- variable at runtime, yet there's nothing to set it. I thought about XML, or some kind of other text format, but that would just seem like ripping off A1. Out of all the 384 maximum objects (default value - for compatibility), half of them can be rendered at once. Memory problems are arising from what I'm doing. I can't even load a 200KB map file, without Marathon running out of RAM -- at 9MB application heap, and about 3MB system memory (I know, my computer doesn't have enough RAM). But hey, if I can get it to do amazing things, then what the f***. My Marathon has gotten very reliable, and despite all of the dprintf junk, warnings, and assertion failures - all put on screen, a lot of the time interrupting gameplay... it barely crashes at all, now.
You should see the translucent player weapon (when invisibility is on at the minimum), and the overhead map; variable scale (not the stupid four-position constants), anti-aliased, colorful, and every line and every object with their facing angles (when cheating, of course).

   6/24/2004

Now with a more precise render viewing angle, and a better point transforming function (it rounds and checks for overflow), the walls map on much better. And something I did, has given me a whole new horizontal polygon clipping bug (or set of bugs), that leaves most of the screen points completely incorrect. The polygon points just kind of, fly all over the place like a motherf***er, sometimes sticking to corners of the view, most of the time rendering over walls, and usually leaving whitespace. A snapshot right now wouldn't look pretty.

   6/21/2004


Check it out. Seein' the hummers before they surface. Anti-aliased!


If you look closely on the left, you can see the Water Yetti, runnin' around like an idiot.

I'm increasing precision with the 12-bit angle system I made last year (updated). Imagine being able to look at an object, almost exactly straight at it. Well, that will come. The new Marathon will look a little different than these recent images. With the use of long vectors (32+ bit accuracy), long distance will become a reality on the 68k Mac (those Aleph bastards threw away their chance of compatibility with 68040 Macs). Rather lengthy wall textures will map without any noticable error up close, when looking at it far to the right. Maybe I should've uploaded some images with some 'splosions.
If were to release any build, it would have to contain a PPC compilation as well, and some damn good improvements from the Marathon 2 source -- releasing the source code with it is required (and of course the gpl license, as well). A build... might come.

   6/18/2004

Semi-transparent texturing has been implemented. It uses a fast method (not supported under 8 bit depth). Simply take the destination pixel, clip off the lower bits in all three channels (bitand), then shift and add the texture pixel. It's not as accurate as the standard method, but having over 13fps when filling the view on my 68k Mac is worth it. I still have a problem getting the opposite media bound part of objects to render. It's good being able to see the floor underneath the water, but stupid not seeing complete objects or any objects that are truly there.
I also modified the screen mode stuff, so the graphics world won't update if it's not neccessary, also sounds won't be stopped in this case. The offscreen buffer is always at least at 320x200 to make it possible to switch the overhead map (low resolution is supported). As soon as I'm able to upload anything, I'll get my results up in images. The anti-aliasing option is for people with faster computers. Plus, you need to give the program over 9MB just to get fullscreen super-sampling under 16 bit -- the offscreen buffer takes most of the allocated memory in that case (duh). I used to able to get the highest resolution under 9MB, but now, due to larger default storage for redundant map data (platforms, polygons, endpoints... etc.), nine megabytes is not enough. The only problem with the screen mode modifications is that the terminal mode keeps screwing up the view dimensions. I had to put in code to catch these, so the renderer won't draw outside the buffer (and potentially create a system crash on "exit").
Since the version record in films weren't used, manual modifications are necessary in order to get- oh, I don't know how it will look, once I - or if I - get film playback fixed. Inconsistencies can arise from bug fixes alone! I wonder what film playback in AlephModular looks like... I need a new computer!!! Maybe I should shake AlephModular's hand, considering I downloaded the original Bungie release of the marathon2src source code off their CVS, I think -- I could've downloaded it from somewhere else.... I need sleep!! I need sleep and a new computer!!! AHHHHHHHHH!!!!!!!!!!!!!!!!
Fades have been updated (a long with damage types, of course - they have been maxed out), and now transparency can overflow in all fade functions. The update started from the explosion fade; to make the initial point of the fade a little white before it goes to yellow (more convincing). The last damage types will be used as monster flags (mortality, simulated fear, etc.), and projectiles can do that type of damage, as well -- make a monster commit suicide from a distance!

