The Pick of the Hacks of MacHack

By Scott T. Boyd, Apple Computer, Inc., The MacHax™ Group, Contributing Editor

It’s midnight in Ann Arbor. A calm, clear night, the world is at rest. I reach for the Special Menu and choose Shutdown. A flick of the wrist and the disk drive slows to a rest. Snap, the SE screen goes black. Unplugging everything, I pick up the Mac and the drive and leave the hotel room. Walking down the quiet hall, I wonder to myself how much longer I’ll be up. I’ve been up since early that morning, and it looks like it’s going to be a long night.

Turning the corner and opening the door, I hear the first sounds of life since we got back from dinner. Ted Nelson had spent the evening regaling us with stories of Xanadu, but he hadn’t once mentioned Kubla Kahn. Odd. The din at dinner as a couple of hundred hackers had consumed dish after dish of oriental food had been something. Dinner had just been the beginning. That was then, this is now, and the noise is tremendous.

Both doors of the suite are open. Inside, the walls are lined with machines. It’s midnight and about a hundred people are crammed into the room, staring at small phosphor screens. “Oohs” and “ahhs!” fill the room as twenty people cluster around one particular machine. “Have you seen this one?” asks Greg. Leonard jumps in with, “What about this one?” “Wait, you’ve gotta see this one first!” someone else shouts as he rushes up with a diskette.

In a far corner, Sean is setting up the SE I just brought in. He’s not done yet and is still working to be able to show off his hack. The noise in here is matched only by the heat. Another Sean is setting up an HP ink-jet printer on the other two feet of the counter top. There’s no other space anywhere near a plug. He is showing off 300 dpi printing that is nearly silent. Of course, you have to use your imagination, but I presume you’re already doing that.

_______________________

What possible excuse can there be for such a bizarre circumstance? Why, MacHack ’88’s HACK of the Year Contest, of course!

MacHack, sponsored for three years now by Expotech, Inc., brings together about two hundred of the best hackers from the Macintosh community. This year, it attracted several celebrities, including Jim Blinn, who brought us the NASA fly-by sequences and The Mechanical Universe animations on a shoe-string budget. It also brought Ted Nelson, who gave the keynote talk about Xanadu. Xanadu is a proposed global docuverse made possible through the magic of hypertext and tumbler arithmetic. Of course, most of the attendees actually program the Macintosh, and quite a few make their livings with that beige or platinum box.

______________________

Twenty-two hacks are in this room. Several are still in development as their authors compile and run them for us. It’s about two o’clock and I’ve finally made the rounds and seen most of the hacks.

Where’s Sean? He and I spent most of the day building our hack. It still isn’t working, but we know we’re close. The SE’s gone. Not that I’m panicked or anything, but I run back to Greg’s room, which had been serving as hacker central for the past couple of days. Sure enough, here’s Sean, madly coding. He babbles something about madness and noise and paces the room briskly, much too briskly for two in the morning. He turns back to the keyboard and resumes the quest. With everything under control, I head back to the fray

_______________________

Earlier in the day, I was sitting in the back of one of the conference sessions, hacking on my SE in the dark. Roy Leban (of Ann Arbor Softworks, now of Ashton-Tate) came over and shoved a disk in my machine. Rather leery of unknown software, I hesitated. “Trust me, I’m a programmer,” he said. “Reboot and behold marvels!” Right in the middle of a conference session, “Bing!” Then, the strangest thing happened. When Roy opened the floppy in the Finder, one of the icons began to spin. Spin!?! Finder icons don’t spin! Wrong again, McGillicuddy! This one does, and I don’t even get to keep a copy of the hack! “Animator’s not ready to give out yet,” Roy said as he sounded a theme to become quite familiar as the day progressed.

Throughout the day, people kept cornering me with, “You’ve got to see my hack. Where’s a machine? When do you have a minute?” Finally, at dinner, the thought hit, why not a midnight hack fair? Let’s get everyone in one place and do it just like back in high school at the science fair? One simple announcement and the stage was set. Of course, simple is an interesting word. You know what I mean if you’ve ever tried to talk over Dave Feldt when he’s talking to two hundred well-fed people, all at once.

