INFO-MAC@SUMEX-AIM.STANFORD.EDU (Moderator Dwayne Virnau...) (05/09/87)
INFO-MAC Digest Saturday, 9 May 1987 Volume 5 : Issue 87
Today's Topics:
Help wanted: ResErrProc hook.
Multiple stacks and messing with HeapEnd
MacTerminal
Direct connection of Laserwriter to Mac
ScrapSaver 1.2
small mistake in file-name in INFO-MAC-5.084
Sneaky Lightspeed Pascal "Feature"
File Fixer .05
Piezo fan for the Mac, Cheap!
data sampling on SE
Postscript <--> imPRESS
TCP/IP on the Mac
Multisync monitors on MacIIs
RE: Using an NEC MultiSync with a Mac II
3 Questions & 1 Flame
SoundWave Review
----------------------------------------------------------------------
Date: Fri, 8 May 87 12:46:30 est
From: munnari!csadfa.oz!jlo@seismo.CSS.GOV (John O'Neill)
Subject: Help wanted: ResErrProc hook.
Does anyone know how best to use the ResErrProc hook (IM 1-116)?
I was hoping to use it as a general error trap mechanism, combined with
a setjmp/longjmp exception handler to avoid having to explicitly call
ResError() after every trap which might generate an error.
IM-1-116 suggests that ResError() also returns error codes for Operating
System errors such as File Manager errors or Memory Manager errors.
This is misleading, as many file manager errors (e.g. dupFNErr) do not
set ResErr, and the ResErrProc is therefore not called.
A further problem is that some ROM routines deliberately (?) induce
Resource Manager errors. i.e. When I tried using the ResErrProc,
it seems to get called with resNotFound (-192) quite a lot, even when
the resource (*is*) found. For example,
SetMenuBar(GetNewMBar(MenuBarID))
results in one call of ResErrProc for each menu in the menu bar, even
though the menus do exist, are found, are displayed OK etc.
Furthermore, ResError() will return noErr in confirmation that all was well!
It turns out that:
GetNewMBar->GetMenuBar->...->SetResLoad(false)->GetResource->GetResInfo
where GetResInfo is passed a handle which is not a handle to a resource,
and hence, as IM-1-121 tells us, ResErr is set to resNotFound.
I currently work around the problem by nullifying my ResErrProc around
such code, but that is hardly a general solution.
Advice on (*exactly*) what errors do invoke ResErrProc would be welcomed,
as would advice on better error management strategies.
John O'Neill Phone ISD: +61 62 68 8818
Dept. Computer Science Telex: ADFADM AA62030
University College ACSNET/CSNET: jlo@csadfa.oz
Aust. Defence Force Academy UUCP: ...!seismo!munnari!csadfa.oz!jlo
Canberra. ACT. 2600. ARPA: jlo%csadfa.oz@SEISMO.CSS.GOV
AUSTRALIA JANET: jlo@oz.csadfa
------------------------------
Date: Thu, 07 May 87 11:26:03 EST
From: sbm@purdue.edu
Subject: Multiple stacks and messing with HeapEnd
I don't remember seeing my last question even posted in a digest
yet, but I have solved it. The problem was that, after my application
started up, everything worked correctly, but the floppy disk never
stopped spinning, even when I quit back to the Finder. The answer was
that, in my overzealous initialization, I was removing the VBL task that
the floppy disk driver installs at system startup.
My new problem is actually an old one. Since I am doing
multitasking, I need multiple stacks, and I have found some more ROM
traps that scribble in random places between HeapEnd and SP, namely,
InitWindows and NewWindow. Do these call MapRgn for some reason
(remember that MapRgn was the routine that was originally found to
display this behaviour)?
Anyway, this is obviously a problem that can't be worked around,
but must be dealt with, and I don't remember seeing any appropriate
suggestions. These types of routines seem to need lots of memory, and
take the region between HeapEnd and SP to be usable stack space. What I
need is a way to tell them just how much stack space they actually have
available, so I tried the obvious. As part of my initialization, I
allocate a big chunk out of the application heap and make it the current
heap zone, so that the application heap is actually never used in my
program. Hence, I thought it wouldn't hurt too much just to set HeapEnd
in the context switch to the topmost (lowest in memory) location
available for stack space for the new process.
To my surprise, it seems to work. There is no more scribbling in
forbidden zones in memory, and, apparently, the ROM routines are not
confused by HeapEnd's bouncing around. Besides that, there is a
benefit: the Mac's stack sniffer will actually tell me when a stack has
*really* overflowed. It all seems too good to be true. Could some
knowledgeable person find out for me whether this radical use of HeapEnd
can mess things up, and, if so, what the correct way is to tell these
routines where the end of the stack zone is?
