THIS THREAD PERTAINS TO THE TEXT GRABBER AND FONTS IN GENERAL!

 ************
Topic 30        Sat Jan 02, 1988
WC.COLEMAN [Geos*Sysop]      at 03:58 EST
Sub: Bug in new Text Grabber located!       

I located the bug in the new Text Grabber that prevented it from loading in
the forms. See message #1 this topic for a discription/fix.
15 message(s) total.
 ************
 ------------
Category 19,  Topic 30
Message 1         Sat Jan 02, 1988
WC.COLEMAN [Geos*Sysop]      at 04:09 EST
 
There is a bug in the new Text Grabber that prevented it from loading in the
forms. The problem is simple -- all the forms are stored as Geos filetype #3
(data). Once you select the form it is loaded with GetFile. Unfortunatly this
routine will not load type 3's, it will simply ignore them! The solution is
simple - all the forms must be changed to type #7 (application data) and then
a byte must be change in the Text Grabber program so that it will search the
disk for 7's instead of 3's. Here's how:

Use a disk doctor to change the forms to type #7. Offset #22 in the directory
entry is the Geos Filetype. Change it from a three to a seven. Now chain to
the header block and change offset 69 the same way.

Now comes the hard part. Chain to the first sector in the Text grabber. The
program is VLIR (it only has one record but the paranoid SOB's are afraid of
people uploading their programs). Just chain to record #1 (the first two bytes
past the link are the T&S of the program. Now count to the 24th sector. The
index sector doesn't count. Go to offset #121. This should be a three. Change
it to a seven.

That's it! The program will now work properly. If anyone has any problems or
comment leave messages in this topic. I will start off the questions --  Why
doesn't BSW test their frigging programs before they sell them to us poor
suckers????????????
 ------------
Category 19,  Topic 30
Message 2         Sat Jan 02, 1988
P.SPEARING                   at 09:02 EST
 
Why doesn't BSworks test their programs before selling them? Beats me, but I
suspect it's that they have a videogame mentality-- they claim they're    ve
written an operating system, but they act as though it was the latest space
shoot-em-up (HEAVY copy protection, bugs, inaccurate documentation, poor or
absent support).  The   GEOS is    makes a whole new machine out   t of the  
64 or 128, but much of the power is lost because of these stupid  things.  I
truly hope they g                   They treat us almost a well as the guys
who built our comop  puters do.   Great machines, but try getting support from
.
 ------------
Category 19,  Topic 30
Message 3         Sun Jan 03, 1988
P.ROSSING                    at 10:31 CST
 
Thanks for the the fix, but what if we don't play around with sectors and
tricky things like that?  any  way that we can get a fixed one uploaded
online?  or would that be copyright violation.  hardly seems fair for those of
us who paid  for the programs to begin with.. BSW should allow fixes to be
uploaded.  especially things like "text grabber" that don't have much use
unless you've bought GEOSwrite. peg

 ------------
Category 19,  Topic 30
Message 4         Sun Jan 03, 1988
WC.COLEMAN [Geos*Sysop]      at 23:35 EST
 
I could write something but it will take a while till I get to it. Don't you
have any friends who know how to use a Disk Doctor. Hell, you only have  to
change a few bytes. I'll try to get something written but I'm in the middle of
a grunch of projects at the moment (geos of course)!
 ------------
Category 19,  Topic 30
Message 5         Mon Jan 04, 1988
P.ROSSING                    at 21:07 CST
 
I'm sorry but I live up in the north woods of new hampshire where i bet it's
150 miles to the next commodore.. that's what online services like GENIE are
for.   I have a disk doctor, it so happens, and i tried to use it, but i
couldnt for the life of me figure it out.  I found the "3s" in the directory,
but where re the headers?  then what happens next?  
   sorry to pester you, but not all of us are techies, or even know any.
 ------------
Category 19,  Topic 30
Message 6         Mon Jan 04, 1988
WC.COLEMAN [Geos*Sysop]      at 22:35 EST
 
Ok i'll try to make this as painless as possible! Here it goes:

First you must understand 'offset' in whole sectors it means the position in
the sector (the first byte is 0 *not* 1). All the offsets will start at 0. In
a Geos directory   entry there are 2 bytes after the name (just past the
shifted spaces if the name is less then 16 chars). These are the track and
sector of the header block. Just put the cusor on the 1st byte (which is the
trk) and Jump to the sector (in most Disk Doctor's you will use the 'J' key).
Now go to position 69 (decimal) in the header. This will be the Geos filetype.
In the forms this will be a chr$(3). just change the byte to a chr$(7) and
save the sector back to the disk.

