[net.micro.mac] Delphi Mac Digest V2 #21

shulman@topaz.RUTGERS.EDU (Jeff Shulman) (06/01/86)

Delphi Mac Digest          Sunday, 1 June 1986      Volume 2 : Issue 21

Today's Topics:
     RE: Hard Disks
     RE: ThinkTank and Acta
     RE: Spelling checkers
     RE: Spelling checkers
     System 3.2
     Re: Network Hardware
     RE: Usenet Mac Digest V2 #41 (Re: Msg 8454)
     RE: Usenet Mac Digest V2 #41 (Re: Msg 8468)
     RE: Usenet Mac Digest V2 #41 (Re: Msg 8490)
     prolog II
     RE: LINT again
     Re: Alternative keyboards for Macintosh
     RE: PackIt II problems
     Reply to Pseudo-DMA Question
     Re: Laserwriter + MS Word Problem on Print Merge (with fix)
     Lisa ROMSwitch
     RE: List Manager (Re: Msg 8396)
     DIZero problems
     RE: DIZero problems (Re: Msg 46)
     RE: DIZero problems (Re: Msg 84)
     system enhancement?
     RE: system enhancement? (Re: Msg 8594)
     Health software development
     3.1.1 HFS Bug
     RE: 3.1.1 HFS Bug (Re: Msg 66)
     Re: MacDraw upgrade wishes
     KanjiTalk (TM)
     Pro-Tek from First Byte
     Internet reply: interrupt button hang
----------------------------------------------------------------------- 

From: MACINTOUCH (8466)
Subject: RE: Hard Disks
Date: 25-MAY 20:34 MUGS Online
 
to: From: lewis@harvard.UUCP (Harry Lewis)
 
We have been using the SuperMac DataFrame 20 for about a month and it
works as advertised, quickly, quietly, and reliably from our
experience.  It's also relatively inexpensive (under $900), has a
1-year warrantee, appears to be shipping in quantity, and is faster
than a Hard Disk 20.  There are virtually no software utilities
included with the unit.
 
Ric Ford

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

From: MACINTOUCH (8467)
Subject: RE: ThinkTank and Acta
Date: 25-MAY 20:37 MUGS Online
 
In response to questions about ThinkTank and Acta:
 
We found Acta to be a fine alternative to ThinkTank 512, with advantages and
none of the memory management problems.  We have a 4 page review in the June
issue of "MacInTouch."  Symmetry is at 761 E. University Dr., Mesa, AZ  85203;
800-624-2485 (602-844-2199).  It lists for $59.95.
 
Ric Ford

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

From: MACINTOUCH (8468)
Subject: RE: Spelling checkers
Date: 25-MAY 20:39 MUGS Online
 
In regard to spelling checkers, we have found Spellswell from Greene,
Johnson to be among the best, and to be far better than the first
generation of Mac spell checkers.  Among its features are the ability
to add and *delete* words from the dictionary, support for both Word
and MacWrite, and many usability features.  It does not have
interactive spell checking, which MacLightning does.
 
ric ford

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

From: BILLIAM (8471)
Subject: RE: Spelling checkers
Date: 25-MAY 20:49 MUGS Online
 
when you say that Spellswell DOES NOT have interactive spell checking
does that mean that I'd have to look up all the misspelled words by
hand in a dictionary?  Sounds like a HUGE step backwards to me.
 
What's the word on MACLIGHTNING? sounds great but so did The RIGHT
WORD (200,000 word dictionary) and it sucks. Is Maclightning a good
purchase? Any word if they plan to make it compatable with future CD
ROM disks as Borland's Turbo Lightning claims it will be?
 
I've been using Hayden Speller (don't even ask) with WORD and I was
shocked to find that it doesn't reconize optional hyphens. It treats
them as separate words which is useless. I have to spellcheck first
and then go back and divide the words with hyphens.
 
