MacTech 1 to 12/MacTech-vol-1-12.toast/Reference/the cmsp digests ('94-'97)/csmp digest Vol 4 No 036

Text File | 1997-02-19 | 118KB | 3,370 lines

C.S.M.P. Digest             Tue, 18 Feb 97       Volume 4 : Issue 36

Today's Topics:

        1-bit depth GWorld has colour?(!)
        3D Info?
        Application-like CPs
        Controlling freePPP
        Determining if serial port in online?
        Faster line drawing routines?
        Floating palettes
        Is LINUX available for Macintosh?
        Is palette animation w- > 256 colors possible?
        Need INIT source code
        Objective C
        Registering app signatures and file types
        Serial port arbitration with OT
        Virtual Memory Question
        What does "register" means on variable declaration?
        Window Lists for Other Apps
        must scriptability be so hard?



The Comp.Sys.Mac.Programmer Digest is moderated by Mark Aiken
([e-mail]).

The digest is a collection of article threads from the internet
newsgroups comp.sys.mac.programmer.help, csmp.tools, csmp.misc and
csmp.games. It is designed for people who read news 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. If you don't have access to news, you
may still be able to post messages to the group by using a mail server
like anon.penet.fi (mail [e-mail] for more information).

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
ee.mcgill.ca).  Article threads are not added to the digest until the last
article added to the thread is at least two weeks 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 digests can be obtained by email, ftp or through the World Wide Web.

If you want to receive the digest by mail, send email to
[e-mail] with no subject and one of the following commands
as body:

    help                        Sends you a summary of commands
    subscribe csmp                      Adds you to the mailing list
    unsubscribe csmp                    Removes you from the list

Once you have subscribed, you will automatically receive each new
issue as it is created.

Back issues are available by ftp from Info-Mac mirror sites in the
per/csmp subdirectory, e.g.

  ftp://sumex-aim.stanford.edu/info-mac/per/csmp/

The contents of all back issues can be searched by accessing the
following URL, courtesy of Andrew Barry ([e-mail]):

    http://marvin.stattech.com.au/search.html

They can also be searched through the following URLs, thanks to
Tim Tuck ([e-mail]):

    http://wais.sensei.com.au/searchform.html
    wais://wais.sensei.com.au:210/csmp?

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

>From [e-mail] (Ian McCall)
Subject: 1-bit depth GWorld has colour?(!)
Date: Tue, 28 Jan 1997 17:48:29 +0000
Organization: The MultiMedia Corporation

Hi.

I have a colour picture, which I'd like to get down to mono. I believe I
can do this by creating an offscreen 1-bit depth GWorld, then drawing my
picture into it. Quickdraw should then take over and dither the picture
accordingly, or at least so I believe.

The trouble is, the picture doesn't seem to be getting dithered. When I
dump it out as a PICT file, it still contains colour.

Here's the code (the 'watermark' variables rather give away what the final
aim is...):




PicHandle     watermarkImage   = somePassedInParameter;

GDHandle      mainDevice;
GWorldFlags   worldSettings;
GWorldPtr     watermarkImageWorld;

PixMapHandle  watermarkImagePixMap;

Rect          watermarkPictureFrame = (**watermarkImage).picFrame;

OSErr         theResult;



// Get details about the main graphics device, and keep track of the current
// GWorld and specify how we'd like our GWorlds to behave
GetGWorld(&originalWorld, &mainDevice);
worldSettings = noNewDevice | useTempMem;

// Allocate an offscreen mono GWorld to hold the watermark image
if((theResult = (OSErr)NewGWorld(&watermarkImageWorld, 1,
    &watermarkPictureFrame, NULL, mainDevice, noNewDevice)) == noErr)
{
     // Keep track of this world's pix map, and lock it down whilst we're using
     // it
     watermarkImagePixMap = GetGWorldPixMap(watermarkImageWorld);
     LockPixels(watermarkImagePixMap);

     // Draw the watermarking image we've been given in this GWorld
     SetGWorld(watermarkImageWorld, NULL);
     ClipRect(&(watermarkImageWorld->portRect));
     DrawPicture(watermarkImage, &watermarkPictureFrame);
}



As far as I'm concerned, watermarkImagePixMap should contain a mono
version of watermarkImageWorld by now. It doesn't though.

Any clues?


Cheers,
Ian

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

>From [e-mail]
Date: Thu, 30 Jan 1997 08:50:03 -0600
Organization: (none)

In article <ianm-2801971748290001@194.70.62.242>, [e-mail] (Ian
McCall) wrote:

> Hi.
>
> I have a colour picture, which I'd like to get down to mono. I believe I
> can do this by creating an offscreen 1-bit depth GWorld, then drawing my
> picture into it. Quickdraw should then take over and dither the picture
> accordingly, or at least so I believe.
>
> The trouble is, the picture doesn't seem to be getting dithered. When I
> dump it out as a PICT file, it still contains colour.
>
> Here's the code (the 'watermark' variables rather give away what the final
> aim is...):
...
snip
...
> // Allocate an offscreen mono GWorld to hold the watermark image
> if((theResult = (OSErr)NewGWorld(&watermarkImageWorld, 1,
>     &watermarkPictureFrame, NULL, mainDevice, noNewDevice)) == noErr)
> {

The problem is the "noNewDevice" flag. Since you're not using a new device,
it ignores the depth you specify and uses the depth of the main device. Try
it without "noNewDevice." It's okay to make an offscreen device. And you
shouldn't need to do anything else besides remember to dispose of it later.
That's one reason GWorlds/GDevices are nice.

- Graham


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

>From Dave Howell <[e-mail]>
Date: Thu, 30 Jan 1997 23:19:53 -0800
Organization: Pablo Media

Ian McCall sez:
>
> Hi.

Hi Ian.

> I have a colour picture, which I'd like to get down to mono. I
> believe I can do this by creating an offscreen 1-bit depth GWorld,
> then drawing my picture into it. Quickdraw should then take over
> and dither the picture accordingly, or at least so I believe.
>
> The trouble is, the picture doesn't seem to be getting dithered.
> When I dump it out as a PICT file, it still contains colour.
>
> Here's the code (the 'watermark' variables rather give away what the
> final aim is...):
>
> [code brutally snipped]

The trouble is that you're using DrawPicture to draw into your 1-bit
PixMap. DrawPicture doesn't dither. You need to tell DrawPicture to draw
into a nice 16-bit or 24-bit direct-color PixMap, and then call CopyBits
(don't forget to add the constant ditherCopy to your copy mode
parameter) to perform a dithered copy into your 1-bit PixMap.
Alternatively, if you can require QuickTime for your application, you
might call DrawTrimmedPicture, which accepts a Boolean doDither
parameter.

Dave

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

>From David Snowdon <[e-mail]>
Subject: 3D Info?
Date: Thu, 30 Jan 1997 18:58:00 +1000
Organization: Concussive Software

Hi,

I am looking to start in 3D programming (with a view to eventually
making some cool games). I definitely want a true 3D environment...
Could someone please tell me what the best books to get are? The black
art of mac game programming is one name that has some up... What
should I get? I don't exactly have that much money to spend on books,
so what should I and what do I have to get?

Dave.

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

>From [e-mail] (Dan Posluns)
Date: Sun, 02 Feb 1997 19:49:17 -0400
Organization: TotalNet Inc.

In article <[e-mail]>, [e-mail] wrote:

> Hi,
>
> I am looking to start in 3D programming (with a view to eventually
> making some cool games). I definitely want a true 3D environment...
> Could someone please tell me what the best books to get are? The black
> art of mac game programming is one name that has some up... What
> should I get? I don't exactly have that much money to spend on books,
> so what should I and what do I have to get?
>
> Dave.

Black Art is probably your best place to start if you want to do true 3-D
as opposed to the more limited versions imposed on the Maradoom Forces
genre, since  that is the main focus of the entire book. Tricks of the Mac
Game Programming Gurus also has a pretty indepth section on QuickDraw 3D,
but having the limited computational resources that I do (680LC40) I
haven't really bothered to read through it. Still, there might be some
good stuff there.

Good luck,

Dan.

--
Dan Posluns - [e-mail], http://www.magic.ca/~kefka/domain.html
"What's the most important thing in life? To be free of obligations! Otherwise you lose the ability to gamble..." - Setzer, FF3
"Play with fire and you get burned." - Magus, Chrono Trigger
"I don't believe there are any bad people in this world. Just bad hearts and weak people. It's difficult for weak people to become strong. That's why we have to get rid of the bad hearts. Please understand." - Jean, Breath of Fire 2

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

>From [e-mail] (nick.c MacTech)
Date: Tue, 04 Feb 1997 00:04:03 -0800
Organization: MacTech Magazine



[e-mail] wrote:

>I am looking to start in 3D programming (with a view to eventually
>making some cool games). I definitely want a true 3D environment...
>Could someone please tell me what the best books to get are?


   Two books that you might find useful are:

     _Engines of Creation_ by Jonathan Blossom
       ISBN 1-878739-90-5

     _Fast Algorithms for 3-D Graphics_ by Georg Glaeser
       ISBN 0-387-84288-2

    Engines has a nice, gradual paced tutorial on basic
      3D programming that leads from the very basic mathematics
      of projection and graphics worlds, up through wireframes,
      polygon rendering, clipping, shading, and inline performance
      optimization for the mac.

    Glaeser, has some great code including reflections, depth
      buffering, complex objects, shadows, etc.  The code is
      object based, platform independent C++.  Nice stuff.

    Hope this helps,

____Nicholas C. DeMello, Ph.D.________________________________________

   Online at MacTech Magazine, the Journal of Macintosh Programming
     http://www.MacTech.com/
                                       _/   _/  _/  _/_/_/   _/   _/
  Chemistry: [e-mail]       _/_/ _/  _/  _/   _/  _/_/_/
    MacTech: [e-mail]      _/ _/_/  _/  _/       _/ _/
       http://www.chem.ucla.edu/~nick/   _/  _/   _/_/_/  _/   _/


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

>From [e-mail] (Nobukazu Toge)
Subject: Application-like CPs
Date: Sun, 19 Jan 1997 16:27:21 GMT
Organization: Netcom On-Line Services

Some control panels such as "AppleTalk" and "TCP/IP" have filetype
'cdev' and do have 'cdev' resources; yet they have their own menus
and appear to act like applications. In fact they have 'CODE'
resources as well. How do they make a combined use of those
resources? Is there any documentation on this?  I've scanned
thru dev CDs, but it's not obvious, and I'm curious.

--
Nobu Toge
     e-mails to  : [e-mail]

#include <StandardDisclaimers.h>

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

>From [e-mail] (Dave Polaschek)
Date: Tue, 21 Jan 1997 07:16:42 -0600
Organization: Polaschek Publishing

In article <[e-mail]>, [e-mail] (Nobukazu Toge) wrote:

> Some control panels such as "AppleTalk" and "TCP/IP" have filetype
> 'cdev' and do have 'cdev' resources; yet they have their own menus
> and appear to act like applications. In fact they have 'CODE'
> resources as well. How do they make a combined use of those
> resources? Is there any documentation on this?  I've scanned
> thru dev CDs, but it's not obvious, and I'm curious.

They were using undocumented tricks. The "correct" way to do this,
starting with System 7.6 is documented in Technote 1090, which is
at <http://17.126.23.20/dev/technotes/tn/tn1090.html>

-DaveP

--
Dave Polaschek - personal: [e-mail] or [e-mail]
PGP key and other spiffy things at <http://www.best.com/~davep/>

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

>From [e-mail] (chris-b)
Date: Thu, 23 Jan 1997 17:44:48 +1300
Organization: Auckland University HMU

In article <[e-mail]>,
[e-mail] (Dave Polaschek) wrote:

>In article <[e-mail]>, [e-mail] (Nobukazu Toge)
wrote:
>
>> Some control panels such as "AppleTalk" and "TCP/IP" have filetype
>> 'cdev' and do have 'cdev' resources; yet they have their own menus
>> and appear to act like applications. In fact they have 'CODE'
>> resources as well. How do they make a combined use of those
>> resources? Is there any documentation on this?  I've scanned
>> thru dev CDs, but it's not obvious, and I'm curious.
>
>They were using undocumented tricks.

These CPs simply have a stub 'cdev' code resource that calls
_LaunchApplication (not undocumented!) on their own file and then returns
to the Finder with an error that requests no error dialog.

Chris B
- ---------------------------------------------------------------------
NewZealand:AucklandUniversity:ComputerScience:HyperMediaUnit:ChrisBurns
Internet: [e-mail]
Phone:    [phone] x8976
Fax:      [phone]                         Async, Therefore I Am.
- ---------------------------------------------------------------------

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

>From [e-mail] (Robert E. Williams, Jr.)
Date: Thu, 23 Jan 1997 03:19:21 -0500
Organization: Enterprise Software

In article <[e-mail]>, [e-mail] (Nobukazu Toge) wrote:

> Some control panels such as "AppleTalk" and "TCP/IP" have filetype
> 'cdev' and do have 'cdev' resources; yet they have their own menus
> and appear to act like applications. In fact they have 'CODE'
> resources as well. How do they make a combined use of those
> resources? Is there any documentation on this?  I've scanned
> thru dev CDs, but it's not obvious, and I'm curious.

