NeOS

DS> Hello to you, Roman !!!

DS> One day on Wednesday April 7, 1999 at 14:31:41 Roman Petrov wrote to Kirill
DS> Frolov
DS> on the topic 'NeOS for Scorp' Nefurychit.':

RP>> PS: I ask everyone who is willing and able test the functions of the
RP> NeOS kernel, read the docs and catch any possible misunderstandings so that
RP>> I can fix them all.

DS> Is there any software for it at all? or a purely bare system?

This system is most reminiscent of is-dos in ideology.
Ideological differences are visible only in:
1.crammed only under #0000,
2.added work with upper memory through the system,
3.added libraries that are still incomprehensible (let me at least
look at them).

now about disadvantages:
1.the kernel is built incorrectly from the point of view of encapsulation.
2.the command interpreter has non-standard commands (drive A
is worth it; it would be possible to do A:, as everywhere in similar
systems).
3.it is buggy when inserting left drives (this is not a drawback, but
a defect, as far as I understand).
4.the presence of system management commands: STACK, CALL, etc.
gives reason to believe that this is a debugger. But where are
breakpoints, normal dumps, registers?processor, etc.
5.lack of at least test software.

In conclusion, I would like to read the authors' opinion regarding
the positioning of the system on the spectrum of tasks, on the spectrum of
hardware, on the spectrum of users. And also their thoughts on
the openness of their system.

+-Good luck, All!
|I was with you
|Valentin Pimenov aka Valker/Style_Group
+---------------

Once Mon 12 Apr 1999 at 00:27:00, while wiping dust from the monitor, I noticed how
Valentin Pimenov was arguing with All on the topic of NeOS and I decided to join the bazaar:

VP> now about the shortcomings:
VP> 1. the kernel was built incorrectly from the point of view of encapsulation.

What is this? word ^^^^^^^^^^^^?

VP> 2. the command interpreter has non-standard commands (drive A
VP> is worth it; you could do A:, as everywhere else in similar
VP> systems).

I agree with this.

VP> 3. glitches when inserting left drives (this is not a drawback, but a
VP> defect, as far as I understand).

I have driver problems I would say...

VP> 4. the presence of system control commands: STACK, CALL, etc.
VP> gives reason to believe that this is a debugger. But where are
VP> breakpoints, normal dumps, processor registers, etc.
VP> 5. lack of at least test software.

;) but this is not a debugger, but an OS!
And managing the stack from the OS is quite convenient.

Sincerily yours,LEO/SG. [Beat the Jews - save Russia!]

Well, what can you do - the system is not quite finished yet.

VP> In conclusion, I would like to read the authors’ opinion on
VP> positioning the system on the spectrum of tasks, on the spectrum
Tasks:
1. Make the work of system programmers easier, because especially lately, they
begin to lack normal functions for working with the disk (byte-byte access).
2. Mitigate the incompatibility problem. Memory over 128Kb is no longer uncommon, but it’s just implemented differently for everyone. So programmers
have to support all memory expansion schemes, and if the authors suddenly do not have access to
a particular model, then the owners of such machines find themselves disadvantaged. And not only memory problems. With the same mice - some people fundamentally do not support Kempston Mouse, some AY-Mouse, few people support both of them. As a result, all mouse owners suffer. And with NeOS, such a problem doesn’t even arise: there is a standard for the mouse driver and the program simply uses the driver, and does not itself access the mouse.

The hardware is any computer compatible with the Spectrum 128. The system works fine with a disk drive, but, Naturally, it will be more pleasant to work with it with
hard drive.

VP> equipment, on the spectrum of users. And also their thoughts on
Users... It’s difficult to talk about _users_, while there is nosoftware.
What can be said about programmers is that they are mainly systems programmers,
authors of network software. They are the ones who need such a system. But this is only at the beginning, then, if the system takes root, other authors will switch to it.
NeOS is, first of all, an _alternative_ to TR-DOS and iS-DOS.

VP> about the openness of their system.
The system is completely open. Absolutely
nothing is hidden from the user/programmer. Everything is documented, there is not a single undocumented function, the system is freely complemented by drivers.

Bye, Valentin! Megus/Brainwave^X-Project. [ZX] [NeOS]

Today I fixed the oldest glitch in NeOS with the command line editor (there were
terrible glitches when reaching the edge of the text window). And also at the request of the workers :) I finally implemented batch files (scripts or bat files, whatever you prefer to call it). Because The space in the ROM is very limited, there are no
additional commands in the scripts, except for enabling/disabling the display of
executable commands and the comment symbol.
So the question arose: what would be better to run when starting NeOS - a code
file or a script?

PS: I accept applications for posting less buggy and updated version of NeOS :-)

Bye, All! Megus/Brainwave^X-Project. [ZX] [NeOS]

Fri Apr 16, 1999, Roman Petrov squeezed the keyboard, hoping that the following
words would reach All:

RP> is very limited, there are no additional commands in the scripts, except
RP> enabling/disabling the display of executable commands and the
RP> comment symbol. So the question arose: what would be better to run when starting
RP> NeOS - a code file or a script?

Of course, a script!

RP> PS: I accept applications for posting a less buggy and updated version
RP> NeOS :-)

And whose word will be the last? ;)))

[Team Kelly] [I ZX]
Hasta la vista.
Tim Kelly

... E-Mail: tim_...@mail.ru ICQ UIN: 34684028

LM> ;) but this is not a debugger, but an OS!
LM> And managing the stack from the OS is quite convenient.

From the OS, the standard command interpreter allows you to manage
the stack, as well as change memory cells, I have only seen
Spectrum BASIC :) Why is this needed! The user
is not interested in what kind of stack there is, and the format of the executable file must
be such that the system at startup can correctly allocate
space for the stack, etc. The OS is a layer between the hardware
and the user. This is taught in the first year of any major university. And such things as memory, stack, music coprocessor, etc.
are hardware from the OS point of view.

+ - All the best, Leonid!

Friday April 16, 1999 16:10, Roman Petrov wrote All:

RP> So the question arose: what would be better to launch at the beginning

TK> Whose word will be the last? ;)))
Probably a moderator :-) By the way, how about allowing me to freely (within reasonable
limits) throw out the new NeOS? This will happen no more often than once every 2 weeks,
and then less often, probably.

