Help me get EMUzWIN 0.27
ZXNet echo conference «zxnet.pc»
Hello, All
I want to use his asm!
But I can’t start it!
And having started it, it is not known whether it is correct, the asm hangs itself brazenly and completely!
What am I doing wrong? Are there any nuances? Or who will share the tuned under
Pentagon512, to sit down and write, and not bang...
Hello, Witch Doctor
I answer (ask) again: what version is it? The site has 2.7 build 2.5. Build watch
in the html file what is in the archive or ? in the main menu (although there is already a message that I am with
this? I was too clever in the last build, so html is better).
And immediately: what kind of machine, what version of Windows. I was able to check on XP SP2, 2K SP4,
HT4 SP5 - from what is at hand. I checked it at 98 on vmware, it seems to be the same
works. Maybe it creates some files in the directory. Somehow he swears.
I was in a bit of a hurry: I haven’t tested build 2.5 on NT4 yet. Today I checked:
crashes (actually it doesn’t crash, it just doesn’t start, but it still doesn’t work).
There may still be problems with live 9x, and it probably won’t work with Athlone 64 either.
I will be solving new incompatibility problems in the near future. But XP and 2K with
A regular PS-compatible PNE-4 should go. Today I'll collect two rubbish from gone
get it, you need to look at the living stump100 and at stump2 too. (Here
improved, the code began to depend on the processor).
Hello, Vladimir Kladov
Now build 2.5a is being applauded. At least it should work under NT (but check
I can only repeat). But under 9x it was not possible to check - there is nothing to transfer to
old stump-100 with 98 on board (there is a floppy there, but I don’t, I have a network,
usb, but that doesn’t happen there. What can I say, there is no mouse there - because only
Komovsky input, and I have all 5 pieces - ps/2 or usb... I'll look for junk using
friends to check).
Hello, Vladimir Kladov
Vla> Build 2.5a is now applauding
menu? it worked :v2_clapp:
Hello, [bETA]mEN
I want to clarify right away: I decided to ask around here because... I'm sure it will give a lot
greater results, including in terms of time spent on sorting out what and to
why... And I have to complain, there is very little time :(
So don’t call me lazy/idiot - just give me a hint! :)))
So:
1. launched the emulsion. Everything is OK. I want to change the ROMs to my own - I go to the settings, indicate
paths to ROMs. All rums are tested, 16384 and used in UHreal. I'm closing
settings window - nothing changes... I go to the ROM settings - it’s already there
empty. Those. not a single path is remembered when closing the emulator under any circumstances
circumstances...
In the end, I realized that for a 512k Pentagon you need a 32kb ROM in which first
basic128, then 48. I ended up replacing the ROMs in the emulator directory
(those that he unpacks with 7zip). Those. slipped his own under the same names
ROMs - it finally worked... So the question is: it should be made normal
by ?
2. Keyboard layout: I would like it to be like in UHreal. I think it's there somewhere
I just didn't get there...
3. Why was this done? For the miracle assembler! With backlight and bookmarks!
I type:
org 26000
ld h
I click compile and the emulsion hangs. Absolutely. Those. after testing I realized that when
the correct, so to speak, commands do not hang themselves. Otherwise it hangs itself... I haven’t tried itneither label nor DEFI - there is no point if it hangs almost in the first line.
All of the above is true for 3 machines: home, home junior and work.
The configurations are different - i.e. iron disappears 100%. Windows? Well, on one machine
XPsp2 ENG, on the second XPsp1 RU, and at work finally w98se...
This is such a rake :(
Hello, Witch Doctor
hey people! The healer doesn’t see my messages at point-blank range. Either he has a filter and he
he banned me... I tell him: update the version, please, someone give it to him,
huh?
I can't hang myself. It simply says in the lower window that Syntax unknown ld h.
The ld command (at least in my ASMA) requires both operands.
Hello, Vladimir Kladov
BLAME!!! :) DOWNLOADING the updated version. I'll report the results immediately :)
Thank you!!!
Hello, Witch Doctor
> I want to clarify right away: I decided to ask around here because... I'm sure it is
> will give much greater results, including in terms of time
> spent on sorting out what and why... And time, should
> complain, very little :(
>
> Therefore, don’t call me lazy/idiot - just tell me! :)))
>
> So:
> 1. launched the emulsion. Everything is OK. I want to change ROMs to my own - I go to
> settings, specifying the paths to the ROMs. All rums have been tested, according to 16384 and
> are used in UHreal. I close the settings window - nothing happens
> is changing... I go to the ROM settings - it’s already empty. Those. none
> the path is not remembered when closing the emulator under any circumstances
> circumstances...
