Celestin Apprentice 2/Apprentice-Release2.iso/Information/Digests/CSMP Digest/volume 1/csmp-v1-215.txt

Text File | 1994-12-08 | 36.9 KB | 1,050 lines | [TEXT/R*ch]

C.S.M.P. Digest             Thu, 19 Nov 92       Volume 1 : Issue 215

Today's Topics:

    Event Manager Too Slow or izzit just me?
    Dynamic Change of Mouse Tracking Speed
    Anyone use SoftPolish?
    DisposeHandle vs. ReleaseResource
    Forcing TextEdit to set scrpAscent and scrpHeight
    Forcing Sys6 on a Q700?



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

The digest is a collection of article threads 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.  You can post articles to any newsgroup by
mailing your article to [e-mail].  So, to post an
article to comp.sys.mac.programmer, you mail it to
[e-mail].  Note the '-' instead of '.'
in the newsgroup name.

Each issue of the digest contains one or more sets of articles (called
threads), with each set corresponding to a 'discussion' of a particular
subject.  The articles are not edited; all articles included in this digest
are in their original posted form (as received by our news server at
cs.uoregon.edu).  Article threads are not added to the digest 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 digest).  Article threads that
consist of only one message are generally not included in the digest.

The entire digest is available for anonymous ftp from ftp.cs.uoregon.edu
[128.223.8.8] in the directory /pub/mac/csmp-digest.  Be sure to read the
file /pub/mac/csmp-digest/README before downloading any files.  The most
recent issues are available from sumex-aim.stanford.edu [36.44.0.6] in the
directory /info-mac/digest/csmp.  If you don't have ftp capability, the sumex
archive has a mail server; send a message with the text '$MACarch help' (no
quotes) to [e-mail] for more information.

The digest is 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 issue as it is created.  Sorry, back issues
are not available through the mailing list.

Send administrative mail to [e-mail].


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

From: [e-mail] (GILPIN, W.W.)
Subject: Event Manager Too Slow or izzit just me?
Organization: Texas A&M University, Academic Computing Services
Date: Tue, 13 Oct 1992 04:49:00 GMT

Here is some code from a little game thingy that I'm writing. It's part
of the event loop I'm using for when the game is actually being played. The
idea is that the little ship moves around on the screen and the player can
control whether it moves up, down, left or right. From the discussion here
I was expecting flicker problems but I haven't really experienced any of those.
(yet...knock on wood)

The problem I *am* having, though, is the movement is jumpy. I've traced
the jumpiness to this code:

    while (! gameDone)
    {
        if (gWNEImplemented)
            WaitNextEvent(everyEvent, &gTheEvent, SLEEP,
                NIL_MOUSE_REGION);
        else
        {
            SystemTask();
            GetNextEvent(everyEvent, &gTheEvent);
        }
        if (gTheEvent.what == keyDown)
            gameDone = HandleGameKey(&newDirection);
        .
        .
    (move the ship and/or change directions)
        .
        .
     }

Note:
#define SLEEP            0L
#define NIL_MOUSE_REGION 0L


Without this code the ship moves smooth as glass...but I can't control it,
which doesn't help. I've tried just using keyDownMask instead of everyEvent but
that didn't help either. My questions are (1) Is there something I'm doing
wrong or something missing that should be painfully obvious? and (2) Is there
a faster way of getting keyboard input?

If I left anything out that would make the problem easier to slove lemme
know...

Thanx
- -------------------------------------------------------------------------------
| Wes Gilpin              | I'd kill myself for you...I'd kill you for myself |
| WWG2101@TAMZEUS(BITNet) |------------------------------------------         |
| [e-mail]   | Southwest Knights of Nee Paintball Team | PANTERA |
- -------------------------------------------------------------------------------

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

Date: 13 Oct 92 09:50:12 GMT
Organization: Royal Institute of Technology, Stockholm, Sweden

> [e-mail] (GILPIN, W.W.) writes:

   The problem I *am* having, though, is the movement is jumpy. I've traced
   the jumpiness to this code:

           if (gWNEImplemented)
               WaitNextEvent(everyEvent, &gTheEvent, SLEEP,
                   NIL_MOUSE_REGION);
           else
           {
               SystemTask();
               GetNextEvent(everyEvent, &gTheEvent);
           }


Okay, why are you checking for WaitNextEvent? Want to run
under system 4.2 without MultiFinder? WaitNextEvent is ALWAYS
implemented if
1) you run system 6.0 or higher
or
2) you run MultiFInder

I did away with checking a LOOOONG time ago. (Now, people, what's
a good way of checking for MultiFinder? :-) :-) :-)

           if (gTheEvent.what == keyDown)
               gameDone = HandleGameKey(&newDirection);
           .

