8bitfiles.net/archives

home *** CD-ROM | disk | FTP | other *** search

/ 8bitfiles.net/archives / archives.tar / archives / genie-commodore-file-library / Information / 02CAT5MSGS.SFX / 05cat5msgs < prev

Wrap

Text File | 1990-02-12 | 56.3 KB | 1,513 lines

************ Topic 14 Mon Dec 05, 1994 P.VACULIN at 07:43 EST Sub: Print Text from Internet Sources This a problem of printing text downloaded or buffered from Internet sources. 7 message(s) total. ************ ------------ Category 5, Topic 14 Message 1 Mon Dec 05, 1994 P.VACULIN at 07:47 EST I have had a problem with text I download or buffer from all sources from the Internet. I guess that there are no carriage returns in the text I receive. By this I mean that the print-out looks like one continuous 80 column print job. Is there some utility for the C=64 that will properly format this text so it is in a readable form? If so, please let me know. I very much need the info so that I can have presentable documents for me and my work. << Patrick >> ------------ Category 5, Topic 14 Message 2 Mon Dec 05, 1994 THE.OUTLAW at 18:17 EST If you terminal program doesn't format text properly, you may have to use a Word Processor like TWS (The Write Stuff) to fix it. ------------ Category 5, Topic 14 Message 3 Mon Dec 05, 1994 C128.JBEE at 23:17 EST Velvetta64 might do the trick, I believe what you need to add is linefeeds. ------------ Category 5, Topic 14 Message 4 Wed Dec 07, 1994 P.VACULIN at 00:38 EST I wonder....is there possibly a program on the UNIX host that could do this (add linefeeds). If so, it might be much faster and possibly even automatic. BTW thanks to all of you who responded!!!! The reason I want to add the linefeeds automatically is that I frequently get e-mails regarding work that are in excess of 100 blocks. This is a lot of text to have to edit:> ------------ Category 5, Topic 14 Message 5 Wed Dec 07, 1994 H.HERMAN1 at 02:28 EST huh? I don't understand this.... I thought UNIX uses LF's. Howie ------------ Category 5, Topic 14 Message 6 Wed Dec 07, 1994 BEN.GELAMP [Ben] at 17:59 EST I'm not certain, myself. I just took a cursory look at a FAQ from the Internet, and its line terminators do indeed appear to be LF's. In any event, the fix isn't difficult, and all Commodore people have at their disposal the means to apply it: "real" programmers might laugh at Commodore BASIC, but I *still* claim it's a nicely capable language. Perhaps someone would be kind enough to translate the following program into Commodore BASIC? I'm very much afraid with the Commodore's keyboard in such sad shape, I can't try the code that's in my head :( The program, as is, works very nicely on the IBM. The advantage of doing it in BASIC is that it's so EASY to change if needed. 100 INPUT "Translate what file "; i$ 110 INPUT "To what file "; o$ 120 : 130 OPEN i$ FOR INPUT AS #1 ' rem open 1,8,3,i$+",s,r" 140 OPEN o$ FOR OUTPUT AS #2 ' rem open 2,8,4,o$+",s,w" 150 : 160 IF EOF(1) THEN 220 ' something involving ST 170 a$ = INPUT$(1, 1) ' get #1, a$ 180 v = ASC(a$) ' v=asc(a$+chr$(0)) 190 IF v = 10 THEN PRINT #2, CHR$(13); : GOTO 160 200 PRINT #2, a$; : GOTO 160 210 : 220 CLOSE 2 230 CLOSE 1 240 PRINT "Done." ------------ Category 5, Topic 14 Message 7 Tue Dec 27, 1994 B.KURSON [Zotar] at 14:10 EST One option is if you can try a straight ascii download instead of using kermit or z-modem, etc... kermit or another protacol. Another thing you can do is D/L the program ACE from the library ghere. It has a conversion utility. Also set file type to binary if FTP'ing even tho it is a text file. Hope this helps, Zotar ------------ ************ Topic 15 Sun Jul 17, 1994 D.TUOMI [Doctor] (Forwarded) Sub: Several Questions on Programming I'm working on a menu program and have several programming questions. Doc. 21 message(s) total. ************ ------------ Category 5, Topic 15 Message 1 Sun Jul 17, 1994 D.TUOMI [Doctor] (Forwarded) Okay, here is the details on my menu program. It's designed to be the menu for my user group's disk of the month. The current version's features are as follows: - Able to operate in either 64 or 128 mode - Operates in either 40 or 80 colummn mode on the 128 - If menu is loaded in 128 mode, it will automatically load 64 software and switch mode of operation and run - Displays and Prints text files This is what version 1.14 is able to do. I'm about to begin work on version 2.0 of the software, but I have some questions which are stumping me that I'd like to ask. Rather than overwhelm with questions, I'll ask one. Wait for a hopeful reply or an acknowledgement that no one knows, then I'll move on. If the program is successful, perhaps you'll see a PD version of it. First question. I currently have the program where it will run in either 40 or 80 column mode on the 128. The menu program can tell by the index file it loads whether the selected program to load is supposed to run in either 40 or 80 column mode and the program will instruct the person to switch the monitor accordingly. Is there anyway of telling if a user has a 16k VDC or a 64k VDC? That way if a program REQUIRES 64k VDC, I could tell an unupgraded flat 128 owner that he cannot run the program selected. Any ideas? Doc. ------------ Category 5, Topic 15 Message 2 Sun Jul 17, 1994 CBM-ED [e.g.bell] (Forwarded) Doc: DT> Is there anyway of telling if a user has a 16k VDC or a 64k DT> VDC There may be. I don't know for sure, but if I were trying to do it, I think what I would do would be to write a character to an address in the second 16K, 16384 for example, and see if it appeared on the screen. That means a simple poke/peek using the routines at $cdcc/$cdda to do that. Actually, it would only take a poke cuz if the character showed up it would suggest to me that the VDC is only 16K. You *might* need to twiddle bit 4 of register 24 in the VDC to do this tho. It is supposed to be 1 for 16K, but that is something done by software. You may have to do it tho to make sure your poke works correctly if there *is* 64K vdc installed. Try it like this: 10 sys dec("cdcc"),64,18:rem set for address 16384 20 sys dec("cdcc"),0,19 30 sys dec("cdcc"),65:rem poke the letter 'a' into memory there 35 : 40 sys dec("cdcc"),0,18:rem now set for address 0 45 sys dec("cdcc"),0,19 50 sys dec("cdd8"):rreg a 60 if a=65 then <16K VDC chip> Line 60 will be your decision line. I can't guarantee this will work. It did on my 128 w/16K VDC. What it does is poke the letter into 16384, which doesn't exist, then read 0, and see if that character is in there. Let me know if it works. If so, this is a nice q&d routine to do what you want. The only