857 lines
36 KiB
Plaintext
857 lines
36 KiB
Plaintext
AppleLink
|
|
APPLE II DEVELOPMENT FORUM CONFERENCE LOG
|
|
January 3, 1989 10:00 pm Eastern
|
|
Topic: An open discussion
|
|
Forum Leader: James Luther (AFL Jim)
|
|
|
|
|
|
AFL Jim Hi Mike!
|
|
Mike Lutz Hi Jim!!
|
|
AFL Jim Howdy, Gary
|
|
AFL Jim Like the new buttons?
|
|
AFL Jim :)
|
|
AFA Gary J Hi, Jim, Mike
|
|
AFA Gary J YES!
|
|
AFL Jim We have the power!
|
|
AFA Gary J :)
|
|
Mike Lutz You can do some really weird things with my software and
|
|
keywords....
|
|
AFL Jim Mike: we can finally release files real time now!
|
|
AFL Jim Like what, Mike?
|
|
Mike Lutz Hu??
|
|
Mike Lutz I can do two things at once... It is weird...
|
|
AFL Jim What wierd things?
|
|
AFA Gary J hehe...I thought that banner said, "Learn the best software
|
|
BUGS..." at first
|
|
AFA Gary J :)
|
|
AFL Jim What software?
|
|
Mike Lutz Ok....I'll run you though....
|
|
AFA Parik Hello all.
|
|
AFA Gary J Hi Parik.
|
|
AFL Jim Hi Parik
|
|
Mike Lutz I logon, do an apple - k for a key word and do adv...
|
|
AFA Parik I feel happy, I have software library power now. :)
|
|
AFA Gary J I enjoyed your upload.
|
|
AFA Gary J YES! Power!
|
|
AFA Gary J :)
|
|
AFA Parik I can validate files... yippeee...:)
|
|
AFA Parik Thanks Gary.
|
|
AFL Jim I enjoyed your source code :) (ga Mike, I'm listening)
|
|
AFA Parik Hi Scott
|
|
AFL Jim hi Scott
|
|
AFA Gary J Yes, the source code is very interesting, Parik.
|
|
AFA Gary J Hi Scott
|
|
ScottG25 Hi Parik, Jim, Gary, Mike!
|
|
Mike Lutz then I get the menu bar and scute over to the post offist...
|
|
Mike Lutz and pick read mail
|
|
AFL Jim and..??
|
|
Mike Lutz But it (as in alink) pulls down the ADV menu under the post offise
|
|
menu...
|
|
AFL Jim Hi Steve
|
|
AFA Parik HI STEVE
|
|
SEGlass Good Evening.
|
|
AFA Parik (damn cap locks key :)
|
|
AFA Gary J Hi Steve
|
|
Mike Lutz so then I pick enter chat or what ever it is...
|
|
Mike Lutz then my mail widow pops up...
|
|
AFL Jim Hi John, Coach
|
|
Coach101 Good evening Jim
|
|
AFA Gary J Hi Coach, John
|
|
AFL Jim IIGS or IIe/IIc software, Mike?
|
|
Mike Lutz but before I can read it I get put in here.....weird hu :|
|
|
AFA Parik Hi Coach & John!
|
|
AFA John Evening everyone!
|
|
Mike Lutz so here I am..
|
|
Mike Lutz IIe/c
|
|
AFL Jim does it happen every time?
|
|
AFA Parik Hi Dave
|
|
Dave Lyons howdy
|
|
AFA Gary J Maybe this software bug is just a good thing for the Development
|
|
Forum, Mike?
|
|
AFA Gary J :)
|
|
AFL Jim Hi Dave
|
|
AFA Gary J Hi Dave
|
|
AFA John Hello, Dave
|
|
Mike Lutz Hiya Dave!!
|
|
AFA Parik A mail window shouldn't pop up, should it?
|
|
Dave Lyons Howdy, all....
|
|
AFA Parik Hi John!
|
|
Mike Lutz Nope it happens with every keyword...
|
|
AFA John Hi Parik.
|
|
THE GIBBER Yep....A slow Node to Alink :)
|
|
AFL Jim Sounds like your disk is glitched, Mike.
|
|
AFA Parik Other John, John. But hi anyways. :)
|
|
THE GIBBER I hate delays in the net.... I have to think about what I am going
|
|
to say :)
|
|
Mike Lutz if I pick something in the ADV menu right when I get here I just
|
|
get another..
|
|
Mike Lutz copy if the menu...
|
|
THE GIBBER Hi All.....How is everything going in Development Land????
|
|
AFA Parik That happens to me too Mike, you've got to let the software finish
|
|
what its
|
|
AFA Parik doing currently. When you hit ADV, just let it put the menu on
|
|
the screen
|
|
AFA Parik before using up/down arrows.
|
|
Mike Lutz I don't know ... Alink is doing a lot of the work...
|
|
Dave Lyons Mike, I've had that happen to me, too. There _are_ some weird
|
|
things that
|
|
Dave Lyons can happen w/ ALPE & the timing of windows appearing.
|
|
AFA Parik (hehe, ever get Flashnotes in the middle of a window? :)
|
|
AFL Jim I always hit open-apple-period to kill the menus as they drop.
|
|
Mike Lutz I'll talk to you voice later dave I'll tell you it then...
|
|
AFA Gary J I wish there was a mouse location that would be the same as
|
|
hitting open-apple
|
|
AFA Gary J period, so I could stay off of the keys.
|
|
AFA Parik Hi Mark
|
|
Mark1000 Hiya Parik
|
|
AFL Jim Parik, I looked thru your source code quickly for UNPACK. It
|
|
doesn't handle
|
|
AFL Jim anything but the MAIN information block. Right?
|
|
AFA Parik Right.
|
|
AFA Parik It just works with the graphics screen, not SCBs or such. It
|
|
always unpacks to
|
|
AFA Parik a $C1 file.