>
> In the end, I realized that for a 512k Pentagon you need a 32kb ROM in which
> first bass 128, then 48. I ended up replacing the ROMs in
> emulator directories (those that it unpacks with 7zip). Ie
> I slipped my ROMs under the same names - in the end it worked... Here> and the question: should it be done the normal way?
>
>
> 2. keyboard layout: I would like it to be like in UHreal. I think it's somewhere
> there is, I just didn’t get there...
>
>
> 3. Why was this done? For the miracle assembler! With backlight and
> bookmarks!
>
> I type:
>
> org 26000
> ld h
>
> I click compile and the emulsion hangs. Absolutely. Those. after testing I realized
> that with the correct, so to speak, commands it doesn’t hang up. Otherwise
> hangs... I haven't tried either the label or DEFI - there's no point if it hangs
> almost in the first line.
>
>
> All of the above is true for 3 machines: home, home junior and
> working. The configurations are different - i.e. iron disappears 100%. Windows? Well
> so on one machine XPsp2 ENG, on the second XPsp1 RU, and finally at work
> w98se...
> > This is such a rake :(
AWW!!!!
Hello, Witch Doctor
I didn't understand. More than 3 hours have already passed. What speed are you running? Ibid 1
a meter in total! I download this from a 53.2Kbps dial-up in 3-5 minutes.
Hello, Vladimir Kladov
We, Volodya, are pumping at work... And therefore, not everything is so simple, not everything is so fast
:(
But I pulled it off and launched it (w98) - error :(. I’ll change the hardware now. In the evening
I'll try it at home. God grant that everything will be OK.
Hello, Witch Doctor
So... Well, at work it doesn’t work under w98. It immediately knocks me out... I stay at HOME under XP...
Hello, Vladimir Kladov
It doesn’t work - immediately such an original and informative AKHTUNG:
>>>
EMUZWIN crashed while accessing memory page
in module at address 0000:00000000.
Registers:
EAX=00000000 CS=0000 EIP=00000000 EFLGS=00000000
EBX=00000000 SS=0000 ESP=00000000 EBP=00000000
ECX=00000000 DS=0000 ESI=00000000 FS=0000
EDX=00000000 ES=0000 EDI=00000000 GS=0000
Bytes at CS:EIP address:
Stack contents:
<<<
Moreover, Unreal also crashes on 2Gforce, etc. - this is decided by selecting a video card
- as a rule, any radeon. Test machine - the hardware is always different in some places...
So here such numbers did not work - stable akhtung
Hello, Witch Doctor
That is Build 2.5 still doesn't work for 9x. Well, let's think about how to overcome it.
The most important thing is where to get something to test/debug on (testing is okay,
you tested it, it would be necessary to be able to debug it).
Hello, Witch Doctor
I confirm that Bild2.5 works under Win98. Falls out with exactly the same
message, but as a task (visible by Ctrl+Alt+Del) it continues to hang.
Hello, Witch Doctor
is this still on 9x? Don’t worry, if it doesn’t plow it means it doesn’t plow. We need to figure it out.
Let it be on XP and 2K for now. And I still need to figure out how to get 98. Only
what under vmvarey if.
Hello, Vladimir Kladov
That's it, applaud 2.7 bld 2.5b. Now it should work on 9x and athlone 64.
There was applause. Try it.
Hello, Vladimir Kladov
So, you probably have a day off. Well, I already put 2.7 bld 2.5c (something on
It was falling out, I corrected it a little).
Hello, Vladimir Kladov
It's starting! Well, I’ll drive a little more - I’ll report back
Hello, Witch Doctor
Well, that's good. (And why, at 98, will you torture him? And don’t feel sorry for him... :) )
Hello, Vladimir Kladov
We'll torture him :) If it's 98 at work...
Now a question about the work of ASMA, because... everything for the sake of his beauty:
1. FILE "c:1.scr" - doesn't work - what's wrong? Should be downloaded according to the description
binary... Same with include
2. How does a macro differ from a procedure - can I do it on my stupid fingers?
3. There are still problems with departure from ASMA. :( But already under certain conditions.
4. where is the error during compilation - it doesn’t show? (bug or system function?). That is I
So I realized that setPC just doesn’t light up and that’s it. And where is the problem - he doesn’t say
Hello, Vladimir Kladov
I’m ready to wait and I need to write (or rather, complete) my long-term construction projects in it. For they
somewhat cumbersome, but if there is such a thing, then what the hell.
And a bunch of files can now be stored in a directory as a person...
Listen, is there an analogue of Alasmov’s DISPLAY?
Hello, Witch Doctor
There is no DISPLAY (what does it do? Maybe it’s because I have ERROR ‘text’?).
Hello, Vladimir Kladov
I quote from the description of Alasma:
1. include "filename.ext",2048
Including a code block from disk of size "length" into a compiled program -
really needed
2. DS number of repetitions[,list of bytes]
A list of bytes (zero byte by default) is copied to the current
compilation address. Number of cycles = "number of repetitions".