Now back to the directory entry: The next byte past the header T&S is the
structure 0=Geos Seq and 1=VLIR. Leave this alone! The very next byte is the
Geos Filetype. Again change the 3 to a 7 and save the sector.

Just do this to all (I think) 7 forms. Now find the dir entry for the text
grabber. The 2 bytes before the 'T' in Text will be the TRK and SEC to the
index sector (normally these would point to the 1st block of the program but
this is a VLIR file). When you get to this sector you will see four bytes at
the beginning and the rest will be zero. The first two are 0,255. The next two
are the TRK and SEC you want. Jump to it. This is sector #1. Follow the links
(usually the 'L' command ---or put the cursor at the beginning of the sector
and use 'J'). count each sector until you get to 24.
 By the way on my disk this was track 8 sector 3, you might want to jump
straight there -- you might get lucky! Anyway go to offset 121, you should be
sitting on a chr$(3). If not go back and forward one block -- you may have
miscounted. Again change the byte to a 7 and save the sector. That's it!

Do I need to tell you to do this on a COPY??? It  's not really that hard - I
can do it in less then 5 minutes. You may want to just do the Text Grabber
itself and a single form. Then boot geos and see how you did.

If you need more help just let me know (also tell me what kind of Disk Doctor
you are using so I can be more specific about the commands). --WC
 ------------
Category 19,  Topic 30
Message 7         Fri Jan 08, 1988
ALBERTMCCANN                 at 21:38 EST
 
While you are playing with the disk doc on Text-Grabber, change the 26th
sector (track 17, sector 18 ???) positions 20 and 21 from 09 20 to 8A 06. This
maks the default font 'COMMODORE 10 POINT' instead of the berkley font . This
makes it easier to print in NLQ mode. Has anybody decoded the word-processor
files that text grabber uses? Also an other tip: To inport from Superscript,
output to disk instead of printer in ASCII. Then use the '128 GENERIC II'
selection to convert it. This preserves the document formating with spaces.
 ------------
Category 19,  Topic 30
Message 8         Sun Jan 10, 1988
E.R.FLINN                    at 01:21 EST
 
     Hmmm, I *knew* there had to be some sort of reward for trying to learn to
navigate this message-base.  In one day I've got Text Grabber working AND can
make boot disks to my heart's content.  Very nice!  <ed> _
     The fix for the form files worked like a charm.  The only quibble I have,
and that's all it is, is that it would have been easier for me if the offsets
were given in hex as well.  No big deal.  The change mentioned in message 7,
though, to force Commodore 10 font, seems not to be for the 64 version, as the
expected data isn't there.  Or else I was looking in the wrong place.  I'm
going to be out of town for a week, so if I don't reply to any response, it's
not out of rudeness or lack of interest.
     It *is* kind of spooky to realize that somehow nobody at BSW bothered to
actually run the released version of the grabber on a real 64 running real
GEOS.  I haven't seen a patch on Qlink yet either, but that's no surprise.
 *s
 ------------
Category 19,  Topic 30
Message 9         Sun Jan 10, 1988
WC.COLEMAN [Geos*Sysop]      at 03:49 EST
 
There is no version of Geos that contains a Commodore 10 font that I know of
either. Anybody know what he is talking about???
 ------------
Category 19,  Topic 30
Message 10        Sun Jan 10, 1988
P.ROSSING                    at 08:08 CST
 
The instructions sound helpful, THANKS!  will try them as soon as i get done
here and let you know.  
    The "Commodore 10" font is on my Geowrite workshop disk (Geowrite version
2.1.  On screen it looks just like good  old Commodore 64 print, like you were
using Easyscript or something.  The letters are not proportionately spaced
either.   t  On printout, it's proportioned to be the same as printer output
from "Easyscript" or any other conventional word processor. In Geos mode, you
get the same "graphics" printing as with any Geos font. In NLQ mode, you can
use your printer's normal font.  The advantage of the "Commodore 10" font is
that you can get WYSWYG screen display for NLQ printing.
   does that help?