|
|
AFA Parik Hi Tim, Marty, Bruce, and Matt
|
|
THE GIBBER Hi Marty!!!
|
|
AFA Gary J That would be the only way it could work, then. (Being $C1
|
|
only)
|
|
AFL Marty Hi all
|
|
AFA John Hi Marty, Matt, Tim, Bruce, Jack and Mark.
|
|
AFA Parik (It DOES unpack the SCBs and such, it just doesn't unpack SCB type
|
|
blocks that
|
|
AFA Parik doesn't contain pictures)
|
|
AFA Parik Hi Deadhead.
|
|
DeadHead13 Hi all
|
|
AFL Jack hiya :)
|
|
DeadHead13 What's the current talk about?
|
|
AFA Parik Hiya Sarge!
|
|
AFL Jim Nothing much yet, Deadhead. We were talking about the source code
|
|
Parik
|
|
AFL Jim uploaded for unpacking Apple Preferred format pictures.
|
|
DeadHead13 Mind if I jump in
|
|
AFA Parik Lot of files released today, eh? Wonder why...:)
|
|
AFL Jim GA
|
|
AFA Gary J :)
|
|
AFA Gary J Yeah, Parik!
|
|
THE GIBBER Which files parik
|
|
AFA Parik Hi Kathy
|
|
DeadHead13 O.K. I develop for a firm using UCSD (ugh...), and I recently got
|
|
my hands
|
|
AFA Kathy Hi!
|
|
AFA Parik John, files all across AppleLink are being unbonded from their
|
|
hidden status
|
|
AFA Parik as we talk.
|
|
DeadHead13 on Kyan pascal, BUT...
|
|
AFL Jim I case anyone wonders, we finally got the online tools to release
|
|
files anytime
|
|
AFL Jim during the day or night.
|
|
DeadHead13 but the but is that I cannot reach them in any way to get the
|
|
extensions,
|
|
Dave Lyons Keen!
|
|
AFA Parik Hi Jono & Mike!
|
|
DeadHead13 latest version, or know even if it is worth my while to deal with
|
|
it.
|
|
JonoS Hi Parik..
|
|
THE GIBBER Ohhhhhhhh.......that is great..... Now I have DL them....:(
|
|
Dave Lyons I don't think anybody's heard a peep out of Kyan lately....
|
|
AFL Jim I haven't been able to get Kyan to respond in months either,
|
|
DeadHead.
|
|
DeadHead13 Anyone know about them?
|
|
THE GIBBER All these files and so little time to Down load ;)
|
|
AFA Parik Hello Mark.
|
|
MarkHulme1 Hiya
|
|
AFL Jim Last I knew, they were going to pay for articles printed in their
|
|
newsletter,
|
|
AFA Parik Hello Omnitreant and Oli.
|
|
DeadHead13 I was thinking of maybe sending them a registerd letter, because
|
|
they haven't
|
|
AFL Jim they printed a couple of mine and then disappeared. They have an
|
|
answering
|
|
DeadHead13 responded to my other letters.
|
|
AFL Jim machine that catches the phone, but they never call back.
|
|
AFL Jim (they = Kyan)
|
|
DeadHead13 Should we get the Better Business Bureau after them??
|
|
AFA Kathy brb
|
|
DeadHead13 This behavior is dispicable
|
|
AFA Parik Hi Floyd & Paul
|
|
PGauthier Good Evening.
|
|
AFL Floyd Hi, all.
|
|
THE GIBBER Hi Floyd!!! What is up??
|
|
AFA Parik (bet Floyd is sick of the number $0046 :)
|
|
DeadHead13 brb
|
|
AFL Floyd Not much. Still trying to recover from New Years! ;)
|
|
AFL Jim DeadHead, which toolkits from Kyan are you looking for? I can
|
|
probably tell
|
|
AFL Jim you if the search is worth your time or not (I have most of them
|
|
if not all).
|
|
THE GIBBER <------Still recovering from New Years :(
|
|
THE GIBBER <-------3 day of Hangover and still counting :(
|
|
DeadHead13 Ok, I realy don't know -- I want to develop in double hi res with
|
|
full control
|
|
DeadHead13 of the serial port
|
|
AFL Floyd (Parik: You got that right! I'm in the process of doing my own
|
|
documentation for ACU to replace...
|
|
AFL Floyd what Quantum has up)
|
|
JonoS Floyd: I was wondering why ACU was created when it packs only a
|
|
few blocks..
|
|
AFA Parik heh
|
|
AFA Gary J Re-hi, Kathy
|
|
JonoS smaller than BLU.
|
|
AFL Jim Kyan's MouseText and Graphics toolkits both eat so much memory
|
|
that you don't
|
|
AFA Kathy :)
|
|
AFL TimB Floyd, you're typing seems slower too! You really are recovering
|
|
from New...
|
|
AFL TimB Years!
|
|
AFL TimB you're=your
|
|
AFL Floyd Jono: ACU doesn't use the Binary II file format. It uses it's own
|
|
more flexible one. AppleLink
|
|
AFL Jim have enough left to write any kind of application. To be exact,
|
|
you have 8.25K
|
|
Omnitreant Hi everyone. Anybody here familiar with programmin the SHR
|
|
screen?
|
|
DeadHead13 AS far as double-hi res, I don't see it taking much more since
|
|
they don't
|
|
AFL Floyd doesn't require Binary II since it maintains all file
|
|
attributes.
|
|
AFA Parik I've programmed it quite a bit Omnitreant.
|
|
DeadHead13 support the upper bank anyway
|
|
AFL Jim left after the MouseText kit and the runtime library are loaded.
