TOSEC Amiga Correction Thread

[dlfrsilver]

ok

[mai]

Originally Posted by Crashdisk

@ mai
you have no problem with the new disk crc $D504DC58 ?

I think your uploaded dump is fine, also both trained versions, which are in TOSEC looks good.

[Crashdisk]

@ mai

[10DB8EF6] Putty (1992)(System 3)(Disk 3 of 3)[cr QTX][a3][CBM A600 bundle]
=> Putty (1992)(System 3)(Disk 3 of 3)[cr QTX][b corrupt file][CBM A600 bundle]
...fixed, mai
[95D84AD6] Computer Third Reich (1992)(Avalon Hill)
=> [cr NMS]


[BA787637] FatTracks V2.0 (1990)(MSI)
=> [h TT]
...fixed, Crashdisk

[mai]

Originally Posted by Crashdisk

[95D84AD6] Computer Third Reich (1992)(Avalon Hill)
=> [cr NMS]

Sorry, cant find this crc32.

Originally Posted by Crashdisk

[BA787637] FatTracks V2.0 (1990)(MSI)
=> [h TT]

Its currently in application dat, yourself are busy with this dat!

[Crashdisk]

Originally Posted by mai

Sorry, cant find this crc32.

Its currently in application dat, yourself are busy with this dat!

as I said to prowler : Today, I have two left hands...

[prowler]

Yep, it's not your day, is it?

[Crashdisk]

Originally Posted by prowler

Yep, it's not your day, is it?

Indeed...
I'm going to bed early tonight

[mai]

@Crashdisk:
Your DMS-Workshop can detect several Amiga-DOS errors.
How to handle such errors, do we need a special [b] descriptor.
For example "Wrong data block sequence".
I suppose for any reason there is an error in file header block, but probably the whole datablocks are intact on disk.

[Crashdisk]

Originally Posted by mai

@Crashdisk:
Your DMS-Workshop can detect several Amiga-DOS errors.
How to handle such errors, do we need a special [b] descriptor.
For example "Wrong data block sequence".
I suppose for any reason there is an error in file header block, but probably the whole datablocks are intact on disk.

Wrong data block sequence:
Most of the time, when a copier/archiver (old version of DMS) can not read a track, it transfers the sectors previously present in the buffer. This results in duplication of sectors. ADFRen (but also AmigaDOS) sees no error because the checksums of the blocks are good but the file is still corrupt.
I use the generic descriptor [b corrupt file] that has the advantage of being true, unlike [b checksum error]. In addition, FFS disks do not have checksums to verify data integrity, so there is never a system error to report data corruption.
Add a descriptor for this specific case, I'm not sure how useful...
We could open a topic, here or on tosecdev, to consolidate and update the specific descriptors on Amiga?


EDIT

[XXXXXXXX] Fantastic Voyage (1991)(Centaur)(Disk 3 of 3)[cr SR]
=> [bootable]
It's just a cracktro but at least, something is happening...

[19CEA2AA] Fantastic Voyage (1991)(Centaur)(Disk 3 of 3)[cr SR][a]
=> [docs]
...fixed, mai
[XXXXXXXX] Neuronics (1989)(Thalion)[cr Atlantis]
=> [b corrupt file]
The "MotSam" sample is corrupt. There are 11 sectors duplicated, which explains the error detected in DMS-Workshop.
A fixed copy is in your FTP!
Note: [7FC13C4B] The file is "repaired", but as bad as the others ...
...fixed, mai
[CEE94ECD] Zyconix (1992)(Accolade)[cr CSL][a highscore]
=> Original CSL release
...fixed, mai
[76385A35] Zyconix (1992)(Accolade)[cr CSL]
=> Fixed CSL release (with DMS patch)
==> Cracktro/loader was changed but I do not know why
Cracktro extracted and posted in "the Zone"
...fixed, mai
[8B77B41A/BF276909/BC891217] Lethal Xcess (1991)(Eclipse)(Disk 2 of 2)[cr SR]
=> [b missing file]
"$00000015" has been supplied separately and should be transferred to the disk 2
...fixed, mai
-------

Other stuff uploaded :
All Dogs Go To Heaven (1991)(Merit)[h CPY - Main Source BBS][Electric Crayon]
World Cup USA 94 (1994)(U.S. Gold)(M4)(Disk 1 of 2)[cr CLS][f NTSC][cracktro disable]

[PandMonium]

The idea about consolidate Amiga descriptors sounds good, we have a bunch of words used all across tosec for the same meanings (although Amiga is not that bad and is hard to avoid them in the future).

[Crashdisk]

Originally Posted by PandMonium

The idea about consolidate Amiga descriptors sounds good, we have a bunch of words used all across tosec for the same meanings (although Amiga is not that bad and is hard to avoid them in the future).

For a while, I wanted to clarify a charter expanded to amiga, but since we are no more than two to work on the Amiga databases...