On with the story. What wondrous marvels did we behold? One category that emerged addressed the need for managing limited screen space, a problem worsened by multiple programs using the same screen under MultiFinder.

ShrinkWindows by Shane Looker adds a way to shrink any or all of the windows in your current layer. When shrunken, they are about an inch big and can be moved anywhere on the screen. When you make them big again, they go back to where they were before you shrank them.

Holey, by Larry Rosenstein, takes an altogether different approach. Holey removes the content region from all the windows in the front layer. Your windows don’t move, they just become transparent, leaving behind just the frames.

Slider, by Bill Johnson and Ron Duritsch, takes yet another approach and lets you drag all or some of your windows as a group, just as you might drag a bunch of objects in a drawing program.

The last entry to fit into this category is RearWindow, by Allan Foster. This one-patch hack lets you drag an icon from one window to another while in the Finder without changing the window order. This comes in really handy when you have a small window in front of a big window and want to move an icon from the rear window to the front. Normally, the Finder would select the rear window, covering up the front window, leaving you with nowhere to drop your icon. Now, hold down some modifier keys while you click, and the windows don’t change.

The rest of the hacks don’t categorize so neatly, but here they are.

ShutDownSoundINIT, by Leonard Rosenthal, plays a sound at shutdown time. You may already have this one. I’ve used the “I’ll be back!” sound for months now.

Leonard also submitted Edit, the text-editing DA with nine hundred or so features. I’ve also been using this one for several months.

There were two time-related submissions. Larry Rosenstein entered TimeKeeper, a MultiFinder-compatible clock with a resizable round window and selectable chime sounds. My favorite is the Big Ben sound.

Darin Adler and Sean Parent started from scratch at the conference to bring us OverTime, a clock in a rectangular “window” that stays on top of everything by removing itself from the window manager’s regions. Someone called this “simple but creative.” Does the label “simple” allow for seven patches and breaking all the rules?

Greg Marriott brought out FaceLift, an INIT that lets you give any of your disks its own icon. This is truly a simple but creative hack. All of the source code fits in less than a page, and the user interface couldn’t be simpler.

The hack that clearly required the least amount of code came from William Leininger. He brought the only hardware entry, a miniMac Screwdriver which consisted of a Torx™ bit with a slot cut in the back so you can open Macs with a regular screwdriver.

Small, specific utility hacks included DTR Resetter from Stuart Schmukler, BNDL Rx from yours truly (it checks all the files on a disk and tells you which files have bundle bits set and whether they have a BNDL resource), and Disk Switch from Larry Rosenstein. Disk Switch lets you eject a disk using -1 or -2 while the disk-switching alert is up. Those of us who use only hard disks these days may have forgotten, but there were days when a hack like Disk Switch would have saved us some grief. This fixes a bug in the ROM that checks for ASCII 1 or 2 instead of ‘1’ or ‘2’ (ASCII 49 or 50). People with a Control key can use Control-A or Control-B without the patch.

In a category by itself, Darin Adler’s Buffered Menu Bar stands as the hack that nobody can be really sure that it does anything. It works by making sure that the menu bar never flickers; it does its job by making sure you don’t notice it.

Christoph Sold brought the hack from the furthest away by bringing Dissolve? all the way from Germany. Dissolve? puts away a dialog by doing a color fade, where the dialog gets dimmer and the background gets brighter. He plans to generalize this to fade from any color picture to any other color picture. A pretty neat hack.

NetTrain runs a train across the screen of any participating machine on the network. Sean Flynn and I got it running at about 4 a.m. the night of the Hack Fair. I’m still amazed that no one called the hotel management to complain when it ran the first time Gordon and Mike Sheridan worked hard to get us some train sounds with Gordon going so far as to walk to the highway to get the sound of passing semi’s.

Kent Johnson showed RainDance, which does some pretty dazzling things with circles to get the effect of raindrops on a pond.

Dumbo, by Jay Zipnick, showed an animated Dumbo flying around a window. Appearances can be deceiving, and Dumbo is a good example. The animation alone is nice and smooth, but the real stuff of this program lies in the fact that the animation almost never stops running, even when a modal dialog is in front, when Dumbo isn’t the front MultiFinder layer, when Dumbo’s window is being dragged around, and even when a menu selection is in progress.