|
|
AFL Floyd (TimB: Flashnote city! ;)
|
|
DeadHead13 That is ridiculous
|
|
AFA Parik (protocol, anyone? :)
|
|
AFA Gary J Q = DeadHead, Omnitreant
|
|
AFL TimB Floyd: %)
|
|
ScottG25 ?
|
|
DeadHead13 Gary, what is Omnitreant?
|
|
THE GIBBER Hi Dennis!!!
|
|
THE GIBBER *
|
|
Omnitreant Parik: We're trying to do the absolute fastest
|
|
copy-buffer-to-screen routine,
|
|
Omnitreant and were wondering if LDA,STA pairs or MVN would be fastest...
|
|
AFL Jim Hi Dennis
|
|
Omnitreant I am an Omnitreant, that's who!
|
|
AFA Parik Using LDA with direct page is the fastest Omnitreant.
|
|
AFA Gary J :)
|
|
DeadHead13 Oh, sorry
|
|
Matt DTS LDA/STA pairs are faster, Omnitreant.
|
|
Omnitreant But doesn't one pair take about 12 cycles compared to 7*n for
|
|
mvn?
|
|
JonoS <-- Got dumped.
|
|
CecilFret What's happening?
|
|
AFA Gary J Hi Cecil
|
|
CecilFret Hey, Doms. Cleaning up my phone book. Last digit of your home
|
|
phone = 6??
|
|
AFA Parik Omnitreant, LDA [0],y takes 7-8 clock cycles.
|
|
Omnitreant Actually, tonite WE are Omnitreant...my friend is here on the
|
|
other keyboard
|
|
Omnitreant Yes, but you need a store too< the friendo
|
|
SEGlass The fastest way to move depends on where you are going.
|
|
DeadHead13 Gary... could we discuss my problem
|
|
AFA Parik (holding off till Deadheads problem is finished)
|
|
AFL Floyd What about setting the screen area as the stack and using PHA?
|
|
THE GIBBER Hi Cecil
|
|
CecilFret Nice to "see" you again gibber.
|
|
Omnitreant That sounds interesting Floyd...
|
|
DeadHead13 I don't thinkl it is that big a deal, but I realy NEED a real
|
|
development
|
|
DeadHead13 system for the //e that IS NOT assembler ... I don't mind but my
|
|
job wants
|
|
DeadHead13 something that is easy to maintain. I personally find aztec the
|
|
pits, and have
|
|
THE GIBBER Nice to see you.... I am still working on that problem
|
|
ScottG25 That would only work for pure word transfers, tho, you would waste
|
|
time if you
|
|
ScottG25 had a mask, no?
|
|
DeadHead13 been using UCSD. I recently purchased Kyan to see if I could use
|
|
it, but I am
|
|
DennisDoms (Where did you find Kyan?)
|
|
DeadHead13 unable to contact them. Therefore I was wondering if any of you
|
|
would help me
|
|
DeadHead13 with Kyan??? :)
|
|
DeadHead13 Believe it or not, off the dusty shelves of the APDA store (they
|
|
had one in the
|
|
AFL Jim Deadhead, as I said before, I've tried to contact Kyan as have
|
|
others and they
|
|
DeadHead13 store on the shelf and sent it to me!)
|
|
AFL Jim don't answer.
|
|
AFA Gary J Q = DeadHead
|
|
ScottG25 ?
|
|
AFA Gary J Q = DeadHead, Scott
|
|
AFA Gary J (Sorry Scott!)
|
|
AFA Parik (guess what folks... this room is full!)
|
|
ScottG25 <No Problem!:)>
|
|
AFL Jim If I lived in San Francisco, I'd go beat on their door, but I'm in
|
|
Kansas City
|
|
AFL Jim so....
|
|
AFL Oli (No Parik... limit here is 50
|
|
AFL Floyd DeadHead: In my opinion the ONLY development system for a //e is
|
|
assembler. Sorry.
|
|
AFL Floyd Aren't these 50 person rooms?
|
|
AFA John (No -- 50 people is the limit Parik. :)
|
|
AFA Parik (Try getting in Oli. My other account just did. I'm in Developer
|
|
Chat A)
|
|
JonoS <-- Lives half a mile from Kyan. But never bought anything from
|
|
them.
|
|
DeadHead13 Well, I guess I'm pretty much done. If anyone wants to help me
|
|
outide of the
|
|
AFA Gary J Q = Deadhead, Scott
|
|
DeadHead13 forumn, thanks. Jonos... could you possibly see if they are still
|
|
alive??
|
|
AFL Jim Dennis, waht to give us you list of other high level languages for
|
|
the //e?
|
|
AFL Floyd <Oops, are we using protocol?>
|
|
DeadHead13 Or if they have a local phone beside the one they advertise I
|
|
could actually
|
|
JonoS If you give me the address # I'll check tommorow...
|
|
DennisDoms I came late...what has been covered?
|
|
DeadHead13 reach them at and leave me Email?
|
|
Omnitreant ?
|
|
AFA Gary J Q = Deadhead, Scott, Omnitreant
|
|
DeadHead13 I'll Email the address to you, thanks all. I agree with you Floyd,
|
|
but the boss
|
|
AFL Jim Deadhead is looking for a high level language for the //e.
|
|
DeadHead13 is the boss. All done, next question...scott
|
|
ScottG25 The question Omnitreant asked brings up another interesting thing,
|
|
how _does_
|
|
AFL Jim OK, Scott GA
|
|
DennisDoms Orhaned - ZBASIC, WSM C, Kyan Pascal, ???.
|
|
ScottG25 one calculate the total time in cycles an instruction executing in
|
|
fast memory
|
|
AFA Parik Hi Stowell.
|
|
AFA Gary J Q = Scott, Omnitreant
|
|
ScottG25 takes when the operand is in slow ram? I was thinking that it is
|
|
the same
|
|
DennisDoms (orhaned=orphaned) Only "major" ones I know of are Manx/Aztec C
|
|
and PROMAL...