Bye, Tim! Megus/Brainwave^X-Project. [ZX] [NeOS]

On Saturday April 17 1999 Alex Noman wrote to Roman Petrov:

RP>> So the question arose: what would be better to launch when starting
RP>> NeOS - a code file or a script?
AN> Script.
Brevity is the sister of talent :-) Persuaded, I’ll make scripts. The file name will be
neosboot.scp

Bye, Alex! Megus/Brainwave^X-Project. [ZX] [NeOS]

RP> Well, what can you do - the system is not quite finished yet.

Well, at least they would do some kind of CheckIt.

RP> Tasks:
RP> 1. Make the work of system programmers easier, because especially lately
RP> them

"With good intentions..."

RP> 2. Mitigate the problem of incompatibility. Memory over 128Kb is no longer
RP> a rarity, here

Here is-dos, imho, has been very successful.

RP> Equipment - any computer compatible with the Spectrum 128. The system works great
RP> with a floppy drive, but, naturally, it’s more pleasant with it will work with
RP> hard drive.

What file system?

Only script. Or you get an axis for slave-hackers, everywhere you need an assembly.

Kirill Frolov. [ZX]

Sat Apr 17 1999, Roman Petrov squeezed the keyboard, hoping that the following words would reach Tim Kelly:

RP>>> comment symbol. So the question arose: what would be better
RP>>> to launch when starting NeOS - a code file or a script?
TK>> Of course, a script!

RP> No problems - five minutes of work.

That’s good...

[Team Kelly] [I ZX]
Hasta la vista.
Tim Kelly

... E-Mail: tim_...@mail.ru ICQ UIN: 34684028

One day on Friday April 16, 1999 at 17:10:28 Roman Petrov wrote to All on the topic 'NeOS':

RP> So the question arose: what would be better to run when starting
RP> NeOS - a code file or a script?

batnik.

With best wishes, Dmitry.

RP>> The name of the file will be neosboot.scp
TK> I would call neosboot.spt ;)

Het, namely .scp. He we first. ;)

TK> Tim Kelly

C regards, Alex Noman.

KF> Only script. Or it turns out to be an axis for cool hackers, you need
KF> assembler everywhere.

That’s not the point. If you always think about what is for “cool hackers” and what is for lamers, then you can go far. A script is better because calling a code file
from a script is somewhat more correct than calling a script from a code file. :) It’s not for nothing that Vasik initially starts on spec, and Autoekzek starts on spec. :)

KF> Kirill Frolov.

With respect, Alex Noman.

ëءث-شد 19 Apr 99 × 07:43:54 ذزدةعدغإج ثدخئجةثش
Tim Kelly ة Roman Petrov سذدزةجة خء شإحص NeOS
ٌ زإغةج ذزدؤدجضةشط ـشدش سذدز:

AN>>> َثزةذش.
RP>> ëpءشثدسشط - سإسشpء شءجءخشء :-) ُاد×دpةجة, سؤإجءہ سثpةذشظ. Hءع×ءخةإ
RP>> ئءتجء آصؤإش neosboot.scp
TK> ٌ آظ خءع×ءج neosboot.spt ;)
ل ر آظ autoexec.bat, شءث إسشط × حسؤدس ة ةسؤدس شءث عءقإح ؤإجءشط قإاد شد س×دإ?

---€€€              ² َ ص×ءضإخةإح, لخؤزإت ¯              €€€€--

[Team Kelly] [I ZX]
Hasta la vista.
Tim Kelly

... E-Mail: tim_...@mail.ru ICQ UIN: 34684028

TK> Then autoexec.bat ... ;)

Well, it depends on what is closer to it - to the script or to the batch file. The capabilities of scripts
are somewhat different from batch files. :) If there are no scripts as such in neos and there won’t be any, then there’s no point in fooling people. :) And if there are or will be,
a compromise is possible: autoexec.scp. ;)))

Is it the name? ;) All that matters is what happens.

[Team Kelly] [I ZX]
Hasta la vista.
Tim Kelly

... E-Mail: tim_...@mail.ru ICQ UIN: 34684028

On Friday April 16 1999 Valentin Pimenov wrote to Leonid Mishankov:

[ Skip ]
VP> interest in a specific FS and will select the FS driver to work with the disk,
VP> if both drivers, when testing the disk, declare that it is “their
VP> native”?
Huh... Indeed, an interesting situation...

VP> But there is a way out of this situation.
Which one? Tell me, I'll try to implement it.

VP> From the OS, the standard command interpreter allows you to manage
VP> the stack, as well as change memory cells, I have only seen
VP> Spectrum BASIC :) Why is this needed! The
VP> user is not interested in what kind of stack there is, and the format of the executable file should
This is all correct for the OS of more powerful machines. For Specky, it is still more convenient to add
some functions that are specific only to him.

On Tuesday April 20 1999 Alex Noman wrote to Kirill Frolov:

AN> That’s not the point. If you always think about what is for “cool-hackers” and
AN> what is for lamers, then you can go far. The script is better, because
Well, this is Frolov :-) He really likes to find fault with everything :-)

AN> calling a code file from a script is somewhat more correct than a script
Yes, you're right :-) And besides running the code file, you can write something else in the
script. Or you can pick up some drivers.

RP>> works great with a floppy drive, but, naturally, more pleasant
VP> What file system?
Arbitrary. Whichever driver you install, this is the one it will work with. I have 2 types of drivers for working with disk: disk device drivers and file system drivers. That. you can install drivers for several file
systems and all of them will be transparently supported.

VP> Openness is not “not a secret”, but first of all
VP> OSI compliance is a standard for connecting open systems.
But with this, please, in more detail.

Or maybe the usual AUTOEXES.BAT is better?

Kirill Frolov. [ZX]

TK> To the point.

Can you make an installer, even a simple one? And then there are many different types of computers, and some don’t have a mouse, others don’t need the is-dos-640 driver, etc. You
will get tired of making all sorts of different versions and posting here...

All the best.
Sincerely Roman.

