| Author |
Message |
|
WikiGuru
ADOS (Attention deficit... Oh! Shiny!)
Elite
Joined: 15 Sep 2005
Posts: 923
|
 Posted: 09 Nov 2006 05:42:05 pm Post subject: |
|
|
As the poll question asks, I was thinking about increasing the 83p's capabilities to calculate similar to an 89's through either an asm program or a flash app. Functions I'm thinking about are simplifying equations, 'smart function display', and other math functions not in the 83p.
Is this feasible and good idea, and why/why not? |
|
| Back to top |
|
|
IAmACalculator
In a state of quasi-hiatus
Know-It-All
Joined: 21 Oct 2005
Posts: 1571
|
| Posted: 09 Nov 2006 06:07:56 pm Post subject: |
|
|
Good idea, but not feasible. A CAS is nearly impossible to write. Even TI had to buy a company to make one for them. Plus it would take up a ton of space.
Oh, you might also want to make the topic title a leetle more specific.  |
|
| Back to top |
|
|
nitacku
Advanced Member
Joined: 23 Aug 2005
Posts: 408
|
| Posted: 09 Nov 2006 08:42:02 pm Post subject: |
|
|
| It's a good idea even if only a few functions could be added. |
|
| Back to top |
|
|
todlangweilig
Advanced Member
Joined: 14 Feb 2006
Posts: 470
|
| Posted: 09 Nov 2006 08:47:52 pm Post subject: |
|
|
How about these two for a start?
Symbolic
PrettyPrint |
|
| Back to top |
|
|
Fallen Ghost
Elite
Joined: 15 Jun 2006
Posts: 955
|
| Posted: 10 Nov 2006 08:54:21 pm Post subject: |
|
|
what do you need to calculate other than prettyprint and symbolic that a 83+ can't do? (and that there isn't already a TI-flash apps that does that)
Or maybe a 1-byte TI-BASIC usable variable, so that it can go much faster to calculate (even if the result is stuck to be between 0 and 255)
Last edited by Guest on 11 Nov 2006 11:40:35 am; edited 1 time in total |
|
| Back to top |
|
|
*DEAD*
Newbie
Joined: 17 Nov 2006
Posts: 5
|
| Posted: 17 Nov 2006 11:24:13 am Post subject: |
|
|
Hi, im new here, but i will tell you that its an excellent idea, and probably more feasible than you might think.
Ive lived in the US for 4 years, and i used a TI 83+. In Australia however, I was told that TI's were not allowed for various reasons which turned out to be a bunch of bs anyway, so I had to go out and get a HP 39g+. It has an excellent software system, but its build quality is absolutely crap. Apparently they have a 50% return rate, which is huge considering that most companies consider .5% to be an issue. So now taht my calculator is falling to pieces (its honestly shocking, half the keys have loosened themselves from their sockets, the delete key is the worst and its probably the most used). So I now have the option of buying a new model HP (which already has a 20% return rate), a casio which i hate with a burning passion, or a TI 84+ (TI 89's arent allowed due to the cas system). What put me off the TI 84+ however is the absolutely pathetic calculations environment (the main screen, whatever you want to call it). Both the HP and casio allow you to scroll up and select your previous inputs and answers. This proves to be surprisingly useful, and is the key reason I have a new HP instead of a TI.
So, I would definitely say go ahead and make something like the TI 89s calculation interface where you can scroll up and select and copy previous inputs/answers. This would be an excellent start.
As for the cas, yes any top cas system would be complicated, i mean look at maple and mathematica they charge several 100 for their software, but on a simple level I think it could be done. What would you want it to do, some things are rather easy, others are rather complex. For example, theres an algorithm, couldnt remember what its called, to symbolically differentiate any equation by introducing a new set of operations for multiplication (product rule), division (quotient rule), brackets(chain rule).
