The TI-84 Plus CE, which replaces the TI-84 Plus C Silver Edition, has been released and various TI community members such as tr1p1ea, author of xLIBC, have finally gotten their calculator in the mail!
http://www.underwooddistributing.com/shop?page=shop.browse&category_id=66
This means that other stores will most likely have it in stock soon and eventually the new school device will make its way to brick and mortar stores. The MSRP is $129.99 and the calculator features the following:
-eZ80 processor
-150 KB of RAM (huge improvement over previous TI-84 Plus series calculators)
-3 MB of Flash
-320x240 color LCD that supports 16bpp, 8bpp and more
-OS with more features
-Considerably faster screen refreshing (especially in menus and the program editor)
-BASIC, C and ASM support (a third-party language called ICE is also available and is similar to Axe on monochrome models)
However, no SDK yet and some ASM functionalities are rumored to not be available. The biggest advantage of this new calculator model will lie in the much faster screen and larger user RAM, so pure TI-BASIC and pure ASM programmers will benefit the most. UPDATE (November 11th 2016): The calculator supports BASIC, ASM, C and a third-party language called ICE, all in 8xp program format. Flash APPs are not available, as TI has decided to not release the 2048 bits RSA key and current technology makes it impossible to factor a key this large. But this should not be a problem, as people managed to find workarounds, such as for parser hooks.
-TI-BASIC: TI-83/84+/CE Basic documentation (http://tibasicdev.wikidot.com)
-Hybrid TI-BASIC: Sprites v3.3 (https://tiplanet.org/forum/archives_voir.php?id=539203) | CE Textlib (http://www.ticalc.org/archives/files/fileinfo/465/46522.html)
-ASM: eZ80 assembly documentation for the CE (http://wikiti.brandonw.net/index.php?title=Calculator_Documentation#TI-84_Plus_CE.5B-T.5D.2C_TI-83_Premium_CE)
-C: C Programming Toolchain, libraries and resources for the TI-84 Plus CE (https://github.com/CE-Programming)
-ICE: ICE Compiler and its documentation (https://tiplanet.org/forum/archives_voir.php?id=587211)
-Sprite graphics converters: ConvPNG (for use in ASM/C/ICE) (https://github.com/MateoConLechuga/ConvPNG) | Sprite Ripper (for use in TI-BASIC with Sprites v3.3) (https://tiplanet.org/forum/archives_voir.php?id=659602)
-Emulation: CEmu emulator (requires a ROM dump from a real calculator) (https://github.com/CE-Programming/CEmu)
I invite owners of the TI-84 Plus CE (as well as owners of the TI-83 Premium CE when it comes out) to share their discoveries about this new calculator series in this topic or on the forums, be it ASM tricks, bugs or other useful information.
Work on the OS and the relevant goodies that we need is underway.
The sad thing is that TI has locked us out of direct hardware access (like setting the LCD to 8-bit mode and palettes for graphics speed+free buffering), but we will continue to investigate.
I've only had time to mess with the LCD a little. I'll post some more info as it arises!
I want to get a Prime first then get the CE :( but I still plan on the CE too :)
Quote from: tr1p1ea on March 25, 2015, 11:55:16 PM
Work on the OS and the relevant goodies that we need is underway.
The sad thing is that TI has locked us out of direct hardware access (like setting the LCD to 8-bit mode and palettes for graphics speed+free buffering), but we will continue to investigate.
I've only had time to mess with the LCD a little. I'll post some more info as it arises!
does the OS block us our or the hardware?
Also i hope an emulator will be released soo, i really wan't to play with it but i don't have the money to buy random calculators :(
Quote from: tr1p1ea on March 25, 2015, 11:55:16 PM
Work on the OS and the relevant goodies that we need is underway.
The sad thing is that TI has locked us out of direct hardware access (like setting the LCD to 8-bit mode and palettes for graphics speed+free buffering), but we will continue to investigate.
I've only had time to mess with the LCD a little. I'll post some more info as it arises!
Yeah that's a shame that they did that. I guess it's better than nothing, though (eg the Nspire series where ASM is locked down entirely). Hopefully ASM coders can find workarounds. Maybe if they can take full advantage of the CPU they can find faster alternatives and limitations won't be as much of an issue anymore?