ëءث-شد َزإؤء 21 لذزإجط 1999 × 19:53:57 ذزدةعدغإج ثدخئجةثش
Alex Noman ة Tim Kelly سذدزةجة خء شإحص NeOS
ٌ زإغةج ذزدؤدجضةشط ـشدش سذدز:

TK>> ôداؤء صض autoexec.bat ... ;)
AN> Hص, ـشد سحدشزر ث قإحص آجةضإ - ث سثزةذشص ةجة ث آءشخةثص. ÷دعحدضخدسشة سثزةذشد×
AN> خإسثدجطثد زءعخرشسر س ذءثإشخظحة ئءتجءحة. :)
ٍءعرسخة حخإ ؤصزءثص زءعخةأص حإضؤص 'سثزةذشءحة' ة 'ذءثإشخظحة ئءتجءحة'.

RP> Not weak, but broken ;-) IMHO, CD is quite normal.

Half of the code will have to be revised/rewritten? ;)

RP> Bye, Valentin! Megus/Brainwave^X-Project. [ZX] [NeOS]

With respect, Alex Noman.

AN>> Well, it depends on what you are closer to - to the script or to the body file. The capabilities of
AN>> scripts differ somewhat from batch files. :)
AI> Explain to me, a fool, the difference between 'scripts' and 'batch files'.

I wish I could come here... He wouldn't teach you anything. :))

I don’t think anyone will be interested in reading, which is already obvious. Okay... All
'fools' (ts) wrinkled their minds. I'm starting to explain. :)

There is practically no difference, except for a trifle. ;) A batch file, as
the name suggests, is designed to work with a group of files and most often
is simply _executed_line by line_ by the system. In MS-DOS, for example, _all_ lines of a file
are considered as OS commands, which are divided into internal (path, set, ver,
date, etc.) and external (the same as file names). Let me emphasize once again, the OS
executes the batch file line by line, nothing is compiled.

The script, on the contrary, is most often written in a language close to a Javascript language, in which, although
not always, it is possible to declare various data types and functions.
Quite often, the language is generally an OO. The complexity of the problems solved by OS commands
and by those provided by an independent programming language, you must admit,
varies like day and night. In addition to all of the above, scripts are usually
compiled before execution.

Here's what I thought... Maybe you have an idea aboutscripts ends with:

waitfor "..."
send "..."
goto x:

??? :) Or do you think you don’t need more from _scripts_? :) (I'm not talking about neos yet.)

AI> -+-───══════════════╡ Sincerely, Andrey

With respect, Alex Noman.

ps/ 2EM: Teacher, don’t kick the lamer, plz... :)))

AN> scripts differ somewhat from batch files. :) If in NeOS
I agree :-) Scripts in NeOS will be full-fledged, but only in the form of an external
interpreter due to the large capabilities. In the meantime, these are just small
batch files that simply allow you to combine several NeOS commands into one
group.

AN> there are no scripts as such and there won’t be any, so there’s no point in fooling people
AN>. :) And if there are or will be, a compromise is possible: autoexec.scp.
AN> Is it really in the name? ;) The only important thing is that it works out.
:-) Oh, these advisers ;-) We persuaded you, you will get autoexec.bat :-)

AN> I wish I could come here... He wouldn't teach you anything. :))
Where did you put it? tied to a stool near the spec so that he could finish the S-terminal
quickly? ;)

AN> I don’t think anyone will be interested in reading, which is already obvious. Okay...
AN> All
AN> 'fools' (ts) wrinkled their minds. I'm starting to explain. :)
I'm starting to read...

AN> There is practically no difference, except for a trifle. ;). representations.

AN> ??? :) Or do you think you don’t need more from _scripts_? :). ╞══════════════────--

AI> Where did you put it? tied to a stool near the spec so that he
AI> S-terminal would finish writing faster? ;)

Yes, he doesn’t read email on weekends. And he seemed to have added the s-terminal.
Called today. Now we need to test, and this is the most difficult thing. ;)) We work with him in opposite phases. ;)

AN> Half of the code will have to be revised/rewritten? ;)
No, enter ~10-15 in the command interpreter :-)

PS: I have already changed my mind about autoexec.bat. There will be neosboot.bat :-)

On Wednesday April 21 1999 Roman Milukov wrote to Roman Petrov:

RM> Maybe you can make an installer, even a simple one? And then there are a lot
RM> of all kinds of computers of different types, and some don’t have a mouse, others don’t have a driver
So now I’m torn between the installer and the commander. I have already started writing both, and I need to continue writing both :-) So I don’t know what to give priority to. The installer currently only has an interface that is 80% ready,
the commander has learned to read and display the directory (not yet fully functional - without
highlighting groups of files with colors and without displaying the file size).

RM> is-dos-640 is not needed, etc. You'll get tired of making all sorts of different versions
RM> and
The driver is dead... Because it's absolutely unnecessary. It was generally written in the absence of an MS-DOS driver, and work with several file systems had to be checked somehow.

Bye, Roman! Megus/Brainwave^X-Project. [ZX] [NeOS]

Somehow 21 Apr 99 at 09:47:29 Roman Milukov and Roman Petrov were beating each other up
The reason was NeOS

RM> Maybe you can make an installer, even a simple one? And then there are a lot of different
RM> computers of different types, and some don’t have a mouse, others don’t need the is-dos-640
RM> driver, etc.
Golden words! Self-automated kernel assembly with customization for specific hardware and the user's desire - that would be just wonderful! Moreover
there is no invention of the wheel here - let's remember Unix.

RM> You'll get tired of making all sorts of different versions and posting here...
That's it. And so the user needed something - he just took the installer and
configured the resulting kernel for specific purposes.

With best regards, Yuri Voynalovich AKA Combinator/PH7.

VP>> In the commander, the command line is not required.
VP>> If you can exit from the commander back to the interpreter
VP>> of the command line from where the commander was called.

KF> I don’t know isdos - with every error a window falls out and again
KF> run mon.com (mon+.res).

I had in mind that the interpreter should not be part of the kernel
of the system. Whichever file is registered in the system settings:
​​or , let it run.
And from it, run at least , at least , at least
. Those. all .exe files are not part of the kernel.

KF> Sachs. IMHO you need to make the command line the most important part of the system,
KF> and
KF> so that errors are written on the screen like MS-DOS.

The program sends errors to the top, i.e. to where it was called from.
And what this system/program will do with them is the business of this
system/program.

VP>> 2authors_of_NeOS:
VP>> Sew upthe interpreter into the kernel is a gross mistake.