I'm not sure how Apple did it with the current crop of control panels, but
I do know that 7.6 has a couple of new file types, including 'appc' (or
something like that) that allows you to make your program an application,
but have it routed to the Control Panels folder when it's dragged to the
System Folder, just like 'cdev' files. I believe that they're even checked
for 'cdev' code during startup. They also have a new one that again is for
normal applications, but causes them to be routed to the Apple Menu Items
folder; this is intended as a replacement for the traditional DA.


Regards,
Bob
--
Robert E. Williams
Macintosh Developer
Enterprise Software
[e-mail]

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

>From [e-mail] (Magnus Thoor)
Subject: Controlling freePPP
Date: Thu, 30 Jan 1997 17:10:24 +0100
Organization: -

Hi,

I'm a novice-programmer attempting to do a control-strip module to start
and stop connections with freeppp (I know this already exists, but it's
more of an exercise). I've trouble finding what calls I should make to
open/close freeppp. I downloaded the source for MacPPP but for my
limited knowledge of C++ it was too sparsly documented to be of use. I
would appreciate any help I could get regarding this.

BTW, I'm trying to do this in C++ with Codewarrior 8, if that matters.

Regards,
Magnus

--
To reply via email, please edit from-field.

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

>From [e-mail] (Rob Friefeld)
Date: Thu, 30 Jan 1997 22:35:07 -0800
Organization: Delta Internet Services, Anaheim, Ca

In article <[e-mail]>,
[e-mail] (Magnus Thoor) wrote:

>Hi,
>
>I'm a novice-programmer attempting to do a control-strip module to start
>and stop connections with freeppp (I know this already exists, but it's
>more of an exercise). I've trouble finding what calls I should make to
>open/close freeppp. I downloaded the source for MacPPP but for my
>limited knowledge of C++ it was too sparsly documented to be of use. I
>would appreciate any help I could get regarding this.

Get "PPP Interfaces 1.2" from info-mac if you are using FreePPP 2.5 or
earlier. If you are using FreePPP 2.6, look in
<ftp:members.aol.com/freepppgrp/> for public interface routines.


                                      Rob Friefeld

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

>From [e-mail] (Magnus Thoor)
Date: Sat, 1 Feb 1997 16:56:19 +0100
Organization: -

Rob Friefeld <[e-mail]> wrote:

> In article <[e-mail]>,
> [e-mail] (Magnus Thoor) wrote:
>
> >-del- I've trouble finding what calls I should make to
> >open/close freeppp. - del-
>
> Get "PPP Interfaces 1.2" from info-mac if you are using FreePPP 2.5 or
> earlier. If you are using FreePPP 2.6, look in
> <ftp:members.aol.com/freepppgrp/> for public interface routines.
>

Thank you very much. "ppp interfaces" was of great help.

Thanks again,
Magnus

--
To reply via email, please edit from-field.

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

>From Steve Collins <[e-mail]>
Subject: Determining if serial port in online?
Date: Fri, 24 Jan 1997 09:31:53 -0800
Organization: Netcom

Hi, sorry if this is a FAQ, but...

How do I tell if a serial port is online?  I found some code
which depends on the SCC which looks something like:

I found some code that tells me if the "Modem Port" is online:

#define SCCRd            (*(Ptr *)0x1D8)

// PORTNO is 1 for Modem Port, perhaps 2 for Printer Port?

Boolean IsOnLine()
{
    register Byte r;

    r = *(SCCRd + ((PORTNO == 2) ? 0 : 2));    /* SCC channel A/B control */
    r = *(SCCRd + 2);    /* SCC channel A/B control */

    return (r & 0x8) ? TRUE : FALSE;

}

This seems to work for the Modem Port, haven't tried Printer, but I
guess
it'd work.   So here's the questions

1. Where is this documented? I'd like to understand it a bit better.

2. What do I do for PCMCIA slots?


Thanks in advance.

--
- -------------------------------------------------
Steve Collins                 [e-mail]
- -------------------------------------------------

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

>From [e-mail]
Date: Sat, 25 Jan 1997 06:42:16 -0600
Organization: (none)

In article <[e-mail]>, [e-mail] wrote:

> Hi, sorry if this is a FAQ, but...
>
> How do I tell if a serial port is online?  I found some code
> which depends on the SCC which looks something like:
>
> I found some code that tells me if the "Modem Port" is online:
>
> #define SCCRd                   (*(Ptr *)0x1D8)
>
> // PORTNO is 1 for Modem Port, perhaps 2 for Printer Port?
>
> Boolean IsOnLine()
> {
>         register Byte r;
>
>         r = *(SCCRd + ((PORTNO == 2) ? 0 : 2)); /* SCC channel A/B control */
>         r = *(SCCRd + 2);       /* SCC channel A/B control */
>
>         return (r & 0x8) ? TRUE : FALSE;
>
> }
>
> This seems to work for the Modem Port, haven't tried Printer, but I
> guess
> it'd work.   So here's the questions
>
> 1. Where is this documented? I'd like to understand it a bit better.
>
> 2. What do I do for PCMCIA slots?
>
>
> Thanks in advance.
>
> --
> ---------------------------------------------------
> Steve Collins                 [e-mail]
> ---------------------------------------------------

1. It's not, really. There's some info in some early issues of MacTech I think.

2. I have no idea about that one. But you could try using SRsrcInfo in the
Slot Manager. If that returns an error of -300 it means there's no card in
the slot you specify. If you really want to get into it, though, I'd read
the Slot Manager section of Inside Macintosh (or talk to someone who knows
more about slots).

Even though there's not much documentation on the above thing. Here's what
it means and also here's a better way to do it:

SCCrd is a low-memory global that stands for "SCC ReaD base address". It is
the starting address for SCC. As "HS" points out, this is reading from
hardware.

You should use a low-memory accessor to get to SCCrd. But since you use it
to read directly from hardware anyway, you might not be too concerned about
reading directly from low memory!

As long as you're using low-memory, here's an easier way that gives you
both ports (without reading from hardware):

Check the byte at 0x01FB (SPConfig) using the low mem accessor:

LMGetSPConfig
(defined as UInt8 LMGetSPConfig(void) in LowMem.h Universal Header)

the high 4 bits are for port A (modem) and the lower for for port B (printer).
Here's what they mean:
0 = free
1 = used by AppleTalk
2 = used by another serial driver

So, here's what some code might look like:

Boolean modemPortInUse = false, printerPortInUse = false;
Byte r = LMGetSPConfig();

if (r & 0xF0) modemPortInUse = true;
if (r & 0x0F) printerPortInUse = true;

That's it unless you want to get more specific i.e.:

switch(r & 0x0F)
{
   case 0: // printerPort isn't being used
      break;
   case 1: // printerPort in use by apple talk
      break;
   case 2: // printerPort is being used by something else
      break;
   default: huh?
}

If you want more info about the printer port, check out the byte at 0x0291
(PortBUse). Bit 5 means it's being used by the XPP driver. Bit 7 once again
says if it's currently in use at all (0 == in use), and the other bits I'm
not sure about but bits 4,5,and 6 are used by various driver's at least
(perhaps MPP and ATP?).

Anyway, if all you're wondering is if the two serial ports are being used
by some driver, checking SPConfig should work okay.

Best,

Graham
[e-mail]


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

>From [e-mail] (Harun Scheutzow)
Date: Fri, 24 Jan 1997 23:50:13 GMT
Organization: Village Tronic Marketing GmbH

This code checks the GPi input pin, connected to DCD of the SCC.
But unfortunately,
* not all Macs have this connection.
* here are many cables in use without the neccessary wire to
the modem
* not all Macs allow direct access to SCC
* conclusion: even if Modem is online, you may get "false".

This direct hardware access is ugly. It does not work on
other serial ports (on cards etc).

But there is no free available documentation about a driver
status call which should be introduced in the future to
poll the status of this line. Using this at the moment
probably non-existing status call would work on other serial
ports in the future.

Sorry Apple, if there should be already such a doc, please
tell me where it is.

Regards, HS
<Opinions are my own>


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

>From Steve Collins <[e-mail]>
Date: Mon, 27 Jan 1997 12:10:12 -0800
Organization: Netcom

Hi, Harun,

Harun Scheutzow wrote:
>
> This code checks the GPi input pin, connected to DCD of the SCC.
> But unfortunately,
> * not all Macs have this connection.
> * here are many cables in use without the neccessary wire to
> the modem
> * not all Macs allow direct access to SCC
> * conclusion: even if Modem is online, you may get "false".

Actually, I don't think the standard Mac cable supports DCD
(carrier detect) *ever*.  Hardware weenies, feel free to jump
in and explain how this might be?  I do know that the code
posted does in fact work on at lease *one* Mac.
>
> This direct hardware access is ugly. It does not work on
> other serial ports (on cards etc).
>
> But there is no free available documentation about a driver
> status call which should be introduced in the future to
> poll the status of this line. Using this at the moment
> probably non-existing status call would work on other serial
> ports in the future.
>
> Sorry Apple, if there should be already such a doc, please
> tell me where it is.
>

Of course, there must be a way to accomplish this, it is my
current holy grail.

Cheers
--
- -------------------------------------------------
Steve Collins                 [e-mail]
- -------------------------------------------------

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

>From [e-mail] (Harun Scheutzow)
Date: Tue, 28 Jan 1997 23:26:59 GMT
Organization: Village Tronic Marketing GmbH

Steve Collins ([e-mail]) wrote:

: Actually, I don't think the standard Mac cable supports DCD
: (carrier detect) *ever*.

What is "the standard Mac cable"?

My personal impression is that Apple ignored the GPi for a long time.
They even ignored (like some other large companies) the serial
ports. That means the API is not good. And even "now" they do not
really think about what they do: Adding the PBControl calls with
the csCodes 115 and 230 for 115200 and 230400 bps was a worse
idea. Do they like to add additional csCodes for every higher
speed which might come in the future???

The normal way would have been one call, taking a bps-value as
input (in a 32-bit variable, and not 16 bit!) and returning
information about available bps-rates if the requested speed
can not be provided. Some UNIX(-like) systems do it this way.

Regards, HS


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

>From Steve Collins <[e-mail]>
Date: Wed, 29 Jan 1997 09:20:48 -0800
Organization: Netcom

Harun Scheutzow wrote:
>
> Steve Collins ([e-mail]) wrote:
>
> : Actually, I don't think the standard Mac cable supports DCD
> : (carrier detect) *ever*.
>
> What is "the standard Mac cable"?
>
OK, what I meant is explained in IM:Devices in their explanation of
the RS-422 Serial Interface as used by "all" Macintoshes.

"Data Carrier Detect (DCD). The DCD signal is not used by Macintosh
computers. In systems that use the signal, it is asserted by the DCE
when a carrier signal is received."



Incidentally, I have *still* not found an answer to my original
question, which is how to detect carrier signal while using the
PC Card slots, if any one knows the answer.  The "Serial GPi
(General-Purpose Input)" method described in  TechNote DV-16
works ok for the A and B ports, happily, unless as noted there,
one is using a Mac Plus, etc.

Cheers
--
- -------------------------------------------------
Steve Collins                 [e-mail]
- -------------------------------------------------

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

>From "Don E. French" <[e-mail]>
Subject: Faster line drawing routines?
Date: Tue, 28 Jan 1997 23:01:33 -0700
Organization: Montana Communications Network

    Just wondering if anyone has come up with a faster way to draw
lines, other that QuickDraw's lineto, moveto, ... etc.  Perhaps using the
pixmap directly?

Any code appreciated,
Thanks in advance



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

>From [e-mail] (Tom Andersen)
Date: 29 Jan 1997 16:18:49 GMT
Organization: Sienna Software

In article <Pine.BSD/[e-mail]>, "Don E.
French" <[e-mail]> wrote:

>         Just wondering if anyone has come up with a faster way to draw
> lines, other that QuickDraw's lineto, moveto, ... etc.  Perhaps using the
> pixmap directly?
>
> Any code appreciated,
> Thanks in advance

You can draw lines faster than quickdraw if you want, but I found that
quickdraw is pretty fast at lines that
a) are longer than say 30 or 40 pixels long, and
b) actually have to be drawn.

Quickdraw Lineto is slow compared to direct pixmap manipulation if you
have to do many short lines. Also Quickdraw will take almost as long to
draw a line that completely misses the screen as it will to actually draw
a line of the same length on the screen.

I was able to speed things up a lot by doing my own clipping first, then
only drawing what was necessary using LineTo.

--
Tom Andersen
Sienna Software
www.siennasoft.com
[e-mail]
Makers of Starry Night for the Macintosh.

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

>From [e-mail] (Tricky S)
Date: Wed, 29 Jan 1997 23:06:10 -0600
Organization: Tricky S Software Factory

In the book 'Black Art of Macintosh Game Programming' by Kevin Tieskoetter
there is very good coverage of Bresenham's Algorithm for fast line drawing
(which is actually what LineTo() is based on) with implementations that
are about 30% faster that QuickDraw's LineTo().