You inline the event handling dispatching. That's bad style,
because if you call WaitNextEvent some other place, or get an
update event within a filterProc, you want to handle that event
using your normal event handling functions.

Separate out the dispatching (and use a switch for chrissake :-)
into a HandleEvent ( EventRecord * event ) function instead.

Now, about your jerkyness problem; you'll lose background tasking
one way or the other:

1) System 6 way:
    Call GetOSEvent instead of WaitNextEvent. This will totally
    freeze up all background tasks, but will yield snappy
    performance for you. Update events and other such unimportant
    stuff won't get in your way, and clicks outside your window
    won't switch you out.
2) System 7 way:
    Ask the user "Would you like jerky motion and living processes,
    or should I kill the f*cking cycle-stealers ?"
    Then, if the user opts for violence, use the process manager
    to kill off all tasks except yourself. WaitNextEvent should
    be snappy enough after that.

Hope this helps,

- --
 -- Jon W{tte, [e-mail], Mac Hacker Suedoise (not french speaking) --

   There's no problem that can't be solved using brute-force algorithms
   and a sufficiently fast computer. Ergo, buy more hardware.

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

From: [e-mail] (Scott Bronson)
Date: 13 Oct 92 18:11:39 GMT

In <[e-mail]> [e-mail] (Jon Wtte) writes:

>> [e-mail] (GILPIN, W.W.) writes:

>Okay, why are you checking for WaitNextEvent? Want to run
>under system 4.2 without MultiFinder? WaitNextEvent is ALWAYS
>implemented if
>1) you run system 6.0 or higher
>or
>2) you run MultiFInder

Hey, I still check simply because it takes no time, is not hard at
all, and I don't really like leaping off unimplemented traps if the
user is running under an insufficient system.  But, that's not the
purpose of this followup.


>1) System 6 way: GetOSEvent.
>2) System 7 way: Kill all apps but me.

Or, finally, a technique that works under all systems that I know
of: don't use an event loop.  Simply use GetKeys and Button to
figure out what you want to do.  This gives absolutely no time
to anyone else, so you'll cause modem programs to overflow buffers
and wreak all kinds of havoc, but your game will be as smooth as
silk.  I may catch flak for this for some reason, but it has worked
for me.

    - Scott

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

From: [e-mail] (Jamie R. McCarthy)
Organization: Kalamazoo College
Date: Tue, 13 Oct 1992 19:48:21 GMT

[e-mail] (Scott Bronson) writes:
>[e-mail] (Jon Wtte) writes:
>>
>>Okay, why are you checking for WaitNextEvent?
>
>Hey, I still check simply because it takes no time,

You are obviously using a meaning of the word "no" with which I am
not familiar.  :-)

>>1) System 6 way: GetOSEvent.
>>2) System 7 way: Kill all apps but me.
>
>Or, finally, a technique that works under all systems that I know
>of: don't use an event loop.  Simply use GetKeys and Button...

Allow me to provide the obligatory reminder to anyone who's going to do
this:  it's considered polite to inform the user that you're deactivating
the rest of her computer.  I suggest a notice on a splash screen or
options screen, something like:  "To obtain maximum smoothness, this
arcade game will temporarily stop the rest of the computer.  For
example, file sharing and modem transfers will not work while the game
is in play." (This suitably-newbie message should suffice both for games
that don't give time, and for games that kill all other processes, I
think.)
- --
 Jamie McCarthy      Internet: [e-mail]      AppleLink: j.mccarthy
 "Magic Johnson says he attended two scheduled meetings [of] the president's
  AIDS commission, one more than Mr. Bush credited him with during Sunday's
  debate [and] one more than any of the three members of the Bush cabinet..."

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

From: [e-mail] (Jim Reekes)
Date: 13 Oct 92 19:57:15 GMT
Organization: Apple Computer, Inc.

