GMX

People, if anyone has a subject, then tell me how it emulates the scan
of the Pentagon? That is Are border effects 100% identical to screen ones in the
"Rage" demos and in the hidden part of "Insult megademo" (run: RUN "INSULT".I LOVE
CODEBUSTERS )?

We're just doing a demo on CC1, in which this compatibility is relevant in some
effects.

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

AM> on-screen in the demos
AM> "Rage" and in the hidden part of "Insult megademo" (run: RUN
AM> "INSULT".I LOVE
AM> CODEBUSTERS) ?
Everything is perfect! Even MMA said “RULEZ!”, although he didn’t admit it later. ;)

AM> We're just making a demo on CC1, in which this
AM> compatibility is relevant in some
AM> effects.
Anything goes! Some demos don't even work on real Pentagons. :( There are glitchy options.

ZXNet: 500:8462/1 ~/ <~ http://www.zx.ru/zs/
Fido: 2:5057/21.93 /_ _> ftp://zx-museum.org.ru/zs/

Wed 16 May 2001, at 15:10:59 Sergey Zotov told Aleksey Malov about GMX.

SZ> Perfect! I don’t have a subject, but on CC999 I saw with my own eyes what it
SZ> can do!
Then I’m calm.

Bye, Sergey.

One day on Monday, May 14, 2001 at 17:58:45 Aleksey Malov wrote a letter
to All about GMX:

AM> People, if anyone has a subject, tell me how it emulates
AM> the Pentagon scan?

During the time that I have been using it, there have been no glitches with emulation saw.

AM> Ie. do border effects coincide 100% with screen
AM> in the "Rage" demos and in the hidden part of "Insult megademo"
AM> (run: RUN "INSULT".I LOVE CODEBUSTERS )?

Rage is definitely running fine, somehow I didn't launch the hidden part of Insult, and
I don't have it at hand right now.

I wish you good luck!
With respectfully Fyodor Odegov.

FO> Rage is definitely going fine,

It’s a shame that GMX was on it during the development process ;)

All the best.
Dmitry.

... LD..X-TRADE... [ZX] [Quake]

DL> of course, GMX was based on it during the development process ;)
By the way, if it’s not a secret, how in Rage does scrolling along the border not jerk,
or do you use your own player while it’s running, running at the same
same time, and store the music in an unlocked form?
By the way, I have glitches, or in the last part (fullscreen) rotator)
is it really 25 fps, not 50?

Bye, Dmitry.

SZ> Anything goes! Some demos don't even work on real Pentagons. :( There are
SZ> glitchy options.

What kind of demos don’t work at the Pentagon????

▌┘│ ■ ▐ ┬─┐▌ ┬──
Sorry, they brought fresh beer... ▌│└─│┌┬┐ ▌ ▌_│▌-┤▌_┬

One day, Fri May 18 2001 09:19, Aleksey Malov wrote to Dmitry Lomov:

AM> By the way, if it’s not a secret, how in Rage scrolling along the curb does not
AM> jerk, or do you use your own player while it’s running,
AM> running at the same time, and store the music in undeclared
AM> form?

yes. and the whole program is aligned to 4 clock cycles.

AM> By the way, am I having glitches, or is the last part (fullscreen rotator)
AM> really 25 fps and not 50?

Is it bothering you. there is 50 fps, otherwise everything would jerk. and in general, non-frame
border effects are nonsense ;)

All the best.
Dmitry.

AM>> is twitching, or do you use your own
AM>> player for the duration of its operation, which runs at the same time, and store the music in
AM>> uncluttered form?

VK> unbundled player is not enough, you need to have exactly
VK> 71680t in each frame and there can be (when using halt at the end) from 71577 to
VK> 71680 depending on which cycle of the halt the interrupt pulse will come.

probably means that from one to three clock cycles can jump from one
frame to another. the amount will remain unchanged ;)

VK> To always have
VK> 71680 you need a special fix before the effect and after that the so-called
VK> fix'a the number of cycles in the interrupt must always be a multiple of four!

wrong. for this you don’t need any fix, you just need to align by 4.

VK> fixes can be found in the blue review for the 96th year and in the last issue of
VK> zx-guide (by alone coder).

I don’t know what it’s about, but I can believe that that fix solves the problem of
synchronizing the first command executed after HALT relative to the absolute
beginning of raster formation, because these two moments can oscillate relative to each other for several (1...3) cycles. The fix makes it so that the phase of the first command is always the same and does not depend on the previous progress of the program. Am I right?

Listen, post this fix here if you can.It's interesting to see what the
method is. At one time I invented my own, rather complex method. maybe this one will be
simpler and cooler ;)

and more. where were such fixes actually used? maybe there are demos that I don't know about? ;) after all, neither when scrolling along the border, nor with multicolor by
attributes, this is not necessary, but only when the need arises to
combine something on the screen with something on the border with an accuracy of one
beat.

AM> By the way, if it’s not a secret, how in Rage does scrolling along the curb
AM> not twitch,
AM> or do you use your own player while it’s running, running at
AM> the same time, and store the music in an unlocked form?

An unlocked player is not enough, you need to have exactly the same content in each frame 71680t
and there can be (when using halt at the end) from 71577 to 71680
depending on at what clock cycle the interrupt pulse will come. To always have
71680 you need a special fix before the effect and after that the so-called. fix'a number
cycles
in an interrupt must always be a multiple of four!
fixes can be found in the blue review for the year 96 and in the latest issue of zx-guide
(by alone coder).

cYa, J7n/4D.

Fri May 18, 2001, at 19:03:16 Dmitry Lomov told Aleksey Malov about GMX.

DL> yes. and the whole program is aligned by 4 measures.
Well, we know that.

DL> crushes you. there is 50 fps, otherwise everything would jerk. and in general, non-frame
DL> border effects are nonsense ;)
So it just seemed to me that each frame holds two ints ;)

Sat 19 May 2001, at 02:32:43 Vassili Klimov told Aleksey Malov about GMX.

VK> an unlocked player is not enough, you need to have exactly 71680t
VK> in each frame, and there can be (when using halt at the end) from 71577 to 71680
..
VK> in the interrupt must always be a multiple of four!
I already know that. I was just interested in the correctness of my suspicions about RAGE.

VK> fixes can be found in the blue review for the year 96 and in the last issue of zx-guide
VK> (by alone coder).
I read the review.

Bye, Vassili.

DL> and more. where were such fixes actually used? maybe there are demos that
DL> I don't know about? ;) after all, neither when scrolling along the border, nor with
DL> multicolor by attributes, this is only necessary when
DL> there is a need to combine something on the screen with something on the border with
DL> accuracy
DL> up to one clock cycle.

What is the latest version of the subject? And if possible, throw it at me, please!