In article <Pine.BSD/[e-mail]>, "Don E.
French" <[e-mail]> wrote:

>         Just wondering if anyone has come up with a faster way to draw
> lines, other that QuickDraw's lineto, moveto, ... etc.  Perhaps using the
> pixmap directly?
>
> Any code appreciated,
> Thanks in advance

 _____________
\ ____________\
||  _________  | Stefan "Tricky S" Sinclair
|| /@--------\ | <mailto:[e-mail]>
|| |         | | <http://cse.unl.edu/~sinclair/SpriteFight/sf2k.html>
|| |         | |
|| \_________/ | The label said "Requires Windows95
||       ____  | or better" so I bought a Mac.
|| o Mac       |
\|_____________|
 \|___________|

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

>From David Matiskella <[e-mail]>
Date: Wed, 29 Jan 1997 22:09:39 -0800
Organization: Another Netscape News Server User

Tricky S wrote:
>
> In the book 'Black Art of Macintosh Game Programming' by Kevin Tieskoetter
> there is very good coverage of Bresenham's Algorithm for fast line drawing
> (which is actually what LineTo() is based on) with implementations that
> are about 30% faster that QuickDraw's LineTo().
>
> In article <Pine.BSD/[e-mail]>, "Don E.
> French" <[e-mail]> wrote:
>
> >         Just wondering if anyone has come up with a faster way to draw
> > lines, other that QuickDraw's lineto, moveto, ... etc.  Perhaps using the
> > pixmap directly?
> >
> > Any code appreciated,
> > Thanks in advance
>

    If all you want is simpe 1 pixel wide lines there are a bunch of
algorithms out there that are much faster than Bresenhams. Wether they
are faster than quickdraw might be tough to measure. I know just about
every VGA card has built in Accelerated line drawing. Assuming (yes a
big assumption) that qd uses this hardware (ATI's cards for sure) you
are going to have a real tough time being faster. Anyone else every
wonder why Applethere own video chips rather than using ATI's, S3, or
any of the other ones? A the cool would also gain pageflipping with less
than 4 megs of ram:)
David Matiskella

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

>From [e-mail] (Scott Bassett)
Date: 30 Jan 97 06:34:39 GMT
Organization: Student

In article <Pine.BSD/[e-mail]>, "Don E.
French" <[e-mail]> wrote:

>         Just wondering if anyone has come up with a faster way to draw
> lines, other that QuickDraw's lineto, moveto, ... etc.  Perhaps using the
> pixmap directly?
>
> Any code appreciated,
> Thanks in advance

Try buying the book Computer Graphics: Principles and Practice, you should
find the midpoint line drawing algorithm in there, it's the same one as
lineto's, but if you directly address the pixmap to plot the pixel you
will usually get a 30% speed increase, but also you have to take care of
clipping by yourself then.  And also, it is pretty difficult to understand
the algorithm at first.  Hope this helps.

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

>From Maf Vosburgh <[e-mail]>
Date: 30 Jan 1997 16:59:04 GMT
Organization: MultiMedia Corporation

In article <[e-mail]> Tricky S,
[e-mail] writes:
>In the book 'Black Art of Macintosh Game Programming' by Kevin Tieskoetter
>there is very good coverage of Bresenham's Algorithm for fast line drawing
>(which is actually what LineTo() is based on) with implementations that
>are about 30% faster that QuickDraw's LineTo().

It is really easy to make the Bresenham line drawing code in this book go
faster on PPC. Just make the "short" variables "int" or "long".

BTW it's a great book.


Maf

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

>From [e-mail] ( nick.c @MT )
Date: Tue, 04 Feb 1997 13:24:16 -0800
Organization: MacTech Magazine



Maf Vosburgh <[e-mail]> wrote:

>In article <[e-mail]> Tricky S,
>[e-mail] writes:
>>In the book 'Black Art of Macintosh Game Programming' by Kevin Tieskoetter
>>there is very good coverage of Bresenham's Algorithm for fast line drawing
>>(which is actually what LineTo() is based on) with implementations that
>>are about 30% faster that QuickDraw's LineTo().
>
>It is really easy to make the Bresenham line drawing code in this book go
>faster on PPC. Just make the "short" variables "int" or "long".


    Interesting.  Do you know how much this speeds up the
      drawing?

    BTW, folks can make a quick $25 by sending in tips like
      this (with a few lines of example code) to <[e-mail]>.
      The tips editor pays $25 a shot for each tip he uses,
      and $50 for the best tip each month.  Easy pay for
      few quick lines of eMail...




____Nicholas C. DeMello, Ph.D.___________________________________________
"MacTech Online"--MacTech Magazine, for Mac OS Programmers and Developers
     http://www.MacTech.com/
                                        _/   _/  _/  _/_/_/   _/   _/
   Chemistry: [e-mail]       _/_/ _/  _/  _/   _/  _/_/_/
     MacTech: [e-mail]      _/ _/_/  _/  _/       _/ _/
        http://www.chem.ucla.edu/~nick/   _/  _/   _/_/_/  _/   _/

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

>From [e-mail] (Matthew Mucker)
Subject: Floating palettes
Date: Tue, 21 Jan 1997 18:19:09 +0500
Organization: Internet America

How does one program them?

I'm sure there are references around, if someone could please point me in
the right direction, I'd appreciate it.

Thanks,

-Matt

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

>From [e-mail] (Brian Gray)
Date: Thu, 23 Jan 1997 09:59:08 -0700
Organization: Macromedia, Inc.

[e-mail] (Matthew Mucker) wrote:

> How does one program them?

There is an excellent description of how to do this and 6 pages of example
source in "Mac Programming FAQs" published by IDG Books.  Generally
speaking, this book has the answer to just about anything that has to do
with Mac programming.

 -- B

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

>From Steve Makohin <[e-mail]>
Date: 23 Jan 1997 18:05:37 GMT
Organization: Water's Edge Software

In article <[e-mail]> Matthew Mucker,
[e-mail] writes:
>Subject: Floating palettes
>From: Matthew Mucker, [e-mail]
>Date: Tue, 21 Jan 1997 18:19:09 +0500
>>How does one program them?
>
>I'm sure there are references around, if someone could please point me in
>the right direction, I'd appreciate it.
>
>Thanks,

Check out the Infinity Windoid. Freeware, netbout, source included in C.

-Steve Makohin
 Water's Edge Software
(Makers of Tools Plus & SuperCDEFs)
 http://www.interlog.com/~wateredg

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

>From Roger Wade <[e-mail]>
Date: Fri, 24 Jan 1997 23:44:13 +0000
Organization: Atlas Internet

Develop 15 Has excellent article of floating
windows

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

>From [e-mail] (James Harvey)
Date: Wed, 29 Jan 1997 23:28:42 -0600
Organization: Internet America

In article <[e-mail]>,
[e-mail] (Matthew Mucker) wrote:

>How does one program them?
>
>I'm sure there are references around, if someone could please point me in
>the right direction, I'd appreciate it.
>
>Thanks,
>
>-Matt

I'm surprised no one has mentioned using NewServiceWindow() as a means for
creating a floating palette.  This is documented in IM: Text.
There are a couple of problems with it , though.  It doesn't create a
color window unless you close its port and open a color port in its place
and window clicks may pass thru it into the next wndow unless you  ignore
them with a null event.  However these windows will *always*
stay in front  with no tricks until you close them.  I don't know how you
intend to use them but hope this helps.  BTW, Infinity Windoids
is not much help since the windoids created by it are already available by
the proper floatproc calls to the System ; ).  Someone please
correct me if I'm wrong.

Regards,
James

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

>From [e-mail] (Scott Kevill)
Date: 30 Jan 1997 06:46:38 GMT
Organization: Curtin University of Technology

In article <[e-mail]>,
[e-mail] (James Harvey) wrote:

:In article <[e-mail]>,
:[e-mail] (Matthew Mucker) wrote:
:
:>How does one program them?
:>
:>I'm sure there are references around, if someone could please point me in
:>the right direction, I'd appreciate it.
:>
:>Thanks,
:>
:>-Matt
:
:I'm surprised no one has mentioned using NewServiceWindow() as a means for
:creating a floating palette.  This is documented in IM: Text.
:There are a couple of problems with it , though.  It doesn't create a
:color window unless you close its port and open a color port in its place
:and window clicks may pass thru it into the next wndow unless you  ignore
:them with a null event.  However these windows will *always*
:stay in front  with no tricks until you close them.  I don't know how you
:intend to use them but hope this helps.  BTW, Infinity Windoids
:is not much help since the windoids created by it are already available by
:the proper floatproc calls to the System ; ).  Someone please
:correct me if I'm wrong.

Noone has mentioned these because they are a bad idea for applications
since they will *always* stay in front. Sure you can hide and show them
when your application suspends and resumes, but what about alerts and
modal dialogues that appear - ones that the application does not know
about (eg. notifications). What happens if a floating palette happens to
cover an alert so that you can't see what it says? As said above you need
to hack to get a colour port. You need to write a gnarly GNE Filter and
worry about A5 worlds and the current heap zone to even get mouse clicks
for the palette. The GNE filter will sometimes miss clicks altogether
unless you have a system patch. You have to continually check the update
regions of the palettes since you will not get update evens. You need to
create a termination proc or an ExitToShell patch in case your app crashes
or gets a force-quit since the windows will be left up and can cause
crashes when the mouse moves over them. There are problems setting the
cursor when the mouse is over a TSM window.

Is this enough? I'm sure there are more bad reasons. It is less work (and
more compatible) to do it yourself the standard way.

I believe the system floating windoid is only present in MacOS 7.5 and
later. So you may still need something like Infinity Windoid.

Cheers,

Scott

Scott Kevill
[e-mail]

:Regards,
:James

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

>From Philipp Krueckel <[e-mail]>
Subject: Is LINUX available for Macintosh?
Date: Wed, 22 Jan 1997 20:42:24 +0100
Organization: University of Technology, Aachen, Germany

I want to use UNIX software on my Mac.
Is there a version of Linux for Macintosh?

Philipp

--

Philipp Krueckel                                 \\://
University of Technology Aachen, Germany         (o o)
[e-mail]          ---ooO-(_)-Ooo---

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

>From [e-mail] (Jude Giampaolo)
Date: Wed, 22 Jan 1997 15:48:32 -0500
Organization: CyberDrugs

In article <[e-mail]>, Philipp Krueckel
<[e-mail]> wrote:

> I want to use UNIX software on my Mac.
> Is there a version of Linux for Macintosh?

http://www.mklinux.apple.com/

--
Jude Charles Giampaolo        'I was lined up for glory, but the
[e-mail]                    tickets sold out in advance'
[e-mail]      http://prozac.cwru.edu/jude/JudeHome.html

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

>From [e-mail] (Michael Murry)
Date: Thu, 23 Jan 1997 15:41:10 -0600
Organization: Gridnet News Service

In article <[e-mail]>, Philipp Krueckel
<[e-mail]> wrote:

> I want to use UNIX software on my Mac.
> Is there a version of Linux for Macintosh?
>
> Philipp
>
> --
>
> Philipp Krueckel                                 \\://
> University of Technology Aachen, Germany         (o o)
> [e-mail]          ---ooO-(_)-Ooo---

Yes there is. It only runs on PowerMacs. See http://www.mklinux.apple.com/

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

>From [e-mail] (Kevin M. Cutts)
Date: Thu, 23 Jan 1997 13:05:54 -0600
Organization: Motorola

In article <[e-mail]>, Philipp Krueckel
<[e-mail]> wrote:

> I want to use UNIX software on my Mac.
> Is there a version of Linux for Macintosh?
>
There are currently two ports of linux to the mac. Try
www.mklinux.apple.com for apples current version of linux. I'm running that
and its pretty cool. In addition there is another version but I don't have
the web site handy. (Its linked from the apple site.)
Good luck...

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

>From [e-mail] (Simon Fraser)
Date: Thu, 23 Jan 1997 16:18:36 -0700
Organization: Santa Fe Institute

In article <cutts-ya023580002301971305540001@129.188.137.103>,
[e-mail] (Kevin M. Cutts) wrote:

: In article <[e-mail]>, Philipp Krueckel
: <[e-mail]> wrote:
:
: > I want to use UNIX software on my Mac.
: > Is there a version of Linux for Macintosh?
: >
: There are currently two ports of linux to the mac. Try
: www.mklinux.apple.com for apples current version of linux. I'm running that
: and its pretty cool. In addition there is another version but I don't have
: the web site handy. (Its linked from the apple site.)

http://www.cs.wisc.edu/~tesch/linux_info/

Simon

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

>From [e-mail] (Jerome Chan)
Date: Sat, 25 Jan 1997 19:20:28 -0500
Organization: TofuSoftÅ

In article <[e-mail]>, [e-mail]
(Simon Fraser) wrote:

> In article <cutts-ya023580002301971305540001@129.188.137.103>,
> [e-mail] (Kevin M. Cutts) wrote:
>
> : In article <[e-mail]>, Philipp Krueckel
> : <[e-mail]> wrote:
> :
> : > I want to use UNIX software on my Mac.
> : > Is there a version of Linux for Macintosh?
> : >
> : There are currently two ports of linux to the mac. Try
> : www.mklinux.apple.com for apples current version of linux. I'm running that
> : and its pretty cool. In addition there is another version but I don't have
> : the web site handy. (Its linked from the apple site.)
>
> http://www.cs.wisc.edu/~tesch/linux_info/
>
> Simon

For non-ppc machines, you can use netbsd.
<http://www.netbsd.org/Releases/formal-1.2/mac68k.html>

- -
 The Evil Tofu (Only Human)

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

>From [e-mail] (Me)
Date: 28 Jan 1997 23:20:35 GMT
Organization: Kaaos

In article <[e-mail]>, [e-mail]
(Jerome Chan) wrote:

> In article <[e-mail]>, [e-mail]
> (Simon Fraser) wrote:
>
> > In article <cutts-ya023580002301971305540001@129.188.137.103>,
> > [e-mail] (Kevin M. Cutts) wrote:
> >
> > : In article <[e-mail]>, Philipp Krueckel
> > : <[e-mail]> wrote:
> > :
> > : > I want to use UNIX software on my Mac.
> > : > Is there a version of Linux for Macintosh?
> > : >
> > : There are currently two ports of linux to the mac. Try
> > : www.mklinux.apple.com for apples current version of linux. I'm
running that
> > : and its pretty cool. In addition there is another version but I don't have
> > : the web site handy. (Its linked from the apple site.)
> >
> > http://www.cs.wisc.edu/~tesch/linux_info/
> >
> > Simon
>
> For non-ppc machines, you can use netbsd.
> <http://www.netbsd.org/Releases/formal-1.2/mac68k.html>
>
   Yeah, but don't work on Quadras and/or IIfxen...:(
> ---
>  The Evil Tofu (Only Human)

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

>From [e-mail] (Davi Bock)
Subject: Is palette animation w- > 256 colors possible?
Date: 31 Jan 1997 16:50:00 GMT
Organization: Brown University

Is it possible to do palette animation in 16 or 24-bit environments?

