Omnimaga
Omnimaga => News => Topic started by: DJ Omnimaga on February 23, 2013, 05:36:40 am
-
Before the calculator has even come out, the first few TI-84 Plus C Silver Edition programs are starting to appear in Cemetech and Ticalc.org archives in newly-created sections: WFRNG and Quadratic Solver!
Just kidding. :P For the first time ever, a new calculator has seen a game come out before a quadratic solver, with KermMartian's port of Frogger 83+ (http://www.ticalc.org/archives/files/fileinfo/452/45215.html), a TI-BASIC clone of the popular arcade classic, retouched to fill the entire 84+CSE screen. Besides that, the first ASM programs have arrived as well: Adriweb found the hexadecimal code to enable lowercase on this calculator and posted the program here (http://tiplanet.org/forum/archives_voir.php?id=11342)!
In addition to that, a ball demo was posted as the 1st ASM program to ever display sprites on the screen: PCSEBall (http://www.cemetech.net/programs/index.php?mode=file&id=856)!
This program uses direct LCD access to display content. While it is a simple program, it is the first of its kind for this calculator and it runs at pretty good framerate. UPDATE: Replaced GIF with video above to show the real speed.
Also, ticalc.org staff Phero has posted two videos comparing the speed of a TI-83 Plus and a TI-84 Plus. The first video in ticalc.org news (http://www.ticalc.org/archives/news/articles/14/148/148260.html) just ran a counter using the Disp command, but the second one shows an old Omnimaga RPG in action: Mana Force:
The game appears to run only slightly slower than 83+ speed and is fully playable, except the battle transition animation, which seems to be half slower. This further confirms that outputing small strings of characters and sprites to the LCD will have acceptable speed (as demonstrated in Kerm's bouncing ball demo), but updating large chunks of the screen at once will be challenging. Illusiat 6 was supposed to be tested, but 84+ pics cannot be sent to the TI-84+CSE, so Mana Force was tested instead. UPDATE: Critor posted a video showing the calculator menus. The speed is pretty acceptable when selecting stuff or switching through menus.
Also, there are reports that the calculator might feature horizontal Z-addressing, meaning that ASM games that requires horizontal scrolling might still run at high frame rates if programmed correctly (basically, only updating the tile map row that the game is scrolling in instead of the entire LCD like what vertical scrolling requires). The TI-83 Plus also featured Z-adressing, but it was vertical instead of horizontal. In other words, if horizontal z-adressing allows high speed gains, then is anyone interested into making a side-scrolling platformer like Super Mario? :)
-
Another performance test:
The menus are slow.
Don't type the keys too fast, or the calculator will skip some of your keypresses...
Source:
http://tiplanet.org/forum/viewtopic.php?f=41&t=11302
-
Actually, I started to get used to the javascript emu, which made it even slower, sadly, so when I actually see it for real, it doesn't look so slow anymore :P
Of course, when put next to another z80 calc, it's clear that's it's slower.
However : (and don't get me wrong, I obviously think TI should have put a better CPU >.>, but that's my personal opinion) :
We are *used* to our other z80 and prizm/nspire/etc. So any comparison we make is kind of biased already.
The thing to see would be to make someone who's never used a graphing calc before try out the 84C, and ask him for any feedback, and that is in all the domains the calc is made for : math, programming features, color-screen, big screen, etc etc.
Then, I wouldn't know if the speed issue would really come first (or at all). Maybe the lack of CAS (yes, I am talking of someone who never used a graphing calc before, so....) would come first, but then we would explain him that's normal on this type of calc, and then maybe he would tell feedback about things we wouldn't even have though about ourselves, since we cannot speak as "new" users.
As you know, TI is insisting a lot on Math and Science etc., so if we consider that part for the 84C and as experimented-users, try to "ignore" the slowness issue, it will probably get the 84C to appear as the best non-Nspire calc out there. And that's because the reaction of the reviewer isn't focused on programming easeness etc. And obviously we, programmers, are at the opposite side, since a lot of us don't even use the calc mostly for math (we probably represent less than a few percents....) Any company (especially big ones) have to go with the majority, whether that's sad or good for its end-users. I guess that's why the 84C is seen as revolutionary to the math teachers I've talked with or heard, and that do not focus on programming with it.
(the best exampel they tell you is how you can differentiate the equations drawn on the graph screen, with the color, and how "beautiful" it is compared to older models. Of course, how can you even go against that argument with other calcs ?
In conclusion, it's all a matter of point of view, and that's pretty much unarguable.
My opinion again on this calc is not really interesting since I'm not a high-schooler anymore, and pretty much all the calcs I got don' get much attention nowadays, since I actually neither don't have much time to use them, nor my school allows any for exams. Sometimes I still use my Nspire CX CAS during lessons, for practical reaons, but that's all.
For me, if I'm asked whether I'd buy this calc, I'd probably say no because I don't *need* it. If I needed a calc and was looking for math, I'd seriously consider it (ignoring the Nspire series, again). As a programmer, I would probably lean towards an 84+ Pocket SE, *for now*. ; But as a "collector", I certainly can't wait to own a 84C, so....
