Nov 86 Mousehole
Volume Number: | | 2
|
Issue Number: | | 11
|
Column Tag: | | Mousehole Report
|
Mousehole Report
By Rusty Hodge, Mousehole BBS
WORD
Tax Free
Thought you might like a peek on the progress of Microsoft Word 3.0. I know, I know. Up to now it was Word 2.0, but it appears Microsoft feels the jump all the way to 3.0 is warranted. I think they are right. Even the first beta shows the outstanding level of design and performance improvements. Some of the features are already public record (built-in spell checking, outlining, etc.) but a lot of the subtleties must be seen to be believed. The outlining is completely interactive with the text, making idea editing not only easy, but promoting a new level of working with writing tools. It is fantastic. Goodbye and good riddance to ThinkTank, Acta. (Welcome to More, though.)
The spell checker has dictionary-based hyphenization. There is moderate kerning control. There are style sheets (like in the upcoming Pagemaker 2.0) that make for powerful formatting ability. There is a preview page function that really gives you a lot of control over your document's final appearance, and a clever way to insert page numbers anywhere you want automatically, as well as built-in indexing, table of contents, line numbering, etc. Built-in graphics ability for lines, boxes, etc. Nice pagination, and some of the best features: real interactive rulers and customizable menus.
And much, much, more, of course. Very well done, very well thought out. And it's a joy to use (when it stops crashing). All indications are for a long and careful test period to fix everything right. Microsoft really wants this one to be right, like Excel. There's lot of competition, what with Word Perfect and all, but the competition's going to have a hard time of it given Microsoft's expertise with the Mac interface showing up so superbly. This is one incredibly powerful program and is thrilling to use. It is also hungry: it uses up 625K of RAM and takes up 2 disks. If they have an offer for an upgrade....do it!
Hard Disk Shutdown
Rusty Hodge
Corsair showed me something on the Dataframe. If you do a shutdown and then turn off the Mac after the beep, when you re-turn it on, the Dataframe and the ProApp and apparently all the other drives will boot in about 20 seconds less time. *If* you turn off without shutting down (or even just reset), it will take 20-30 or so seconds to do something, who knows what, before it boots. What magic thing happens if you do a shutdown? You got me, but not doing it must cause some sort of disk media check.
That's Why
Randy Saunders
For all HFS volumes there is a thingy in the volume header called the "clean dismount" indicator. This tells the file system that the disk was correctly updated before it was ejected. If this is not on then when a volume is mounted, the file system goes out and checks the catalog B-tree and extents B-tree to make sure they don't have any apparent damage. I don't think it can do much to fix it, and none of my disks have ever failed, but you probably get a cute bomb box.
When a disk is first written to, the "clean dismount" is reset. If you turn your Mac off it never gets set back on, and the file system does its little number for 10-15 seconds (for 20 MB). If you don't have RAM cache on, you almost assuredly have good stuff on disk, but the power could have failed while you were copying folders (or some busted program could have hosed things) so it's really not safe to assume the disk is OK when it only takes 10 seconds to check.
By the way, I never turn my hard disks off. They take 30 seconds to spin up and that's a lot more time than the mount check takes. I just select "Shutdown" and hold the mouse down and turn the MAC off. One blink of the active light it probably a seek to the index track because the controller has noticed that the MAC is off-line. I have had a lot more disks fail at start-up than any other time (we have lots of ST-506 disks where I work, not only in Macs).
TOPMAIL & Mac Rumors
Macowaco
Finally got it and it was worth the wait. It's got all that Inbox has except for the expensive and unnecessary bells and whistles. AND nobody has to give up a Mac! Only, it won't install on a HD20 and sometimes it won't install on a floppy. No matter what I do I can't find why...oh well. It works fine with MacServe, which by the way seems to be on its way to being the one to beat in the server game. Nobody can out do it for the price. A good solid combination, MacServe, TopMail.... and Mazewars (which won't work with Macserve!) [Due to the fact that MacServe has problems with HFS, we think TOPS from Centram Systems may be best. -Ed]
As for rumors, I haven't dropped any lately so I thought now is a good time.... Apple is gonna stop on release of the new big one, but will go ahead with the new smaller one with the 68020 in it. The rumor is Apple is gonna put the 68030 in the biggie. Oh my, 8 MIPS.... The talk now is of two product lines. The first is the compact line which consists of the closed machines; Mac 512E, Mac+, IIc.... small footprint and low cost.
