Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Path: utzoo!utgpu!water!watnot!watmath!clyde!rutgers!husc6!seismo!rochester!cornell!uw-beaver!apollo!weber_w
From: weber_w@apollo.UUCP
Newsgroups: comp.sys.atari.st
Subject: Re: GEMBOOT tested on US ROM's
Message-ID: <342b571c.1f6@apollo.uucp>
Date: Thu, 9-Apr-87 09:11:00 EST
Article-I.D.: apollo.342b571c.1f6
Posted: Thu Apr  9 09:11:00 1987
Date-Received: Sat, 11-Apr-87 13:51:34 EST
References: <8704031501.AA13923@inria.inria.fr> <341cab34.1f6@apollo.uucp> <788@viper.UUCP>
Reply-To: weber_w@apollo.UUCP (Walt Weber)
Distribution: world
Organization: Apollo Computer, Chelmsford, MA
Lines: 41

In article <788@viper.UUCP> john@viper.UUCP (John Stanley) writes:
>In article <341cab34.1f6@apollo.uucp> weber_w@apollo.UUCP (Walt Weber) writes:
>>
>>Rebooting without GEMBOOT after that caused Show Info on the partition
>>to trash a portion of the drive FAT, and could even be made to cause a folder
>>copy to "two-bomb" crash the desktop.
>
>  Walt, I suspect part of the show-info problem is that it uses repeated
>fsfirst/fsnext calls to Tos which will probably bomb any time you try it
>on a drive with over 50 folders on a non-GemBoot enhanced system.

Agreed - the statement was more in the line of "explaining the testing" than
in "Gee, look, I've never seen THIS before" :-)

>  BTW, does the trashing of the FAT always come from doing a show-info?
>This seems a bit odd since showinfo itself doesn't write -at-all- to the
>disks.  Are you sure the order you listed cause-and-effect is correct?

I mis-spoke when I wrote that. ( "Mis-spoke" == word used when attempting to
distance oneself from obviously bad information.  Alternative to "shin kicking"
used in certain HQ facilities. :-)  If the FAT had actually been trashed, the
subsequent reboot using GEMBOOT would not have shown the partitions as healthy.

What I should have said was that "Show Info would report erroneous results in
the folder count, file count, bytes used, bytes available fields with the
folder/file counts LOWER than normal, bytes used LOWER, bytes avail LOWER, and
( used + available ) not equal partition capacity".  Furthermore, subsequent
copies to ANOTHER partition could either trash the FAT or two-bomb the desktop;
this might be related to how far over the limit the system had gone during a
session.  In one case, 79 folders on D: were all accessible, and additional
files and folders could be created by carefully tracking the count of folders
"seen", but it unfortunately blew off all of E: (empty at the time).  Show Info
on E: would report 0 files, 0 folders, 0 bytes used, 0 bytes free.  Easisest 
solution was to re-format the disk, since recovery would be inconclusive without
a lot of work.

My apologies again for distributing bad information.
-- 
Walt Weber               PHONE: (617) 256-6600 x7004
Apollo Computer          GENIE: W.WEBER
Chelmsford, Mass.   COMPUSERVE: 76515,2423