|
|
SEGlass !
|
|
AFA Parik Hi Don.
|
|
ScottG25 cycle count, but the time is different... I am doing the same
|
|
thing as OT
|
|
ScottG25 is and would like my routine to be the fastest possible, too!
|
|
ScottG25 Done
|
|
AFL Jim GA Steve
|
|
AFA Gary J Q = Scott (Steve), Omnitreant
|
|
DennisDoms oops, Micol BASIC..I'd have to agree with Floyd that assembler
|
|
seems to end..
|
|
DennisDoms up being more viable in the long run.
|
|
SEGlass Code running in slow ram, runs slow (at 1megh) no matter what
|
|
addresses
|
|
SEGlass it accesses.
|
|
AFL Jim (thanks Dennis :)
|
|
SEGlass The number of cycles is always the same.
|
|
AFA Parik ?
|
|
SEGlass The length of the cycles just differs.
|
|
SEGlass ga
|
|
AFA Gary J Q = Scott (Steve), Omnitreant, Parik
|
|
ScottG25 To limit time, then limit references to screen memory, no?
|
|
AFL Jim So, leave your code in fast RAM, Scott.
|
|
SEGlass To limit time (ie go fast) run code in fast RAM and access slow
|
|
RAM
|
|
ScottG25 It IS in fast ram, Jim..
|
|
SEGlass as infrequently as possible.
|
|
SEGlass The most common accesses to slow RAM are writes to the screen.
|
|
SEGlass If we are talking text screen, I don't think counting cycles will
|
|
make
|
|
SEGlass a difference. But in Graphics, there can be a lot of accesses.
|
|
SEGlass For example to draw a dot on the screen you have to read the
|
|
screen and write
|
|
SEGlass the screen, both accesses go slow for some of the read and write
|
|
cycles.
|
|
SEGlass The number of slow cycles depends on the addressing mode you
|
|
use.
|
|
SEGlass I don't remember all the differences off hand, but the more direct
|
|
the mode
|
|
SEGlass the faster it is. GA
|
|
ScottG25 so then this would save time, no? Lda [Screen],y Eor
|
|
[Offscreen],y BEQ Skip
|
|
AFA Gary J Q = Scott (Steve), Omnitreant, Parik
|
|
SEGlass As apposed to storing anyway?
|
|
ScottG25 Skip terminates all Screen memory references in this time through
|
|
an indexed
|
|
ScottG25 loop... No?
|
|
SEGlass I'm still confused. What are you doing?
|
|
ScottG25 Yes, Steve, as opposed to storing.... I'm moving contents of an
|
|
offscreen port
|
|
ScottG25 to the screen, through a mask... and only writing words to the
|
|
screen that are
|
|
ScottG25 different... GA
|
|
SEGlass Is the mask usually FFFF?
|
|
SEGlass Or is it only rarely FFFF.
|
|
AFL Jim Scott, if a lot of the contents are different, then your code will
|
|
run slower.
|
|
ScottG25 Yes in areas where it is ok to store... in areas that are masked
|
|
it is zeros...
|
|
ScottG25 I know... but not much slower... I think...
|
|
AFA Gary J Q = Scott (Steve), Omnitreant, Parik
|
|
SEGlass What I mean to say is that if you can test the mask for FFFF
|
|
quickly, you can
|
|
SEGlass avoid reading the screen to begin with.
|
|
ScottG25 And [Mask]y
|
|
SEGlass Just store with the mask is FFFF. Avoiding the read saves time.
|
|
SEGlass (That should read: Just store when the mask is FFFF.
|
|
Matt DTS Even just storing when the mask is non-zero will save a lot of
|
|
time by...
|
|
Matt DTS ...eliminating the write to slow RAM when you won't be changing
|
|
the value there
|
|
ScottG25 That's what I do, only the EOR [Screen],y is testing for
|
|
difference in the
|
|
ScottG25 screen and offscreen buffers, This serves as kind of a look ahead
|
|
function
|
|
DeadHead13 Bye Y'all, thanks for all the help (I'm not from down south!)
|
|
DeadHead13 :)
|
|
AFA Parik :)
|
|
AFA Gary J Q = Scott (Steve), Omnitreant, Parik
|
|
SEGlass You can copy data through a mask to a destination in five
|
|
instructions.
|
|
AFA Parik All these ANDs, ORs, they need a WHY opcode.
|
|
SEGlass Adding any testing usually slows you down. The big win comes if
|
|
you can
|
|
SEGlass decide if a whole line will not be masked.
|
|
SEGlass This happens in QuickDraw a lot. When the clip and vis regions
|
|
are rects,
|
|
SEGlass the drawing is much faster than when they are not.
|
|
AFA Kathy Day job calls. :( Byeeeeeeeeeeeeeeeeeeeeeeeeeeeee.... :<<<<<
|
|
AFA Gary J Bye, Kathy
|
|
ScottG25 Ok that makes sense.... my loop is seven instructions, including
|
|
the BEQ, I can
|
|
ScottG25 therefore trim the BEQ and the LDA [Screen],y, right?
|
|
SEGlass My thinking is a little fuzzy. I think I need to see the whole
|
|
thing...
|
|
AFA Gary J Q = Scott (Steve), Omnitreant, Parik
|
|
SEGlass Let me show you the five instructions: lda src, eor dest, and
|
|
mask, eor dest,
|
|
SEGlass and sta dest.
|
|
SEGlass This copies src to dest through mask.
|
|
SEGlass If you do any testing in here you slow down the case where you
|
|
must copy
|
|
SEGlass through the mask.
|
|
ScottG25 Same as mine except the first eor... !
|
|
ScottG25 My eor [Screen],y is at the top of the five instructions you just
|
|
mentioned..