The second are the soon-to-be-expandables that we've all been drooling about.... 68020 at 16 (!!!) Mhz (sounds like 2 mips). The ROM will include color quickdraw routines and will be 256K. It will come with 1 meg. 256K SIMMS, but enough room for expansion up to 1.4 GIGABYTES (!). The internal bus ain't VME (sigh) instead they're going with a bus developed by MIT and bought by TI.... it don't need no dipswitches. They're gonna use a Foley sound chip which has only four voices and won't be as powerful as the GS's: they think this machine won't need it. It'll come with a 12" monochrome, 73dpi at 640/480, a later release of a color 13" monitor with the same resolution. 6 slots with one used for the monitor leaving five for the user. 2 SCSIs with one internal (!) and the external using the DB25. Two serials ext. using the Din8s. Another later release will be a 19" color with 1280/960 for 88dpi. Expansion cards will include a 68551 Motorola MMU and a 6881 at 16 Mhz. A 5.25" DD will be available with SW for MS-DOS file compatibility. A third party is working on a 286 card which will use the 5.25DD for full MS-DOS application compatibility.
They are also working on a 386 and an RT emulator (WHY RT??? It's a flop!) It'll have Unix v5.2 with the Berkeley 4.2 stuff in it, and it is AT&T approved. Quickdraw commands in portable C for UNIX compatibility.
ALL BY MARCH 1987!
Other things include a definite Appletalk File Server which will not be anything like the 3COM. Confirmed laying of Appletalk protocol on top of Ethernet. They are looking more at the DEC market than the IBM (smart). MSWord 3.0 will have a DCA filter and I'm told it will allow one to create full document compatibility with Mainframe systems.
Now for what I think is Science Fiction....
A fiberoptic Appletalk setup which will deliver Mazewars at 20 Mbits/sec (!) due late 87 or early 88.... the 68030 one year from next release, and.... 68040 two years from next release.
I'm sure there was a lot more, BUT SINCE I WASN'T THERE I DON'T KNOW. HINT! However, if I was there I would have signed something that would have forbidden me from dropping all these shoes.
NuBus
Frank Henriquez
Oh please! Not the NuBus! That's the only MIT-TI bus I know of, and it doesn't compare to VME. Well, you wouldn't expect Apple to follow any established standards...
Roll Out Date for New Mac
Mike Steiner
When I was at Apple, Phoenix for RSP training on the //GS, I said to the trainer, "See you in January." He asked, "Why?" and I replied, "for the new Mac training." He said that he'd see me in March in that case. Guess that March will be when the new Mac sees public light.
Laserwriter Trix
Richard Clark
Most of you LaserWriter hacks out there know about the "command-F" trick to get the PostScript command file for your document instead of the document itself. (Oops - clumsy phrasing. Command-F puts the PostScript file on disk that the LaserWriter would otherwise use in printing your document.)
But, I haven't heard any mention of "Command-K", which dumps the current PREP file into a text file called PostScript. (Try it - you'll get a TEXT file containing the code sent from Laser Prep.) This means no more fooling around with FEdit to see the prep file. [Interestingly enough, you can download the cmd-F postscript file with a postscript dumper program and it will run correctly, calling the LaserPrep file if necessary. But if you download the cmd-K file, thinking it contains both the LaserPrep header and your file, the LaserWriter hangs up! Anybody know why? -Ed]
Final Revision of MacDrive Software
Beaker
I can confirm that (as published in MacTutor) there WILL be a version 2.3 of the Tecmar MacDrive software. I know because I have it. I'll be receiving the second beta version of it this week.
As reported by me earlier, it does "fix" the Imagewriter problem. What that means, unfortunately, is that now it doesn't hang when the Imagewriter tries to talk back, it only absorbs the data. So you STILL can't do color printing, etc., but you CAN use the Imagewriter 2.3 (which is faster).
The driver software DOES support larger drives (up to the capacity of the ZEBEC controller). Unfortunately, it is still not known whether the Volume Manager will be fortified to handle the larger drives. If not, then forget it, since the Vmgr is the only way to create the necessary partitions. I'm urging them on HARD. (Dave told me that if you replace the controller, you could do something disgusting, like shove a Maxtor 1140 or even a new 380MB Maxtor in there - YUM!)
Tecmar will be distributing a pretty neat backup program called FullBack with the new software. You'll like it. It was written for their large drive, which may or may not ever see the light of day, but that's a whole other story.
The final (bad) news is that this won't be a free upgrade. It'll cost, but I'm not sure how much (it won't be bad though). I'll let you know on the saga of larger drives.