From the bottom of my heart Ciao, Dmitry. [I ZX] [SHADOW MAD]

DL> is incorrect. this does not require any fix, you just need to align it at 4.
sorry, I was wrong;)

DL> I don’t know what it’s about, but I can believe that this fix solves the problem
DL> of synchronizing the first command executed after HALT relative to
DL> the absolute start of raster formation, because these two moments can
DL> fluctuate relative to each other for several (1...3) cycles. The fix makes
DL> so that the phase of the first command is always the same and does not depend
DL> on the previous move of the program. Am I right?
here I didn’t understand you a little;) the fix runs only once before the effect,
then you don’t need to fix anything - just the number of cycles in the interrupt should be a multiple of 4m
DL> listen, post this fix here if you can. It's interesting to see what
DL> method there is. At one time I invented my own, rather complex method. maybe
DL> this one will be simpler and cooler ;)
hehe, I saw it myself and was very surprised:

;Setting procedure with an accuracy of 1 clock
;for border effects
;(c) AlCo 09/08/2000 (ZX-Guide 3 SUPPLEMENT)
LD BC,#6FF
;no more than 6 passes!
;first 2 passes - blanks:
;1st - to hit IMER during INT,
;2nd - to score D.
DI
LD HL,IMER
LD (#BEFF),HL
LD A,#BE
LD I,A
IM 2
BACK XOR A
LD HL,LOOP ;+2t
EI
LOOP DEC A
JP (HL)
;-1t (INT=11t)
IMER POP HL ;+2t
LD E,(HL) ;-1t
RL E ;"JP (HL)">128
RLA
LD E,D
LD D,A
SUB E
SUB C ;has D increased?
SBC A,A
OR 4 ;-1t 4=EXIT,#FF=NO EXIT
DEC B ;total +1 cycle per frame
AND B ;For idle NO EXIT:)
JR NZ,BACK
END ;39 bytes in total!

DL> and more. where were such fixes actually used? maybe there are demos that
DL> I don't know about? ;) after all, neither when scrolling along the border, nor with
DL> multicolor by attributes, this is needed only when
DL> there is a need to combine something on the screen with something on the border with
DL> accuracy
DL> to one clock.
the other day I received 16k from linoleum, there are also similar things in the confusion demo from
scene and in 7UP with paradox '99 it seems.
By the way, regarding Insult, there is no fix - you need to press enter so that
the screen letters coincide with the border ones.

cYa, J7n/4D.

GK> What demos don’t go to the Pentagon????
don’t you remember how many demos didn’t go to tsts000?
but there was a real pent...

cYa, J7n/4D.

ZXNet: 500:8462/1 ~/ <~ http://www.zx.ru/zs/
Fido: 2:5057/21.93 /_ _> ftp://zx-museum.org.ru/zs/

AM>>> I.e. Are border effects 100% identical to the screen ones
AM>>> in the "Rage" demos and in the hidden part of the "Insult megademo"
AM>>> (run: RUN "INSULT".I LOVE CODEBUSTERS )?

DL>> demos I don't know about? ;) after all, neither when scrolling along the border,
DL>> nor with multicolor by attributes, this is needed only when
DL>> when there is a need to combine something on the screen with
DL>> something on the border with an accuracy of one clock.

AM> 7up and Confusion boasted that they have auto_fix applied there,

how is that?

AM> but on the real Pentagon there was no coincidence of the curb with the screen
AM> picture.

good fix ;)

DL> of course, GMX was on it for a while during the development process ;)

Maniac ;-)
Full screened rotator - you can write THIS if it’s cracked... I wouldn’t be able to.
Probably.

WBR, Jason.

[Team Obituary][Team Friday 13th][Team Evil Maniacs][Team Huge Axes]

DL> how is this?
I meant setting it to the first interrupt cycle, then leveling for 4
cycles, as usual. At 7 up we decided to use the ld a,i command for this matter, but to no avail.

Are you not going to CC1? At least to be present as spectators.

By the way, I would like to hear your opinion, as one of the elite coders of
Spectrum, about our Tryptomine Dream with CC'000 ;)

AM>>> applied,
DL>> and how is this?
AM> What was meant was the setting for the first interrupt cycle, then the alignment
AM> for 4 clock cycles, as usual. At 7 up they decided to plow the ld a,i
AM> team for this matter, but to no avail.

What were the prerequisites for them to believe that this team was leveling something out?

AM> Are you not going to CC1? At least as a spectator, be present.

as a spectator - a must.

AM> By the way, I would like to hear your opinion, as one of the elite
AM> Spectrum coders, about our Tryptomine Dream with CC'000 ;)

somehow I missed it. Is it on the internet somewhere? I'll check it out ;)

DL> I missed it somehow. Is it on the internet somewhere? I'll check it out ;)
You could find it on www.brainwave.dax.ru, but so far the site is in
several indecent condition (there were reasons for this). He will remain in this state for two weeks. And then you can steal it from there.

Maybe there’s also one on zx.da.ru. Not sure.

ZXNet: 500:8462/1 ~/ <~ http://www.zx.ru/zs/
Fido: 2:5057/21.93 /_ _> ftp://zx-museum.org.ru/zs/

One day, Mon May 21 2001 08:50, Aleksey Malov wrote to Dmitry Lomov:

AM> Well, if an interrupt occurs during the execution of this command, then the
AM> parity/overflow flag is set as if interruptions were
AM> disabled. Probably, their interrupt handler looked something like
AM> like this:

AM> isr jp pe,int1 ;or po (in the sense that the transition to in1 is performed if
AM> interrupt
AM> ;came between two commands ld a,i
AM> ei
AM> ret

AM> int1 ... ;the first interrupt cycle is caught.

AM> For some reason, this The feature worked, apparently, only
AM> for the authors.

if it all worked at all ;)

Sat 19 May 2001, at 12:25:50 Vassili Klimov told Dmitry Lomov about .

VK> hehe, I saw this myself and was very surprised:
Everything ingenious is simple!

VK> the other day I received 16k from linoleum, there are also similar things in the confusion demo from
VK> scene and in 7UP from paradox'99 it seems.
Don't tell my slippers! In the linoleum 16k intro there is no alignment at 4
beats, and, moreover, no coincidence of the screen with the border in the middle. While
on the same machine Rage works in the same way.

In confision and 7up on this machine there was no coincidence between the screen and the border.

The word “this machine” meant the real Pentagon-1024. The demos were launched
with the top memory turned off.

AM> In 7up and Confusion they boasted that they had auto_fix applied there, but on
AM> the real Pentagon there was no coincidence of the border with the screen image in them
AM>.
And on mine it matches (albeit with accuracy not to the pixel, but to 2x - read one
tact, but more on that already wrote...)

cYa, J7n/4D.

AM>> is in a somewhat obscene condition (I had my own
DL> for this, I found it on ZX.DA.RU. Nothing crazy, just a lot of plasma ;)
Now Lesha will be offended and say, that there is no plasma there at all :)
And it really is not there. We have long had an aversion to plasmas :)

Bye, Dmitry!
Megus/Brainwave^X-Project. http://brainwave.dax.ru

VK> Maniac ;-)
VK> Full screened rotator - it’s cracked, you can write THIS... I wouldn’t
VK> be able to. Probably.
heh, there are only 2 frames - each on its own screen, but with the border it’s obviously
the same thing...

cYa, J7n/4D.

