CRC errors?

John,

thanks for re-testing. From my point of view, the results do not yet absolve FreeAXP ! Do the VBN numbers reported by BACKUP/COMPARE show some recognisable pattern (like every n blocks) ? Could you try DUMP/BLOCK... from one of those 'corrupted' blocks from the good and bad file, so we might be able to see the corruption pattern ?

Bruce,

note that John was even testing with a different .ZIP file (i.e. ALPHA_FOR0801.ZIP instead of BASIC016.ZIP as used by Astrodanco) and certainly a different CD ! And he did see the same kind of behaviour.

Astrodanco,

did you try to copy the .ZIP file with BACKUP from the CD to the FreeAXP disk and unzip from there ?

Volker.<hr>

John,

if you happen to be running PersonalAlpha as well, could you try this test with PersonalAlpha ?

Volker.

John and I are testing with the same zip file from the same CD. I've ordered an 8.3 Hobbyist CD to do a second check.<hr>

I installed the latest ECOs to the Alpha VMS 8.3 and made a new copy of the zip file. The new copy was corrupt.

So for the next test, FTP the same .ISO file from the Mac OS-X host that was mounted by the Virtual Box Windows XP VM to the VMS host.

I then mounted it as a logical disk and copied the zip file.
That copy tested fine with unzip.

So it appears that there is nothing wrong with the CD-ROM media or the MacBook Pro DVD drive.

Interestingly I was only able to get about 200KiB/s FTP transfer rate between the Mac OS-X host and the AlphaVMS emulated machine.

John,

please dump and compare the 'good' and 'bad' blocks. This may tell us something about the type of corruption.

Volker.

The behavior of the corruption has changed with build 283.

Before random data was wrong in the copied file.

Now for the file appdev010.zip, blocks 118, 119, 242, 243 are all zeros.

For for the file alpha_for0801.zip, the copied file has block 243 as all zeros.

I copied up the diff to Eisner and sent you an e-mail.


I was also able to get 7.5 Mb/sec on the upload.