down side is that it will put an 'a' in the upper left corner of your 80 col screen if in that mode. Another way to do it would be to try another address in the 2nd 16K that corresponded with the zero bytes of padding between characters.... totally transparent to the user that way.... just a little more figuring for the addresses, but no real change in the routine other than that. ------------ Category 5, Topic 15 Message 3 Sun Jul 17, 1994 H.HERMAN1 (Forwarded) Doc, This is the way Lou Wallace did it: 20010 poke 23013,0:poke23026,0 20020 @walrus,1:@color,0,0,0:@clear,0,0,0 20030 scrdef,1,0,0,640,400,16000,0:@screen,1,0:@clear,255,0,0 20040 @screen,0 20050 vr=0:tq=@pixel,0,0,0:if tq=0 then vr=1 20060 poke23013,134:poke23026,134 20070 return It would be exceedingly helpful, however, before you run the above subroutine, if you've booted BASIC 8. :) Howie ------------ Category 5, Topic 15 Message 4 Sun Jul 17, 1994 H.HERMAN1 (Forwarded) Doc, Thinking about it, a B8 RTL module should do it. Howie ------------ Category 5, Topic 15 Message 5 Sun Jul 17, 1994 D.TUOMI [Doctor] (Forwarded) Thanks Mr. Bell, I'll try your routine and see what it does. Difficult getting around on the VDC, there doesn't seem to be any central location for information on this chip. It's kind of scattered from the Programmer's Reference Guide, 128 Internals, etc. Wish I had a Mapping the 128, probably some useful stuff in there too. Anyone have a copy they want to sell? Are there any other books which have this sort of technical data in it? Howie, I appreciate the info, but my program is written to run stand alone. BASIC-8, while a nice language, produces copyright problems as well as compatibilty problems with my routine to swtich to 64 mode to run a program. Thanks for the info though, might need it in the future. Anyone have any other ideas? Doc. ------------ Category 5, Topic 15 Message 6 Mon Jul 18, 1994 CBM-ED [e.g.bell] (Forwarded) You do need to get Mapping the 128.... good luck! Them that has it keeps it! It has an excellent coverage of the VDC registers. However the routines I posted for you may be what you need. ------------ Category 5, Topic 15 Message 7 Mon Jul 18, 1994 H.HERMAN1 (Forwarded) Doc, The BASIC 8 package includes a RunTime Library (RTL), which allows any user to create his/her own stand-alone BASIC 8 programs. Any programs created with the RTL can be distributed however the programmer wants, including commercially. I _think_ that IPAINT and IPORT might be examples of this. Not sure. Perhaps someone might know? More to the point, however, for your purpose you would need to first determine that the user has a 128, then you could call in a RTL module to read the vdc, based on the earlier post. I am pretty sure that Ed's suggestion is the way to go. However, I did want to point out an alternative, even if it is a bit more, abstract. Howie ------------ Category 5, Topic 15 Message 8 Tue Jul 19, 1994 D.TUOMI [Doctor] (Forwarded) I don't actually own a copy of BASIC-8. Nor do I know where to get one. It seems to be out of print. Bell, I'm having problems with your routine. On both my 128's the value returned is a 32. This isn't the value expected, and one 128 has a 64k and the other has a 16k, so there should be some difference. Have you tested this routine yourself? If so, then I'll go back check it again, might of missed something myself. Doc. ------------ Category 5, Topic 15 Message 9 Tue Jul 19, 1994 CBM-ED [e.g.bell] (Forwarded) Doc: I looked at the stuff I sent you and there is at least 1 error in it. Try this.... 10 bank 15 20 sys dec("cdda"),,28:rreg a 30 a=a or 16:sys dec("cdcc"),a,28:sys 65378 35 : 40 sys dec("cdcc"),64,18 45 sys dec("cdcc"),0,19 50 sys dec("cdca"),65 55 : 60 sys dec("cdcc"),0,18 65 sys dec("cdcc"),0,19 70 sys dec("cdd8"):rreg b 75 sys dec("cdcc"),a and 239,28:sys 65378 80 : 85 if b <> 65 then <128 has 64K VDC> You must make SURE that when you run this the first position in the 80 column screen does not contain a lower case 'a'. I can't test this for you, but I'm reasonably sure it will work. If you want a blow-by-blow: Line 10 makes sure all SYS calls are into the screen editor routines Lines 20 and 30 set the machine for 64K VDC and resets the character set. Lines 40 thru 50 poke a lower case 'a' into address 16384, one byte above the standard 16K chip RAM. Lines 60 thru 70 peek address 0 to see if it contains that lower case 'a' If it does, the assumption is that address 16384=0 in the 16K VDC Line 75 restores bit 4 of register 28 to 'off' and again calls the character set installation routine Line 85 tests to see which size RAM is installed. if 'a' is in address 0 it is assumed that you have a 16K VDC. Hope that all makes sense. Sorry for the error in my earlier post. On the upside the program is small. Incidentally, this should work fine in either 40 or 80 column mode. Let me know if this one does the trick. egb ------------ Category 5, Topic 15 Message 10 Wed Jul 20, 1994 THE.OUTLAW (Forwarded) Doc, I just came across exactly what your looking for.... Fred Bowen wrote a utility to do what you ask. ------------------------------------------ 1 REM FRED'S NIFTY PROGRAM TO DETERMIN SIZE OF 8563 DRAM 10 BANK 15: AD=DEC("D600"): DA=AD+1: GOSUB 900 :REM SETUP ML 20 POKEAD,28: S=PEEK(DA): POKEDA,63 :REM SELECT 64K 30 I=16896: SYSW,I/256,18:SYSW,IAND255,19SYSW,85,31 :REM WRITE $55 40 I=16896: SYSW,I/256,18:SYSW,IAND255,19:SYSR,,31:RREGC1 :REM READ HERE 50 I=17152: SYSW,I/256,18:SYSW,IAND255,19:SYSR,,31:RREGC2 :REM AND HERE 60 I=16896: SYSW,I/256,18:SYSW,IAND255,19:SYSW,170,31 :REM WRITE $AA 70 I=16896: SYSW,I/266,18:SYSW,IAND255,19:SYSR,,31:RREGC3 :REM READ HERE 80 I=17152: SYSW,I/256,18:SYSW,IAND255,19:SYSR,,31:RREGC4 :REM AND HERE 90 POKEAD,28: POKEDA,S: SYSDEC("FF62") :REM RESTORE 16/64K 100 IF C1=C2 AND C3=C$ THEn PRINT "16K": ELSE PRINT "64K" :REM DID IT ECHO? 110 END 900 FORI=0TO13: READA$: POKEDEC("180")+DEC(A$): NEXT 910 R=DEC("1800"): DATA 8E,00,D6,AD,01,D6,60 :LDA $D601: RTS 920 W=DEC("1807"): DATA 8E,00,D6,8D,01,D6,60 :STA $D601: RTS 930 RETURN ------------ Category 5, Topic 15 Message 11 Wed Jul 20, 1994 THE.OUTLAW (Forwarded) Sorry about that.. I straightened it out... 1 REM FRED'S NIFTY PROGRAM TO DETERMIN SIZE OF 8563 DRAM 10 BANK 15: AD=DEC("D600"): DA=AD+1: GOSUB 900 :REM SETUP ML 20 POKEAD,28: S=PEEK(DA): POKEDA,63 :REM SELECT 64K 30 I=16896: SYSW,I/256,18:SYSW,IAND255,19SYSW,85,31 :REM WRITE $55 40 I=16896: SYSW,I/256,18:SYSW,IAND255,19:SYSR,,31:RREGC1 :REM READ HERE 50 I=17152: SYSW,I/256,18:SYSW,IAND255,19:SYSR,,31:RREGC2 :REM AND HERE 60 I=16896: SYSW,I/256,18:SYSW,IAND255,19:SYSW,170,31 :REM WRITE $AA 70 I=16896: SYSW,I/266,18:SYSW,IAND255,19:SYSR,,31:RREGC3 :REM READ HERE 80 I=17152: SYSW,I/256,18:SYSW,IAND255,19:SYSR,,31:RREGC4 :REM AND HERE 90 POKEAD,28: POKEDA,S: SYSDEC("FF62") :REM RESTORE 16/64K 100 IF C1=C2 AND C3=C$ THEn PRINT "16K": ELSE PRINT "64K" :REM DID IT ECHO? 110 END 900 FORI=0TO13: READA$: POKEDEC("1800")+DEC(A$): NEXT 910 R=DEC("1800"): DATA 8E,00,D6,AD,01,D6,60 :LDA $D601: RTS 920 W=DEC("1807"): DATA 8E,00,D6,8D,01,D6,60 :STA $D601: RTS 930 RETURN ------------ Category 5, Topic 15 Message 12 Wed Jul 20, 1994 D.TUOMI [Doctor] (Forwarded) Well, Mr. Bell, I was unable to get your version of the program operational. Once again no matter what I do it comes back with the same value, in this case a "32." After much decyphering I did have more luck with Outlaw's routine. It actually worked perfectly on my flat 128, and my 128d. It successfully was able to tell the difference between the two. Does anyone have an upgraded flat 128 that we could try this program on? It wouldn't be of much help if the routine only worked on 128d's and 16k flat 128's and not on upgraded 64k flat 128's. Doc. ------------ Category 5, Topic 15 Message 13 Thu Jul 21, 1994 CBM-ED [e.g.bell] (Forwarded) Well, I can't speak for the 64K VDC, but the 16K VDC of my 128 returns the 65. I have the chips now and will have to install them and give that a try. But if Fred Bowen's routine works, I'd suggest you go with that. I can't imagine why it would not work on an upgraded VDC, tho CAM's posts were unreadable for me. (*sn Cam.... keep repeating *sn) ------------ Category 5, Topic 15 Message 14 Thu Jul 21, 1994 H.HERMAN1 (Forwarded) Ed, Ah, good! They arrived!!! >..... tho CAM's posts > were unreadable for me. (*sn Cam.... keep repeating *sn) Thank goodness.... I was about to return my new glasses to the optician..... Howie ------------ Category 5, Topic 15 Message 15 Thu Jul 21, 1994 THE.OUTLAW (Forwarded) That'll teach me to do things late at nite! At least the 2nd time it was readable! :> ------------ Category 5, Topic 15 Message 16 Thu Jul 21, 1994 THE.OUTLAW (Forwarded) Uh Ed? I think you misplaced this post. Should be cat 3 top 18? Glad I'm not the only one who makes mistakes! <g> ;> ------------ Category 5, Topic 15 Message 17 Thu Jul 21, 1994 D.TUOMI [Doctor] (Forwarded) I'm not sure why it doesn't work. I read once in the C128 Internals book though that BASIC is too slow for most VDC operations and that their programs written in BASIC didn't operate reliably. Perhaps that's the cause. The Outlaw's routine worked fine, although it took a bit of work to figure it out. I actually rewrote the routine posted, bcause it was a little "over- engineered" for my tastes. I got it down to about a quarter of its original size and it still works the same. Doc. ------------ Category 5, Topic 15 Message 18 Fri Jul 22, 1994 CBM-ED [e.g.bell] (Forwarded) Here is the program The.Outlaw posted from Fred Bowen. :) 1 REM FRED'S NIFTY PROGRAM TO DETERMIN SIZE OF 8563 DRAM 10 BANK 15: AD=DEC("D600"): DA=AD+1: GOSUB 900 :REM SETUP ML 20 POKEAD,28: S=PEEK(DA): POKEDA,63 :REM SELECT 64K 30 I=16896: SYSW,I/256,18:SYSW,IAND255,19SYSW,85,31 :REM WRITE $55 40 I=16896: SYSW,I/256,18:SYSW,IAND255,19:SYSR,,31:RREGC1 :REM READ HERE 50 I=17152: SYSW,I/256,18:SYSW,IAND255,19:SYSR,,31:RREGC2 :REM AND HERE 60 I=16896: SYSW,I/256,18:SYSW,IAND255,19:SYSW,170,31 :REM WRITE $AA 70 I=16896: SYSW,I/266,18:SYSW,IAND255,19:SYSR,,31:RREGC3 :REM READ HERE 80 I=17152: SYSW,I/256,18:SYSW,IAND255,19:SYSR,,31:RREGC4 :REM AND HERE 90 POKEAD,28: POKEDA,S: SYSDEC("FF62") :REM RESTORE 16/64K 100 IF C1=C2 AND C3=C$ THEn PRINT "16K": ELSE PRINT "64K" :REM DID IT ECHO? 110 END 900 FORI=0TO13: READA$: POKEDEC("1800")+DEC(A$): NEXT 910 R=DEC("1800"):DATA 8E,00,D6,AD,01,D6,60 :LDA $D601: RTS 920 W=DEC("1807"): DATA 8E,00,D6,8D,01,D6,60 :STA $D601: RTS 930 RETURN [Reformatted w/RTCM 2.04 Editor Overlay]:) ------------ Category 5, Topic 15 Message 19 Fri Jul 22, 1994 CBM-ED [e.g.bell] (Forwarded) Incidentally, I just noticed a but in the prog from Fred, tho it was probably just mistyped when posted. In line 100, C$ should be C4... someone was just holding the shift key when they hit the 4. :) ------------ Category 5, Topic 15 Message 20 Fri Jul 22, 1994 THE.OUTLAW (Forwarded) Your Right Ed! Thanks for RE-Redoing the post! I was about to try it a 3rd time and see if I could avoid screwing it up again! :> ------------ Category 5, Topic 15 Message 21 Sat Jul 23, 1994 D.TUOMI [Doctor] (Forwarded) I got it going, but I'm curious. This program basically makes the check twice with two different values. Is there a reason for this redundancy? Doc. ------------ ************ Topic 16 Sun Jul 24, 1994 CBM-ED [e.g.bell] (Forwarded) Sub: Printer Questions I have run into a problem that my manuals do not cover. Can someone help me program my printer. (A printer Programmer's forum) 5 message(s) total. ************ ------------ Category 5, Topic 16 Message 1 Sun Jul 24, 1994 CBM-ED [e.g.bell] (Forwarded) This seems to be an uncovered topic, and I have a need that I can't find the answer to in my printer manual. I am hoping someone may be able to provide a solution. In GWBASIC on IBM machines, there is a command called WIDTH which is used to set the width of printout. What are the escapes this command sends to the printer. I'm having a problem with my printouts in transparent mode doing line feeds at column 80. I know that the WIDTH command fixes this when I do IBM BASIC stuff. Just can't find the right codes to send in Commodore BASIC. Thanks! ------------ Category 5, Topic 16 Message 2 Mon Jul 25, 1994 H.HERMAN1 (Forwarded) Ed, I do not know exactly what WIDTH sends. However, the following should work, with Epson emulating printers: print#1,chr$(27)+chr$(81)+255 If the above does not work, try this variation: print#1,chr$(27)+chr$(81)+"255" One or the other will define an Epson printer's right margin to 255. Therefore, _only_ after every 255 characters the printer will insert a CR, or CR/LF, depending upon how you set things up. If this doesn't do it, let me know, and I'll hunt through some manuals for more ideas. Howie (who has not tested this, so can offer no gaurantees that it will work......) ------------ Category 5, Topic 16 Message 3 Mon Jul 25, 1994 H.HERMAN1 (Forwarded) Oooopsss... Assuming those lines do it, be sure to add a semi-colon: print#1,chr$(27)+chr$(81)+255; or: print#1,chr$(27)+chr$(81)+"255"; Howie ------------ Category 5, Topic 16 Message 4 Mon Jul 25, 1994 CBM-ED [e.g.bell] (Forwarded) Howie: I had already tried those commands unsuccessfully. Thanks for the input tho. There