shulman@topaz.RUTGERS.EDU (Jeff Shulman) (02/02/86)
Delphi Digest Sunday, 2 Feb 1986 Volume 2 : Issue 4
Today's Topics:
IM error
RE: MicahDrive (Re: Msg 5345)
RE: MicahDrive (Re: Msg 5346)
Where's CurDirStore in HFS
How large A5 globals area?
RE: How large A5 globals area? (Re: Msg 5370)
Mini-DIN-8 connectors
TK!Solver
RE: TK!Solver (Re: Msg 5401)
RE: TK!Solver (Re: Msg 5401)
MacTutor
RE: Making my MacSCSI disk the startup disk
RE: Usenet Digest Volume 2 Issue 8 (Re: Msg 5421)
RE: SCSI HD20
Future Macintoshes
MacEqn
Megamax DA/Standard File Bug
RE: Megamax DA/Standard File Bug (Re: Msg 5494)
----------------------------------------------------------------------
From: RAYSANDERS (5351)
Subject: IM error
Date: 26-JAN 18:35 Mousing Around
Those of you who have IM (all three version unfortunately) may want to make a
note in the TextEdit appendix, The TE global variable TEScrpLength is a [word]
and NOT a [long] as documented. This is correctly reported in the toolbox
equates and the May 85 low memory map.
Ray (took an hour to find this one) Sanders
------------------------------
From: BRECHER (5352)
Subject: RE: MicahDrive (Re: Msg 5345)
Date: 26-JAN 18:50 Hardware & Peripherals
Software is my game, and I run the MicahDrive in "development
configuration," i.e., with everything (Mac logic board, MicahDrive
unit) spread out on a table for easy access to EPROM sockets.
However, the following is my understanding of the mechanical mounting
situation. The MicahDrive does not attach to the Sony enclosure; it
sits in back of it, on the platform which covers the Mac's logic
board, and attaches to the vertical rear "lip" of that platform. The
same unit is installable in (and has been installed in) 512K Macs and
Mac Plusses. The Micah PC board has connectors for Apple ROMs which
are position-adjustable via a slide gizmo to accommodate the different
relative spacing of the ROM sockets between the 512K and the Plus. (I
really don't know how this works -- it seems like magic to me, but
then I've always believed that electrons are a hoax perpetrated by
hardware engineers. For hardware mounting details, please contact
Micah).
Backup software is coming "real soon now" from Micah as well as
independent software vendors. I think Micah is planning -- or at
least hoping -- to have their backup software available with first
drive shipments, but given the schedule I'd have to consider the
program "beta" at that point regardless of whether it's labelled as
such. There is no reason for backup to be specific to the medium
being backed up, i.e., a backup program should be able to back up (or
restore) any HFS or (if it has MFS capability) MFS volume. In other
words, the backup software should use standard Mac I/O requests to
access the hard disk, even if it is specific to the backup medium
(tape, floppy, etc.).
I think Micah is planning a tape unit (but I'm not sure when it'll be
available) .
My preferred backup system would be streaming tape. Or, if I had
bucks to burn, I might go for a 20MB removeable-cartridge SCSI
Bernoulli box ($3095) and use it for additional operating storage as
well as internal hard disk backup -- you could back up just by
dragging a Finder icon.
------------------------------
From: BRECHER (5353)
Subject: RE: MicahDrive (Re: Msg 5346)
Date: 26-JAN 18:53 Hardware & Peripherals
No, Micah will not be providing a recover-trashed-directories utility. Fedit
itself is being enhanced for complete HFS compatibility and automatic recovery
in the upcoming version 4.0. However, the MicahDrive, like many hard disks (
with the exception of the Apple HD20), does not use sector tags and such tags
are required for automatic recovery. (Implementation of tags requires a custom
controller, which is too expensive for companies which do not have Apple's
magnitude of sales volume.) So if directories get trashed your only recovery
possibility (other than restoring from backup!) is to try to find and
reconstruct data on a block-by-block basis using Fedit -- not a pleasant
prospect. As has been said once or twice before: always back up your data!
------------------------------
From: MARSHG (5362)
Subject: Where's CurDirStore in HFS
Date: 27-JAN 00:01 Programming
The "New, Improved Standard File Package" document talks about a place in low
memory called CurDirStore where, if HFS is running, the current directory is
placed. Unfortunately none of the include files in the HFSPP define that
variable or anything that looks like it. Anybody know where CurDirStore is?
Marsh Gosnell
------------------------------
From: SBOAG (5370)
Subject: How large A5 globals area?
Date: 27-JAN 02:28 Programming
What happens when you go over the default A5-globals area in MDS? I
just added a bunch of blank subroutines and a bunch of A5 globals to
my code and I seem to be getting errors for no reason. How large can
the A5 globals area be anyway? Any help on this would be appreciated.
thanx-- Scott Boag
------------------------------
From: MACLAIRD (5396)
Subject: RE: How large A5 globals area? (Re: Msg 5370)
Date: 28-JAN 21:17 Programming
Scott, Referring to the Bible (latest IM) p. II-19, you'll see that the Globals
end where the Stack begins. This is all pretty automatic and based on what the
application load module says.