VK> why don’t you remember how many dems didn’t go to tsts000?
VK> but there was a real pent...
Fate and Hell Raiders didn’t go because their files turned out to be, for unknown
reasons, overwritten files of some other demo.
Kag - because of the curvature of the car.
SKY4k - apparently because im2

I appeal to all potential demomakers reading this echo:
Switch pages using port #7ffd. Switching over the #fd port with 6
bits enabled solves the problem of working on Scorps, but on the Pentagon-512/1024 it clogs up the
RAM-Disk and causes the absence of the 7th screen (the #57 ban is turned on, and the hardware
displays the screen from the #17 can).
But don’t say that it’s faster this way. In the vast majority of cases
it is enough to optimize the code so much that there are enough cycles to switch
pages via #7ffd.

The latest demos from Real Masters, 4k SKY and many demos from
Enlight'96-97 suffer from this problem.

Bye, Vassili.

Sat 19 May 2001, at 14:35:13 Sergey Zotov told Grigory Klimenko about GMX.

SZ> Why did the demo from Real Masters on CC000 fail? Because the Pentagon
SZ> was glitching.
SZ> :) Well, she went on a scorpion. And it was on KAY too. Do you know why
SZ> didn’t go?
SZ> Because you need a stable bus, and 100%
SZ> of Pentagons cannot boast of such a bus. But a Scorpio, and even with a subj, can. :)
Do you mean "stable bus"? Is this so that im2 works properly? So they taught us to fight this starting from the ZX Review. But, still, it’s not a good idea to exhibit such a glitchy machine at a demo party. At CAFe'2000, in general, the Pentagon didn't work on the first day of the party.
I also want to express my deepest thumbs up to the authors of Fate trackmo, but this time
not for the squeaky music, but for the crooked multiloader. On my Scorpio I was never able to watch Fate to the end due to the fact that the second
part did not have time to load.

To all: Do you think it’s worth throwing here an improved version of our multiloader since Tryptomine Dream? The reading speed on computers in which interrupts
arrive synchronously with disk rotation has increased by 2 times. Min. speed -
2Kb/sec (this is the case when the disk rotation speed is 300 rpm (i.e.
5 rps), and the interrupt length is 70000 clock cycles + -100 clock cycles, so the interrupt
occurs during readingthe same sector for 20-30
turns. The reading speed on such machines is increased by reading
several tracks at once (the head moves back and forth along the tracks
of the file being read)), the maximum is 9Kb/sec.

Bye, Sergey.

One day, Saturday May 19 2001 10:02, Nikolay Maslennikov wrote to Dmitry Lomov:

NM> What is the latest version of the subject? And if possible, throw it in my
NM> soap, please!

let there be no subject newer than the old one. I'm tired of coding :(

DL>> problem of synchronization of the first command executed after HALT
DL>> relative to the absolute start of raster formation, because these two
DL>> moments can fluctuate relative to each other by several
DL>> (1...3) cycles. The fix makes it so that the phase of the first command
DL>> always becomes the same and does not depend on the previous move
DL>> of the program. Am I right?

VK> here and I didn’t understand you a little;) the fix runs only once before
VK> the effect, then you don’t need to fix anything - just in the interrupt
VK> there should be a number of cycles that is a multiple of 4m

judging by your words, you understand everything ;)

VK> Maniac ;-)
VK> Full screened rotator - it’s cracked, you can write THIS...

there was a certain fuse ;) and there was no laziness then. there was just a desire to do
cool. As far as I understand, it worked :-)

VK> I would not have been able to. Probably.

oh, stop it. you're a maniac like me ;)

All the best.
Dmitry.

... LD..X-TRADE... [ZX] [Quake]

NM> ëÁËÁÑ ÓÁÍÁÑ ÐÏÓÌÅÄÎÑÑ ×ÅÒÓÉÑ ÓÁÂÖÁ?é ÅÓÌÉ ÍÏÖÎÏ ÔÏ ËÉÎØ ÅÇÏ ÍÎÅ ×
NM> ÍÙÌÏ,ÐÌÚ!
íÎÅ ÔÏÖÅ ÜÔÏ ÉÎÔÅÒÅÓÎÏ.
óÅÊÞÁÓ Ñ ÐÏÌØÚÕÀÓØ Storm 1.3
ôÁÍ ÉÓÐÏÌØÚÏ×ÁÎÁ ËÏÄÉÒÏ×ËÁ ËÌÁ×ÉÁÔÕÒÙ - "ñ÷åòôù",
Á ÍÎÅ ÂÏÌØÛÅ ÕÄÏÂÅÎ "êãõëåH" (ÒÅÁÌØÎÙÊ SPECTRUM). ñ ÐÒÏÓÔÏ ÎÁÒÉÓÏ×ÁÌ ÒÕÓÓËÉÅ
ÂÕË×Ù ÎÁ ËÌÁ×ÉÛÁÈ. ðÒÏÂÏ×ÁÌ ×ÌÅÚÔØ × ËÏÄ. ôÁÍ ×ÉÄÅÌ Ä×Å ÔÁÂÌÉÃÙ Ä×ÕÈ ËÏÄÉÒÏ×ÏË.
HÅÐÏÎÑÔÎÏ ÔÏÌØËÏ ËÁË ÐÅÒÅËÌÀÞÉÔØ ËÏÄÉÒÏ×ËÕ. ëÔÏ ÚÎÁÅÔ ?

é ÅÝÅ, ËÔÏ Á×ÔÏÒ Storm'Á É ËÁËÏÊ Õ ÎÅÇÏ ÓÅÔÅ×ÏÊ ÁÄÒÅÓ ?

VK> why don’t you remember how many dem didn’t go to tsts000?
VK> but there was a real pent...

There was a really buggy pent! Upon arrival home, I looked
everything on my Pentagon without any problems.

▌┘│ ■ ▐ ┬─┐▌ ┬──
Sorry, they brought fresh beer... ▌│└─│┌┬┐ ▌ ▌_│▌-┤▌_┬

AM>>> You could find it on www.brainwave.dax.ru, but for now the site
AM>>> is in somewhat indecent condition (I had my own
DL>> I found it on ZX.DA.RU.

One day, Tue May 22 2001 12:27, Aleksey Malov wrote to Vassili Klimov:

AM> I appeal to all potential demomakers reading this echo:
AM> Switch pages using port #7ffd. Switching over the #fd port with the 6th bit enabled solves the problem of working on Scorps, and on the
AM> Pentagon-512/1024 it clogs up the RAM-Disk and causes the absence of the 7th
AM> screen (the #57 ban is turned on, and the hardware displays the screen from the #17 bank).
AM> Just don’t say that it’s faster this way. In the vast majority of
AM> cases, it is enough to optimize the code so that there are enough clock cycles for
AM> switching pages via #7ffd.

AM> This bug suffers from the latest demos from Real Masters, 4k SKY and many
AM> demos from Enlight'96-97.

so it is said: the compo machine is Pentagon128 ;)
and Pentagon512 and Scorpov's memory expansion beyond 128 is
ideological pornography ;)

in system programs - yes, of course, only full addressing. in demos - as
more convenient for the encoder.

Sun 20 May 2001, at 09:19:48 Valerij Kozhevnikoff told Dmitry Lomov about GMX.

VK> Full screened rotator - it's cracked, you can write THIS... I wouldn't
VK> be able to. Probably.
It’s not even possible.

Bye, Valerij.