Steve Munson
sbm@Purdue.EDU
sbm@Purdue.CSNET
------------------------------
Date: Fri, 8 May 87 09:49+0100
From: Ralph <MartinRR%cardiff.ac.uk@Cs.Ucl.AC.UK>
Subject: MacTerminal
I would like to configure Macterminal so that it always starts up with
settings appropriate for my use. Obviously, one way to do that is to
save the settings in a document. Another would seem to be to edit the
Untitled document, then save it, but there seems to be problems with
this - each time I quit, it then asks if I want to save the document.
Another possibility seems to be to cut out the resources which do the
trick from a MacTerminal document, and paste them into MacTerminal
itself. On looking at one, there are 4 types of resources, including
one called CNFG. Does anyone out there have a specification of what
these resources contain, and can they tell me which I need to paste into
MacTerminal itself to customize it, if this is possible ? Or does
anyone have any other good suggestions for customizing ?
Ralph
------------------------------
Date: 08 May 87 16:19:55 BST
From: gorry_fairhurst%aberdeen.ac.uk@Cs.Ucl.AC.UK
Subject: Direct connection of Laserwriter to Mac
Some time ago I made a cable up to directly connect two apple talk
devices. (i.e. a LaserWriter and a Mac), now I seem to remember it
was pretty simple, but can't remember how to connect the differential
driven lines together.
Some one out there must have a better memory than me, could you please
mail me, I've got the wire and DB-9, and I've just put the soldering iron
on to warm.
Gorry (Mac)Fairhurst
------------------------------
Date: Thu, 7 May 87 09:41 EDT
From: Jeffrey Shulman <SHULMAN%slb-test.csnet@RELAY.CS.NET>
Subject: ScrapSaver 1.2
This is ScrapSaver 1.2 (written by Paul Mercer (LOGICHACK) on Delphi.)
Fixes a problem with the clipboard and System 4.1.
[
also archived as [SUMEX-AIM.Stanford.EDU]<INFO-MAC>INIT-SCRAPSAVER-12.HQX
DoD
]
(This file must be converted with BinHex 4.0)
:#P0MFQ&`8f&fCA)!58j*9%e"3e-!N!J#(jQ[!*!%!3#3!`(L!*!$iJ#3!chrN#!
+8f0bBA"6BACPFQ8#!*!$58j*9%e"3e-!N!0S!#-!N!4*6NP868&$8`#3!fJ!)`#
3%T[9d6i!N!B#(rq3JJ#3!pj)jamqB4"K3%cII2K1G8+RUITBMdje-$bSPD&'XIJ
#VQ`HF!LP(QBB)NJ[#%(krq"`#+!Z2c`!"$mm!!1SP8je2c`!!DR)6R8`2!!63UG
4b2rm)%p#U!!-%A`!!3!E-AJ#%!!@3rS!5#&*!"+J!'Bb)A`!N!-%!#3aI!!"!#`
KI!#3!`3!,N2i#@!K53!JS!*Q%+!"8,J*B%*i#@VHr!"36R@J!8+i#@$Hr!"36R8
13faTF'*[BA*N)%CTE'8,9Q9bFfP[EL!a,M)6+%-T)$%j1$FX)&!Z)%ePFQ0PFJ#
3!`%!N!-"iJ#3!q)!N!-p!!1l&J@%!*!$(!!b!!"*6NP8!*!$#J2S!!!J!*!%!lS
U#P0MFQ&`8f&fCA,b[J:
------------------------------
Date: Thu, 7 May 87 09:40:14 CDT
From: CMP.WERNER@r20.utexas.edu
Subject: small mistake in file-name in INFO-MAC-5.084
you announced:
archived as [SUMEX-AIM.Stanford.EDU]<INFO-MAC>UTILITY-DUPTREE-11.HQX
in fact the file is version 1.0 and appears in the archives as:
UTILITY-DUPTREE-10.HQX
no big deal - but it defeated my "semi-automatic" mechanism of fetching
new files ....
Werner
[
Ooops. The file is indeed version 1.0, and is archived as
[SUMEX-AIM.Stanford.EDU]<INFO-MAC>UTILITY-DUPTREE-10.HQX
DoD
]
------------------------------
Date: Fri, 8 May 87 10:34:29 PDT
From: PUGH%CCC.MFENET@nmfecc.arpa
Subject: Sneaky Lightspeed Pascal "Feature"
Well, I sure feel foolish. I was nabbed by a feature of Lightspeed Pascal
and it ends up making me look pretty inept. However, I will swallow my
pride and tell what I have learned in order to prevent it from happening to
someone else.