My understanding is:
   ÄPixels in 8-bit color contain references to color tables, and that
when you do palette animation, you're really messing with these color
tables.
   ÄPixels at greater depths each contain the actual color they depict,
rather than an index into a clut, so there's just nothing to animate.

Is there any way to work around this?  I would really like to experiment
with palette animation at greater color depths.

Any help would be greatly appreciated.

--Davi Bock
  [e-mail]

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

>From [e-mail] (Harun Scheutzow)
Date: Sun, 2 Feb 1997 20:07:19 GMT
Organization: Village Tronic

In article
<[e-mail]>,
[e-mail] (Davi Bock) wrote:

>    ÄPixels at greater depths each contain the actual color they depict,
> rather than an index into a clut, so there's just nothing to animate.

On many display hardwares, every color component of such a pixel, that means
Red, Green and Blue, is independently taken to look up in a conversion
table for Red, for Green, and for Blue.

You may change the intensity of each color, but you may not convert a red
pixel into a blue one, by doing palette changes (unless you use only
specially prepared pixels: assume R=20, G=0, B=10. At index 20 for Red
there is "Full Red", at index 0 for Green there is "No Green", at index 10
for Blue there is "No Blue". Now you change the value at Red-index 20 to
"No Red", at Blue-index 10 to "Full Blue". This would convert a Red to a
Blue pixel. But it would change the appearance of all pixels which have
R=20 or(/and) B=10.

Regards, HS
<Opinions are my own>

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

>From [e-mail] (Davi Bock)
Date: Mon, 03 Feb 1997 12:49:20 -0400
Organization: Brown University

In article <[e-mail]>,
[e-mail] (Harun Scheutzow) wrote:

> On many display hardwares, every color component of such a pixel, that means
> Red, Green and Blue, is independently taken to look up in a conversion
> table for Red, for Green, and for Blue.
>
> You may change the intensity of each color, but you may not convert a red
> pixel into a blue one, by doing palette changes (unless you use only
> specially prepared pixels [rest of example deleted]

Thanks very much for your reply.

Do you know of any resources I could pursue to find out more about this?
Inside Mac is pretty sparse on the organization of graphics devices.  How
do I tell if a graphics device uses this kind of mapping?  How do I get in
and change this mapping to produce on-screen color changes?

I would greatly appreciate any further help.

--Davi Bock
  [e-mail]

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

>From [e-mail] (Harun Scheutzow)
Date: Tue, 4 Feb 1997 15:18:58 GMT
Organization: Village Tronic

Designing PCI Cards & Drivers includes a description of the Display Device
Driver API.

Regards, HS
<Opinions are my own>

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

>From Darrin McCarthy <[e-mail]>
Subject: Need INIT source code
Date: Sun, 02 Feb 1997 14:09:08 -0500
Organization: Parametric Technology Corporation

Hello!

Does anyone know of a simple INIT example (with source code) that will
work on a PPC machine? I tried hacking the AFI INIT in Dave Mark's Mac C
Primer (Vol II) but it is for a 68K machine and uses calls specific to
THINK C (like CallPascal()). Thanks in advance!

--dmc

--
Darrin McCarthy

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

>From [e-mail]
Date: Sun, 02 Feb 1997 12:51:22 -0800
Organization: (none)

In article <[e-mail]>, Darrin McCarthy <[e-mail]> wrote:

> Does anyone know of a simple INIT example (with source code) that will
> work on a PPC machine?

There are a large variety of such at the URL <http://www.machack.com>

--
(Pointers to other Mac programming web sites at
<http://devworld.apple.com/dev/geeks.html>)

To reply personally, remove the anti-spam "NOSPAM." from the email address
in the header.

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

>From [e-mail] (bill coderre)
Date: Mon, 03 Feb 1997 12:40:52 -0800
Organization: This posting does not represent the official opinion of Apple Computer, Inc.

In article <[e-mail]>, Darrin McCarthy <[e-mail]> wrote:

|   Hello!
|
|   Does anyone know of a simple INIT example (with source code) that will
|   work on a PPC machine? I tried hacking the AFI INIT in Dave Mark's Mac C
|   Primer (Vol II) but it is for a 68K machine and uses calls specific to
|   THINK C (like CallPascal()). Thanks in advance!
|
|   --dmc
|
|   --
|   Darrin McCarthy

Dave Mark's Ultimate Mac Programming Book has a whole chapter on INITs.
Very recommended.
--
This posting in no way represents the official opinion of Apple Computer,
Inc. Contact me at [e-mail] or [e-mail] to discuss my personal
opinions and those of Apple Computer, Inc. Have A Nice Day.

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

>From [e-mail] (Don Arbow)
Date: Mon, 03 Feb 1997 12:41:38 -0800
Organization: Wolfe Internet Access, L.L.C.

In article <[e-mail]>, Darrin McCarthy <[e-mail]> wrote:

:  Hello!
:
:  Does anyone know of a simple INIT example (with source code) that will
:  work on a PPC machine? I tried hacking the AFI INIT in Dave Mark's Mac C
:  Primer (Vol II) but it is for a 68K machine and uses calls specific to
:  THINK C (like CallPascal()). Thanks in advance!
:

Take a look in the Mac archives for Extension Shell.  This is a complete
INIT shell, all of the code you need is there.  You just need to add the
specific code for your INIT.  You can read about it on the author's home
page at:

http://www.kagi.com/authors/dair/purpleshark.html

Don

--
- ------------------------------------------------------------------------
    \ | /    Don Arbow, Partner, CTO |  [e-mail]
  -- EDO --  EveryDay Objects, Inc.  |     ^ delete underscore
    / | \    Seattle, WA             |   http://www.edo-inc.com
- ------------------------------------------------------------------------
  My Prographing web page:   http://www.wolfe.net/~donarb/Dataflow.html
        "The fix is only temporary, unless it works" - Red Green
- ------------------------------------------------------------------------

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

>From [e-mail] ( nick.c @MT )
Date: Tue, 04 Feb 1997 21:01:06 -0800
Organization: MacTech Magazine



[e-mail] (bill coderre) wrote:


>Dave Mark's Ultimate Mac Programming Book has a whole chapter on INITs.
>Very recommended.


    Second that recommendation:

    _The Ultimate Mac Programming Book: Methods of the Macintosh Masters_
    by Dave Mark,  Programmers Press/IDG Books, 1994 ISBN: 1-56884-275-5



    Also, check out Zobkiw's book, which is specifically about code
     fragments such as inits:

    _A Fragment of Your Imagination_ by Joe Zobkiw
    Addison Wesley, 1995 ISBN: 0-201-48358-0  $40




____Nicholas C. DeMello, Ph.D.___________________________________________
"MacTech Online"--MacTech Magazine, for Mac OS Programmers and Developers
     http://www.MacTech.com/
                                        _/   _/  _/  _/_/_/   _/   _/
   Chemistry: [e-mail]       _/_/ _/  _/  _/   _/  _/_/_/
     MacTech: [e-mail]      _/ _/_/  _/  _/       _/ _/
        http://www.chem.ucla.edu/~nick/   _/  _/   _/_/_/  _/   _/

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

>From [e-mail] (Alexander Chow)
Subject: Objective C
Date: Thu, 16 Jan 1997 14:23:27 -0700
Organization: The University of California at San Diego

Though I am sure that somebody has prior to me posted something about
this, what exactly is Objective C, versus C or C++?

- ---------------------------- Ä ------------------------------
                Alexander Chow Ä [e-mail]
         -- PGP Public Key available upon request --
- ---------------------------- Ä ------------------------------

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

>From [e-mail] (nick.c)
Date: Mon, 20 Jan 1997 19:46:01 -0800
Organization: MacTech Magazine



[e-mail] (Alexander Chow) wrote:

>Though I am sure that somebody has prior to me posted something about
>this, what exactly is Objective C, versus C or C++?


1   What is Objective-C?

    Objective-C is an object oriented computer programming language.  It is
    a superset of ANSI C and provides classes and message passing similar to
    Smalltalk.

                         (from the comp.lang.objective-c FAQ)

         see also:

                   <news:comp.lang.objective-c>

                          and

         <http://web.xplain.com/mactech.com/magazine/urls/next.html>

____Nicholas C. DeMello, Ph.D.________________________________________

   Online at MacTech Magazine, the Journal of Macintosh Programming
     http://www.MacTech.com/
                                       _/   _/  _/  _/_/_/   _/   _/
  Chemistry: [e-mail]       _/_/ _/  _/  _/   _/  _/_/_/
    MacTech: [e-mail]      _/ _/_/  _/  _/       _/ _/
       http://www.chem.ucla.edu/~nick/   _/  _/   _/_/_/  _/   _/


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

>From [e-mail] (Zeppenwolf)
Date: Tue, 21 Jan 1997 11:52:34 -0700
Organization: http://Opion.com


> 1   What is Objective-C?
>
>     Objective-C is an object oriented computer programming language.  It is
>     a superset of ANSI C and provides classes and message passing similar to
>     Smalltalk.

   Nuts, man.

   Without wading through the entire documentation that explains about
'classes' and 'objects', can someone cut to the chase for us and give us
in a nutshell exactly what makes Obj-C DIFFERENT/BETTER/CRAPPER than C++?

Z

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

>From [e-mail] (Dr. Benway)
Date: Tue, 21 Jan 1997 19:55:44 -0600
Organization: Interzone

In article <Zeppenwolf-2101971152340001@207.67.197.77>,
[e-mail] (Zeppenwolf) wrote:

> > 1   What is Objective-C?
> >
> >     Objective-C is an object oriented computer programming language.  It is
> >     a superset of ANSI C and provides classes and message passing similar to
> >     Smalltalk.
>
>    Nuts, man.
>
>    Without wading through the entire documentation that explains about
> 'classes' and 'objects', can someone cut to the chase for us and give us
> in a nutshell exactly what makes Obj-C DIFFERENT/BETTER/CRAPPER than C++?

Well, for one thing, Objective-C programs don't look like they're written
in a language planted in the brains of abductees by malevolent aliens.

In other words, IMHO, the syntax of Objective-C is much cleaner...

Doc

--
Dr. Benway .................. Interzone
=======================================
*      And you may ask yourself:      *
*         How do I work this?         *
=======================================
*        [e-mail]        *
* http://www.cu-online.com/~browning/ *
=======================================

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

>From "Thomas L. Ferrell" <[e-mail]>
Date: 26 Jan 1997 17:23:37 GMT
Organization: Oak Ridge National Lab

[e-mail] (Dr. Benway) wrote:
>In article <Zeppenwolf-2101971152340001@207.67.197.77>,
>[e-mail] (Zeppenwolf) wrote:
>
>> > 1   What is Objective-C?
>> >
>> >     Objective-C is an object oriented computer programming language.  It is
>> >     a superset of ANSI C and provides classes and message passing similar to
>> >     Smalltalk.
>>
>>    Nuts, man.
>>
>>    Without wading through the entire documentation that explains about
>> 'classes' and 'objects', can someone cut to the chase for us and give us
>> in a nutshell exactly what makes Obj-C DIFFERENT/BETTER/CRAPPER than C++?
>
>Well, for one thing, Objective-C programs don't look like they're written
>in a language planted in the brains of abductees by malevolent aliens.
>
>In other words, IMHO, the syntax of Objective-C is much cleaner...
>
>Doc
>
>--
>Dr. Benway .................. Interzone
>=======================================
>*      And you may ask yourself:      *
>*         How do I work this?         *
>=======================================
>*        [e-mail]        *
>* http://www.cu-online.com/~browning/ *
>=======================================

Another point is that Obj-C lacks multiple inheritance, although there are commands that circumvent this.
tom

Delete NoSpam in above address to reply.



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

>From Bob Foster <[e-mail]>
Date: Fri, 31 Jan 1997 23:40:05 +0000
Organization: Symantec Corp.

Dr. Benway wrote:
>
> In article <Zeppenwolf-2101971152340001@207.67.197.77>,
> [e-mail] (Zeppenwolf) wrote:
>
> > > 1   What is Objective-C?
> > >
> > >     Objective-C is an object oriented computer programming language.  It is
> > >     a superset of ANSI C and provides classes and message passing similar to
> > >     Smalltalk.
> >
> >    Nuts, man.
> >
> >    Without wading through the entire documentation that explains about
> > 'classes' and 'objects', can someone cut to the chase for us and give us
> > in a nutshell exactly what makes Obj-C DIFFERENT/BETTER/CRAPPER than C++?
>
> Well, for one thing, Objective-C programs don't look like they're written
> in a language planted in the brains of abductees by malevolent aliens.
>
> In other words, IMHO, the syntax of Objective-C is much cleaner...