Would Maclightning be a breath of fresh air for me?
 
thanks Billiam

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

From: MACINTOUCH (8470)
Subject: System 3.2
Date: 25-MAY 20:49 Mousing Around
 
Just got it thru the grapevine:  the long-awaited System 3.2 had a fatal bug
discovered at the last minute this week, and won't be shipped.  Look for 3.3 or
higher...  (groan)
 
ric

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

From: BMUG (8475)
Subject: Re: Network Hardware
Date: 26-MAY 03:36 MUGS Online
 
To: {allegra,ihnp4,seismo}!uwvax!uwmacc!dan Re: Network Hardware
 
A couple of suggestions:
 
a) Look into having more than one printer available to any given station (maybe
have 10 macs to 2 printers networked), so that someone can print to the 2d
printer if the 1st is occupied.  Maybe have more networked... I've heard of up
to 40 macs on a net, no problem.
 
b) Have LaserWriters (at least one) available.    The ability to get thesis-
quality print is really worth it, and it is much quieter... people will even pay for printing (if needed to subsidize cost)
 
c) Look into using PhoneNET(tm), an AppleTalk-compatible network that
is about half the cost of AppleTalk per node and uses ordinary phone
cable (much cheaper) , and more reliable!  Call Farallon Computing
Berkeley at (415 ) 849-2331.
 
-- Raines Cohen BMUG

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

From: DDUNHAM (8490)
Subject: RE: Usenet Mac Digest V2 #41 (Re: Msg 8454)
Date: 26-MAY 22:17 MUGS Online
 
To: lewis@harvard.UUCP I'm using a LoDOWN 20, and I'm very happy with
it.  Only problem is its noise.  It has no fan but it does have a very
audible motor.  But it is very fast.  (Boots in 9s, return to Finder
in 4s.)  Inexpensive (for Mac hard disks), but no software provided.
 
To: ephraim@wang.UUCP Using a dynamic reserved slot doesn't sound like
a real problem...at least if you reboot (and don't use them all).  And
if you only have 1 SCSI device, should be no trouble there, either.
Besides, I thought the dynamic slots were 28-31, not the ones above...
 
To: bart@reed.UUCP I've had troubles with the MacPaint clipboard, too.
The cure seems to be to use DiskInfo to Set Default to the System
Folder.  You may have to do a Copy before you can import anything from
a DA.
 
To: werner@ut-ngp.UUCP The advantage of the current zoom box is that desk
accessories can use it and thuse keep a standard interface.  DAs are _not_
passed any other sort of title bar events.  Also, you don't have to remember
which modifier key makes the box zoom.
 
To: robb@mentor.UUCP I know MacInTouch has already answered you, but I don't
think they said that Acta includes a program to convert your existing TT512
documents.  Acta takes approx. 32K of memory as a desk accessory, so it's
compatible with everything.

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

From: DDUNHAM (8491)
Subject: RE: Usenet Mac Digest V2 #41 (Re: Msg 8468)
Date: 26-MAY 22:17 MUGS Online
 
You might be interested to know that Michael Greene is planning to add
the ability to proof Acta documents.  It's not at the top of his list,
however, so if you'd find it useful, give him a call.  BTW, Spellswell
is possibly the only speller to know about typographer's apostrophes
(opt-}).

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

From: PEABO (8493)
Subject: RE: Usenet Mac Digest V2 #41 (Re: Msg 8490)
Date: 26-MAY 23:45 MUGS Online
 
Concerning the advantage of zoom boxes ... I dislike the zoom box
greatly, since Microsoft's 'double-click the title bar' seems to be a
vastly superior way of doing it.  Are you saying that there is a good
technical reason why Apple has picked up the DRI Zoom Box instead of
the Microsoft technique?
 
peter

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

From: RICKLEPAGE (8502)
Subject: prolog II
Date: 27-MAY 15:39 Developer's Corner
 
Well, all you AI freaks out there, we got a press release from our buddies at
Expertelligence, stating that they are going to sell PrologIA's Prolog II for
the Mac in this country.
 
