Mousehole Report

By Rusty Hodge, Mousehole BBS

What: System 4.2

From: Jim Reekes (CMS)

Okay, I spent about a hour getting my Mac II set up with System 4.2/Finder 6.0 and the same exact problem is still there. In PageMaker 2.0a, all Menus, Dialogs, and fonts within the document are totally garbaged. It is as if PageMaker cannot discover the proper font to be using, including the System font. So far this is the only application that I have had a problem with and I do blame it on PageMaker, not Apple’s system. But, as I did mention before, this is a bug that I saw in the beta version of the System 4.2.

The other major problem I had with upgrading to System 4.2 was that the boot blocks have been changed. The first time I copied the new System/Finder to my system folder, the boot blocks did not get updated. This presented me with the shaking “Welcome to Macintosh” dialog at startup. So I check the original system disk’s boot blocks with FEdit and modified my hard disk to match. Now it boots up properly. [Looks like the best way to update is to use the Installer. We used it and had no problems. Also, I have not seen your Pagemaker problem, except on certain down-loaded fonts. As we reported last month, Pagemaker 2.0 cannot place formatted MacWrite text correctly which uses certain down-loaded fonts, like the thin saratoga we use for all our source code listings! -Ed]

What: MF Interactions

From: Randy Saunders (Honeywell Inc.)

I have two interesting MultiFinder interactions to report. I can work around them, but I would like to hear hypotheses on what causes them.

1) Start up the new DA/Font Mover (3.6?) from the MultiFinder. The first display (fonts in System) has all my fonts. Click the radio button for DAs, only 1 DA appears. Click the radio button for fonts and NO fonts appear. Clearly the message here is don’t run the mover when the MF is using this data, but why the lost fonts? On second thought, why shouldn’t the mover work?

2) When running FreeTerm 2.0 under Finder 6.0 everything seems neat. When you run FreeTerm under the MF, and the host clears the screen, the program freezes up. Even the I-beam cursor flickers very dimly. Unless the host sends you a character, you seem to be stuck. I honestly don’t have a guess what is going on here. [Lots of things crash under MultiFinder. The Lightspeed compilers bomb miserably. How about MPW? -Ed]

What: ProApp and Wabash Computer

From: Dirk (American Zettler, Inc.)

I just found out yesterday that ProApp is owned and operated by Wabash Computer. Boy is it hard keeping up with the ‘players’.

What: ICONs for INIT notification

From: Jaff (Lake Forest, California)

I’m writing a cdev that uses an INIT as well, and I’ve noticed that some of the newer ones (QuicKeys, ColorDesk) show an ICON when their INIT runs. Mine shows an ICON, too, and I’m wondering if there’s a mechanism whereby I can offset the position of my ICON depending on what’s been run before. In other words, can I tell who else’s ICON is already on the screen so I don’t overwrite it.

What: ICONs for INIT31 notification

From: Paul Mercer (SOL Systems)

I wrote ShowINIT last Summer so I didn’t have to resort to sicko windows for notification at INIT31 time (like TOPS). I have made the source code public domain hence its use in QuicKeys and other INITs. Leave me an address and perhaps I can get it to you. I’m on Delphi, AppleLink, and MCI Mail. How does one get on the MH dowload board anyways? [Register and pay the dues on the Mousehole board. -Ed]

What: Tape Backup 40 SC Blues

From: Maxbug (Ashton-Tate)

Apple’s Tape Backup 40 SC ( ver 1.1) doesn’t like FINDSWELL! Volume Backup will ID=01 on a Mac II. If you have to split a backup across two tapes on an SE it never recognises the first tape again to write the catalog. The workaround is to move FINDSWELL out of your system folder and re-boot. After the backup you can put FINDSWELL into the system folder and all is well. ( SYSTEM 4.1, FInder 5.5 ). [Try it again under the new system and finder. We have had no problems with the tape backup 1.1 under the new OS on a Mac II. But be sure to have a formatted tape ready. It apparently forgets in the middle of a backup if it has to format a tape before continuing, making you start all over again. -Ed]