is still a happy ending to this story though. The problem was my stupidity. I thought I was setting my interface into transparent mode, but I inadvertently set it for 1525 mode. All the dips go one way for that mode, the other for transparent... thus I was doomed to 80 cols at that setting no matter what. The upside is that I found and fixed the situation and all is working well now. duhhhhhh! ------------ Category 5, Topic 16 Message 5 Tue Jul 26, 1994 H.HERMAN1 (Forwarded) Ed, I am happy to hear you have the prob solved. It is strange how, at times, we can get so involved with the nitty-gritty details, that it is easy to forget a basic, like a correct interface setting... Howie ------------ ************ Topic 17 Fri Aug 12, 1994 D.HARKER1 [Dan] (Forwarded) Sub: PETASCII printer codes Help needed with the differences between PETASCII and ASCII 1 message(s) total. ************ ------------ Category 5, Topic 17 Message 1 Fri Aug 12, 1994 D.HARKER1 [Dan] (Forwarded) I need some help with the PETASCII printer codes for a program I'm working on. I need to know if the PETASCII codes (dec)0 - (dec)31 are the same as in standard ASCII. I especially need to know the ones for paper movement (formfed, vertical tab, etc.). Can someone who has a manual for a Commodore printer help me and/or upload the codes? On another question (sorry, don't know where else to ask this), is there a 'standard' number of lines to a printed page on a printer? If so, what is the number? Thanks in advance for any help anyone can give. Dan ------------ ************ Topic 18 Fri Feb 10, 1995 D.BELTER (Forwarded) Sub: Basic to ML and Back again... Just some routines that will you program in ML and Basic! 4 message(s) total. ************ ------------ Category 5, Topic 18 Message 1 Fri Feb 10, 1995 D.BELTER (Forwarded) I thought the following would be helpful for anyone who programs on the C64 in ML and Basic. Some will let you SEND NNumbers or Text to ML and also RECIEVE text from ML. One last thing, I couldn't figure out how to save the TEXT to a Basic area of memory, so it would be good if you did if you did a C$=MID(C$,1) which will move C$ from beyond $C000 to normal basic memory. If someone has an SYS (JSR) that would do this please tell me since I could never get it to work. print"1. asm":print"2. run":inputa:ifa=1then1070 z$=chr(0):a=2000:b=125:a$="var. test"+z$:b$="ptrget test":q$=chr(34) printq$;:sys49152+3,b$,c$:c=len(c$):c$=mid(c$,1) printq$:printc$,c,len(c$):gosub1061 sys49152,a$,a,a+a,a+a+a,a+a+a+a,b+b,b print sys49152,"i read this"+z$,1000,-3000,5000,7000,125,255 end sys45195c$ vp=peek(780)+peek(782)*256:printvp;peek(vp+1)+peek(vp+2)*256:return sys 999 .org 49152 .mem ;******* c o n s t a n t s ******* ; < - lo ; > - hi chrget =$0073 ;get comma chkcom =$aeff ;get comma frmevl =$ad9e strout =$ab1e ;a-lo,y-hi ptrget =$b08b ;a-lo,y-hi conint =$b1aa ;y-lo,a-hi len =$b77c frestr =$b6a3 ;a-len,x-lo,y-hi comp'len =$b3a2 getbyt =$b79b ;x-lo getnum =$b7eb ;get 16 and 8 bit getadr =$b7f7 ;y-lo,a-hi linprt =$bdcd ;print a-hi,x-lo getinx =$e200 ;same as - getbyt chkcmm =$e20e ;same as - chkcom chrout =$ffd2 movstr =46728 ;a-len;x-lo;y-hi ;******* v a r i a b l e s ******* linnum =$14 index =$22 datlin =$64 var =$fb pointer =$fd length =$02 ;******* start of program ****** : jmp big ;ptrget ascii test : jsr chrget : jsr ptrget : sta index : sty index+1 : ldy #00 : lda (index),y : sta length : iny : lda (index),y : sta pointer : iny : lda (index),y : sta pointer+1 : ldy #00 - lda (pointer),y : jsr chrout : iny : cpy length : bne - : jmp send'var ;*** get asc text big = * : jsr chrget : jsr frmevl : jsr frestr : sta length : lda index ;lo : ldy index+1 ;hi : jsr strout : jsr space ;*** get two bytes ;positives : jsr chkcmm : jsr frmevl : jsr getadr ;y-lo a-hi : ldx linnum ;x-lo : lda linnum+1 ;a-hi : jsr linprt : jsr space ;*** get two bytes ;negatives : jsr chkcom : jsr frmevl : jsr conint ;y-lo ; a-hi : lda datlin ;hi than lo byte : ldx datlin+1 : jsr linprt : jsr space ;*** get two bytes ;positives : jsr chkcom : jsr frmevl : jsr getadr ;y-lo a-hi : ldx linnum ;or linnum : lda linnum+1 : jsr linprt : jsr space ;*** get two bytes than one byte : jsr chkcom : jsr getnum : stx var : ldx linnum : lda linnum+1 : jsr linprt : jsr space : lda #00 : ldx var : jsr linprt : jsr space ;*** get one byte : jsr getinx : lda #00 : jsr linprt : rts space lda "," : jmp chrout : jsr linprt send'var = * : jsr chrget : jsr ptrget : sta index : sty index+1 : ldy #00 : lda #estore-store : sta (index),y : iny : lda <store : sta (index),y : iny : lda >store : sta (index),y : rts store = * : .asc "send this stuff" estore = * ------------ Category 5, Topic 18 Message 2 Fri Feb 10, 1995 E.GBELL [e.g.bell] (Forwarded) Dave: could you resend this message, but save it with '*sn' so that the format does not get trashed. :) I love source code of any size, but this would be a bit of work getting into a format I could understand and work with easily. Thanks! ------------ Category 5, Topic 18 Message 3 Fri Feb 10, 1995 C128.JBEE (Forwarded) You need to use *sn when sending formatted text to GEnie :) ------------ Category 5, Topic 18 Message 4 Wed Feb 22, 1995 D.BELTER (Forwarded) Ok lets try this again, that was strange since all it was formatted with was spaces, never had a problem before and it looked ok before I saved it. So here it again... print"1. asm":print"2. run":inputa:ifa=1then1070 z$=chr(0):a=2000:b=125:a$="var. test"+z$:b$="ptrget test":q$=chr(34) printq$;:sys49152+3,b$,c$:c=len(c$):c$=mid(c$,1) printq$:printc$,c,len(c$):gosub1061 sys49152,a$,a,a+a,a+a+a,a+a+a+a,b+b,b print sys49152,"i read this"+z$,1000,-3000,5000,7000,125,255 end sys45195c$ vp=peek(780)+peek(782)*256:printvp;peek(vp+1)+peek(vp+2)*256:return sys 999 .org 49152 .mem ;******* c o n s t a n t s ******* ; < - lo ; > - hi chrget =$0073 ;get comma chkcom =$aeff ;get comma frmevl =$ad9e strout =$ab1e ;a-lo,y-hi ptrget =$b08b ;a-lo,y-hi conint =$b1aa ;y-lo,a-hi len =$b77c frestr =$b6a3 ;a-len,x-lo,y-hi comp'len =$b3a2 getbyt =$b79b ;x-lo getnum =$b7eb ;get 16 and 8 bit getadr =$b7f7 ;y-lo,a-hi linprt =$bdcd ;print a-hi,x-lo getinx =$e200 ;same as - getbyt chkcmm =$e20e ;same as - chkcom chrout =$ffd2 movstr =46728 ;a-len;x-lo;y-hi ;******* v a r i a bl e s ******* linnum =$14 index =$22 datlin =$64 var =$fb pointer =$fd length =$02 ;******* start of program ****** : jmp big ;ptrget ascii test : jsr chrget : jsr ptrget : sta index : sty index+1 : ldy #00 : lda (index),y : sta length : iny : lda (index),y : sta pointer : iny : lda (index),y : sta pointer+1 : ldy #00 - lda (pointer),y : jsr chrout : iny : cpy length : bne - : jmp send'var ;*** get asc text big = * : jsr chrget : jsr frmevl : jsr frestr : sta length : lda index ;lo : ldy index+1 ;hi : jsr strout : jsr space ;*** get two bytes ;positives : jsr chkcmm : jsr frmevl : jsr getadr ;y-lo a-hi : ldx linnum ;x-lo : lda linnum+1 ;a-hi : jsr linprt : jsr space ;*** get two bytes ;negatives : jsr chkcom : jsr frmevl : jsr conint ;y-lo ; a-hi : lda datlin ;hi than lo byte : ldx datlin+1 : jsr linprt : jsr space ;*** get two bytes ;positives : jsr chkcom : jsr frmevl : jsr getadr ;y-lo a-hi : ldx linnum ;or linnum : lda linnum+1 : jsr linprt : jsr space ;*** get two bytes than one byte : jsr chkcom : jsr getnum : stx var : ldx linnum : lda linnum+1 : jsr linprt : jsr space : lda #00 : ldx var : jsr linprt : jsr space ;*** get one byte : jsr getinx : lda #00 : jsr linprt : rts space lda "," : jmp chrout : jsr linprt send'var = * : jsr chrget : jsr ptrget : sta index : sty index+1 : ldy #00 : lda #estore-store : sta (index),y : iny : lda <store : sta (index),y : iny : lda >store : sta (index),y : rts store = * : .asc "send this stuff" estore = * ------------ 5 ? ************ Topic 20 Sun Sep 11, 1994 T.BROWN25 [Tim] (Forwarded) Sub: Hi-Res Printing I need help compacting character data stored in RAM 5 message(s) total. ************ ------------ Category 5, Topic 20 Message 1 Sun Sep 11, 1994 T.BROWN25 [Tim] (Forwarded) Greetings to one and all, And thanks for reading this help request! I've encountered a bit of a dilemma while coding a paint program and would appreciate any advice on how to deal with it. The system I'm working with is a C64 & 1541 II, no JIFFYDOS or any other enhancements. Basically, the program is a utility to create multicolor bitmap screens for use in other programs. RUNPaint, DOODLE, and GEOS may be similiar. You can't easily use GEOS pics in non-GEOS programs, however. I want to prevent the tool icons from cluttering the drawing area, and to not have to scroll. The whole screen as it appears in the using program needs to be visible. The program will have several features, but for now, this is what is coded: - multicolor bitmap - keyboard & joystick control - line draw routine - 16 x zoom - printing keyboard characters in 8,10,12,14, & 16 point sizes Future enhancements: mouse, REU, patterns, etc. Everything is coded in machine language and bug free. The program works fine for now. The problem that I have is that the character data eats too much memory KERNAL ROM is the only ROM I use, no BASIC or CHARACTER ROM involved in the program. For now there is plenty of available RAM. However, if a selection of styles to print is desired, (and it is), another set of character data would break the bank. What I need is a way of compacting my character data. What follows is an example of how I print characters to the screen. The size of the character will determine the number of pixels tall, or rows. Using bit-pairs, more than one screen byte may have to be used per row. An 8- point character will have 4 bit-pairs in each of 8 rows. A 10-point character will have 5 bit-pairs in each of 10 rows, etc. 8-point: 00000000 00001100 00111111 00110011 00111111 00110011 00110011 00110011 10-point: 00000000 00000000 00001111 00000000 00111111 11000000 00110000 11000000 00110000 11000000 00111111 11000000 00111111 11000000 00110000 11000000 00110000 11000000 00110000 11000000 00000000 00000000 00000000 00000000 00000000 00000000 00000000 00000000 00000000 00000000 00000000 00000000 What I would like to be able to do is have one size defined in RAM and be able to expand it at will to fit the other sizes. But, as can be seen from the above examples, taking the data from the 8-point character and manipulating it to form the 10-point character is not exactly feasable. Therefore, a seperate set of data is required for each size character set. At present, I've compacted the 20 bytes that define the 10-point "A" in the above example so that it takes only 12.5 bytes to store the data. The data is stored 10 bits at a time like so: 00000000 00000011 11000011 11111100 11000011 00110000 11001111 11110011 11111100 11000011 00110000 11001100 00110000 During printing, a subroutine breaks-up the data and aligns it into columns, then plots it to the screen. So, for one font with all sizes, the memory requirements for storage is as follows: 8pnt- 8 x 91 char = 728 bytes 10pt-12.5 x 91 char = 1137.5 12pt- 18 x 91 char = 1638 bytes 14pt-24.5 x 91 char = 2229.5 16pt- 32 x 91 char = 2912 bytes ------------- total = 8645 bytes What could I do differently to save memory and still accomplish the same thing? I'd like to have different styles in memory at the same time; at present I can't. If anyone can understand this spaghetti that I've written, please give me whatever advise that you can. I thank you in advance. Sincerely, T.J. Brown ------------ Category 5, Topic 20 Message 2 Sun Sep 11, 1994 CBM-ED [e.g.bell] (Forwarded) TB: TB> 8-point: 00000000 TB> 00001100 TB> 00111111 TB> 10-point: 00000000 00000000 TB> 00001111 00000000 TB> 00111111 11000000 TB> taking the data from the 8-point character and manipulating it TB> to form the 10-point character is not exactly feasable Why not? It seems to me that the only problem you are running into here is locking yourself into thinking in terms of 8 bit bytes. I can understand why, but from what you are showing here, except for the height, it would be a fairly straight forward procedure to create wider or narrower characters. Perhaps it is the 'spaghetti' you have written. I know I've talked to you about this before. What algorithm did you use to finally compact the font you have compacted. I can see one tricky part would be the fact that you are working with multi-color, but your point sizes are all in multiples of 2, right? Here is what I suggested to you before, as I remember it. Decide on one or more 'key' bits in a letter's byte, perhaps decided by the point size being rotated. For example: 10 pt. bits 6/7 14 pt. bits 4/5 12 pt bits 2/3 16 pt bits 0/1 Then, your expansion routine could duplicate the appropriate bit pairs while expanding... for example, using the above scheme, your routine for a 10 point font would duplicate bits 6/7 in the expansion, giving you a 10 point 'emulation' of th original character... it would have 2 extra bits in it, which I don't think you are going to be able to avoid under any circumstance, for each row of the character in the bitmap, but they would be 'predictable' throughout any character set,so to speak. Such a scheme would probabl work from the middle instead of the ends of course (in terms of bit pairs duplicated). The height