| |
AlexC
Registered: Jan 2008
Posts: 299 |
best cruncher running on c128
While everyone is using Exomizer those days on PCs I am wondering what was the best cruncher that could utilize fast mode of c128 (in c64 mode). |
|
| |
The Shadow
Registered: Oct 2007
Posts: 304 |
Darksqueezer. There was a 2mhz C=128 version of Timecruncher created in the late 80s. |
| |
Ninja
Registered: Jan 2002
Posts: 411 |
I remember using some version of CruelCruncher which used the 2MHz-Mode. I think ByteBoiler does use it, too. But come to think of it, I wonder if there isn't a cruncher for C128-mode? The extra memory might even more help than the extra speed. |
| |
Mace
Registered: May 2002
Posts: 1799 |
Found another one: Time Cruncher V3.3 (Special Version) |
| |
Stryyker
Registered: Dec 2001
Posts: 468 |
I don't know about sizes but Abuze Crunch supports 2 MHz mode. |
| |
MaD ][
Registered: Sep 2004
Posts: 23 |
Time cruncher v4.2 (Network version) or V5.0 uses 2MHz C128 |
| |
algorithm
Registered: May 2002
Posts: 705 |
I know this may be off topic as the initial query was in regards to which cruncher performs well speed/compression on c128. But one thing is the following..
Why use a C64/C128 for compressing data? The inconvenience of having to wait a huge amount of time when on the other hand you can have the data compressed in seconds on a PC system.
For exploring new vic/audio tricks etc or running the final demo etc on C64 to see if there are any glitches, flicker etc is fair enough, but all the other stuff (compressing/converting) etc should IMO be done on a PC or similar system |
| |
Frantic
Registered: Mar 2003
Posts: 1648 |
Algo: You're not allowed to speak until you've released all those unreleased PC tools that you promised from your resurrected harddrive. ;) |
| |
algorithm
Registered: May 2002
Posts: 705 |
The trimode trucolor conv, imci fmv converters etc (Ouch. given too much away :-} ) will be released when the trackloader is complete ;-} Have barely any time left (9-5 job, girlfriend etc - excuses excuses) |
| |
Mace
Registered: May 2002
Posts: 1799 |
Algo: AlexC clearly asks what WAS the best cruncher that uses 2MHz mode. He doesn't ask what IS the fastest way to crunch today ;-) |
| |
algorithm
Registered: May 2002
Posts: 705 |
Quote: Algo: AlexC clearly asks what WAS the best cruncher that uses 2MHz mode. He doesn't ask what IS the fastest way to crunch today ;-)
I know :-} Hence why i stated that it may be offtopic to mention it. But whats the point really? No advantage at all unless the c64/c128 is the only computer available |
| |
Mace
Registered: May 2002
Posts: 1799 |
Quote:But whats the point really? Expansion of knowledge, I'd say. ;-) |
| |
algorithm
Registered: May 2002
Posts: 705 |
Yes, I agree that the person wants the answer to the query and gets answers which is great but almost feels like something such as.. 'How can i improve my typing speed to type the below faster..'
poke 49152,169
poke 49153,0
poke 49154,141
poke 49155,32
poke 49156,208
poke 40157,96
when a quick alternative would be something such as
m $c000 a9,00,8d,20,d0,60
Hihih. Just me in a funny mood. ignore ;-}
|
| |
Mace
Registered: May 2002
Posts: 1799 |
Quote:How can i improve my typing speed to type the below faster.. I believe there were people who could type this in screencodes, which might be even faster, but harder to know from the top of your head :-)
[inverse on]shift+q[inverse off]cbm+k ...etc... |
| |
algorithm
Registered: May 2002
Posts: 705 |
Quote: Quote:How can i improve my typing speed to type the below faster.. I believe there were people who could type this in screencodes, which might be even faster, but harder to know from the top of your head :-)
[inverse on]shift+q[inverse off]cbm+k ...etc...
By the time the commodore/shift key and combo keys are pressed together with the required key it would be faster just to type something such as 'a9' Just all for fun really. for productivity and anything decent, then assembler it is |
| |
Richard
Registered: Dec 2001
Posts: 621 |
Didn't Crest's version of the Cruncher AB use 2MHz? |
| |
The Shadow
Registered: Oct 2007
Posts: 304 |
There is a functional reason for using a Commodore based cruncher. It saves space. The Darksqueezer's decruncher uses only slightly more than a block of memory. If you are going to spend hours composing music, cracking a game, creating a demo then to spend an extra few minutes crunching it with a good Commodore cruncher is not that big a deal. Using a PC based cruncher only serves for a specific need such as the Gold version of Ultima IV. |
| |
chatGPZ
Registered: Dec 2001
Posts: 11386 |
hu? and what exactly stops you from using exomizer for...anything else too? o_O |
| |
Mace
Registered: May 2002
Posts: 1799 |
Yeah, what Groepaz says.
