INFO-ATARI16 Digest Fri, 6 Apr 90 Volume 90 : Issue 418 Today's Topics: Acc DTA Malloc AES Amiga/Atari help (2 msgs) Digitized Sound Player Available at Terminator (Even More Fun!) MIDI at 500KHz ... IT WORKS!!! NeoDesk problem No Sound? Sozobon fix for Gemini TEX/DVI... are they married? ---------------------------------------------------------------------- Date: 6 Apr 90 12:48:24 GMT From: uupsi!sunic!dkuug!imada!micro@nyu.edu (Claus Pedersen) Subject: Acc DTA Malloc AES Message-ID: <732@imada.dk> I hope that some of you could help me. I have 3 problems - no one really, and thats ACC's. No one can ever have written an ACC without having these problems - and I just don't want to reenvent the wheel... First two simple questions : 1) Does 'fsel_input' use 'Fsfirst', 'Fsnext' ? 2) Is there, way to find the size of a file - without using 'Fsfirst' ? If the answer is 'NO' to 1) or if 'fsel_input' has a private DTA, and 'YES' to 2), then please tell me how find the size of a file, without using low level comands (eg sector read). The reason is that I don't want my ACC break the current running program by using or setting the DTA. If the ACC uses the 'default' DTA, it will write in some 'undefined' memory which might be used by the running program (or an other ACC). In fact it will use the DTA from the current running program (Defaults to the comandline!!, in a real program). It shuld be real easy to do, as the basepage contains an pointer to the DTA (p_dta), which means that every process can have it's own DTA. As 'Fsetdta' and 'Fgetdta' are relying on the '*SYSHDR->_run' pointer, which points to the basepage of the active process, there should be no problems, BUT '*SYSHDR->_run' never points to the basepage of an ACC, not even when it is supposed to do it's initiation! The second problem - 'Malloc' - yes I guess you know, GEMDOS allocates memory to an ACC under the name (pointer to the basepage) of the current running process, which means that the allocated memory dies with the process. (If this is done to minimize fragmentation, I say it is a bit drastic!). Again it appears that all the features are there. As things are now, ACCs are merly fancy TSR program, and not real processes. If there was a way of getting '*SYSHDR->_run' to be updated correctly, it appears that the two problems would be fixed or ??? Also problems with 'Pexec'ing programs from an ACC, could probaly get fixed too (that is, if ACCs was real processes). As to Matthew Lodge's experiment with AC_CLOSE, it appears that EVERY? AES call automatically makes a call to the process switcher. Try to make an ACC which opens a window, and the AC_CLOSE looking something like this : case AC_CLOSE: Cconws("acc closed"); Cconin(); form_alert(1,"[1][|Closed][OK]"); break; Is there anybody who can help me getting the DTA and Malloc working. I would also like to hear about alternative ways to follow. I am not a developer, and have "The Hichhikers Guide ...", "The GEMDOS Manual", "Pexec cookbook", "Atari ST Profibuch" and "TC v1.1" but none of these fine documents could help me. I am using a Mega2 with TOS1.4. Thanks for the time Klaus Pedersen (micro@imada.dk). ------------------------------ Date: 7 Apr 90 00:45:13 GMT From: cunixf.cc.columbia.edu!cunixb.cc.columbia.edu!es1@rutgers.edu (Ethan Solomita) Subject: Amiga/Atari help Message-ID: <1990Apr7.004513.7141@cunixf.cc.columbia.edu> In article <39@solist.htsa.aha.nl> erwinh@solist.htsa.aha.nl (Chaos Conquerer) writes: >I can imagine that you think the Commodore Amiga is much better than the Atari ST. Fine with me, maybe you are right (could be). But please stay in comp.sys.amiga. People, please watch your cross-postings! One person posted a flame to both and now all replies are being sent everywhere. When you reply, delete the cross-posting! -- Ethan Ethan Solomita: es1@cunixb.cc.columbia.edu "If Commodore had to market sushi they'd call it `raw