would be done similarly as far as deciding which 'row' of bit pairs was duplicated in the final character (or rows) to get the desired extra number of rows to bring the character up to final height. Does this make sense? If I were doing the project, I *think* this is the approach I would take as I understand things right now. From you demo characters, the only fly in the ointment above would be the 16 point characters, believe it or not, because one row of your example (one column actually) is all '00' pairs, naturally, to allow for intercharacter spacing. The obvious solution to this (to me) is to make the column affected by a 16 pitch the same as the one for 14 pitch, or else make it user-selectable between pairs 2/3, 4/5, and 6/7... if you get my drift... this could even be presented as a feature, allowing a minor difference between characters in a given font. (You could evene extend the principle, I suppose to include 10, 12, and 14 point fonts, AND to the height expansion as well). For example: For a 10 point font, you need 10 bits for your character. Create it using ASL:ROL combinations, but when you get to the specified 'duplicate' bit pair, do it twice. When you get to the specified 'duplicate' row of bit pairs, do the whole row twice. For a 12 point font, you would have 2 specified bit pairs and 4 specified rows of bit pairs. For 14, you would have 3 specified bit pairs and 6 specified rows of bits, and for 16, you would have all 4 bit pairs and all 8 rows for duplication. In each of the expansions, you could allow user selection of which bit pairs and rows were slated for double expansion in both directions, even allowing them to be the same ones... lots of effect with no RAM overhead, even with a single font. The down side is the same as what I mentioned before. You have to track the position of each character so you will know where to start next. You have to stop thinking in terms of 8 bit 'cards', tho this might just involve only keeping a record of where the previous character started in the bit map and adding an appropriate amount to get to the next location.... same for the height. You would have to do this anyway unless you are sticking exclusively to starting on 8 bit boundaries, which I can't imagine you would be doing.... you said you already have plot routines. Incidentally, a down side of all of this is that it would probably preclude proportional font, but that might be kind of limited anyway by the multi-color thing. But if it was important, you *might* want to take a clue from the way proportional font characters are defined for printers that support user-creatable fonts... they have specifiers I believe, right in the character set, to tell the printer how much spacing is required by the character, etc. I guess all this is enough to consider for now. Remember, the above is only theory, not any practical experience. It may be too confining for what you want to do. In theory, though, I believe it does what you want. It eliminates the need for compaction entirely... it would use normal font files and just work with the data from there, giving you also a very predictable location for data for all characters, which would allow you a way to set up a table to use to determine absolute inter- characer spacing required for each character... perhaps 2 tables, one for space 'before' the character and one for number of actual pixels in the character. That seems like it would work quite nicely at a cost of no more than 512 bytes. (I like that idea. hehehe :) ------------ Category 5, Topic 20 Message 3 Sun Sep 11, 1994 CBM-ED [e.g.bell] (Forwarded) BTW, why ar you using only 91 characters? And how are you arriving at half-bytes (14pt-24.5 x 91 char = 2229.5)? ------------ Category 5, Topic 20 Message 4 Thu Sep 15, 1994 T.BROWN25 [Tim] at 21:59 EDT Thanks for replying e.g.bell, Believe me, since we first talked, I've taken your advice to heart and have been working on it. There are problems, however. (And just to get this out of the way, the spaghetti I was referring to was the BB message, not my code:) Aesthetics is the primary goal, then compactness. I'm trying to find a way to comply with your suggestions, but looking at the examples in the first post: To convert the 8-point to 10-point, bits 2 & 3 could be duplicated and then inserted at that position for each row to create the 10-point; rows 3 & 4 duplicated, also. (Counting bits R to L starting w/0) The same algorythm wouldn't work on the letter "B", however. The bit duplication is fine, but different rows. I'm beginning to see some light in the dark recesses of my mind after reading your suggestion of two tables; one for the data and perhaps one to indicate which rows/bits to manipulate. I do appreciate your suggestions, Ed. Like Johnny5, though, I need lots of input. I'm hoping that my post will draw several schemes that I might tweak. As to your question of my present algorythm, I've compacted the 20 bytes that define the 10-point "A" into 12.5 bytes: 00000000 00000011 11000011 11111100 11000011 00110000 11001111 11110011 11111100 11000011 00110000 11001100 00110000 Every ten bits is a row in the letter. The subroutine reads the first two bytes and stores them. The bits in the template are then ROL'ed twice. Bytes 2 & 3 are then stored, etc. After ten iterations, the stored data is then manipulated based upon the position it is to be plotted to, AND'ed with the appropriate value, and then plotted to the screen. Your other question was pertaining to 1/2 bytes. You're right. 20 into 12.5 should be 20 into 13. 91 characters is another space saver. Since this is a paint program, I left out the graphic characters from the character sets. Also, reverse characters are accomplished by EOR'ing the data rather than a seperate set of data. Thanks for taking the time to reply, Ed. You've been quite a help and I'm looking hard at you're solution. Hope to hear from you again. T.J. Brown ------------ Category 5, Topic 20 Message 5 Fri Sep 16, 1994 CBM-ED [e.g.bell] at 04:51 EDT TB> To convert the 8-point to 10-point, bits 2 & 3 could be TB> duplicated and then inserted at that position for each row to TB> create the 10-point; rows 3 & 4 duplicated, also. (Counting TB> bits R to L starting w/0) The same algorythm wouldn't work on TB> the letter "B", however. The bit duplication is fine, but TB> different rows. Then go back to a table lookup message for each of the characters... From your scheme, it seems to me that each table would take only 91 bytes, and each byte could be a bit mask of which bits get duplicated. That way you 1] have a map for the algorithm to use when duplicating the characters, and 2] a way to allow other fonts to be created... namely saving the TABLES instead of the characters.. see what I mean. Your program need not deal with character sets at all, just the bit map tables... except for the 1 8 point character set. I never figured my post would answer all of your questions, just jog the thought process. I thought later about the reversing of chars as far as my asking about only 91 chars/set. :) Tho what I'm suggesting here would allow all of them at a minimal addition of bytes, and nowhere near what you are currently using to do the same thing. Does this all seem to clarify anything? :) BTW, you would need 2 91 byte tables for each point size other than 8, which is still only like 1400 bytes or so for all point sizes for any font. ------------ 5 ? ************ Topic 22 Sat Oct 08, 1994 W.OVE at 15:02 EDT Sub: Boot sector Information I would like to know the format of the boot sector. 