KF> Well, it turned out to be a masdai.

What about the glued shell? I agree.

+-All the best, Kirill!

RP> Which one? Tell me, I'll try to implement it.

/---+-MSDOS ;subdirectory associated with the ms-dos driver
| |
| +-A ;access to floppy disk ms-dos
| +-B
| +-C
| +-G ;access to the ms-dos subsection on the screw
+-TRDOS
| |
| +-A
| +-B
+-ISDOS
| |
| +-B
+-BIN
+-TMP
+-WINDOWS

I.e. the main file system includes
subdirectories associated with the corresponding file systems.
Moreover, they do not necessarily have to be located in the root
directory.
Due to the refusal of autodetection of the FS, the concern for the correspondence between the
directory and the real disk falls on the user.

RP> This is all correct for the OS of more powerful machines. For Specky, it is still more convenient
RP> to add
RP> some functions that are specific only to him.

this is needed by _programmers_, not _users_! Most, I
think, don’t even know what it is.

+ - All the best, Roman!

RP> Could you suggest what kind of information such a
RP> test program should output?
RP> It’s just that I’m busy with the commander right now, and I still need other software. And I would be
RP> very
RP> glad if they gave me specific advice for other software.
RP> Develop
RP> a test program, throw it in the echo (or with soap, if you want), and we will implement it.
RP> Or,
RP> if you want, you can write it yourself. We will be only glad!

1.Processor type :)
2.Memory capacity,
3.system version,
4.relative speed of the machine,
5.availability of peripherals: joysticks/mice/CMOS/GS/modems/printers/
scanners/etc.,
6.available file systems,
7.free memory on all external media.

[skip]

RP> No, enter ~10-15 in the command interpreter :-)

10-15 what? :)

RP> PS: I have already changed my mind about autoexec.bat. There will be
RP> neosboot.bat :-)

And that's right. ;) I just don’t understand, will it be a script or a batch file? ;)

RP> Bye, Alex! Megus/Brainwave^X-Project. [ZX] [NeOS]

With respect, Alex Noman.

Sunday April 25, 1999 02:51, Roman Petrov wrote Roman Milukov: :)

Regards, Alex Noman.

On Sunday April 25 1999 Alex Noman wrote to Andrey Isaev:

AN> Yes, he doesn’t read mail on weekends. And he seems to have added the c-terminal like
AN>. Called today. Now we need to test, and this is the most difficult thing. ;))
AN> We work with him in antiphases. ;)
So let him throw it into the echo - we'll test it with all the echoes!

AN> I would like this to happen on spec, although it's practically not realistic.
In theory it's possible... I want to make an interpreter of full-fledged scripts (with
variables, with conditional branching, etc.) in the form of a separate program.
By the way, I don't even need it here They suggested doing something like the Amiga REXX. Now
this is unlikely to happen.
AN> All this is slow. Let there be at least a parody in neos. :) (No offense
AN> to the authors.)
No problem.

[Team Kelly] [I ZX]
Hasta la vista.
Tim Kelly

... E-Mail: tim_...@mail.ru ICQ UIN: 34684028

25 Apr 99 14:47, Valentin Pimenov wrote to Roman Petrov:

VP> 1.Processor type :)

Z80 or stump with emulator.

VP> 2.Memory capacity,

128k always.

VP> 4.relative speed machines,

100%

VP> 5. presence of peripherals: joysticks/mice/CMOS/GS/modems/printers/
VP> scanners/etc.,

It’s not possible to do plug and play. Here the GS test hangs if there is a cache.

Kirill Frolov. [ZX]

AN> I would like this to happen on spec, although it’s practically impossible.
AN> It’s all slow
AN> all this. Let there be at least a parody in neos. :) (No offense to the authors.)

Here's a parody, please, no need! There are already enough of them.
If it is impossible to implement this in a normal way, then we need to solve these problems
in other ways.

AN> Yes, he doesn’t read mail on weekends. And he seemed to have added the s-terminal.
That’s how he is. :) I haven’t finished writing yet... I don’t understand something about the upload. Melon and PC
for some reason have a slightly different “dialect” of the modem.

AN> I called today. Now we need to test, and this is the most difficult thing. ;)) We work with him in antiphases. ;)
2ALL: Basically I can only do it on weekends, but on weekends he leaves for the Volga. :)

Good luck! Evgeny.

■ The night is shorter than the day... The day will kill me... The world of illusions burns down in it. (C) Aria.

AN>> really.
RP> In theory, you can... I want to make an interpreter of full-fledged scripts (with
RP> variables, with conditional branching, etc.) in the form of a separate program.

In principle, the user menus in IS-DOS work in the same way. There are
no abstruse things there, you can still make a discount on the left code, but just remember
what kind of brakes it is. :)) I wonder why? :)

RP> By the way, I was even offered to do something like Amiga
RP> REXX. This is unlikely to happen.

Sorry, I’m not familiar with Rex. :)

RP> Bye, Alex! Megus/Brainwave^X-Project. [ZX] [NeOS]

With respect, Alex Noman.

EM> I am like that myself. :) I haven’t finished finishing it yet...

I finished it, then I didn’t finish it. :)) Let's decide. ;)

EM> I don’t understand something with the upload. Melon and PC, for some reason, have a slightly
EM> different “dialect” of the modem.

Well, the excuses have begun. :) You said that some other packages
are used there and that you understood everything.

EM> 2ALL: I can basically only do it on weekends, and he leaves for the Volga
EM> on weekends. :)

He will tell the whole country... everything. :)

EM> Good luck! Evgeny.

With respect, Alex Noman.

VP>> 1. Processor type :)

KF> Z80 or a stump with an emulator.

Z180/Z380. I think they can be tracked.

VP>> 2.Memory capacity,

KF> 128k always.

VP>> 4.relative speed of the machine,

KF> 100%

is this all a joke? then you need to put ":)". otherwise I don’t understand.

Tuesday April 27, 1999 13:53, Roman Milukov wrote to Alex Noman:

AN>> really. It's all slow. Let there be at least a parody in neos. :)
AN>> (No offense to the authors.)
RM> Here's a parody, please, no need! There are already enough of them.