What: Mac II Behavior

From: Cpettus (Nova Development Corporation)

I’ve been noticing some strange behavior in my Mac II, and I was wondering if anyone else has had the same problem or has an insight into it. Very occassionally (0.5-2 times per day) the Mac will freeze, emitting a high pitched whine. It seems to happen only during extensive disk access. The system will be frozen for 5-20 seconds, after which it will unfreeze and everything seems to be fine. While it doesn’t seem to hurt anything, it is annoying and perplexing, to say the least ... Thanks in advance for any help or information. [This is a new one! Anybody? What we have seen is the color monitor display getting a case of the jiggles when first turned on. -Ed]

What: ClipSaver

From: Patrick Coffin (Ucla)

Paul, ClipSaver is an init so that when a you quit a program or shutdown the mac the clipboard will be saved. I dont think it is Scrapsaver because it is only 543 bytes. It is a nice feature except that is does not work with MF using a Mac II.

What: RR 10.0

From: Patrick Coffin (Ucla)

I just received RR10.0 today and just trying it out. It has some new features, but I believe that you still have to click after a file transfer which I feel is a real hassle if you are receiving multiple files from someone, and you want to leave your mac unattended for any reason. The person on the other end tries to send the next file and he gets no response, until you send a mouseclick. Why have autoreceive if you cannot receive multiple files?

What: AppleShare Sharing

From: Dave Goss (Mcdonnell Douglas)

Sam Choi........Yes it is true. Serial Baseband networks like AppleTalk can carry only one message at a time. If a second (or third, etc) us tries to send or receive something while that message is going down the pipe he/she must wait. The key to the thing’s success is that network activity is really pretty random, and it is more unusual than you may think for there to be conflicts (or contention, as the network designers call it). Just think of the cables connecting your stand-alone Mac to the disk drive and printer. How often are they ‘busy’ during a normal session. Most of the time you are pounding the keys, scratching your head, chatting with someone, etc. and those wires just sit there doing nothing. If you were on a network, others could be using them and you would never know it. Furthermore, most data transfer sessions are fairly short, so that even if someone does have to wait, it should not be for long. It generally comes down to how many users you have on the network and what kinds of activities they are engaged in. The folklore is that ten or twelve people can share an AppleTalk network without noticable degradation, but after that things can start slowing down if everyone is really busy. Again, the chances are that not everyone will be busy at once. I have seen up to 20 workstations coexist without serious problems. Naturally things like games running on the net will have some affect, because they are using part of it’s capacity. The most realistic suggestion that I can make is to try it and see what happens. There are a lot of people using networks successfully, so don’t worry about it so much. Just go for it!

I neglected to mention in the above, another very important factor that makes sharing a network practical, is packet data transfer. When a transaction, such as sending a file to the printer, is in progress that transaction does not hog the network exclusively. The network manager software breaks each message up into packets of data, which I think are about 600 bytes long. Between packets, other transactions have an opportunity to grab the bus and have one of their packets sent. The various transactions are thus interleaved so that nobody is left out in the cold. When things get busy the interval between packets of a given transaction get longer, but the data is still flowing. The guy trying to download a file off the server while the print file is being transfered may (or may not) notice that it takes a little longer, but he won’t be left staring at a blank screen and wondering what is happening.

What: RR 10.0

From: Patrick Coffin (Ucla)

Well RR has a lot of bugs, he is already coming out with a patch for some of them. The biggest problem is the pathnames, he only allows 39 chars. Which if you have a hard disk and RR is somewhat burried this becomes a problem. Other problems are with the proceedure commands; either he changed the command name or the syntax. And all the old proceedures have to be recompiled to be used with the new version.

What: RR10.1

From: The Dumacker (National Medical Homecare)

Disregard the 10.1 upgrade completely. Patch for 10.2 is on the way.