Edit : Oops, I forgot to talk about the slowness impacting key responses issues : that's definitely a bad point. I have kind of 0 knowledge on ASM programing, but unless the skilled community programmers find a way to fix that, it's going to be a major problem with users who get used to the calc and start typing faster than what the calc can handle (and that's getting ridiculous...). But yet again, this is addressing a very minor percentage of all users TI's targeting.. :(
-
Here's a video demo of PCSEballs.
It's of course faster and smoother than the animated GIF above taken on the jsTIfied emulator! :)
So nothing to be afraid of: the hardware is perfectly able, and the OS will just have to be optimized by TI.
Source:
http://tiplanet.org/forum/viewtopic.php?t=11302
-
Hopefully, they will consider that Critor.
Else, we will have to port Axe to it. :D
-
We can probably set up some kind of an interrupt with a keyhook that records keypresses, stores them to a stack of some sort, and then executes each keypress when the screen is updated properly and there is time. I made a program that did the second half of this, so I could probably make a program to do the key recording, too :D
-
So nothing to be afraid of: the hardware is perfectly able, and the OS will just have to be optimized by TI.
Forgive me if I am a bit lacking in my confidence of TI at this point. Also, a short demo that shows only a few objects moving around on screen isn't that reassuring to me. When I see some full screen action with other things going on as well perhaps I will rest easy.
-
There was a little glitch so I remade the video with the fixed program.
Quite impressive for now ;)
-
Another performance test:
The menus are slow.
Don't type the keys too fast, or the calculator will skip some of your keypresses...
Source:
http://tiplanet.org/forum/viewtopic.php?f=41&t=11302
Hmm, at first glance: Menu loading seems pretty acceptable. As for the missing keypresses, this reminds me of the PRIZM. It can be a tad annoying early on but we get used to it.
Can you try scrolling through menus? I also wonder how really slow it is to scroll through a large program (for example, can you quit with ON then Goto to see how long it takes to scroll through code?), now that we can see that the emulator seems slower than the real calc?
Also the ball demo seems really great and show promising results. I definitivelly think that we could see CaDan on a TI-84 Plus C Silver Edition (as long as bullets aren't 16x16). :D
EDIT: I updated the news with your PCSEBall video.
-
I also wonder how really slow it is to scroll through a large program (for example, can you quit with ON then Goto to see how long it takes to scroll through code?), now that we can see that the emulator seems slower than the real calc?
I allready did that as I've started porting some of my huge math programs...
Seriously, while the calculator scrolls your program to the line on which the error/break occured, you can go doing something else...
Same speed as the TI-Basic performance test from ticalc.org: it scrolls by less than 2 text lines per second...
We'll have to develop TI-Basic programs on emulators...
Hm sorry, I forgot it was forbidden now! :P
-
The Frogger link is wrong, and it was created by me, not Elfprince. :)
-
Oh lol thanks for pointing out. I was sure it was Elfprince's game. I'll check that now.
-
Nice, we already have games for it! :D
I am so struck inbetween between getting it and not >.<
-
We'll have to develop TI-Basic programs on emulators...
Hm sorry, I forgot it was forbidden now! :P
We can create an editor for the PC.
-
How about we call the emulators now simulators? XD
-
We'll have to develop TI-Basic programs on emulators...
Hm sorry, I forgot it was forbidden now! :P
We can create an editor for the PC.
A good alternative will be Doors CS8 instant-goto feature, but might be better if ALPHA scrolling behavior is also modified so that it scrolls all 7 lines instantly instead of line by line.
-
ALPHA scrolling was always a page, so it must be 9 lines with the bigger LCD, right?
-
I wouldn't be surprised if it was. The problem is that when it scrolls, it does it gradually, meaning that the screen has to be almost entirely updated 9 times in the process. The best solution would be an hook-based feature that removes that gradual scrolling entirely.
-
Yeah, we probably need some awesome flash apps that mod the os enough or install hooks to increase speed...../me pokes thepenguin
-
Hi,
Made a speed test for Disp and Output instructions.
Source:
http://tiplanet.org/forum/viewtopic.php?f=41&t=11309
-
How is the TI-83 Plus.Fr managing to get Disp speed almost as fast as a TI-84 Plus in classic mode?? O.O
-
Because 2.53/55MP classic mode was allready slower than 2.43 ?
Let's make another test with 5 calculators together :P
-
Isn't it hard to press the buttons on the calcs at the same time? XD
-
Not if you have 3 hands.
-
Because 2.53/55MP classic mode was allready slower than 2.43 ?
Let's make another test with 5 calculators together :P
Don't the 83 Plus.Fr in this vid come with OS 1.19? It's the blue design that was actually a real TI-83 Plus rather than a TI-84 Plus. (Unless you swapped the cases?) ???
(http://www.trendysdog.com/img/p/10-50-thickbox.jpg)
If I recall correctly, if MathPrint mode is disabled, the calc gets as fast as OS 2.43.
Could you try with a regular 83 Plus from United States? (The one with English keypad) I am curious about this whole fast Disp on original 83 Plus.Fr thing.