I was trying to fix a couple of little bugs in Randomizer, as you probably
noticed. I had placed the Randomizer source, project, and resource files
in a folder deep in the my disk for permanent storage. When I wanted to
work on them, I would copy them into the System Folder to do the
development and then copy them back when I was done. So far so good.
Well, I noticed that I had lost some of the changes I had made. So, after
some experimentation I discovered that LSP was saving the pathname of the
source code and modifying the copy that was back in the original directory,
and not the copy in the System Folder, although it was using that project
and that resource file.
A picture would be appropriate here, but you don't get one, you get a table
instead. These are the files used by LSP.
HD:System Folder: HD:LSP:Projects:Randomizer:
Random Project <--- Random Project
Random.pas Random.pas <---
Random Resources <--- Random Resources
Needless to say, this is a bit sneaky of LSP, but to be expected since it
manages the pathnames for you. It is related to the problem where LSP
cannot find a file and asks you to look for it, only to reveal it in the
default directory. LSP should look in the default directory first, and
then search it's saved pathname. That would remove some of the ambiguity
from the issue and would prevent it from asking you to look for all the
files when you move a project to a new folder or disk.
The simplest fix is to modify the project file to use the source copy by
holding down the Option key and double clicking on the source name in the
project window. This is equivalent to Remove and Add. I had merely been
double clicking to bring up the source and getting the wrong copy
unknowingly.
Well, since I have been copying the source, project, and resource files
back to their original locations (replacing the ones there), I have been
losing my changes. This means that the Randomizer in the archives is a
goof and here is the correct one. Lord, please help me so that I don't
have to post this again. I apologize profusely for the inconvienience and
promise that everything now works as advertised. I promise not to do this
again.
People using Randomizer should be aware that the old BeepInit crashes
frequently. You should be using the one included with Randomizer, which is
also archived as INIT-SOUND-BEEP.HQX on Sumex-Aim.Stanford.Edu. People
start using this and then blame Randomizer when the BeepInit crashes.
Honest, it's not my fault. Randomizer just makes sure that BeepInit
eventually gets the one file that will crash it.
If you pass Randomizer along, I would appreciate it if you would keep the
BeepInit and SoundInit files with it. Thanks for your support.
Jon
N L pugh@nmfecc.arpa
M A L National Magnetic Fusion Energy Computer Center
F T N Lawrence Livermore National Laboratory
E L PO Box 5509 L-561
C Livermore, California 94550
C (415) 423-4239
[
once again archived as
[SUMEX-AIM.Stanford.EDU]<INFO-MAC>INIT-RANDOMIZER.HQX
DoD
]
------------------------------
Date: Fri 8 May 87 12:26:00-EDT
From: TN07@TE.CC.CMU.EDU
Subject: File Fixer .05
Attached is the latest version of the program File Fixer. This program repairs
400 or 800-K Mac disks, MFS or HFS. The program was written by Frank Beatrous
at the University of Pittsburgh.
If you recover a trashed disk send your thanks directly to Frank at
110343@PittVMS.BITNET.
Enjoy,
Tom
[
archived as [SUMEX-AIM.Stanford.EDU]<INFO-MAC>UTILITY-FILEFIXER-05.HQX
DoD
]
------------------------------
Date: 7 May 87 17:12:00 EST
From: <bouldin@ceee-sed.arpa>
Subject: Piezo fan for the Mac, Cheap!
A cheap source for the piezo fan that is used in a lot of Mac upgrades is
thru Edmund Scientific. It's $12.00 in quantity one. Stated to move 7 cfm
of air. Connect to the 110 V on the inside of the Mac case and you are in
business. It's on page 41 of the 1987 catalog, under "Educational Aids".
Edmund part number E36,653.
------------------------------
Date: 7 MAY 87 09:15-N
From: INDUGERD%CNEDCU51.BITNET@wiscvm.wisc.edu
Subject: data sampling on SE
Hello,
Can anybody help me about data acquisition with mac SE ?
A friend of mine need to acquire data in a neurophysiology experiment.
The sampling rate is 10 khz.
Are there any hardware and software available on Mac SE for such task ?
Thanks for help.
Philippe
------------------------------
Date: 7 May 87 14:45:00 EDT
From: "STEPHEN_TETHER" <tether@mitlns.mit.edu>
Subject: Postscript <--> imPRESS
Two questions:
(1) Is there a general method of capturing the Postscript commands
that would normally be sent to a Laserwriter when a document is
printed? Ideally the method should work for ALL applications.