What: Capture to file by append (help)

From: Mysteray (American Zettler)

I’d like to be able to capture files from BBS sessions. I can do it now (using Microphone w/ scripts), but always it is a new file. The Mac allows no spiffy way to copy multiple files to a single file (even CP/M could do THAT), as far as I know. Microphone will always open a new file on capture, even deleting the old one if it existed.

Anyone know of a communications program that allows appending to an existing text file? This way I could put an entire months worth of MH (or other) into one file without having to manually append these using QUED or some other editor. Or another solution, other than to have ten or fifteen separate little files that make finding things difficult.

What: Capture File by Append

From: Arnold Woodworth (Sterer Engineering And Mfg.)

There is a way to make Red Ryder 9.4 (I don’t have my 10.x update yet) append incoming text to an existing file. DON’T use “Capture Incoming Data to Text File” in the File menu. If you use the name of an existing file, it will be replaced, NOT appended to. Instead, you must write a procedure which uses the RECA command. RECA will open an existing text file and append incoming data to it. For example:

RECA MouseHole:Messages

will append incoming data to the file Messages on the disk MouseHole.

What: RR 10.2 and MF

From: Macowaco (Rockwell Int.)

I’m running under MF now and it seems to send the output to the window even when the window is under others. One has to set the SIZE resource in the same way as the one on the MHDL or one can just copy it in. I’ll try a DL next. Too bad the Supercharged DL doesn’t seem to work either. In fact the only DL I could do was with 1K blocks.

From: Macowaco (Rockwell Int.)

Well, I tried DLing in the Background and while it seemed to work okay, the DL’ed file was missing all it’s finder attributes as well as some other data leaving it useless and with a blank icon. Too bad. Now I wonder if Hyperhype has the same size resource and if I can get it to chug away in the background.

What: More MF - Problems

From: Golden Ears (Seal Beach, Ca)

Under MultiFinder, running SmartCom II v2.2B (regardless of whether or not it is the sole task) it will just assign the ol’ TEXT/MACA to all files downloaded ..... it isn’t just Scott Watson. All the finder attributes were blown away. Hayes supposedly will be showing a MF version with all kinds of upgrades at COMDEX and will ship shortly after. I’ll believe it when I see it. I still like it better than anything else. If VersaTerm worked as it could, I’d use it (haven’t seen this v3.1 tho....)

What: Release MultiFinder

The release of MultiFinder 1.0, System 6.0 and Finder 4.2 are on the MouseHole Download now. Somebody had uploaded the beta version yesterday claiming it was the release. It has been replaced with the real thing. Larry

What: MultiFinder Funnies

Just for yucks, do a COMMAND-OPTION select of “About MultiFinder” from the Apple Menu. True Holy Grail style. (it’s also in STR#129 if you are poking around)

What: Dynamac

Look for a large “financial services” company to buy a LOT of Dynamacs in the next few months. Their salespeople will apparently be using them to do fascinating things like show how rich you’ll be if you buy their insurance, etc etc.

What: New MacII ROMs

The new ROMs have ONLY the Slot Manager 24/32 bit addressing mode bug fix incorporated in them -- that’s it -- to avoid drastic compatablility problems between ROM revs. [As reported in the press, the Mac II ROM’s slot manager cannot properly handle a card with over 1 Meg of memory on the NuBus since it only runs in 24 bit mode. Certain NuBus memory cards over 1 meg, like National Semi require this bug fix. There may be other 24 bit bugs as Apple attempts to take the Mac II into full 32-bit mode. The memory manager is one that comes to mind. You apparently can get the bug fix ROMS by griping at your dealer. -Ed]

What: Soundwave/SoundDesigner

From: Tycho (Contel Business Systems)

