shulman@topaz.UUCP (05/09/86)
Delphi Mac Digest Thursday, 8 May 1986 Volume 2 : Issue 18
Today's Topics:
PCPC's HFS Backup
RE: PCPC's HFS Backup (Re: Msg 7688)
RE: Need help with JCLOCK utility
Mac SCSI connection problem
MacPaint Pasting Problem
good disk prices
LaserWriter and VMS
Re: Apology to Micah
ROM Resources
Bug in LightspeedC
RE: Warning about HP disks (Re: Msg 7731)
To John Bass
cache setting
RE: cache setting (Re: Msg 7775)
RE: cache setting (Re: Msg 7779)
Microsoft Updating Word...
Re: Changing the default font
Mac+ games
HFS Backup bugs
MacPCBridge
Desperately Seeking Software
LoadSeg problem
Printing Errors
Setting file protect bit in ResEdit
RE: Setting file protect bit in ResEdit (Re: Msg 7883)
RE: Setting file protect bit in ResEdit (Re: Msg 7893)
----------------------------------------------------------------------
From: RICFORD (7688)
Subject: PCPC's HFS Backup
Date: 1-MAY-16:18: Business Mac
We have spent some more time with HFS Backup ($40) from PCPC
(1-800-MAC-BUTT) and we're finding it very useful. Some of the key elements:
- ability to do full, incremental, or selected-file-and-folder backups on HFS
and MFS volumes
- ability to back up files larger than a floppy disk
- ability to save configurations of files and folders to backup ("Backup sets")
- automatically chooses 400K/800K for Mac-formatted disks (otherwise
assumes 800K on a double-sided drive)
- Macish, friendly, flexible, and quick
- ability to scan and print disk directories (can't print directories without
going thru backup/restore procedure)
- no selection by file _type_; otherwise, flexible selection options
- requires 128K ROMs (haven't tried it with RAM-based HFS)
- does only one volume at a time
- flexible options for checking disk integrity during backup
- backs up to floppies only
Ric Ford
"MacInTouch" newsletter
------------------------------
From: JOHNCRANE (7696)
Subject: RE: PCPC's HFS Backup (Re: Msg 7688)
Date: 1-MAY-20:34: Business Mac
Ric, I've also been checking out HFS Backup, and would like to footnote your
observations. While it's true you go through the backup procedure, or at least
part of it, to get a catalog, you can then abort before actually carrying out
the process. It worked fine with 64K ROM, and is now working on 128 ROM on the
same machine. I backed up about 10MB on 14 400K floppies; that's about a 60
percent condensation factor. It took about an hour. I'm sure with the new
drive and 800K floppies it would be much faster. It is _very_ picky when
testing floppies; I had a large number rejected that are otherwise giving no
problems - and of course this adds to backup time. Unless anyone gets the wrong
impression, the program does not modify the backup floppies when one does an
update of the backup but rather adds the changes onto additional floppies.
Of course, the real test would be to see how well it restores.
- John Crane "MacNews"
------------------------------
From: PEABO (7697)
Subject: RE: Need help with JCLOCK utility
Date: 1-MAY-21:12: MUGS Online
I think JCLOCK installs itself into your System File as an INIT resource that
has the title JCLOCK. It should be easy to find and eradicate using ResEdit!
peter
------------------------------
From: SCHAFE (7698)
Subject: Mac SCSI connection problem
Date: 1-MAY-21:43: Hardware & Peripherals
Help! I have read somewhere that the Macintosh Plus's SCSI port is
externally convertable to the 50 pin differential specified in the
ANSI definition of SCSI. Is this true, and if so, how is this
accomplished? I thought the task was as simple as matching up the
signals from each side and ignoring all other lines, but I doubt this
is the correct method. Any replies greatly appreciated.
Thanx in advance.
------------------------------
From: DSACHS (7703)
Subject: MacPaint Pasting Problem
Date: 1-MAY-22:36: Creative Pursuits
About the problem with cutting and painting with MacPaint reported in the last
info-Mac. Is it possible that the SYSTEM disk is either locked or low on space
- The latest software forces the clipboard to reside on the system disk.