- something similar, I think, already exists?
3. When loading text from disk, it is scanned for the presence of "+"
at the beginning of each line. During the compilation process, all encountered
"+" will be replaced with "-", which will cause the compiler to skip
lines during the next assembly.
- not very important
4.LOCAL
ENDL
Declaring local labels. All marks within this
operator bracket will be invisible outside it (except for marks
starting with the @ symbol are global tags). Allowed
nested structures.
- is this an analogue of PROC?
5. DISPLAY [/L,]["any text",][[/D|/H|/A,][/T,]expression]
Displays text or numbers during assembly. Keys
/D, /H and /A set the output format for the next one
numbers. Next, the numbers are displayed in the current number system. /D -
decimal number, /H - hexadecimal, /A - in both
systems.
/L indicates that printing should not be on a new line.
/T prints the expression itself first, then its value.
Examples:
DISPLAY "code size: ",/D,end-startDISPLAY /A,"abc+xyz=",abc+xyz
DISPLAY "simple","text"
6. Expressions
{a} - a word located in memory at address a
$$ - current code page
Now the problems that AlCo spoke about:
In all operations except + and -, the use of
labels that will be defined below (due to single pass)
- it happens, by the way. Isn't there something like this in emuZasme?
Hello, Vladimir Kladov
1. I already have FILE "name"[,start[,len]]. I left INCLUDE to include a piece
sorts as before. (Hmm, I recommend putting a checkbox immediately after INCLUDE.)
2. There is DUP n ... EDUP, FOR i=val_list ... EFOR
3. about + - I didn’t understand. It's not up to you to change the source. According to the rules, he must
work in read-only mode. It's iron.
4. PROC ... EPROC - All the same and for the same. Optionally prefixed with local
the tags may be @, _, !, , I don’t remember something else - in the ASMA options. Local
you can duplicate as much as you like and set the direction for the transition.
5. Yes, this is approximately the same as ERROR, but I did not provide for the output of the number.
Well, you can do DISPLAY, but the format would be simpler, like
DISPLAY 'txt'[,#expr[,#passes]] (mask on which passes to output, as in
ERROR).
I have a two-pass one, and if C-- -like is used, then a third one is also done
pass, I don’t see any problems with this. My speed turned out to be good, 48K from
disassembled code compiles in less than a second. Labels should be higher
are defined only in special cases: for repeaters in DEFB (#expr)n..., for
ORG. Looks like it's for something else. Macros must be defined above, DEFINE's. All
the rest (sort of) can link down.
In general, everything is written in the rules and must comply.
Hello, Witch Doctor
I put the version down, you can take it. Now, by the way, I’ll read and respond to the post. (A
It wasn’t possible before, or at least with soap? Now after the fight it will be necessary
watch).
Hello, Vladimir Kladov
Now I'll put the first update. And, oh, as always, the first pancake is lumpy, at least
something is wrong. I didn’t test the EQU, while I was fixing one thing, I missed another.
Nowadays it’s more formal than that. Link to update ZXAsmPP.dll here
[http://bonanzas.rinet.ru/zx/ZXAsmPP.zip], and then it will be here (there only
assembled dll, about 40K in the archive, if you really need the source code, I’ll do it too
post it there).
I tested on real code, and at the same time wrote the most unrealistic (in
sense using all the special things). So this version of asma should
to be more precise.
Although I was happy in vain with the speed. I put a cycle FOR i=0 TO 255 and in it
nested FOR i=j TO 8 and there by IF #expr THEN macro1 ELSE macro2 EIF, and how...
he got lost in thought... In the end, after half a minute everything is OK, it compiles, but that’s what made me
not satisfied. We need to think something else about how to speed it up. Maybe (if
brakes in this) you will have to prohibit nesting INCLUDE inside FOR/DUP.
Hello, Vladimir Kladov
Oh yes. I forgot to remove the “prohibition” on nested inclusions (although the brakes were not the same)
in this, but calling macros). However, this is (certainly) not the last update.
In general, it kills how sluggishly a bunch of them work in all these nuclear weapons and how sluggishly they work
In particular, huge strings work. However, I have now completely gotten rid of everyone
huge strings, added hashing, special high-speed storage objects
name tables, your own pseudo-memory manager (why bother freeing it up?
in pieces, then in this heap there are only brakes - it’s easier to single them out to the point of numbness
while the compilation is in progress and do not release anything, and at the end everything at once
release, everything will be faster).
Write letters. I went to read a fairy tale to my son, it’s time for him to go to bed.
Hello, Vladimir Kladov
Speed won (the new one is lying). It seems there were also problems with stability.
You may have to rebuild the emulator itself and transfer it. If that version is what
He’s lying at the exit now, swearing (is he swearing or not swearing?).