11 message(s) total. ************ ------------ Category 5, Topic 22 Message 1 Sat Oct 08, 1994 W.OVE at 15:07 EDT I know that the first 3 bytes are cbm. It is followed by the file name to be booted. If the file name is followed by machine code, is that code loaded into the drive or is it loaded into the computer. I am trying to make an auto boot disk not auto boot. but removing the boot sector keeps the program from loading. It seems to me I could write a program to get the program to boot up, But I am not sure what to do with the code following the file name. I have been unaable to find any information on this is the libs. thank you w. ove ------------ Category 5, Topic 22 Message 2 Sat Oct 08, 1994 CBM-ED [e.g.bell] at 17:01 EDT Bill: I'll have the information for you hopefully by tomorrow. The code is loaded into the compter, though, for that part of uour question. There is also a checksum or CRC calculated from the code in the boot sector. I have a very good reference I just have to locate it so I can give you the correct information, and I'm not going to have a chance today. BTW, have you tried to BOOT your program from the drive instead of loading it? i.e. boot "program",u<device> ------------ Category 5, Topic 22 Message 3 Sat Oct 08, 1994 CBM-ED [e.g.bell] at 18:12 EDT Bill: Here is hopefully enough information to help you solve your problem: Boot sector data: Bytes Function 0-2 CBM identification code 3-4 Memory address for following boot sectors 5 Configuration index for following boot sectors 6 Block counter for number of following boot sectors 7 to 1st 0 Boot message - null terminated next byte to 0 Name of program to load - null terminated Bytes following the filename null-terminator is any machine langauge code to be called by the boot process. Finally... "The bytes following the filename are interpreted as a machine language program and the boot routine passes control to this program. From this point on, the programmer is responsible for starting the program loaded, or for loading another program, or for branching to another of the boot blocks." The above information is quoted from the Abacus book '128 Internals', pages 188-190. There is more information there, but it seems the above will get you where you want to go. Let us know how you make out. ------------ Category 5, Topic 22 Message 4 Sun Oct 09, 1994 W.OVE at 11:28 EDT Thanjk you for your prompt reply. I have been able to modify the superbase v3 boot disk so that it can be used for storing other programs etc. without interfering with superbase loading. I have moved other programs to the disk and am able to load them and run them. The problem is that every time I reset or quit superbase it reloads itself. Superbase will not reload unless you reset the system or you simply type "boot" <return>. I was hoping that I could make so type of loader program that would provide the information that is currently in the boot sector and so eliminate the boot sector so that I would not be bothered with always removing the disk from the drive. The boot sector is set up exactlly in the manner that you indicaated. In the 1581 manual it talks of booting information being loaded into the drive when using a file named CBM of the type "user". I think I need to know where that machine code get loaded. Are the first two bytes of the code a load address? I am perhaps in way over my head. thank you w. ove ------------ Category 5, Topic 22 Message 5 Sun Oct 09, 1994 CBM-ED [e.g.bell] at 18:51 EDT Bill: WO> In the 1581 manual it talks of booting information being loaded WO> into the drive when using a file named CBM of the type WO> "user". I think I need to know where that machine code get WO> loaded. Are the first two bytes of the code a load address I believe that what you are talking about is someting totally unrelated to boot sectors. There is a file you can create for the 1581 that will auto execute but it is unrelated to specific software. It is a feature anyone can use, and in fact, someone even published a way to use such a file to actually move the boot sector to another area on the disk. This peculiar feature is used to boot CP/M disks, if I remember correctly, because they cannot give up the normal boot sector due to their directory location. I'm not saying that nothing is loaded into the drive by superbase. I don't know that. If it is, you probably are in over your head. I only say that because I've never done any drive RAM programming, and I feel personally I'd have to work pretty hard to get it done. The file you are talking about on 1581 disks will always be named 'COPYRIGHT CBM 86',and does not get tied to specific software unless they have made it do that. Is there such a file on your disk. If so, read the 1581 manual on page 87. It says the first 2 bytes contain the load address, the 3rd byte contains the length of the code/chars to follow, and finally a trailing checksum. It also says the load address is the start address for the code. Are you saying that when you just QUIT from superbase that it reboots itself?? WO> every time I reset or quit superbase it reloads itself You may have to disassemble the boot sector to do what you want to do. That is unless the boot sector is only booting the program and nothing more. Still, that is not such a daunting task, I wouldn't think. How many sectors could it be? ------------ Category 5, Topic 22 Message 6 Sun Oct 09, 1994 H.HERMAN1 at 22:43 EDT Here's one thing that you may want to try, but if you do, do it carefully: (1) Using a sector editor, block copy the boot sector from the SuperBase disk, onto another. (2) Then fill the boot sector on the SuperBase book with nulls. (3) Try