Cleaner if you like writing Smalltalk. Objective-C is a language
hybrid/mutant.

Bob

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

>From [e-mail] (Dave Newman)
Date: Sat, 01 Feb 1997 13:58:17 -0600
Organization: CRL Network Services

[e-mail] wrote:
>
> Dr. Benway wrote:
> >
> > Well, for one thing, Objective-C programs don't look like they're written
> > in a language planted in the brains of abductees by malevolent aliens.
> >
> > In other words, IMHO, the syntax of Objective-C is much cleaner...
>
> Cleaner if you like writing Smalltalk. Objective-C is a language
> hybrid/mutant.

All languages borrow concepts from the langauages that preceeded them and
research into languages that were ongoing at the time. Objective-C is no
more a hybrid/mutant than any other programming language.

Better to say that Objective-C takes an approach towards Object Oriented
programming that is different from C++. Objective-C had two goals in mind.
Increase in the complexity of the language syntax only minimally and
support runtime binding of classes and operations instead of compile time
binding.

--Dave
--
Dave Newman
[e-mail]

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

>From "Jonathan W. Hendry" <[e-mail]>
Date: Sat, 01 Feb 1997 16:21:39 -0500
Organization: By Displacement in Metric Tonnes.

Bob Foster wrote:
>
> Dr. Benway wrote:

> > Well, for one thing, Objective-C programs don't look like they're written
> > in a language planted in the brains of abductees by malevolent aliens.
> >
> > In other words, IMHO, the syntax of Objective-C is much cleaner...
>
> Cleaner if you like writing Smalltalk. Objective-C is a language
> hybrid/mutant.
>
> Bob

If Objective-C is a hybrid, then C++ must be a cancer. ;)

--
       "Liar. Liar. Liar. Liar. Liar. Hateful liar. That's what she
       is." - ELLIOT J. ABELSON
       A Los Angeles lawyer who represents Scientology speaking
              about the Pinellas-Pasco medical examiner

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

>From [e-mail] (William Edward Woody)
Date: Sat, 01 Feb 1997 13:59:07 -0800
Organization: In Phase Consulting

In article [e-mail] (Dave Newman) wrote:
> Better to say that Objective-C takes an approach towards Object Oriented
> programming that is different from C++. Objective-C had two goals in mind.
> Increase in the complexity of the language syntax only minimally and
> support runtime binding of classes and operations instead of compile time
> binding.

[thread delurk]

The runtime binding of classes is independant of syntax. Which makes
a 'Direct to Objective C' solution possible (a 'la Direct To SOM, where
you write code using a subset of the C++ syntax and get runtime binding
out the back-end).

Which hopefully for those of us who don't care for method calls that
look like complex and misform array lookups happier...

                                                - Bill

--
William Edward Woody - In Phase Consulting - [e-mail]
                 http://www.alumni.caltech.edu/~woody

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

>From Thomas Okken <[e-mail]>
Subject: Registering app signatures and file types
Date: Sun, 26 Jan 1997 12:19:17 +0100
Organization: World Access, Internet, E-mail and Videotex

Hi all,

I have just written a little program that I would like to publish, but
from long ago I remember that you should register your application
signature and its file types first (to avoid conflicts with other apps).
I've been away from Mac development for a while and I don't remember
the e-mail address I used for this, back then.
Please, could someone give me the e-mail address or URL or whatever it
is one uses these days?
Thanks in advance!

 - Thomas.

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

>From [e-mail] (James W. Walker)
Date: Sun, 26 Jan 1997 11:03:42 -0800
Organization: Nisus Software, Inc.

In article <[e-mail]>, Thomas Okken
<[e-mail]> wrote:

>Please, could someone give me the e-mail address or URL or whatever it
>is one uses these days?

<http://devworld.apple.com/dev/cftype/>
--
  --  Jim Walker <http://members.aol.com/jwwalker/>

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

>From David Scott <[e-mail]>
Date: Sun, 26 Jan 1997 22:22:53 +0000
Organization: West.Net Communications

Thomas Okken wrote:
>
> Hi all,
>
> I have just written a little program that I would like to publish, but
> from long ago I remember that you should register your application
> signature and its file types first (to avoid conflicts with other apps).
> I've been away from Mac development for a while and I don't remember
> the e-mail address I used for this, back then.
> Please, could someone give me the e-mail address or URL or whatever it
> is one uses these days?
> Thanks in advance!
>
>  - Thomas.


Try  http://gemma.apple.com:80/dev/cftype/main.html


--

 David Scott


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

>From [e-mail] (Pratap Reddy P.)
Subject: Serial port arbitration with OT
Date: Fri, 24 Jan 1997 03:35:07 +0530
Organization: Apple Developement India

Hi All,

I need do serial port arbitration and this has to work with OT(future
releases also). As for as I know arbitration means "not letting others use
the port which is already in use".

If it is only with the built-in (presently available) 3 ports, I think this
can be done in 2 ways. one is try opening .AIn, .AOut,...and if it gives
portIsInUse. second getting DCtlEntry ptrs and checking for open in the
flags.
I don't think this is the preferred way with OT as we might have any number
of serial ports(?) and might not work also.

Is there any better way to do this??

Thanks for any kind of help,

Pratap

--
Pratap P. Reddy
Apple development India
email : [e-mail]

phone - (91) - 80 - 2259495

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

>From [e-mail] (Mike Fahl)
Date: Sun, 26 Jan 1997 20:56:50 +0100
Organization: Dataton =?ISO-8859-1?Q?=AD=A0TRUE?= MULTIMEDIA

Pratap Reddy P. <[e-mail]> wrote:

> I need do serial port arbitration and this has to work with OT(future
> releases also). As for as I know arbitration means "not letting others use
> the port which is already in use".
>
> If it is only with the built-in (presently available) 3 ports, I think this
> can be done in 2 ways. one is try opening .AIn, .AOut,...and if it gives
> portIsInUse. second getting DCtlEntry ptrs and checking for open in the
> flags.

I haven't done any serial port programming through OT. But talking
directly to the serial drivers (ie, using OpenDriver, Read/Write, etc),
you have to do exactly as you say. If a serial pport is open by
LocalTalk when you try to open it you get the portInUse error. However,
if it is in use by another application, OpenDriver will not return an
error code, for some obscure historical reason. In this case, you have
to check the DCtlEntry yourself before calling OpenDriver.

I susect having the "Serial Arbitrator" extension (part of ARA)
installed may affect this. But as you can't rely on users aving this
extension installed, I tend to do it the safe way, and check the
DCtlEntry.

Hopefully, OT handles the "multile apps trying to open the same serial
port" problem better. But I don't know that for a fact.

Hope this helps.

By the way, why do you have to access the serial port through OT instead
of using OpenDriver?

Mike Fahl
- --------------
Dataton - TRUE MULTIMEDIA integration and show control systems.
Check it out at http://www.dataton.com

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

>From [e-mail] (Pratap Reddy P.)
Date: Mon, 27 Jan 1997 12:03:24 +0530
Organization: Apple Developement India

In article <[e-mail]>, [e-mail]
(Mike Fahl) wrote:

>
> By the way, why do you have to access the serial port through OT instead
> of using OpenDriver?

Thanks for you help, Mike Fahl. It's of great use to me.

I need to list all the available serial ports with their names and icons
and have to do arbitration.

If I do it thru OT, I am sure it will be supported method and there won't
be any inconsistencies. OT 1.5 provides an API which makes this very
trivial and straight forward job to me.

I do not want to do this in classic method as this (some of the calls)
might not be supported after sometime, if I am correct.

Anyway, i got no choice. I have to use OT :-))

Pratap

--
Pratap P. Reddy
Apple development India
email : [e-mail]
        [e-mail]

phone - (91) - 80 - 2259495

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

>From [e-mail] (Rob Brooks)
Subject: Virtual Memory Question
Date: Thu, 16 Jan 1997 14:27:16 -0600
Organization: The University of Texas

Hello, this isn't exactly a programming question but I thought some of you
might be able to help.  How exactly does VM work on the Mac?  What is the
algorithm for determining what is written to disk and what isn't?

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

>From [e-mail] (Timothy Oliver)
Date: 22 Jan 1997 18:40:44 GMT
Organization: The University of Western Australia

Rob Brooks ([e-mail]) wrote:
: Hello, this isn't exactly a programming question but I thought some of you
: might be able to help.  How exactly does VM work on the Mac?  What is the
: algorithm for determining what is written to disk and what isn't?

Hmmm. Apples VM, eh?

That would be something like. Gee, that page was the most recently used.
I'll just page that out right now...

I know. I know. I should have just kept my mouth shut.


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

>From [e-mail] (Jump Long)
Date: 24 Jan 1997 06:15:10 GMT
Organization: AOL http://www.aol.com

>>Rob Brooks ([e-mail]) wrote:
>>: Hello, this isn't exactly a programming question but I thought some of
you
>>: might be able to help.  How exactly does VM work on the Mac?  What is
the
>>: algorithm for determining what is written to disk and what isn't?
>
>Timothy Oliver replied:
>Hmmm. Apples VM, eh?
>
>That would be something like. Gee, that page was the most
>recently used. I'll just page that out right now...
>
>I know. I know. I should have just kept my mouth shut.

Rob, The basics of how Apple's VM system is implemented can be found in
the Motorola MC68030 Users Manual since that's the Memory Management Unit
model used in Macintosh systems.

Apple's VM replaces physical pages that don't have the MMU used bit set
(the 68k MMU sets the "used" bit in a page's page table entry whenever
there's a memory access in that page) and that aren't held in physical
memory by HoldMemory or LockMemory. If all unheld physical pages have been
"used" since the last page fault, only then does VM simply use the next
unheld page it sees. If the logical page being replaced is dirty (the
memory has been written to), then the logical page's data is written to
the backing store file before the physical page is remapped to a new
logical page. If the logical page being brought into memory is initialized
(that is, it had valid data in it when it was last written to disk), then
it read from disk - otherwise, we leave whatever "garbage" was in the
physical page there.

Timothy, try installing System 7.5.5 or 7.6 (when it's available). I've
done a lot to improve the performance of Apple's VM in the last two system
releases and I'm wooking for ways to improve the code further.

- Jim Luther


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

>From Wulf Hofbauer <[e-mail]>
Date: Mon, 27 Jan 1997 14:10:16 +0100
Organization: Max-Volmer-Institut, TU Berlin

Jump Long wrote:
> (the 68k MMU sets the "used" bit in a page's page table entry whenever
> there's a memory access in that page) and that aren't held in physical
> memory by HoldMemory or LockMemory. If all unheld physical pages have been
> "used" since the last page fault, only then does VM simply use the next
> unheld page it sees.

Could you elaborate on what "the next page it sees" exactly means? Is it
a random choice? LRU (least recently used) is not feasible without
special hardware devoted to it, but at least an approximation like NRU
(not recently used) or even FIFO should be used. I'd be really
interested in this. I've always attributed poor VM performance on the
Mac to thrashing due to heap compaction, but it could also be a bad
replacement strategy. Can you shed some light on this or is it just
another one of Apple's secrets?

> If the logical page being replaced is dirty (the
> memory has been written to), then the logical page's data is written to
> the backing store file before the physical page is remapped to a new
> logical page. If the logical page being brought into memory is initialized
> (that is, it had valid data in it when it was last written to disk), then
> it read from disk - otherwise, we leave whatever "garbage" was in the
> physical page there.

Does this mean that no supply of clean pages is maintained (which would
often halve latency on a page fault)?

- Wulf

--
 ________________________________________________________
! Dipl. Phys. Wulf Hofbauer  ([e-mail]) !
! Max-Volmer-Institut     Technische Universitaet Berlin !
! Strasse des 17. Juni 135     10623 Berlin      Germany !
!________________________________________________________!

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

>From [e-mail] (Miguel Arroz)
Subject: What does "register" means on variable declaration?
Date: Sun, 2 Feb 1997 23:18:57 +0000
Organization: Home

Hi!

 What does regsiter means before a declaration? For example,

 register WindowPeek myWindow;
 int a;
 register WindowPtr otherWIndow;

 Please answer by mail too, because I don't receive every article on
this group.

  Yours

Miguel

--
Miguel Barreto da Silva Arroz - Carnaxide, Portugal - [e-mail]

www.geocities.com/SiliconValley/3232 - How to put your Mac on the net
http://www.geocities.com/SiliconValley/3232/softpage.html - SoftPage

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

>From [e-mail] (Mark Hartman)
Date: 2 Feb 1997 20:45:06 -0700
Organization: Mark Hartman Computer Solutions

In article <[e-mail]>, [e-mail]
(Miguel Arroz) wrote:

> What does regsiter means before a declaration? For example,
>
> register WindowPeek myWindow;
> int a;
> register WindowPtr otherWIndow;

A "register" declaration means that, if possible, the program will assign
a CPU register to hold this value.  There are some implications of this:

1)  It can't be guaranteed (there may be no registers available when needed).
2)  A register has no address, so - depending upon the compiler - trying to
    take the address of a "register" variable may result in either a run-
    time or compile-time error, or it may result in an undefined value. YMMV.
3)  It's very fast, especially as a pointer from which other data is
    referenced, because it's already in a register so the CPU doesn't have
    to do (as many) indirect memory references.
==========================================================================
  Mark Hartman Computer Solutions - specializing in all things Macintosh
     C  C++   4th Dimension   Networking   System design/architecture
tel [phone] -+- fax [phone] -+- e-mail [e-mail]
==========================================================================
          A virus for Windows '95?  Isn't that a bit redundant?