cold fish'" -- The Bandito, inevitably stolen from someone else ------------------------------ Date: 6 Apr 90 21:31:56 GMT From: ndsuvm1!mtus5!ldshaner@cunyvm.cuny.edu (Leon D. Shaner) Subject: Amiga/Atari help Message-ID: <90096.163156LDSHANER@MTUS5.BITNET> >If I wanted to do MIDI, I'd buy an Atari ST. Since I like what the Amiga >does for me, that's what I have. Considering the fact that a MIDI interface with 1 in, 1 thru and 3 outs can be built for $5-$10, the fact that an ST has one built in, really doesn't mean much...A SUN has a built-in sound digitizer, but considering the fact that it is only 8 bits at a max of 8KHz, it is of little use...Just how many in outs and thrus does a stock 520ST have? MIDI software is also of little worry here, since Dr. T's makes nearly every title they have for both machines...(my local computer dealer, who does not sell Amiga, told me that Dr. T's is about fed up with the ST because of piracy, and from what I hear, so are many other software vendors...) +--------------------------------+-----------------------------------+ | LEON D. SHANER :BBS: AMIGA BITSWAP CENTRAL DISPATCH: | EE UNDER GRAD. : (906)487-0006 <-AMIGA ONLY! : | MICHIGAN TECHNOLOGICAL UNIVERSITY : FREE! - 24 HOURS - 80 MB : +------------------------------------+-----------------------------------+ ------------------------------ Date: 7 Apr 90 01:05:57 GMT From: haven!wam!dmb@ames.arc.nasa.gov (David M. Baggett) Subject: Digitized Sound Player Available at Terminator (Even More Fun!) Message-ID: <1990Apr7.010557.23101@wam.umd.edu> I've posted a program to terminator and to comp.binaries that lets you play samples at their correct speeds (specifiable in Hz). It will also run from an AUTO fodler, giving those of you with Mac envy the ability to play a digitized sound at bootup. Look in ?atari/music/samples/ for play.arc. (terminator == 35.1.33.8) Dave Baggett dmb@cscwam.umd.edu ------------------------------ Date: 6 Apr 90 20:31:47 GMT From: att!oucsace!bwhite@ucbvax.Berkeley.EDU (Bill White) Subject: MIDI at 500KHz ... IT WORKS!!! Message-ID: <1266@oucsace.cs.OHIOU.EDU> In article <9004051235.AA00350@Argus.Stanford.EDU>, SA44@liverpool.ac.UK (Kevin Maguire) writes: > A keyboard interrupt/VBL interrupt will cause > the polling loop to drop bytes. 500000 bits per second = 50000 bytes > (1 start bit/8 data bits/1 stop bit) which means only 160 clock cycles > for the 68000 to get the byte, any delay will be disasterous. You can disable the interrupts quite easily; just mask off the appropriate bits on the MMU (I think that's the chip; it's one of those) and all interrupts will be ignored. Or you can re-vector them to short little routines that increment a counter and return. I'd suggest the former, because it probably doesn't matter whether the machine keeps track of, say, mouse movement while data is being transmitted. BTW, I've had a problem re-enabling interrupts, though; things keep crashing every now and then. -- | Bill White Internet: bwhite@oucsace.cs.ohiou.edu | | This .sig file temporarily under construction. | ------------------------------ Date: 6 Apr 90 23:04:21 GMT From: zephyr.ens.tek.com!orca.wv.tek.com!pogo!bluneski@uunet.uu.net (Bob Luneski) Subject: NeoDesk problem Message-ID: <8833@pogo.WV.TEK.COM> In article <1990Apr6.204104.14336@ucselx.sdsu.edu> coffey@ucselx.sdsu.edu (pat coffey) writes: > >I have been using Neodesk for about a year now and lately it has been >acting strangly. When I am looking for a file, and the file window >says that the mask is '*.