|
|
SEGlass If enough words have a mask of zero, you might save time by
|
|
checking it first,
|
|
ScottG25 then there is the beq to the dey dey bpl loop part!
|
|
SEGlass but if most have to draw anyway, I think this is as fast as you
|
|
can get.
|
|
AFA Gary J Q = Scott (Steve), Omnitreant, Parik
|
|
SEGlass A better way to speed it up is to unwind the loop. Its a clear
|
|
win every time.
|
|
SEGlass Just takes code space.
|
|
SEGlass GA
|
|
AFL Jim Scott, Do you have it now?
|
|
ScottG25 I'll post a facsimile in the quickdraw folder after the chat!
|
|
ScottG25 Yes! I was thinking to unwind the loop, too...! Thanks, Steve!!
|
|
GA
|
|
AFL Jim GA Omnitreant with your question.
|
|
Omnitreant The Goal: Fastest Screen Refresh!
|
|
AFA Gary J Q = Omnitreant, Parik
|
|
Omnitreant Under what conditions are a LDA/STA faster than MVN? We though
|
|
MVN was faster.
|
|
SEGlass !
|
|
Omnitreant !?
|
|
AFA Gary J Q = Omnitreant (Steve), Parik
|
|
Omnitreant (did I lay an egg???)
|
|
AFL Jim GA Steve
|
|
AFL Jim ! is for a comment, Omni
|
|
SEGlass The MVN tends to be faster than unwound loops unless the slow
|
|
cycles
|
|
SEGlass get in the way.
|
|
SEGlass For example an mvn from screen memory to screen memory is 7 cycles
|
|
per
|
|
SEGlass byte but 4 of the cycles are slow.
|
|
SEGlass An unwound loop of lda[],y sta [],y is 8 cycles per byte but only
|
|
2 are
|
|
SEGlass slow.
|
|
Omnitreant We are copying a screen image from bank 07 to E1 direct. Should
|
|
we go with
|
|
SEGlass Depending on your loop overhead, I think you can go faster without
|
|
the mvn.
|
|
SEGlass GA
|
|
Omnitreant the unwound pair or the mvn?
|
|
AFA Parik !
|
|
SEGlass Are you copying a full screen?
|
|
Omnitreant Ga
|
|
Matt DTS !
|
|
Dave Lyons [Eek! You shouldn't _know_ your image is in bank 7. Use the
|
|
memory mgr!]
|
|
Omnitreant Let's say half a screen.
|
|
AFA Parik :) dave
|
|
Matt DTS [dave zapped me again. Withdraw the !]
|
|
SEGlass Is it always in the same place? or can it move (code is easier if
|
|
you
|
|
AFA Gary J Q = Omnitreant (Steve, Parik), Parik
|
|
SEGlass can hard code it)
|
|
Omnitreant Well, we put the image there (heh!). APW crashes merrily!
|
|
SEGlass Did I really say that?
|
|
AFA Gary J :)
|
|
AFA Parik OmniTreant, I did the same type thing, and I find the LDA/STA
|
|
combo faster. I
|
|
SEGlass I mean hard coding the part of the screen to refresh.
|
|
Omnitreant We are hard coding it. First, we had inline mvns that did the
|
|
move. It was
|
|
Omnitreant pretty smooth.
|
|
AFA Parik use Y as the offset for the source in the direct page (LDA [0],y)
|
|
and X for the
|
|
AFA Parik destination offset (STA $E12000,x)
|
|
SEGlass There are ways to be faster. First don't use sta []y or even sta
|
|
$e12000,x.
|
|
Omnitreant For a scrolling game, this kind of refresh might not look good for
|
|
side-side
|
|
Omnitreant movement.
|
|
SEGlass set the data bank to $E1 and use 2000,x addressing.
|
|
Omnitreant Wow! Sounds like a good idea!
|
|
SEGlass Or if you want to get really tricky, you can reserve $2000 to
|
|
$9FFF in bank 1
|
|
AFA Gary J Q = Omnitreant (Steve), Parik
|
|
SEGlass and try the following: turn shadowing on for the super hi res
|
|
screen.
|
|
AFL Jim Let the hardware move the bytes for you from fast to slow...
|
|
SEGlass Turn off interrupts and Set the aux-mem switch. so that your stack
|
|
is now
|
|
AFA Parik Steve stole my question...:)
|
|
AFA Gary J (I was thinking the same question, Parik! :)
|
|
SEGlass really bank 1. Finally, set the dbr to your source bank and lda
|
|
xxxx,x pha in
|
|
SEGlass an unwound loop. It can only be faster by compiling the image
|
|
into pea XXXX.
|
|
SEGlass and then runing the compiled code. GA
|
|
Omnitreant (awe)
|
|
Matt DTS !
|
|
Omnitreant GA Matt
|
|
AFA Gary J Q = Omnitreant (Matt), Parik?
|
|
Matt DTS (wow a new forum leader) Be SURE you turn off interrupts before
|
|
doing that, if
|
|
AFA Parik Yes (keep me)
|
|
AFA Gary J Ok! :)
|
|
Matt DTS you're going to. An OS call or other heartbeat or interrupt
|
|
handler can really
|
|
Matt DTS muck things up in a hurry. (this is obviously not for standard
|
|
applications)
|
|
Matt DTS GA.
|
|
SEGlass I think the tech support sample code gives an example of all
|
|
this.
|
|
SEGlass If it doesn't it should (or will in the future).
|
|
DwightW1 WHICH ONE?
|
|
SEGlass Matt will hit me now.
|
|
AFL Jim hehehe
|
|
AFA Gary J Q = Omnitreant (Steve/Matt), Parik
|
|
Matt DTS I don't think it does currently, but it could in the future (when
|
|
we dig out).
|
|
Omnitreant Thanks. You've all given us stuff to think about! One more quick
|
|
one...