Something like integration on the other hand is much much more complicated, im sure if you looked at some of the open source cas systems you could see how they do it, but im not sure if you could fit it all onto the calculators themselves.
Then theres things like exact values. Instead of multiplying numbers, you need a new set of operators to multiply expressions. For example, sin(60)*2 quite obviously = 3^.5. But for the calculator to do this, it would have to first check a table of exact values to determine that sin(60)=3^.5/2. now, you dont want to loose this exact value, so i guess youd store the expression almost like strings so they dont get lost. Then multiplying it by 2 would give the expression 3^.5*2/2. That would be the easy bit, then you have to simplify it which would require more thought than i can give at midnight.
As you can see, its a good idea but extremely complicated, however i think a much improved interface would be an excellent start. Just having something which would allow you to scroll though your past answers saves a remarkable amount of time. I guess the job might be a bit tedious because if you were going to make an entirely new program youd have to re implement all the functions ranging from adding to logs (im unfamiliar with TI programming, i know Sys rpl which is used on the HPs but id imagine mirageos would provide all the tools youd need), but i seriously believe this is a good idea. Maybe by next year when my HP breaks again ill be able to get myself another of my beloved TIs |
|
| Back to top |
|
|
leofox
INF student
Super Elite (Last Title)
Joined: 11 Apr 2004
Posts: 3562
|
| Posted: 17 Nov 2006 04:33:07 pm Post subject: |
|
|
| I really would love this idea, but i can image that it's hard to make.. I barely got multipcsjdkjatiocnt working in asm... |
|
| Back to top |
|
|
Fallen Ghost
Elite
Joined: 15 Jun 2006
Posts: 955
|
| Posted: 17 Nov 2006 07:56:37 pm Post subject: |
|
|
leofox wrote: multipcsjdkjatiocnt
\/\/h47???
Or maybe just having the ability to use 1-byte variables, so that calculations can go much faster (like for list elements or just for screen coordinates), so that it can go much faster to calculate (even if the result is stuck to be between 0 and 255). That would help alot to increase Basic game performance. |
|
| Back to top |
|
|
*DEAD*
Newbie
Joined: 17 Nov 2006
Posts: 5
|
| Posted: 19 Nov 2006 07:30:36 am Post subject: |
|
|
| i still think that just having a better environment to work in, even if it was only being able to copy answers, youd wind up with something that would be used almost as much as mirageOS itself. Have any of you guys ever used a ti-89/hp/casio. It is an extremely handy feature. |
|
| Back to top |
|
|
DarkerLine
ceci n'est pas une |
Super Elite (Last Title)
Joined: 04 Nov 2003
Posts: 8328
|
| Posted: 19 Nov 2006 12:19:47 pm Post subject: |
|
|
Fallen Ghost wrote: leofox wrote: multipcsjdkjatiocnt
\/\/h47???
Or maybe just having the ability to use 1-byte variables, so that calculations can go much faster (like for list elements or just for screen coordinates), so that it can go much faster to calculate (even if the result is stuck to be between 0 and 255). That would help alot to increase Basic game performance.
[post="91198"]<{POST_SNAPBACK}>[/post]
What would be feasible in that direction is to either write your own non-Basic interpreted language (which has in fact been done) or switch to assembly.
It also has nothing to do with improving the way the TIOS does math (moreover, you're actually sacrificing even more functionality for the sake of speed).
Last edited by Guest on 19 Nov 2006 12:22:44 pm; edited 1 time in total |
|
| Back to top |
|
|
threefingeredguy
Advanced Member
Joined: 01 Sep 2005
Posts: 479
|
| Posted: 21 Nov 2006 03:17:42 pm Post subject: |
|
|
| Why don't you guys just work on Vera? |
|
| Back to top |
|
|
IAmACalculator
In a state of quasi-hiatus
Know-It-All
Joined: 21 Oct 2005
Posts: 1571
|
| Posted: 21 Nov 2006 05:41:32 pm Post subject: |
|
|
Good idea, but no one knows where to start. That's been the whole problem. Your post of 1:20 PM has been one of the few human posts in forever. The last ones were Super Speler marvelling at Vera's fallen state, and me poking fun at spam bots. Since they were in spam topics, however, they've been deleted.