Well, I’m not an expert, but here goes... In response to a previous query about the differences between these two packages, is a _very_ rough generalization: Soundwave is a newer version of another product called “SoundCap”, which is to be used with the old MacNifty 8 bit sound digitizer hardware for the Mac. This box will convert audio input from either a line or mic source into a digitized “sound sample”. SoundCap allows you to edit and enhance (reverb, flanging, etc.) the sound, to be then played back on the * MAC *, resulting is relatively low quality sound. SoundWave adds a few new features, most notably the ability to “hook into” filtering and processing modules which YOU can write in C or Pascal. If you want more details, I have the package and can look them up. ONE WARNING...Soundwave has a BUG which crashes the Mac. According to Impulse, the publisher, this should be fixed “real soon now” (I think Apple Tech support is helping them but they seem to think it’s not moving fast enough). Impulse says it happens only on some machines and not others, but MINE is one of the unlucky ones. SoundDesigner is a different animal altogether. It accepts digitized samples which are generated on professional quality audio “sampling keyboards” such as the Emulator II, Ensonique Mirage, Akai S900, etc. Sound Designer allows you to manipulate these samples in a much more sophisticated way, and then to download these back into the sampling keyboard, to be reproduced using the high quality playback hardware that they offer. Most of the signal processing can also be done in the keyboards, but what they provide as a user interface is obviously less friendly than our favorite computer. The bottom line between these two packages is that Soundwave is a neat introduction to sound sampling and manipulation, and may be useful for specific applications, limited to its lack of interface to MIDI (and hence the inability to play the sounds back via keyboard, etc.) and the low sound quality of the Mac audio hardware. SoundDesigner is an expensive, pro-music “peripheral” for the $1000+ sampling keyboards used by your favorite musical artists, and is worth every penny if you have such gear.

What: favorite MIDI app

From: Tycho (Xerox)

I have been trying to use the latest version of Performer (Version 2.1) and it does offer lots of enhancements over previous versions. Looping and looping within loops is supported, new dialog interface, lots of cutesie user interface tricks. However, I have two major complaints. It crashes, and it has an interface which is looking more like “M”/Jam Session/UpBeat with each release (and I’m repulsed by their idea of user friendly). I’ve given up and gone back to my Opcode Sequencer (2.5) after each attempt to work with Performer. It’s easy to work with, powerful enough, and gets the job done (without the frustration of locking up in the middle of a * HOT * recording session.

What: Denver MUG

From: Richard Hyman (Comp.Tech)

For those in the Denver area, a new MUG is forming for the south metro area. The next meeting will be on Tues, 3 Nov. at 6 PM. The location is the SAIC offices at 6021 S. Syracuse Way, Englewood, Suite 300. For more info, call Mike Phelps (SAIC) at 773-6900, or Blanche Cohen at 755-1525, or Rick Hyman at (w) 889-1200, (H) 973-8028.

What: Dove Port

From: Jim Reekes (CMS)

I had talked with the VP of Marketing and a couple other tech support people at Dove. I informed them of the problems of using their port with a standard Seagate drive. They were baffled and could not provide a solution. They promised to send a SCSI port for me to test with. As of yet, I have never seen a Dove SCSI port. I called back a month later, and again they promised to send me one. This goes all the way back to the MacWorld show in Boston. At this time, I do not recommend the Dove port.

I have been recommending the Julian System upgrade. So far nobody has reported to me any problems using it. I do not know, for sure, if it works. If anyone does indeed get the Julian port, please let me know how it works out.

What: DataFlame....

From: Golden Ears (Seal Beach, Ca)

While trying to assist a good friend who was duped into buying a DataFrame20 (not an XP) I discovered that things in the lands of SMS are not kosher. The bozo that answered the phone was obviously out of her depth even with fairly simple questions (although she did have an error sheet at hand, when I told her that the Manager 3.03 app crashed to an id=02, she said hmmmm, let’s see, that’s an address error....) furthermore she admitted that not only was that version of the manager a bad release but the subsequent 3.05 release was also known to be faulty, they SHIPPED BOTH of those versions to some poor unsuspecting customers. What galls me is that they could only offer to mail me a new disk which might arrive next week, no-one around there either had a modem or the brains to use it. After digging further into their documentation I discovered that they supposedly put their current apps on Compuserve/Genie/Delphi. I guess I will have to start up an account again. Why didn’t THEY know about this ? But the real kicker is this, after digging up their Initializer v2.6 (which was known to be good) the bloody drive is faulty. This is the last straw with DataFrame and SMS, we had somewhere in the vicinity of 18-24 of their drives at my last location and they were consistently a pain in the neck. (not as bad as HyperDrive, but then, is that possible?) Which of you have had a large number of units of a particular manufacturers SCSI hard drive in heavy use in house ? What are your experiences ? Whose would YOU recommend or purchase ? Why are so many manufacturers quality assurance programs seemingly non-existent? As bad as their software was, and outrageously priced as it was, the AST-4000 was the most stable combo I have used (I don’t think they even sell that model anymore) I had two of them running almost full time for over 18 months and no hiccups.