I quote :
 
"Exper-Prolog II [had to get that in somewhere, I guess] is the
product developed from the work of the _father_ of Prolog, Alain
Colmerauer.  This Prolog interpreter includes real numbers, string
manipulations, advanced process control, and allows you to load,
extend and modify interactively Prolog programs.  All this while
taking advantage of all of the facilities of the Mac interface."
 
they say it will be available in June for $495, though if you buy now (before
people scream about bugs, I guess), it is yours for $395.  existing
Expertelligence customers can get Exper-Prolog for $295 during the month of
June.
 
no comments about copy-protection or anything else.  anyone have any experience
with the developer, PrologIA??  they are a French outfit.
 
Rick LePage MacInTouch

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

From: PEABO (8522)
Subject: RE: LINT again
Date: 27-MAY 23:16 MUGS Online
 
From: CML5A9%IRISHMVS.BITNET@WISCVM.WISC.EDU
Subject: LINT again
 
I would also like to see LINT for the Mac.  ANY lint.
Of ANY kind.  I remember a while back someone was talking
about something along these lines and pointed out something
interesting.  In order to have a really USABLE lint for
the mac, someone also has to take the time to type in
all of the function definitions with types as per Inside
Mac into header files.  That is, lint without the toolbox
 
--------------
 
Don't type it in ... get a copy of Tech Note 45 instead.  It's all there in
machine readible form.
 
peter

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

From: MACINTOUCH (8519)
Subject: Re: Alternative keyboards for Macintosh
Date: 27-MAY 22:56 MUGS Online
 
to: Pat Mcconnel <mcconnel@nrl-b43.ARPA> Subject: Alternative keyboards for
Macintosh
 
I think you'll need more than software.  The only hardware that's applicable
that I've seen is from Computer Parts Mart in San Jose, CA.  408-993-9011. For
$99 (90 day guarantee) they have an adaptor that's supposed to make IBM PC
keyboards compatible with the Mac.  I know nothing about the company and I'd
love to hear from anyone who tries this.  I don't like the "feel" of the Mac
keyboard at all.
 
Ric Ford

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

From: CHESLEY (0000)
Subject: RE: PackIt II problems
Date: 27-MAY 21:43 MAIL
 
If anyone else sees any bugs in PackIt II, PLEASE let me
know, either at the address given in the program or at my BBS (415-562-2491).
I can't fix 'em unless I know about 'em -- this is the first I've heard of
a problem with PackIt running under Switcher. I did test it under Switcher,
but obviously missed something. I'll fix it for PackIt III.

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

From: BRECHER (8570)
Subject: Reply to Pseudo-DMA Question
Date: 30-MAY 03:13 MUGS Online
 
To: ronr@sphinx.UChicago.UUCP (Ron Rusnak)
Subject: Mac+ SCSI pseudo dma question
 
A9 set asserts DACK; reset asserts CS.  EOP (an input to the 5380) is probably
connected to some uninteresting place, e.g., a pull-up (did I say that right?
It's in my Berlitz Hardware-English Dictionary, Programmer's Pocket Edition).

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

From: MACINTOUCH (8575)
Subject: Re: Laserwriter + MS Word Problem on Print Merge (with fix)
Date: 30-MAY 09:48 MUGS Online
 
to: derek@gucis.OZ (Derek Austin) 
 
LaserWriter/Laser Prep 3.0 and higher eliminate this problem.  They no longer
print out instruction sheets for each page printed manually.
 
Ric Ford

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

From: DOCGRINNELL (8568)
Subject: Lisa ROMSwitch
Date: 30-MAY 01:17 Hardware & Peripherals
 
I have some news for Lisa/XL owners!  (y'all remember them?)  I was
over at a company called Orphan Technology the other day, and they
were testing ( successfully) a device called the ROMSwitch that allows
Lisa/XL owners to switch the original Lisa roms andd the new Macworks
screen fix by one simple toggle switch.  This gives the user a choice
between the *original* roms and the *official* new roms!  What do you
all think?
 
-=> Doc

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

From: LOGICHACK (8569)
Subject: RE: List Manager (Re: Msg 8396)
Date: 30-MAY 02:31 Programming
 
I have used the string displayed functions of the List Manager.  To
add data to a list, I used _LSetCell.  The List Manager allocates its
own storage for the data.  I implemented a _SFGetFile type list by
looping thru each row and setting the contents individually.  Seems to
work ok.  One annoyance I have is that upon a _LActivate for a
deactivate event, the list's scroll bars, if any, are 'hidden' I would
rather have the controls hilited as inactive I have worked around this
but it is quite a hack.  One last thing, check the sample source code
(in Pascal) that Apple distributed with the List Manager information;
it is pretty straightforward.
 
