🚀 go-pugleaf

Rainbow EchoMail Digest                     Mar 27, 1990

In this issue:
  RE: MS-DOS 3.X FOR RAINBO      (3 messages)
  RE: TURBO PASCAL
  RE: OPUS 1.10 / RB'S


  Articles posted to either INFO-DEC-MICRO or comp.sys.dec.micro are
currently gatewayed automatically to the Rainbow Echo on Fidonet; thus,
you do not need to take any special action to respond to articles in
these digests.

  Please send reports of problems or suggestions for improvement of this
digest to GTHEALL@PENNDRLS.UPENN.EDU (ARPA Internet).
------------------------------------------------------------------------
Date:     03-25-90 (09:57)
To:       DAVID MAROUN
Subject:  RE: MS-DOS 3.X FOR RAINBO
From:     FRANK MALLORY

DM> Try running CHKDSK under MS-DOS 3.10 in a subdirectory on a hard disk.
DM> I tried it on a 10 megabyte hard disk (an ST412) and thought that the
DM> computer had crashed.  After reboting and trying again, I realized that
DM> the program really did take a long time.

I don't understand why CHKDSK is taking so long for you.  Perhaps your HD
needs to be defragmented?

I just timed it on my RD53 and find it takes 53 secs; that's with 32MB, 60
directories, 4,898 files.  I'm going to try it again after defragging; also,
with BUFFERS=1 as used by you.


--- msged 1.99L MSC
* Origin: Silver Bullet - Silver Spring, Maryland - 301-622-2247 (1:109/123)

------------------------------
Date:     03-25-90
To:       DAVID MAROUN
Subject:  RE: MS-DOS 3.X FOR RAINBO
From:     FRANK MALLORY

DM> I have made the tests, and I have stated the facts.  As for the buffers
DM> assignment, I use BUFFERS=1 for MS-DOS 3.10, since I have seen tests
DM> involving different values for the assignment, and the results are so
DM> slightly different that setting up buffers is not worth the trouble.
DM> I would use BUFFERS=0, but MS-DOS does not allow that assignment.

I tried CHKDSK on both partitions of my Miniscribe 3650 with the results
shown below.  The 8MB partition has 28 dirs, 447 files; whereas the 32MB
partition has 41 dirs, 1066 files.  This was done both before and after
defragging, with no significant time differences (however, my disk was not
very heavily fragmented to start).  I was originally using BUFFERS=50; after
getting those figures, I rebooted with BUFFERS=1 and retimed CHKDSK.  Tried
again with BUFFERS=20.  Here are the results:

8MB                        32MB

Buffers:        1    20   50             1      20      50
Time (sec)     ---    7    9           228      18      22

I didn't run CHKDSK with BUFFERS=1 on the 8MB partition, because when I ran
it  on the 32MB it sounded like the drive was about to self-destruct.  But
boot-up  and SEDT, both of which run from the 8MB partition, were noticeably
slower.

My conclusion is that the time taken by CHKDSK varies directly with the number
of files in the partition and is an inverse function of the number of buffers
up to some point.  After these experiments, I set the number of buffers
to  20 (which is the same as what I was using on my RD53).

I recommend you set BUFFERS=20, reboot, and retry CHKDSK.  Let us know the
results.


--- msged 1.99L MSC
* Origin: Silver Bullet - Silver Spring, Maryland - 301-622-2247 (1:109/123)

------------------------------
Date:     03-24-90 (20:05)
To:       DAVID MAROUN
Subject:  RE: MS-DOS 3.X FOR RAINBO
From:     VAN VANHORN

OK, I decided that this nonsense had gone on long enough.  I shut the board
down and ran CHKDSK from MS-DOS 3.10, 2.11, and 2.05.  The hard disk capacity
is 38.5 Megs, partitioned as 32 Megs on E: and 6.5 Megs on F:.

Under 3.10, CHKDSK reported full stats on the E: partition in 1:02, and the
F: partition in less that :02.

Under 2.11, CHKDSK scrolled like mad on the E: partition, and reported properl
in :04.5 on the F: partition.

Under 2.05, CHKDSK reported propable non-DOS disk on E:, then immediately
found 22 megs of lost clusters.  On F: it found 9 megs of lost clusters,
on a 6.5 Meg partition!

The long and short of it is that CHKDSK is SLOWER under 2.11 when it works.

Of course, that isn't really all that meaningful, since you only need to
run CHKDSK occasionally.

I don't pretend to be the world's foremost DOS expert, but I know there are
reasons why OPUS can't run under DOS 2.xx.  8 Meg partitions would be a nightm
to keep track of when users occasionally upload quarter-meg files.

But that doesn't mean YOU need to run DOS 3.xx.  You can do what you like.
I would prefer that if you truly want to keep your head stuck in the sand
that you limit your calling to BBS systems that run under DOS 2.xx.  That
way I won't ever, ever have to read any of your errant nonsense again.

Van Van Horn,
Horsefeathers BBS.


--- Opus-CBCS 1.10
* Origin: Horsefeathers + The Census?  Be creative! + (1:343/30.0)

------------------------------
Date:     03-24-90 (23:22)
To:       CHRIS RYAN
Subject:  RE: TURBO PASCAL
From:     DAVID MAROUN

Thanks for filling in the details that slipped my mind when I wrote to you
before.


I had Turbo Pascal 3 in mind when I wrote my message to you. Note, though,
that more recent versions of MS-DOS Turbo Pascal WILL run
on a Rainbow in command-line mode;  that is, you do not get the built-in edito
and other features of the "integrated environment", but you can still compile.


---
* Origin: Glacier Peak Rainbow, Bellevue, WA - 206/644-8431 (Opus 1:343/3)

------------------------------
Date:     03-24-90 (20:17)
To:       BILL HLIWA
Subject:  RE: OPUS 1.10 / RB'S
From:     VAN VANHORN

Just go for it.  I was up and running 1.10 in less than two hours.
I don't want to clutter up this echo on the subject, but the things
that didn't fly right off the bat were easier to find than I thought they
would be.  One of the great things is that I only have to keep one nodelist
now, instead of a full list for BINK and a tiny one for OPUS.

If you want more details, net mail, or watch for my comments in the MEADOW.

Van.


--- Opus-CBCS 1.10
* Origin: Horsefeathers + The Census?  Be creative! + (1:343/30.0)

------------------------------