Nautilus 1992 July/Nautilus-3-8/Nautilus-3-8.bin/Tools & Utilities/Techy Stuff/Doco ƒ/CSMP ƒ/CSMP-V1-081.TXT

Text File | 1992-06-30 | 43.1 KB | 1,233 lines

C.S.M.P. Digest             Fri, 15 May 92       Volume 1 : Issue 81

Today's Topics:

    TextWidth and LaserWriters
    MacBinary source
    Problems with ResEdit2.1
    Think C, TCL, AppleTalk Async calls and a question
    Anybody know the SE "Easter Egg?"
    debugging tools for lisp
    Network copy protection problems
    Problem with PBHGetFInfo
    Regular reading of a serial device-how?
    How to read 10 sectors per track on floppy? (Repost)
    WWDC Dinner?
    archive site for Frontier stuff?
    What is Discipline?
    32K boundary problem?


The Comp.Sys.Mac.Programmer Digest is moderated by Michael A. Kelly.

These digests are available (by using FTP, account anonymous, your email
address as password) in the pub/mac/csmp-digest directory on ftp.cs.uoregon.
edu.  This is also the home of the comp.sys.mac.programmer Frequently Asked
Questions list.  The last several issues of the digest are available from
sumex-aim.stanford.edu as well.

These digests are also available via email.  Just send a note saying that you
want to be on the digest mailing list to [e-mail], and you will
automatically receive each new digest as it is created.

The digest is a collection of articles from the internet newsgroup comp.sys.
mac.programmer.  It is designed for people who read c.s.m.p. semi-regularly
and want an archive of the discussions.  If you don't know what a newsgroup
is, you probably don't have access to it.  Ask your systems administrator(s)
for details.  (This means you can't post questions to the digest.)

The articles in these digests are taken directly from comp.sys.mac.programmer.
They are not edited; all articles included in this digest are in their original
posted form.  The only articles that are -not- included in these digests are
those which didn't receive any replies (except those that give information
rather than ask a question).  All replies to each article are concatenated
onto the original article in the order in which they were received.  Article
threads are not added to the digests until the last article added to the
thread is at least one month old (this is to ensure that the thread is dead
before adding it to the digests).

Send administrative mail to [e-mail].

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

From: [e-mail] (Paul L Merchant Jr.)
Subject: TextWidth and LaserWriters
Date: 9 Apr 92 13:16:40 GMT
Organization: Dartmouth College, Hanover, NH

I'm having trouble spacing text properly on a LaserWriter.  I'm trying
to draw two strings, the second following the first by x "pixels".  The
procedure I'm using looks like this:

MoveTo(startx, starty);
DrawText(text1);
Move(x, 0);
DrawText(text2);

The problem is that apparently the LaserWriter's idea of the pen
location after the first DrawText is different from the GrafPort's idea
of the pen's location.  So the space between text1 and text2 is much
larger than I want.  The effect seems to vary with the font choice with
the reproduction being best if I have chosen a LaserWriter font such as
Helvetica.  (This code is part of an application which allows the user
to select any available font.)

The problem is really dramatic if the following lines are added after
the first DrawText:

Line(0, -10);
Move(0, 10);


This should draw a vertical line immediately after the text, right?
Wrong.  Instead I get a diagonal line whose width is equal to the error
between the different measurements.

I have tried using SetFractsEnable set to both true and false, and I
have tried Font Subtitution and Text Smoothing enabled and disabled.
All have an effect on the problem, but none cure it.

My questions then are:

1. Is there any way to discover the "true" pen location when printing
since it does not appear to be available using GetPen?

2.  If not, is there a way to discover what the actual printed text
width is?  I have compared the results of TextWidth(text1) within the
printing GrafPort and a screen GrafPort and they are the same.  I have
also used the global font width table to compute my own text widths and
still found no difference.

3.  Is there any way to find out what font the LaserWriter is
substituting when it is drawing?

4.  Is there another way to make the spacing between text1 and text2
accurate?  (There must be some way, or word processors would never be
able to do justification.)

5.  Is there any documentation about this anywhere?

And on a related topic:

6.  How can I find out which fonts are "fixed width"?  I thought about
looking at the fontType field of the font record until I discovered
that it described Monaco as a proportional font even though it appears
to be fixed width.

Thanks in advance for any help you can give me!

Paul Merchant
[e-mail]

+++++++++++++++++++++++++++

From: [e-mail] (Christopher Tate)
Date: 14 Apr 92 18:19:27 GMT
Organization: Penn State University

You need to read, carefully and patiently, the Tech Notes that discuss
printing in general, and LaserWriters in particular.  I don't recall just
which Tech Notes they are, but the numbers 92 and 86 stick in my mind.
Get the index (or the Tech Notes Stack), and go from there.

Basically, when you print text to the LaserWriter all manner of strange
and wonderous things may happen, such as Font Substitution, which will
play merry hell with the actual position of text and other graphics on
the page.  Get the tech notes and study them, and experiment a lot.