|
|
Matt DTS I'll throw fruitcake; I really will.
|
|
Omnitreant Never mind. I'm done. Bye
|
|
AFL Jim Should I bring a big shovel, Matt?
|
|
AFL Jim OK Parik... your turn.
|
|
AFA Parik I'm doing what steve said basically (used bank $01) except I
|
|
didn't think about
|
|
AFA Gary J Q = Parik
|
|
AFA Parik using the stack! Thanks Steve! :) My question. Is there
|
|
anything I will
|
|
Matt DTS I'll have you know my mother's fruitcake got rave reviews from
|
|
the...
|
|
AFA Parik stomp on in bank $01? For those who will use this technique, you
|
|
need to set
|
|
Matt DTS ...IIgs System Software group.
|
|
AFA Parik a ORG in your APW/Orca source code to somewhere, otherwise your
|
|
program will
|
|
Matt DTS !
|
|
AFA Parik probably get stomped on by your scrolling (using bank $01 with
|
|
shadow on)
|
|
SEGlass Hey, I have to leave and terrorize some children. See you all
|
|
soon.
|
|
AFA Gary J Q = Parik (Matt)
|
|
AFA Gary J Bye Steve
|
|
DougDavies Hi Dave
|
|
AFA Parik bye steve
|
|
AFA Parik GA
|
|
AFL Jim Bye Steve
|
|
AFL Jim GA Matt
|
|
ScottG25 Bye and Thanks!
|
|
Matt DTS That makes it my turn. If you're going to shadow into bank 1, you
|
|
don't...
|
|
Dave Lyons Howdy, Doug.
|
|
Matt DTS ...need to ORG the code. You just try to NewHandle for $01/2000 -
|
|
$01/9FFF...
|
|
Dave Lyons !
|
|
Matt DTS (what, did I type the wrong address? go ahead, dave <sheesh>0
|
|
AFA Gary J Q = Parik (Matt, Dave)
|
|
AFL Jim Go on, Dave
|
|
AFA Parik Will it move my program though Matt?
|
|
DougDavies What's the subject tonight? I have a good one if anyone cares to
|
|
chat about it
|
|
Dave Lyons And if NewHandle gives you an error, DON'T go ahead and use the
|
|
memory anyway
|
|
Dave Lyons the way TETRIS does!
|
|
Matt DTS (oh, that's benevolent enough) He's right. The problem you might
|
|
hit, Parik..
|
|
AFL Jim Is that why Tetris crashes GS/OS, Dave?
|
|
Dave Lyons [Gee, why do people always think I'm going to correct them?
|
|
Sheesh... :-) ]
|
|
Matt DTS ...is that your program might have been loaded there. OMF 2.0 has
|
|
a bit you...
|
|
AFA Parik :)
|
|
Matt DTS ...can set to specify that the loader not load your application in
|
|
Special...
|
|
AFA Parik Thats the problem i hit every time Matt. Orca loads my program
|
|
into $01. :)
|
|
AFA Gary J Q = Parik (Matt)
|
|
Dave Lyons Generally the Tetris crashes I've seen are when you Quit
|
|
Tetris--yes, that's
|
|
Dave Lyons what's happening then.
|
|
Matt DTS ...memory (bank 0 or 1). Also, you could make an initial segment
|
|
that...
|
|
Matt DTS ...loads the rest of the program not in special memory, and then
|
|
disposes...
|
|
Matt DTS ...of the initial segment (which might have loaded in bank 1).
|
|
The rest of the
|
|
Matt DTS program should then be able to NewHandle the shadow memory with
|
|
few problems.
|
|
Matt DTS (whew!) GA
|
|
AFA Parik aha. Thanks. :)
|
|
AFA Parik FEW problems, eh? ;)
|
|
Dave Lyons 0 is "few", isn't it?
|
|
Matt DTS It might fail. A DA might get it, for instance, but that's less
|
|
likely.
|
|
AFA Gary J Q = Parik (Matt)
|
|
Dave Lyons Grab the memory before calling DeskStartup to increase the chance
|
|
of success.
|
|
Matt DTS i'm out
|
|
AFA Parik Heh, was very interesting using the debugger... debugger loaded
|
|
into $01, the
|
|
Matt DTS (done, that is)
|
|
AFA Parik program loaded into $01, and graphics were in $01. :)
|
|
AFA Parik Doug, whats up? I want to hear the interesting topic...
|
|
AFL Jim So, what else is up with everyone tonight?
|
|
DougDavies You can tell the debugger to load where ever you want!
|
|
DougDavies oh ok Parik
|
|
Dave Lyons How?
|
|
AFA Parik BLOAD DEBUGGER,A$... :-)
|
|
AFA Gary J hehe
|
|
Dave Lyons BLOAD the debugger? I think we're talking about different
|
|
debuggers....
|
|
AFA Parik Just kidding Dave
|
|
Dave Lyons (Oh...didn't see your ":-)" :-)
|
|
Matt DTS Calm down, Dave. Breathe deeply. Think of Valium.
|
|
DougDavies Just go in with an editor and change the org address in the load
|
|
information
|
|
AFA Parik Developers going blind? Oh no! ;)
|
|
Matt DTS (Oh yeah: :-) )
|
|
Dave Lyons I know: you just allocate all the RAM except for where you want
|
|
the debugger
|
|
DougDavies i load it in bank 7 a lot
|
|
Dave Lyons to load, and then you dispose of it later [making Mem Mgr calls in
|
|
the monitor
|
|
Dave Lyons thru NIfty List, for example].
|
|
Matt DTS Or through the mangler, for those who have it
|
|
DougDavies No just change the OMF info
|
|
Matt DTS Doug: What was your whiz-bang gonzo topic?
|
|
DougDavies Well any way my question, dave already knows it!