------------------------------
From: RICFORD (7718)
Subject: good disk prices
Date: 2-MAY-11:55: Mousing Around
Now that we're double-sided, it's time for new disks. We found
double-sided Sonys, packaged 10 to a box, for $2.50 and under at
ComputerWare in Palo Alto, CA. They come with labels and everything.
There are further price breaks for larger quantities and bulk Sonys
without the labels. 2nd day shipping to Boston cost us $6.00 for 100
disks. Although our experience is limited, this seems like a good
store.
800-235-1155 (800-323-1133 in CA)
Ric
------------------------------
From: JIMH (7735)
Subject: LaserWriter and VMS
Date: 3-MAY-15:19: Programming
Is anyone out there using the Apple Laserwriter on a VAX under the VMS
operating system? Adobe sells a product to use it under UNIX but not
under VMS, does anyone sell an equivelent product for VMS? I can buya
laserwritted for my mac at work if i can share it with the VAX, so i
am kind of desperate for such a product. Thanks Jim Hopper
------------------------------
From: BRECHER (7737)
Subject: Re: Apology to Micah
Date: 3-MAY-16:05: MUGS Online
> I don't know if this means that Micah's hostadapter just works
> with a mac plus or that their hostadapter provides a 100% hardware
> and software compatable external mac plus scsi port on a 512k system like our
> MacSCSI Plus hostadapter (anybody who has a Micah drive know?).
I have a MicahDrive. MicahDrive is compatible with 512K Macs (plain or
enhanced), and Mac Plus. It does not provide an external SCSI port on a 512K.
> [Brecher's] goal is to build THE BENCHMARK that clearly states WHO IS BEST
> on HIS definition of a STANDARD WORKLOAD.
I do think a benchmark based on a "typical" workload approach would be
of value for many prospective hard disk buyers. But "THE BENCHMARK,"
"clearly states WHO IS BEST" and "HIS definition" cannot have been
derived from anything I've said.
With respect to a Delphi message in which I said, "If speed is a criterion,
forget Warp 9; it uses basically the highly-inefficient MacSCSI driver software
(it also requires a boot floppy)," Bass comments:
> There was NO need to single us out ...
My message was a response to a query by someone who solicited comments on hard
disks specifically including the Warp 20, and who stated that speed was a
criterion.
> Mr Brecher's coments about "the highly inefficient MacSCSI driver" are both
> WRONG and OUT OF LINE.
Following is what the Aztec C (1.06g) compiler generates for the data
transfer (read) routine in the DDJ article. I have confirmed with
someone who owns a copy of the recent release of the MacSCSI driver
that this routine has not changed significantly (if at all) since the
DDJ version. Since only in the past few weeks has there been talk
about Warp/Mirror converting the driver to assembly language, it is
reasonable to assume that this code is what was running in the
Warp/Mirror products I referred to in the message Bass cites. The
numbers in brackets are 68000 clock periods required for the
instruction:
SCSIIn:
Link A6,#-4 ;[16]
MoveM.L D4-D7/A3-A4,-(SP) ;[56]
Move.L $0A(A6),A3 ;[16] A3 = ptr (buffer addr)
Move.L #$10,D4 ;[12] constant
Move.L #1,D5 ;[12] constant
Move.L #0,D7 ;[12] constant
Move.L #$500000,-4(A6) ;[24] NCR base addr
Move.L -4(A6),D6 ;[16]
Add.L #$11,D6 ;[16] D6 --> Current SCSI Data register
Move.L -4(A6),A4 ;[16]
Add.L #2,A4 ;[16] A4 --> Initiator Command register
Move.L -4(A6),A2 ;[16]
Move.B $0F(A6),4(A2) ;[20] Target Command register := phase
; -----
; [248] for setup
; Wait for REQ
;
@0 Move.L -4(A6),A2 ;[16] A2 = NCR base addr
Move.B $19(A2),D3 ;[12] D3 = (Current Bus Status register)
Ext D3 ;[4]
And #$20,D3 ;[8]
Cmp D7,D3 ;[4]
Bne @1 ;[10 -- assume branch]
Bra @0 ;[0]
;
; bail out if not phase match...
