INFO-MAC-REQUEST@SUMEX-AIM.ARPA (Moderator David Gelphman...) (08/31/86)
INFO-MAC Digest Saturday, 30 Aug 1986 Volume 4 : Issue 107
Today's Topics:
Symbolic Debuggers/Lightspeed Pascal
Lightspeed Pascal + Mac/XL = WOW!
Pernicious bug in Videx Mail Center, clobbers HFS Backup, etc.
July tech1 notes
file recovery tools
digital circuit simulator
APDA product list
MAC+ upgrade for Lisa owners?
Smalltalk on the Macintosh
About the regular vco conference...
----------------------------------------------------------------------
Date: Wed, 27 Aug 86 13:45:46 edt
From: rs4u@andrew.cmu.edu (Richard Siegel)
Subject: Symbolic Debuggers/Lightspeed Pascal
As. Mr. Gelphman says, the only two >compilers< that are shipping with symbolic
debuggers are MS Fortran and Lightspeed Pascal. The Macintosh Pascal
interpreter also has a debugger, to which Lightspeed's is very similar.
For those of you who don't know, get this: Lightspeed's
debugging capabilities are impressive. There is an "observe" window,
into which you can type variable names, function references, or expressions.
There's an "Instant" window in which
you can execute any valid statement (except
no declarations). The "LightsBug" window lets you look at registers and memory.
You can step through a program, trace through it (high-speed stepping), set
breakpoints, and then either run and stop at each breakpoint, or just pause
briefly at each one. You can tell the debugger to stop whenever you call an
A-Trap (basically a ROM routine).
In addition to all this, the run-time protection is amazing. If you have a
fatal system crash in your program somewhere, Lightspeed DOESN'T die. Instead,
it puts up its little error box, describes the system error (Illegal address,
autointerrupt, nil pointer, whatever), then it goes to the line of your code
where the error occurred! It's rather nifty.
I do have two gripes about Lightspeed Pascal, tho'. It is really inefficient
of disk space; the project file probably has all the library code loadded in
it too. Also, the linker does
not strip out unused code, like TML Pascal's linker
does (which is why I still like TML). I would be really happy to trade a few
seconds of link time for lots of disk space (I am probably the only Certified
Developer alive that still uses a 512K Mac (old roms) and two 400K (horrors!)
disk drives). An identical application, when linked in TML Pascal, took up 6K
on disk, including resources. The same program,, with resources, took up 16K
using Lightspeed. A smart linker sure would be nice..
But that compile speed is addictive. I wrote a program that generated 10000
lines of dummy Pascal code. Lightspeed compiled that program at a rate of 11000
lines per minute (yes, that is eleven thousand).
Anyone out there used the Prodigy 4 board and want to talk about it?
--Ric
P.S.
Just in case you were wonder, I am not using a prerelease or beta version of
Lightspeed Pascal. I am using the Genuine Article (it went on sale at the Expo)
------------------------------
Date: Fri 29 Aug 86 10:33:56-PDT
From: David L. Edwards <DLE@SRI-STRIPE.ARPA>
Subject: Lightspeed Pascal + Mac/XL = WOW!
After receiving alerts such as "That feature not supported" or "Go buy
a REAL Mac" on my Mac/XL, I was prepared for the worst when I ran
Lightspeed Pascal. I knew it had a built-in debugger and the outside
of the package doesn't mention the XL. But I had faith in Think after
using Lightspeed C so I bought it sight unseen. I was pleased to see
that it ran OK so I quickly modified their example so that I could get
the bad news about the debugger. The source debugger worked and it
was fun! As excitement started to build, I entered LightsBug and not
only NO alert but also lots of useful information present in an easy
to use Mac-style interface. I was dumbfounded! The more I worked
with it the more excited I became. Think Technology has done
something that even Apple hasn't done faithfully... supported all Macs.
XL owners should buy Lightspeed Pascal and forget the trade-in 'deal'.
After an evening of use, I fell in love with it. It stands alone in
the face of other development systems. It fully supports the toolbox,
has a true source level debugger, integration with Macsbug (which you
will probably never use), LightsBug (you have to see), an external
module interface (similar to Modula-2), REL converted for MDS support,
a code profiler, a desk accessory debugging interface (DAShell), code
resource support...
The only thing that I was disappointed in is their editor. Lightspeed
C could have a more powerful editor but Lightspeed Pascal has fewer
features (but some enhancements). What was nice was that they did
include an Editor program which is fun to use and modify.