And be glad you don't have to pay $0.25 per page when experimenting,
the way I did....  :-(

- -------
Christopher Tate        |  "Computer Science has more real-world
                        |   applications than rotisserie baseball."
[e-mail]    |
Bitnet:  CXT105@PSUVM   |       -- Ben Liblit, next-door neighbor

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

From: [e-mail] (Francois Schiettecatte)
Subject: MacBinary source
Date: 10 Apr 92 21:02:32 GMT
Organization: Johns Hopkins Univ. Welch Medical Library

Hi

I am looking for C source to convert MacBinary files downloaded
from a Unix box to a proper mac file. Ideally this source
should be public domain, can anybody give me any pointers?


francois



Francois Schiettecatte
Internet: [e-mail]
Phone    : [phone]

+++++++++++++++++++++++++++

From: [e-mail] (Mel Walker)
Organization: Novell, Inc. - Walnut Creek
Date: Tue, 14 Apr 1992 14:40:26 GMT

Does anyone out there know of a site from which I can download the source
code for doing MacBinary file conversion?  I would much appreciate it.
Thanks,
- --Mel Walker                                 [e-mail]

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

From: [e-mail] (taylor)
Subject: Problems with ResEdit2.1
Date: 13 Apr 92 08:40:38 GMT
Organization: Brighton Polytechnic, UK

I have ResEdit 2.1 which seems to work fine except when I try to change the
name or ID of a resource. Attempting to do this results in the machine
locking up. I am using system 7.0 - does anyone else have this problem or
have I got a duff version? Help!

Paul Taylor
Brighton Polytechnic
England

+++++++++++++++++++++++++++

From: [e-mail] (John E Bush)
Date: 13 Apr 92 12:08:30 GMT
Organization: The Ohio State University

In article <[e-mail]> [e-mail]
 (taylor) writes:
>I have ResEdit 2.1 which seems to work fine except when I try to change the
>name or ID of a resource. Attempting to do this results in the machine
>locking up. I am using system 7.0 - does anyone else have this problem or
>have I got a duff version? Help!
>
>Paul Taylor

Paul,

The latest version of ResEdit is 2.1.1 and is available on Apple's ftp site.
I used to have the same problem that you are experiencing, but have not had
the problem since I upgraded versions.

Eric Bush

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

From: [e-mail] (Johnathon L. Suker)
Subject: Think C, TCL, AppleTalk Async calls and a question
Organization: Univesity of California at Irvine Main Library
Date: 10 Apr 92 22:37:50 GMT

Hey AppleTalk and Think Class Library,

    I have a programming question regarding AppleTalk programming and
the think class library.  I am going to write (am writing) a programming
that uses appletalk and sends syncronous calls.  I want to change them to
Async. calls.  According to Inside AppleTalk, to should check for async
calls in the main event loop of your program.  In TCL I do not have access
to this and inorder to put something in their I must changing the source
code of TCL (which Symantec says is BAD).  Has anyone (or does anyone) know
how to make Async calls within TCL.  I would really like to use the TCL (my
first big project using it) and this is the only thing holding me back.

                Thanks in advance,

                    Johnathon
_______________________________________________________________________________
Johnathon Suker           |
UC Irvine                 |   .signature file under construction
[e-mail]  |
_______________________________________________________________________________

+++++++++++++++++++++++++++

From: [e-mail] (Lance 'Moof' Kimes)
Date: 10 Apr 1992 19:17:45 -0700
Organization: University of Southern California, Los Angeles, CA


In article <[e-mail]>, [e-mail] (Johnathon L. Suker) writes:
|> Hey AppleTalk and Think Class Library,
|>
|>     I have a programming question regarding AppleTalk programming and
|> the think class library.  I am going to write (am writing) a programming
|> that uses appletalk and sends syncronous calls.  I want to change them to
|> Async. calls.  According to Inside AppleTalk, to should check for async
|> calls in the main event loop of your program.  In TCL I do not have access
|> to this and inorder to put something in their I must changing the source
|> code of TCL (which Symantec says is BAD).  Has anyone (or does anyone) know
|> how to make Async calls within TCL.  I would really like to use the TCL (my
|> first big project using it) and this is the only thing holding me back.
|>
|>                 Thanks in advance,
|>
|>                     Johnathon
|> _______________________________________________________________________________
|> Johnathon Suker           |
|> UC Irvine                 |   .signature file under construction
|> [e-mail]  |
|> _______________________________________________________________________________


When I first started TCL programming I made the same mistake, but remember that TCL is a class library!!

CMyClass : public CApplication


override Process1Event

void CMyClass::Process1Event( void)
{


my stuff & checking


inherited::Process1Event();
}

then call CApplication method for processing events



Hope it helps

Lance Kimes
USC

+++++++++++++++++++++++++++

From: [e-mail] (Jon W{tte)
Date: 11 Apr 92 10:30:48 GMT
Organization: Royal Institute of Technology, Stockholm, Sweden

.uci.edu> [e-mail] (Johnathon L. Suker) writes:

   calls in the main event loop of your program.  In TCL I do not have access
   to this and inorder to put something in their I must changing the source
   code of TCL (which Symantec says is BAD).  Has anyone (or does anyone) know

Of course you do !

Your gopher will get a dawdle message, or you can install~
CIdleChores.

- --
[e-mail]; Jon W{tte, the Diplomat - NOT!

