Survey posted 2004-04-26 06:25 by Morgan.

Contribute ideas to surveys by sending a mail to survey@ticalc.org.

Yeah, just to avoid some confusion really.

It wont do any harm, but will make things easier, seeing as the programs written for the 84+ will have more processor speed in mind.

[by the way, just curious, by saying backwards compatible.... does that mean ASM too, or just basic progs?]

Why don't group similar arquitecture calculator in one Dir Tree?

That would make more sense :)

68K ROCKS!

>That makes more sense.

I agree completely. Make a new 84 directory, and watch as the entire contents of the the 83(+) section is copied into the 84 directory. Perhaps some programs will run faster on an 84, but users will either have to deal with it or someone will have to create a program that wastes clock cycles on the patched program (the program is patched on-calc). While it does not always work, there is one such program for the 89; I can't seem to remember the name of it though.

The 83+SE already has a 15 MHz processor, just like the 84+ and 84+SE will have. TI fixed this potential incompatability by putting in a switch that controls whether the processor runs at 6 MHz (just like the 83- and 83+), or 15 MHz.
There is no reason I can think of why this same switch wouldn't be in the 84+. Everything will run on 6 MHz by default, unless it wants to run on 15 MHz (and can therefore be assumed to be capable of running at that speed).
Sure, as time goes by, and the 83+BE is used less and less, it will become standard practice to bluntly set 15 MHz in one of the first lines of your program, but that doesn't matter; for now, they'll be compatible, even for Asm programs.

For the last time, the 84+ and 84+SE are no faster than the 83+SE. They are only faster than the 83+.

Would it be possible to rename the 83+ directory to something like "83+/84+"? That way, it will not be confusing to new users, and we will not have to worry about the troubles of having two directories.

MWM

They should simply make the "84plus" directory hard linked to the "83plus" directory

good idea. That way 84 users can see what they want and so can 83 users, without having to go to a directory with multiple "/" in its title. If i was an average student with a brand-new 84+SE i might not know what to make of the 83+/84+/SE directory, id want to go to the directory clearly labeled "84+" knowing ill find what i want there. Another idea is to make 83+, 83+SE, 84+, and 84+SE directories, hardlinked so that the last 3 are all the same and the 83+ directory does not have those programs that are in there that say, "doesnt run well on regular 83+, need SE". It would provide ease of browsing without having a strain on the server with maintaining multiple directories.

I was thinking more of a soft link (symlink) than a hard link. :-)

ln -s 83plus 84plus

Most UNIXes do not allow directories to be hard linked. This includes BSD, Linux and Darwin (Mac OS X). In fact, POSIX.1 specifically disallows this.

Would you happen to be the same David Phillips that is a moderator at LinuxQuestion.org?

MWM

But as soon as the look in the address bar of their browser, they'll see they are in some subdir of http://www.ticalc.org/pub/83plus/ , and think they've found a broken link...

so make the directory

http://www.ticalc.org/pub/83_84/

or

http:// www.ticalc.org /pub/83plus_84plus/

or somthing so that people know that it's the right place. Just add a character (duh, **valid** characters) and put everything in there. It might be some work re-naming everything but how often is that required? besides, if it's 100% compatable technically it dosn't matter what happens.

Anyhow, I was all for making the catagories more specific anyway back when that debate was up.

NO FILES DELETED OR REJECTED!!!

:)

You should post that if you own the 84, consult the 83 archive files. This way, new members will be able to see that they are compatible, and this will also increase downloads for people who make 83 files. If you make an 84 section, people will just duplicate every file and re-upload it to keep their download statistics up. I'm all up for not having this section.

Yes, because there are several new functions added that BASIC programs can use, involving the clock. This would cause incompatibility issues alone.

ASM programs will also have additional B_CALLs and other things to facilitate using the USB port and the clock.

Very true. The new features need to be separated from the old.

A TI-89 with hardware version 2 also has a clock, and if a program required that, it wouldn't work on a TI-89 with hardware version 1. Yet somehow we don't have separate directories for these two variations of the TI-89. Probably because even though differences can affect a few programs, almost everythingis compatible. Which is the same situation as with the TI-83 Plus vs. the TI-84 Plus (as far as we know).