In article <[e-mail]>,
[e-mail] (Jon Wtte) wrote:
>
> > [e-mail] (GILPIN, W.W.) writes:
>
>    The problem I *am* having, though, is the movement is jumpy. I've traced
>    the jumpiness to this code:
>
>            if (gWNEImplemented)
>                WaitNextEvent(everyEvent, &gTheEvent, SLEEP,
>                    NIL_MOUSE_REGION);
>            else
>            {
>                SystemTask();
>                GetNextEvent(everyEvent, &gTheEvent);
>            }
>
> [ some stuff deleted ]
>
> Now, about your jerkyness problem; you'll lose background tasking
> one way or the other:
>
> 1) System 6 way:
>     Call GetOSEvent instead of WaitNextEvent. This will totally
>     freeze up all background tasks, but will yield snappy
>     performance for you. Update events and other such unimportant
>     stuff won't get in your way, and clicks outside your window
>     won't switch you out.
> 2) System 7 way:
>     Ask the user "Would you like jerky motion and living processes,
>     or should I kill the f*cking cycle-stealers ?"
>     Then, if the user opts for violence, use the process manager
>     to kill off all tasks except yourself. WaitNextEvent should
>     be snappy enough after that.

I'd like to point out at this time that while some developers want this,
others are asking Apple for pre-emptive multitasking.  The above would not
be possible with a pre-emptive OS.  Now, personally I'm not in favor of
having pre-emptive OS on my personal computer.  I just have to point out
that one cannot have it both ways.  The above idea is a strong argument in
favor of not using a pre-emptive OS.

- -----------------------------------------------------------------------
Jim Reekes, Polterzeitgeist  |     Macintosh Toolbox Engineering
                             |          Sound Manager Expert
Apple Computer, Inc.         | RAll opinions expressed are mine, and do
20525 Mariani Ave. MS: 81-KS |   not necessarily represent those of my
Cupertino, CA 95014          |       employer, Apple Computer Inc.S

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

From: [e-mail] (GILPIN, W.W.)
Organization: Texas A&M University, Academic Computing Services
Date: Tue, 13 Oct 1992 13:17:00 GMT

>Okay, why are you checking for WaitNextEvent? Want to run
>under system 4.2 without MultiFinder? WaitNextEvent is ALWAYS

Never underestimate a business major!  :) :) That's not where the slow down is
anyway. I tried it without checking and just using WNE too.

>           if (gTheEvent.what == keyDown)
>               gameDone = HandleGameKey(&newDirection);
>
>You inline the event handling dispatching. That's bad style,
>because if you call WaitNextEvent some other place, or get an
>update event within a filterProc, you want to handle that event
>using your normal event handling functions.

>Separate out the dispatching (and use a switch for chrissake :-)
>into a HandleEvent ( EventRecord * event ) function instead.

Okay, I feel the need to defend myself here cause I'm not always this sloppy.
I wrote this code quick and dirty and since right now the only thing I'm
interested in is the 4 keydowns for the direction of the player and the 1
other keydown for stopping the game. I just wanted to see if I could get the
bugger to move around on the screen, dontchaknow? :)  My other event loop
for when the game ain't runnin' has a nice big switch right where it
should be.

>Now, about your jerkyness problem; you'll lose background tasking
>one way or the other:

Not a problem.

>1) System 6 way:
.
.
>2) System 7 way:
.
.
>Hope this helps,

Most definately! Thanks!

- -------------------------------------------------------------------------------
| Wes Gilpin              | I'd kill myself for you...I'd kill you for myself |
| WWG2101@TAMZEUS(BITNet) |------------------------------------------         |
| [e-mail]   | Southwest Knights of Nee Paintball Team | PANTERA |
- -------------------------------------------------------------------------------

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

From: [e-mail] (Dana S Emery)
Date: 15 Oct 92 11:15:51 GMT
Organization: Personal

In article <REEKES-131092125541@90.10.20.67>, [e-mail]
(Jim Reekes) wrote:
>
> In article <[e-mail]>,
> [e-mail] (Jon Wtte) wrote:
> > [ some stuff deleted ]
> >
> > 1) System 6 way:
> >     Call GetOSEvent instead of WaitNextEvent. This will totally
> >     freeze up all background tasks, but will yield snappy
> >     performance for you. Update events and other such unimportant
> >     stuff won't get in your way, and clicks outside your window
> >     won't switch you out.
> > 2) System 7 way:
> >     Ask the user "Would you like jerky motion and living processes,
> >     or should I kill the f*cking cycle-stealers ?"
> >     Then, if the user opts for violence, use the process manager
> >     to kill off all tasks except yourself. WaitNextEvent should
> >     be snappy enough after that.
>

How about the Miss Manners way?  check for ongoing processes, and refuse to
torpedo the user untill the machine is clean (7), or is prounced clean (6,
- - ).  This will force the user to recognize what the machine state really
*is*, rather than what she *thinks* it is, and will allow an orderly USER
CONTROLLED shutdown of possibly interdependant processes.
- --

Dana S Emery <[e-mail]> | "Novo, de Novo,
                                  |     de novo, de no-o-o-o-o---,
                                  | Novemba come an' dey gonna go home."