+++++++++++++++++++++++++++

From: [e-mail] (Martin Minow)
Date: 11 Apr 92 20:33:37 GMT
Organization: Digital Equipment Corporation


In article <[e-mail]>, [e-mail]
(Johnathon L. Suker) asks about asychronous I/O and the Think library.

The simplest solution is to create a CChore to watch for I/O completion.
If your I/O needs are more elaborate, you might want to create your
own action queue using roughly the following algorithm:

Your I/O routine (PBControl or similar) establishes an I/O completion
handler.  This is a short assembly-language stub that queues the I/O
block into a private queue.  Your CChore then looks for a non-empty
queue and, if it finds something, dequeues the first element and
processes it.

To save you much head-scratching, create your I/O request as follows:
    struct MyIORecord {
        long    applicationA5;
        ParamBlock    pbRec;
        ... other data for the I/O request ...
    };

Create the I/O request by

    anIORequest.applicationA5 = (long) CurrentA5;
    ... other parameters ...
    anIORequest.pbRec.ioCompletion = (ProcPtr) MyCompletionRoutine;
    status = PBControlAsync(&anIORequest.pbRec);

The completion routine is as follows:

QHdr        gProcessQueue;

pascal void
MyCompletionRoutine()
{
    asm {
        move.l    a5,-(sp)        ; Save current A5
        move.l    -4(a0),a5        ; Restore application A5
        lea        gProcessQueue,a1    ; My queue header
        _Enqueue                ; Put buffer in my queue
        movea.l    (sp)+,a5        ; Restore A5
    }
}

Your CChore or idle loop should call a "dequeue" function defined as follows:

/*
 * Get an element from the queue, return NULL if none.
 */
struct MyIORecord *
GetFromQueue(void)
{
    QHdrPtr        theQueue;
    QElemPtr    qElemPtr;

    theQueue = &gProcessQueue;        /* Parameter, perhaps    */
    if ((qElemPtr = theQueue->qHead) == NULL)
        return (NULL);
    else {
        Dequeue(qElemPtr, theQueue);
        return ((struct MyIORecord *) (((long) qElemPtr) - sizeof (long)));
    }
}

This was retyped from working code: apologies for any errors that crept in.
Note that your CChore should set a short maxSleep time if the queue is
non-empty. (The method that calls PBControlAsynch must also set a
small value in gSleepTime.)

Martin Minow
[e-mail]

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

From: [e-mail] (M. Scott Marcy)
Subject: Anybody know the SE "Easter Egg?"
Organization: Colorado SuperNet, Inc.
Date: Sat, 11 Apr 1992 05:44:15 GMT

I was just wondering if anybody out there knows and remembers the address of
the "Easter Egg" in the original SE ROMs?  It's the one which draws the
pictures of the development team.

I guess while I'm at it, any other "Eggs" would be appreciated, too (other
Macs, any software, etc.).  I know the one about the Zebra Lady in
Mac Paint II.

I'll summarize to the net if I get anything new.

Thanks!

- -Scott

- --
M. Scott Marcy                            Internet:  [e-mail]
First Class Systems, Inc.                 Applelink: FIRST.CLASS
PO Box 50005
Colorado Springs, CO  80949 USA           Disclaimer: I own the company!

+++++++++++++++++++++++++++

From: [e-mail] (Neal Trautman)
Date: 11 Apr 92 18:55:48 GMT
Organization: Farallon Computing, Inc.

In article <[e-mail]>, [e-mail] (M. Scott Marcy) writes:
> I was just wondering if anybody out there knows and remembers the address of
> the "Easter Egg" in the original SE ROMs?  It's the one which draws the
> pictures of the development team.

Here's the THINK C source code for a small
application I wrote years ago to show the pictures.

Enjoy...
Neal Trautman
Timbuktu Lead Software Engineer
Farallon Computing, Inc.
[e-mail]


main()
{
    SysEnvRec        MacInfo;
    EventRecord        theEvent;
    long            time;
    int                i;

    InitGraf(&thePort);
    InitFonts();
    FlushEvents(everyEvent, 0);
    InitWindows();
    InitDialogs(0L);
    InitCursor();
    SysEnvirons(1, &MacInfo);
    if (MacInfo.machineType != envSE)
        {
        ExitToShell();
        }
    asm
        {
        jsr        0x41D8B4
        }
    for(;;)
        {
        asm
            {
            addq.l    #4,D7
            andi.l    #0x0C,D7
            jsr        0x41D8D6
            }
        for (i=0; i<=8; i++)
            {
            if (Button())
                ExitToShell();
            Delay(20L,&time);
            }
        }
}

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

From: [e-mail] (Chen-Kuang Yu)
Subject: debugging tools for lisp
Date: 11 Apr 92 17:49:26 GMT
Organization: DSG, Stanford University, CA 94305, USA

I've been using lisp on the NeXT machines and recently have to do some
programming on the mcl.  However I find the STEP and TRACE debugging
macro rather crude on the mcl.  I wonder if it's mentioned in the
manual that there are some other more powerful debugging tools
available.  (I'm using the program running on the server of our schoo.
The consultant said the manual was lost, so there's no way for me to
check it out.)