Quote from: Cumred_Snektron on March 26, 2015, 09:44:14 AM
Also i hope an emulator will be released soo, i really wan't to play with it but i don't have the money to buy random calculators :(
Same I am saving for a Prime right now :(
Ooh, glad you are getting an HP Prime too. It would be nice to see more HP PPL games out and if you ever get interested in ASM/C, then maybe hacking :)
By the way people have started documenting the calculator on WikiTI and the new sections are reachable via http://wikiti.brandonw.net/index.php?title=Calculator_Documentation .
Some notable mention about how ASM and BASIC programs being 64 KB max each and a Flash unlock sequence is still present. As for ports, to avoid pissing TI off, the wiki discourages the creation of any program that access the hardware directly for the time being.
This Cemetech thread might be handy as well: http://www.cemetech.net/forum/viewtopic.php?t=11405
Of course I also encourage CW users to share their discoveries and tricks in this thread too, as always.
EDIT: Also basic matrices are now limited to 400 elements. This could pose some issues to BASIC programmers, although there are alternative to matrices anyway (such as lists or strings).
64KB for a program, nice. It didn't scale too well because the grapism will take much more place ( we're talking from 1bit/pixel to 1/2btes per pixel, unless you do 16 or 4 colors per pixel).
What about the matrices? Wasn't that limit higher on monochrome calcs?
i doubt anyone would need a 20*20 matrix for serious applications
Quote from: Duke "Tape" Eiyeron on March 28, 2015, 06:56:52 AM
64KB for a program, nice. It didn't scale too well because the grapism will take much more place ( we're talking from 1bit/pixel to 1/2btes per pixel, unless you do 16 or 4 colors per pixel).
What about the matrices? Wasn't that limit higher on monochrome calcs?
Yeah the larger graphics were a bit of an issue on the CSE, although at least they were often archived. The problem was mostly when using inline sprites with xLIB or inside pure ASM programs.
Quote from: Cumred_Snektron on March 28, 2015, 12:14:26 PM
i doubt anyone would need a 20*20 matrix for serious applications
It depends. Remember how xLIB for monochrome calcs used xLIB for map engines? Scrolling tilemappers might use much larger matrices than that. But I don't really like using matrices for maps anyway because they're so large. I prefer strings (although slower) or ASM data.
Quote from: DJ Omnimaga on March 28, 2015, 04:29:59 PM
Quote from: Cumred_Snektron on March 28, 2015, 12:14:26 PM
i doubt anyone would need a 20*20 matrix for serious applications
It depends. Remember how xLIB for monochrome calcs used xLIB for map engines? Scrolling tilemappers might use much larger matrices than that. But I don't really like using matrices for maps anyway because they're so large. I prefer strings (although slower) or ASM data.
Not really actually :P im quite new to the community
anyway i'd use strings too, or just store them in a program.
On a side note, one reason why matrices can be good for maps is that they're easy to alter with boolean logic. For example, after you progressed far enough into a game you can make different tiles display instead in specific squares. With strings, that is easy too, but the problem is that the TI-OS has bugs with concatenating a very large amount of strings together, resulting into memory leaks much worse than when you don't use Goto properly (in one case I had a program cause a memory leak worth 8 KB in one go, with a slowdown worse than regular 83+ speed).
OH yeah i see why you would use a matrix now :) Though i didn't know TI was that bad with memory leaks and such <_<
What about lists ? I know you need more math but aren't these smaller/faster than matrices ?
They are the same size and the only advantage they offer over matrices is that their width can be altered as you wish, just like xLIBC and Axe map data. You can also use custom list names, which is impossible with matrices, and the limit is 999 elements instead of 400 (unless that was changed too?)
Not sure if they are faster, since I haven't tried that.
Earlier today calc84 found that some IO is mapped to RAM! This means that we can access with LCD for example, without needing and exploits or hax!
This is VERY good news as it will allow us to achieve greater performance and keep data sizes down.
I have also located the parser hook and can confirm that it works stock which could mean that an xLIB on the CE is likely possible.
Here is a vid of a palette test:
https://www.youtube.com/watch?v=zFVD8od6qdk
Ill post more info later today if i can :).
Awesome tr1p1ea! I'M really glad to read that. My main worry was the parser hook, because without them this severely limited some possibilities, especially hybrid BASIC programmers. What about Flash apps, though? I thought that parser hooks could only be installed from a Flash app?