Actually, from looking at the stuff on TI's website, it seems that the 83+ and 84+ (or whatever) are actually more compatible than HW1 and HW2 on the 89.

I would hope that is the case, but I can't be sure. By "compatible" TI was probably referring mainly to basic programs (and maybe also flash apps that TI itself distributes). I won't be convinced they really are compatible until extensive tests are done.

Also, the HW version incompatibilites on the TI-89 aren't really that high these days. But of course they would be if people did things like requiring the clock in lots of programs.

That's true...I hadn't thought of that. Maybe there could be a separate directory for the 84+ ONLY programs, and they could have a hyperlink to the 83+ programs...

It would be about as stupid as creating a TI-89 Titanium tree.

That won't be happening :-)

Awwwwwww... j/k :)

But the TI-89 and TI-89Ti are really the same thing... only one has added memory, USB support, and APPS. So, really, nothing has changed.

As for the TI-84+... well... again, nothing really has changed, so I can see why people would not want to make another directory.

Now if there were only some way to prevent cluttering AND people getting 2x the credit for the same program...

While the TI-84 Plus seems compatible with TI-83 Plus assembly programs, the TI-89 Titanium seems *not* to be currently compatible with most TI-89 programs (although it usually should be easy to fix them). Thus a TI-89 Titanium directory seem to be a better idea, since it won't run the same programs. I don't really think it is a good idea, though; I'd prefer to simply see identification of which programs are compatible.

How do you know this?! All I know is that the AMS on the 89 Titanium is v3.00 and they didn't change anything! Well, nothing big anyways..

I know it mainly from information I read on the message board at tict.ticalc.org, which comes from testing done by someone who has a prototype TI-89 Titanium.

It's not compatible? Would could they have changed? They claim the TI-89Ti is compatible with the TI-89 on education.ti.com... (but maybe they're talking about BASIC and such)

Anyway, I'm sure somebody could make some sort of launcher that would keep compatiblility... or something.

There are a lot of things that could have been changed. Two of the things that they did change (and which in themselves break most ASM stuff) are:

1. The RAM is actually located in the first 256K of address space (0x000000 to 0x03ffff) but on the regular TI-89 you could access "ghost space" beyond this (such as 0x040000 to 0x07ffff) which would wrap around and refer to the RAM that was meant to appear at lower addresses. Since the interrupt table at the very start of memory is "protected" against writing (though this is very easily switched off) some programs would write to the higher addresses which wrap around and access the table, but don't trigger the protection. This was also used to beat the size limit, since if the last part of real RAM is allowed execution, so is the ghost space.

Unfortunately this won't work on the TI-89 Titanium. Older versions of TI-GCC used the ghost space to set interrupts, although this has recently been fixed.

2. Also, TI has moved the starting location of the ROM to 0x800000. Many programs (included TI-GCC ones, until this was recently fixed) would try to find the start address by reading the function table and ANDing with 0x600000. Now this has to be changed to 0xe00000 to accomodate all version. Only one character of source to change, but will still break without it. Since calculating this is often part of HW version checking (necessary for grayscale among other things), programs not aware of the TI-89 Titanium may assume the calculator is HW1 (meaning grayscale won't appear correctly) or perhaps even crash.

Fixing programs should be easy in most cases, but will need to done separately for each program. This could be a problem if the author is no longer around and source code is unavailable. Some work has been done on automated patchers but I don't think they are completely effective yet.

You're right, just giving identification as to what calculator will best run a program would be easiest.

Isn't the reason it's not compatible because of the USB port? Obviously that would make the original link port functions inoperable.

Several things seem to be incompatible, click the Web Page link to go to a forum host by the TI Chess Team. If the link is broken, the topic is on the second page of the Off Topics section. Also, I know I typed Titanic, jsut read the forum to find out why.

The link does seem to be broken, just go to tict.ticalc.org. When there beneath the title there are several links, click Message Board. When the forum appears, go to Off Topics, then on to the second page where you should find a thread called TI-89 Titanium.