BTW, are there any ftp sites where they keep a bunch of useful lisp
codes such as a modified TRACE or STEP?

Thank in advance.

/Chen Yu ([e-mail])

+++++++++++++++++++++++++++

From: [e-mail] (Kent Sandvik)
Date: 14 Apr 92 19:36:02 GMT
Organization: MacDTS Mongols

In article <[e-mail]>, [e-mail]
(Chen-Kuang Yu) writes:
>
> I've been using lisp on the NeXT machines and recently have to do some
> programming on the mcl.  However I find the STEP and TRACE debugging
> macro rather crude on the mcl.  I wonder if it's mentioned in the
> manual that there are some other more powerful debugging tools
> available.  (I'm using the program running on the server of our schoo.
> The consultant said the manual was lost, so there's no way for me to
> check it out.)

As I know STEP and TRACE are the default Common Lisp utilities under
MCL. There's an inspector which you could use for inspecting data
structures.

> BTW, are there any ftp sites where they keep a bunch of useful lisp
> codes such as a modified TRACE or STEP?

cambridge.apple.com has MCL code, you might check their archives.

Cheers,
Kent Sandvik

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

From: [e-mail] (Chris D. Nebel)
Subject: Network copy protection problems
Organization: University of Maryland, College Park
Date: Sat, 11 Apr 1992 23:20:23 GMT

I'm trying to write one of those things that checks to make sure that you're
only running one copy of the same program at once on an Appletalk network.
(Please, no flames from folks that think this practice is evil.  This is my
job we're talking about, here.)  Anyway, I've got a little thing written that
registers the "name" (all right, the serial number) of the program using
NBPRegister; if NBPRegister comes back with a duplicate name error, then it
decides that another copy is being run somewhere and exits.  (The program
unregisters the name when you quit.)

This works pretty well, but there are two major problems.  One, it only checks
the current zone.  I'd like it to check ALL the zones, but I can't seem to find
anything at all on zones in IM except for an occasional mention of '*' -- where
does Apple hide this information, anyway?

Problem number two is more serious: if the program terminates abnormally (read:
crashes! (or the user kills it, or whatever)), then the name never gets
unregistered.  So when they try to launch again, it says "This copy is already
being used!"  Oh, joy.  I tried turning off self-send during the inital name-
check, and that seems to work, but when it tries to unregister the name,
something fairly horrible must happen, because the Finder almost always
crashes a few seconds after I quit.

Any tips/pointers/useful comments would be greatly appreciated.

Thanks,
Chris Nebel
[e-mail]

+++++++++++++++++++++++++++

From: [e-mail] (Michael Peirce)
Date: 12 Apr 92 16:30:37 GMT
Organization: Peirce Software


In article <[e-mail]> (comp.sys.mac.programmer), [e-mail] (Chris D. Nebel) writes:
> I'm trying to write one of those things that checks to make sure that you're
> only running one copy of the same program at once on an Appletalk network.
> (Please, no flames from folks that think this practice is evil.  This is my
> job we're talking about, here.)  Anyway, I've got a little thing written that
> registers the "name" (all right, the serial number) of the program using
> NBPRegister; if NBPRegister comes back with a duplicate name error, then it
> decides that another copy is being run somewhere and exits.  (The program
> unregisters the name when you quit.)
>
> This works pretty well, but there are two major problems.  One, it only checks
> the current zone.  I'd like it to check ALL the zones, but I can't seem to find
> anything at all on zones in IM except for an occasional mention of '*' -- where
> does Apple hide this information, anyway?

You need to use the Zone Information Protocol, ZIP, to find out what
zones exist in your network.  With it you can lookup the name in each
zone.

Of course, this will take FOREVER in a large network environment -
and there are many large AppleTalk networks out there.

You need to be a little more clever than just checking everywhere.
Do something like only picking a couple of random zones each time you
start up.

Before you release your product do some serious testing of your scheme
on a real network (at least a few dozen zones with real traffic) and
make sure you aren't hogging all the resources or causing other problems
for people.  Your customers will be happier and so will you - you
don't want people dumping your product because it nukes their network!

> Problem number two is more serious: if the program terminates abnormally (read:
> crashes! (or the user kills it, or whatever)), then the name never gets
> unregistered.  So when they try to launch again, it says "This copy is already
> being used!"  Oh, joy.  I tried turning off self-send during the inital name-
> check, and that seems to work, but when it tries to unregister the name,
> something fairly horrible must happen, because the Finder almost always
> crashes a few seconds after I quit.

Patch the ExitToShell trap and do your unregistering there.

Not only are you getting a duplicate name, but worse, NBP still thinks
it owns the block of memory that contains the original name.  This
chunk of memory could very well be part of another program's memory
space now!

- --  Michael Peirce         --   [e-mail]
- --  Peirce Software        --   Suite 301, 719 Hibiscus Place
- --  Macintosh Programming  --   San Jose, California USA 95117
- --           & Consulting  --   voice: [phone] fax: [phone]
- --                         --   AppleLink: peirce & America Online: AFC Peirce

+++++++++++++++++++++++++++

From: [e-mail] (Scott Bronson)
Date: 12 Apr 92 22:20:23 GMT

In <[e-mail]> [e-mail] (Michael Peirce) writes:


>>[On a crash, names aren't being unregistered]

>Patch the ExitToShell trap and do your unregistering there.