And it's good that direct LCD access is still possible. Hopefully we can do extremely fast games on this platform.
Btw, could you test the following code?
While 1
ClrHome
Output(1,1,"0123456789ABCDEF0123456789ABCDEFGHIJKLMNOPQRSTUVWXYZ0123456789ABCDEF0123456789ABCDEFGHIJKLMNOPQRSTUVWXYZ0123456789ABCDEF0123456789ABCDEFGHIJKLMNOPQRSTUVWXYZ0123456789ABCDEF0123456789ABCDEFGHIJKLMNOPQRSTUVWXYZ0123456789ABCDEF0123456789ABCDEFGHIJKLMNOPQRSTUVWXYZ
End
I am curious about how fast this refreshes compared to the CSE? That might give me an idea of what enhancements a CE version of First Fantasy could benefit from.
Also since this calc lacks a 160x120 mode, will xLIBC have a mode where the GRAM is the entire screen rather than one half at a time?
The GRAM thing might work a bit different, as there is no LCD interlacing and no windowing. The buffer is still divided into 2 halves when using 8-bit palettes, however instead of being side-by-side like CSE, its top-bottom.
Ill test that when i get home!
Aah ok. I guess I don't mind, though, if the xLIBC syntax isn't made entirely different unless necessary by any limitation and stuff.
Awesome! It could be years before I get one though since I don't ever buy things new - my CSE came from winning a contest. But I love drooling over what other people get to experience :P like for the Nspire.
Nice tr1plea! What does the fox say though ???
Quote from: DJ Omnimaga on March 31, 2015, 12:50:51 AMBtw, could you test the following code?
While 1
ClrHome
Output(1,1,"0123456789ABCDEF0123456789ABCDEFGHIJKLMNOPQRSTUVWXYZ0123456789ABCDEF0123456789ABCDEFGHIJKLMNOPQRSTUVWXYZ0123456789ABCDEF0123456789ABCDEFGHIJKLMNOPQRSTUVWXYZ0123456789ABCDEF0123456789ABCDEFGHIJKLMNOPQRSTUVWXYZ0123456789ABCDEF0123456789ABCDEFGHIJKLMNOPQRSTUVWXYZ
End
I am curious about how fast this refreshes compared to the CSE? That might give me an idea of what enhancements a CE version of First Fantasy could benefit from.
Each full "refresh" takes 0.7s on average, on my 84+CE.
Ok thanks. That seems a bit faster than on my CSE, although not by much. Still an improvement, though.
Bwahahahahaha! http://tiplanet.org/forum/viewtopic.php?p=179794#p179794
http://zippy.gfycat.com/InbornConcernedFulmar.webm
Basically, the following code
While not(getKey
"abcd...........
"Searc.."+Ans
{1→⌊A
End
Will freeze your calc.
I hope TI fix this >.< (not that it's easily triggerable due to being useless code, but you never know)
Ok so I managed to get some parser stuff running and made a small tilemap demo:
www.youtube.com/watch?v=iNrhNgNze3o
The entire map is redrawn each frame and the scrolling is 8-pixels at a time. The code an be optimised for more speed as well.
Very encouraging when compared to the CSE!
I love the speed, it seems on-par with the monochrome calcs. :D
Nice! It's actually way better than CSE! Now with vertical scrolling?
Yup. Also this will be easier to port games between the CSE, HP Prime and Nspire, since speed on the ez80 calc won't be a serious issue . Mario is definitively possible for sure, now, even if not at 60 fps.
It'll be hard to do very graphical effects though, 20MHz is hard. THink of the Prizm (but welp, it had a strange FPS lock) about speed and quality.
Quote from: tr1p1ea on April 03, 2015, 12:31:33 AM
Ok so I managed to get some parser stuff running and made a small tilemap demo:
www.youtube.com/watch?v=iNrhNgNze3o
The entire map is redrawn each frame and the scrolling is 8-pixels at a time. The code an be optimised for more speed as well.
Very encouraging when compared to the CSE!