;
@1 Move.L -4(A6),A2 ;[16] A2 = NCR base addr
Move.B $1B(A2),D3 ;[12] D3 = (Bus and Status register)
Ext D3 ;[4]
And #8,D3 ;[8]
Cmp D7,D3 ;[4]
Beq @2 ;[12 -- assume no branch]
; ----
; [108] each time through loop, total = [6912]
; transfer 8 bytes...
;
.Repeat 8 ;assume this assembler directive to save listing space
Move.L D6,A2 ;[4] A2 --> Current SCSI Data
Move.B (A2),(A3) ;[12] move data byte to buffer
Add.L D5,A3 ;[8] increment buffer addr
Move.B D4,(A4) ;[8] assert ACK
Move.B D7,(A4) ;[8] deassert ACK
.EndRepeat
; ----
; [320] each time through loop, total = [20480]
; Loop if count not exhausted
;
Sub #8,8(A6) ;[20]
Cmp #0,8(A6) ;[16]
Bgt @0 ;[10 -- assume branch]
; -----
; [46] each time throuh loop, total = [2944]
;
@2 Move.L #0,D0 ;[12] return 0 to caller
MoveM.L (SP)+,D4-D7/A3-A4 ;[60]
Unlk A6 ;[12]
Rts ;[16]
; -----
; [40] for termination
; =======
; Grand total [30624] clock periods for 512 bytes transferred
Following is my hand-coded version. It is functionally identical
except that it further unravels the loop:
SCSIIn:
MoveM.L D1-D4/A3-A4,-(SP) ;[56]
Move.L $0A(A6),A3 ;[16] A3 = ptr (buffer addr)
Move.L #$500011,A2 ;[12] A2 --> Current SCSI data register
Lea 2-$11(A2),A4 ;[8] A4 --> Initiator Command register
Lea $19-$11(A2),A1 ;[8] A1 --> Current SCSI Bus Status register
Lea $1B-$11(A2),A0 ;[8] A0 --> Bus and Status register
Move.B $0F(A6),4-$11(A2) ;[20] Target Command Register := phase
MoveQ #0,D0 ;[4] constant (and returned value)
MoveQ #$10,D1 ;[4] constant
Move 8(A6),D2 ;[16] D2 = len
Lsr #5,D2 ;[16] D2 = len/32
SubQ #1,D2 ;[4] adjust for Dbra
MoveQ #5,D3 ;[4] constant
MoveQ #3,D4 ;[4] constant
; ----
; [180] for setup
;
; Wait for REQ
;
@0 Btst D3,(A1) ;[8]
Beq.S @0 ;[8 -- assume no branch]
;
; bail out if not phase match...
;
Btst D4,(A0) ;[8]
Beq.S @1 ;[8 -- assume no branch]
; ----
; [32] each time through loop, total = [512]
;
; transfer 32 bytes...
;
.Repeat 32 ;assume this assembler directive to save listing space
Move.B (A2),(A3)+ ;[12] move data byte to buffer
Move.B D1,(A4) ;[8] assert ACK
Move.B D0,(A4) ;[8] deassert ACK
.EndRepeat
; ----
; [896] each time through loop, total = [14,336]
;
; Loop if count not exhausted
;
Dbra D2,@0 ;[10 -- assume branch]
; ----
; [10] each time through loop, total = [160]
;
@1 MoveM.L (SP)+,D1-D4/A3-A4 ;[60]
Rts ;[16]
; -----
; [76] for termination
; =======
; Grand total [15264] clock periods for 512 bytes transferred
The compiled code occupies 110 words; my assembler version occupies 128 words (
due to the further loop unravelling). If the entire driver were coded in
assembly, the total net requirement would be much reduced from that of the
compiled version.
I think it reasonable to call a driver that takes twice as long as it
should to transfer data "highly-inefficient." The inefficiency of its
data transfer code, similar inefficiencies elsewhere in the driver,
and the lack of provision for pseudo-DMA in the MacSCSI design combine
to necessitate two disk revolutions to transfer each sector as
indicated by the DiskBench results. Those results cannot be explained
by reference to 11:1 interleave.
> We showed Apple one of our boards early last summer and they
> quickly added the NCR5380 onto the main logic board and offered it
> as the MacPlus (clearly defining SCSI as the standard interface).