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

Date: 16 Oct 92 11:59:13 GMT
Organization: Royal Institute of Technology, Stockholm, Sweden

> [e-mail] (Dana S Emery) writes:

   > >     Ask the user "Would you like jerky motion and living processes,
        ~~~~~~~~~~~~

   *is*, rather than what she *thinks* it is, and will allow an orderly USER
   CONTROLLED shutdown of possibly interdependant processes.

Maybe one could add a list of processes to the question,
but as long as the question has a "don't kill" button,
it's a user-in-control situatino.

After all, if the user is unaware of that 400k dissertation
in that hidden Word window, does he/she need it? :-)

- --
 -- Jon W{tte, [e-mail], Mac Hacker Suedoise (not french speaking) --
  _/~|   Yellow
 / * \_  Shark                      (This signature has won the "Worst ASCII
 ~~~~\/  Software                    Logo of the Year" award)

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

From: [e-mail] (Repenning Alexander)
Subject: Dynamic Change of Mouse Tracking Speed
Date: 13 Oct 92 19:30:09 GMT
Organization: University of Colorado, Boulder


I am dealing with the visual representation of complex data structures.
I like to experiment with the Mouse Tracking Speed to make it easier
for users to track on important information: "soft snapping".

Is there some simple way to change the mouse tracking speed (e.g. via
a trap)?  What I need is a function allowing me to adjust mouse
tracking to speeds like the ones avaivable in the MOUSE CDEV (very
slow, .., fast). How is the MOUSE CDEV doing it?

Any ideas, clues?

  Thanks,  Alex

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

From: thepope@bigboy (Michael Kohne)
Date: 15 Oct 92 00:27:33 GMT
Organization: Temple University

[e-mail] (Repenning Alexander) writes:
:
: I am dealing with the visual representation of complex data structures.
: I like to experiment with the Mouse Tracking Speed to make it easier
: for users to track on important information: "soft snapping".
:
: Is there some simple way to change the mouse tracking speed (e.g. via
: a trap)?  What I need is a function allowing me to adjust mouse
: tracking to speeds like the ones avaivable in the MOUSE CDEV (very
: slow, .., fast). How is the MOUSE CDEV doing it?
:
: Any ideas, clues?
:
:   Thanks,  Alex

I don't know how to change the mouse tracking speed, but I do know that it is
more complex than 1 number. If you can find someone with a Kensington Turbo
Mouse, and the Turbo Mouse control panel, have them show you the 'custom'
settings. Basically, the mouse tracking speed is a series of 8 thresholds
that tell the system how to repond to mouse movements of varying lengths and
speeds. I don't know where these numbers reside in memory during run-time,
however. It may be that you could simply use macsbug to figure out what
resource the Mouse Control panel is messing with in the system.

"Pope" Q.E.D
Michael Kohne
[e-mail]
House of the Techno-Discordians, Archery Amusements Division
<twang!>

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

From: [e-mail]
Date: 16 Oct 92 07:16:10 GMT
Organization: Wright State University

> [e-mail] (Repenning Alexander) writes:
> :
> : I am dealing with the visual representation of complex data structures.
> : I like to experiment with the Mouse Tracking Speed to make it easier
> : for users to track on important information: "soft snapping".
> :
> : Is there some simple way to change the mouse tracking speed (e.g. via
> : a trap)?  What I need is a function allowing me to adjust mouse
> : tracking to speeds like the ones avaivable in the MOUSE CDEV (very
> : slow, .., fast). How is the MOUSE CDEV doing it?
> :
> : Any ideas, clues?

Alley & Strange, _ResEdit Complete_, from the Macintosh Inside Out series,
discusses mouse thresholds stored in 'mcky' resources. There's a nice
explanation of what the values mean, but nothing about how to change
them on the fly (except to edit the active system file and touch the
control panel).

I notice that booting System 5.1 messes up mouse tracking, as does the
return to System 7 after fixing it in 5.1. I'm going to go out on a limb
here and speculate that mouse tracking is system dependent.

After many folks begged to be able to change screen depth on the fly,
DTS produced HasDepth and SetDepth. I wonder if something analogous
for mouse tracking might be possible.

o----------------------------------------------------------------------------o
| John B. Matthews, [e-mail], disclaimer:= myViews <> WSU |
|      "Whom the gods would destroy, they first invite to program in C"      |