:o it's like a proper done CSE :D
Quote from: DJ Omnimaga on April 03, 2015, 04:58:03 PM
Yup. Also this will be easier to port games between the CSE, HP Prime and Nspire, since speed on the ez80 calc won't be a serious issue . Mario is definitively possible for sure, now, even if not at 60 fps.
if only there was a common language, that would be awesome O.O
(Source portin' time)
There was an attempt to make a common langue between Casio and TI : the MLC. It didn't got very popular though...
Actually it was somewhat popular back then. It's just that the project founder quit calc stuff early in the project, which pretty much led to its demise. Also, the one in charge of the TI-86 version ran into issues with memory constraints (such as a 8 KB executable limit)
A lot of the MLC discussion was on Epic Programming Studio programming team forums, which I was staff at. But that team disbanded in late 2005.
It's amazing how much of this stuff you remember :).
Another parser test, this time TIOS getKey vs ASM lib getkey. All code is the same besides the key function:
I believe the TIOS getKey actually isnt slow, rather its the key-repeat.
Meeeh, I want to get back on game programming! When I seee such graphisms, I'm amazed by the richness of details of this spritesheet.
Lol tr1p1ea thanks. I even remember some exact community event dates or member nicks from the old MC board. O.O
And woah BASIC getkey is so slow compared to ASM. That makes me want to investigate in FF to see if dungeon movements aren't actually throttled by getKey XD.
But everything is faster in ASM so that's not really a surprise :P
Yeah, but the result is still pretty pretty, so, whatever the language to make this, it's cool.
^this. As long as the code is optimized whenever possible and isn't slower than what the BASIC language is capable of.
Imagine making a demo like this in Brainc o.o
Can we even display anything other than text in brainc, such as sprites?
Yeah, you'd have to add some instructions (like pixel-on or something) or define a specific location in the ram as the screen (kinda like PlotSScreen)
Better : you map some sort of opcode parser in the BF memory, like the first 8 bytes to detect the operation and the arguments of the operation.
It's almost ASM that way :P
We should stick to C and ASM I think. Let's not make stuff even more complicated for people who have troubles learning low-level languages. :P
Also
@tr1p1ea, For xLIBC, I think there should be an extra option allowing you to use map data made of tiles from 0 to F instead of 00 to FF. The 16 tiles you would be able to use would either be defined in a list before launching the real command, a much slower string or if none is specified, then only tile 0 through 15 in tilepic 0 or 1 would be used. I'm thinking that a list would be better because TI-BASIC string storage (even in Ans) is very slow.
{tile1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16:real(3,0,X,Y,MAPWIDTH,MAPSTRING,XSTART,YSTART,XEND,YEND,UPDATELCD,TILEMODE
Axe Parser had ways to use such map data and it was much smaller in size. It could be handy for people who have simpler tilemap data who wants to save space.
Yeah, you mean using 4 bytes for each tile? YEah that could be feasible as even Axe Parser does have a function for dat "nibble{}"
Yeah 4 bits actually, although in TI-BASIC, unsquished form, it would be more like 8 bits instead of 16.
The custom palette feature would come to the rescue so maybe some map data could even be reused for multiple dungeons like in Illusiat 11.
So I just got my TI-84+CE and only a few seconds after turning it ON I ran into a funny bug: Basically, if you perform a full memory reset, it says RAM Cleared instead of Mem Cleared O.O
Also, some stuff to know about TI-BASIC: It appears that strings cannot be larger than 32768 characters. I haven't tested with 2-byte tokens, though. An ERR:UNDEFINED will be thrown when attempting to concatenate two strings that exceeds that amount when combined.
Fun fact: Memory means RAM:P its a common misconceptions that memory means storage aswell :)
though i guess it should say "storage cleared" or something
Yeah I meant mem as in both RAM+archive.
EDIT: This is an edited CSE screenshot of about how fast a graph grid (in line mode) can render:
(https://img.ourl.ca/griddraw.gif)
This could be handy for some animations (eg magic) in pure-BASIC
EDIT: New screenshot with the right frame rate.
Good news, at first glance, it appears that hacked pic vars still work on the TI-84+CE!
With the BASIC program included with http://www.ticalc.org/archives/files/fileinfo/279/27979.html , you can have up to 255 pictures. But I don't think you can send them between calcs.
Yay for Kevtiva making it possible to have 255 pictures instead of 40 15 years before the calc existed. :P
(https://img.ourl.ca/255pic.gif)
Is the that a CE emulator?