To round out the bunch, Rod Magnuson and Steve Kiene show Tear-Off Menus. You can tear off any regular menu and leave it anywhere on your screen. Menus stay on top of other things. If the menu is too big for the screen, it appears as a tab that pops up a menu when you click on it. These guys have done some good work. They continued to improve it throughout the conference.

Which hack won? That’s hard to say. Any one of these hacks, had it been the only one there, would have provided a topic for lively discussion. These were all excellent hacks, each having interesting solutions to problems. Some looked especially good because they solved a tough problem with a simple insight. Others looked deceptively simple, hiding amazing complexity behind their interfaces. Some were just plain fun. All of the hackers in this contest were winners because a good time was had by everyone.

Perhaps the most convincing measure of all is that not a single person asked or seemed to care what the prizes were. The thrill of the competition and the accolades and camaraderie of their peers were sufficient.

So, do you feel like you missed something fun? I hope so. In the tradition of MacHack’s past, a good time was had by all. Get your hacks started, next year isn’t that far away.

For these hacks, see the following mini-articles. Holey, and FaceLift are available on the MacTutor disk.

_________

For more information about MacHack, contact

Aimée Moran or Carol Lynn

Expotech, Inc.

1264 Bedford Rd.

Grosse Pte Pk, MI 48230

313-882-1824

ARPA: aimee_moran@um.cc.umich.edu

Compuserve: 72000,40

GEnie: Expotech

Applelink: X1156

_________

MacHax™ is a trademark of The MacHax Group. MacHack is a trademark of Expotech, Inc., an unaffiliated but friendly firm.

Tear-Off Menus

Rod Magnuson & Steve Kiene

An Overview

“Tear-Off Menus” is an INIT that allows menus in most applications to be “torn” off. Menus then become floating menus that stay in front of all other windows, we will refer further to these menus as TOMs (Tear Off Menus). DAs and the hiliting of TOMs are handled correctly-TOMs are only hilited if the application is the front-most window. If a DA is the front-most window then only the DAs menu can be “torn” off.

“Tear-Off Menus” works under MultiFinder. “Tear-Off Menus” includes a cdev that allows customization of the INIT and a FKEY that allows hiding/showing the TOMs. The FKEY is installed/removed by the cdev for user convenience. “Tear-Off Menus” works on anything from a Mac 512KE up, as long as the latest System is running. It also supports color.

“Tear-Off Menus” provides for two variations of the TOMs. The first looks like a standard menu with a close box to the left of the inverted menu title. The second is only the inverted menu title with a close box to the left and a triangle to the right; clicking in the triangle will pop up the TOM’s menu. Variation one is what you typically get when you simply tear off a menu. If you hold down Command or the menu is too large to fit on the screen, then variation two will be used.

Holding down Option while choosing a menu from the menubar will restrict menus from being “torn” off. Holding down Option and clicking on a TOM’s titlebar will send it to the back of the TOMs. Holding down Option and clicking in the TOM’s close box will close all TOMs. You get some options about whether “Tear-Off Menus” is installed at boot time, depending on your cdev choices. You also get control of whether “Tear-Off Menus” is installed in each application at launch time. Holding down Backspace at boot time will remove any traces of “Tear-Off Menus”. The INITs icon reflects your choices.

Problems

“Tear-Off Menus” has problems running in most paint-type application’s. Paint application’s don’t seem to care about their window’s visRGN and draw on the TOMs too [custom blitters generally want speed, so they often ignore regions -- ed]. Surprisingly, “Tear-Off Menus” gets along well with HyperCard. [Well, since Bill A. wrote HyperCard and created regions, is it really that surprising? -- ed]. When we have both HyperCard tear-offs and our tear-offs, we don’t see any problems. Personally we prefer our menus to HyperCard’s. Ours allow better selections; you can click and hold the mouse down and select (this is often called tracking). You don’t necessarily have to select the first item you clicked in.