o----------------------------------------------------------------------------o

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

From: [e-mail] (Carl W. Haynes III)
Subject: Anyone use SoftPolish?
Organization: Purdue University
Date: Fri, 16 Oct 1992 17:46:20 GMT



I've been thinking about purchasing SoftPolish, but have never
seen any comments about it.

Has anyone used it? I'd be interested in any comments, especially:

1) Is the dictionary customizable (can I add entries?)
2) Can you spell check only part of a resource?, I have a number
   of resources that are part text and part bitmap.
3) Is it generally easy to use, do you get lots of useless
   comments about your resources?
4) Can you specify which resource types to check (i.e. check
   some but not others?

As you can see, I'm mainly interested in the spell checking part of
it. So any comments about that aspect are especially appreciated.

carl


- --
Carl W. Haynes III             ||  "She'll see through me like Grandma's
Buddy can you spare a job?     ||   underpants."
[e-mail]      ||                  -- Bart Simpson
[e-mail]        ||


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

From: [e-mail] (Brian Bechtel)
Date: Fri, 16 Oct 1992 21:35:07 GMT
Organization: Apple Computer Inc.

[e-mail] (Carl W. Haynes III) writes:

>I've been thinking about purchasing SoftPolish, but have never
>seen any comments about it.

>1) Is the dictionary customizable (can I add entries?)

Yes.

>2) Can you spell check only part of a resource?, I have a number
>   of resources that are part text and part bitmap.

>3) Is it generally easy to use, do you get lots of useless
>   comments about your resources?

You set the level of comments you want to receive.

>4) Can you specify which resource types to check (i.e. check
>   some but not others?

Yes.

In general, this is a WONDERFUL product.  It polishes up your grammar
and spelling.  It checks for resource validity.  It checks for missing
parts of a set of resources (e.g. DLOG without a DITL, DITL without a
DLOG.)  It does a lot more.

For more information, send a message to [e-mail] or
call them at [phone].

- --Brian Bechtel     [e-mail]     "My opinion, not Apple's"


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

From: [e-mail] (Marc Lesure)
Subject: DisposeHandle vs. ReleaseResource
Organization: Arizona State University
Date: Thu, 15 Oct 1992 20:20:20 GMT

I have a routine that deletes, adds, and updates a resource in a
preference file.  The function performs what it is suppose to do, however
it doesn't seem to release the memory (if I call the routine enough times,
it dies with a dsMemFullErr).  I understand that you need to use
ReleaseResource(), if the allocated handle came from GetResource().  But
what if the handle is created via a NewHandle() call?

Here is the snippet:

writePrefs(void)
{
    Handle        aH;
    int        resFile,tmpResFile;

    tmpResFile = CurResFile();
    resFile = openPrefs();
    UseResFile(resFile);

/* delete old resource */
    aH = Get1Resource(TYPE, baseID);
    if (aH != NIL)
    {
        RmveResource(aH);
        UpdateResFile(resFile);
        ReleaseResource(aH);
        aH = NIL;
    }

/* add new resource */
    if (aH == NIL)
    {
        aH = NewHandle(...);
        if (aH == NIL)
            return;
        AddResource(aH, TYPE, baseID, NAME);
    }

/* update resource */
    if (ResError() == noErr)
    {
        HLock(aH);
        ... fill handle ...
        HUnlock(aH);

        ChangedResource(aH);
        WriteResource(aH);
    }

    ReleaseResource(aH); /* or DisposeHandle or Both? */

    UseResFile(tmpResFile);
    CloseResFile(resFile);
}

- -----------------------------------------------------------------------
Marc Lesure / Arizona State University / Tempe, AZ
"Between the world of men and make-believe, I can be found..."
"False faces and meaningless chases, I travel alone..."
"And where do you go when you come to the end of your dream?"

UUCP:       ...!ncar!noao!asuvax!lesure
Internet:   [e-mail]

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

From: [e-mail] (Marshall Clow)
Date: 15 Oct 92 20:55:45 GMT
Organization: Hewlett Packard San Diego Printer Division

In article <[e-mail]>,
[e-mail] (Marc Lesure) wrote:
>
> I have a routine that deletes, adds, and updates a resource in a
> preference file.  The function performs what it is suppose to do, however
> it doesn't seem to release the memory (if I call the routine enough times,
> it dies with a dsMemFullErr).  I understand that you need to use
> ReleaseResource(), if the allocated handle came from GetResource().  But