DL> wow, maybe I looked at something wrong? ;)
Judging by the effects described, that’s exactly what you looked at.

DL> or is what I saw called something else? if colored
DL> chaotic squares are spinning - this is called a plasma rotator (or something like that
these are not chaotic squares, but a whirlpool. This is done in an elementary way (algorithm by
Megus): the first 4 lines will scroll cyclically to the left, the next 4 to the right,
the first 4 columns will scroll cyclically up, the next 4 - down.

DL> spirit). if all this is also moving away/approaching - somehow,
DL> for example,
It’s even a shame, so much was done with multicolors (especially with voxels and
mirror-rotator, because there we had to dynamically insert rets to
switch screens) and on you: they say there is only one plasma!

DL> plasma funnel ;) is it squashing me? :-) then what exactly?
And where the sky and mountains are, is it mountain plasma?
And where there is a tunnel, is it tunnel plasma?
And in Rage, where the pig is spinning, is it pig plasma? ;)

Gone are the days when plasma was made in multicolor. Almost any
chunky 2D effect can be easily done in multicolor (if you are not too lazy to mess with the code). In addition, in multicolor it looks more beautiful and faster than chunky
.

Or were you probably confused by the plasma-like textures?
So this is Tigrr using themI painted on Photoshop filters.

Here is a list of effects (for those who still think that TD has a lot of plasma):
1. 2 Bitplane multicolor (nothing special) 50 fps
2. Inertial tunnel 25 fps
3. Pulsing & sucking tunnel 25 fps
4. Voxel Landscape 16 fps
5. Whirlpool (well, not plasma) 50 fps
6. Perverted tunnel 25 fps
7. Mirror rotator 25 fps
8. 4 mirror rotators 50 fps (where is the plasma here?
Innerloop is almost like in a rotator)
9. Phong shaded plushka 4-6 fps
10. Environment mapping 4-6 fps
11. Scroll 50 fps

Perhaps there are too many tunnels.

EA/Antares, after watching Power Up, generally stated at CAFe'2000 that everything there
is built on the basis of two effects ;) Well, at least SerzhSoft and I
convinced him.

Never mind, we're making a demo on CC1. It will be beautiful and bright. As they say:
"Movin fasta, doin betta!"

Bye, Dmitry.

Tue 22 May 2001, at 19:52:34 Dmitry Lomov told Aleksey Malov about GMX.

DL> so it is said: the compo machine is Pentagon128 ;)
And if the compo machine on the PC is Celeron-650, then this means that the deme has the right
not to go to P-III? ;)
The fact is that the Pentagon-128 is not the only Spectrum-128-compatible
computer in our country. And if the lack of synchronization of multicolors can still be tolerated, then refusal to work at all is too much.

DL> and the Pentagon512 and Scorpio memory expansion over 128 is
DL> ideological pornography ;)
I don’t know how much you suffered debugging Rage, but it’s very convenient for me to debug my
demos on Scorpion-1024 (according to the Pentagon standard): ZASM sits in 128 memory,
the debug version is in the last pages of the RAM-DISK. Is everything hanging? No problem
- I press reset, but I have MadROM, press the R button and oops - I'm already in ZASM with live
text. Everything takes 2 seconds. And with 128 kilobytes, I have nowhere to put the STS,
not to mention the successful return to assembler when debugging a program
using over 48 kilobytes.

DL> in system programs - yes, of course, only full addressing. in demes -
DL> as
DL> is more convenient for the encoder.
If the encoder is good, then he can do everything with full addressing. I say this as a good coder. My Tryptomine Dream is full everywhereaddressing
is used, and there are enough clock cycles for everything.

Tue 22 May 2001, at 01:45:33 Vassili Klimov told Aleksey Malov about .

VK> and on mine it matches (albeit with an accuracy not to the pixel, but to 2x - read
VK> one clock cycle, however, this has already been written about...)
Apparently, the fix there is a curve.

Bye, Vassili.

Tue 22 May 2001, at 09:48:27 Vassili Klimov told Valerij Kozhevnikoff about
GMX.

VK> heh, there are only 2 frames - each on its own screen, but with the border
VK> obviously
VK> the same thing...
Nothing like that, one color goes into another there no less than for 8 frames.

Bye, Vassili.

AM> Fate and Hell Raiders did not work because their files turned out to be, for
AM> for unknown reasons, overwritten files of some other demo.
AM> Kag - because of the curvature of the machine.
AM> SKY4k - apparently because of im2
and what were the signs in all the others?

AM> I appeal to all potential demomakers reading this echo:
AM> Switch pages on port #7ffd. Switching over the #fd port with
AM> the 6th bit turned on solves the problem of working on Scorps, and on
AM> Pentagon-512/1024 jams the RAM-Disk and causes the absence of the 7th screen
AM> (they turn on the #57 ban, and the hardware displays the screen from the #17 bank).
AM> Just don’t say that this is faster. In the vast majority of cases
AM> it is enough to optimize the code so much that there are enough clock cycles to switch
AM> pages through #7ffd.
bla bla bla.... :(
read adventurer 12 and you will not have problems with this,
I would say that in some cases it is very necessary:
or PUSH BC:LD BC, #7FFD:OUT (C),A:POP BC
or OUT (#FD),A
make your choice, gentlemen! ;-)

AM> The latest demos from Real Masters, 4k SKY and many demos
AM> from
AM> Enlight'96-97 suffer from this problem.
it's very useful in the intras;)

cYa, J7n/4D.

AM> To all: Do you think it’s worth throwing here an improved version of our multiloader since Tryptomine Dream
AM>? The reading speed on computers in which
AM> interrupts arrive synchronously with disk rotation has increased by a factor of 2. Min.
AM> speed is 2Kb/sec (this is the case when the disk rotation speed is 300
AM> rpm
AM> (i.e.
AM> 5 rev/sec), and the interrupt length is 70,000 cycles +-100 cycles, so interrupt
not +-100, but exactly 71680
As far as I understand, all modern demos (Refresh, Anamnesis) do not use
interrupts at all: they load a portion, play music, etc.
I saw with my own eyes: the refresh was loaded from the screw, the music played several
times faster, and the fps was probably about 50 really ray cut...

cYa, J7n/4D.

DL> let there be no subject newer than the old one. I'm tired of coding :(
and rightly so;) tasm forever...

cYa, J7n/4D.

AM> Don't tell my slippers! In the linoleum 16k intro there is no alignment at 4
damn, and I thought my pro was buggy;)
right now I pulled the pipe and it became ok.

AM> clock, and, moreover, the coincidence of the screen with the curb in the middle. While
AM> on the same machine Rage works in the same way.
AM> In confision and 7up on this machine there was no coincidence between the screen and the border
AM>.
AM> The word “this machine” meant the real Pentagon-1024. The demos
AM> were launched with the top memory turned off.
so things are like this;) my Pentagon has been lying on the top shelf for months without
moving, but when he worked on it everything coincided in rage and in confusion and in
7up. Now I'm using a pro board version 5.02 with 100% Pentagon
synchronization (synchronized to the exact rage clock - just like GMX;),
and the above-mentioned things also work fine

cYa, J7n/4D.