[I have had nothing but good results with DataFrame hard disks and with the company’s response on service problems. I highly recommend the drive and the company. We currently have five drives in house. Admitedly this was before the merger. Things may have changed. -Ed]

What: SuperMac

From: Jack Howarth (Rice University)

I believe the problem is that their tech support wizard left the company and no one there can fill his shoes.

What: HD MTBF

From: Macowaco

Well, my experience with lots and lots of HDs has been that the Apple HDs have always been reliable and worthy of a major gamble. Although, and quite surprisingly the ZapSS10s that were purchased (large quantities) have also shown to be fairly reliable too. If I had my choice of whatever I wanted regardless of price I’d go with the ProApp or CMS, both of which use the same HD HW. Lets face it, if you had to weight HD factors in making purchases- reliability is infinately more important then high performance. A note on Zap; while their SS10 has been purchased enmass at my workplace the only reason after their proven relibility is the low price. If one had a HD that had comparable reliability and under $500 then I’m sure my co. would look seriously at it.

What: Dead SuperMac 19'

From: Max (Microtimes Mag)

Well, it really happened this morning... I was using the ol’ Mac II system, equipped with a 19" SuperMac Spectrum (color) monitor. Turned my back on it for about two minutes, and when I came back, the screen was black. Blank. Dead. Pulled out the fuse, and it was fried. Installed a new fuse, and it too, instantly fried. For the record, the monitor had been in daily use since last June, and has never had a problem. Anybody else have a SuperMac Spectrum go down? Or is it my Karma? Or what? What’s a bummer is the dang thing weighs a whopping 85 pounds, so you can’t even UPS ship it back to SuperMac! I think rush- fragile-freight for something this big and heavy will come to about $75... just to ship it in for repair!

Since I have the unit for a magazine evaluation, we’ll be able to find out exactly what went wrong, and I’ll report back. In the meantime, many thanks to Mike Jewitt at Computer Partners in Stanton, CA. When the screen blew, I was on a skin-tight magazine deadline, and he’s provided me with a “loaner” mono screen. You can’t beat service like that... which means mucho thanks to Mike, and Kudos to Computer Partners.

What: MPW 2.0 Pascal

From: Venus (Jet Propulsion Laboratory)

My copy of MPW Pascal 2.0 finally arrived -- sans any documentation, of course. grrr. Anyway, seems mostly everything is v1.0 compatible. And I still don’t understand: whenever you compile a program, it creates a “%a5Init” CODE segment. What is this? Apple’s example, first thing, does an “UnLoadSeg(@_DataInit);” to get rid of it. What’s in this segment and is there anything useful you can do with it? Is it still an artifact from Lisa Pascal? What is the “_DataInit” item? TML Pascal never had such stuff.. Luckily, I find most of my TML code ports straight in, just change the compiler directives.

From: Rich Siegel (Think Tech., Inc.)

I think that the %a5init is for initializing data; the compiler and linker both support initializing data, although the only usage if this I’ve seen is initalizing everything to $D2D7 for debugging purposes.