   6/12/2004

After reverting to my modified version of the source code archived in April 14 (I back this shtuff up for good reason), I was able to fix the rendering bugs I made -- mostly that my rewrite of "memmove" was bad, effecting the placement of clipping vectors that go up to "render floor or ceiling". Big chunks of code were reverted back to the public release Marathon 2 code. Except for films and saved games not working, and 8 bit depth not copying to screen correctly at "high resolution" (color table stuff that was handled in the non-public functions "myNewGWorld" and "myUpdateGWorld"), everything else is better than Marathon Infinity -- except for maybe the assimilated fusion civilian's bleeding effect; I still can't get it to set up a valid object effect, and I had to do idiot proofing down the line up to even "texture_rectangle", or else Marathon would crash.
There are still rendering problems (that Infinity had) with texture mapping that doesn't line up, leaving a handfull of edge pixels un-rendered (shows up white in the debug compilation), and the left and right rays that make up the clipping leave out a verticle line that's one-pixel wide on one or both sides (comes out white as well).

I've sort'a given up on larger screens for now, and decided to go for supersampling (antialiased view!). I wrote the scaling routine so it's faster than QuickDraw scaling (only 15 bit depth has been tested) -- and it's the C version! Either way, it's doing a maximum of 1280x960 rendering, yet the offscreen scaling by itself takes a pretty noticable percentage of the whole frame rate -- toooooo slow on my 68K Mac (about 0.38 fps). A PPC version... is still not present because I don't have a non-MPW compiler for PPC. So I would have to backdate the code style a little in order to do a PPC compilation. GetOSEvent was replaced early on (for future use under Carbon) with GetNextEvent, so as a side-effect, if I could get films working again, I could take snapshots of frames -- and even sequence them together to make movies! - Of course, I could just add an offscreen playback method of generating movies with motion blur with by just command-option clicking the playback button on the main screen, but someone would have to actually SPEAK TO ME!!!!! I will have my anti-aliased results up here...... SOON! That's if I can get the ability to upload anything again.

I'm also adding other features in a backward compatible manner; adding flags and tricks that are usually not present in the old definition formats. I'm updating monster intelligence attributes (even simulated fear), tweaks to projectiles, and updating the overhead map to show just about everything. I also added cheat codes... I even made sure that NONE indices and NULL data pointers are handled correctly (if an object goes off the map, surely a piece of code will pass -1 to "get_polygon_data"). Right now, it's in the middle of modification, and is not in a compile-able form. I can't legally post any source, but I might be able to put up my notes and/or a compilation with a secret password lock.

   6/1/2004

Yes. I beat DOOM II under the "Ultra-Violent" difficulty level. It took a me about a week, and it drained me a little... but I did it. I can't remember exactly when I started -- and I started over again at around the fifth level. The earlier levels seemed easier at that point -- and they are.
Hints to DOOM II: check around corners beforehand, save your game before climactic points, and remember to save your ammunition for the next level! Otherwise, you die. Don't be afraid to go nuts and try out dumb things after you saved. You might learn a new strategy on getting the f**k out!

Marathon is not a DOOM clone. It may have some of the same technology, but that wasn't the major part of it. Marathon was supposed to be politically correct; about the adventure and plotline, not about the carnage. They just needed a general 3D-like engine that'll give a good enough framerate on a 68040 Mac (since the fps genre was hot at the time).