No crashes, no unexpected problems and a good user interface makes
this my nomination for development system of the year. My only
warning is that it is addictive and you will find yourself sitting at
work wondering why you didn't call in sick today!
-dle
------------------------------
Date: Fri, 29 Aug 86 13:55:45 pdt
From: Mike Wirth <mcw@lll-crg.ARPA>
Subject: Pernicious bug in Videx Mail Center, clobbers HFS Backup,
Subject: etc.
I had a terrible time trying to use PCPC's HFS Backup program--80% of the
time when I inserted a disk, it wouldn't be recognized by the program (which
would just sit and wait). I'd have to use command-1 to eject it and try
again and again and ... This got pretty tiring when it took 28 disks to
back up my HD20. After searching many blind alleys and some very helpful
discussions with Robin at PCPC, I discovered that the problem wasn't in
HFS Backup at all! It was in Videx's Mail Center software. (A gold star
for PCPC; they were extremely helpful, a fact made all the more remarkable
since HFS Backup only costs $39 less discounts).
Here's what happened. When you install Mail Center, two items are installed
in the system file:
1) a desk accessory called Mail Preferences which is used to turn on and
off the mail receiver and set preferences.
2) an INIT resource which executes at boot time and installs a background
process to watch the AppleTalk port for mail traffic (i.e., executes the
Name Binding Protocol for the named receiver, according to Jim Berry of
Videx).
The background NBP process apparently executes whether or not(!) the mail
receiver is turned on. It causes a problem with the handling of disk
insertion events roughly 80% of the time (sounds like a timing dependent
problem to me). You can verify this by running any program's file open
dialog (I used Font/DA Mover 3.2). Select "Open..." from the file menu
and then watch what happens in the dialog window when you insert a floppy.
If everything works OK, the EJECT and DRIVE buttons should light up and
the file window should update to show the files on the floppy. If Mail
Center is installed, most of the time this will NOT happen; only the
DRIVE button will light up. Twenty percent of the time, it will work OK.
I talked to the folks at Videx (including Bill Rea in Tech. Support and
Jim Berry, the product engineer for Mail Center). Jim even sent me a beta
test copy (version 1.54a) of the receiver, all to no avail. He was able
to reproduce the problem on a 512K Mac and noted that it does NOT occur on
Mac Pluses or 512K Macs with the new ROMs. I verified that it does NOT
occur on a Mac XL running MacWorks (weird! since it uses the same code as
the old ROMs). He said he might look into the problem further, but was
not highly motivated to do so since it only occurred on old 512K machines.
(A bronze star to Videx, slightly tarnished, for their help so far, but
without a firm commitment to chase the problem to its source).
So those of you out there with 512K Macs and Mail Center, be forewarned.
We're upgrading our Macs with new ROMs this weekend. I'll let you know if
the problem goes away. But my technical curiosity is aroused. My quick
reading of Inside Macintosh led to the information that excess events will
get lost if the event queue is full. I surmised that the Mail Center NBP
was generating lots of events and filling the queue; and that since the
Mac Plus and the Mac XL have 1 Meg of memory, maybe they have bigger event
queues. But Jim Berry contends that the disk insertion events are always
getting posted to the queue properly; the problem is downstream from there.
Anybody else out there seen this sort of problem?
Mike Wirth
Standard Oil Production Co., Houston
713-552-8670
mcw@lll-crg.arpa
------------------------------
Date: Sun, 24 Aug 86 22:38 EDT
From: Michael.Fryd@A.CS.CMU.EDU (X435MF0E)
Subject: July tech1 notes
This is the july 1986 tech note release.
When you decode with BinHex 4.0 you will get
a PackIt II (Compressed) file.
When you unpack things you will have the individual tech notes.
-Michael Fryd
[archived as
[SUMEX-AIM.Stanford.EDU]<INFO-MAC>TECHNOTES-JULY86-PART1.HQX
[SUMEX-AIM.Stanford.EDU]<INFO-MAC>TECHNOTES-JULY86-PART2.HQX
[SUMEX-AIM.Stanford.EDU]<INFO-MAC>TECHNOTES-JULY86-PART3.HQX
[SUMEX-AIM.Stanford.EDU]<INFO-MAC>TECHNOTES-JULY86-PART4.HQX
[SUMEX-AIM.Stanford.EDU]<INFO-MAC>TECHNOTES-JULY86-PART5.HQX
DAVEG
]
------------------------------
Date: Fri, 29 Aug 86 17:59:47 PDT
From: <DAVEG@SLACVM.BITNET>
Reply-to: DAVEG%SLACVM.BITNET@Forsythe.Stanford.Edu
Subject: file recovery tools
[ brought over from usenet- DAVEG]
The following is a program I wrote to help me recover trashed documents
and disks that get brought in to me. You might consider it a poor man's
disk and file recovery program with a few added features. It is called
File Tricks and is neither elegant nor flashy, but does its job. I could
have spent a lot more time on user interface and added utility, and I
probably will. As it stands, it works so I feel others might benefit from
it. The accompanying documents are general discussions on recovering files
and disks and they refer to the use of File Tricks. At presnt, they are
the only documentation of the program. The file must be unpacked with
UnPit (or Packit) after BinHex'ing.