> what if the handle is created via a NewHandle() call?
>
> [ code deleted ]

Rule of thumb: If a handle refers to a resource, then use ReleaseResource.
If it doesn't, then use DisposHandle. It doesn't matter how it was created.

DetachResource and RmveResource turn resource handles into a
non-resource handle. AddResource turns a non-resource handle into a
resource handle.

Even bigger rule of thumb: Check for errors!!!

Here lies one problem in your code:
>        RmveResource(aH);
>     UpdateResFile(resFile);
>     ReleaseResource(aH);
ReleaseResource does nothing if aH is not a handle to a resource.
ResError will tell you this. (See rule #2)


Marshall Clow
San Diego Printer Division      Hewlett Packard
Internet: [e-mail]   AppleLink: HP.Marshall    AOL: MClow

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

From: thepope@bigboy (Michael Kohne)
Date: 15 Oct 92 23:38:59 GMT
Organization: House of the Techno-Discordians


[e-mail] (Marc Lesure) writes:
: I have a routine that deletes, adds, and updates a resource in a
: preference file.  The function performs what it is suppose to do, however
: it doesn't seem to release the memory (if I call the routine enough times,
: it dies with a dsMemFullErr).  I understand that you need to use
: ReleaseResource(), if the allocated handle came from GetResource().  But
: what if the handle is created via a NewHandle() call?
:
<Some code deleted>
: /* add new resource */
:     if (aH == NIL)
:     {
:         aH = NewHandle(...);
:         if (aH == NIL)
:             return;
:         AddResource(aH, TYPE, baseID, NAME);
At this point you will need to ReleaseResource. I always do a ChangedResource,
then a WriteResource first, but the ChangedResource might not be necessary.
:     }

: -----------------------------------------------------------------------
: Marc Lesure / Arizona State University / Tempe, AZ
: "Between the world of men and make-believe, I can be found..."
: "False faces and meaningless chases, I travel alone..."
: "And where do you go when you come to the end of your dream?"
:
: UUCP:       ...!ncar!noao!asuvax!lesure
: Internet:   [e-mail]

Michael Kohne
[e-mail]
House of the Techno-Discorians, Ammunition Amusements Division
<bang!>

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

From: [e-mail] (Ivan M CaveroBelaunde)
Organization: The World Public Access UNIX, Brookline, MA
Date: Fri, 16 Oct 1992 04:13:33 GMT

[e-mail] (Marc Lesure) writes:

>I have a routine that deletes, adds, and updates a resource in a
>preference file.  The function performs what it is suppose to do, however
>it doesn't seem to release the memory (if I call the routine enough times,
>it dies with a dsMemFullErr).  I understand that you need to use
>ReleaseResource(), if the allocated handle came from GetResource().  But
>what if the handle is created via a NewHandle() call?

Couple of problems here.  First of all, you don't need to call
ReleaseResource if you've removed it already from your resource map.
In fact, ReleaseResource will do nothing. Thus you need to call
DisposHandle.

Second, just to be on the safe side, you probably should SetResLoad(false)
before getting the resource you're removing. Otherwise you might get
NIL back from Get1Resource, not because it didn't find it, but because it
could not be loaded into memory.  If this happens, then you won't be
able to perform the AddResource properly (in fact, you'll end up with
two resources with the same type and ID in that file - a BAD thing).
Don't forget to SetResLoad(true) immediately after the Get1Resource.

>/* delete old resource */
    SetResLoad(false);
>    aH = Get1Resource(TYPE, baseID);
    SetResLoad(true);
>    if (aH != NIL)
>    {
>        RmveResource(aH);
>        UpdateResFile(resFile);
        DisposHandle(aH);
    /*    Do even if aH points to a purged handle to free the
        master pointer.    */
/*    Comment out
>        ReleaseResource(aH);
*/
>        aH = NIL;
>    }

>/* add new resource */
>    if (aH == NIL)
>    {
>        aH = NewHandle(...);
>        if (aH == NIL)
>            return;
>        AddResource(aH, TYPE, baseID, NAME);
>    }

You might want to build this handle before toasting the old Prefs.
In general, I'd say it's bad form to nuke the old prefs and then be
unable to write out the new prefs because you were unable to allocate
space for the new prefs handle.
>
>/* update resource */
>    if (ResError() == noErr)
>    {
>        HLock(aH);
>        ... fill handle ...
>        HUnlock(aH);
>
>        ChangedResource(aH);
>        WriteResource(aH);
>    }

>    ReleaseResource(aH); /* or DisposeHandle or Both? */

Just use release resource.