I don't see why you wouldn't be able to use PuCrunch or Exomizer for anything that you'd want to do with a C64-based cruncher.
In fact, using Exomizer might be even easier if you're developing on a PC. |
| |
The Shadow
Registered: Oct 2007
Posts: 304 |
I have always admired the power of the Darksqueezer. Antitrack sent me this email a few months ago. I asked him how Darksqueezer compares to the PC based crunchers and he said...
Andy,
The idea for DSQ-REU-version emerged after going to the universities' "data
structures and algorithms" lectures. It flashed up in an instant, after learning what a
"chained list" is. Technically speaking, it is nothing new - data structures have been
around for much longer ; the inventors of LZW77/78 surely also had more memory
than an average c-64. ARJ, and LHA were around for *years* on Amiga and PC
computers before I transfered a tiny part of their ideas to the commodore world.
The first coding of it (i.e. most of what you are seeing in the article) just took 20
minutes on a weekend. Fascinated by the new gained crunching speed, I crunched
dozens of testfiles over and over and refined a few parameters for DSQ on the way,
whilst working on it. All REU-improved DSQ versions were done on a weekend in just
a couple of hours.
Exomizer can compete very well, but the problem is the decruncher. Its code is so
long, due to own Huffmann (and other) tables, that he almost fills the whole area
$0400-$0800. How many multiload games offer so much free memory?
Anyhow, the difference are typically just a few blocks or even just a few bytes, thus it is not really worth the trouble of transfering the data from PC to 64 and back IMHO.
Long time ago, Rockstar modified DSQ a bit so the decruncher would be smaller
than $0100 bytes. That's nice enough for extreme cases of lack of memory.
Compression wise, Exomizer can compete very well.
The bad thing about other PC crunchers like ARJ and ZIP is, they need extra
memory to decrunch. For example. if you want to decrunch a 60 kb file on the PC
using ARJ, he will need 60kb RAM for output and 60kb RAM as a scratchpad. We
don't have so much RAM inside the c-64. For the PC however, there is no such a
problem.
Yours
ATT
|
| |
chatGPZ
Registered: Dec 2001
Posts: 11386 |
your point beeing? exomizer certainly gives a better result than pretty much any c64 based packer. (in a fraction of the time) and the depacker isnt quite as ridiculously huge as he says, i wonder where he got that number from :) (just look at the decruncher sources that come with it). and the same is true for pucrunch too (slightly worse pack results, but even shorter decruncher, and much faster packing and decrunching) |
| |
The Shadow
Registered: Oct 2007
Posts: 304 |
For a multi-load game with very limited memory, I would use the Levelsqueezer for its good squeezing results and short decruncher. For some games, there is only a tiny amount of memory to work with. Pucrunch is not familiar to me. It deserves to be examined and experimented with. Darksqueezer has proved to be very reliable. I have found cases which Cruel Cruncher will corrupt data. Darksqueezer is the most incredible tool to emerge from the crackers of the old scene. |
| |
AlexC
Registered: Jan 2008
Posts: 299 |
Quote: Quote:But whats the point really? Expansion of knowledge, I'd say. ;-)
Thanks for all the answers. My point and the reason behind the question was... curiosity about how cruchers and packers had evolved during years. Use of c128 fast mode and REU was an obvious step.
So yes, Mace is right: expansion of knowledge. I don't think anyone is doing crunching on real machine today, especially if you look at Exomizer. |
| |
chatGPZ
Registered: Dec 2001
Posts: 11386 |
Quote:For a multi-load game with very limited memory, I would use the Levelsqueezer for its good squeezing results and short decruncher. For some games, there is only a tiny amount of memory to work with.
the exomizer decruncher isnt really bigger, and it certainly packs better. same for pucrunch. |
| |
Burglar
Registered: Dec 2004
Posts: 1101 |
exo decrunch is ~226 bytes, excluding the ~100 byte decrunchtable.
hardly 0400-0800 ;) |
| |
chatGPZ
Registered: Dec 2001
Posts: 11386 |
and you can make it even shorter if you disable literals (and its still better than squeezer then =P) |
| |
Fungus
Registered: Sep 2002
Posts: 686 |
Exom rules.
Levelab2 2nd place. (c64/128+REU)
Squeezer is ok for stuff with lots of little files. 4.0 is best, but you had to be special to get the unpacker for it. (I have it :P)
I used Level Crusher sometimes too. Shortest unpacker there is.
Also, if the unpacker is too big, RECODE IT like the rest of us did... Always useless crap attached to it if you use your own loader.
Almost everyone I did was under $0100, except levelab2 and exom (about $0128 or so). Unless I made a special version which didn't use zeropage (some game can't disturb it, and there is no memory to buffer it)... etc
Depends on the situation of course!
Timer was ok for 1986, cruel (cruncher or level) was always bugging crap. Also can get better results by EC/RLE packing the data first with C64 cruncher (levels too!) code your own unpacker for such ;)
Damn, this basic stuff... back to nursery school.
|