Future

We have lots of ideas for expanding “Tear-Off Menus”. We would like to see things such as movable menubars and other things that would make it easier for the user to access menus. We are developing a “Tear-Off Menus” developers kit that will allow developers to implement “Tear-Off Menus” easily by adding one global and two procedure calls to their existing programs. We are currently in negotiations with a publisher for “Tear-Off Menus”. We would be glad to answer any questions you may have or receive any of your comments/suggestions. Give us a call.

TOM/cdev: “Tear-Off Menus’” cdev

TOM/ReadySetGo!4

Practical application of “Tear-Off Menus” in

ReadySetGo!4

TOM/Finder

“Tear-Off Menus” running under the Finder

Holey, the FKEY

Larry Rosenstein

Holey is an FKEY that empties the content region of all the windows in a layer. This is useful with MultiFinder. With Holey, you can make your windows transparent to see through to and clcik on a Finder icon or another application’s window.

It is a simple variation of the technique described by Greg Marriott in the October 87 MacTutor. It works by modifying the strucRgn and contRgn of all the windows, which is a very bad thing to do, especially under MultiFinder (where this FKEY is most useful). This puts the FKEY into the category of extra-slimey hacks.

The basic routines can be recombined to achieve a variety of effects. For example, instead of toggling, one FKey could make the windows transparent and another restore them.

Note that we do no error checking on the region calculations, so we could crash in low memory situations.

Known Bug: In MPW if you activate a transparent window, the shell computes the active region of the window. Since the window is transparent, the region is empty. If you make the window normal, the shell becomes confused until the window is activated again. -- Larry Rosenstein (lsr@apple.com)

Before Holey

After Holey from FullWrite

After Holey in the MacPaint layer

MPW Build commands (to make FKEY 9)
Pascal Holey.p -o Holey.p.o 
link Holey.p.o -o ‘Holey FKEY’ 
  -rt FKEY=9 -sn Main=”Holey” 
  -m ‘ENTRYPOINT’ -t ‘FKEY’ -c ‘FKEY’

Holey Code
UNIT Holey;

INTERFACE
USES MemTypes, Quickdraw, OSIntf, ToolIntf;
 
 PROCEDURE EntryPoint;

IMPLEMENTATION
PROCEDURE HoleyFKey; FORWARD;

{ The main program.  We need this entry point here because we use nested 
procedures.  Otherwise the nested procs would be first in the segment. 
}
PROCEDURE EntryPoint;
BEGIN
 HoleyFKey;
END;

PROCEDURE HoleyFKey;

 VAR  w:WindowPeek;
 v:INTEGER; 
 savePort:GrafPtr; 
 topChange: WindowPeek;
 { topmost windows that changed }
 clobber: RgnHandle;
 { union of all changed windows }
 
 PROCEDURE Setup;{call this to setup things}
 BEGIN  
 GetPort(savePort);{ Save the current port }
 topChange := NIL; { Init some variables }
 clobber := NewRgn;
 END;
 
 PROCEDURE WrapUp;
 { call this after munging the windows }
 BEGIN
 IF topChange <> NIL THEN
 { have system refresh things }
 BEGIN
 CalcVisBehind(topChange, clobber);
 { cause update events }
 PaintBehind(topChange, clobber);
 { refresh window structures }
 END;
 DisposeRgn(clobber);
 SetPort(savePort);
 END;
 
 PROCEDURE NoteWindow(w: WindowPeek);{ remember info about window }
 BEGIN
 IF topChange = NIL THEN
 topChange := w; 
 UnionRgn(clobber, w^.strucRgn,clobber);
 END;
 
 PROCEDURE MakeTransparent(w: WindowPeek);
 BEGIN
 WITH w^ DO { windows are nonrelocatable }
 IF visible THEN 
 { don’t call this on invisible windows }
 BEGIN
 NoteWindow(w);
 
{ empty content rgn and fix structure rgn accordingly }
 DiffRgn(strucRgn, contRgn,strucRgn);
 SetEmptyRgn(contRgn);
 END;
 END;
 
 PROCEDURE MakeNormal(w: WindowPeek);
 BEGIN
 IF w^.visible THEN
 { don’t call this on invisible windows }
 BEGIN
 WITH grafPtr(w)^.portRect DO
 SizeWindow(windowPtr(w), 
 right-left, 
 bottom-top, TRUE);
 NoteWindow(w);
 END;
 END;
 
 
BEGIN
 Setup;
 w := WindowPeek(FrontWindow);
 { Loop through all the windows }
 WHILE w <> NIL DO
 BEGIN
 IF EmptyRgn(w^.contRgn) THEN
 { already transparent }
 MakeNormal(w)
 ELSE { window is normal }
 MakeTransparent(w);
 
 w := w^.nextWindow;
 END;
 WrapUp;
END;
END.

