I have both, and they have their strengths and weaknesses. I tend to go more for the 89, though, because I just like it more. (And it has better games.)

This is interesting. In my experience, everyone gets TIs, i almost never see anything else for sale where i live (santa cruz ca). Ive only ever seen two calcs that arent ti. One is some HP that someone i know owns, referred to by my math teacher (an 86 user) as the "HP monster." The other ive seen was an unfortunate friend who had a casio. Looking at the text and comparisons on the page, it looks like the HP is infinitely(approx. value, too big to be exact :)) better than the 89, but looking at the screenshots id say i would rather have the 89 just for the looks. But i am biased towards the 89, and having a place like ticalc to get tons of free 3rd party programs is a bonus. Besides, i never think of the 89 as slow, or not having enough memory, so i dont think it would make a difference to have the SD card or faster processing speed. And i love the menus on the 89, they are so intuitive, and somewhat windows-esque and look really good, so i would prefer the 89, but would not be disappointed in an HP. This is getting kind of long.
BTW, i didnt even know there was an hpcalc.org, which came first? i get the impression it was ticalc.
And is there a casiocalc.org, or maybe badcalc.org, or something like that?

the ticalc.org copyright says 1996. The hpcalc.org copyright says 1997.

thanks, i guess that would have been a good place to look...

I agree. The TI-89 has never seemed slow, to me. And even if you do have to wait three seconds to solve a complex problem... so what? That's not a large amount of time. Though, for programming purposes, having a faster calc is better.
LOL to your comment about badcalc.org--
C:\>ping www.badcalc.org
Unknown host www.badcalc.org.

Nope, doesn't exist. :)

From what I've read on comp.sys.hp48 it seems like the speed increase varies in two distinct areas: For floating point calculations the arm has to emulate BCD instructions so the speed increase is a factor of 3. For other operation the speed improvement is a factor of 6.

The CAS of the 49g has superior integration, but the ti89's solve command is superior to that on the 49g. You can search comp.sys.hp48 for examples. The 49g might fail to solve for a variable if the formula involves log() (base 10), but it will succeed if you manually rewrite it using ln(). I forget the example but the factor command on the 49g once returned to me a factored expression involving: (-1)^2.

There are several equation writer programs that can be installed on the ti, not just 1. Also MY ti89 can do 299! in 2 seconds, I don't know about anyone elses ;)

overclocked it?

In the AMS the slowest part of doing 299! is not the math, but the display code. I have my own display routines that display the result of 299! in 1/3 a second, whereas the AMS might take 10 seconds.


-Samuel

alpage.ath.cx/hptute/arm.htm
Explains how to use the ARM CPU of the 49g+ directly.

alpage.ath.cx/hptute/hpprog.htm
FFAC - Fast Factorial V3.00
FFAC runs about 85 times faster then the built in routine.
299! 0.04 Seconds
3000! 3.99 Seconds
5000! 11.9 Seconds

<<Also MY ti89 can do 299! in 2 seconds, I don't know about anyone elses ;)>>

Well, my V200 w/ AMS 2.09 can do it in less than a second ;P

Well, I don't know if anybody else thought of this or not, but the battery power in your calculator greatly affects the performance. If you have low batteries, of course, it's going to be slower. My TI-89 has been saying "BATT" for about a week now, and it's getting a lot slower. So, the time tests may not be perfectly balanced. Also, as stated on the website, the HW1 is slower than the HW2.

my 89 today returned 5*e^0, i think it was in solving diff eqs

LOL!!! That's REALLY bad...

I was testing that idea of 2 sec for 299! and on my HP49G+
and it was instantaneous on the screen after I pushed enter

bah ... down with HP!!

Do you realize that if HP starts making great calcs again, TI will have to start making better ones in order to keep up? Even if you are very fond of TIs, you should be very glad that HP is making good calcs!

MWM

The problem is, they don't need to keep up. If they did, we would be seeing improvements and upgrades all the time.

All I ask is that TI does a platinum/Titanium/(possibly call it chrome, lol) 86 edition.

NO! I would hate that, since my 86 friends would start bragging. My only argument now is that my 83+SE is faster, has archive space, more memory, and more popular. I try not to admit that the 86 has better math capabilities. If he had a better one, I wouldn't be able to take it ;)!

Man, ive seen a lot of 86 bashing going on here at ticalc, and it just seems out of place. Theres nothing wrong with the 86. I dont own one, and dont plan to, but ive used them a lot and they are very good calculators. They are the most powerful z80s (i think) and the menu system is not all that bad. It only seems bad to those who are used to the menus on the 83 or on the 89, which are kind of similar. I think the big benefit of the 86 menu is in basic programming. That kind of menu system being already present and easy to use is a godsend in making most basic programs, even in games.

If you look closely, the TI83+SE has very similar hardware to the TI86

The only thing it laks it the screen resolution.

Someone should make an OS that gives it the same functionality as one (with more customisability).

Currently, I am working on an Os for the TI83+se and TI83+.

When it gets released, I may write a shell that has some of the TI86 interface.

i knew the 86 was more powerful than the 83 and the 83+, but i wasnt sure about the SE, so thats why i added the "(i think)". You're writing an os that would be 86-ish? I think that would be useful, but im not sure if it would be appreciated, because most of the 83+ users feel that their current menu system is superior, and the 86 guys probably wouldnt have an 83+ to take advantage of an 86 interface...

TI-86 - 6mhz cpu, 128k RAM (i think), 256k ROM (I think), 128x64pxls

TI-83+SE - 15mhz cpu, 128k RAM, 2048k ROM, 96x64pxls

My initial release of the OS will probably take up about 48k (pages 0,7c,7d), plus 32k of RAM (pages 0 and 1), but page 1 will only be used for ASM execution.

This will leave alot of extra space to write a parser and a shell.

I plan to write an 83-style shell, and an 86-style shell (there is not enough screen space for an 89-style interface)

I have both a TI-86 and a TI-83+ & I would be very interested in such a project... it sounds like a really good idea if you can pull it off.

That looks like fun. However, make sure not to use anything from TI's OS, or they'll most likely sue you.

It won't (doesn't) have much. Just a command prompt, with no floating-point capabilities, so there isn't much to take.

I'm rewriting the B_CALL routine and the memory management, to take advantage of the SE's extra mem.

My main barrier is the PutC and VPutC routines and the linking protocol.

Does anyone know how the character graphics are stored?

Currently the characters are 5x8 sprites and the only way to link is to use the ROM's OS-loading (only allows you to reload TIOS)

I should add that processor speeds can be deceptive when they aren't the same flavor of processor. For example a Pentium 4 2Ghz is obviously slower than a Pentium 4 4Ghz, but it's much more difficult to compare a Pentium 4 4Ghz processor to a Mac G5 3Ghz processor. There are internal differences that make pure comparisons less than pure.

As it relates to this conversation:
10 is slower than 12, but it's uncertain the precise difference between the TI 12Mhz and HP 75Mhz. How many transistors are in each processor? How many "bits" is each processor? It's more complex than people think. It is likely however that the HP is faster...

I agree completely. I have a long list of good reasons why the 86 is better than the 83+ (I own both), and I would post it if it wouldn't be off-topic...

It would be nice to see some improvements to the 86. It's the best Z80 calc, and TI just leaves it behind.

See my comment on the first page.