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

>From [e-mail]
Date: Sun, 2 Feb 1997 23:57:01 -0500
Organization: Netcom

Miguel Arroz <[e-mail]> wrote:

> Hi!
>
>  What does regsiter means before a declaration? For example,
>
>  register WindowPeek myWindow;
>  int a;
>  register WindowPtr otherWIndow;
>
>  Please answer by mail too, because I don't receive every article on
> this group.
>
>   Yours
>
> Miguel

  Register is a keyword that tells the C compiler to use a register if
avaiable, for this auto variable. Many optimizing C compilers ignore
this as they do place in regs heavily used variables.  The number of
avaiable register variables is cpu and compiler dependent.

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

>From [e-mail] (William Edward Woody)
Date: Sun, 02 Feb 1997 21:33:06 -0800
Organization: In Phase Consulting

[e-mail] (Miguel Arroz) wrote:
> Hi!
>
>  What does regsiter means before a declaration? For example,
>
>  register WindowPeek myWindow;
>  int a;
>  register WindowPtr otherWIndow;
>
>  Please answer by mail too, because I don't receive every article on
> this group.
>
>   Yours
>
> Miguel

It's a hint to the compiler that the variable should be put into an
available processor register. Sometimes this helps the code run
faster: it allows someone who knows he will be using a variable
often to keep it in a register (which is faster) than put the
variable on the processor's stack (which is slower).

Note that you only have a few registers: on the 68K Mac, you have
about a half-dozen, and on the PowerPC, you only have a dozen. And
for each variable you put into a register, it's one less register
the compiler has to use for itself.

                                        - Bill

--
William Edward Woody - In Phase Consulting - [e-mail]
                 http://www.alumni.caltech.edu/~woody

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

>From [e-mail] (Don Arbow)
Date: Tue, 04 Feb 1997 10:14:03 -0800
Organization: Wolfe Internet Access, L.L.C.

In article <woody-0202972133070001@192.0.2.1>, [e-mail]
(William Edward Woody) wrote:

:
:  Note that you only have a few registers: on the 68K Mac, you have
:  about a half-dozen, and on the PowerPC, you only have a dozen. And
:  for each variable you put into a register, it's one less register
:  the compiler has to use for itself.
:

The PowerPC has 32 general purpose registers.  Two are used for program
maintenance, one for the stack and one for the TOC.  That leaves 30
general purpose registers for use by a procedure, much more than enough.
There are also 32 floating point registers.

Don

--
- ------------------------------------------------------------------------
    \ | /    Don Arbow, Partner, CTO |  [e-mail]
  -- EDO --  EveryDay Objects, Inc.  |     ^ delete underscore
    / | \    Seattle, WA             |   http://www.edo-inc.com
- ------------------------------------------------------------------------
  My Prographing web page:   http://www.wolfe.net/~donarb/Dataflow.html
        "The fix is only temporary, unless it works" - Red Green
- ------------------------------------------------------------------------

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

>From [e-mail] (William Edward Woody)
Date: Tue, 04 Feb 1997 17:54:28 -0800
Organization: In Phase Consulting

[e-mail] (Don Arbow) wrote:
> [e-mail] (William Edward Woody) wrote:
>
> :
> :  Note that you only have a few registers: on the 68K Mac, you have
> :  about a half-dozen, and on the PowerPC, you only have a dozen. And
> :  for each variable you put into a register, it's one less register
> :  the compiler has to use for itself.
> :
>
> The PowerPC has 32 general purpose registers.  Two are used for program
> maintenance, one for the stack and one for the TOC.  That leaves 30
> general purpose registers for use by a procedure, much more than enough.
> There are also 32 floating point registers.

Eh, a dozen, a couple dozen; what's the difference?

But seriously, I thought the Macintosh ABI reserved more than just two
of the general purpose registers for other purposes. On the other hand,
it does leave more than a dozen registers for general purpose use.
My flippant remark stands corrected.

Note that the 'register' keyword works best when you have quite a
few references to a particular variable and no subroutine calls. Otherwise,
with each subroutine call all those registers have to be pushed and
restored--that's because the C compiler has no way to know which
registers may be changed across the subroutine.

                                                - Bill

--
William Edward Woody - In Phase Consulting - [e-mail]
                 http://www.alumni.caltech.edu/~woody

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

>From Tim Humphrey <[e-mail]>
Subject: Window Lists for Other Apps
Date: Mon, 3 Feb 1997 18:24:40 -0500 (EST)
Organization: Info Avenue INTERNET Access

How do you get a window list for other open applications.  I know it's
possible because the extension TitlePop can do it.  Anyone here know how?

..._Tim_...
--=[There is great power in the act of visualization.]=--
http://www.winthrop.edu/~humphret
[e-mail]


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

>From Tim Humphrey <[e-mail]>
Date: Tue, 4 Feb 1997 10:05:59 -0500 (EST)
Organization: Info Avenue INTERNET Access

On Tue, 4 Feb 1997, Olivier Beaudoux wrote:

> If you write an INIT, the best way is to patch the trap that call
> NewWindow()
> (and patch the Dispose/CloseWindow() to...). By this way, you can creat
> your own global
> window list (and their related application). This is not really a hard
> work.

Hey, now there's an idea!  One thing though, would I have to
implement this as an INIT, because wouldn't switching out my app,
well, prevent me getting the window list?  I'll probably try it out
and see what happens.

..._Tim_...
--=[There is great power in the act of visualization.]=--
http://www.winthrop.edu/~humphret
[e-mail]



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

>From Olivier Beaudoux <[e-mail]>
Date: Tue, 04 Feb 1997 14:33:11 -0800
Organization: Universite d'Angers, France.

There is no direct way to access all application window list :-(

If you write an application, the only way (I think...) is to use Process
Manager to
scan all open applications (by using SetFonrtProcess()) and get their
own window list
(by using FrontWindow()).Therefore, it's not really nice: you have to
switch to all opened
applications AND you have to access private window data (WindowPeek)

If you write an INIT, the best way is to patch the trap that call
NewWindow()
(and patch the Dispose/CloseWindow() to...). By this way, you can creat
your own global
window list (and their related application). This is not really a hard
work.

Have fun.

Bod

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

>From [e-mail] (Dave Nebinger)
Date: Tue, 04 Feb 1997 14:28:36 -0600
Organization: KHP Services, Inc

In article <[e-mail]>, Tim Humphrey <[e-mail]> wrote:

// On Tue, 4 Feb 1997, Olivier Beaudoux wrote:
//
// > If you write an INIT, the best way is to patch the trap that call
// > NewWindow()
//
// Hey, now there's an idea!  One thing though, would I have to
// implement this as an INIT, because wouldn't switching out my app,
// well, prevent me getting the window list?  I'll probably try it out
// and see what happens.

You could try the layer manager (documented at the alt.sources.mac
archive site).  The LM still works, even though Apple said it might
go away...

Dave.

==========================================================
Dave Nebinger                             [e-mail]
             The Alt.Sources.Mac Archivist
     <http://www.AmbrosiaSW.com/alt.sources.mac/>
    <ftp://ftp.AmbrosiaSW.com/pub/alt.sources.mac/>

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

>From Joseph Strout <[e-mail]>
Subject: must scriptability be so hard?
Date: Tue, 14 Jan 1997 14:34:02 -0800
Organization: The University of California at San Diego

I have a little app (http://www-acs.ucsd.edu/~jstrout/macvol/) which just
takes a data file, some parameters from the user, and generates a picture.
I'd like to do it properly, i.e., include balloon help, speech
recognition, and make it scriptable/recordable/extensible.

Now, balloon help and SR are pretty easy.  But the Inside Mac book on
making something scriptable completely befuddles me (and I haven't even
gotten to the parts on being recordable or calling scripts from within the
app yet).  All I want is to let the user get and set a host of variables
(e.g., "set xangle to yangle + 4"), and Render.  My app already responds
to the Required suite of AppleEvents (thanks to MacZoop!), so I thought it
would be fairly easy to add these few additional commands.  But instead,
it looks like a very frightening task.

Can anyone help?  Is there a simplified C++ wrapper somewhere, or an
automatic 'aete' generator perhaps?  Or maybe an article that at least
explains it all in a more concrete manner?

Many thanks for any help you can provide.
-- Joe Strout

,------------------------------------------------------------------.
|    Joseph J. Strout           Department of Neuroscience, UCSD   |
|    [e-mail]           http://www-acs.ucsd.edu/~jstrout/  |
`------------------------------------------------------------------'


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

>From [e-mail] (Michael Schuerig)
Date: Wed, 15 Jan 1997 13:31:09 +0100
Organization: RHRZ - University of Bonn (Germany)

Mark Hartman <[e-mail]> wrote:

> In article <Pine.SGI.3.94.970114142643.3872E-100000@knack>, Joseph Strout
> <[e-mail]> wrote:
[...]
> >Now, balloon help and SR are pretty easy.  But the Inside Mac book on
> >making something scriptable completely befuddles me (and I haven't even
> >gotten to the parts on being recordable or calling scripts from within the
> >app yet).  All I want is to let the user get and set a host of variables
> >(e.g., "set xangle to yangle + 4"), and Render.  My app already responds
> >to the Required suite of AppleEvents (thanks to MacZoop!), so I thought it
> >would be fairly easy to add these few additional commands.  But instead,

>
> The key to doing it is to conceptually break your app into two parts: the
> user interface and the engine.  The engine should be controlled by
> AppleEvents; the user interface should only send/receive AppleEvents
> to/from the engine. In this way, ANYONE can control the engine using
> AppleEvents.  You just have to resist the temptation to directly control
> your engine.

That's not sufficient, unfortunately. Even after factoring the app this
way you're still a long way from scriptability if you want to do it
right. Object model scriptability requires you to implement the
hierarchy of your user-leve objects and its interaction with the
"engine". This will probably force you to go deeper into the guts of
your app than you like and it's a pain to implement.

Your best bet is to go with a framework that supports scriptability:
MacApp, ODF, PowerPlant, TCL. Sorry to say, but with MacZoop you're out
of luck, I guess.

Michael

- -
Michael Schuerig
mailto:[e-mail]
http://www.uni-bonn.de/~uzs90z/

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

>From [e-mail] (Mark Hartman)
Date: 14 Jan 1997 22:17:04 -0700
Organization: Mark Hartman Computer Solutions

In article <Pine.SGI.3.94.970114142643.3872E-100000@knack>, Joseph Strout
<[e-mail]> wrote:

>I have a little app (http://www-acs.ucsd.edu/~jstrout/macvol/) which just
>takes a data file, some parameters from the user, and generates a picture.
>I'd like to do it properly, i.e., include balloon help, speech
>recognition, and make it scriptable/recordable/extensible.
>
>Now, balloon help and SR are pretty easy.  But the Inside Mac book on
>making something scriptable completely befuddles me (and I haven't even
>gotten to the parts on being recordable or calling scripts from within the
>app yet).  All I want is to let the user get and set a host of variables
>(e.g., "set xangle to yangle + 4"), and Render.  My app already responds
>to the Required suite of AppleEvents (thanks to MacZoop!), so I thought it
>would be fairly easy to add these few additional commands.  But instead,
>it looks like a very frightening task.
>
>Can anyone help?  Is there a simplified C++ wrapper somewhere, or an
>automatic 'aete' generator perhaps?  Or maybe an article that at least
>explains it all in a more concrete manner?

The key to doing it is to conceptually break your app into two parts: the
user interface and the engine.  The engine should be controlled by AppleEvents;
the user interface should only send/receive AppleEvents to/from the engine.
In this way, ANYONE can control the engine using AppleEvents.  You just have
to resist the temptation to directly control your engine.

Does that help?
==========================================================================
  Mark Hartman Computer Solutions - specializing in all things Macintosh
     C  C++   4th Dimension   Networking   System design/architecture
tel [phone] -+- fax [phone] -+- e-mail [e-mail]
==========================================================================
            Windows '95: How often do you want to crash today?

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

>From Joseph Strout <[e-mail]>
Date: Wed, 15 Jan 1997 14:50:59 -0800
Organization: The University of California at San Diego

On 14 Jan 1997, Mark Hartman wrote:

> >making something scriptable completely befuddles me ...
> >  All I want is to let the user get and set a host of variables
> >(e.g., "set xangle to yangle + 4"), and Render.  My app already responds
> >to the Required suite of AppleEvents (thanks to MacZoop!), so I thought it
> >would be fairly easy to add these few additional commands.  But instead,
> >it looks like a very frightening task.
>
> The key to doing it is to conceptually break your app into two parts: the
> user interface and the engine.  The engine should be controlled by AppleEvents;
> the user interface should only send/receive AppleEvents to/from the engine.
> In this way, ANYONE can control the engine using AppleEvents.  You just have
> to resist the temptation to directly control your engine.
>
> Does that help?

I'm afraid it does not.  I understand the theory.  I just can't get a
grasp of the nuts and bolts of it.  There's a LOT of stuff apparently
involved: 'aete's, 'aeut's, AppleEvents, etc.