>Not only are you getting a duplicate name, but worse, NBP still thinks
>it owns the block of memory that contains the original name.  This
>chunk of memory could very well be part of another program's memory
>space now!

Yes, but this won't help you on a crash.  I don't know how comfortable
you might feel with patching _SysError...

You will probably want to periodically want to poll the names in use,
and simply ping each one to make sure they are still there.  If there
was a crash, you wouln't get a ping, and you might want to take some
action there.

If the user aborts (cmd-opt-esc), I don't know if _ExitToShell gets
called.  Can you rely on it always being called in the future?  Perhaps
a more future compatible way of doing this would be to make a small
stub application that simply launches the main application.  The small
one would accept ChildDied events, then know when to release the name.
(when its child died).  If you had the main app register the name and
the stub app remove it, I think you can be virtually guranteed future
compatibility.

    - Scott

+++++++++++++++++++++++++++

From: [e-mail] (Joe Zobkiw)
Organization: The World Public Access UNIX, Brookline, MA
Date: Mon, 13 Apr 1992 14:04:07 GMT

Chris -

I learned about Zones from IM VI I think...also check IM V.

As far as your problem...I do the exact same thing, from within an INIT
and it works without a hitch. You mention that if the application crashes,
the name doesn't unregister...this could be a problem...but you would hope
that whenever someone's machine crashes, they will restart it just to be
safe. When the machine is restarted, the registration will disappear and
everything will be back to normal.

- --
<--------------------------------------------------->
 joe zobkiw                     [e-mail]
 mac.synthesis.MIDI.development.C.asm.communications
>---------------------------------------------------<

+++++++++++++++++++++++++++

From: [e-mail] (Mel Walker)
Organization: Novell, Inc.
Date: Mon, 13 Apr 1992 14:48:13 GMT

>> Problem number two is more serious: if the program terminates abnormally (rea
d
>> crashes! (or the user kills it, or whatever)), then the name never gets
>> unregistered.  So when they try to launch again, it says "This copy is alread
y
>> being used!"  Oh, joy.  I tried turning off self-send during the inital name-
>> check, and that seems to work, but when it tries to unregister the name,
>> something fairly horrible must happen, because the Finder almost always
>> crashes a few seconds after I quit.
>
>Patch the ExitToShell trap and do your unregistering there.
>
>Not only are you getting a duplicate name, but worse, NBP still thinks
>it owns the block of memory that contains the original name.  This
>chunk of memory could very well be part of another program's memory
>space now!
>
>--  Michael Peirce         --   [e-mail]
>--  Peirce Software        --   Suite 301, 719 Hibiscus Place
>--  Macintosh Programming  --   San Jose, California USA 95117
>--           & Consulting  --   voice: [phone] fax: [phone]
>--                         --   AppleLink: peirce & America Online: AFC Peirce


How does one patch the ExitToShell trap to do something? Sample or psuedo code,
please...

- --Mel Walker
[e-mail]
** I have not yet developed a snazzy disclaimer **

+++++++++++++++++++++++++++

From: [e-mail] (Rick Gordon)
Date: 14 Apr 92 00:09:41 GMT
Organization: Aldus Corporation, Seattle WA


regarding "pinging" every node periodically: this has a couple of faults.
first it generates a lot of extra traffic, AT is bad enough with broadcasts
why make it worse?
second it's non-deterministic: network loads can vary significantly depending
upon hops through routers. what you might this is a failed name is not.
of course, even if you detect the name doesn't answer, you cannot remove the
name from that nodes name table so now you have to keep a list of those nodes
that contain bogus names.

what i would suggest if NBP Register detects a match is doing an NBP Query
and testing the node id against the local host. (You should be doing this
anyway in case the user has SetSelfSend turned on.) If you detect this
situation at startup, you know it's because the name didn't properly get
removed the last time, in which case you should probably advise the user
to reboot. (Because NBP still has that block of memory - or thinks it does -
containing the name that was originally registered. And who knows what's
happened between the time your application crashed and the time it was
relaunched!)

[e-mail]

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

From: [e-mail]
Subject: Problem with PBHGetFInfo
Date: 13 Apr 92 12:46:22 GMT
Organization: The MITRE Corporation

Can anyone think why this returns -43 (file not found)

    gMultiFiles[index]->parID  is 1610667778
    gMultiFiles[index]->vRefNum is 0
    kFSAsynch is true
- ----
    HFileParam        fpb;

    /* prepare a hParamBlock */
    fpb.ioCompletion = nil;
    fpb.ioFVersNum = 0;
    fpb.ioFDirIndex = 0;
    fpb.ioNamePtr = CtoPstr(gMultiFiles[index]->name);
    fpb.ioVRefNum = gMultiFiles[index]->vRefNum;
    fpb.ioDirID = gMultiFiles[index]->parID;
    err = PBGetFInfo(&fpb,!kFSAsynch);

when this get s info on the file.  (I want to use this so I can look at files
in different directories)