Paul

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

From: DWB (46)
Subject: DIZero problems
Date:   30-MAY 05:55 Developers' Corner
 
A short while ago somebody mention having problems with Mount.  I have
just finished writing/debugging a SCSI driver for a new hard disk and
have now encountered problems with DIZero.  It goes for a while,
writes lot's of things to the hard disk, does a PBStatus and then goes
away as the status routine exits (via RTS) Anybody have any ideas?
 
Just for the record the code being used to load, etc. the drive is:
OpenDriver("\p.SCSI", &rn);
dn = GetFreeDrive();            ; is returning 32, the first SCSI drive
DetachResource(GetResource('DRVR', "\.pSCSI"));
 
AddDrive(rn, dn, dqel);
 
DIZero(dn, "\p.HardDisk");
 
Any help or suggestions would be appreciated.
 
David

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

From: BRECHER (74)
Subject: RE: DIZero problems (Re: Msg 46)
Date:   31-MAY 07:41 Developers' Corner
 
Unless you're doing it somewhere else not shown in your message, you
need to put some things in the drive queue element, such as the drive
size.
 
GetResource wants an ID, not a name, as the second argument.
 
32 is a peculiar drive number, though I suppose it would do no harm.  Don't
confuse unit numbers (w/r the unit table) with drive numbers.
 
Ask the Mac Tech Support Group at Apple for a copy of the SCSI Developer's
Package.  It's extrememly useful.

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

From: DWB (85)
Subject: RE: DIZero problems (Re: Msg 84)
Date:   1-JUN-05:25: Developers' Corner
 
Hurray!  I got it working.  Unfortunately I don't remember what all I
had to fix.  Some things I do remember are that you must make sure the
drive queue element is completely filled out.  I had copied code that
did a NewPtrSysClear, which LightSpeed doesn't support and had turned
it into SetZone and NewPtr.  End result being I had a block of memory
with garbage in it.  Anyway I now have everything working except for
the auto boot part.  I'll wait to get the software from Levco for that
one.
 
If there is enough interest I will post the set of programs that I
wound up with.  Included are a program to execute miscellaneous SCSI
commands and test the driver, the driver itself, and a program to
format the drive as a Mac Volume or mount it depending on a compile
time flag.
 
David

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

From: ROWLAND (8594)
Subject: system enhancement?
Date: 30-MAY 21:25 Mousing Around
 
In a number of applications I would find it extremely useful to be
able to enter a file name into an Open File dialog box by pasting from
the clipboard. For example in an editor in which I had a file which
was a sort of directory I could double click a file name (contained in
the text), copy, select openfile, and then paste to open that file
(with Tempo of course the operation could be made even simpler).
 
Is there some reason this could (or should) not be made a general feature in a
future ROM ? Is there any way to patch it into the current system ? Is there an
even easier way to do what I want ?
 
Mike Burns

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

From: PEABO (8598)
Subject: RE: system enhancement? (Re: Msg 8594)
Date: 30-MAY 22:22 Mousing Around
 
The problem is that the dialogs you are referring to are "Modal
Dialogs", which means that the system is in a mode where it will
accept only a limited number of actions.  Cut and Paste are not
included in the eligible operations.
 
Contrast this with Modeless Dialogs, such as the Find/Change dialog in
MacWrite.  Here the dialog can coexist with other activities, such as
setting the cursor position and making ancillary edits in the text
near where the strings are found, and cutting and pasting.
 
Apple says in their User Interface Guidelines that "Modes are bad",
and you have discovered a good illustration of why: you want to use
tools that you are accustomed to using in other situations within a
mode where they are not supported.
 