Apple had prototype designs with SCSI interfaces in 1984 and early
1985. Perhaps Bass's use of the NCR 5380 influenced the final design
as Bass claims; fortunately, Apple's design provides for pseudo-DMA.
> [T]he Mac filesystem/OS or finder crawls into a hole with filesystem that
> have 10mb or more of allocated files -- above 18mb bootup to the finder desk
> top takes more than a minute on a system that took 7sec at 3mb.
My MicahDrive, with 525 files totalling 8.5MB, takes 12 seconds from reset
"bong" to a quiescent Finder desktop showing the startup volume's window
containing 16 folders. With 18.5MB, 1078 files, 17 folders it also takes 12
seconds. (512K, 128K ROM, HFS volume, System 3.2b3, Finder 5.2)
------------------------------
From: DSACHS (7729)
Subject: ROM Resources
Date: 2-MAY-21:02: Programming
Does anyone know which resources can be safely deleted from a system file that
is to be used ONLY with a Mac plus. I have the following partial list of the
resource listed in TN 57
Safe to delete: FONT 12, PTCHs for other ROMs, MDEF 0, WDEF 0
Cannot be deleted: PACK 7, CDEF 0, 1, CURS 1, 2, 3, DRVR 2
------------------------------
From: PEABO (7742)
Subject: Bug in LightspeedC
Date: 3-MAY-20:49: Programming
Two bugs found in LightspeedC:
ctype.h: treats iscsym() and iscsymf() the same. iscsymf() should not include
_digit_ in its definition.
ctype.c: treats the value 128 as ASCII. This is wrong, since the high order
bit is set. The documentation on page 13-32 of the manual is also wrong.
peter
------------------------------
From: RSTICKLE (7754)
Subject: RE: Warning about HP disks (Re: Msg 7731)
Date: 4-MAY-12:19: Hardware & Peripherals
Easy fix folks, Take your HP disk and turn it to the back side, you'll see a
small plastic catch that fits in the sliding door cut-out. Now take a knife,
something like an X-acto knife and cut the catch off so it's even with the rest
of the disk, and PRESTO, the sliding door can't catch and no more disk drive
jams!! <grin> Rick
------------------------------
From: RDETCHELLS (7772)
Subject: To John Bass
Date: 4-MAY-23:00: MUGS Online
This is directed to John Bass (DMS Designs), and regards the "new,
improved" driver software I just got from Mirror Technology for the
MagNet 20 (Which is, I think, identical to the Warp20). I don't know
whether he had anything to do with the coding of this release, but if
he did, I'm curious as to why things were done the way they were...
The disk doees now indeed launch applications MUCH faster (about a
factor of two in launch times). HOWEVER, it also now has the most
incredibly extravagant disk allocation granularity that I've ever
heard of: The smallest file I can find on the disk occupies 35K bytes
(a rare occurrence), and most occupy at least 69K!~ I could understand
a desire to reduce the size of the FAT (file allocation table, or
whatever it's called on the Mac), to cut search times, but this is
ridiculous! A disk contaiining 362K of MacWrite files occupied nearly
3 MEG when transferred to the MagNet20! At that rate, the MagNet
could only hold the equivalent of 7 or 8 floppies! Is this the way
the new software release is actually supposed to work, or is it a bug?
(I have version 1.6 of the driver software from Mirror Tech.) Thanks
for whatever info you might be able to pass on. BTW, kudos to you for
your gentlemanly apology to Micah -- it was very well taken. Thanks,
David Etchells, Perspect Systems (213) 821-7884
------------------------------
From: PIZZAMAN (7775)
Subject: cache setting
Date: 4-MAY-23:20: Hardware & Peripherals
I'm about to upgrade my 512 with the new ROM and 800 drive. I have
system 3.1 and wonder whether the cache setting is anything I will
need to play with. Is this something I will use with my 512, or is it
just for the Mac+ with the extra memory? barry
------------------------------
From: BMUG (7779)
Subject: RE: cache setting (Re: Msg 7775)
Date: 5-MAY-02:34: Hardware & Peripherals
It will work on a 512 though not as well. It loads certain things into
the cache first then goes on to programs and such later. So it doesn't
seem to make much differnce useing it. Your programs will run faster
if they use alot of fonts ect. Steve Costa/BMUG
------------------------------
From: RICFORD (7790)
Subject: RE: cache setting (Re: Msg 7779)
Date: 5-MAY-10:40: Hardware & Peripherals
I had the same experience on a 512K Mac. I kept forgetting whether the cache
was on or off, because the speed gave no clues.