HyperDrive FX-20
The Anarchist
After using the new Hyperdrive FX-20 for a couple of weeks, I can tell all of you that I am very pleased with it. If the case design doesn't offend you (about half my friends like it and the other half can't decide) then this is one heck of a drive.
It comes with the typical HyperDrive 'Security', 'Backup/Restore' (which is now HFS-Rated) and a *NEW* 'LaserSpooler' DA. It retails for $1195.00, and can be had for about $1000.00 at your local store. It is about the same price as many of the other 20 meg drives, but I would suggest this one since GCC is a manufacturer that will be around to support you when the others are gone.
Also, the new HyperDrive 10 and 20 software release is out, and it is pretty nice. It is version v3r1, and has the LaserSpooler DA too! If you have an internal HyperDrive in your Mac or Mac+ drop by your dealer with three fresh disks and get this worthwhile free software upgrade.
Protect Bit
Chief Wizard
The protect bit works differently on MFS and HFS. To change it with MFS, you have to read the catalog directly and adjust the flag byte manually and write it back out. In HFS, you have to do something similar, but I think they've MOVED the protect bit. You should play with FEdit and find out where the protect bit is now located. There is no way to clear the protect bit with any SetFileInfo type calls. The resource file name is NOT kept around anywhere. The only thing I could recommend would be to scan the files (you can find out what volume it's in) and compare the location of each file to the locations associated with the path of the file you want.
[HD Util allows you to change the protect bit, unlike ResEdit and Fedit, which only let you see the state of the protect bit. - Laser Dolphin]
Paper Jamming
Jim Reekes
If you have an ImageWriter // and the paper jams, it's probably getting stuck under the paper out sensor. Have your favorite dealer order a BUNCH of "paper out sensor frames" from Apple. They'll most likely be on back order, so you'll have to wait, but they only cost $10-15. This bugger has caused just about every ImageWriter // owner to kick and spit while trying to get a single page to print.
Button Help
The Atom
Here's the new question: I have a dialog (made with rmaker) that loads up with GetNewDialog. I've got my OK and Cancel buttons as the first two items. When you hit return, it does the OK button just likes it's supposed to. The question is, how do I tell it to put a double line border around the default button (the OK)?
The only way I know how to get a button to have a double border is to draw the outer (heavy) border with DrawRoundRect; it is not part of a button definition. (Boy, if I'm wrong, I'm gonna get a lot of flak about this one!)
- Mike Steiner
There are a couple of issues involved here. (And you thought it was a simple cosmetic matter. Oh well, that's the Mac). First, if you are SURE you're never going to get any update events, you can just draw the roundRect manually by setting the port to the dialog window and doing it. But you should first do a DrawDialog so it won't get wiped out when the dialog manager draws the dialog later.
If you ARE going to have to deal with normal update events (and solve for the general case anyway), you have to make the thick line a user item. BUT, make sure to place it in the item AFTER the okay button itself. Otherwise, when the dialog manager scans the item list to find out where a mousedown occurred, it will hit the disabled user item before finding the okay button and it will stop.
To make a user item, do your GetNewDialog, then do GetDItem and SetDItem. For the itemHandle parameter, pass a pointer to a procedure to draw the outline. The parameters for a user item procedure are listed in IM.
- Chief Wizard
Of course if the default button is the "OK" button no border is necessary. On the other hand if you want a border anyway, and your dialog could be an alert instead, you will automatically get either the first "OK" or second "Cancel" outlined depending on your alert template.
- Don L
Here's a routine I use to outline my OK buttons in Modal dialogs. Just pass in the DialogPtr, item number and corner radius for the outline (16 is std.) and you've got it. As far as the updates go, you could call this from a filter routine everytime the filter gets an update event. This way you'll be sure to always have the outline around.
- Rick Boarman
PROCEDURE OutlineButton(theDialog: DialogPtr; itemNo,
cornerRad: integer);
{ Draws a border around a button. 16 is the normal
cornerRad for small buttons }
VAR
itemType: Integer;
itemBox: Rect;
itemHdl: Handle;
tempPort: GrafPtr;
BEGIN
GetPort(tempPort);
SetPort(theDialog);
GetDItem(theDialog, itemNo, itemType, itemHdl, itemBox);
PenSize(3, 3);
InSetRect(itemBox, -4, -4);
FrameRoundRect(itemBox, cornerRad, cornerRad);
PenSize(1, 1);
SetPort(tempPort);
END;
Lightspeed...