If you write letters, write something specific and shorter in the subj field. Like
EmuZWin - bla-bla-bla. Otherwise I don’t know now whether I received the letter and killed him or
it just hasn't arrived yet.
Hello, Vladimir Kladov
nested includes are nonsense. I completely forgot that for DISPLAY I changed the number
parameters in callback - functions for transmitting a message (or not changed? exactly
I don’t remember anymore), so I’m already knocking out build 2.6a of version 2.7 of the emulator in
full archive, if anything happens, take the meter again. At the same time coloring for some
directives in the source code have been corrected, and DISPLAY with END have been added to the painting.
It's ready, it's there.
Hello, Vladimir Kladov
Half a minute is not enough time :) Let's test it.
Hello, Vladimir Kladov
Somehow KAY was left completely without a menu to select the path to the ROM, and it would also be desirable
put a separate menu under his ROM with tyr dos.
And by the way, the reassigned ROM paths are either not saved, or are reset to
default.
ZY It is better to store all settings as a simple text file.
Hello fan
Somehow KAY was left completely without a menu to select the path to the ROM, and it would also be desirable
make a separate menu for his ROM with Tyr Dos.
And by the way, the reassigned ROM paths are either not saved, or are reset to
default.
ZY It is better to store all settings as a simple text file.
Hello, Vladimir Kladov
The Saturday ZXAsm++ update is ready. Corrected multiplication in operations (* not in
the beginning of a line was also mistakenly considered the beginning of a comment in some
cases), generally counting expressions with more than 1 operation of type x+y*z. Added
sampling the substring "stroka"[ n[,L] ] from a string, comparing strings "s1" = "s2".
If you want, you can download the entire EMuZWin 2.7 bld 2.6b - I fixed it there
coloring of the lines, added the background to the coloring.
From what was planned, the only thing left to do seems to be to add tighter control
walking on the second and 3rd (if available) passes, so that the compilation follows the same
branches and used the same values in DUP, FOR (and I don’t even know that yet
how to make it cheaper). Well, this is actually not critical. If
The code is written intelligently, it should follow the same branches, no problems. But
two-pass is useful for counting forward references. (A three-pass
allows optimization of transitions in additional IFcond-type structures)
Hello fan
As for the models, for now I have settled on the Pentagon with Scorpio (and
then Scorpio has the wrong tact, as they write to me). There will be a time (and a place) -
I will bring to mind other models and video modes, both hdd and TR-DOS. In the meantime, I'm at war with
compilers. Now ZXAsm, then I return to Pascal. This will take a long time.
Hello, Vladimir Kladov
Yesterday I put build 2.6f (fix for loading the palette in the 256 editor,
nothing else).
Hello, Vladimir Kladov
build 2.6g (fixed issues: toolbar in asme, splitter missing).
Hello, Vladimir Kladov
Version 2.7 build 2.7 - everything that was discussed in the threads about the Bit n, (HL) command +
corrected multicolor for scorpio.
command Bit n,(hl) [http://zx.pk.ru/showthread.php?p=40912#post40912]
request to realists from emulators [http://zx.pk.ru/showthread.php?t=2586]
There may still be changes, but at the moment all tests agree, and there is a choice
of 2 processors. Thank you!
Hello, Vladimir Kladov
I posted the final release of version 2.7 - build 2.8, in anticipation of version 2.8. Just
I’m finally fixing a couple of bugs in the debugger so that 2.8 is “cooking” until
condition, there was an alternative.
The speed control ladder works, but the contents of IX are not checked (forgot,
I was thinking about video modes). So you can use it if you want. From next
versions will already be checked, so it’s better to enter. (This is me about the thread
“Managing the emulator from the Spectrum program”, if anyone is wondering what kind of traps
there are more like that).
The finale is the finale, but if the bug is serious, then I will edit it - the sources are all from
I left the version in case of fire.
Hello, Vladimir Kladov
thanks for the detailed explanation.
Hello, goodboy
Why do you even need to look into the emulator directory? Sort files by
types, and it won’t fit in well. ROM's can be conveniently stored in a separate folder. (A
Now I’ll grumble 8-EE) It’s my business whether I should place files in the heap or not in the heap, but
not the user. In addition, there will always be a certain number of users
who will actively object to my creating a multitude
subdirectories. There is no need to refer to the Amiga, I have never seen it. you
better look at other emulators, 1 is used almost everywhere
directory. Finally, a simple consideration: in the text of the program, to address
to a file from the startup directory, I write GetStartDir + filename. And the code is shorter.
You should have seen my emulator directory, all the sources are lying around there, and
all garbage files that the compiler creates: 2225 files, 147 MB, not
counting another 97 subdirectories = 0.98Gigabytes 8-]]