    FileParam        fpb;

    fpb.ioCompletion = nil;
    fpb.ioFVersNum = 0;
    fpb.ioFDirIndex = 0;
    fpb.ioNamePtr = CtoPstr(gMultiFiles[index]->name);
    fpb.ioVRefNum = gMultiFiles[index]->vRefNum;
    err = PBGetFInfo(&fpb,!kFSAsynch);

+++++++++++++++++++++++++++

From: [e-mail] (Jon W{tte)
Date: 13 Apr 92 19:33:16 GMT
Organization: Royal Institute of Technology, Stockholm, Sweden

.mitre.org> [e-mail] writes:

   Can anyone think why this returns -43 (file not found)

       gMultiFiles[index]->parID  is 1610667778
       gMultiFiles[index]->vRefNum is 0

I have a hard time believing you ever got that many folders
created on your disk.

Instead, it looks like a large negative number or something
like that; probably the result of assigning a wdRefNum to
a long.

Try swapping the two around, so the dirID is 0 and the
vRefNum is whatever the parID was assigned.

- --
"You should meet yourself someday. I'm sure you would hate it."
- - Me: [e-mail]; Jon W{tte (The Diplomat - NOT!)

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

From: [e-mail] (Michael Kohne)
Subject: Regular reading of a serial device-how?
Date: 13 Apr 92 01:57:08 GMT
Organization: University of Maryland Baltimore Campus, Computer Science Department

I've got an interesting situation. I have come by a graphics tablet that is a
serial device, and I want to hook it to my mac. This should be simple, as I
know the format of the data it's sending, and I know how to use the serial
ports on the mac. There's only one problem: I need to check the serial ports
pretty regularly in order to do this, but the Control and Status driver
calls might move or purge memory, and therefore can't be called from any sort
of interrupt-level routine like a VBL task. Then I thought of using an
interrupt routine directly, but since that would involve playing with the
hardware directly, it's not a good idea.

What I'd really like is either:
1) a way to call the serial drivers from a VBL task.
2) a way to get some code called regularly (every 10 ticks or so), but NOT
at interrupt level. I can't think of any way either of these can be done,
so I ask you: any ideas?

respond email, and I'll post a summary.

Michael Kohne
"Pope" Q.E.D

- --
"Pope" Q.E.D         Lord High Jabberwocky
Michael Kohne        [e-mail]
Coded Message: FIME NDUXXMS MZP FTQ EXUFTK FAHQE SKDQ MZP SUYNXQ UZ FTQ IMNQ.

+++++++++++++++++++++++++++

From: [e-mail] (Jon W{tte)
Date: 13 Apr 92 08:09:17 GMT
Organization: Royal Institute of Technology, Stockholm, Sweden

> [e-mail] (Michael Kohne) writes:

   1) a way to call the serial drivers from a VBL task.

Can't help you there

   2) a way to get some code called regularly (every 10 ticks or so), but NOT
   at interrupt level. I can't think of any way either of these can be done,
   so I ask you: any ideas?

Hey, how about a simple application that passes a value < 10 to
WaitNextEvent ? Could even make it faceless background...

Or you could write a driver (DRVR) and have an INIT install it;
DRVRs get regular time from SystemTask.

Or you could patch one of the "regularies" like _OSEvent or
_SystemTask - but that degrades system performance a lot.

Hope this helps !

- --
"You should meet yourself someday. I'm sure you would hate it."
- - Me: [e-mail]; Jon W{tte (The Diplomat - NOT!)

+++++++++++++++++++++++++++

From: [e-mail] (Scott Bronson)
Date: 14 Apr 92 18:35:40 GMT

In <[e-mail]> [e-mail] (Jon W{tte) writes:

>> [e-mail] (Michael Kohne) writes:

>   2) a way to get some code called regularly (every 10 ticks or so), but NOT
>   at interrupt level. I can't think of any way either of these can be done,
>   so I ask you: any ideas?

>Or you could write a driver (DRVR) and have an INIT install it;
>DRVRs get regular time from SystemTask.

>Or you could patch one of the "regularies" like _OSEvent or
>_SystemTask - but that degrades system performance a lot.

DRVRs get time, but I wouldn't exactly call it regular.  If you NEED to
be called every 10 ticks, right now a VBL task is the only way I know
of to ensure you get the time.  Most applications (and much system
software) that I've seen don't release processor time when tracking
mouseDowns, so if the user holds down the mouse button for more than
ten ticks (which is not uncommon), you could be left high and dry...

    - Scott

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

From: [e-mail] (Steve Berkley)
Subject: How to read 10 sectors per track on floppy? (Repost)
Date: 13 Apr 92 14:55:22 GMT
Organization: University of Rochester - Rochester, New York

I'm posting this question again, in case anyone didn't see the
original.

I'm looking for a way to read sectors off of diskettes for a
keyboard sampler, which are formatted quite like IBM diskettes-
10 sectors per track, 80 tracks on two sides.  The mac drive
(.sony driver) reads at 8-12 sectors per track, depending on the
location of the track on the physical disk (for optimum storage
space).  What I'd like to do is be able to tell the driver to
read sectors 0-9 on *each* track.