Saturday 19 May 2001 14:35:13, Sergey Zotov -> Grigory Klimenko:

SZ> Because you need a stable bus, and 100%
SZ> of Pentagons cannot boast of such a bus. But a Scorpio, and even with a subj, can. :)

Hmm... And in what zoo did you find a pentagon with an unstable tire? I
for example, have never come across something like this... :-)

With respect, Dmitriy.

VK> and that all the others had signs?
Apparently, there were. My friend's SKY didn't work precisely because of this on
Pentagon-128.

AM>> pages via #7ffd.
VK> bla bla bla.... :(
VK> read adventurer 12 and you won't have problems with this,
VK> I would say that in some cases this is very necessary:
VK> or PUSH BC:LD BC,#7FFD:OUT (C),A:POP BC
VK> or OUT (#FD),A
VK> make your choice, gentlemen! ;-)
In most cases, it is enough:
ld a,page
call lkl

lkl push bc
or 0
screen equ $-1
ld bc #7ffd
pop bc
ret

Only 1 byte longer. You can pervert it to such an extent:
ld a,page
rst 8
Only 3 bytes. True, such a feature may not work on Scorp.

In multicolors it is enough:
exx
out (c),e / d
exx

AM>> interrupt
VK> not +-100, but exactly 71680
Just in the case of 71680 clock cycles everything is ok (the multiloader in TD loads 8.5k/sec),
but with an int length of 69888 clock cycles the speed drops to 1.5k/sec

VK> As far as I understand all modern demos (Refresh, Anamnesis) do not
VK> use
VK> interruptions at all: they load a portion, play music, etc.
To do this, you need to restore the disk, otherwise it will either be a bummer, as in Anamnesis,
or the music will howl terribly as in Napalm/Dogma/Refresh.
In TD, we have done it this way - 8.5 k/sec on the Pentagon. And the music doesn’t howl
at all.

VK> I saw with my own eyes: the refresh was loaded from the screw, the music played
VK> several
VK> times faster, and the fps was probably about 50, although the beam cut...
It’s funny: how it loaded from the screw, because this requires complete interception
of the main reading procedures. Maybe I looked under emulsion?

Wed 23 May 2001, at 02:38:17 Vassili Klimov told Dmitry Lomov about Storm.

VK>> and on mine it matches (albeit with an accuracy not to the pixel, but to 2x - read
VK>> one clock cycle, however, this has already been written about...)

AM> The fact is that the Pentagon-128 is not the only Spectrum-128-compatible
AM> computer in our country. And if the lack of multicolor synchronization is still
AM> can be tolerated,
and this is the holder of the Pentagon speaking;)
have you read Subliminal Extacy? In my opinion, there are 2 topics: f$$$ing beta disk and f$$$ing
xUSSR faster's computers owners.

AM> their demos on Scorpion-1024 (according to the Pentagon standard): ZASM sits in 128
AM> memory, the debug version is in the last pages of the RAM-DISK. Everything hangs?
AM> No problem - I press reset, but I have MadROM, press the R button and oops - I'm already in
AM> ZASM with live text. Everything takes 2 seconds. And with 128 kilobytes, I
AM> have nowhere to put the STS, not to mention the successful return to assembler when
AM> debugging a program that uses over 48 kilobytes.
so you're always welcome, and you also debug other people's demos :)?

AM> If the encoder is good, then he can do everything with full addressing. I say this
AM> as a good coder. In my Tryptomine Dream, full
AM> addressing is used everywhere, and there are enough clock cycles for everything.
although someone CJ considers intros compo for masochistic coders;), but try
write 1k with pages...

cYa, J7n/4D.

AM> Nothing like that, one color goes into another there in no less than 8
AM> frames.
well, there are only 2 pixel frames/delays along the border, the rest is all AGA
plasma;-)

cYa, J7n/4D.

cYa, J7n/4D.

One Tuesday May 22, 2001 at 5:52:43 pm there was a conflict
Sergey Karachentsev and Nikolay Maslennikov were arguing about Storm
I decided to argue too

SK> And also, who is the author of Storm and what is his network address?
The author is subzha Dmitry Lomov, he regularly appears in REAL.SPECCY
Toko no longer wants to bother with coding...tired...:(((

With all my heart Ciao, Sergey. [I ZX] [SHADOW MAD]

One day on Wednesday May 23, 2001 at 01:38:17 there was a conflict
Vassili Klimov and Dmitry Lomov were arguing about Storm
I decided to argue too

VK> heLLo, Dmitry!

GK> It was really buggy! Upon arrival home, I'm on my
Dak and I'm talking about the same thing!

GK> Pentagon looked through
GK> everything without problems.
And how many glitchy Pentagons still exist! And all their owners will swear.

ZXNet: 500:8462/1 ~/ <~ http://www.zx.ru/zs/
Fido: 2:5057/21.93 /_ _> ftp://zx-museum.org.ru/zs/

VK> the music played several
VK> times faster, and the fps was probably about 50, although the beam
VK> cut...
Can you share this version of the demo?

On Wednesday May 23 2001 Vassili Klimov wrote to Aleksey Malov:

AM>> Spectrum-128-compatible computer in our country. And if
AM>> the lack of synchronization of multicolors can still be tolerated,
VK> and this is said by the owner of the Pentagon;)
Sir, you will laugh, but Vivid has Scorpio :)
And he wrote all his monstrous multicolor effects in Tryptomine Dream in
Scorpio. And then I made the final fix for the Pentagon from Tiger (Tigrr/BW) :)
This is how real coding geniuses get perverted :)

VK> have you read Subliminal Extacy? In my opinion, there are 2 topics: f$$$ing beta disk
VK> and f$$$ing xUSSR faster's computers owners.
And they are right to be indignant. I support SE on this.

AM>> 2 seconds. And with 128 kilobytes, I have nowhere to put the STS, not to mention
AM>> not to mention the successful return to assembler when debugging a program
AM>> using over 48 kilobytes.
VK> so you’re always welcome, and you also debug other people’s demos :)?
Do you know that you can learn from other people’s mistakes?

Bye, Vassili!
Megus/Brainwave^X-Project. http://brainwave.dax.ru

DL> colored chaotic squares - this is called a plasma rotator (or
Colored chaotic squares - these are the ones in Greetings? When I came up with this effect, and it was either 1995 or 1996, I didn’t know such a word in relation to
demo effects :) This is not plasma rotator, and a bunch of scrollers. Plasma, in
my understanding, is not made at all like that.

DL> something else, for example, a plasma funnel ;) is it crushing me? :-)
Listen, if the effect uses a terrible plasma-like texture, does the effect
also turn into plasma? Well, imagine that on the walls of the tunnels there is not a plasma
texture, but the face of Bill Gates/Vladimir Putin/adjust to your taste :) Will there be plasma too?