Anyway, let's stay on topic. </hypocracy>
Last edited by Guest on 21 Nov 2006 05:44:28 pm; edited 1 time in total |
|
| Back to top |
|
|
*DEAD*
Newbie
Joined: 17 Nov 2006
Posts: 5
|
| Posted: 22 Nov 2006 09:44:56 am Post subject: |
|
|
yep, looks pretty dead to me. Btw, after reading a few things about how the TI works, keeping in mind i could be completely wrong, i have a whole new idea bout how this could work.
This is almost a complete guess, and it should probaly be a question, but how does the ti OS actually work. Would each part of the OS, be like a separate function in the rom. For example, from what weve been talking about so far, completely rewriting a working environment would require a complete rewritting of every feature the calculator has. But would it be possible, say for the graphing, to call the address of the graphing procedure in the rom to use it. As a best guess with out really looking into it, id say that when something is called, such as the graphing procedure, execution is passed from the main screen procedure to the graphing procedure, and when clear is pressed back to the main screen. Now, if a program were to be called from the main screen, for simplicities sake say you go Asm(TEST) from the main screen, execution is now passed to whatever TEST is. Then, if from within TEST the graphing procedure is called, execution is passed to the to the graphing thing, and then when the person presses clear it finally reaches RET and execution goes back to TEST.
like
Code:
homescreen
call test
call grapher
RET
still in test
Would this work, i realize it probably wouldn't, but its worth a shot. If it did all the original feature could be retained and the main screen just replaced with a new, better working environment which could have a simple CAS system added to it if anyone ever gets around to it, lol.
Also, wed be able call the function that takes what id imagine would be nothing more than a string, and converts it into an answer, which would also save a hell of a lot of work.
Last edited by Guest on 22 Nov 2006 09:48:42 am; edited 1 time in total |
|
| Back to top |
|
|
leofox
INF student
Super Elite (Last Title)
Joined: 11 Apr 2004
Posts: 3562
|
| Posted: 22 Nov 2006 10:25:45 am Post subject: |
|
|
Fallen Ghost wrote: leofox wrote: multipcsjdkjatiocnt
\/\/h47???
Or maybe just having the ability to use 1-byte variables, so that calculations can go much faster (like for list elements or just for screen coordinates), so that it can go much faster to calculate (even if the result is stuck to be between 0 and 255). That would help alot to increase Basic game performance.
[post="91198"]<{POST_SNAPBACK}>[/post]
Quote: Would this work, i realize it probably wouldn't, but its worth a shot. If it did all the original feature could be retained and the main screen just replaced with a new, better working environment which could have a simple CAS system added to it if anyone ever gets around to it, lol.
Also, wed be able call the function that takes what id imagine would be nothing more than a string, and converts it into an answer, which would also save a hell of a lot of work. You'd have to disassemble and rewrite parts of the os, which is hard, and illegal.
If you want to make a new graphing environment, you'd make the button y= call your function (with a keyhook). You can make it very much like the original if you use lots of romcalls (aka ugly and slow)
Last edited by Guest on 22 Nov 2006 10:29:39 am; edited 1 time in total |
|
| Back to top |
|
|
Fallen Ghost
Elite
Joined: 15 Jun 2006
Posts: 955
|
| Posted: 22 Nov 2006 05:45:05 pm Post subject: |
|
|
Maybe ROM calls are slow, but the slowness mainly comes from the coding under the B_Call (aka routine), not from the calls themselves.
Rom calls can be good, it's just the way it is used now that isn't good. |
|
| Back to top |
|
|
|
Project Ideas/Start New Projects => Your Projects