Well, if the author is going to leave the name neosboot.bat, then we can
look at the problem from a different angle. Maybe it's the PC's body files that are poor compared to the neo-Osov ones? ;)))

RM> If it is impossible to implement this in a normal way, then we need to solve these
RM> problems in other ways.

Yeah. ;) In neos it is implemented differently. :))

RM> All the best.
RM> Best regards, Roman.

Regards, Alex Noman.

AN> 10-15 what? :)

Litpov beer.

A: B: C: D: there is on all computers.
Should be and in neose.

Kirill Frolov. [ZX]

Thursday April 29 1999 00:13, Kirill Frolov wrote to Alex Noman:

KF> A: B: C: D: is on all computers.
KF> It should be in neos too.

You tell the author that... I'm for it. ;)

KF> Kirill Frolov.

With respect, Alex Noman.

On Sunday April 25 1999 Valentin Pimenov wrote to Kirill Frolov:

VP> I meant that the interpreter should not be part of the kernel
VP> of the system. Whichever file is registered in the system settings:
VP> ​​or , let it run.
VP> And from it, run at least , at least , at least
VP> . Those. all .exe files are not part of the kernel.
This is all correct if you have a hard drive. Moreover, you can launch an external
interpreter.

On Wednesday April 28 1999 Alex Noman wrote to Roman Petrov:

AN> In principle, the user menu in is-dos works in the same way. There
AN> there are no abstruse things, you can still make a discount on the left code, but
AN> you remember what the brakes are. :)) I wonder why? :)
And we will do it without any brakes. This can be done, it can be done.

[ Skip ]

VP> That is. the main file system includes
VP> subdirectories associated with the corresponding file systems.
VP> Moreover, they do not have to be located in the root
I have already figured out how to get around even the case when the file systems are similar. You just need to do a more stringent check of the system. Then a file system similar to
TR-DOS, but different from it, will simply not be recognized by TR-DOS drivers.

On Thursday April 29 1999 Alex Noman wrote to Roman Milukov:

AN> Well, if the author was going to leave the name neosboot.bat, then you can
AN> look at the problem from a different angle. Maybe these PC batch files are pathetic
AN> compared to neo-Soviet ones? ;)))
You flatter me :) Ptssh batch files are more powerful compared to neosovskie ones, there
even the rudiments of a script structure are there, but I don’t have them _at all_, just a
sequence of commands.

PS: The new firmware is not being thrown due to the fact that I gave it to Andrey Isaev to finalize the drive driver. By the way, he says that under TR-DOS 5.03
it will be quite difficult to do :( You must have TR-DOS in a,(#1f):ret.

RP>> user? Let's consider these functions as a graphical
RP>> library :-) Will it work then? :-)
VP> 1. the kernel does not work with hardware other than the processor,
VP> memory (interrupt controller/DAP). Those. it can
VP> function in the absence of a screen/video memory,
VP> music processor, disk drives, etc.
I'm stuck, damn it :) You're now talking about the _classical_ way of organizing
operating systems. But what you say does not take into account the specifics of the Spectrum.

VP> 3. would make a real library, for example , where
VP> would put these non-encapsulated functions of the screen/music process -
Yeah, and everything would slow down wildly... I tried it - I know... Character-by-character output to the screen
worked with at the speed of a turtle, but everything is correct... I don’t want to do that.

[ Skip ]

KF> A: B: C: D: is on all computers.
KF> Should be in neos too.

On Saturday May 01 1999 Yuri Voynalovich wrote to Roman Petrov:

RP>> under TR-DOS 5.03 it will be a bit difficult to do :( You need to have it in ROM
RP>> TR-DOS in a,(#1f):ret.
YV> Is it worth it then to pervert under 5.03??? Probably it will turn out slower :(
Not particularly, but in terms of volume - it will be more _significant_. Andrey will make
an appeal using the standard address. And which standard - let him say for himself :)
I just don’t understand TR-DOS ROM at all.

YV> Anyway, almost everyone has >=5.04, and if anyone flashes Neos, then flashing a new trdos will not be a problem.
Uh-huh. I’m also planning to flash NeOS on Scorp, I’ll install TR-DOS 5.13 right away. By the way, you can’t just flash NeOS on Scorp - you’ll lose the service monitor. put.

Bye, Yuri! Megus/Brainwave^X-Project. [ZX] [NeOS]

VP>> No one wants to do P&P. CheckIt existed before the advent of P&P.

KF> A checks for no devices that have an impact on its further
KF> performance,
KF> does not try to identify.

Yeah, I’ll now make a reboot port - out (#1234), #ff equ Reset,
and then I’ll try to test it with a checkit:

“is our port working? reboot?..."
...
"does our reboot port work?..."
...
"does it work... "

Devices that have such a serious impact should be tested
with their special means, and not with general, universal ones.

27 Apr 99 you wrote here:

AN>> I called today. Now we need to test, and this is the most difficult thing. ;)) He and I work in antiphases. ;)

30 Apr 99 22:31, Roman Petrov wrote to Alex Noman:

RP> TR-DOS 5.03 will be a bit difficult to do :( You must have TR-DOS in
RP> a,(#1f):ret.

Better in a,(c) : ret.

Read A sector/track/data register is also sometimes necessary.
You can, of course, put the stack in the ROM, but why do that when it can be simpler?

Kirill Frolov [ZX]

RP> Yeah, and everything would slow down wildly... I tried it - I know... Character-by-character output on
RP> the screen worked at the speed of a turtle, but everything is correct... I don’t want it like that.

What prevents you from entering an entry point in the driver of a character output device for
printing a string of characters?

WBR, Oleg.

Fri 30 Apr 1999 at 22:31:48, Roman Petrov => Alex Noman:

RP> finalizing the drive driver. By the way, he says that under TR-DOS 5.03
RP> it will be quite difficult to do :( You need to have TR-DOS in a,(#1f):ret.

Why? (From my own experience - it’s not necessary).

WBR, Oleg.

YV> Anyway, almost everyone has >=5.04, and if someone flashes Neos, then trdos
YV> new
YV> flashing will not be a problem.

Who decides for everyone here? On scorps there is no need to change anything at all, but in tr-dos there is in a, (ts)/out (ts), a. So
I advise you to first figure out what and how.

+-All the best, Yuri!