Bye, Dmitry!
Megus/Brainwave^X-Project. http://brainwave.dax.ru

On Wednesday May 23 2001 Aleksey Malov wrote to Dmitry Lomov:

AM> I was offended.
Yes, I predicted this :)

AM> these are not chaotic squares, but whirlpool. It’s done in an elementary way
AM> (algorithm by Megus): the first 4 lines scroll cyclically to the left,
Now I remember the first time I implemented this effect... Nostalgia, damn it! I was still writing at GENS back then. Well, this effect, naturally, was not in multicolor; the raster image of 8x8 familiar places was scrolled. By the way, this effect is a good
example of how you can achieve interesting
results using very simple means. And I wrote it when I was in either the 7th or 8th grade :)

AM> And in Rage, where the pig is spinning, is there pig plasma? ;)
Pig plasma :)))

AM> not too lazy to pervert the code). In addition, in multicolor it is more beautiful and
AM> it looks faster than chunky.
But just in terms of writing time, it doesn’t look any faster. What can you do? :)

AM> Never mind, we’re making a demo on CC1. It will be beautiful and bright. As
AM> they say: “Movin fasta, doin betta!”
Just say one more word and Tigger and I will kill you :) Although no, we won’t kill you, who
will write the demo then? Okay, live :)

Bye, Aleksey!
Megus/Brainwave^X-Project. http://brainwave.dax.ru

NM>>> him
NM>>> I’m interested, plz!
SZ>> I have 1.3, but it’s been around for a long time.

DL> damn, if only we could write a new one. editor 2.0 is almost
DL> there;) for the
DL> compiler there is an ideology and pieces of code.
It doesn’t seem to work with the screw, and it takes up 2 banks. I generally use XAS,
although it doesn’t work with a screw, it only takes up 1 jar of memory. Yes, and frame
imho is useless in ASMA... And for some reason it has 2 modes: ext and break. But despite all this, if you share the source code, I will be very grateful. :)
And it’s worth adding asm. He has a lot of fans... And in conferences they regularly ask about him. Maybe you can add more? Huh?

VK>>> and on mine it matches (albeit with an accuracy not to the pixel, but to 2x - read
VK>>> one clock cycle, however, they already wrote about this...)
AM>> Apparently, the fix is ​​crooked.

AM>> still
AM>> can be tolerated,
VK> and this is the owner of the Pentagon speaking;)
Nothing like that. I have Scorpio-1024. From the pentagon there is only switching
cans via #7ffd, and from the scorpion - frame scanning. Instead of a shadow
service monitor, I have a MadROM, trdos with support for 896k ramdisk.
I wrote all the multicolors in Tryptomine Dream on my scorpion, and then threw them to Tigrr's
modem and then told him over the phone where to put what delay.

VK> have you read Subliminal Extacy? In my opinion there are 2 topics: f$$$ing beta disk and
VK> f$$$ing
VK> xUSSR faster's computers owners.
I read it. In this regard, everything is fine with us. TAP versions of Tryptomine Dream and
Technogen exist. The truth is in TD i=#5e - well, you can’t put it anywhere else.

Wed 23 May 2001, at 16:18:01 Roman Petrov told Aleksey Malov about .

RP> effect - a good example of how you can achieve
RP> interesting results using very simple means. And I wrote it back when I was studying either in 7th or 8th grade :)
And then I was very good at typing programs from the book “How to write a game in BASIC” and inserting “Cracked by Malov Alex (c) 1995” into the loaders for them in Bill
Gilbert’s frame ;)

RP> But that’s just not the time of writing it comes out faster. What
RP> can you do? :)
Sometimes it was managed overnight. Sometimes it takes more than a month. And so - 2-3 days.

NM> ^^^^ slow shit

[+] - ннрршеенисмммо, севсме назад...

[I ZX] [Team Guardian] [Team Kelly]

Up to the view.
Tim Kelly

... tim_...@mail.ru 500:622/1@zxnet http://kgs.dax.ru

Once Sunday May 20, 2001 at 23:38:20, Dmitry and Aleksey were arguing on the topic

AM>> for 4 measures, as usual. At 7 up they decided to use the command ld a,i
AM>> for this purpose, but without success.
DL> and what were their prerequisites for the opinion that this command
DL> aligns something?

on their pentagons, one of the actions of this command was to set one of the
flag bits in the case of interrupt alignment... they used this feature
in their Deme...

because of this, on my scorpion it stops at multicolor and until you can correct it with magic
prog... it will go further...

[Times of Lore] For this I say goodbye, Yuri. /I \
[No RAR allowed] [SoHm] [Conceptual Minimalism] [AF] \ZX/

AM> Tell this to someone who doesn’t like your next intra. It's not worth saving
AM> on
no one has complained
AM> on matches yet. If you really can't bear it, make a version with full switching
not possible, apparently for those who don't have
AM> pages working, even if it will affect the synchronization of multicolors.
we have some kind of flame ;)
you're trying to convince me that full addressing #7ffd is needed, however
there are already a (very) considerable number of programs that use a short addressing
(by the way, it seems to me that the #40 mask was the first to be made by the bourgeoisie), moreover
programs (not mine) will soon appear that will use #fd again. I propose to close
the topic.

I also agree with LD regarding the use of computer resources: on the compo - 100%
of the capabilities of the compo machine, and everything else is system/games/... it is advisable to
configure/comply with generally accepted standards.

Regarding the CC's computer, there is a version that there was no M1 on the AY circuit,
since resistors are located on all boards 128, and circuits with AY are usually soldered
at random.

cYa, J7n/4D.

Somehow Wednesday May 23, 2001 at 01:23:21, Vassili and Aleksey were arguing about GMX

VK> bla bla bla.... :(
VK> read adventurer 12 and you won’t have problems with it,
VK> I would say that in some cases it is very necessary:
VK> or PUSH BC:LD BC,#7FFD:OUT (C),A:POP BC
VK> or OUT (#FD),A
VK> make your choice, gentlemen! ;-)

and if bit 6 is not installed, then the only joy is on Pro or Scorp...

but here the code must be written so that it is #7FFD that is in BC, and not #hzch... this is where the skill of the coder comes into play...

[Times of Lore] For this I say goodbye, Yuri. /I \
[No RAR allowed] [SoHm] [Conceptual Minimalism] [AF] \ZX/

AM> It's funny: how it loaded from the screw, because this requires complete interception
AM> of the main procedures for reading. Maybe I was looking under the emulsion?
I have my own screw controller.... I was
emuli sax. If the frame rate is >50hz, then the beam cuts, and at 50hz it flickers terribly.

cYa, J7n/4D.

AM> ZASM rules. True, it is somewhat slow.
I myself once wrote in version 3.0..., the new one is okay, but the macros are rather weak

cYa, J7n/4D.

NM> ^^^^ slow shit
NM> From the bottom of my heart Ciao, Vassili. [I ZX] [SHADOW MAD]
cool the last two lines are combined...;(

cYa, J7n/4D.

RP> Sir, you will laugh, but Vivid has Scorpio :)
Siril also has a pentagon... whiteless and multicolors are almost the same
RP> And all his monstrous multicolor effects in Tryptomine Dream he
RP> wrote
RP> in Scorpio. And then I made the final fix for the Pentagon from Tigger
RP> (Tigrr/BW) :)
RP> This is how real coding geniuses pervert :)
I also pervert: I write source code in DN on PC, and then compile it in real life.

VK>> the music played several
VK>> times faster, and the fps was probably about 50 though the beam
VK>> cut...

VK> is not possible, apparently for those who do not work
It is not their fault that they do not have Pentagon-128.