Ric
------------------------------
From: MACINTOUCH (7793)
Subject: Microsoft Updating Word...
Date: 5-MAY-12:35: Business Mac
Infoworld, 5 May '86:
Microsoft Updating Word for the Macintosh
Microsoft Corp. is reportedly working on a new release of the Macintosh version
of Word that will make the system similar to Word, version 3 for the PC,
according to independant sources.
Initial test copies of the new Macintosh version are now being completed, and
beta test copies are expected to be delivered this summer, according to a close
source at the firm. "It's safe to assume that Microsoft could deliver a new
release of Word sometime late this fall or early 1987," the source said.
Microsoft is considering adding greater capabilities in the word
processor, such as spell check, outline processing, sytle sheets,
desktop publishing capability, and automatic index and table of
contents generation, according to the source.
josh - MacInTouch
------------------------------
From: DSACHS (7838)
Subject: Re: Changing the default font
Date: 6-MAY-19:33: MUGS Online
It is very easy to change the defaut (system font if you have a
current version of the Resource Editor. Simply move your replacement
font onto the system file with an ID OTHER THAN 12. Then edit the
FOND 0 resource to point to the new font. This of course works with a
Mac plus only.
------------------------------
From: LOGICHACK (7821)
Subject: Mac+ games
Date: 6-MAY-00:51: Bugs & Features
This is for the people complaining about compatibility problems. I have been
able to get both Megaroids and MazeWars to work on a Mac+ by using Fedit to
change the code that refers to the hardware buffers. The primary screen is at
$7a700, that should be changed to $3fa700 to be compatible with up to 4MB of
memory (the max). The other strings to change are: $72700 to $3f2700 for the
secondary screen, and $7fd00 to $1ffd00 for the primary sound buffer. I don't
think too many people use the second sound buffer...
Other advice: don't rename megaroids since it hard codes the name to do a re-
launch to use the second screen. Also, for software that use the alternate
buffers, be sure to turn your cache setting OFF or the system will bomb sooner
or later. It seems that Apple doen't acknowledge you buffer requests in the
launch block in respect to the cache memory usage.
Hope this helps,
Paul
------------------------------
From: RICFORD (7832)
Subject: HFS Backup bugs
Date: 6-MAY-14:03: Bugs & Features
We have discovered some bugs in PCPC's HFS Backup program which
prevent restoring some backed up files. This is in version 1.0b.
PCPC is shipping an updated version which corrects some of the bugs.
We'll have more details later.
Ric Ford
"MacInTouch"
------------------------------
From: MUSICWORKS (7875)
Subject: MacPCBridge
Date: 7-MAY-22:20: Hardware & Peripherals
I just installed a MacPCBridge on one of the IBM XT on a client's
office network with appletalk/Mac/Laserwriter... Installation is a
snap just plug in a half slot card and plug in appletalk cable and the
software is excellent ! just run the program and you can choose what
kind of document (i.e. wordstar/multimate etc) and it will print it
out to the laserwriter, although (most IBM wordprocessor) doesn't let
you see what kind of fonts you're about to print but by specifying
some control commands you can change fonts,size,style etc... It takes
some memory about 30K from your main memory ( big deal... ) with the 2
programs you have to exe cute first before running the laserScript
Program ( the printing and mail box program)... those two program will
search for the available appletalk devices (like laserwriter/plus) so
far it's been working like a charm !
Jundi
------------------------------
From: BKV (7899)
Subject: Desperately Seeking Software
Date: 8-MAY-18:05: Business Mac
Am desperately seeking software that runs on a Mac for use in a
video/audio retail business. It must be able to handle customers by name AND
number, preferably be multi-user, bar-code compatible, password protection
compatible, be able to generate multiple types of reports including mailing
labels; customer updates; catalog updates and modifications; category,
customer, format, etc. code updates; batch checkins; customer activity; title
activity (slow titles, top titles, titles listed by format, price, etc.);
system codes; weekly, monthly & yearly summary reports and a few more. It
must also be able to send to an on-line printer a sales receipt for each sales
and rental.