[skip]
RP> I have already figured out how to get around even the case when the file systems are similar.
RP> Just
RP> we need to make it more rigid system check. Then a file system
RP> similar
RP> to TR-DOS, but different from it, will simply not be recognized by
RP> TR-DOS drivers.

Damn, I clearly said that this tricky new system
is located right on top of the TR-DOS logical structure of the
disk. Those. there is both a 9th sector and a tr-dos directory. So
your tr-dos driver _always_ recognizes this disk as native.

+-All the best, Roman!

+-All the best, Roman!
|I was with you
|Valentin Pimenov aka Valker/Style_Group
+---------------

[ZX] [LETI]

RP> worked at a snail's pace, but everything is correct... I don't want to do that.

If calling a library service is so much slower than
using kernel code, then most likely the very idea of ​​libraries
is not successfully implemented. How I imagine calling through a
library:

1.Before using any library service, you need to
install it. This can be done by the kernel code. Based on the name/number of the
library, it allocates memory to it and loads it from an external
device. The usage counter of this library is incremented
by "1".

2. The calling code passes the library manager a pointer to a table of custom addresses of the form:
N_adrs db 3 ;number of custom addresses
ncall1 db #f1 ;library procedure number
acall1 dw #6789 ;custom address
ncall2 db #12 ;...
acall2 dw #891a
ncall3 db #31
acall3 dw #9a1b

3.After which, calling the library is _no_ different from calling a
procedure.

4.After finishing the work, you need to notify the library manager
that this library is no longer needed. The manager reduces the
use counter, and if it reaches zero, the library is removed from
memory.

With this scheme, the increase in time occurs only due to
installation & address settings. Because These operations are carried out
relatively rarely, so the results are not slow at all.

On Sunday May 02 1999 Kirill Frolov wrote to Roman Petrov:

RP>> in a,(#1f):ret.
KF> Better in a,(c) : ret.
I believe, I agree. I’m telling you that I don’t understand TR-DOS at all.

KF> Sometimes you also need to read the sector/track/data register.
KF> You can, of course, put the stack in the ROM, but why do that when it can be simpler?
Why such perversions? There are normal firmwares.

RP>> 5.03 will be a bit difficult to do :( You need to have TR-DOS in
RP>> a,(#1f):ret.
OG> Why? (From my own experience - it’s not necessary).

So as not to touch the hidden areas of memory. =) This is not tyrdos. :)

OG> WBR, Oleg.

With respect, Alex Noman.

RP>> finalizing the drive driver. By the way, he says that under TR-DOS
RP>> 5.03 it will be difficult to do :( You need to have TR-DOS in
RP>> a,(#1f):ret.
OG> in the ROM Why? (From my own experience - it’s not necessary).

RP>> output to the screen worked at a snail's speed, but that's all
OG> What prevents you from entering an input point
OG> in the driver of a character output device to print a string of characters?
The idea is good, this will give some speedup for printing lines. It's true that it's too late to add all the graphical functions to the library - work on the kernel is completed.

2All: add such a function?

ëءث-شد ÷دسثزإسإخطإ 2 يءت 1999 × 19:25:01 ذزدةعدغإج ثدخئجةثش
Oleg Grigoriev ة Roman Petrov سذدزةجة خء شإحص NeOS
ٌ زإغةج ذزدؤدجضةشط ـشدش سذدز:

RP>> ؤدpءآدشثص ؤpءت×إp ؤةسثد×دؤء. ïخ, ثسشءشة, اد×دpةش, قشد ذدؤ TR-DOS 5.03
RP>> شpصؤخد×ءشد سؤإجءشط آصؤإش :( Hءؤد ةحإشط × ًُْ TR-DOS in a,(#1f):ret.
OG> ْءقإح? (ًد س×دإحص دذظشص - خء ئةا خإ خءؤد).
‏إاد 'عءقإح'? ‏إاد 'خإ خءؤد'? ïغةآثة قشد جة دآزءآءشظ×ءشط خإ خءؤد? حدضإش شظ إفإ
سثءضإغ ؤزءت×إز ثدشدزظت سإتقءس × خإدسإ خدزحءجطخد زءآدشءإش?

---€€€              ² َ ص×ءضإخةإح, لخؤزإت ¯              €€€€--

RP> The idea is good, it will give some speedup for printing lines. It's true that it's too late to add all the graphical functions to the
RP> library - work on the
RP> kernel has been completed.

RP> 2All: add such a function?
If it speeds up the output, and it will, then of course yes!
In Ms-Dos, by the way, this was the case initially, because This is a reasonable decision, and for
Spek it is even more necessary. (how did you not do this right away?)

With best regards, Yuri Voynalovich.
[Powered by Z80] [intel outside]

RP> Yep. I’m also planning to flash NeOS for Scorpio, and I’ll install TR-DOS 5.13
RP> right away. By the way, you can’t just flash NeOS on Scorp - the
RP> service monitor will be lost. You need to install an additional ROM.
Hmm.. but what about Scorpio people... probably not everyone will want the additional one. install the ROM.
Maybe we can find some kind of compromise for them? For example, keep Neos in their
abdominal area, or somehow use the shadow area...

In the only case it is necessary to read the contents of register 1F without
entering a new command into the VG. You can get by, but there will be a small brake.
And I couldn’t read the readiness from FF, everything is buggy there (or I was buggy).

Kirill Frolov. [ZX]

YV> Anyway, almost everyone has >=5.04, and if someone flashes Neos, then flashing a new trdos
YV> will not be a problem.

In my opinion, the person says very correctly: we will flash all the ROMs, i.e.
all 4 pages. It wouldn’t be a sin to enter IN a,(c), it’s time to stop tormenting
with reading Beta-Disk registers via ...
For example, Scorp’s ROM has long been inserted at address #3ff3 (correct me if I lied in
the address).

Sincerily yours,LEO/SG.

Monday May 03 1999 22:11, Yuri Voynalovich wrote to Roman Petrov:

YV> Hmm.. what about Scorpio people?. probably not everyone will want extra. ROM
YV> install. Maybe we can find some kind of compromise for them? For example
YV> keep Neos in their sub-arousal area, or somehow use
YV> the shadow area...

Memory can have only one owner. If neos works in the sub-arbitrary
area, then you will have to part with the shadow guy... :( Or are there options?