   5/25/2004

Beat "Oh no! More Lemmings". It took a while. I am drained. I also fixed a bug in DOOM II 1.01, allowing me to use different WAD files, played some "Hoover Dam" thing -- it doesn't look like anything recognizable. 'Finally beat DOOM II (in GOD mode). You have to aim a little lower, when it comes to that red hole (supposed to be the exposed brain). The map maker didn't set the ceiling height high enough, so that even if the rocket passes the rectangular hole... it still may hit an upper side inside. I first saw that (turning clipping off) and stood on the sector that connects the so-called outside world to the big monster. Even though Marathon is better in most of everything, the DOOM II has less bugs and drawbacks -- well, except for the engine vertically aiming for you, and (saying the same thing) sh**. Behhlahhhhhhhhhhh.....

   5/11/2004

Absence of mind... almost completely deadpan...
Learning PowerPC opcodes... I got RISC technology down, as far as learning the 32-bit-come-64-bit architecture. IBM made G5... and whatever. Because every instruction is 32 bits, it requires two instructions to attain 32 bit values in registers -- immediate values are nested in the lower 16 bits of the "load immediate" instruction, and other means, like absolute addressing require the high 16 bits be put first. I plan to scare away the Wintel mo fos that may be reading this page!
Pict fontier...   **eyrgggghhtttttt**    last of it -- "beta" version. Bug fixes, and sh**. Go to the website for more information and a new download. Two visible bugs visible- must... try... to fix...

   4/25/2004

"Splinter Cell" looks good, sort'a (they took out the gore to get it to a teen rating). At least it has a plot, unlike some'a'tha crap out there.
Not much sleep... rewriting this shtuff for Marathon has proven difficult. I've turned it into Barfathon. I feel like ripping out a chainsaw and turning everything (especially my head) to pulp -- anyone feel like that these days? Everyone that is "Marathon" is paying attention to the Aleph One project (which is painfully slow - years after years). Duke Nukem 3D has an engine far better than Marathon Infinity's -- actual 3D permutations, however slow. The only reason I don't play Duke Pukem is because the people that developed it set up such a sh***y plot, and are reluctant to take real advantage of the engine. And so, they made it unworthy. So... my job is to climb that mountain, and push Duke, sorry, Puke, off that ledge.
Oh, and for what I said yesterf**k -- it's 64-pixel increments - not 32 (for PPC cache boundaries - 6 bits). I'll see if I can get a screen shot up here (that was mainly the purpose of putting the feature in).

   4/22/2004

The key to B&B, slopes and rotated 2D sprites in Marathon is to remove all of the "texture vertical polygon" stuff. I did that. Now I have to rewrite all this sh** to get it to work with the shading tables. I may try to use lightmapping techniques... which is not very fast... on a 68k Mac. It's all torn apart, so I need my own team to accomplish this, yet, no one is listening...
For screenshots, I'm dynamically changing the F-Keys to use F13 to take a render screen shot -- it doesn't take the black background, just the front (bottom) panel, and the terminal or view and/or overhead map*. The screen size now goes by 32-pixel width increments, instead of "full_screen"-"75%"... etc. The limitation of the screen size is limited to only what you start the program at.

I'm setting up the overhead map to be allowed for translucent crossfading with the view, as well as rotation around the player, texture mapping, and facing angles of monsters. Oh, and a preferences panel for all the overhead options (except for cheats - which I also changed).

   4/18/2004

I got to the full gameplay! Only two problems: film playback does not work, and in somewhere along the line, there's a horizontal polygon line bug. And even further... after more optimizations... I broke it. That's all folks.

Check out the benchmarks of the Dual G5 2.0GHz with Radeon 9800Pro (Mac, retail edition) -- at 1024x768, it delivers (at least) a whopping 339 frames per second in Quake 3 (higher than the P4):
http://www.barefeats.com/g5b.html

   4/12/2004

Finally, after all of the buggy improvements and optimizations, I got to the gameplay. I am now ready to test the "fixed" texture_vertical_polygon_lines_16. It's slightly faster, because it uses bitfield operations.