 ------------
Category 19,  Topic 30
Message 11        Sun Jan 10, 1988
WC.COLEMAN [Geos*Sysop]      at 15:28 EST
 
I knew it was a stand alone font but that is something you select from
GeoWrite, modifing the Text Grabber....Ah wait a sec! I see what you are
doing. You are changing the font ID that TG places in the grabbed file. Great
idea! Actually you should be able to change the default font to any valid font
ID. Of course if the font isn't on the disk the geoWrite will revert back to
BSW. Hmmm...technically you should be able to change the default Tabs and
margins also. I never looked for where they where, I was just trying to get
the damn thing to work!
 ------------
Category 19,  Topic 30
Message 12        Sun Jan 10, 1988
P.ROSSING                    at 20:00 CST
 
WC:  The instructions were beautiful.  they were easy to follow, and once I
did everything you said, I had a working program.  I repeat the curses of
everyone else -- doesn't BSW try this out before they unleash it on us
consumers?  
    I didn't fool with the defaults since I have different fonts on different
working disks, since with only one  disk drive, there isn't much you can get
on the same disk at once.
    Now that the text grabber works, it is a very useful program, one of the
few BSW applications that doesn't have to be on the same disk as the thing
being converted.   BSW must have stock in 1764 expansions, no?  
    thanks again.
    peg
 ------------
Category 19,  Topic 30
Message 13        Sun Jan 17, 1988
ALBERTMCCANN                 at 09:37 EST
 
That FONT patch for TEXT GRABBER is for the 128 version. You would have to
hunt around the 64 version for the font variable location. BTW, the 3/7 file
type problem is fixed in the 128 grabber. I've disassembled the 128 grabber
and discovered it supports EASY SCRIPT and SPEED SCRIPT also. Just no
conversion file on the disk. Are these files on the 64 workshop disk?
 ------------
Category 19,  Topic 30
Message 14        Sun Jan 17, 1988
WC.COLEMAN [Geos*Sysop]      at 20:54 EST
 
Yes I seem to recall that they are. The 3/7 bug as near as i've been able to
find out is not on all the 64 version disks, just some of them.
 ------------
Category 19,  Topic 30
Message 15        Mon Feb 15, 1988
WC.COLEMAN [Geos*Sysop]      at 18:32 EST
 
Just wanted to leave a quick note about the GeoPublish Text Grabber. The
reason it will work for some people and not for others is that it will work
fine for V1.3 and above but not for V1.2. If you ask BSW they will tell you
that it is not compatable with V1.2. This is untrue. If you just change the
filetypes of the forms as I explained above it will work fine with all
versions. Don't ask me why they did it that way (probably an oversight -- I
doubt they ever tested it with V1.2)!  -WC
 ------------

NEXT IS A LIST OF SOME UNDOCUMENTED SHORTCUTS!

 ************
Topic 12        Sat Jan 13, 1990
D.LARIVEE                    at 22:48 PST
Sub: Undocumented shortcuts                 

Has anyone found some undocumented shortcuts to use with GEOS 64 or 128? Lets
document them here.
1 message(s) total.
 ************
 ------------
Category 19,  Topic 12
Message 1         Sat Jan 13, 1990
D.LARIVEE                    at 22:53 PST
 
I have discovered by accident three shortcuts that I can't find in the manual.
When I tested them some more, I discovered that they are only available on the
GEOS 128 V2.0 (sorry 64 2.0 users). 1 and 2.  When you have selected the files
you want to copy to another  disk, press the F1 key to copy them to disk A or
F2 to copy to disk B. By magic, the files are automatically copied over
without having to  pick up and drag the icon over to the drive. The third
trick is moving files to the border.  You can highlight the files you want to
move to the border and then press the F4 key and they will be moved down to
the border for you.