Unfortunately, Modal Dialogs are easy to do and people use them a lot,
so we are stuck with the situation unless someone does some major
surgery to the Mac User Interface.  One workaround you alluded to was
the use of Tempo to pretend to type the entry in.  Another possibility
would be to notice a command-V keyboard entry while in a dialog that
has a TextEdit box and presume that the user wants to paste from the
clipboard.  This would require some reprogramming of applications, and
therefore would only work in limited cases.
 
peter

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

From: LOCKE (8601)
Subject: Health software development
Date: 31-MAY 02:19 Developer's Corner
 
The Center for Clinical Computing at Harvard Medical School is
interested in reviewing software targeted for the health care field
for possible publishing/distribution.  If you have a software product
in development or planned, contact Steven Locke, M.D., Beth Israel
Hospital, 330 Brookline Avenue, Boston, MA 02215, 617-735-4664.

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

From: MARSHG (66)
Subject: 3.1.1 HFS Bug
Date:   31-MAY 00:11 Programming Techniques
 
I just discovered a bug in System 3.1.1 that will cause some HFS calls to
randomly fail.  I say "some" because the bug is in a piece of common code that
is probably shared among most of the calls. GetWDInfo is definetly affected
(that's how I found the bug).
 
The symptoms of my problem were that regardless of what WDRefNum I
would pass in to GetWDInfo, I would usually get the right answer but
sometimes (fortunately reproducably) I would get a dirid of 2.  The
problem lies in the conversion of the WDRefNum into the offset on the
working directory table.  The ROM loads the WDRefNum into the lower
half of D1 and then does a DIVU $10,D1.  When the upper half of D1
(never set by the system) was non-zero and big enough, the DIVU failed
because the result was larger than 16 bits.  My workaround was to
change my HFS trap routine to always clear D1 before doing the trap.
 
Marsh

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

From: BRECHER (75)
Subject: RE: 3.1.1 HFS Bug (Re: Msg 66)
Date:   31-MAY 07:41 Programming Techniques
 
Sounds like you meant a ROM bug rather than a System 3.1.1 bug.  At any rate,
System 3.2 patches GetWDInfo to pre-clear the high word of D1.

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

From: BMUG (8618)
Subject: Re: MacDraw upgrade wishes
Date: 31-MAY 20:15 MUGS Online
 
To: borton@sdcc3.UUCP
Subject: Re: MacDraw upgrade wishes

a) the SIZER and ADJACENCY DA's, from Johnson & Johnson, Design/Build,
677 Grove, Newton, MA 02162, costs $69 (and your MacDraw) and gives you:
SIZER:  You can specify the dimensions of any object to the hundredth of an
inch or 1/100 cm or architectural notation.
ADJACENCY:  Select any two objects, makes them touch but not overlap.
 
Used by the Wall Street Journal and New York Times art departments, and
by BMUG.  Really useful, and a terrific deal.  Now Mac+ & Switcher-compatible.
 
-- Raines Cohen, BMUG Operations Manager

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

From: FRIED (8610)
Subject: KanjiTalk (TM)
Date: 31-MAY 13:45 Mousing Around
 
Hot off the wires <grin>:
 
Received: from UTAH-CS.ARPA by MIT-ZERMATT.ARPA via INTERNET with SMTP
id 39463; 30 May 86 17:15:58-EDT Received: by utah-cs.ARPA
(5.31/4.40.2) id AA20102; Fri, 30 May 86 15:16:33 MDT Received: by
utah-orion.ARPA (5.31/4.40.2) id AA21453; Fri, 30 May 86 15:16:28 MDT
Date: Fri, 30 May 86 15:16:28 MDT Message-Id:
<8605302116.AA21453@utah-orion.ARPA>
 
Macintosh Plus with KanjiTalk(TM) Introduced
 
Tokyo, Japan.  May 19,1986.  Apple Computer Japan today introduced the Japanese
Macintosh Plus personal computer complete with a state-of-the art Japanese
operating system.
 
Two years in development, the KanjiTalk(TM) operating system gives Japanese
users access to the three traditional Japanese alphabets plus the English
alphabet.
 
"KanjiTalk is unequaled in the Japanese marketplace because of Macintosh's
ease-of-use, advanced features and powerful graphics," according to Alexander
van Eyck, General Manager, Apple Japan.
 