VK> you are trying to convince me that full addressing #7ffd is needed, however
VK> there are already a (very) considerable number of programs that use short
VK> addressing (by the way, it seems to me that mask #40 was the first make bourgeois),

Well guys, let's learn to write COMPETENTLY. Time and technical progress
(albeit very slow) dictate their own rules for writing programs. It used to be
it was fashionable to insert turbo loaders that loaded only from drive A:.
They began to install several disk drives on the Spec and turbo loaders became smarter.
Hard drive controllers appeared and work with RamDisk in TRDOS and system software
began to work through #3d13 (especially since in the latest versions of TRDOS and so
the error when switching to the next track was corrected and the speed became
almost the same as that of turbo loaders) and full addressing of the ports for switching
pages. But in demes, everyone does it in their own way, even when there is no
REAL need for it. For example, it would be convenient for me to see the same dogma with
RAMDisk without any howls of Mouzon (in TD, a multiloader with RamDisk through
#3d13 reads 2 sectors per interrupt (you can actually increase it to 8
sectors)). But no. Bummer.

VK> Moreover, programs (not mine) will appear soon, which will againthey will use #fd.
And what REAL benefit will this give to this program, besides the moral
satisfaction of the author of the program, who realizes that in order to launch his brainchild you need to
either go into the computer with a soldering iron and solder the bells and whistles switch, or break it
and correct the author’s “mistakes”.

VK> I suggest closing the topic.
Yeah, it’s time. But I want to say the last word: There must be VERY compelling
reasons that the program refuses to work on a computer with more
memory than the author of this program.

VK> I also agree with LD regarding the use of computer resources: on compo -
VK> 100% of the capabilities of the compo machine, and everything else system/games/... is desirable
VK> configure/comply with generally accepted standards.
The #fd port, in my opinion, should be used as a last resort, when ALL
other optimization methods have already been tried and do not give the required speed.
In most cases, you can get by with full addressing.
Especially in chunk effects.

VK> Regarding the central computer, there is a version that there was no M1 on the AY,
VK> circuit since resistors are wired on all 128 boards, and circuits with AY are usually
VK> soldered haphazardly.
What is this signal responsible for, otherwise I’m not very knowledgeable about hardware?

VK> Siril also has a pentagon... whiteless and multicolors are almost the same
^^^^^^^^ Did you mean to say Scorpio?
Or I didn’t understand you. I don’t see a logical connection between my and your phrases.

RP>> he wrote in Scorpio. And then I made the final fix under
RP>> Pentagon at Tigers (Tigrr/BW) :) This is how the real ones are perverted
VK> I also pervert: I write the source code in DN on PC, and then compile it on
But this is completely perverted. It’s easier to use the Mednonogovsky assembler and test under the emulsion. During final debugging, everything is done in real life
test.

VK>>> disk and f$$$ing xUSSR faster's computers owners.

I won’t argue.

DL>> spirit). if all this is also moving away/approaching - somehow,
DL>> for example,
AM> It’s even a shame, so much was done with multicolors (especially with
AM> voxels and mirror-rotator, because there we had to dynamically insert rets
AM> to switch screens) and on you: they say there is only one plasma!

By the term “plasma” I mean something, perhaps, beautiful, but not having
a clearly defined form;)
if for someone this term sounds offensive - we did not understand each other: (

DL>> plasma funnel;) is it crushing me? :-) then what exactly?
AM> And where the sky and mountains are, is it mountain plasma?

no, there are mountains ;)

VK> I myself once wrote in version 3.0..., the new one is okay, but the macros are rather weak
That's true. But that's enough for me for now. It’s much shorter to write:

stw #fc21

than
ld (hl),#21:inc hl:ld (hl)#fc:inc hl

It’s a pity that there probably won’t be a newer version. Never.

VK> and who is stopping you from restoring?
Anamnesis, Refresh, 7up and Kag are not restoring on my 3.5~ drive. At 5.25~ -
restoration. But this, again, is hemorrhoids, and besides, there is a possibility that they will not be restored at the party either.

VK> well, you screwed it up;)) I correct other people’s mistakes, and I learn by debugging good
VK> programs.
I like to look into the code from Accept Corp and laugh at
TOTALLY NOT_OPTIMIZED code. How can you write something like this? Or did they compile BASIC effects at Tobos?

VK> PS: try to trace SE
What is it?

RP> Listen, if the effect uses a terrible plasma-like texture -
RP> does the effect also turn into plasma? Well, imagine that on the walls
RP> of the tunnels there is not a plasma texture, but the face of Bill Gates/Vladimir
RP> Putin/put it to your taste :) Will there be plasma too?

oh well ;) no, it won’t be plasma, it will be a textured funnel.
And you, then, have a funnel textured with plasma :-)

SZ> It doesn’t seem to work with the screw,

are there any problems to teach?

SZ> and it takes up 2 banks.

Yes? I've already forgotten ;)

SZ> Yes, and IMHO there is no need for framing in ASMA...

You, IMHO, are mistaken. however, there was a lot of arguing about this, I got tired of it...

SZ> And for some reason it has 2 modes:
SZ> for ext and break.

there weren’t enough buttons.

SZ> But despite all this, if you share
SZ> the source code, I will be very grateful. :)

I’ll wash it one of these days.

SZ> And it’s worth adding asm. He has a lot of fans... And in conferences about
SZ> people are regularly interested in him. Maybe you can add more? Eh?

no. all the summer before last I gave birth to ideology and wrote. and all last summer I tried to persuade myself to take up and do something... it didn’t work out. Why do people write
software on the Spectrum? to prove to myself that “I can.” to prove to others that “I can.” to prove to others that “I’m cooler than most.” to learn
something new. They write just for fun. they write out of despair, because
there is nothing else to write on. they write for themselves so that they have something to work on. Did I
forget something?

Unfortunately, I went through all this. interest remained, but it was sluggish.

One day, Tue May 22 2001 18:52, Sergey Karachentsev wrote to Nikolay
Maslennikov:

SK> but “YTSUKEH” (real SPECTRUM) is more convenient for me. I just drew
SK> Russian letters on the keys. I tried to get into the code. There I saw two
SK> tables of two encodings. The only thing that is not clear is how to switch the encoding.
SK> Who knows?

Knew it when I wrote it.

SK> And also, who is the author of Storm

me ;)

damn, the source code will probably still fly here. edit yourself ;)
I’ll just ask the moderator...

One day, Wed May 23, 2001 02:38, Vassili Klimov wrote to Dmitry Lomov:

NM>>> What is the latest version of the subject? And if possible, throw it at me
NM>>> in the soap, please!

AM> Spectrum-128-compatible computer in our country. And if the lack of
AM> multicolor synchronization can still be tolerated, then refusal to work
AM> at all is too much.

possible. what if a couple of bars are missing? ;)))))))))

DL>> and the Pentagon512 and Scorpov memory expansion over 128 is
DL>> ideological pornography ;)
AM> I don’t know how much you suffered debugging Rage,

there’s nothing to debug there. I practically didn’t run the debugger.
it was all just built there, there were procedures that translated geometry into clock cycles
and wrote programs. and I wrote this using a clever process, so that I didn’t have to align it by 4 during the writing process. so there was not so much fuss.

and when other software was written, the debugger didn’t hurt to launch. at least
I didn’t sit in it - that’s for sure.