From: Geoff Bryan (Latham & Watkins)

What: Getting the screen size in a DA

To get the screen size in a DA without using the A5 register, declare a “rect” type variable and call this procedure:

procedure GetScreen(var r : rect);     
const       
 CurrentA5 = $0904;    {CurrentA5 lomem global}          ScrnBitsOffset 
= 116; {offset ScreenBits.Bounds}     
type       
 IPointer = ^longint;       
 IHandle = ^IPointer;       
 RPointer = ^rect;     
var       
 Ptr1 : RPointer;       
 Handle1 : IHandle;    
begin      
 Handle1 := IHandle(pointer(CurrentA5));      
 Ptr1 := pointer(ord(Handle1^^) - ScrnBitsOffset);    
 r := Ptr1^;    end; {screen rectangle} 

You can then examine the rect, r, for top, left, bottom, right values. Good luck.

From: The Cloud (U C L A)

What: Test for Multifinder Friendly

Here’s how I do it in Pascal:

VAR  
 MfPresent: Boolean;   {global boolean variable}
CONST  
 WaitNextEventTrap = $60;  {determine whether}           UnImplementedTrap 
= $9F; {trap exists or not}   
BEGIN     
 IF GetTrapAddress(WaitNextEventTrap) =      GetTrapAddress(UnImplementedTrap) 
THEN  
 MfPresent:= FALSE     
 ELSE  
 MfPresent:= TRUE;  {rest of init code here}   
END;  
 
PROCEDURE MainEventLoop;   
VAR  
 Event: EventRecord;        
 Foo: Boolean;     
FUNCTION WaitNextEvent(mask: integer; VAR theEvent:      EventRecord; 
sleep: longint; mouseRgn:RgnHandle): boolean;  INLINE  $A860;   
BEGIN {procedure}     
 REPEAT       
 IF MfPresent THEN           
 Foo:= WaitNextEvent(EveryEvent, Event, 0, NIL)
 ELSE           
 Foo:= GetNextEvent(EveryEvent, Event);       
 IF Foo THEN  {handle the event!}
 UNTIL HellFreezesOver;   
END;  {MainEventLoop} 

Thanks to Jim Reekes for documenting the above technique in his “Zippy” MPW source...

What: Screen Size from a DA

From: Lsr (Apple Computer, Inc.)

To get the screen size from a DA, I might try looking at the port rect of the Window Manager port. You get this port with the trap GetWMgrPort. On a multi-screen system, this should be the size of the screen with the menu bar. If you are interested in the total screen area, look at the low memory global GrayRgn, which will be the total screen area, minus the menu bar.

Also, am I correct in assuming that when a MacII is set (for example) to 8 bits/pixel mode, the ScreenBits global will be 8 times normal size? I’m doing CopyBits calls from ScreenBits to an offscreen bitmap, and there is a dramatic slowdown in the program. Since screenBits is a BITMAP and not a PIXMAP, does it still contain all the depth info? Tech Note#120 (color pixmaps) demonstrates how to check the pixel depth of the screen, using the new routines....can I just calculate this info from ScreenBits instead?

What: DrawPicture

From: The Cloud (U C L A)

I also have a lot of pictures, accessed through picHandles and drawn in my window via DrawPicture. I’d like to be able to find a Q&D way of displaying them in color, preferably the “lazy” way, by calls to ForeColor and BackColor. Naturally, these only affect drawing done by the Quickdraw pen...or do they? Is there a way to do it without a major rewrite? Also, what happens if I set myPort^.bkColor to, say, blue, and then call InvertRect? will the result be black, or the color complement of blue? Thanks in advance for any experienced advice...

What: Color Quickdraw Answers

From: Chief Wizard (Apple Technical Support)

First of all, the QuickDraw global screenBits doesn’t have enough information to let you calculate the depth of a Mac II screen. For example, using a standard Apple monitor, screenBits.bounds would be (0,0,480,640) (top,left,bottom,right). Therefore, you would assume that a single row was 640 pixels across. This is correct.

Now, to get the depth, you would probably divide screenBits.rowbytes by the width of the screen to see how many bytes/pixel were being used. To get the number of bits/pixel, you would multiply by 8. This won’t work.