The framework I'm using (MacZoop) already receives and responds to the
standard 4 AppleEvents, so I suppose I can extend that to receive
additional ones.  But there's more to making it scriptable than that,
right?  I need to define the scripting dictionary, at least.

After about 3 hours of searching, I finally found an 'aete' editor
(/devworld/Tool_Chest/Interapplication_Communication/AE_Tools_/ResEdit_'aete'_Editor_1.0b4.sit.hqx)
...this helps quite a bit.  But it's still not clear what will happen when
one of these aete-defined verbs is invoked.

This is probably easy stuff for professional Mac programmers, but for a
hobbyist, it seems very complicated.  I was hoping that somewhere, someone
had boiled it down to a "for a simple scriptable app, follow these steps"
document.

,------------------------------------------------------------------.
|    Joseph J. Strout           Department of Neuroscience, UCSD   |
|    [e-mail]           http://www-acs.ucsd.edu/~jstrout/  |
`------------------------------------------------------------------'


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

>From Joseph Strout <[e-mail]>
Date: Thu, 16 Jan 1997 10:31:05 -0800
Organization: The University of California at San Diego

On Wed, 15 Jan 1997, Michael Schuerig wrote:

> > The key to doing it is to conceptually break your app into two parts: the
> > user interface and the engine....
>
> That's not sufficient, unfortunately. Even after factoring the app this
> way you're still a long way from scriptability if you want to do it
> right. Object model scriptability requires you to implement the
> hierarchy of your user-leve objects and its interaction with the
> "engine". This will probably force you to go deeper into the guts of
> your app than you like and it's a pain to implement.

I'm not sure what you mean by "object model" scriptability.

> Your best bet is to go with a framework that supports scriptability:
> MacApp, ODF, PowerPlant, TCL. Sorry to say, but with MacZoop you're out
> of luck, I guess.

Could be; MacZoop's primary attraction is that it's simple.  But I'm not
convinced that it can't do it.  Yesterday I finally found an 'aete'
editor, and today I figured out how to receive the AppleEvents that result
from AppleScript calls.  These are caught by a generic handler, and passed
up the command chain like other events, until somebody handles it.

It seems to work fine so far; I even made a command recordable by sending
myself an AppleEvent for the command instead of calling it directly.  (And
we may talk Graham into doing the same for the core MacZoop commands.)
But then, I haven't yet tried anything complicated like the "get" and
"set" AppleEvents.  Do these require some deeper support architecture?  In
other words, what is it that PowerPlant, TCL, etc. have that makes
handling this stuff easier?

Many thanks for your patience,
-- Joe

,------------------------------------------------------------------.
|    Joseph J. Strout           Department of Neuroscience, UCSD   |
|    [e-mail]           http://www-acs.ucsd.edu/~jstrout/  |
`------------------------------------------------------------------'


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

>From "Thomas L. Ferrell" <[e-mail]>
Date: 17 Jan 1997 03:22:59 GMT
Organization: Oak Ridge National Lab

The AppleScript guru Cal Simone, has a series of articles in Develop on
Scriptability in apps. I believe issue 22 has "Designing a Scriptable
Implementation". You can access all of Develop (tho it's better to get
the CD) at Apples dev site.
tom



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

>From [e-mail] (Edmund K. Lai)
Date: Thu, 16 Jan 1997 15:09:26 -0800
Organization: Apple Computer

In article <Pine.SGI.3.94.970114142643.3872E-100000@knack>, Joseph Strout
<[e-mail]> wrote:

> I have a little app (http://www-acs.ucsd.edu/~jstrout/macvol/) which just
> takes a data file, some parameters from the user, and generates a picture.
> I'd like to do it properly, i.e., include balloon help, speech
> recognition, and make it scriptable/recordable/extensible.
>
> Now, balloon help and SR are pretty easy.  But the Inside Mac book on
> making something scriptable completely befuddles me (and I haven't even
> gotten to the parts on being recordable or calling scripts from within the
> app yet).  All I want is to let the user get and set a host of variables
> (e.g., "set xangle to yangle + 4"), and Render.  My app already responds
> to the Required suite of AppleEvents (thanks to MacZoop!), so I thought it
> would be fairly easy to add these few additional commands.  But instead,
> it looks like a very frightening task.
>

>From what I understand, you have just a bunch of variables in your
application, xangle, yangle, zlength etc, they are all at the same
level, i.e. you don't have something like zlength of the third rectangle
in the fifth part. And you just want to be able to set or get these
varaibles. So it is just Set [xangle|yangle|zlength] to value or
Get [xangle|yangle|zlength]

That would be trival.

In your aete define a bunch of enum, mapping xangle, yangle, zlength
etc each into a unique 4 letter codes like 'XANG', 'YANG', 'ZLNG'.

Then define a 'setd' Apple Event, the event name is 'SET' and there
is a direct parameter that is a enum and a parameter named 'TO' with a
four letter code 'TO  '.

Also define a 'getd' Apple Event with a direct parameter that is a enum.

Then in your application, install a 'setd' event handler and also a
'getd' event handler. In the event handler, get the direct parameter.
It should be one of four letter code in the enum, that defines which
variable it is referencing. In the 'getd' event, just put the result
in the reply Apple Event. In the 'setd' event, fetch the 'TO  ' parameter
and then set the variable to this value.

That is all you have to do.

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

>From [e-mail] (P. Mortensen)
Date: 20 Jan 1997 01:01:34 GMT
Organization: EMBL

> Can anyone help?  Is there a simplified C++ wrapper somewhere, or an
> automatic 'aete' generator perhaps?  Or maybe an article that at least
> explains it all in a more concrete manner?

Resorcerer has a very good aete editor.


>> >making something scriptable completely befuddles me ...
>> >  All I want is to let the user get and set a host of variables
>> >(e.g., "set xangle to yangle + 4"), and Render.  My app already responds
>This is probably easy stuff for professional Mac programmers, but for a
>hobbyist, it seems very complicated.  I was hoping that somewhere, someone
>had boiled it down to a "for a simple scriptable app, follow these steps"
>document.

If you use PowerPlant or TCL it becomes much simpler. You don't even
have
to understand what is going on, just override the appropriate member
functions. The only thing that is missing is example code.
I provide source code directly from two "production" applications:

<http://www.mann.embl-heidelberg.de/Development/ImplementAppleScript.htm
l>
  "Implement AppleScript support in your Macintosh application"



>But then, I haven't yet tried anything complicated like the "get" and
>"set" AppleEvents.  Do these require some deeper support architecture?  In
>other words, what is it that PowerPlant, TCL, etc. have that makes
>handling this stuff easier?

The frameworks take care of the generic stuff: installing call-backs,
object specifiers, etc. To add one more property put in one more case
in DoPropertySetDataEvent(), DoPropertyGetDataEvent() and
AccessObject() in the C++ object that represents the element  (for
TCL)


>The AppleScript guru Cal Simone, has a series of articles in Develop on
>Scriptability in apps. I believe issue 22 has "Designing a Scriptable
>Implementation". You can access all of Develop (tho it's better to get
>the CD) at Apples dev site.

That (excellent) article only describes design of the aete resource
and not the code implementation. The aete resource describes
the mapping between user terms seen in the Script Editor and the 4
byte code used internally in the program to refer to properties,
elements, enumerations, etc.



Regards,
Peter Mortensen


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

Peter Mortensen,                    E-mail:
[e-mail]
Software Engineer, M.Sc.E.E.                    FAX: [phone]
306
European Molecular Biology Laboratory (EMBL)  Phone: [phone]
560
http://www.mann.embl-heidelberg.de/PPG/PersonalPages/PeterM/PeterM.html

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

>From [e-mail] (Zeppenwolf)
Date: Mon, 20 Jan 1997 13:47:21 -0700
Organization: http://Opion.com


In article <5bug5e$[e-mail]>, [e-mail]

> Resorcerer has a very good aete editor.

   No, it doesn't.  It has a very good aete 'viewer', but you're
out of luck if you want to do serious 'editing'.  There's no
substitute for EightyRez.  No, I'm not associated.

   And that hypercard stack is buggy, and therefore almost useless...

> If you use PowerPlant or TCL it becomes much simpler. You don't even

   Amazingly, I found, the TCL AE code is vastly more complete than the
MW AE code, which sticks it to me right up the kazoo, cuz I'm currently
trying to port a TCL app to PP.  The AE code is by far the biggest hurdle...

Z

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

>From [e-mail] (P. Mortensen)
Date: 21 Jan 1997 09:11:04 GMT
Organization: EMBL

In article <Zeppenwolf-2001971347210001@207.67.197.76>
[e-mail] (Zeppenwolf) writes:

> > Resorcerer has a very good aete editor.
>
>    No, it doesn't.  It has a very good aete 'viewer', but you're
> out of luck if you want to do serious 'editing'.  There's no
> substitute for EightyRez.  No, I'm not associated.

I have been using Resorcerer's aete editor for the last 2 years with
great success but if EightyRez is so much better then I will buy it. Do
you have ordering information? (company name, address, fax, etc.)


How do you define "serious 'editing'?"




Regards,
Peter Mortensen


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

Peter Mortensen,                    E-mail:
[e-mail]
Software Engineer, M.Sc.E.E.                    FAX: [phone]
306
European Molecular Biology Laboratory (EMBL)  Phone: [phone]
560
http://www.mann.embl-heidelberg.de/PPG/PersonalPages/PeterM/PeterM.html

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

>From [e-mail] (Zeppenwolf)
Date: Tue, 21 Jan 1997 11:45:16 -0700
Organization: http://Opion.com

In article <5c2178$[e-mail]>, [e-mail]

> I have been using Resorcerer's aete editor for the last 2 years with
> great success but if EightyRez is so much better then I will buy it. Do
> you have ordering information? (company name, address, fax, etc.)

   Actually, I remember going through hoops trying to find/buy it-- I
actually had to make inquiries through the US mail!!  But I found this
in the readme:

   (From EightyRez readme)
 ] Report bugs and other problems to [e-mail].


   I originally got the first piece of it by downloading a demo version
from a MacWareSite somewhere.  Then you send away for the number.

   Anyway, for the nominal cost, there's no excuse not to have it if
you have to work with aete's...

> How do you define "serious 'editing'?"

   You say you've been using Res: do you regularly create/delete
sub-objects in the aete?

   I remember that Resorcerer would let me do a few things but not
some others...?  I think that I could cut/paste, but I couldn't create
a new element of one of the sub-objects in that template; I don't
really remember now-- I ended up only using it on my aete's to fix the
buggy parts of the Hypercard stack output.

   About that, there was some problem in creating 'plurals', so
making a class 'widget' and the same class but with plural flag,
'widgeaux', (French), wouldn't create the element in the aete, and I
always had to do it by hand...

Z

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

>From Doug McKenna <[e-mail]>
Date: Tue, 21 Jan 1997 21:38:58 +0000
Organization: Online Network Enterprises, Inc.

> > Resorcerer has a very good aete editor.
>
>    No, it doesn't.  It has a very good aete 'viewer', but you're
> out of luck if you want to do serious 'editing'.  There's no
> substitute for EightyRez.  No, I'm not associated.
>
>    And that hypercard stack is buggy, and therefore almost useless...

Resorcerer indeed edit's 'aete's using its standard template mechanism,
and a great many developers have bought it solely to do 'aete' editing.

Resorcerer 2.0 will have a dedicated 'aete' editor that will be far
superior to the current method of using the data editor with the
template.  Please check out our website for further details on the
upgrade.

Doug McKenna
Mathemaesthetics, Inc.
PO Box 298
Boulder, CO 80306-0298
USA
[e-mail]      <-- New internet email address
http://www.mathemaesthetics.com      <-- New website
[phone] (vox)
[phone] (fax)

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

>From [e-mail] (Joshua Juran)
Date: Wed, 22 Jan 1997 06:04:14 -0500
Organization: People who haven't founded organizations (Who founded it?)

In article <Pine.SGI.3.94.970114142643.3872E-100000@knack>, Joseph Strout
<[e-mail]> wrote:

> Now, balloon help and SR are pretty easy.  But the Inside Mac book on
> making something scriptable completely befuddles me (and I haven't even
> gotten to the parts on being recordable or calling scripts from within the
> app yet).  All I want is to let the user get and set a host of variables
> (e.g., "set xangle to yangle + 4"), and Render.  My app already responds
> to the Required suite of AppleEvents (thanks to MacZoop!), so I thought it
> would be fairly easy to add these few additional commands.  But instead,
> it looks like a very frightening task.

   It is.

   I was confronted by this when I chose to Do Things Right with regard to
Apple events in the two applications in TinyMUSH/Mac. The Netmush is the
server, and Game Wiz is a client that controls the server. The current
shipping version just defines custom Apple events with no parameters for
"Startup", "Shutdown", and "Make New Database". My intent here was just to
make something which worked, so I could toss the AppleScript hack, which
didn't.
   But I wanted to support multiple 'game instances' per server, which
meant that Game Wiz needed to know their names. In AppleScript, this would
be written as 'get the name of every instance'. In AEGizmos,
"core\getd{want:type(prop), form:prop, seld:pnam, from:obj{want:type(cIns),
form:abso, seld:all, from:null()}}". I realized that a full-blown Object
Model implementation would be the Right Thing to Do, rather than some
half-baked code to see if the event was requesting the names, and if so,
return them. It was at this point I decided to Do Things Right and roll my
own support library. I don't think C++ is the right tool here, but it's
what I've got.