AM> and it’s very convenient for me to debug my demos on the Scorpion-1024 (according to
AM> the Pentagon standard): ZASM sits in 128 memory, the debug
AM> version is in the last pages of the RAM-DISK. Is everything hanging? No problem -
AM> I press reset, and I have MadROM, I press
AM> the R button and oops - I’m already in ZASM with live text.

I have the same thing, only on Pentagon 128 ;)
I press reset - there is text on the screen ;)

AM> Everything takes 2 seconds.

I have the same amount.

AM> And with 128 kilobytes for me There is nowhere to put the STS, let aloneabout
AM> successful return to assembler when debugging a program using
AM> over 48 kilobytes.

well, did you somehow get out? suffered, but not much.

I'm not saying that upper memory is not needed. if it's transparent and doesn't cause anything to glitch - go ahead. but every time you need to watch something, pull the toggle switch - thank you. although, I must admit, I also had to pull when I had to copy a lot. 256 in this case is a necessity.

DL>> in system programs - yes, of course, only full addressing. in
DL>> demos - as is more convenient for the encoder.

One day, Tue May 22 2001 09:48, Vassili Klimov wrote to Valerij Kozhevnikoff:

FO>>>> Rage is definitely going fine,
DL>>> of course, GMX has been using it for a while in the development process ;)
VK>> Maniac ;-)
VK>> Full screen rotator is but if you crack, you can write SUCH... I
VK>> would not have been able to. Probably.
VK> heh, there are only 2 frames - each on its own screen,

four, if I'm not mistaken. two on each screen ;) one at a time would be
too fast.

Thu 24 May 2001, at 01:20:47 Dmitry Lomov told Aleksey Malov about .

AM>> I was offended.
DL> come on, stop it ;)
I'm kidding ;)

DL> by the term "plasma" I mean something that is perhaps beautiful, but not
DL> bright expressed form ;)
DL> if for someone this term sounds offensive - we did not understand each other :(
By plasma I mean effects based on:
1. Palette shifts.
2. Effects like “fire”, i.e. those in which the color of a point depends on the color
of neighboring ones on the previous frame.
3. Sine plasm (as in Eye Ache - 2)

Thu 24 May 2001, at 02:26:25 Dmitry Lomov told Aleksey Malov about GMX.

DL> pussies are a completely different story ;) there has historically been respect for
DL> standards and compatibility. and on the Spectrum they are full of crazy people.
That’s what I’m talking about. It's time to learn to write COMPETENTLY.

DL> perhaps. what if a couple of bars are missing? ;)))))))))
If other optimization methods do not help, then this is an extreme case.

DL> there is nothing to debug there. I practically didn’t run the debugger.
DL> it was all just built there, there were procedures that did geometry in clock cycles
Well, I had such and such procedures.

DL> they translated and wrote programs. and I wrote this using a clever process, so that
DL> would not need to be aligned by 4 during the writing process. so there wasn't that much fuss
DL>.
In the sense of "on a tricky process" what is it like?

DL> and when other software was written, it didn't hurt to launch the debugger. at least
DL> didn’t sit in it - that’s for sure.

DL> I have the exact same thing, only on the Pentagon128 ;)
DL> I press reset - there’s text on the screen ;)
Did I change the ROM or something?

DL> Well, did you somehow get out? they suffered, but not much.
This is exactly the same pornography as debugging 48 programs on a 48
machine.

DL> you just approach this very seriously and responsibly. maybe it
DL> and
DL> are correct.
Icould write like “all the damn demomakers”, i.e. I don’t give a damn about
non-Pentagogues, as Codebusters did, but I don’t do this out of respect
for them, and also because I myself am a Scorpio.

Thu May 24, 2001, at 01:51:20 Dmitry Lomov told Roman Petrov about .

DL> come on ;) no, it won’t be plasma, it will be a textured
DL> funnel.
DL> and you have a funnel textured with plasma :-)
On PC and Amiga it’s called Mirror rotator. What are the reasons to move away from
this terminology on the Spectrum? One person even saw “fractal plasma” in this effect.

VK>> will use #fd.
AM> And what REAL benefit will this give to this program, other than the moral
AM> satisfaction of the program author, who realizes that in order to launch his brainchild
it is still a benefit to the program or its author?;)
AM> you must either go into the computer with a soldering iron and solder the bloat switch, or
AM> break it and fix it "mistakes" of the author.
first place on the demo compo, for example

AM> Yes, it's time. But I want to say the last word: There must be VERY compelling
AM> reasons that the program refuses to work on a computer with more
AM> memory than the author of this program.
first a ban on #fd, then on multicolors, then on multiloaders...
you can go so far

AM> Port #fd, in my opinion, should be used as a last resort, when ALL
AM> other optimization methods have already been tried and do not give the required speed.
note, I just didn’t think that this was such a serious problem...

FO>>>>> Rage is definitely going fine,
DL>>>> of course, GMX was on it for a while during the development process ;)
VK>>> Maniac ;-)
VK>>> Full screened rotator - it can crack SUCH

RP> Do you want to say that they are all saying it wrong? That's right. Everything they say is very correct
RP>. It may be rude, but it’s correct. Raver doesn’t know any other way - he
RP> always expresses everything very emotionally.
from swearing to assault is one step, and as for correctness, I don’t want to start
demagogy, so I’ll refer you to the recently released Inferno #2, everything is very simply explained there
why we make programs in Russian and for beta disk.

VK>>>> and on mine it matches (albeit with an accuracy not to the pixel, but
VK>>>> up to 2x - read
VK>>>> one clock cycle, however, they already wrote about this...)
AM>>> Apparently, the fix is crooked.
VK>> how can you fix a pixel = half a clock cycle?;)))
AM> Only hardware. You write “..accurate not to the pixel, but to 2x”
so the fix has nothing to do with it

cYa, J7n/4D.

AM> They are swearing not at us, but at those who croon about the bourgeoisie. We don't give a damn.
AM> We provide TAP versions for them. Let them do the rest themselves.
I'm happy for you, but this is a drop in the ocean... by the way, the conclusion is the same: this correspondence
still will not help the matter.

AM> I like to look into the code from Accept Corp and laugh at
AM> COMPLETELY NOT_OPTIMIZED code. How can you write something like this? Or were they at Tobos
AM> compiling BASIC effects?
and you want them to do nothing at all? and in the end we must respect
the work of others, there were so many lies about the games of the 8th department, etc., and the scolders themselves
did anything?

cYa, J7n/4D.

DL> pussies are a completely different story ;) there has historically been respect for
DL> standards and compatibility. and on the Spectrum they are full of crazy people.
they have no less problems: all sorts of blasters, geese, drivers, video cards.....
on the same CC, everything worked no better than the spec...

AM>> work

NM>>>> in soap, plz!
DL>>> let there be no subject newer than the old one. I'm tired of coding :(
VK>> and rightly so;) forever...

RP>> coding geniuses :)
VK> I’m also perverted: I write the sources in DN on PC, and then compile in
VK> real.

yeah, there’s the last step left ;) you can compile in some M80.
and then run it on an emulator :-)