|
|
Dave Lyons Hmmm...hadn't thought of that. Make it an absolute-address
|
|
segment.
|
|
DougDavies Anyone written a program with scroll bars not allocatted with
|
|
TaskMaster..?
|
|
DougDavies that act the same a task masters?
|
|
Dave Lyons Oh, _that_ topic...I was going to try to figure out how SizeWindow
|
|
or TaskMstr
|
|
DougDavies We relocate the debugger to break out a lot of games...
|
|
Dave Lyons resizes the scroll bars--haven't done it yet.
|
|
DougDavies found out how to debug prodos 8 applications with it too.
|
|
Matt DTS It's not SetDataSize, Dave?
|
|
DougDavies Dave looks like your correct, need to dispose and then reallocate
|
|
control
|
|
Dave Lyons He wants to know how to move the bounding rectangle when the
|
|
window changes
|
|
DougDavies Need to change bounds rect, not data represented by scroll bar
|
|
Matt DTS (I'm really rusty on this; I've been buried in OS for several
|
|
months now)
|
|
Dave Lyons size--SetDataSize will just recalc the size of the indicator.
|
|
DougDavies And then how do you stop your Draw Content routine from clobbering
|
|
your
|
|
DougDavies scroll bars
|
|
Matt DTS Doug - send an Email through MCI and we'll tackle it tomorrow.
|
|
DougDavies with the ClipRect? Seems logical
|
|
AFA Parik Buried in OS? Sounds like Dorothy & Toto's problem, 'cept they
|
|
were trapped.
|
|
DougDavies ok matt, i'll check with dale
|
|
Matt DTS Thanks.
|
|
AFL Floyd Doug: You might want to look at some Mac source code for this
|
|
since the Mac doesn't have TaskMaster
|
|
DougDavies Guess you never fool with the origin then, leave it at (0,0) all
|
|
the time
|
|
Matt DTS Anything else someone wants me to ignore and not answer? <grin>
|
|
DougDavies Where floyd, i've looked everywhere, no one here at WordPerfect
|
|
has any
|
|
AFL Floyd I have a stupid little scrolling program written in LightSpeed C
|
|
if you want to look at it.
|
|
Dave Lyons Doug, if your data size will always be less than 16K pixels, you
|
|
can use the
|
|
Dave Lyons Origin as usual. Having a super-large data area is sort of a
|
|
separate problem
|
|
Dave Lyons from doing your own scroll bars, except that I guess that's the
|
|
only reason you
|
|
Dave Lyons want to do yours yourself.
|
|
ScottG25 What was the solution for the super large area drawing, anyway?
|
|
AFL Floyd Doug: I'll upload the source to my little program to you
|
|
tonight.
|
|
DougDavies Yeah, really want to be able to display 50,000 lines if possible
|
|
DougDavies ok floyd, uses control managers scroll bars, not taskmasters?
|
|
AFA Parik hire someone to do it for you is a good solution Scott.
|
|
AFL Floyd Doug: It is Mac source code. LightSpeed C, but you should be able
|
|
to figure it out.
|
|
DougDavies ok, great, no 68000, GOOD
|
|
AFL Floyd The Mac doesn't have TaskMaster, you have to do everything
|
|
yourself.
|
|
AFL Floyd I'll send it to ya as a straight text file.
|
|
AFA Parik Next time I see a Mac user, I'm going to rub THAT in his face.
|
|
DougDavies there use to be a call called SizeControl, it dissappeared i
|
|
guess?
|
|
DougDavies ok thanks floyd
|
|
AFL Floyd BRB... uploading to doug.
|
|
DougDavies No explanation how to do it now in any books though
|
|
Matt DTS parik: Like that will mean anything to an average Mac user.
|
|
AFA Parik :-)
|
|
Matt DTS Try it on AFL Tom and see what he says.
|
|
DougDavies Could pull apart task master
|
|
Matt DTS Only if you have a strong stomach, doug. <grin>
|
|
AFA Parik "I have Taskmaster and YOU don't!!!" "I have a 25 megahertz
|
|
machine..." :-)
|
|
DougDavies Too bad apple didn't make the Data size bigger than $3FFF
|
|
Matt DTS Remember, the Window Manager can do what it likes in TM because
|
|
it's...
|
|
DougDavies yah, sucks doesn't it
|
|
Matt DTS ...using it's own data structures. It can "cheat" if it wants to,
|
|
iow.
|
|
Matt DTS Doug: You got an enhancement request, send it in. We read 'em
|
|
all.
|
|
DougDavies enhancement request?
|
|
Matt DTS "Gee, it would be nice if the data size could be bigger than
|
|
$3FFF..."
|
|
Dave Lyons (Request an infinite Data Size, Doug! This is your big chance!
|
|
:)
|
|
DougDavies would be nice, but they wouldn't even listen to half my ideas on
|
|
custom menus
|
|
DougDavies at least Jim wouldn't, he did help quite a bit though...
|
|
DougDavies what happened to Jim
|
|
Matt DTS Jim's gone to work for Steve in Apple II System Software (making
|
|
the toolbox).
|
|
DougDavies yes, wish we could get a journalling driver working too?
|
|
AFA Parik Hi again Dwight
|
|
DougDavies it's all GetMouse's fault
|
|
Matt DTS ReadMouse, isn't it?
|
|
DwightW1 MODEM PROBLEM AGAIN
|
|
DougDavies yah, almost works
|
|
Matt DTS It will work better on a future system disk. Jim says so.
|
|
<grin>
|
|
DougDavies great, guess or new product won't support it for a while then.
|
|
Matt DTS Well, it almost works. <sighing>
|
|
DougDavies but Jim is a big help.....cleared me up on a lot of things
|
|
Matt DTS (For the audience: Journalling doesn't work because it doesn't
|
|
journal...
|
|
Matt DTS ...the ReadMouse call, which is made by many of the tools quite
|
|
often.)