I would like to use Omnis 3 or Double Helix since they offer
password protection, but will consider others if they meet the above criteria.
Any leads or info will be greatly appreciated as the store will be in
operationshortly after June 1st.
Brad Vautrinot
------------------------------
From: JEFFS (7869)
Subject: LoadSeg problem
Date: 7-MAY-19:51: Bugs & Features
I finally tracked down the bug that had been bothering me for some time now. I
have an application that would go off to Mac heaven whenever I opened a certain
file. It would only go off the deep end whenever I would select a particular
menu item. If I selected this menu item FIRST and then opened the file, all
worked! (Very frustrating!)
As I investigated further I discovered that it was hung in in the
LoadSeg trap (the code for the menu item was in a different segment.)
Now, why did this happen?? Well, it just so happens that the file I
opened that caused the crash was another application. What was
happening was the LoadSeg was looking for the CODE resources in THAT
application and not mine!
The moral of the story is that if your application opens other application
resource files you should put something like the following before you call
functions residing in another code segment:
curRes = CurResFile();
UseResFile(myResNum);
loadFooSeg();
UseResFile(curRes);
Where loadFooSeg can be an empty function in the segment you want to load.
Jeff
------------------------------
From: JEFFS (7870)
Subject: Printing Errors
Date: 7-MAY-20:07: Programming
I want my application to be "friendly" about reporting printing errors
like "disk full" or "disk locked". By this I mean it should tell the
user the name of the disk that is full or locked. This seemed easy
enough to do since the Printing Manager says it uses the default
volume for the spool file if the iFileVol field of the TPrJob record
is 0 (the default.) You should then be able to call CurVol to get the
name of the volume that caused the error.
To test this out I locked my boot disk and tried to print from my program. The
error message said my OTHER disk (where the program resided) was the disk that
was locked! What gives?? (When I lock that other disk and not the boot disk
all proceeds OK indicating that the boot disk *is* the disk it is trying to
spool from.)
What do I do now? Does iFileVol = 0 *really* mean the System volume
and not the default volume?
Jeff
------------------------------
From: JIMWEINRICH (7883)
Subject: Setting file protect bit in ResEdit
Date: 8-MAY-06:27: Developer's Corner
How can I set the Protect bit using ResEdit? I can set most of the
file bits in the Get Info box, and the File Protect bit is displayed
there, but when I click in the check box it doesn't toggle. Is there
some special option-key combo I have to press? Some magic certified
developer's password that I must chant in order for Apple to let me
mess with things I shouldn't touch? SetFile, etc., don't work
properly under HFS, and that's why I need to use ResEdit. Thanks!!
--JimWeinrich
------------------------------
From: RICFORD (7893)
Subject: RE: Setting file protect bit in ResEdit (Re: Msg 7883)
Date: 8-MAY-13:33: Developer's Corner
HFS directories don't support the same file info bits. I'm pretty
sure the file protect bit is gone, making it impossible to use old
techniques for defeating copy protection hassles on programs such as
MS Word. The zap programs such as HDUtil and MacZap handle it though.
Ric
------------------------------
From: OPPENHEIM (7910)
Subject: RE: Setting file protect bit in ResEdit (Re: Msg 7893)
Date: 8-MAY-19:48: Developer's Corner
Seem to support them for me -- no disk editor I have seen gives access
to them w/o dire straits, but set the byte two before TYPECREA and the
byte immediately after to $C1. Than protects, locks and renders
invisible the file. Set the first of those to $C9 for bozo, if
necessary. Fedit 3.52 doesn't allow looking at the volume directory,
but you can find the file with an ascii search for the name or
type/creator string. If you're doing this on the Plus with floppies
eject the volume BEFORE quitting to the Finder. Otherwise, for some
unexplained reason, the changes are not recorded. Now I have one --
does anyone get 'Quit' in the 'Special' menu (Finder)? No Apple docs,
of course...
------------------------------
End of Delphi Mac Digest
************************