In 8-bit mode, screenBits.bounds is still (0,0,480,640). And screenBits.rowBytes is $C000. The high bit is set to indicate that it’s a PixMap instead of a BitMap, so we remove the high bit and get a rowBytes of $0400, or 1024. Using the formula above, you would get an answer of 12.8 bits/pixel. Oooops. What happened?

The Mac II video card hardware is set up so that there is ‘extra’ memory to the right and below the actual screen ram. This memory MUST be included in rowBytes for the math to work, but it is not included in bounds because it isn’t actually part of the displayed area. This was done to make the hardware simpler.

Although all video cards may not do this, you can see why you can’t use screenBits to calculate screen depth. (See TechNote #117 for more information about using the fields of a BitMap.)

On Pictures: The pen color used when creating a picture is actually stored within the picture. So if you call RGBForeColor (or just ForeColor) before drawing a picture, it won’t have any effect; the pen color that was saved with the picture is used instead.

There is, however, a fairly simple way to “colorize” pictures and bitmaps that were originally just Black and White. The operation actually only works on BitMaps, so if you want to “colorize” a picture, you would first draw it into an offscreen BitMap and then CopyBits it to your final destination.

There will be a TechNote released in early November that explains this in detail. It’s a little long to go into here. But try this:

DrawPicture into an offscreen BitMap.

RGBForeColor(someColor);
RGBBackColor(someColor);

CopyBits from offscreen to your color window.

What: Color with a K

From: The Cloud (U C L A)

Mr. Wizard: thanks for the detailed reply. Yes, I was indeed thinking exactly as you described...guess I’ll have to GetMaxDevice, etc. to determine the screen depth after all. On the subject of “colorizing”, what effect do source transfer modes have on the color? Let me explain what I’m trying to do: I have a single picture (b&w) that serves two purposes: it’s either drawn normally (a “white” game piece) or inverted (a “black” game piece). Now, if I want to have “red” and “blue” pieces on a color monitor, I would need to first draw the picture into an offscreen bitmap, set the on-screen window’s ForeColor to red, then CopyBits the picture onto the screen in “invert” mode (notSrcXor), so that the “white space” in the original picture becomes red. Am I on track still? Hopefully this is of general interest... As I don’t have a MacII w/color here in front of me, I have to solve the problem “in theory” first....

What: Screen depth

From: Lsr (Apple Computer, Inc.

To find the screen depth you have to examine the pixelSize field of the pixMap data structure.

I think if you want to display red and blue pieces, you can set the RGB fore color to red or blue and use CopyBits to display the piece. CopyBits will make all the black bits in your offscreen bitmap be the same as the foreground color.

What: Bitmap SLowdown

From: Lsr (Apple Computer, Inc.)

Copying from an offscreen bitmap (1 bit per pixel) to an 8-bit screen will be slow because Color Quickdraw will have to expand each of the offscreen bits to the appropriate byte (8-bits) corresponding to black or white. You will get the maximum copying speed when the source and destination are the same depth and have the same color tables.

What: Red Game Pieces

From: Chief Wizard (Apple Technical Staff

I’ve had a lot of success using CopyBits to copy a 1-bit BitMap to a multi-bit PixMap. This is especially nice when trying to add color.

To color your pieces, you can do as you suggest except you don’t need to use the funky srcNotXor mode. Instead, you can just colorize the white part of the picture. In other words, do these steps:

DrawPicture into an affscreen BitMap/PixMap. TechNote #120 shows how to create an offscreen PixMap. It’s easy cause there is a call that creates the PixMap data structure for you. You just have to fill in a few fields.

Next, when you want a red piece, Set RGBBackColor to red and CopyBits the piece. If you wanted a blue piece, you’d set RGBBackColor to blue and then CopyBits.

It might be easier to use CopyMask, using an inverted version of the piece bitmap as the mask. That would let you CopyBits the piece without obliterating the entire CopyBits rectangle. (Does that make sense?)

Also, if you are relying on Xor mode to draw and erase your pieces, I’m afraid that you will find the colors less than desireable. Xoring and inverting colors usually results in less-than-appealing colors.