(2) Are there any Postscript <--> imPRESS translators out there?
The idea is to use the Imagen laser printers connected to the Lab
VAX-cluster for printing Mac documents. I know about PaintImp, DrawImp, but
these are written in C for Unix, and our machines run VMS and don't have a
C compiler. Translating Postscript to imPRESS seems easier than recoding
the above programs, besides being more general.
Stephen Tether
MIT Laboratory for Nuclear Science
Cambridge, MA 02139
tether@mitlns.mit.edu
tether@mitlns.bitnet
satether@BIX
------------------------------
Date: Thu, 7 May 87 15:22:30 PDT
From: Ray_Davison%SFU.Mailnet@umix.cc.umich.edu
Subject: TCP/IP on the Mac
Our university is in the progress of establishing a campus wide network
to tie together the various networks that are cropping up. We are seriously
looking at TCP/IP as the interconnection protocols. We also have a fairly
large number of Macs around campus. I would like any information anyone
has on TCP/IP products for the Macintosh. This would include any Mac LAN
to TCP/IP gateways, hardware, software and vapourware.
Please reply directly to me and I will summarize the responses for the list.
Ray Davison
Simon Fraser University
Computing Services
Burnaby B.C. Canada V5A 1S6
USERRAYD@SFU.bitnet
Ray_Davison%SFU@um.cc.umich.edu
(604)291-4448
------------------------------
Date: Thu, 7 May 87 15:09:55 EDT
From: Dan Magorian <magorian@umd5.UMD.EDU>
Subject: Multisync monitors on MacIIs
Leornard Li of Xerox posted a query about the use of NEC Multisyncs on the
MacII. Basically, NEC or anybody else's autosyncing unit will work. I
have had a NEC hooked up on and off since February. You need to wire a
conversion from the DB9 connector used by the MEC and the DB15 used by the
MacII (the new IBMs use DB15 as well, probably not the same pinout). The
configuration is:
MacII Monitor Signal
1 ------- 6 ground
2 ------- 1 red
5 ------- 2 green w/ sync
9 ------- 3 blue
3 ------- 4 composite sync (can be omitted)
This adaptor may be available prewired from vendors as well, but it's easy
to sold to solder up. The monitor must be set to analog and manual mode (it
syncs at 35KHz) The quality of the picture isn't as good as the Sony the
Apple sells, but it's OK. Also, use the DB15 to DB15 monitor cable from
Apple (in other words, wire only a short adaptor instead of the whole
cable) to avoid an impedance mismatch that rounds pixels and degrades the
image. If you're buying one outright, buy Apple or another vendor's for
better quality. But the multisyncs are around and easy to get hold of...
Dan Magorian (301) 454-6030
Computer Science Center magorian@umd5.umd.edu
University of Maryland magorian@umdd.bitnet
College Park, MD 20742 Applelink: A0190
------------------------------
Date: Thu, 7 May 87 19:57:54 EDT
From: MacTechnics_User_Group@um.cc.umich.edu
Subject: RE: Using an NEC MultiSync with a Mac II
(Originally from Apple Tech Support)
To hook up an NEC MultiSync monitor to a Macintosh II:
You will need to make an adapter cable from the video card to the
monitor. Here are the pinouts of the video card:
Pins Signal Name
1,4,6,11,13,14 Ground
2 Red
3 CSync- (composite sync)
5,12 Green (with sync)
9 Blue
7,8,10,15 Not connected
Here is the pinout description for the adapter cable (using the
sync-on-green configuration):
Macintosh II NEC Signal Name
1 ----------- 6,7,8,9 Ground
2 ----------- 1 Red
5 ----------- 2 Green (sync)
9 ----------- 3 Blue
The monitor must be set to Analog mode and manual mode as well.
I know someone that tried this without 100% success (fuzzy picture). I'm
not sure what the problem was, it's possible that unused pins should be
grounded or tied high.
Eric Shapiro
MacTechnics
P.O. Box 4063
Ann Arbor, Mi 48106
------------------------------
Date: Fri, 8 May 87 08:35:12 PDT
From: UWACDC.BITNET
Subject: 3 Questions & 1 Flame
1. Does anyone know what the price of A/UX is going to be (AUC or retail)?
2. Does anyone know how to reach LoDown? We're very interested in their
WORM drive for the Mac and can't find any adds in recent Mac magazines.
Any experience with this device?
3. Does anyone know of any GIS (Geographic Information System) software for
the MAC? How about the (shudder) PC?
