<none>

A new information packer is getting ready for release. It currently exists in source form and does not have any interface. The author is not Hrumer, but Roman Petrov (namesake of Megus'a^BW) is unknown to anyone except us. The author lives in the village and studies at our MarSTU, so he only has time to write a packer on
weekends. Brief characteristics of the packer:

Memory requirements - 128 kilobytes.
Maximum file size - #b800 bytes (46 kilobytes).
Search window - #a000 bytes (40 kilobytes).
Compression:
There are 2 compression algorithms: for code and for texts (the depacker is the same).
At 30 kilobyte code files, the latest version of the packer that I saw
won about 200-300 bytes over Hrust and ZX-Zip. During my last conversation with the author last week, he said that the compression ratio has increased even more, and he has now gained about 200 bytes on a 30-kilobyte ArtStudio file compared to the standard RAR. I think the author won’t lie, besides, I’ll check
it myself at the end of the week.

For texts, the advantage over ZXZIP and HRUST is even greater - texts with a volume of about
40 kilobytes are compressed by about 2 kilobytes better.

The compression speed of a 40 kilobyte text file is 2 minutes 40 seconds (for
Hrust - 1.20 seconds). Well, the algorithm is more
more sophisticated than Hrust, and the search window is 2.5 times larger.

Unpacking:
The size of the unpacker is about #118byte. When unpacking, a buffer of #5c0
bytes is required. The unpacker does not change SP during its operation. The unpacking speed is 2-3 times lower than that of HRUST (this is again due to a different packing algorithm), but this is not so noticeable.

On the basis of this packer, with the permission of the author, we will make an archiver like HRIP.

Bye, All.
WBR, Vivid^Brainwave of the X-Project association.

AM> A new information packer is being prepared for release. It still exists in the form
AM> With 30 kilobyte code files, the latest version of the packer that I
AM> saw was superior to Hrust and ZX-Zip by about 200-300 bytes. During my last conversation with the author last week, he said that the compression ratio had increased even more, and he had now gained about 200 bytes on a 30-kilobyte ArtStudio file over a standard RAR file. I think the author can’t lie
AM> On texts, the advantage over ZXZIP and HRUST is even greater - texts with a volume of about 40 kilobytes are compressed by about 2 kilobytes by about 2 kilobytes.

I’ve been using Plastic for about six months now, it compresses 20-30 percent better than
Hrust, but at the same time it also packs longer (about 8-10 kbyte/min).
It also exists in the source code (Alasm), and they probably won’t release it...
It’s a pity...

People! Or maybe you all already use some rare packers, but at the same time, laughing quietly, manage to discuss Hrust? ;)

Ours is for you. Alexey

Tue 17 Apr 2001, at 07:32:22 Alexey Voskresensky told Aleksey Malov about
Packer.

AV> I've been using Plastic for about six months now, it presses 20-30 percent better than
AV> Crunch, but at the same time it packs longer (about 8-10 kB/min).
AV> It also exists in the source code (Alasm), and they probably won’t release it...
AV> It’s a pity...
Is it possible to get these source codes (for home use only, of course)?
By the way, the new version of the subject with 30 kilobyte codes wins more than 1
kilobyte from Hrust, and about 200-300 bytes from RAR. It makes sense
to use the subject even for compressing 4k intros (on 5 kilobyte files, the gain compared to hrust/hrum (naturally, taking into account the length of the unpacker) reaches
30-70 bytes, which is fundamental in 4k intros).

AV> People! Or maybe you all already use some rare packers, but at the same time, laughing quietly, manage to discuss Hrust? ;)
If this remark applies to us, then it does not apply to us ;). For now, let's pack with
what we have (i.e. Hrust/Hrum).

Bye, Alexey.

don’t you want to share?
AV> they won’t release it
AV> It’s a pity...
who is the author?

Max

AV> kbytes/min).
AV> Also exists in source code (Alasm), and probably they won’t release
AV> it...

Why won’t they? can I have the source code then?

AM> A new information packer is getting ready for release. It currently exists in
AM> source form and does not have any interface. The author is not Hrumer, but
AM> Roman Petrov (namesake of Megus'a^BW) is unknown to anyone except us. The author
AM> lives in the village, studies at our MarSTU, so he
AM> has time to write a packer only on weekends. Brief characteristics of the packer:
AM>
AM> Memory requirements - 128 kilobytes.
AM> Maximum file size - #b800 bytes (46 kilobytes).
AM> Search window - #a000 bytes (40 kilobytes).

AM> Compression:
AM> There are 2 compression algorithms: for code and for texts (one depacker and the same
AM>). On 30 kilobyte code files, the latest version of the packer,
AM>, which I saw, beat Hrust and ZX-Zip by about 200-300 bytes. When
AM> during my last conversation with the author last week, he
AM> said that the compression ratio has increased even more, and he has now won
AM> on a 30-kilobyte ArtStudio file about 200 bytes from the standard RAR.

And RAR With what compression ratio? On one file?

Didn’t you try to compress the result with RAR?
One file is, alas, not an indicator. Let's take 30K zeros and write something like this
"unpacking" -

ld hl, start
ld de, start+1
ld (hl),0
ld bc, len
ldir
ret

Press steeperthan RAR!

AM> I think the author won’t lie, besides, I’ll check it at the end of the week
AM> myself.

The public is hungry for adequate tests :)

Alex Raider / Flash Inc.
I’m really looking for the French cartoon “The Lord of Time”! If anyone knows, please respond!

We urgently need subprograms for working with the Kempston mouse, preferably, or rather
necessarily with comments. For HorrorWord II.
ps: You can via netmail.

For this I say goodbye, with respect Alexey Kravchenko AKA Kurleson^HS^CPU

[Speccy] [HorrorSoft] [CyberPunksUnity] [Army SUX!!!]

... np: - - COMPOSED:PHANTOM LORD:26121196

This is a test message...

Aleksey Malov writes:

AM> By the way, the new version of the subject on 30 kilobyte
AM> codes wins more than 1
AM> kilobyte from Hrust, and about 200-300 bytes from
AM> RAR. It makes sense
AM> to use the subject even for compressing 4k intros (on 5
AM> kilobyte files, the gain in
AM> compared to hrust/hrum (naturally, taking into account
AM> the length of the unpacker) reaches
AM> 30-70 bytes, which is
AM> fundamental in 4k intros).

How to contact the author?

PS. This is a test.
--
Sent via the Talk.Ru server - http://www.talk.ru

DP> How to contact the author?
Only through me. He lives in the village, studies at the same university as me.
He doesn’t have a telephone (and, naturally, a modem or e-mail).
If you want to ask anything, ask now. It will be impossible to find it in the summer. I
hope that he will finish his packer by June (in principle, as he says, there
the algorithm remains to be optimized a little and the interface to be screwed in).

He uses the method of repeating strings and uses Huffman coding. As a
consequence, when unpacking, tree memory of size #5c0 bytes is required.

Bye, Dmitry.

AA> And RAR With what compression ratio? On one file?

AA> Didn’t you try to compress the result with RAR?
AA> One file is, alas, not an indicator.

AA> The public is thirsty for adequate tests :)

Indeed, it is! :).

There is a proposal: collect a selection of test files, where you can include a wide variety of files: text, music, code blocks from games, etc.
The length of the files is from 200 bytes to 42 kb. Create a full disk, and carry out
testing on these files.

If you have any files, don’t consider it difficult, throw them at hru...@inbox.ru.
In addition, such a selection of files will make life easier for packer writers - they will have
something to rely on when choosing a particular algorithm.

Happy. Dmitry.