Quite an interesting discovery I think.

Has anyone else discovered something they would like to share?
 ------------

HERE'S ONE FOR ALL YOU REU BUFFS!

 ************
Topic 37        Mon Feb 15, 1988
WC.COLEMAN [Geos*Sysop]      at 22:05 EST
Sub: Geos REU command summary.              

I have had a couple of questions on how to use the REU commands contained in
the Geos jump table. Here's the answer.
6 message(s) total.
 ************
 ------------
Category 19,  Topic 37
Message 1         Mon Feb 15, 1988
WC.COLEMAN [Geos*Sysop]      at 22:24 EST
 
There are five REU functions contained in the V1.3 Kernal. They are:

 $C2C8 - StashRAM. Transfer computer -> REU.
 $C2CB - FetchRAM. Transfer REU -> computer.
 $C2CE - SwapRAM.  REU <-> computer.
 $C2D1 - DoRAMOp.  Send a command to the REU.

For any of these routines pass:

 R0 - Computer destination/source address.
 R1 - REU destination/source address.
 R2 - Length in bytes.
 R3L - REU bank number (each bank is 64K).

For DoRAMOp the command to be sent to the REU should be passed in the Y
register in addition to the above. Nothing is returned by any of these
routines but you can read the REU status register at $DF00 as follows:

 Bit 7 - interupt pending.
 Bit 6 - Transfer Complete.
 Bit 5 - Block Error.
 Bit 4 - Size. 0 = 128K, 1 = 512K.
 Bits 0-3 are the version number.

All of these routines will hang in a loop until either bit 5 or 6 are set. The
X register always contains zero on exit so you will have to read $DF00 to
detect an error condition. That should be enough to get you started at least.
If there are any more questions just leave a message.  -WC
 ------------
Category 19,  Topic 37
Message 2         Wed Feb 24, 1988
M.THOMAS6                    at 21:12 CST
 
  Thanks for the info.  I ought to find some use for this later.

  Mark Thomas
 ------------
Category 19,  Topic 37
Message 3         Thu Feb 25, 1988
WC.COLEMAN [Geos*Sysop]      at 20:51 EST
 
By the way if you want to peek at the code (in V1.3) take a look at $9EAA to
$9F00 (it doesn't take much)!  -WC
 ------------
Category 19,  Topic 37
Message 4         Wed Mar 02, 1988
M.THOMAS6                    at 22:35 CST
 
  Does GEOS just drop the values into the memory locations at $DF00 or  where
ever?

  Mark Thomas
 ------------
Category 19,  Topic 37
Message 5         Fri Mar 04, 1988
WC.COLEMAN [Geos*Sysop]      at 06:16 EST
 
That's about it. Geos simply take the values passed to it and sends them to
the REU:   

 9EAA   LDY #93   ; VerifyRam entry
 9EAC   BNE $9EB8
 9EAE  LDY #$90   ; StashRam entry
 9EB0  BNE $9EB8
 9EB2  LDY #92   ; SwapRam entry
 9EB4  BNE $9EB8
 9EB6  ldy #$91  ; FetchRam Entry
 9EB8   ldx #$0D  ; DoRamOp Entry
 9EBA   lda $08
 9EBC   cmp $88C3
 9EBF   bcs $9F00
 9EC1   ldx $01
 9EC3   lda #$35 
 9EC5   sta $01   ; Basic out, Kernal out, I/O in
 9EC7   LoadW $DF02,$02
 9ED1   LoadW $DF04,$04
 9eDB   LoadB $DF06,$08
 9EE0   LoadW $DF07,$06
 9EEA   LDA #0:STA $DF09:STA $DF0A
 9EF2   STY $DF01
 9EF5   LDA $DF00  ; Wait loop..
 9EF8   AND #$60   ;  ... begins here
 9EFA   BEQ $9EF5  ; Not ready yet
 9EFC   STX $01    ; Config back the way it was
 9EFE   LDX #0     ; NEVER REPORTS ERRORS OF ANY KIND!
 9EFF   RTS

That's it (for V1.3 anyway)!  -WC
 ------------
Category 19,  Topic 37
Message 6         Sat Mar 05, 1988
M.THOMAS6                    at 16:24 CST
 
Well, that sounds pretty straight forward then.  Anyway, thanks for the info
on the calls.