   4/5/2004

Again, not getting much sleep. In fact, I got the most sleep in the car (not mine). The project is going like fudge; it may taste good, but it goes down sllllooooooowwwwwwlly. The sound works, the preferences dialogs work (including the monitors dialog I had to reverse engineer and rewrite), the fades work and show up properly in every bit depth my machine is capable of. Everything works... except gameplay! Just small little problems that cause it to tank. I don't know what I added, or didn't do, that creates this sh**...
I added more extensions definitions - just about everything I could think of to be overwritten: control panels, platforms, player shapes... at least seven of 'em. I merged the export and import definitions into one, making it simpler, yet costly - I have to set up variables for the counts of all of the definitions, since they are now variable in size and can hold a lot more (24 weapons is just one of them!). The default definitions now get loaded differently (instead of at A5 initialization). On tha Mac, it uses resources of the matching tag of ID zero- which forced me to modify a copy of the export_definitions stuff; it is an application program, not an MPW tool, and it uses StandardPutFile to save the Physics (I will have to write in the code for the matching tag resources as well). The A5 world is getting quite large -- over 50K. No one cares about this project, so I sux.

   3/31/2004

Last day o'the month - what I meant to say on the 27th was, "I'd like to see one of the Aleph One team do all'that in 15 days -- I'm pretty sure if they'd put their minds to it the whole team could do over the minimal stuff again (geting it to compile) in that amount of time.
Addin' mo' stuff, like extended shape discriptors (up to 16 color tables, 255 collections, and 65535 shapes), and replacing "indexes" with "indices" (at least I know how to spell that sh**). None of this tested, yet... I've cleaned up most of the source code, and it might be a good idea to make some of the constant width terms universal - eg.: "word" should be "UInt16". - I think I'll do that, now. I will try to keep compatibility (unlike the other guys), making every new thing that would effect the flow, optional. Films should come out the same!

.... and according to the license (I was meaning to say, earlier), I can't display any of the modified code, or make any builds available ('Can't do nothin').

   3/27/2004

I've had hell... or something like it- nah, I deserve to be beaten. I have the Marathon 2 source built, and running... however, of course, it's buggy. So far, the main screen fades in and out correctly and the buttons work and draw correctly -- except only in 8 bit mode. In 16 bit, the picture rectangle is off vertically, and the clipping screws up -- you think you're clicking on the center "info" button, yet it starts a new game - I learned my lesson from that (it crashed, horribly). I also added long-distance junk.. I'd like to see the Aleph One team getting that far in 15 days. (they suck)

   3/16/2004

I got those renderings and their sources in whole. As far as the marathon2src thing (I call the project DurandalClassic), according to the license, I can't display ANY of the source code unless it's made open-source. I don't have the money to make it open-source. I don't have any money. This geocities website is f**kin' free!!!- man... The only legitimate reason that a project would exist for it is 'cause compatibility issues. The 68k platform hasn't vanished completely, you, Mo Fos!!
I got forty hours the other day... 40 hours of being awake. And now I am not sleeping, again. I sleep during the daytime, and this page is being edited in the daytime.

   3/14/2004

Yeah... the cut is getting deeper- I downloaded the marathon2src (source files) in the AppleDouble format a couple of days ago, and I have already got almost half of the sources compiling (due to my programming experience). Of course, with a number of groups already doing this, it is appearent that I have to recreate the "cseries" files; the Bungie programmers intentionally left out a few to make sure that their commercial game wouldn't get ripped of -- for one, the serial number algorithim (makes sense). I am doing this completely without the help of AlephOne. I haven't accessed the directory of the AlephOne sources I downloaded from December. Well... maybe I'm getting ahead of myself. At first, I had to translate the AppleDouble format to MacBinary catalogued format, so I created a translator with HyperXDB (HyperTalk and priceless externals). I've also rewritten some of the ANSI junk (fsprint, assert), and it wasn't even necessary... ahhhh... I spend tooooo much time on being an ass. I'm just gonna have to inlude some of the already programmed mem/string utils stuff.
Out of all of this... of course, I've gotten no sleep, today, on the fourteenth of March of 2004............ Sunday.