Gary Voth
"You've come a long way baby..." That's the first thing I could think of to say after getting a look at Think Technologies' LS Pascal at Dave Kelly's the other night. The source-level tracing, breakpointing, immediate statement compilation and variable observation make this product damn near as easy to work in as an interpreter. The pretty-printing editor is great too! I've long been spoiled by some of the features of MS-BASIC and Macintosh Pascal editors, such as automatic indentation, keyword boldfacing, and so on. Now you can get it in Lightspeed. Still doesn't work with Mac Plus cursor keys, though.
LS Pascal Problem
Yes like Gary said.... it's great. So far I've been really having fun with it. One problem I noticed that maybe you experienced Pascal people can figure out. If not maybe there isn't a fix?? I tried running some of the very first MacTutor Pascal stuff (from Vol. 1) Well, Alan Wooton's 'READING PAINT FILES' program won't compile because Lightspeed won't let you have variables the size of a paint file. The manual states that the largest size a variable can be is 32767 bytes because of something to do with the way the 68000 does offsets (whatever that means). (Alan set up an array to hold the bit map for the screen that is about 55K.) OK, Alan, how about a way to do it? Or should we all complain to Think Technologies? Except for this I see no problems with Lightspeed Pascal.
-Dave Kelly
Lightspeed C is undergoing a revision and should be out in November as version 1.5. It doesn't have the Pascal source code debugger, but a lot of the nice HFS features of the Pascal product have been added to the C product, making it more user friendly.
-David Smith
MPW
Micro Ghoul
I recently ordered MPW (at horrid prices, I might add), and am now wondering (as I anxiously await its arrival) what other people think of it.
I have been working on both the Xerox 1186 using Common Loops (we are a Beta Site) and Sun 3 with Objective C which, of course, have very different approachs and implementations of the OOP paradigm. (Side note: if you're interested in seeing Objective C for the Mac, call PPI: they are very interested in your thoughts!) I have read the Hayden book Object Oriented Programming for the Macintosh, which I thought was interesting though often a bit wordy (I am always in a rush when trying to get info, so just give it to me straight!), but I did not get a strong feel for the programming environment itself. Does anyone out there have enough experience with MPW to comment on it? I purchased all of the parts (for kicks, I suppose, as that gives me 3 C's and 3 Pascal's and 2 assemblers!), but I did it, basically on the thought that the potential for the product was high enough to chance it. Though I did not like being told that I was not entitled to a discount on upgrades or actual purchase!
MPW Feedback
- Rick Boarman
To start with, MPW is not for the light of heart, casual or easily frustrated programmer. [That let's me out! -Ed] Using it is easy, but mastering it can take many months. I've been using it steadily for six months now and still learn new commands and ways of doing things. MPW consists of the MPW Shell (Editor) and numerous Tools. Each tool (command) is activated by entering its name into any window in the Shell and hitting the Enter key. Typical tools are the Compilers, cross referencers, Linkers, editing commands, file IO, folder and volume commands, and custom commands of your own.
A typical command to compile, link, resource compile and execute a Pascal program might look like this: Pascal Sample.p Link Sample.p.o {OBJLibraries} -o Sample.code Rez Types.r Sample.r -o Sample Sample. [You mean I gotta learn to type again? -Ed] If this whole mess is selected and executed, Sample.p will be compiled and linked with the object files specified in the variable {OBJLibraries}. The output of the link will be saved in Sample.code. Then the resource compiler (Rez) will do its job and put the output in Sample. Sample is now an executable Mac application. The last line launches Sample and actually leaves the Shell completely. When !_ExitToShell is executed from Sample, the Shell starts up again right where you left off. All windows are in the same place, selections are the same.....
There is no debugging per se. MacsBug and TMON are still the best for that. One real slick command allows custom menu commands and menus to be added to the Shell. For instance, I have a Menu command to read all .p files in a folder, put each file name in a menu as an item. When the file name is selected from the menu, that file is opened. Real slick, since I have three folders of source code, with about fifteen files in each.
MPW provides 100% toolbox compatibility. All calls are implemented. The version I have (1.0B2) is still fairly buggy and likes to bomb at times. Compile times are fair. About 10% faster than Lisa Workshop, but still very slow compared to LightSpeed. One of my programs has nine UNITS with about 75,000 lines of code. It takes 32 minutes to compile compared with 45+ on the Lisa. The learning curve is several hours. But to master what are called 'Make Files', which handle automatic application-building based on modification dates, takes many hours of frustrating work. At least this was in my case, as I have many folders and files to juggle and NO examples of how this is supposed done. The latest manual has examples of this - too late for me though. [Watch for a special MPW and MacApp issue of MacTutor in the very near future. -Ed.]