|
|
Matt DTS And it can't journal ReadMouse because the Event Manager calls
|
|
that.
|
|
Matt DTS (ALL the time.)L
|
|
DougDavies sniff
|
|
Matt DTS SO: Who's gonna buy me dinner at Boston AppleFest? Positions
|
|
still available!
|
|
DougDavies ok, any more on the scroll bar stuff?
|
|
AFA Gary J :)
|
|
Matt DTS (Like I always say: when depressed, go for the ultimate gall.)
|
|
Matt DTS Not off hand, Doug. Write in tonight; we'll attack it tomorrow
|
|
morning.
|
|
AFA Parik I thought DTS was going to buy dinner for everyone.
|
|
AFL Jim Matt, we'll send Willie to pick you up.
|
|
Matt DTS I think you're mistaken, parik. Isn't it the other way around?
|
|
DougDavies Ok, how bout this one?
|
|
Matt DTS Willie? The Great Missing AFP?
|
|
DougDavies How to have a lot of dialogs without all those structures sitting
|
|
in memory?
|
|
AFL Jim Ask Floyd about Willie's sense of direction :)
|
|
DougDavies how bout out in a text file?
|
|
AFL Floyd Doug: The source code is in your mailbox. Remember this is my
|
|
first program on the Mac. ;)
|
|
DougDavies ok floyd, thanks
|
|
Matt DTS If you want to read them from a text file into a handle, use them
|
|
and then...
|
|
AFL Floyd Jim: I barf just thinking about it. ;)
|
|
DougDavies We don't have resources yet do we?
|
|
Matt DTS ...dispose of the handle, it should work.
|
|
Matt DTS We have resource forks, but no resource manager. Patience.
|
|
DougDavies thats what I thought of doing
|
|
DougDavies shame to waste all that memory on a dialog that only gets called
|
|
once
|
|
Matt DTS I personally, being a filetypes kinda guy, would recommend reading
|
|
them...
|
|
DougDavies No info on resource forks, though!
|
|
AFL Jim Floyd: I remember seeing your breakfast flash before my eyes...
|
|
Matt DTS ...from a BIN file (or an EXE file if you wanna use the loader)
|
|
instead of TXT.
|
|
Dave Lyons Doug, what info on resource forks don't you have?
|
|
DwightW1 excellent conference...looking forward to downloading it....bye
|
|
:)
|
|
AFA Parik bye Dwight.
|
|
AFL Floyd Jim: I didn't have breakfast...that was dinner from the night
|
|
before. ;)
|
|
Dave Lyons Matt, that reminds me...why is the ERROR.MSG file in */SYSTEM an
|
|
EXE file?
|
|
Dave Lyons It ought to be GLF, or something.
|
|
AFA Gary J I get the idea that Willie is NOT the person to get a ride with?
|
|
:)
|
|
DougDavies not sure, I have the most recent GSOS stuff, really haven't looked
|
|
at it
|
|
Matt DTS Why not? There was really no reason to change it to anything
|
|
else, although...
|
|
Dave Lyons The only info there _is_ on resource forks is how to open them,
|
|
and how to
|
|
AFL Floyd Not if you want to get anywhere in a reasonable amount of time.
|
|
:)
|
|
Matt DTS ...it probably ought to be either $BC or $F9. It just wasn't
|
|
important...
|
|
Dave Lyons create a file that has one. The format of data in all resource
|
|
forks is
|
|
Dave Lyons currently reserved by Apple.
|
|
Matt DTS ...enough to delay 4.0 for. (grin; some things always fall
|
|
through the crax)
|
|
DougDavies thats what i thought
|
|
Dave Lyons Matt--unimportant as long as nobody tries to run the thing! It's
|
|
currently in
|
|
AFA Gary J Floyd: Yes, but didn't he deliver your dinner right to your
|
|
doorstep? :)
|
|
Dave Lyons OMF format, so $BC sounds good. I guess $F9 would work if the
|
|
system just
|
|
Dave Lyons OPENed and READ it instead of InitialLoading it.
|
|
DougDavies anyone want to know how to run a prodos 8 program under the APW
|
|
debugger?
|
|
AFA Parik Oh matt, that reminds me, has anyone asked for filetype $CD to be
|
|
used as the
|
|
Matt DTS Yeah. It may be changing around (internally) a bit anyway.
|
|
Matt DTS (it can do that since it's a private data structure.)
|
|
AFA Parik ACE packed files? $Bsomething is sound, but just about everyone
|
|
is using
|
|
AFA Parik $CD.
|
|
Matt DTS The ACE files will be filetype $D8, with a low auxtype.
|
|
Matt DTS Early ASIF, AIFF and ACE documentation referred to filetypes $CA,
|
|
$CB and $CD
|
|
Matt DTS ...which created no end of confusion. The new sound filetypes are
|
|
in the $Dx
|
|
AFL Floyd Doug: Yes to P8 and APW debugger.
|
|
Matt DTS range. ($D8 is "Sampled Sound")
|
|
DougDavies Ok here it is
|
|
AFA Parik Ok.
|
|
DougDavies put a break in the PRODOS file (prodos 8) at byte 0, with an
|
|
editor
|
|
DougDavies now load the debugger and type in a program that does a QUIT call
|
|
to PRODOS
|
|
DougDavies run the program. It will load prodos, hit the break, and your
|
|
debugging
|
|
DougDavies prodos 8, very useful for cracking prodos 8 games, like ELECTRONIC
|
|
ARTS...
|
|
DougDavies who hasn't figured out how to program in a real operating system
|
|
yet
|
|
Matt DTS Shame shame, Doug. <cant believe you said that>
|
|
AFA Parik haha! I love it! Thats great Doug! Thanks a LOT!
|
|
AFL Floyd Neat.
|