>    UseResFile(tmpResFile);
>    CloseResFile(resFile);

Your UseResFiles are a little mixed up. The only reason you'd need to do
the tempResFile bit at top is if the prefs res file was open in the first
place (if it's closed it'd become the current resFile when you open it,
so there would be no need to do UseResFile). If it was already open,
however, then you probably don't want to close it on exit.

Finally, it is also possible to do the following:
    Open the prefs file
    SetResLoad(false)
    prefH = Get1Resource(prefs)
    SetResLoad(true)
    if (!prefH) {
        newlyAdded = true;
        prefH = NewHandle()
        if (!prefH) return(MemError())
        AddResource(prefH, etc)
    } else newlyAdded = false;
    LoadResource(prefH)
    if (StripAddress(*prefH)) {
        fillPrefsHandle(prefH)
        ChangedResource(prefH)
        WriteResource(prefH)
    } else {
        RemoveResource(prefH)
        DisposHandle(prefH)
    }

Something like this would modify the prefs handle already in the resource
map. It seems a little cleaner, IMHO.

Hope this helps,

- -Ivan
- -----
Ivan Cavero Belaunde ([e-mail])
DiVA Corporation

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

From: [e-mail]
Subject: Forcing TextEdit to set scrpAscent and scrpHeight
Organization: Seattle Online Public Access Unix.  [phone] (all lines telebit-equipped)
Date: Sun, 11 Oct 1992 01:32:21 GMT

I am writing a Windows version of a Mac application. It uses the same file
format as the Mac version. This file format uses style scrap records to
store information about the text styles. So on the Windows end, I have to
generate scrap style records myself without the help of TextEdit. The
problem is setting the scrpHeight and scrpAscent fields properly. Since
I have no way of obtaining font metrics, it seems like I am stuck. My
question is this: is there some "magic" value for these fields that will
force text edit to read the font metrics and fill them in properly? I
tried -1, but this did not do it.

If there is no magic value, I will have to modify the Mac program to
get it to reread the font metrics if the file is coming from Windows.
What is the easiest way to get TextEdit to do this? Interestingly
enough, TESetSelect rereads these fields, but this seems like the
wrong function for this purpose.

If you have any answers or ideas, I would be much obliged. If you can figure
out a method that does not require any changes on the Mac end, I will mention
your name in the "Special Thanks To..." section of the manual! :-)

- --
Tony Mann
Spontaneous Software

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

Date: 13 Oct 92 09:53:34 GMT
Organization: Royal Institute of Technology, Stockholm, Sweden

> [e-mail] writes:

   I am writing a Windows version of a Mac application. It uses the same file

Gee, aren't we all? :-)

   format as the Mac version. This file format uses style scrap records to
   store information about the text styles. So on the Windows end, I have to

Sorry, but style scrap records are WHOLLY INSUITABLE for temporary
storage, since it refers to fints by number. Font numbers change
between machines, and even ON machines sometimes. Store font info
by name. When you do this, you'll see that the other things you
do will fall out naturally.

You might also want to set up a mapping from windows/mac fonts; so
the user can choose to substitute appropriate fonts from platform
to platform.

- --
 -- Jon W{tte, [e-mail], Mac Hacker Suedoise (not french speaking) --

   There's no problem that can't be solved using brute-force algorithms
   and a sufficiently fast computer. Ergo, buy more hardware.

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

From: [e-mail] (Dana S Emery)
Date: 15 Oct 92 11:02:29 GMT
Organization: Personal

In article <[e-mail]>,
[e-mail] (Jon Wtte) wrote:
>
> Sorry, but style scrap records are WHOLLY INSUITABLE for temporary
> storage, since it refers to fints by number. Font numbers change
> between machines, and even ON machines sometimes. Store font info
> by name. When you do this, you'll see that the other things you
> do will fall out naturally.


isn't that a bit draconian?  I agree that no code should depend on fond ID
info, but I would use the style scrap record plus a font name table with
the fond id as key, and the font baseline/ascent/descent info as well.

Defering the fondID resolution to the reader seems a nicely conservative
approach.

Readout would involve a resolution of the (possibly) different id's, and
haveing them available helps to clarify not just ID collision resolutions,
but also diferent font/same name/different ID confusions.  (Too bad so many
