trmshob tests

Forwarded by Dima Boyko
Orig.Area: 562.TALKS
Orig.From: Dima Boyko
Orig.To : All
Orig.Subj: trmshob tests

═══════════════════════════ CUT ═══════════════════════════

H a yu sh k i, All !

I’m here, as promised, carried out some tests of the subject programs.
Firstly, a glitch with disk damage REALLY exists.
And it is described by the author of this software himself:

In #3D13 mode, all untested
drives are considered as 80/DS and no checks are made for TR-DOS disks. Using #3D13 with MS-DOS is not possible. Mode
#3D13 was introduced for working with a hard drive, but I recommend using it when
working with disk drives that do not have a delay in engine spin-up.
Using a turboloader on such
disk drives often leads to the destruction of
sectors during writing (reading is normal)
due to failure to achieve the required engine rotation speed values.

I did not have the described mode set, and the drive was a Soviet one.

Secondly, another glitch was discovered, which is not described by the author:(
If you write files to an MS disk in the gluing mode, then they are not normally deleted from it by this program. That is, if you try to delete such a file for the first
time, then by turning the drives the software will not do anything externally. In subsequent
attempts to erase any files, she complains about problems with the file.
But ifrename the file recorded in splicing mode, then everything falls into place :)

Thirdly, it is unknown how records made by the subject program will behave with
Linux. Since the author warns:

a few words about working with MS-DOS
─────────────────── ─────────────────────
Better never write small letters and symbols with codes 0..31
and 128..255 into the file name, because the reaction of MS-DOS to such
files is very strange (when copying from
TR-DOS to MS-DOS, characters with codes 0-31 and
127-255, as well as characters "*+,./:;<=>?[\]
`| are replaced with the underscore character _)
and never format/write under
Windows, because it may happen that
the cluster will have one sector or the file header will have 64 bytes (and not
32) or the FAT table will not be 12-bit
(the program always assumes that on disks
12-bit FAT), and it is best to format here.

And all the listed pranks are typical for Linux copers.
Well, the sabzhe utility writes to disks formatted under Linux. But how
Linux will perceive the disks formatted by it, and how the software will read what is written by
Linux - time will tell. If anyone is interested in the results, please let me know.

Dima
-+- Lara Croft with a 4.50 meter scythe
+ Origin: Don’t kick the kettle!!!!! (500:562/1.14)
═══════════════════════════ CUT═════════════════════════

С наилучшими, Dima Boyko.

Greetings, Dima!

[Monday October 2, 2000] at [18:26:21] Dima Boyko wrote a letter to All.
They discussed the topic "trmshob tests". Let's continue what we started...

DB> Here, as promised, I carried out some tests of the subject
DB> program.
DB> Firstly, a glitch with disk damage REALLY exists.
DB> And it is described by the author of this software himself:

[skip]

DB> My described mode was not set, and the drive
DB> Soviet-golimy.

I now have a “flattened” version of the subject (trmshob+) with an external
driver file for working with Pentagon-512 memory.

============== here we will start the file “trmshob+.C” ==

DIFFERENCES FROM TRMSHOB
───────────────── ───────────────────────
1. Names of MS-DOS files are displayed in small
letters, and subdirectories are displayed in real
characters (in capital letters).
2. When copying from MS-DOS to TR-DOS
, file names (except hobet ones)
are converted to small letters.
3. Ignoring the error "incorrect
hobet file length."
4. Pressing SS+1, SS+2, SS+3 and SS+4
in TR-DOS and MS-DOS windows leads to
re-reading the disk on drives
A, B, C and D.
5. Output to Real Commander is supported.
This is done as follows:
when loading the copier,
all memory pages are checked,which the TRMSDRV driver supports (if it is loaded, otherwise the standard memory Spec-
trum 128) for the presence of RC and, if found, copies it to the very
last logical memory page, and the amount of free memory
will become 16K less.
When you press SS+Q in any the moment
of the RC keyboard poll is copied from
the last page to the zero and
JP #C000.
6. The sound of the keys
on AY has been removed due to lack of memory:
TRMSHOB+ has 0 bytes free.

============== here we will finish the file "trmshob+.C" ================

Yes, indeed, I also sometimes had tr-dos disks beaten while
writing to them with a subject program (and ms-dos' ones, including /720kb/), if a turbo loader/saver was set for them
(drives). This happened _only_ on
the Bulgarian drive EC5323.01. When #3d13 was prescribed for tr-dos drives, the glitches disappeared. Now both TEACs are running, no problems at all. To be on the safe side, I again set '#3d13' for tr-dos drives. I'm in no hurry. ;)

By the way, these same Bulgarian ES5323.01 worked unstable with accelerated #3d13
(reading) in the firmware TR-DOS_v5.13F by Andrey Isaev.

DB> Secondly, another glitch was discovered, which is not described by the author
DB> :(
DB> If you write files to MS disk in thegluing, then they
DB> are then not normally removed from it by this program
DB>. Those. If you try to delete such a file
DB> for the first time
DB>, then by turning the drives the software will not do anything externally
DB>. on subsequent
DB> attempts to erase any files, it complains about problems with the
DB> fat.

I just checked - no problems (ms-dos = 720kb). ;) MS-DOS disks on Speck
format in Honey_Commander_v4.00

[skip]

p.s. "Turbo-VG93" is always on. Disabling Turbo-VG93 when using EC5323.01
did not fix the problem with disk corruption by the turbo saver.

p.p.s. Or maybe they stuck an ancient subject here? ;)

Sincerely, Felix.

[I ZX]