   3/6/2004

Precompiled the Universal 3.4 headers for Symantec C++, to make it more compatible, and such and such and such and such and such. Does anyone out there want a copy? It took a f***load of f****oads to fu****ad - whoops. Also, working on new gaming designs and functions, reinventing the "thinker" (basically, everything that acts), from DOOM; just the name, no actual code was copied -- although it did look a little similar after I looked at the DOOM code after I made my own version of the structure (like I can remember everything from the DOOM code - not much to look at).
Also, 'Testing out Apple Event handling (I should've done that a long time ago for the other programs I've worked on.
I don't get to talk about much personally, 'cause I'm boring. I stop myself... oh, and also 'cause I have no internet access at home.

   3/4/2004

'Tis me Birthday.... Whocares!!!
   At least I'm no longer a f**kin' teenager.
My old Mac is too slow to make POV-Ray a productive tool. I have no "blob" or mesh programs, and I don't feel like making one.
I'm back on my map editor; fixing bugs and implementing features, like gameplay (hell of a lot of work to get that to work - with endless linking errors). 'Back onto my old PixMap fonting program (to get it done, and over with).

   2/25/2004

Rainbow Six 3


Star Wars Jedi Knight: Jedi Academy

   2/18/2004
'Terribly sorry. Persistance of Vision, not "Point". So far, no unwelcome responses to my E-mails, just no responses. The 640x480 rendering of the "Balls" scene took over three hours to render. I will also note that in the image on the left of the two renderings has a bad y co-ordinate for the top of the box, literally clipping off the bottom of the two spheres (for anyone who noticed the inconsistency of the two images), so it was moved down to only touch the spheres.
I will try fix things on this site, and try to reserve space for more stuff to upload. I changed the file name of the marathon page to marathon.htm deliberately to get someone to NOTICE!!!!!

   2/12/2004
One of the best things I've ever stumbled upon. Point Of Vision. A free (non-redistributable) program. Here's are two renderings on my old 68K Mac:


The one on the left shows no anti-aliasing, and a point-sourced light (not smooth). The right, both on. Both images are with radiosity on.
I will upload 64bpp PNGs, soon.

   2/3/2004
Need to work on releasing DB, need to work on releasing DB, need to work on releasing DB, need to work on releasing DB... oh sh**...
Sent an E-mail to David A. Kopf (DAKX compression president), expecting an unwelcome response because of the crap I put in the E-mail, like every E-mail I write... 'cause I'm an ass...

   1/27/2004
Nothing much... just working on my assembler. Got almost every 68LC040 instruction down. It will be written in THINK C (of course). Takin' more off this page, moving 'em to the repository for old crap - old pages, since it's nearly February. Dumbit has changed to HyperXDB, as in, HyperCard Extension DataBase. I sux.

   1/21/2004
I've already proven myself to be an ass. Well a couple files have been updated, and my venture into bullsh** has begun.

C++:

try {
  function_call();
} catch (const char* string_exception) {
  if (!strcmp(string_exception, "the_worst")) {
    handle_it();
  } else {
    throw;
  }
} catch (...) {
  throw;
}

Ada:

begin
  function_call;
exception
  when the_worst => handle_the_f__ker;
  when others          => raise;
end;

Bare:  (name is subject to change...)

{
  -- install a temporary exception routine, just in case
  routine exception oh_sh__();
  -- call it
  function_call();
  switch(failure) -- the most recent 'result'
  {
    case the_worst:
      ;
    default:
      oh_sh__();
  }
}

   1/14/2004
My new resource format to go with a system extension I'm developing. It is weak, now. It's similar to the DFF format, yet, it should deliver more performance than the MacBinary format, allows disk write protection (to prevent and/or detect file corruption), compression, and encryption... plus more.

   1/8/2004
The HTML version is here!! 1