     Mark Thomas
 ------------

HERE'S THE LIST OF BSw REFERENCE GUIDE ERRATA!

************
Topic 38        Mon Feb 15, 1988
WC.COLEMAN [Geos*Sysop]      at 22:56 EST
Sub: BSW Reference Guide Errata Sheet.      

BSW Reference Guide Errata Sheet.
1 message(s) total.
 ************
 ------------
Category 19,  Topic 38
Message 1         Mon Feb 15, 1988
WC.COLEMAN [Geos*Sysop]      at 23:01 EST
 
The following is the most complete list of BSW Reference Guide errata that I
have been able to compile. Here and there I've also thrown in a few of my own
comments:

 PAGE 18 - ResetHandle This is the Geos Cold Start Vector. Once the kernal is
loaded in a call to this routine will start things going. Unfortunatly this
routine calls a subroutine that checksums memory to see if GeosBoot is
resident and then uses this value to set the IRQ vector. Supposedly this will
prevent a user from loading in the Kernal using another (presumedly
unprotected) booter. For those who are interested the proper IRQ vector is
$E2DC. The offending (offensive?) routine is located (in Kernal V1.2) at
$DBAE.

 PAGE 29 - Menus The reference to i-DoMenu is not correct. Currently there is
no inline form of DoMenu. (Note: there may be one in the 128 version - I
haven't checked tha  one yet)!

 PAGE 30 - Menu Figure 2 All of the entries whos comment has the word
'pointer' in it are word length (two bytes) not single bytes. This is all of
them except for the two SUB-MENU entries. 

 PAGE 36 - DoMenu Again, there is no inline form of this routine.

 MENUS IN GENERAL There are also two other jump table Menu routines that are
not mentioned anywhere in the book at all:

 RecoverMenu ($C154) - This routine erases the current  submenu from the
screen. Period. It will not reactive the previous menu.

 RecoverAllMenus ($C157) - This routine calls RecoverMenu repeatedly until you
get back to the main menu. The main difference between this ro1tine and
GotoFirstMenu is that this routine will not re-draw the first menu nor
reposition the mouse. But it's also faster.

 PAGES 49-52 - PRGToGEOS I'm not even going to attempt to list all the errors
here. Do yourself a favor and tear these pages out of the book!

 PAGE 96 - BitOtherClip Third line: BitOtherClip allows the caller to specify
an input routine to return the next compacted byte in R0.
 PAGE 101 - GraphicString Commands This is not a complete list! See page 395
for a list of all the commands.

 PAGE 107 - Control Codes Outline od is CHR$(26). Plaintext is CHR$(27). Also
CHR$ codes from 0 to 7 will crash the system if you attempt to print them!!!
You must be careful to parse all strings that could possibly have these values
embedded in them.

 PAGE 112 - GetString There is no such thing as a MaxCharFault Vector. All
GetString will do is call StringFault Vector. One note about this routine:
there has been alot of confusion about how to use GetString. It is driven by
the interrupts so you should JMP to it after you have all the variables set
up. When the user hits Return control will be passed back to the application
at the address contained in KeyVector. Remember that the user will be able to
use the menu/icons at this time. If you don't want this to happen then
constrain the mouse. 

 PAGE 117 - OnrGetString This routine won't work right for a number of
reasons.

 PAGE 162 - UpdateMouseVels The opcode blt  is actually BCC. The opcode bge is
actually BCS.

 PAGE 219 - Position Command Third line of the paragraph should read: The
first data byte-is the x position in bytes, 0-39.

kPAGE 222 - Dialog Box Commands
 DBGetString - RegNumber should be the address of the zero page variable not
it's number, i.e. for R5 RegNumber would contain $0C not 5.

fDBGetFiles - the code for this command is 16 not 15. Also the maximum number
of files that this command can support is 15 not 16 (15*17=255).

 DBGrphStr - This one is command #15. See page 405 for a description.