However, you might be doing what I did recently, which is assuming that those
globals were really globals. MDS linker seems to assign a separate location to
each global value in each module (i.e. each DS). In order to actually pass the
value among a group of modules, you have to allocate it once and XREF it the
other times. I think this is an error but that's the nutshell of it.
Now, I remember tracing this problem by putting a .Verbose option in
the MDS assemblys, and /Verbose and [ and ( in the linker directives,
after which I inspected (edited) the resulting text file, which really
does show all the nitty gritty. I hope you have a Hard disk (or maybe
RAM disk) to send it to though, otherwise you might go out for a beer.
I hope this helps!
-Laird
------------------------------
From: MOUSEKETEER (5395)
Subject: Mini-DIN-8 connectors
Date: 28-JAN 21:16 Hardware & Peripherals
Apple has recently released names of two sources for cables using the
Mini-DIN-8 connectors, one in CA and one in CT. After talking with
both today, the following:
The CT company is tooling up and making various cable assemblies,
though at present they are concentrating on 8<>8 cables (hooking Mac+
to IWII and Personal Modem, etc.). They claim they will also provide
8<>"DB9", and 8<>DB25 cables. When asked about individual connectors,
they said it would be very difficult for users to make their own
cables, as the pins are very closely spaced. Their prices depend on
quantity, and they are primarily set up to service businesses. The CA
company has not started making cables as yet, but is providing the
connectors to other companies (they will, I was told, start making
cables soon).
They will sell the single connectors in an Assembly Model, but only in larger
quantities. A certain well-loved Mac User Group has samples of the connectors,
and agree that they are a bear to solder up with the wire.
Is there interest in a group buy of just connectors, cables, or ?? I had
thought of co-ordinating such, but would like to know what interest, if any,
there is in such a move. The connectors alone go for from 3.50 at 100 quan or
so to around .75 in MUCH larger quantity (as in 1.50 at 1000 quantity). Alf.
------------------------------
From: DWB (5401)
Subject: TK!Solver
Date: 29-JAN 06:19 Business Mac
Went to my software dealer to find a copy of TK!Solver today and found out that
(according to him, and I havn't caught him lying yet!) Software Arts the people
who wrote TK!Solver were bought out by Lotus who for all intents and purposes
disbanded the company and discontinued TK!Solver. The question this brings up
is if anybody has a copy of TK!Solver they would be willing to sell.
Another interesting question this brings up is, since the original
owner of the copyright is defunct and the apparent current owner of
the copyright (Lotus) has no current, or apparently future, interests
in the product, what are the legal ramifications of using "hot" copies
of the product? I wouldn't want to be caught posting it in a public
place, or selling copies of it, but it doesn't seem there would be
serious objections to copying somebody else's legitimate (or even
illegitmate) version of the software.
Comments anyone?
David (Note: I take copyright protection of software very seriously,
I'm a developer. Acquisition of new products tends to be a two step
process I attempt to get a demonstration of a product to find out if
it will do what I need and it claims, and then I buy it if it will and
does. )
------------------------------
From: PEABO (5403)
Subject: RE: TK!Solver (Re: Msg 5401)
Date: 29-JAN 15:29 Business Mac
Well, the copyright is still in effect. Lotus might choose to prohibit any and
all distribution of TK!Solver because, for example, they may have a competing
product at some point. If I recall, the guys who wrote TK still work for Lotus
in some department for special applications, so they may indeed have something
in the works.
peter
------------------------------
From: RICKLEPAGE (5417)
Subject: RE: TK!Solver (Re: Msg 5401)
Date: 30-JAN 15:54 Business Mac
David: Lotus did indeed buy Software Arts and TK!Solver. After a few
months of floundering, Lotus turned around and sold TK!Solver to a
company in Ohio (i think Ohio). The company, whose name escapes me
right now, has stated that they will continue to sell and support the
product. One interestingaside to this is that Lotus had another
company in Boston handling support for the product. If I can dig up
the references, i will send em along.
Rick LePage
MacInTouch
------------------------------
From: BRECHER (5419)
Subject: MacTutor
Date: 30-JAN 16:04 Developer's Corner
Laura Smith, manager of MacTutor Magazine (and wife of
publisher/editor Dave Smith) called me this morning to ask about
submission of my column. As usual, I asked what the absolute latest
deadline was. She said, in an unsteady voice that suggested tearful
determination, that they'd had a "setback": their office was burgled
last night and all their equipment, software, submissions --
everything -- was taken. (For those who don't know, MacTutor is a
Macintosh programming journal.)
Laura said the next issue of the magazine might be late by a week or two.
The shock and bravery in her voice really affected me; she is one of
those rare people whose purity of spirit and goodness strike one anew
each time one meets her. She and Dave, who quit his job last year to
spend full time nurturing what was then a tiny magazine, have worked
night and day and have succeeded in making MacTutor an important
resource of the Mac community. Their dedication has been and
continues to be a labor of love for the potential of the Macintosh and
an instance of the best kind of entrepreneurial spirit.