YV> With best regards, Yuri Voynalovich.

With respect, Alex Noman.

EM>> The Volga is leaving.

RM> It is necessary for All to pass a resolution prohibiting traveling “beyond the Volga”,
RM> especially on weekends! ;)

Well, yes, and also catch fish using folk remedies (dynamite). :) And then it all comes to fruition. %)

RM> All the best.
RM> Sincerely, Roman.

Respectfully, Alex Noman.

zy. 2all: You should kick EAT, by the way, otherwise I’ll soon be left without a modem =>
there will be no one to test the s-terminal. ;)

YV> If this speeds up the output, and this will speed up, then of course yes!
OK. I'll do it today. The acceleration will be almost unnoticeable on short lines, but on long lines it will be very noticeable (saving about 100-150 cycles per character).

YV> In Ms-Dos, by the way, this was the case initially, because this is a reasonable decision,
YV> for Speck it is even more necessary. (how did you not do this right away?)
I don’t know, apparently I didn’t guess :)

On Sunday May 02 1999 Valentin Pimenov wrote to Yuri Voynalovich:

YV>> Neos, then flashing a new trdos will not be a problem.
VP> Who decides for everyone here? On scorpions there is no need to change anything at all
VP>, but in tr-dos there is in a, (ts)/out (ts), a. So
If Scorpi has a specific standard for placement in a,(c):ret, then
share. But in TR-DOS 5.03 there is only out (c),a.

On Tuesday May 04 1999 Alex Noman wrote to Yuri Voynalovich:

AN> Memory can have only one owner. If neos works in the
AN> sub-routine area, then you will have to part with the shadower... :( Or is
Yes, the shadower disappears. It is by loading into the ROM area that I test NeOS on my
Scorp.

Bye, Alex! Megus/Brainwave^X-Project. [ZX] [NeOS]

RP>> Already there.
VP> They got it after all :)
And there’s nothing to do there :) Even less than 10 teams turned out :)

VP> Damn, I clearly said that this clever new system
VP> is located right on top of the tr-dos' logical structure
VP> disk. Those. there is both a 9th sector and a tr-dos directory. So
VP> that your TR-DOS driver _always_ recognizes this disk as a native one.
So I didn’t explain clearly :) In the 9th sector there are bytes characteristic of TR-DOS,
but it itself does not pay attention to them. This is the kind of byte we will use to indicate another file system. And my TR-DOS driver also looks at this byte,
thus rejecting the left system.

VP> If calling a library service is so much slower than
VP> using kernel code, then most likely the very idea of ​​
VP> libraries is poorly implemented. How I imagine calling through
Nowadays, not everything should be so slow, but calling a library from ROM
is extremely inconvenient :( Calling libraries is done exactly as a procedure.

[ Skip ]

VP> 4. After finishing the work, you need to notify the library manager
VP> that this library is no longer needed. The manager decreases the counter
I don’t have this. I’ll think about how to implement it (there is a possibility - there is just a
reserve byte in the library descriptor).

RP>> they do, IMHO. That is, in principle, it is a matter of time.
VP> This is not a file exchange protocol. It is a protocol for managing and
VP> data exchange. For example, launching a remote process, working with
VP> remote data arrays, etc.
In theory, and such. you can write programs...

So it’s written in Russian in plain English - THIS IS A SPECTRUM! ;-(

VP> With this scheme, the increase in time occurs only due to
VP> installation & address settings. Since these operations are carried out
VP> relatively rarely, the result is not slow at all.

Windows in miniature is obtained.

Kirill Frolov. [ZX]

RP> But this is not for me, this is for Andrey. If it is somehow possible with TR-DOS 5.03
RP> _without_using_ TR-DOS system variables to count #1f, then explain
RP> to him.

It’s absolutely impossible. All existing methods are based on the use of
TR-DOS variables and sending the next command to VG93.

Kirill Frolov. [ZX]

I warn you right away that they will draw on the Spectrum screen without any
neos. And on everyone else too, because your brakes there are terrible. And what
graphical functions?

There is no need to place neos in ROM, it will be a mess. Everyone should voluntarily
install Z180..Z380 and then there will be no problems with access to
any part of the memory.

Kirill Frolov. [ZX]

ëءث-شد ÷دسثزإسإخطإ 2 يءت 1999 × 22:27:00 ذزدةعدغإج ثدخئجةثش
Valentin Pimenov ة Yuri Voynalovich سذدزةجة خء شإحص NeOS
ٌ زإغةج ذزدؤدجضةشط ـشدش سذدز:

YV>> ÷س£ زء×خد ذدقشة ص ×سإب >=5.04, ء إسجة ثشد آصؤإش ذزدغة×ءشط Hإدس, شد ة
YV>> شزؤدس
YV>> خد×ظت
YV>> ذزدغةشط آصؤإش خإ ذزدآجإحء.
VP> üشد ثشد شصش عء ×سإب زإغءإش? Hء سثدزذءب ×ددآفإ خةقإاد ذإزإغة×ءشط
VP> خإ خءؤد, عءشد × شز-ؤدس'إ إسشط ةخ ء,(أ)/دصش (أ),ء. ôءث قشد
VP> سد×إشصہ ×خءقءجإ ذزةثةخصشط, قشد ؤء ثءث.
÷ شدح ؤزءت×إزإ قشد ×قإزء صجإشإج ٍدحءخص ةسذدجطعصإشسر جةآد in a,(#1f):ret
خءبدؤرفةتسر ذد ءؤزإسص #3c30 جةآد in a,(c):ret خءبدؤرفةتسر ذد ءؤزإسص #3ff3 ص
سثدزذد×.

OG>> What prevents you from entering the entry point
OG>> in the driver of a character output device to print a string of characters?

04 May 99 at 11:54:00, Kirill Frolov wrote to Roman Petrov:

KF> There is no need to place neos in ROM, it will be a mess. Everyone put
Is someone forcing you to flash neos in ROM? This is voluntary. You can
be a great original and flash Z180..Z380 into the modem ROM (but I don’t know how many memory meters there are ;))

KF> in a voluntary
KF> order Z180..Z380 and then no problems
KF> with access to
KF> any part of the memory.
One problem - Will you give me this percent? Will you completely remake my computer for it? Of course, I’m not averse to having a computer with a Z380, but only if
for this I only need to go somewhere to get it, and not fiddle around
to the point of complete exhaustion of both myself and my wallet.