Yeah it's TI Smart View. But it's paid (I have a 90 days trial from TI website) and it doesn't emulate undocumented ASM instructions, so games that utilize special modes like Pacman CE does will not work properly. It's also ultra-laggy. (sometimes freezes for 5 seconds when I press like 3 keys in less than 2 seconds)
Sounds TI :P
Quote from: Unicorn on June 23, 2015, 04:45:13 AM
Sounds TI :P
Could you translate your post into English so that I can understand it?
Quote from: DJ Omnimaga on June 23, 2015, 02:07:46 AM
Yeah it's TI Smart View. But it's paid (I have a 90 days trial from TI website) and it doesn't emulate undocumented ASM instructions, so games that utilize special modes like Pacman CE does will not work properly. It's also ultra-laggy. (sometimes freezes for 5 seconds when I press like 3 keys in less than 2 seconds)
Thats what you get for emulating a 12 Mhz CPU in Java <_< it's a fine language, but its not made for that.
They should have used C or something. Also i looked a bit at the code and they use a lot of JS too (since java has abuilt in JS scripting engine). Wouldnt suprise me if
they run an emulator inside JS inside Java <_<
Quote from: DJ Omnimaga on June 23, 2015, 04:46:16 AM
Quote from: Unicorn on June 23, 2015, 04:45:13 AM
Sounds TI :P
Could you translate your post into English so that I can understand it?
I think he meant that it is normal for Ti to sell you something that barely works or laggs allot :P
Quote from: Cumred_Snektron on June 23, 2015, 04:51:01 PM
Quote from: DJ Omnimaga on June 23, 2015, 02:07:46 AM
Yeah it's TI Smart View. But it's paid (I have a 90 days trial from TI website) and it doesn't emulate undocumented ASM instructions, so games that utilize special modes like Pacman CE does will not work properly. It's also ultra-laggy. (sometimes freezes for 5 seconds when I press like 3 keys in less than 2 seconds)
Thats what you get for emulating a 12 Mhz CPU in Java <_< it's a fine language, but its not made for that.
They should have used C or something. Also i looked a bit at the code and they use a lot of JS too (since java has abuilt in JS scripting engine). Wouldnt suprise me if
they run an emulator inside JS inside Java <_<
On top of that they're charging money for it O.O
I bet they would be open to porting it to C or another faster language, but I doubt they'll bother paying staff to do it. It's a shame because the emulator from TI actually has some nice features, such as being able to see the graph, homescreen and table windows all at once.
Quote from: Cumred_Snektron on June 23, 2015, 04:51:01 PM
Quote from: DJ Omnimaga on June 23, 2015, 02:07:46 AM
Yeah it's TI Smart View. But it's paid (I have a 90 days trial from TI website) and it doesn't emulate undocumented ASM instructions, so games that utilize special modes like Pacman CE does will not work properly. It's also ultra-laggy. (sometimes freezes for 5 seconds when I press like 3 keys in less than 2 seconds)
Thats what you get for emulating a 12 Mhz CPU in Java <_< it's a fine language, but its not made for that.
They should have used C or something. Also i looked a bit at the code and they use a lot of JS too (since java has abuilt in JS scripting engine). Wouldnt suprise me if
they run an emulator inside JS inside Java <_<
No no no.
The problem (slowness of the emu inside SV) isn't coming from the fact that the emulator is written in JavaScript, nor the fact that the SmartView itself is written in Java.
The problem is that Java's webview has a bad/slow(er) JS engine in its WebView (compared to the standard set in modern browser, with insane JS "compiling"/performance, so to speak).
As said earlier, if you extract the .js and try the emulator core (the .js file) in Chrome, FF, or whatever, it runs fast. Too fast, even (2x-5x the real calc's speed). I did a benchmark (http://codewalr.us/index.php?topic=486.msg13814#msg13814) on that earlier.
They chose JS as the implementation language of the emu core, and that is clearly not an issue.
Now, if only they can bypass Java's crappy webview and embed v8 or similar, it'd be great. But I don't see TI doing that, as the current speed is already acceptable for teachers (it is faster than the directly native (C++ or C# I think?) emulator previous versions (the 84+ SV)
Oh I see now. That's definitively weird. And yeah I doubt they'll bother doing it. That said, I could see teachers asking for a faster emu if the slow key detection gets in the way. When I type fast, the emulator freezes for a few seconds because key input seems to make the emulator take a considerable amount of CPU resources.