utilitys "protect" the user from such "confusing" information.)  I would
compare the fontname saved to the fontname available for the saved ID, and
ask the user for a substitute when things dont check out.  Noisy? yes, but
I (as a user) like things that way.

's too bad that TE doesn't handle -1 in those fields, I was hopeing for
that to be true for a presently backburnured project.  :-(

- --

Dana S Emery <[e-mail]> | "Novo, de Novo,
                                  |     de novo, de no-o-o-o-o---,
                                  | Novemba come an' dey gonna go home."

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

From: [e-mail] (Kenneth Hughes)
Subject: Forcing Sys6 on a Q700?
Date: 14 Oct 92 04:53:58 GMT
Organization: School of Computer Science, Carnegie Mellon

Anyone know the inside technical details as to why a Q700 (or any other
System-7-only platform) cannot run System 6.0.x ?  I'm wondering how feasible
it'd be to hack something up so that I can test and debug some code under
System 6 without having to pick up another Mac besides my Q700.

Along those lines, does anyone know the current break-down of System 7 vs
System 6 users?  Has it passed 50/50 yet?  Thanks for any info.

Kenneth J. Hughes     ([e-mail])

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

From: [e-mail] (Mats Bredell)
Organization: Uppsala University Computing Center (UDAC)
Date: Thu, 15 Oct 1992 08:12:27 GMT

Kenneth Hughes ([e-mail]) wrote:
: Anyone know the inside technical details as to why a Q700 (or any other
: System-7-only platform) cannot run System 6.0.x ?  I'm wondering how feasible
: it'd be to hack something up so that I can test and debug some code under
: System 6 without having to pick up another Mac besides my Q700.
:
: Along those lines, does anyone know the current break-down of System 7 vs
: System 6 users?  Has it passed 50/50 yet?  Thanks for any info.

There's a version of system 6 that works on the PowerBooks. I think this also
works on the Quadras. The version is called 6.0.8L.

/Mats
- --
Mats Bredell                                   [e-mail]
Uppsala University Computing Center (UDAC)     Ph:  [phone]
Department of medical systems                  Fax: [phone]
Sweden                                         Think straight - be gay!

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

From: [e-mail] (Kenneth Hughes)
Date: 15 Oct 92 20:10:25 GMT
Organization: School of Computer Science, Carnegie Mellon

Here's a summary of the responses I received on this topic:

Since 7.0.1 has ROM bug fixes for the Quadras (that of course are missing
in System 6 since the Quadra's came out after the last version of System 6
had been released), it would not really be feasible to attempt to hack
something to make System 6 run on the Quadras.

[e-mail] wrote:
> There's a version of system 6 that works on the PowerBooks. I think this also
> works on the Quadras. The version is called 6.0.8L.

According the the 10-5-92 article in MacWEEK:
"The new-found system is labeld 6.0.8L.  It supports a limiteed set of Macs:
the PowerBook 100, the LC, LC II, Classic and Classic II.  Other models, such
as the Mac II, IIcx and IIci, are not supported at all."
    I didn't see anywhere else any mention of Quadra support.  If it's in
there (which I doubt) I'd love to hear about it.
    They do later refer to an "acute " desire for System 6, however.
This in fact had prompted my other question regarding what fraction of users
still use System 6.  I didn't get any numbers on this, but one respondent
pointed out that he thought the System 6 hold-outs were probably much less
likely to be buying software.  I'd love to get some real numbers on this if
any developer has any info to share.
    Thanks especially to Jon W{tte and Larry Rosenstein for their comments.

Kenneth J. Hughes     ([e-mail])
Entelechy Corporation

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

From: [e-mail] (Jamie R. McCarthy)
Date: 15 Oct 92 21:24:31 GMT
Organization: Kalamazoo College

[e-mail] (Kenneth Hughes) writes:
>
>...one respondent
>pointed out that he thought the System 6 hold-outs were probably much less
>likely to be buying software.  I'd love to get some real numbers on this if
>any developer has any info to share.

No numbers, but the other day I did see a bumper sticker that read
"I'm pro-System-6...and I purchase!"
- --
 Jamie McCarthy      Internet: [e-mail]      AppleLink: j.mccarthy
 I suppose ya don't think I was run over by a streetcar!

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

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