running SuperBase from the newly changed disk. If it runs okey, you are home free. If not, then write back the boot sector which you earlier saved, back onto the SB disk. One note of caution: If this does work, I would still permanently keep in storage that boot sector, in case it turns out that some rarely used part of SB needs that code to properly run. I did something akin to this on my very heavilly copy-protected Pocket Filer 2 disk. This particular disk always sits in drive 8. RAMLink will usually autoboot "Main Menu", as it is supposed to. But at times, some programs exit by booting unit 8. Therefore, I wanted to change the boot sector to also boot "Main Menu", and not to boot Pocket Filer 2. The boot sector for Filer 2 is safely tucked away on a storage disk, and instead that sector was over-written with the custom boot. To my delight it has no effect on Filer2. However, even if it did, there is no harm done, as that sector could be resurrected. One more word of caution: anyone fooling around with ANY Pocket disk must be very careful, as only some sectors are "user accessible". Most of the Pocket disk cannot be accessed without destroying the disk, due to the copy- protection. So "BE CAREFUL"!!!! Howie ------------ Category 5, Topic 22 Message 7 Tue Oct 11, 1994 W.OVE at 19:33 EDT I have tried the following things without success. 1) I did fill boot sector with nulls. Program won't load. 2) I tried loading the machine code into computer using first two bytes as load address then sysing to it. Locked up computer. Loaded machine code into drive ram using first two bytes as load address Resulted in horrible racket will not be repeated. I would like to make another try. I need to know what the computer does when it boots a program after a cold start. You get a message booting.. Is that message generated by the computer or is that comming from the drive. I did notice on a memory map that an area is identified as boot sector work space. What goes in that space? I tried to dissamble the machine code and it seems to be gibberish. I started the dissambly from the beginning and again from other bytes. But nothing dissambled to anything more that a whole lot of question marks and seldom used instructions. ( not an lda in the lot). Thank your for your help and sugestions. I am not a programer. I merely hack to try and get thing the way I want them to be. w.ove ------------ Category 5, Topic 22 Message 8 Tue Oct 11, 1994 CBM-ED [e.g.bell] at 22:10 EDT Bill: WO> You get a message booting.. Is that message generated by the WO> computer or is that comming from the drive. That message is generated by the computer. I would have to do some research as far as your other questions go. I don't know how you tried to disassemble the machine code either... from the boot sector? If so, can you save it to a file and upload it? How big is it? When you do the disassembly, or I should say instead of doing a disassembly, use the 'm' command of the monitor to see if what you are looking at is text instead of code. WO> 2) I tried loading the machine code into computer using first WO> two bytes as load address then sysing to it. Locked up WO> computer I have found through experence that when a SYS locks up the computer it is very frequently because the proper MMU configuration is not in context. Try this again, but explicitly force the configuration with the BANK command. Try bank 15. What machine code did you load into the drive RAM anyway? From what I see of the info I have on $0b00 block, when an auto boot disk is detected, the sector 0 of track 1 is read into this area, and the 1st 3 bytes examined. If the CBM is detected, the boot process proceeds. "It would be possible to simulate a boot by filling the buffer with the proper values, then jumping into the BOOT_CALL routine at address 63737/$f8f9. Try booting the disk, but pressing STOP before it is complete. Then examine the contents of this area. Duplicate that, then proceed with the call (BANK 15) listed above). Perhaps that will help you emulate what the actual boot sector does. BTW, the book (Mapping the 128) says that there is a Kernal jump table entry for the boot routine at 65363/$ff53, which you should use if you do all of this. Try it first, and if it doesn't work try the other one. Let us know how you make out. egb ------------ Category 5, Topic 22 Message 9 Wed Oct 12, 1994 H.HERMAN1 at 06:32 EDT >I have tried the following things without success. 1) I did fill boot sector >with nulls. Program won't load.... When you did this, did you: dload"filename",u8 <-- or whatever drive? In other words, did you try to DLOAD, instead of BOOT? Howie ------------ Category 5, Topic 22 Message 10 Fri Oct 14, 1994 CBM-ED [e.g.bell] at 06:44 EDT Bill: Try bloading the file called 'sbloader' and tell me what happens. From the demo boot sector you sent me, I can tell several things. First, there seems to be only the one boot sector involved. There is no boot message, and no machine language follows the boot procedure, at least nothing valid. I even checked my list of undocumented opcodes and it does not seem that they are involved. One of the autobooting schemes that is used is to load programs in over the operating system vectors. After the load, then, when the system takes a normal jump through certain vectors, it is directed to code loaded into that area, which in turn would finish up the loading process. This may be what is happening here. You will know this is what is happening if you do the bload above and the full program loads and then auto-starts. If it doesn't auto-start, it may still be that all you have to do is SYS to the start address of this file once it is loaded. Give this a try and let me know how you make out. egb ------------ Category 5, Topic 22 Message 11 Sat Oct 15, 1994 W.OVE at 11:14 EDT I have tried all the sugestions and I really appreciate you taking the time to respond. the program "sbloader" is the same strange jumble of information that is contained in the boot sector. I have broke into the progrm during the boot process and can only report that I am very confused by what it is doing. However on the good news front I solved my problem of the program autobooting, by copying the boot sector the program "sbloader" and a couple other files to different disk and have erased them on the other disk. So now the program doesn't boot and I caan boot it off drive 9. The loader loades in from drive 9 and the it does a normal load of the program on drive 8. It is a left handed solution for sure, but it does what I was hoping. thank you for your assistssnce. w.ove ------------ 5 ?