On (12 May 96) Paul Edwards wrote to Keith Richardson...



KR> unfortunately i didn't do my own test, loading a 34 meg tiff into

KR> photoshop, but it seems noticably faster.



KR> btw, if you want a hires file, you can have this. it is a picture of a

KR> doll scanned off the original object at 400 dpi and enhanced to 800 dpi.

KR> as there was no photography involved, it is the best that can be got

KR> from a normal desktop scanner.



 PE> 6.5 hours at V32bis.  Yes please.  Are you able to send it say

 PE> 12 noon - 6.30pm on a weekday?  That is the slackest time.

 PE> BFN.  Paul.



dunno what happenned, the xfer rate was abysmal. the drop out was when

wendy rang me to find out where i was (sitting in a traffic jam in the

city after some dickhead tried to go through the tunnel in a truck that

was a foot or so higher that the tunnel was), and, by the time i got

home, it was so near the end that there didn't seem much point doing

anything. is there anything in your logs that may give a clue?



i rang annabelle bits today, the main dealers for adaptec, and spoke to

their "tech support". the guy seemed to know fuck all about scsi, he had

heard of scsi select (tm) and that it was capable of verifying disks.

when i asked why it should abort on a single recovered read error he

suggested "looks like your disk is on the way out" when i suggested to

him that  re-assigning blocks with tempory errors is what the verify

program is supposed to do, he suggested that i look in adaptec's web

page. christ, if i was after shitty advice, i'd have asked rod (:



                        Keith



... COBOL programs are an example of artificial inelegance.



--- PPoint 2.00