Shrink Windows

Shane Looker

Small windows for the rest of us... Shrink Windows is a utility package for the Macintosh designed for use under MultiFinder.

One of the biggest problems with using MultiFinder is the clutter left on the desktop when multiple applications are running. Shrink Windows allows the user to reduce the size of windows so they can be moved out of the way on the screen.

The above figure shows Shrink Windows in use. After moving several application windows out of they way, I am able to get to my desktop easily to trash files or launch applications I have on my desktop.

The current version of Shrink Windows is difficult to install. When a final version of Shrink Windows is available, it should be installable as an INIT. I hope to have a release version of Shrink Windows ready by the end of summer. It will be released into the public domain (i.e. it is not available on the MacTutor distribution disk). If you need it immediately, contact me at Shane_Looker@um.cc.umich.edu -- Shane Looker

FaceLift

Greg Marriott

FaceLift, the short short story.

FaceLift is an INIT which assigns user-defined icons to hard disks. It walks down the drive queue in low memory and makes a Control call to each of the disk drivers. Most disk drivers respond to Control calls 21 or 22 by returning a pointer to where their icons can be found. FaceLift (rather rudely) places the new icon in memory over the old one. You can associate an ICN# with a volume by giving it the same name as the volume. Use ResEdit to put your custom ICN#s into the FaceLift file. Use the “Get Info” feature to name the ICN#s.

FaceLift’s methods have several problems which cause it to exhibit strange or confusing behavior. If the driver is in ROM (like the floppy/HD20 driver), then the icon is too. FaceLift has kind of a hard time blasting a new icon into ROM! If two drives share a common driver (such as the latest Apple SCSI driver, which doesn’t load duplicate copies of itself), then they end up sharing the same icon, too. It just happens to be the last one FaceLift blasted in. FaceLift doesn’t use Control code 20, which some drivers understand. Finder-mountable devices should respond to Code 20 by returning the ID number of an ICN# resource which can be loaded (presumably from the system file). FaceLift would have a hard time replacing the bits in a resource that hasn’t been loaded yet!

Bit-O’History

The original idea was realized in an application by Greg Marriott and Scott T Boyd. The program allowed the user to pick any ICN# in the Desktop file and assign it to the boot volume. This only worked with special versions of Carl Nelson & Associates disk drivers that had a special csCode to accept the new ICN#, and was only good until the next restart.

The next step came when Ron Aldrich created the first INIT version. This little piece of code worked at startup and applied the same ICN# to each volume that happened to be mounted when it ran. Ron was the one who came up with the “blast the old icon out from under the driver” idea.

This (almost) published version is by me, Greg Marriott. I introduced the idea of assigning a different named ICN# to each volume at startup time. FaceLift now mounts each drive in the drive queue and looks for an ICN# whose name matches the volume name. If it finds one, the disk gets the new ICN#. If the search fails, the disk retains its default.

Vapor Hack

FaceLift’s problems have been solved in FaceLiftPlus, which I have been madly coding the last few days. The new one works (at least it seems to), but I needs some more time to make sure. FaceLiftPlus will be the subject of an exhaustive article in next month’s issue (since I didn’t quite make the deadline for this month...). The new version is much better than the old one, so I’m not furnishing any source code this month. The new one lets you change disk icons instantly by simply changing the disk’s name in the Finder! The executable version of the original is on the MacTutor™ source code disk. Enjoy (until next month)! -- Greg Marriott

__________

Well, that’s a wrap. Get your hacks ready for next year and the Third Annual MacHack HACK Contest -- scott

---------- Sidebars ----------

Detailed Hack Reports