"KanjiTalk offers immediate access to a powerful library of Macintosh software
which can be easily translated (localized) from its original language to
Japanese. Perhaps more important, KanjiTalk also provides a development
environment which allows local software developers to create Japanese software
products," van Eyck adds.
 
Within a month, Apple Japan will begin to ship fully localized
versions of Apple software, including MacPaint, MacDraw and
MacProject.
 
EgWord, a powerful Japanese word processing program from ErgoSoft, is available
at launch.
 
A localized version of Microsoft Excel, the powerful integrated spreadsheet
program, will be available in July.
 
Historically, the Japanese personal computer market has been driven by word
processing applications. Now the emphasis is beginning to shift towards
spreadsheets and data bases, according to van Eyck.
 
"From a timing point of view, we are announcing the right product at the right
time.  The market is demanding powerful, diverse software solutions. Thanks to
KanjiTalk, third-party developers can localize and ship those solutions in a
fraction of the normal time. We have a unique window of opportunity," van Eyck
adds.
 
The suggested retail price for the Macintosh Plus with KanjiTalk is
648,000 yen.  Apple has no immediate plans to market the KanjiTalk
Macintosh outside Japan.
 
MACINTOSH IN JAPAN
 
The 128k Macintosh (English) was first introduced in Japan in April, 1984.  In
May, 1985, Apple Japan introduced a katakana (phonetic) keyboard for the 512k
Macintosh.
 
EgWord, a powerful word processing package was introduced in
September, 1985, by ErgoSoft.
 
Several other third parties have also introduced various other versions of
Japanese Macintosh software.
 
Apple Japan is offering a variety of upgrade programs for existing Macintosh
owners.
 
KANJITALK FEATURES
 
Japanese input/output using existing software
 
 
Since "KanjiTalk" supports Japanese language input and output at an OS level,
selected software developed in the United States and Europe can be easily
localized in Japanese.  In addition, new software which follows Apple's
development guidelines can also be localized for Japan.
 
KanjiTalk application tools also facilitate the conversion of application
software from English into Japanese.
 
Japanese development tools
 
Japanese software developers can easily access and utilize KanjiTalk
development tools to develop unique Japanese software products.
 
Accurate sentence-level conversion
 
Through phonetic input, the Macintosh can execute kanji conversions, making it
possible to convert up to 40 characters at once.
 
Three separate dictionaries
 
KanjiTalk features three separate dictionaries with different
conversion levels--a basic dictionary, special dictionary and users'
dictionary.  The basic dictionary maintains more than 35,000 words.
 
KanjiTalk desk accessories
 
Through a unique design, the KanjiTalk utilities are built into the Macintosh's
Desk Accessories so any function can be used any time without turning off (or
rebooting) the software.
 
24 X 24 dot printing
 
In addition to the 16 x 16 dot font, KanjiTalk is equipped with a 24 x 24 dot
font for producing enhanced quality output.  KanjiTalk comes with print drivers
for  Apple's Imagewriter II.

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

From: MACINTOUCH (81)
Subject: Pro-Tek from First Byte
Date:   31-MAY 12:37 Developers' Corner
 
We are starting to see information about First Byte's new copy protection
scheme which is quite intriguing.  If it does what they claim it does, it
will have a large impact on copy protection.  We find some of the claims
hard to swallow.
 
It is called Pro-Tek and is being sold by First Byte, Inc., 2845
Temple Ave., Long Beach, CA  90806.  Phone numbers are 800-523-8070,
800-624-2692 (in CA.) or 213-595-7006.  A starter kit is $400 and costs
range from $.15 to .40 per disk.
 
Claims:
- uses normal diskettes
- hasn't been copied in months of testing
- compatible with Mac/XL/MacPlus, 800K and 400K disks, and HFS
- doesn't affect performance or use up memory (uh huh)
- no worms (thank God)
- software-based but uncopyable with nibble/bit copiers
- contains "debugger hunters"
- application is "encrypted" with multiple layers of checking and encryption
- will be available for most personal computers
- uses hardware board for adding protection during duplication process
 
The good side for users is the claim that it supports flexible options for
hard disk installations and the ability to check key disks a certain
percentage of the time (instead of at every launch).
 