Most have said that this requires a new driver...lots of work.
Does anyone have a simple solution?

- -Steve Berkley
[e-mail]

+++++++++++++++++++++++++++

From: [e-mail] (Matthew T. Russotto)
Date: 13 Apr 92 17:09:27 GMT
Organization: College of Engineering, University of Maryland, College Park

In article <[e-mail]> [e-mail] (Steve Berkley) writes:
>I'm posting this question again, in case anyone didn't see the
>original.
>
>I'm looking for a way to read sectors off of diskettes for a
>keyboard sampler, which are formatted quite like IBM diskettes-
>10 sectors per track, 80 tracks on two sides.  The mac drive
>(.sony driver) reads at 8-12 sectors per track, depending on the
>location of the track on the physical disk (for optimum storage
>space).  What I'd like to do is be able to tell the driver to
>read sectors 0-9 on *each* track.

You can't do this on an 800K drive, but if you have an FDHD (and
appropriate support chips), ask the .Sony driver to read it in IBM
format-- the tech notes have the documentation for that driver.

Originally you said that the diskette was recorded in GCR format.  If
so, you have a really strange disk, and will need to write a new
driver.  I didn't know anyone other than Apple used GCR for 3.5"
floppies.


- --
Matthew T. Russotto    [e-mail]    [e-mail]
Some news readers expect "Disclaimer:" here.
Just say NO to police searches and seizures.  Make them use force.
(not responsible for bodily harm resulting from following above advice)

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

From: [e-mail] (Neal Trautman)
Subject: WWDC Dinner?
Date: 9 Apr 92 13:29:36 GMT
Organization: Farallon Computing, Inc.


Is there any info on an Internetters dinner at
Apple's World Wide Developer Conference?

- --
Neal Trautman
Timbuktu Lead Software Engineer
Farallon Computing, Inc.
[e-mail]

+++++++++++++++++++++++++++

From: [e-mail] (Michael Peirce)
Date: 14 Apr 92 00:10:57 GMT
Organization: Peirce Software


In article <[e-mail]> (comp.sys.mac.programmer), [e-mail] (Neal Trautman) writes:
>
> Is there any info on an Internetters dinner at
> Apple's World Wide Developer Conference?

Sure is.  It's been talked about some here already.  Thursday is the
night and Gorden Biersch is the place.  Send me email to have your
name put on the reservation list.

- --  Michael Peirce         --   [e-mail]
- --  Peirce Software        --   Suite 301, 719 Hibiscus Place
- --  Macintosh Programming  --   San Jose, California USA 95117
- --           & Consulting  --   voice: [phone] fax: [phone]
- --                         --   AppleLink: peirce & America Online: AFC Peirce

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

From: [e-mail] (Doug Frick)
Subject: archive site for Frontier stuff?
Date: 13 Apr 92 17:54:24 GMT
Organization: University of Hawaii at Manoa

I'm posting the following in hopes that someone out in Netland is
interested in supporting an archive site for Frontier-related
stuff. It is a pretty cool application, and allows all us Mac
techies to script our environments.
- ---------------------------------------------------------------------------
From: [e-mail] (UserLand SW, David Winer,PRT)

Doug:

We're not on the Internet at all. If you know anyone who might want to act as a
pipeline for demo scripts, toolkits and install files, we'd love to open the
channel.

Is there some place you could re-post this message? We're looking for someone
with both an Internet and an AppleLink or CompuServe account.

This question comes up all the time and we'd really like to get in the loop.
Getting educated on Internet culture isn't an option for us right now, we're
overloaded with supporting AppleLink and CompuServe. That may change, but not
in the next few months...

DW
- --
Doug Frick
[e-mail]

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

From: [e-mail] (Adam Winer)
Subject: What is Discipline?
Date: 13 Apr 92 18:52:04 GMT


Does anyone out there know what Discipline is?  I downloaded it from
ftp.apple.com a couple days ago.  It appears to be some sort of a
debugging tool, but the instructions that came with it were
extremely skimpy.  Is it designed for use with Macsbug (which I have)
or with MPW (which I don't have)?  Does anyone have access to complete
instructions?

Thanks,
Adam
- --
Adam Winer        | The number you have reached is imaginary.
[e-mail]    | Please rotate your phone 90 degrees and
                        | try again.

+++++++++++++++++++++++++++

From: [e-mail] (Mark Lanett)
Organization: University of Illinois at Urbana
Date: Mon, 13 Apr 1992 20:24:55 GMT

[e-mail] (Adam Winer) writes:


>Does anyone out there know what Discipline is?  I downloaded it from
>ftp.apple.com a couple days ago.  It appears to be some sort of a
>debugging tool, but the instructions that came with it were
>extremely skimpy.  Is it designed for use with Macsbug (which I have)
>or with MPW (which I don't have)?  Does anyone have access to complete
>instructions?