Or better yet, you sit on the PC... And enjoy life under Windows :)

With respect, Kirill Maksarov. [Nastya] [I ZX]

Tuesday May 04 1999 22:56, Andrey Isaev wrote Valentin Pimenov:

AI> in a,(#1f):ret located at #3c30

Another standard. :))

AI> -+-───══════════════╡ Sincerely, Andrey

With respect, Alex Noman.

RP>>> REXX'a. This is unlikely to happen.
AN>> Sorry, I’m not familiar with Rex. :)
RP> Do you think I know you? I've never even seen an Amiga.

It's not practical to implement something like Arexx without real multitasking,
when with the help of this thing one program controls another, or a server
several. And for this to work, all programs through which the Arexx server controls them need appropriate ports. Otherwise, you'll end up with
an ordinary script, and it wouldn't be a bad idea to wrap it up with some commands,
something like a programming language, working with variables, etc.

With respect, Alex from FFC COMPUTERS.

>E-MAIL: asde...@softhome.net ICQ: #13064553

AI> What 'why'? What 'don't'?

Do you have problems with your organs of perception of visual information? :) I repeat:
why do you need to read the status register? Errors can be handled without this.

AI> Shouldn't there be any need to handle errors? can you
AI> also
AI> tell me the driver that now works normally in neos?

In total, I spent about ten minutes in neos and I can’t “say” anything about the driver
. However, I've written quite a lot of "turboloaders" to
assert that almost all errors can be handled without reading the
status register. The only exception may be write protection, and even that
is perfectly caught as a disk error.

WBR, Oleg.

Tue 4 May 1999 at 20:34:28, Roman Petrov => Yuri Voynalovich:

RP> OK. I'll do it today. The acceleration will be almost unnoticeable on short
RP> lines,
RP> but on large ones it will be noticeable and very noticeable (saving about 100-150 clock cycles for each
RP> character).

Why so little? The acceleration should be much greater. For example, a driver for
42 characters can print in fours of characters, and a 64-character driver in pairs - this
will give a very large increase.

WBR, Oleg.

p.s. Frame! Frame! :))

ëءث-شد َزإؤء 5 يءت 1999 × 19:10:44 ذزدةعدغإج ثدخئجةثش
Alex Noman ة Andrey Isaev سذدزةجة خء شإحص NeOS
ٌ زإغةج ذزدؤدجضةشط ـشدش سذدز:

AI>> in a,(#1f):ret خءبدؤرفةتسر ذد ءؤزإسص #3c30
AN> هفإ دؤةخ سشءخؤءزش. :))
èح... هاد شزصؤخد خءع×ءشط 'إفإ دؤخةح' ش.ث. إحص جإش آدجطغإ قإح سثدزذةدخءح شءثءر
ثدخسشزصثأةر ةسذدجطعد×ءجءسط إفإ × سءحظب زءخخةب ×إزسةرب بدخةقء ة إسجة إإ خإآظجد
شد بدخةق دآ ـشدح جءسثد×د ذزإؤصذزإضؤءج ة سآزءسظ×ءجسر شءث قشد ×دعحدضخد قشد شءثدإ
إسشط ص حخداةب, ء دخة دآ ـشدح ة خإ ذدؤدعزإ×ءہش :)

KF> I warn you right away that on the Spectrum screen they will draw
KF> without any neos. And on all the others too, because your brakes
KF> are terrible there. And what graphical functions?
Not necessarily, by the way. Windows, for example, can be displayed perfectly with NeOS.
Printing, by the way, is also very convenient. My commander uses a lot of
NeOS functions for the interface.

KF> There is no need to place NeOS in ROM, it will be a mess. Everyone should put Z180..Z380 in
KF> voluntarily and forcibly and then no problems
Uh-huh... You should have developed a computer like that first and distributed it to _all_
Spectrum users.
You have some kind of resource mania... It's bad for you to sit at a PC, oh that's harmful.

Bye, Kirill! Megus/Brainwave^X-Project. [ZX] [NeOS]

AI> In the driver that Roman flew away yesterday, either in
AI> a,(#1f):ret located at address #3c30 or in a,(c):ret located
AI> at address #3ff3 for Scorpios is used.
It works fine on TR-DOS 5.12 (under the emulator), I haven’t tested it on Scorpia yet -
I was deprived of Scorp for a week.

Bye, Andrey! Megus/Brainwave^X-Project. [ZX] [NeOS]

OG> In total, I spent about ten minutes in neos and I can’t say anything about the
OG> driver
OG>.
The current one only catches disk errors.

OG> However, I have written quite a lot of “turboloaders” to
OG> claim that almost all errors can be processed without reading the
OG> state register. The only exception may be write protection, and even that
OG> is perfectly caught as a disk error.
Yes, but at the same time variables are actively used, and this is not worth doing!

---───══════════════╡ Best regards, Andrey ╞══════════════────--

RP> If Scorpi has a specific standard for location in a,(c):ret,
RP> then share. But in TR-DOS 5.03 there is only out (c),a.

ADDR SCORPION TR-DOS 5.03

2A53 OUT (C),A : RET OUT (C),A : RET
3FF0 OUT (C),A : RET RST 38 : RST 38 : RST 38
3FF3 IN A,(C) : RET DB FF,FF,FF

Kirill Frolov. [ZX]

RP> So I didn’t explain clearly :) In the 9th sector there are bytes characteristic of
RP> TR-DOS, but he himself does not pay attention to them. We will use this byte and
RP> to indicate another file system. And my driver
RP> TR-DOS also looks at this byte, thereby rejecting the left system.

The new system is located ABOVE the TR-DOS DISK, for example
in the form of files on the TR-DOS disk! Well, like MOA CP/M disks.

In short, IT IS KNOWN IN ADVANCE that you can only tell the difference by the sticker on the floppy disk.

Kirill Frolov. [ZX]

KF> So it’s written in Russian in plain English - THIS IS A SPECTRUM! ;-(
Spectrum is also a computer. Systems must be built according to general principles, although
Spectrum still has its own specific features that must be taken into account
when writing a system.