> Can anyone help?  Is there a simplified C++ wrapper somewhere, or an
> automatic 'aete' generator perhaps?  Or maybe an article that at least
> explains it all in a more concrete manner?

   I have not concerned myself with the 'aete' really, because Game Wiz is
an application and doesn't need user terminology. I hacked in enough using
the ResEdit editor to test Netmush from the Script Editor. I have much more
to learn before I actually write anything on this. Maybe I can use this as
a paper and get into MacHack for free? :-)
   As for a "simplified C++ wrapper", I think this does indeed fit the
bill. My target is not just the server-side handling that PowerPlant
already does, but the whole deal -- I need a framework that Game Wiz can
use to send Apple events and get replies in the mail er, event queue, :-)
and be able to pick up from where it left off. This has the potential to
create some really nightmarish code, which I'd like to head off before I
write it, by creating this framework. (That part isn't done yet -- I'm
still sending kWaitReply.)
   But I've created a bunch of C++ classes -- for descriptors, tokens,
model objects, event handlers, object accessors, event prototypes, and
event instances. One of the main things these wrapper classes do is that
they encapsulate the error-checking and throw exceptions. This really
cleans up the calling code, A LOT. I was amazed (ow, think I sprained my
arm) :-) when I rewrote a routine to send an Apple event -- the
error-checking is completely gone! The code is much easier to read and
actually makes some semantic sense without having to figure out exactly
what effect each statement will have.
   I'm currently working on the object accessors. I have a CModelObject
class, from which other classes may inherit, like CModelApp (which in turn
is inherited by the app's class). A semi-generic accessor to get model
objects from other model objects calls an overridable class method. A
special accessor for anything from 'null' redispatches the access using the
application as a model object, so it's the only accessor that needs to know
absolutely where the application is (of which it's informed through its
constructor). Lists are coming soon.

   By the way, I have an implementation question: In the model object from
model object accessor, it's possible that the access must return a list.
Instead of abstracting the return type so as to include a list (which means
making a class for it, etc.) I just create an AEDesc and throw that as an
exception -- which the dispatching accessor is ready to catch and deal with
appropriately. Pretty cool, huh? I love exceptions!
   Anyway, is this a good way to do it, or am I abusing the language?

   The only online information about this class library is on the
TinyMUSH/Mac Web page at <http://www.metamage.com/mush/>. If you're really
interested in creating awesome Mac software, I request that you stop by
there and also <http://www.metamage.com/mammal/>, my multi-stage project to
create a Web Browser For The Rest Of Us. And if this excites you, then you
should definitely take a look at
<http://www.metamage.com/softdev/donating.html>. :-)

   I'd like to finish this thing first. Then I'll exhibit it in my products
(first TinyMUSH/Mac, then Mammal, then whatever else I write that makes
sense to make scriptable). At some point I may choose to distribute the
source. However, if someone could enroll me in the possibility of, say,
kick-butt software for the Mac, it might see light of day sooner...

Josh

--
Joshua Juran                              Metamage Software
=)                                     and other creative arts
[e-mail]
<http://www.metamage.com/>             * Home of TinyMUSH/Mac *

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

>From Joseph Strout <[e-mail]>
Date: Wed, 22 Jan 1997 08:50:12 -0800
Organization: The University of California at San Diego

On Wed, 22 Jan 1997, Joshua Juran wrote:

> I realized that a full-blown Object
> Model implementation would be the Right Thing to Do, rather than some
> half-baked code to see if the event was requesting the names, and if so,
> return them. It was at this point I decided to Do Things Right and roll my
> own support library. I don't think C++ is the right tool here, but it's
> what I've got.

I don't think it's a bad tool.  The object model is weakly typed and
rather smalltalky, but C++ can deal with that...

>    But I've created a bunch of C++ classes -- for descriptors, tokens,
> model objects, event handlers, object accessors, event prototypes, and
> event instances. One of the main things these wrapper classes do is that
> they encapsulate the error-checking and throw exceptions. This really
> cleans up the calling code, A LOT. I was amazed (ow, think I sprained my
> arm) :-) when I rewrote a routine to send an Apple event -- the
> error-checking is completely gone! The code is much easier to read and
> actually makes some semantic sense without having to figure out exactly
> what effect each statement will have.

I bet!  This sounds like a great library.  (Why doesn't Apple do things
like this?  The IM code examples are still in Pascal, fer cryin' out
loud...)

>    By the way, I have an implementation question: In the model object from
> model object accessor, it's possible that the access must return a list.
> Instead of abstracting the return type so as to include a list (which means
> making a class for it, etc.) I just create an AEDesc and throw that as an
> exception -- which the dispatching accessor is ready to catch and deal with
> appropriately. Pretty cool, huh? I love exceptions!
>    Anyway, is this a good way to do it, or am I abusing the language?

Hmm, I might have gone for abstraction -- but your solution seems to work
too.

>    I'd like to finish this thing first. Then I'll exhibit it in my products
> (first TinyMUSH/Mac, then Mammal, then whatever else I write that makes
> sense to make scriptable). At some point I may choose to distribute the
> source.

It sounds like a great library, and I encourage you to make the source
freely available as soon as you can.  This will encourage others to write
great Mac software.

As for me, I'm finally getting the hang of this whole AppleEvents and
Object Model kaboodle.  I'm working with Graham Cox to incorporate
easy-to-use support for it into MacZoop, a simple-but-super C++ app
framework.

Keep up the good work!
-- Joe

,------------------------------------------------------------------.
|    Joseph J. Strout           Department of Neuroscience, UCSD   |
|    [e-mail]           http://www-acs.ucsd.edu/~jstrout/  |
`------------------------------------------------------------------'


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

>From [e-mail] (Francois Pottier)
Date: 22 Jan 1997 15:48:58 GMT
Organization: INRIA Rocquencourt, BP 105, 78153 Le Chesnay Cedex, France

In article <[e-mail]>,
Doug McKenna  <[e-mail]> wrote:

>Resorcerer 2.0 will have a dedicated 'aete' editor that will be far
>superior to the current method of using the data editor with the
>template.

Do you have pricing information for version 2.0 (regular price and
student discount)?

Thanks!



--
FranÁois Pottier
[e-mail]
http://pauillac.inria.fr/~fpottier/

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

>From Doug McKenna <[e-mail]>
Date: Wed, 22 Jan 1997 22:11:32 +0000
Organization: Online Network Enterprises, Inc.

Francois Pottier writes:

> Do you have pricing information for version 2.0 (regular price and
> student discount)?
>
> Thanks!

Not yet.  It will be on the website when the time comes.

Doug McKenna
Mathemaesthetics, Inc.
PO Box 298
Boulder, CO 80306-0298
USA
[e-mail]      <-- New internet email address
http://www.mathemaesthetics.com      <-- New website
[phone] (vox)
[phone] (fax)

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

>From [e-mail] (Gary McGath)
Date: Fri, 24 Jan 1997 09:31:00 -0500
Organization: Conceptual Design

In article <5c2178$[e-mail]>,
[e-mail] (P. Mortensen) wrote:

>In article <Zeppenwolf-2001971347210001@207.67.197.76>
>[e-mail] (Zeppenwolf) writes:
>
>> > Resorcerer has a very good aete editor.
>>
>>    No, it doesn't.  It has a very good aete 'viewer', but you're
>> out of luck if you want to do serious 'editing'.  There's no
>> substitute for EightyRez.  No, I'm not associated.
>
>I have been using Resorcerer's aete editor for the last 2 years with
>great success but if EightyRez is so much better then I will buy it. Do
>you have ordering information? (company name, address, fax, etc.)

For full information on EightyRez, see my Web page at

  http://www.ultranet.com/~gmcgath/EightyRez.html

It's available by mail, or can be downloaded from Software Unboxed.
There's a link from my page for doing the latter.

--
   Gary McGath     [e-mail]
   http://www.ultranet.com/~gmcgath

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

>From [e-mail] (Laurent Gasser)
Date: 27 Jan 1997 13:45:52 GMT
Organization: Swiss Meteorological Institute

In article <Pine.SUN.3.94.970115144449.19330A-100000@szechuan>, Joseph Strout <[e-mail]> writes:
>
> I'm afraid it does not.  I understand the theory.  I just can't get a
> grasp of the nuts and bolts of it.  There's a LOT of stuff apparently
> involved: 'aete's, 'aeut's, AppleEvents, etc.
>
> ,------------------------------------------------------------------.
> |    Joseph J. Strout           Department of Neuroscience, UCSD   |
> |    [e-mail]           http://www-acs.ucsd.edu/~jstrout/  |
> `------------------------------------------------------------------'
>

I remember seeing an Apple example at their site (the Developper's side).
It was called 7Edit and offered similar capabilities as the ScriptEditor.
This could be another way to learn it, by the example.

--
Laurent Gasser ([e-mail])
Computers do not solve problems, they execute solutions.



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

>From [e-mail] (Mike Fahl)
Date: Tue, 28 Jan 1997 22:32:23 +0100
Organization: Dataton =?ISO-8859-1?Q?=AD=A0TRUE?= MULTIMEDIA

Laurent Gasser <[e-mail]> wrote:

> Joseph Strout <[e-mail]> writes:
> >
> > I just can't get a
> > grasp of the nuts and bolts of it.  There's a LOT of stuff apparently
> > involved: 'aete's, 'aeut's, AppleEvents, etc.

I agree. I spent a lot of time before I eventually got the hang of it.
Once I did, I discovered that it was actually rather simple and elegant!
I'm sure the original inventors of OSA and the object model went
ballistic when they saw what a mess the documentation department made of
their baby.

Not only does the documentation manage to make it all appear very
complicated and convoluted. It's also plain wrong in many places, such
as this example from the description of the AEGetAttributePtr function:

"You can use the AEGetAttributePtr function to get a pointer to a buffer
that contains the data from a specified Apple event attribute."

This function does not "get a pointer" in any way - rather the opposite.
The function retieves the data from the specified attribute, and stores
that data in a buffer you provide by passing it a pointer to said
buffer. Any reasonably experienced programmer can fugure that out. But
too much of that goofy technobabbel can make even the most determined
decipherer go astray from time to time.

By the way, I get the same feeling from many of the new (black) Inside
Mac books (with the possible exception of the Open Transport book). They
give me the creeps. They contain sooo much text, but for me its all
organized in the wrong way.

I had *no* problems whatsoever understanding the old (white) IM books.
They explain how things work starting off at the right end (the basics).
However, I've also heard people saying the opposite, so I assume it's
just me that's warped. But it would be interesting to know if there's
someone else out there feeling the same way.

All IMHO, of course. And pardon the rambling - I just couldn't resist
the temptation.

Mike Fahl
- --------------
Dataton - TRUE MULTIMEDIA integration and show control systems.
Check it out at http://www.dataton.com

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

>From [e-mail] (P. Mortensen)
Date: 28 Jan 1997 23:21:57 GMT
Organization: EMBL

In article <[e-mail]>
[e-mail] (Laurent Gasser) writes:

> > I'm afraid it does not.  I understand the theory.  I just can't get a
> > grasp of the nuts and bolts of it.  There's a LOT of stuff apparently
> > involved: 'aete's, 'aeut's, AppleEvents, etc.
> >
> > ,------------------------------------------------------------------.
> > |    Joseph J. Strout           Department of Neuroscience, UCSD   |
> > |    [e-mail]           http://www-acs.ucsd.edu/~jstrout/  |
> > `------------------------------------------------------------------'
> >
>
> I remember seeing an Apple example at their site (the Developper's side).
> It was called 7Edit and offered similar capabilities as the ScriptEditor.
> This could be another way to learn it, by the example.

Or have a look at my source code for a script interface (taken directly
from 2 programs in everyday use here at our lab):

<ftp://mac-mann6.embl-heidelberg.de/Saturn/Pub/Development/ImplementAppl
eScript/TCL/ScriptingImpl_TCL1.1.sit.hqx>

and

<ftp://mac-mann6.embl-heidelberg.de/Saturn/Pub/Development/ImplementAppl
eScript/PowerPlant/ScriptingImpl_PP1.0.sit.hqx>     (Be sure to read
the readMe file)

The executables are also available.



Regards,
Peter Mortensen


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

Peter Mortensen,                    E-mail:
[e-mail]
Software Engineer, M.Sc.E.E.                    FAX: [phone]
306
European Molecular Biology Laboratory (EMBL)  Phone: [phone]
560
http://www.mann.embl-heidelberg.de/PPG/PersonalPages/PeterM/PeterM.html

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

>From Bob Foster <[e-mail]>
Date: Fri, 31 Jan 1997 23:42:54 +0000
Organization: Symantec Corp.

Zeppenwolf wrote:
>
> In article <5bug5e$[e-mail]>, [e-mail]
>
> > Resorcerer has a very good aete editor.
>
>    No, it doesn't.  It has a very good aete 'viewer', but you're
> out of luck if you want to do serious 'editing'.  There's no
> substitute for EightyRez.  No, I'm not associated.
>
>    And that hypercard stack is buggy, and therefore almost useless...
>
> > If you use PowerPlant or TCL it becomes much simpler. You don't even
>
>    Amazingly, I found, the TCL AE code is vastly more complete than the
> MW AE code, which sticks it to me right up the kazoo, cuz I'm currently
> trying to port a TCL app to PP.  The AE code is by far the biggest hurdle...

Amazingly? :-]

Bob

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

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