It's a parameter checker for Toolbox calls; i.e. it will complain if you
pass obviously wrong stuctures to the Toolbox. For example, a rectangle
with right < left (the Finder tends to do this!). Turn it on in macsbug
with 'dsca'; the same command toggles it off. 'dsc' will turn it on/off for
the entire machine rather than only the current application, but you don't want
to do this (unless you feel like it complaining about the Finder).
- --
Mark Lanett, Software Tools Group, NCSA           [e-mail] or  NCSA.STG
"People wander in and out of virtual rooms in virtual settings with virtual
characters and virtual money and virtual armor and virtual weapons, which is
virtually a good way to spend time, but not quite." -- Usenet Oracle

+++++++++++++++++++++++++++

From: [e-mail] (John Nagle)
Date: Tue, 14 Apr 92 00:06:22 GMT
Organization: Netcom - Online Communication Services  (408 241-9760 guest)

[e-mail] (Adam Winer) writes:
>Does anyone out there know what Discipline is?  I downloaded it from
>ftp.apple.com a couple days ago.  It appears to be some sort of a
>debugging tool, but the instructions that came with it were
>extremely skimpy.  Is it designed for use with Macsbug (which I have)
>or with MPW (which I don't have)?  Does anyone have access to complete
>instructions?
      Discipline is documented, more or less, in the Macsbug manual,
which can be ordered from APDA.

      Discipline is an attempt to remedy the fact that the Mac toolbox,
operating system, and packages don't adequately check arguments on calls.
When started, Discipline reads in a file of interpretive statements
which define checks to be made for each system and toolbox call.
These checks are then made when checking is enabled with a Macsbug
command.  The checker is slow; turning on Discipline may slow programs
by an order of magnitude.  (If these checks were built into the Toolbox
code, they'd probably only cost a few percent performance).

      When Discipine detects an error, it causes a trap to Macsbug, with
a reasonable error message identifying the problem.

      Discipline doesn't distinguish clearly between calls which are
flatly wrong and calls which are returning an unusual result.  Attempting
to read in a resource that isn't present will cause a Discipline trap,
even for calls where this just returns an error code, without causing
"undefined results", as IM puts it.  This reflects a Lint-like approach
to error handling, where false alarms are a problem.

      Annoyingly, the Finder isn't clean enough to pass Discipline, and
running under MultiFinder (at least prior to System 7) with Discipline
on generates occasional traps.  This really ought to be fixed; there are
only a few places where the Finder fails Discipline checks.  (The
Macsbug manual has an amusing comment on this subject, along the
lines of "nothing is wrong, it's OK for Apple code to violate Discipline)
Haven't tried this under system 7 yet.

       Discipline ranks with "lint" as a concept; it is intended to
catch common errors, not to guarantee that specific classes of errors
have been eliminated.

                    John Nagle

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

From: [e-mail]
Subject: 32K boundary problem?
Date: 13 Apr 92 20:42:24 GMT
Organization: Management Graphics, Inc.

Can someone help me with this problem?

I'm programming in C using MPW 3.0 and have run into a problem related
the size of an array.  The program works as the following:

#define MAXLINES 4096

void routine(void) {
    rcPtr    scanList[MAXLINES];
...

but the program bombs on the call to _routine_ if MAXLINES is declared
as 8192.  (rcPtr is an 8 byte structure by the way).

This has to be a 32K boundary problem.  How do I get around it (short of
redefining my algorithm to eliminate the array, please).

Thanks in advance.
- --
******************************************************************************
* Woody Johnson     * [e-mail]   * I used to paint the princess*
* 10816 Quinn Ave S * Management Graphics, Inc * I used to paint the frogs   *
* Bloomington, MN   * 1401 E 79th St, #6       * Now paint mustaches          *
*           55437   * Bloomington, MN 55425    * on dangerous dogs         *
*             *                    *           - Don Henley *
******************************************************************************

+++++++++++++++++++++++++++

From: [e-mail] (Tim Endres)
Date: 14 Apr 92 02:27:48 GMT
Organization: ICE Engineering, Inc.


In article <[e-mail]> (comp.sys.mac.programmer), [e-mail] writes:
> #define MAXLINES 4096
>
> void routine(void) {
>     rcPtr    scanList[MAXLINES];
> ...
>
> but the program bombs on the call to _routine_ if MAXLINES is declared
> as 8192.  (rcPtr is an 8 byte structure by the way).
>
> This has to be a 32K boundary problem.  How do I get around it (short of
> redefining my algorithm to eliminate the array, please).

This is more likely your stack crawling down over the top of your
heap. Remember that the heap and stack grow towards eachother. What
you need to do is decrease the fire wall that limits the heap's growth,
which allows the stack to get bigger without clobbering it. Here is
the a sample of increasing your "stack size":

    /* Increase stack space by 16,000 bytes */

    SetApplLimit(GetApplLimit() - 16000);

    /* Expand the application heap zone to the maximum. */

    MaxApplZone();

    /* Allocate some master pointers. */

    MoreMasters();
    MoreMasters();
    MoreMasters();

    InitGraf(&qd.thePort);
    InitFonts();
    InitWindows();
    InitMenus();
    TEInit();
    InitDialogs(NULL);
    InitAllPacks();


tim endres - [e-mail]  -or-  uupsi!tbomb!time
ICE Engineering, Inc. - Phone [phone] - FAX [phone]
8840 Main Street, Whitmore Lake, MI  48189
USENET - a slow moving self parody... ph

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

End of C.S.M.P. Digest
**********************