Joseph Buchanan
University of Utah Computer Center
[ archived as
[SUMEX-AIM.Stanford.EDU]<INFO-MAC>UTILITY-FILEFIX.HQX
DAVEG
]
------------------------------
Date: Fri, 29 Aug 86 18:03:58 PDT
From: <DAVEG@SLACVM.BITNET>
Reply-to: DAVEG%SLACVM.BITNET@Forsythe.Stanford.Edu
Subject: digital circuit simulator
A digital logic simulator which allows subcircuits for arbitrary
complexity. Excellent for both learning logic and designing and testing
circuits. Pulled off from USENET.
[ archived as
[SUMEX-AIM.Stanford.EDU]<INFO-MAC>>CIRCUIT-SIMULATOR.HQX
DAVEG
]
------------------------------
Date: Fri, 29 Aug 86 12:26:00 PDT
From: <DAVEG@SLACVM.BITNET>
Reply-to: DAVEG%SLACVM.BITNET@Forsythe.Stanford.Edu
Subject: APDA product list
This is the initial product list for APDA which includes both
Apple ][ stuff and Macintosh products.
[ archived as
[SUMEX-AIM.Stanford.EDU]<INFO-MAC>APDA-PRODUCT-AUG86.HQX
DAVEG
]
------------------------------
Date: Fri, 29 Aug 86 08:06 ???
From: YOUNG%RCSMPA%gmr.com@CSNET-RELAY.ARPA
Subject: MAC+ upgrade for Lisa owners?
In response to the earlier question regarding the trade-in for the
MAC+ for Lisa owners, please be advised that we h ave now been waiting
over two months for the MAC+ supposedly offerred for the Lisa trade-in.
This is another example of Apple's lack of meeting its commitments in
a timely manner, and continued poor treatment of former Lisa owners.
Good luck to current MAC owners when Apple's new machine is announced
next year! And Apple wonders why large corporations haven't been buying
MACs.
-- R. Young, General Motors
------------------------------
Date: Sat, 30 Aug 86 16:48:24 pdt
From: Bernard Aboba <bernard@ararat>
Subject: 68020/68881 Board by SPECTRA Micro Development
In the August 1986 issue of MacTutor (and again in the September issue),
SPECTRA Micro Development of Tucson, AZ advertsed a 68020/68881 board which
'will give you a complete 68020 workstation,' for $1295. Has anyone
seen this board or used it? Would appreciate details, such as clock
speed, memory, software and appletalk compatibility, etc.
------------------------------
Date: Fri, 29 Aug 86 10:56:12 -0200
From: Ole Solberg <O_Solberg%vax.runit.unit.uninett@tor.ARPA>
Subject: Smalltalk on the Macintosh
A year ago (as far as I remember) there was an announcment
on INFO-MAC for Smalltalk on the Mac.
Later it has been rather quiet on this subject.
Does anyone have any experiences to report ????
[ note from moderator: the recently announced APDA organization is selling
Smalltalk version 0.2 ( appropriate for 512K Mac ) for $50 and version 0.3
which only runs in 1Meg or more and is not Mac+ compatible for $50. See
the file <info-mac>APDA-INFO.HQX for more information on how to join
APDA and what it is. DAVEG ]
------------------------------
Date: Fri, 29 Aug 86 15:37 N
From: <INFOEARN%HLERUL5.BITNET@WISCVM.WISC.EDU>
Subject: About the regular vco conference...
Oops, I forgot to mention the vco conference is only on wednesdays. Of course
some of us might be logged on more often, but the regular one only every
wednesday at 0:00 GMT, channel 512 of the relay network on Bitnet.
See you then,
-- Thomas
FRUIN@HLERUL5.BITNET
------------------------------
End of INFO-MAC Digest
**********************