*', some of the files don't show up. > >For example, I created a file in first word. When using the file >selector in first word, the file is there. When I exit from first >word and look at the directory, it is not there. If I then drop >out of neodesk, the file shows up again. Has anyone else experienced >this? Does anyone have any idea of what is causing it? I really >like Neodesk and would like to continue to use it. Neodesk will display up to the number of files you specify in the preferences menu. Enter the preferences and increase the number of files in a directory and your files will return. Don't allocate a lot more than you need, because it does take additional memoery. ____________________________________________________________________________ Bob Luneski Diamond Back Support Hotline: bluneski@pogo.WV.TEK.COM Genie: B.LUNESKI1 The opinions expressed herein are my own and in no way reflect the opinions of Tektronix, Inc. ------------------------------ Date: Fri, 06 Apr 90 13:31:17 CST From: Robert Gambs Subject: No Sound? Dear Atarians, First of all, I am getting a really wierd problem. Every once in a while when I turn my computer on and load either a term program or a WP program, at a random point in time after loading, the sound just dies!!! No keyclick, nobell, no digitized sound, NOTHING! Then, whats weirder, is when I turn the computer off, back on again and load something, the sound returns to normal. Has this ever happened to anyone else? Is something going out in my computer? Next: are there any english docs on Gemini. I have the Gemini.zoo which released about 27 files. I have done everything I can think of, but I can't get itto work. Perhaps someone could send me a short file describing how to get goingon it, I would appreciate it. If someone could tell me what it is, that would be great. What does it do to the desktop? etc.etc.etc. Finally, a NeXT will be here on campus in about a week for a demo and I know nothing about it. I would like to know how it stands up to a TT (?) and how the two are similar and different. We already have a price quoted for NeXT: $13,000+ However, this brings a very nice system (all you could want, right?) to your desk. ************************************************************************* | | | ___________________________________________ | | | / / ROBERT GAMBS | | | /______ / S54394GR@ETSUACAD.BITNET / | \ / / EAST TEXAS STATE UNIVERSITY / | \ ________/ / ************************************************************************* ------------------------------ Date: 7 Apr 90 00:26:51 GMT From: uupsi!ncs.dnd.ca!balkwill@nyu.edu (R. J. Balkwill) Subject: Sozobon fix for Gemini Message-ID: <773@ncs.dnd.ca> To all who have problems running Sozobon-built programs under Gemini. Do not despair. The real problem is in jas, the Sozobon assembler. My memory is a little shaky here but after more than a few minutes investigation I discovered the following: 1. Gemini uses the xArg convention for passing arguments to programs and Sozobon startup code honours this. Gulam and most other shells do not make an xArg entry in the environment so certain portions of your startup code are never executed when calling from those. 2. The section that deals with xArg passing invokes either memcpy or lmemcpy from dlibs. 3. Although the source for these functions looks ok the jas assembler generates erroneous code for some bit-oriented instruction (btst?) therein. Hence your startup code walks into trash of jas' making. The Solution - either reassemble the memcpy.s and lmemcpy.s functions with a compatible assembler, or write tiny versions of your own either in assembler or C, compile them and add them to dlibs replacing the old versions. PS - I love both Gemini and Sozobon - add in Gulam and life is fun! --- Bob -- --- Bob ------------------------------ Date: 7 Apr 90 00:37:03 GMT From: cs.utexas.edu!uwm.edu!ux1.cso.uiuc.edu!ux1.cso.uiuc.edu!cs325ec@tut.cis.ohio-st ate.edu Subject: TEX/DVI... are they married? Message-ID: <16000040@ux1.cso.uiuc.edu> Can someone tell me the relationship between tex and dvi? Do I need both to print a tex document? Does tex drive an LQ printer? Thanks. -- Greg ------------------------------ End of INFO-ATARI16 Digest V90 Issue #418 *****************************************