What: The Red and the Black

From: The Cloud (U C L A)

Actually it does make sense. I won’t need to do any “inverting” of pictures when color is present. I’m still concerned about the performance issue Lsr mentioned, when copying from my 1-bit/pixel offscreen bitmap to the multi-bit/pixel screen. My current strategy is to allocate these bitmaps on the fly, and I discovered something interesting: if there isn’t enough contiguous space in the heap for them, the Memory Manager has to do its thing, compacting the heap. This causes a “wait” of between 1 and 3 seconds! The problem was solved by calling CompactMem at crucial points with a size slightly greater than the combined sizes of the bitmaps I was creating, so that there would always be a contiguous block free for them as they were needed. Naturally, multi-bit modes throw a curve at this approach; this is why I was concerned about finding the pixel depth.. An interesting sidenote to this discussion is the use of SuperPaint or GraphicWorks to create “color” pictures, which can then be pasted into a program (retaining fgColor info). But if you want to change the picture’s color within your program (like I do) its value is limited at best...

What: TML Pascal 2.5

From: The Wombat (Edmonton, Alberta, Canada)

I just received my TML Pascal 2.5 update (finally). There are no apparent changes to the compiler, it still bombs frequently on my Mac II whenever the monitor is in color mode. The compiler doesn’t generate inline 68881 code, a special unit is linked to access the chip directly. Also the bug described in June/87 MacTutor still hasn’t been fixed. I’m rather disappointed.

What: LightspeedC New Stuff

From: Rich Siegel (Think Technologies, Inc.)

Multifinder Niceties are on the way. We’ll have a MultiFinder-compatible version when MultiFinder is released. The source-level debugger will be coming later...

I honestly don’t know what the timeframe for The Big One is, though...

What: ReadySetGo 4

From: Dave Kosiur (Connections)

Just received my upgrade to RSG 4.0 yesterday. It’s great! Style sheets, irregular graphics runaround, facing pages, the list goes on.. I particularly like the ability to place graphics over facing pages for brochure design and the like. I’ll keep you posted if I find any bugs.

From: Keith H. (Etak) #171

I also got my 4.0 upgrade today. It is nice to have a real manual that is nicely done, finally. The package also includes a little book called The Grid Book, which is a nice guide to page layout planning.I think they may have got it right this time as the program seems to perform better than ever. I will also watch for bugs but after the M.S. Word bug fiasco I went through anything else will seem minor!

From: Mysteray (American Zettler)

RSG4 is quite an improvement from the 3.0 version I suffered with (I got to h-a-t-e the “non”-scroll bars utterly). But it still has some shortcomings (or bugs, perhaps).

The double-click selection of a word now selects the trailing space, which is fine for cuts and pastes (like in fixing word order), but quite lousy for doing underline attribute, which nobody can work out so it doesn’t underline spaces if an entire sentence is highlighted.

The 10%, 80% and 90% fill or pen patterns do not give a true gray fill. I noticed that the arbitrary run-around seems to repel more on the right side of text than on the left, perhaps because their algorithm includes the trailing space at the right end of a line, even though it is not selectable on screen.

It is still slow on the first line below a graphic box with run- around on. I have had it do other strange things but not often enough to make conclusions to report.

They fixed the reverse direction shift-click selection of a block (almost), which most Mac text processors fail (including QUED). For those who would like to try this on their favorite editor, here is the test:

(1) set insertion point somewhere (several lines must be above)

(2) shift-click the I-beam say three lines above the first one;

(3) the text between these two points will highlight;

(4) now... shift click anywhere within the block. What do you see?

It should select the text from the lower mark (the first chronoligically). Repeating the shift-click should continue to select text from the lower mark backwards to wherever the click\was made. I only see this in MacWrite. RSG4 does it correctly except when the shift-click point is on the same line as the lower (original) mark point. Anyone else find an editor that does this correctly? (I don’t have PM or Word to try here.) RSG4’s new tab settings are a Godsend.