Path: utzoo!utgpu!jarvis.csri.toronto.edu!cs.utexas.edu!uunet!mcsun!sunic!tut!ra!rosenber From: rosenber@ra.abo.fi (Robin Rosenberg INF) Newsgroups: comp.sys.amiga Subject: Re: MSH Message-ID: <219@ra.abo.fi> Date: 25 Feb 90 09:34:25 GMT References: <23012@usc.edu> <2414@cpsc6a.att.com> Organization: Abo Akademi, Finland Lines: 64 In article <2414@cpsc6a.att.com>, crs@cpsc6a.att.com (Chris "I'm Outta Here!" Seaman) writes: > rosenber@ra.abo.fi (Robin Rosenberg INF) writes: > < >the cyl and side no. won't advance and nothing gets formatted. > < > I use ARP 1.3, but (other than mount) I can't see where that would make a > difference. BTW, the docs instruct you to remove a newly formatted disk > and re-insert it for the handler to 'see' it. I added the line Reserved=0 as someone mentioned earlier and it works now, also with AmigaDos mount. So much for that. > < I have also noted a few other bugs. MSh seems to set the DirEntry type to an > < illgal value (zero ?) so list and dir lists files on a disk as DIR even though > < they are plain files. Most file requesters report them correctly though.> > If you use ARP's Dir, or ls (I use 3.1), there is not a problem. Ls is > better anyway, because it doesn't throw a silly requester at you when you > make a typo in a device name (such as typing fd0: instead of df0:). It > just says "fd0: not found". Hmm. I just checked this. It's amigados DIR and LIST which have the bugs. Thay seem to check for specific values. So it's not a MSH bug. > < Bug#2: Volume labels longer than eight characters are handled incorrectly, > < i.e they are broken into 8+"."+3 character which causes problems. > > This is not a bug, it's a feature :-). Actaully it is a requirement. Volume > labels on MS-DOS are treated as a special file (OK, so they're directory > entries, but not real files), and therefore must conform to the MS-DOS filename > template. This IS a bug since if you get the volume name from the volume node you will get one name and from Examine() yuo will get anither. Anyway, the fact that volume names are stored as directory entries doesn't mean that they should be treated as such under AmigaDos, it's just the way they are stored. If the volume if called LONGDISKNAME the volume node say so. Examine says it's called LONGDISK.NAM. Just the fact that some programs get the full name of a file before opening it and those programs don't work. > < Anyway it's a good attempt. Seems usable if one has a PC to format disks on. > > I use it in conjunction with my office 386 (saves lots of download time > for c.b.a. postings), and have had no problems (well, almost none). I > can format disks on either machine and have them recognized by the other. > > The only repeatable problem I've had so far is that, since MS-DOS disks > don't have a unique time stamp on them (at least through DOS 3.3), if you > have to disks with the same name (or no name), you will only get one WB > icon. This is really a "non-problem", I suppose, since there is no point > to using MS-DOS disks from WB (although you CAN give your MS-DOS disks > a disk icon). I guess this could be fixed by storing the volume data with the volume name since the volume name is stored as a directory entry. There is one more problem if you want to use messyformat disks from Workbench; You can't since it impossible to have a file with the extension '.INFO'. Also, it would be nice if someone could tell me what the compiler options for Manx means so I could try to fix the misfeatures myself if nobody else has. ------------------ Robin Rosenberg