News
ZXNet echo conference «zx.spectrum»
Hello All!
Today I taught the emulator to read and write FDI disks! :)
IsDos works! Hooray! :) Tomorrow we will learn how to format, and
a problem arises here - if the new format does not coincide with the old one -
the information about the tracks and the track data follow each other and
when their length changes, all the data on the image is shifted, in short
in this matter the FDI format is not very good. Then you will need to create your own
format - an image of a physical disk with tracks of a fixed size
and information to the track about A1 labels, etc., by the way, you can add
labels of unstable bytes and track lengths, which will allow you to simulate
some types of protection (if not all). True, this image
will occupy about a megabyte.
PS: Here a question arose.
As I understand it, when writing a sector to the controller, the old
data field is not important, and even if it is missing, a new
will be written on top of the old one or on top of some information (for example, another
header) with the size from the header. Is this true or not?
Then, when reading a sector, the controller reads the header and starts
looking for the data field... what happens if the data field is missing at all?
error: sector not found? and if there is another header after the header,
is it ignored or taken as a data field? In general, I'm interested
in the controller's operating algorithm.
Crespect
┬ ┬┌┐┬┐ ┐┬┬
╨ґ╨╙ґ╨╙ ╨ ╨
Greetings, Vladimir!
─────────────────────────────────── ──────────────────────── ────────────────────
In distant times (13 Dec 97), Ser Vladimir wrote on a scroll All,
Bequeathing "News", and today (14 Dec 97) I decided intervene:
────────────────────────────────── ──────────────────────── ─────────────────────
VY> Today I taught the emulator to read and write FDI disks! :)
VY> IsDos works! Hooray! :) Tomorrow we will learn how to format, and
VY> a problem arises here - if the new format does not coincide with the old one -
VY> information about the tracks and the track data follow each other and
VY> when their length changes, all the data on the image is shifted, in short
VY> in this matter the FDI format is not very good. Then you will need to create your own
VY> format - an image of a physical disk with tracks of a fixed size
VY> and information to the track about A1 labels, etc., by the way, you can add
VY> unstable byte labels and track lengths, which will allow you to simulate
VY> some types of protection (if not all). True, this image
VY> will occupy about a megabyte.
Wonderful! How many opinions, so many formats!
VY> PS: Here a question arose.
VY> As I understand it, when writing a sector to the controller, the old
VY> data field is written on the drum, and even if it is missing, a new
VY> will be written on top of the old one or on topsome information (for example another
VY> header) with the size from the header. Is this true or not?
So .
VY> Next, when reading a sector, the controller reads the header and starts
VY> looking for the data field... what happens if the data field is missing at all?
Everything after is the data field.
VY> error not found sector? and if after the header there is another
VY> header,
The 'sector not found' error will only occur when the header is not found.
But if there is no data field, then there will be a CRC error.
VY> is it ignored or taken as a data field? In general, I'm interested in
It will be like this ^^^^^^^^^^^^^^^^^^^^^^^^^^.
VY> algorithm of the controller.
All the best. /──────────────────────\ [ Fantasy World - Rulez ]
Sincerely, │ Wanderer 'Sagacious' │ [ Getting up early - Must Die]
─────────────────── ────\──────────────────────/ [ Station 106.8 - Rulez ]
... Do what you must, what is meant to happen ...
VY>> interested
VS> It will be like this ^^^^^^^^^^^^^^^^^^^^^^^^^^.
IMHO no, the data field has a different AM.
Best regards
┬ ┬┌┐┬┐ ┐┬┬
╨ґ╨╙ґ╨╙ ╨ ╨
VY> header) with the size from the header. Is this true or not?
Absolutely true. Any information immediately following the sector header
will be overwritten (reminiscent of local formatting :)
Warm greetings, Slava!
And you reserve 5-6 kilos for each track at once (IMHO, there were no examples
of tracks where sectors in total would take up more than 5.125 Kb)
Warm greetings, Slava!
VM> Absolutely so.
More precisely like this:
"The SECTOR WRITE command is executed similar to the sector READ command in the part
of analyzing the index array...
...The DRQ signal is generated by requesting the first byte of data to
be written to the GMD. The chip then calculates 11 bytes with a single (22 bytes
at double) density to provide space between the index area and
data. From the moment of passing 11 or 22 bytes (if the first request of the DRQ
signal is serviced and the data is written to the RGD), a WTSB write strobe is issued and 6 bytes of zeros for
single (12 for double) density are written to disk. This corresponds
to a space record, and then the IAM byte is written. maybe either
FB..., or F8..."
Then the data is written (if the next byte of data is not received, 0 is
written instead), and at the end:
"...KK is written in the form of two bytes, then one byte FF and the WSTB signal is set
to a low level"
(from the previously popular brochure "1818 BG 93")
There was also a good brochure Ilyinets about TR-DOS 5.03 (not as monumental, of course, as P. Fedin, but it had all the salt :))
With warm greetings, Slava!
VY> Then, when reading a sector, the controller reads the header and starts
VY> looking for the data field... what happens if the data field is missing at all?
It doesn’t care about that. There is a standard sector format. The beginning of a header
is accompanied by a synchronization failure. Based on this criterion, the beginning of the sector is determined, and then there is simply a fixed structure (not counting the variable size of the data field).
Vyacheslav is that you?
...why did you tell Vladimir Yudin about News?
VM> And you reserve 5-6 kilos for each track at once (IMHO, there were no
VM> examples of tracks where sectors in total would take up more than 5,125 Kb)
on Amiga = 5632 :)
And finally, to solve this problem you need to do this: reduce
the disk drive speed as much as it will read
normally. Format and check the length of the track - we will get
THE MAXIMUM possible length of the track... Using the knowledge gained
you can not be afraid that there will be a lot or a little...
Good Luck...
... What? The second letter is an S? I thought it was "McDos".
VM> And there was also a good brochure by Ilyinets about TR-DOS 5.03 (not as
VM> of course, monumental as P. Fedin’s, but it had all the salt :))
by the way, yes. It was a very good book. Briefly and in detail. I haven't seen anything better about holes-dos anywhere....
p.s. once Logros even wanted to re-release it ;)
Good Luck...
... Backup not found: (A)bort (R)etry (P)anic
MR> on Amiga = 5632 :)
I'm talking about the Spectrum. On IBM - you can easily format a sector in 6k :)
Warm greetings, Slava!
Vyacheslav is that you?
...why were you telling me about News?
VM> I'm talking about Spectrum. Ha IBM - you can easily format a sector in 6k
VM> :)
I'm talking about DD disks. Well, and you make ONE sector at 6k, so what? For that matter, Vg93 allows you to make sectors of about 16kil :)
Good Luck...
... "Set phasers to 'humiliate,' Mr. Worf."