I'd be interested in hearing more about this product from people who dig
further.
 
Ric Ford, "MacInTouch" newsletter

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

From: BRECHER (8628)
Subject: Internet reply: interrupt button hang
Date: 1-JUN-10:18: MUGS Online
 
To: jer@peora.UUCP (J. Eric Roskos)
Subject: Re: Delphi Mac Digest V2 #20
 
> ... If carrier is
> present, there should be no problem... it's only if a modem is on and
> *negating* the CTS signal (not asserting it).  When no modem is connected,
> or when it's off, the pin "floats" to CTS.  When a modem is on but no
> carrier is present, it is driven to /CTS, causing the "hardware flow
> control" in the serial driver to wait for CTS (which it is supposed to do;
> it is not a "ROM bug" at all).
 
Most Mac serial cables do not connect carrierdetect to HSKi; mine
doesn't -- it connects CTS.  When my modem's on and I'm not running a
comm program, pressing the interrupt button causes a hang.  When I'm
running a comm program, regardless of whether or not I'm online, or
when the modem's off, I get the minidebugger prompt box.
 
When the interrupt button is pressed, ROM code looks at the system global
SPconfig, which contains two 4-bit fields, one for each serial port, indicating
whether the port's in use and if so by what (AppleTalk, asynch, ext. clk.). If
the shift key is held down while the interrupt button is pressed, the printer
port's in-use bits are examined; otherwise, the modem port's. If the port is
free, data is transmitted -- as I press the interrupt button, the SD/RD LEDs on
my Hayes flicker (RD due to its being in command mode, defaulting to echo) and
then the Mac is hung.  If the port is in use the prompt box is displayed.
 
Therefore, my current theory is an extension of my original one (i.e., that CTS
[HSKi] asserted is the hang condition).  If the modem's off (the theory goes)
the code sees lack of CTS and doesn't try to transmit, and puts up the local
prompt.  Likewise if the modem's on  but the port's marked as in use.  But if
the modem's on (CTS asserted) and the the port seems to be free, the code
transmits to the remote and then hangs waiting for a valid resume-execution
debugger command input from the serial port.
 
Regardless, I'm somewhat curious about what you get on the other end
of the serial line if all goes well.  The minidebugger could be a
MacDB nub, or maybe it's just a dumb-terminal interface that outputs a
">" prompt and takes the same commands from the remote as it does when
in local mode.

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

End of Delphi Mac Digest
************************

ephraim@wang.UUCP (pri=8 Ephraim Vishniac x76659 ms1459) (06/03/86)

> From: DDUNHAM (8490)
> Subject: RE: Usenet Mac Digest V2 #41 (Re: Msg 8454)
>  
> To: ephraim@wang.UUCP Using a dynamic reserved slot doesn't sound like
> a real problem...at least if you reboot (and don't use them all).  And
> if you only have 1 SCSI device, should be no trouble there, either.
> Besides, I thought the dynamic slots were 28-31, not the ones above...

Using a dynamic reserved slot *is* a very dangerous thing.  Consider this:
My internal hard disk boots, and selects a dynamic slot which is open.
While running off the disk, I run MacPaint with ClickArt Effects installed.
I choose ClickArt Effects, which is installed in MacPaint as DRVR 31 (a
reserved slot).  The DA *replaces* the disk driver as driver 31 and the
Mac hangs....  When you open a driver with the same number as an open driver,
it's curtains for the older one.  ClickArt Effects comes with its own installer,
which installs only into MacPaint and doesn't give a hoot for the rules.
You can fix it by removing it with the Font/DA Mover (use option-open to
see all applications), then re-installing it.  FDAM follows the rules, so
you're safe.  But most people wouldn't recognize the problem.

Yes, 28-31 are the dynamic slots.  32-39 are the slots for SCSI devices
zero through seven.  But when you're installing DA's, how do you know 
what SCSI devices the user has?  Especially if some of them are used
intermittently, like tape backup devices?