<Flame On>
4. I'm a new developer and have already hit the 32K segment wall using LSP.
How do the developers of LSP, TML, or Turbo expect us to do scientific
programming on the Mac with such a silly limit? I know, I know, the
limit is really caused by the Segment Manager but that didn't seem to
stop Absoft--their FORTRAN compiler doesn't have the 32K limit. Shame
on Apple for such a fundamental shortcoming and shame on all the Pascal
compilers for propogating the mistake. I can only hope that the new
Mac II toolbox doesn't have such a limit and/or software developers
give us a work around. I don't relish the idea of using FORTRAN on
Mac even with MacFace available. I realize that this is an old complaint
but flaming is a great release. Perhaps this will be the flame that
gets something done about this serious problem.
BTW: Prototyper from SmethersBarnes, advertised as 'availablw now' in
MacTutor is just now going to Beta testing! I could really use such
a product right now and my hopes were dashed to find it still vapor.
Should be out RSN (right).
<Flame off>
John Osborne
School of Oceanography, WB-10
University of Washington
Seattle, WA 98195
<OZ001@UWACDC.BITNET>
------------------------------
Date: Thu, 7 May 87 10:23 EDT
From: BOGARTZ%UMASS.BITNET@wiscvm.wisc.edu
Subject: SoundWave Review
I just received my copy of SoundWave, the next generation of sampling
software from the folks who brought us MacNifty SoundCap. (They are
now called Impulse, but are working out of the same location.)
SoundWave is mostly a great improvement over SoundCap. Here is my
impression of the improvements they have made:
It has Undo!
It is written in Lightspeed Pascal.
It uses a much more standard user interface. Where SoundCap took
over the entire screen, SoundWave opens up to four separate sample
windows, each with its own palette of tools for playing, drawing on
the wave, etc.
They have solved the problem of how to view a whole sample while
editing it at a fine level of detail. The scroll bar is a horizontal
"summary" of the sample, scaled both vertically and horizontally to
fit in one screen's width, that resides at the top of the sample's
window. The thumb of the scroll bar is a little transparent
rectangle that corresponds to what is in the main portion of the
window. If you remember the video game Defender, you will find it
very reminiscent of the planet summary at the top of the screen.
It has a digital filter bank with five filters, each of which can
be set at a user specifiable frequency. (I guess this makes it
a parametric equalizer, but it is laid out as a graphic eq.)
It is truly user-programmable. Here's how you do it. The program
provides a menu titled "Custom." In SoundWave there is a resource
called "swci" (for SoundWave custom item). Every CODE resource
pasted into the swci appears as an item under the Custom menu. So
all you have to do is use your favorite language compiler to create a
CODE resource, paste it into SoundWave with ResEdit, and SoundWave
effectively becomes a front-end to whatever digital sound application
you want to write. The CODE resource gets a pointer to
SoundWave's frontmost window that contains a sample (SoundWave won't
let you choose anything from the Custom menu unless the frontmost
window contains a sample). From this it can get a pointer to an
array containing the sample data. Then you can mess around with it
to your heart's desire. Some obvious applications for this: a tone
generator (included as a sample Custom item with SoundWave, along
with LSP project and source file); a routine to dump your sample to a
sampling keyboard over MIDI; some heavy-duty algorithms for digital
processing, such as digital reverb, harmonization (pitch change),
compression, etc.
Gripes:
It doesn't have some of the signal processing features that
SoundCap had, like flanging or scratchbar. (However, I would guess
that because of SoundWave's easy extensibility, people will write them...)
Most of the dialog boxes have "Cancel" on the left and "OK" on the
right, which I think is backwards from most applications.
The filtering operation takes a long time (over 20 seconds for even
small ranges). This is understandable, but the "Cancel" button does
not work during this process, leading you to think your Mac has
hung. There should be a warning that "this may take a while."
I am going to be writing an extensive review of SoundWave for the
June issue of the Boston Computer Society's Computers and Music
newsletter, which is called (coincidentally, and we thought of it
first!) "Soundwaves". What? You've never heard of or seen
"Soundwaves"? Well, just send your name and address to me:
David Bogartz
81 Bridge St., Apt. 2
Northampton MA 01060
(Bitnet: DSBOGARTZ@AMH)
for a complimentary issue.
DISCLAIMER: I am not associated with Impulse in any way except as a
satisfied customer. I am, however, associated with the Boston
Computer Society, being a member and the editor of "Soundwaves".
------------------------------
End of INFO-MAC Digest
*****Thisions oTerminaD