[mai]

Originally Posted by Crashdisk

Wrong data block sequence:
Most of the time, when a copier/archiver (old version of DMS) can not read a track, it transfers the sectors previously present in the buffer. This results in duplication of sectors. ADFRen (but also AmigaDOS) sees no error because the checksums of the blocks are good but the file is still corrupt.
I use the generic descriptor [b corrupt file] that has the advantage of being true, unlike [b checksum error]. In addition, FFS disks do not have checksums to verify data integrity, so there is never a system error to report data corruption.
Add a descriptor for this specific case, I'm not sure how useful...
We could open a topic, here or on tosecdev, to consolidate and update the specific descriptors on Amiga?

Thanks for explaination, so correct datablocks are not on disk anymore.
Same for different AmigaDOS errors, such as
"wrong data block key and sequence"
"file extension not found"?

BTW: Meanwhile i have began to change [b] descriptor "checksum error", because your arguments are valid.

Can you please check this:

Atlantis (1993)(Pisteurs)(fr)(PD)(Disk 2 of 6)(Disk 1)
0FE518D4
Atlantis (1993)(Pisteurs)(fr)(PD)(Disk 2 of 6)(Disk 1)[b corrupt files]
70BB0564

either both have corrupt files or both have no corrupt files, i cant see a difference.

[Crashdisk]

Originally Posted by mai

Thanks for explaination, so correct datablocks are not on disk anymore.
Same for different AmigaDOS errors, such as
"wrong data block key and sequence"
"file extension not found"?

BTW: Meanwhile i have began to change [b] descriptor "checksum error", because your arguments are valid.

"file extension not found"
The "file header" is a sector that contains the first 72 sectors to a file, the following 72 pointers are stored in a sector called "file extension". Cannot read a file with missing a block of this type on FFS disk. In any case, the AmigaDOS does not like this kind of problem...

"wrong data block key and sequence"
The key of a data block refers to the sector where is stored the "header file"
The sequence of a data block is the order number of reading a file.
the sequence of a data block is the order number in which it must be read to load the file. This error indicates that the block is that of another file and this means that the file is considered incomplete. ex: "MotSam" of Neuronics

00071800h: 00 00 00 08 00 00 03 85 00 00 00 07 00 00 01 E8 ; .......….......è
00071810h: 00 00 03 8D 5F 68 0B 56 D9 81 16 5C E9 40 23 70 ; ..._h.Vف.\é@#p

00 00 00 08   Sector id (data block)
00 00 03 85   File header is in the sector 901 (key)
00 00 00 07   This is the seventh block of file data (sequence)
00 00 01 E8   Size of the data in the block : 488 bytes (this is the maximum)
00 00 03 8D   This is the number of the next sector to read
5F 68 0B 56   The block checksum
D9 81 16 5C   First data....

[mai]

There was a case where DMSWorkshop has reported AmigaDOS error on a disk, i cant remember the exact error.
There was also another disk with no error for comparing.
I have copied the corrupted file to another place and got the same crc32 as the good one.

[Crashdisk]

Originally Posted by mai

There was a case where DMSWorkshop has reported AmigaDOS error on a disk, i cant remember the exact error.
There was also another disk with no error for comparing.
I have copied the corrupted file to another place and got the same crc32 as the good one.

it would be interesting to inform me of disks that you seem to be misinterpreted.
it should be noted that the message "Wrong data block key with proper sequence" is not a fatal error, but it is suspect. If the checksum is $00000000, is that there is an error.
I'll make some adjustments for the next version...

[mai]

Originally Posted by Crashdisk

it should be noted that the message "Wrong data block key with proper sequence" is not a fatal error, but it is suspect. If the checksum is $00000000, is that there is an error.
I'll make some adjustments for the next version...

Indeed, could be this error message, i am sure checksum was $00000000,
after copying checksum was correct.
Thats why i was not 100% sure to mark every disk as bad, where a file has checksum $00000000 at this specific error.
Anyway, this DMS-workshop is an excellent program.
In current state not worth to provide to EAB community?

[Crashdisk]

Originally Posted by mai

Indeed, could be this error message, i am sure checksum was $00000000,
after copying checksum was correct.
Thats why i was not 100% sure to mark every disk as bad, where a file has checksum $00000000 at this specific error.

Indeed, we must take into consideration some subjective criteria such as for fake files generated by bootgirl. They are recognized by DMS-Workshop but some customizations not still. And I do not even talk about some exotic protections...
Some cases, similar to Neuronics, are real errors.

Originally Posted by mai

In current state not worth to provide to EAB community?

I am not yet decided about it...

[mai]

Originally Posted by Crashdisk

Indeed, we must take into consideration some subjective criteria such as for fake files generated by bootgirl. They are recognized by DMS-Workshop but some customizations not still. And I do not even talk about some exotic protections...
Some cases, similar to Neuronics, are real errors.

Especially this 297 bytes files!

[Crashdisk]

Originally Posted by mai

Especially this 297 bytes files!

Can you explain to me?

[mai]

Originally Posted by Crashdisk

Can you explain to me?

You know, this is file 297 bytes on sector 11.
There are so much, some early games, which have such file, i dont know why, protection related, pseudo file?
You are the specialist!
example:

Three Stooges, The (1987)(Cinemaware)(US)(Disk 1 of 2)[cr MAAD]
DC6A7DAB