I'm sure Laura and Dave and those who work with them will overcome
this setback. If you have a subscription, I'd suggest that now would
be a good time to extend it. If you've been thinking of subscribing,
now would be a good time to send in your subscription. It is what I
know of Laura and Dave and what I heard in Laura's voice this morning,
rather than my association as columnist, that leads me to make this
suggestion. MacTutor, $30/yr, P.O Box 846, Placentia, CA, 92670.
------------------------------
From: BRECHER (5427)
Subject: RE: Making my MacSCSI disk the startup disk
Date: 30-JAN 23:32 Mousing Around
>Now that I have my MacSCSI disk running with HFS, I need to know what
>I have to do to make my hard disk the startup disk in an automatic way.
You need to write a program that will be the startup program on your
boot floppy. This program should:
See if there's a System and Finder on the hard disk (if not, exit);
Open a Working Directory for the system directory (DirID is contained
in the longword at offset 90 decimal from the parameter block returned
by _GetHVolInfo) and put the WDRefNum into system global BootDrive;
Detach your code resource, close all resource files, eject the floppy;
Switch to the hard disk System by calling _InitResources and _InitFonts;
Launch Finder.
------------------------------
From: PEABO (5428)
Subject: RE: Usenet Digest Volume 2 Issue 8 (Re: Msg 5421)
Date: 31-JAN 00:20 Mousing Around
Re: 800K drives faster?
Presumably also the information is packed on both sides of the disk as
alternating tracks, so the same about of data will fit in a smaller
range of the disk, so that seek time is reduced.
Re: MS-WORD file format
Last time I looked at an MS-WORD file, its data fork seemed to be composed
mainly of the text of the file, with incidental control characters here and
there, and the Resource Fork seemed to have most of the paragraph and division
format info. If the goal is to retrieve the text, it would be worth trying to
read the data fork as a stream of characters.
re: European modem standards
In Europe, most modems conform to CCITT signalling standards, which
are not the same as Bell. However, the 2400 bps modems obey CCITT at
the 2400 bps speed, and one at least (Hayes) offers a CCITT 1200 mode
in addition to Bell 212A mode.
re: Turbo Pascal
Borland exhibited at the Macworld Expo and said that Turbo Pascal for the Mac
would be shipping Real Soon Now.
------------------------------
From: RICFORD (5434)
Subject: RE: SCSI HD20
Date: 31-JAN 09:35 Mousing Around
We understand that Apple (Scully?) claimed that they _wouldn't_ be coming out
with an SCSI version of the HD20, instead leaving that segment to the third
parties.
From our testing of disk caches (and Jan's), it seems that increasing
amounts of cache, up to 256K or more, continue to improve performance,
so it might be worth increasing it well over 64K. Mark Zimmer of
Fractal Software was claiming at the Expo that Apple's cache wasn't
all that well designed, and that his Power Cache (distributed by
MacNifty) was more efficient. I'd like to see a face off between
them.
Ric Ford, MacInTouch newsletter
------------------------------
From: PEABO (5445)
Subject: Future Macintoshes
Date: 31-JAN 21:22 Macintosh In Fact
I see from the description of the new ROM calls HideDItem and ShowDItem that
future Macintoshes will not soon have screens more than 8192 pixels tall or
wide.
;-)
peter
------------------------------
From: GAIGE (5492)
Subject: MacEqn
Date: 1-FEB-23:47: User Supported Software
Does anyone know where I can get ahold of a program named MacEqn (in the
literature the E is a sigma. I desperately need this for some scientists at
work.
Thanks,
Gaige
------------------------------
From: MARSHG (5494)
Subject: Megamax DA/Standard File Bug
Date: 2-FEB-01:55: Programming
I just got finished tracking down what I believe to be a bug in the way Megamax
references globals in a DA. For those of you who don't know, Megamax allocates
a block on the first open call, stuffs the handle in dcltstorage, and also
stuffs the handle into A4. All user code references the globals off of A4.
Here's the bug. If you have dNeedTime set and use the Standard File routines,
you will occasionally crash. Doing the same exact thing again the next time
will work just fine. What happens is that when you're in the SF routines, they
call systemevent which (to my suprise) calls your DA's control routine with a
accrun event. The way megamax is written, when you come in to the control
routine, you lock the globals, do the user control routine, then unlock the
globals. What causes the crash is that once or twice in a blue moon, the block
with the globals gets moved while it's unlocked. All is well until you make
your first access to a global variable after returning from your SF call and
BOOM. Your A4 still points to the old location and you "interesting" things
happen.
The workaround seems to be to not use dNeedTime and the SF routines. If you
don't have dNeedTime set, the control routine isn't called. Unfortunately, I
can't think of a way for Megamax to do it right other than either never
unlocking the global segment.
Marsh
------------------------------
From: BRECHER (5497)
Subject: RE: Megamax DA/Standard File Bug (Re: Msg 5494)
Date: 2-FEB-07:39: Programming
A way to do it right would be to check, on entry, as to whether the
globals are locked. If so, set a flag (which could be in the
globals). On exit, if the flag is set, don't unlock.
------------------------------
End of Delphi Digest
********************