 PAGES 244-245, 259, 284 - LoadOpt Flag Change all references to Bit 1 to Bit
7.

 PAGE 261 - GetFile Synopsis GetFile can load any GEOS filetype with the
notable exception of fonts. If Gou try to load a font with this routine it
won't do anything at all (unfortunatly it won't return an error either).

 PAGE 297 - FreeBlock The address mentioned in the last line as 'to be
supplied' (sic) is $9844. Remember this only applies to versi.n 1.2.

 PAGE 300 - BldGDirEntry One thing that is not metioned here is that you must
also pass the size of the file (in CBM blocks) in R2. This indirectly effects
SetGDirEntry since it will call this routine.

 PAGE 308 - PurgeTurbo This routine marks the turbo code for the indicated
drive as no longer resident in the drive. The next time you try to open the
drive the Turbo code will be have to be resent to the drive. There is also
another routine which will only turn it o f (it is still resident) - ExitTurbo
($C232). The parameters are the same as for PurgeTurbo.

 PAGE 317 - VLIR Files Three routines are mentioned but not listed anywhere.
They are:

AppendRecord - No parameters. This routine will mark the next available record
(usedRecords+1) as empty and increment the total records (usedRecords =
$8497).

InsertRecord - No parameters. Inserts a new (empty) record just after the
current record.

DeleteRecord - No parameters. Deletes the current record and frees up all
sectors the record used. s
 PAGES 347, 372 - GetDimensions There is one feature here that is not
mentioned anywhere except in the source code for the drivers: the accumulator
will be zero if the driver can do ASCII printing and $FF if it can only do
graphics. Keep this in mind if you ever write a printer driver.

 PAGE 392 - Selection_Delay This is the delay in 1/60 ths of a second.

 PAGE 398 - Geos Filetypes Note that System_Boot, Temporary, and Auto-exec are
only  valid in Kernal V1.3 and up.

 PAGE 420 - Drawing Modes This table is a mets! It should read:
 Bit 7 - UNDERLINE Bit 3 - OUTLINE
 Bit 6 - BOLDFACE  Bit 2 - unused
 Bit 5 - REVERSE   Bit 1 - unused
 Bit 4 - ITALICS   Bit 0 - unused

 PAGE 423 - Disk Buffers FileTrScTab is at $8300 not $8400. DirEntryBuf is at
$8400 not $8500.

 PAGE 425 - Vectors ApplicationMain should read $849B. Also note that
MouseVector is at $84A1, it's indented so it's easy to miss.

 PAGE 431 - CallRoutine The proper address for this routine is $C1D8.

 PAGE 432 - Other Routines The proper address of DoneWithIO is $C25F. The
proper address of CopyString is $C265.

 OTHER ODDS AND ENDS
 MENUS -  Do not define menus/submenus with a right margin greater then 255.
All sorts of strange things will occur. I haven't had a chance to track this
bug down yet but it is definatly there.

 ICONS - The maximum number of icons that can be defined at one time is 32
(numbers 0-31). This reason for this is quite simple. The icon table that you
pass to DoIcons is not moved, only a pointer to it is saved (in $3F-$40 to be
exact). To find a given icons position in the table use the following formula -
icon# * 8 + 4. Ok, let's plug in an icon # of 32 (remember this is actually
the 33rd icon): 32 * 8 + 4 = 260. Since the value is greater then 255 the icon
manager can't index into the table properly.

 DB ICONS - There are four icons that are listed as 'future expansion'
(commands #7-10). Do not ever call these icons! The system will crash if you
attempt to use them (no Panic box, it will just lock up). This can be a bear
to track down if you don't realize what these commands will do.

 TABLES - All of the routines that require tables (DoIcons, DoMenus,
InitProcess) will reset to the new table everytime they are called. Any
options that where active from a previous call to the routine will be lost
unless they are also contained in the new table. Technically speaking (I've
never tried it) you could disable menus by pointing to a table that contains
only a zero.  Of course you would still have to erase the old menu from the
screen.

If anybody finds anymore please let me know so I can add them to the ever-
growing list! -WC
 ------------

END OF FILE!!!! Enjoy - WC