I suppose http://codewalr.us/index.php?topic=592 would help :)
Indeed. That said, like I said in the other thread, eventually, a community emulator would be a great thing too :)
Also
@tr1p1ea should share some of the progress he made on xLIBCE here :)
I just searched TI 84+ CE cheap online, not excpecting anything, but look what I found! O.o
(http://i.imgur.com/s966CM3.png)
Site: http://www.overstock.com/Electronics/Texas-Instruments-TI-84-Plus-CE-Graphing-Calculator/9989573/product.html
This has to be a pack of 5-10 calcs. There's no way a store can charge this much unless it was a mistake... O.O
It might be a pack, butthat really suprised me.
I poked around and clicked on package contents and:
(http://i.imgur.com/4VPy29f.png)
You you think it comes with an extra bat? A bat doesn't cost 350+ $
Probably just a store trying to rip people off.
Maybe... I don't know though. :P
By the way, something I found out about the CE is that it doesn't seem to be very good at keeping a charge when unused for extended periods of time. It's not as bad as the TI-Nspire CX, who loses its charge after 2 weeks in deep sleep mode, but now it's running low on battery since its last charge last month, while my CSE once remained full even after half a year without a charge.
So yeah, if you store your calc in the drawer for the Summer, make sure to check every 2 week in case you need to charge it.
Quote from: DJ Omnimaga on July 13, 2015, 05:28:20 AM
By the way, something I found out about the CE is that it doesn't seem to be very good at keeping a charge when unused for extended periods of time. It's not as bad as the TI-Nspire CX, who loses its charge after 2 weeks in deep sleep mode, but now it's running low on battery since its last charge last month, while my CSE once remained full even after half a year without a charge.
So yeah, if you store your calc in the drawer for the Summer, make sure to check every 2 week in case you need to charge it.
I thought they advertised it would stay charged over the summer?
Indeed. But from what I saw, I think by the end of the summer, if my calc was not used for 2 months, it would probably have near dead batteries.
Hmm, maybe theres a ghost in your housing up the batteries :P I wonder if other people have had the same problem.
Does it help if you pull out the battery?
Quote from: Cumred_Snektron on July 28, 2015, 07:31:35 PM
Does it help if you pull out the battery?
That has helped me in the past when my calc wouldn't charge. I've done it ~3 times.
Leave the bat out for 3 minutes.
Do you mean removing the battery everytime the calculator is not in use or simply removing it for a few minutes then putting it back in?
Removing it when youre not using it. To see if the calc just uses a lot in stand by mode or the battery is just too weak
@DJ Omnimaga Removing it for 3 minutes once, that might fix it, until it happens again.
Ah I see. I didn't know that (I heard that depleting phone batteries to around 3% then recharging them again sometimes fixed such issues, but depleting them completely and leaving them like that for an extended period of time can be bad.)
DrDnar made an ASM lib that lets you invert the home screen, change text color and other things https://www.cemetech.net/programs/index.php?mode=file&id=1340
Here is the original topic: https://www.cemetech.net/forum/viewtopic.php?p=238138#238138 . This could be handy for people who wants to port DCSE homescreen games to the 84+CE, although they would obviously not look as good. I'm gonna give the tool a try later.
I should try that out, eventually. Currently, I believe I have one program that can run on the CE.
Cat Nipper? Because if it's pure-BASIC then it's guaranteed to run.
Its pretty nice the CE can run the same basic code as the CSE.
No, LettterDraw, its in Full on TI BASIC :)
Ah right, because Cat Nipper was also pure BASIC, right?
No, I had added some screen inverts.
Ah ok, I didn't see them in the screenshots.
Yeah, it was on the homescreen version ;)
There's an homescreen version? O.O
So I was wrong when I said that strings are maximum 32 KB large. It appears that they can be larger, but I think the max size is 65535 bytes.
Quote from: DJ Omnimaga on August 17, 2015, 10:27:34 PM
So I was wrong when I said that strings are maximum 32 KB large. It appears that they can be larger, but I think the max size is 65535 bytes.
Just like variables, or whatever unknownln used to store his sores with bejeweled?
Nah, just strings it seems. But I assume the largest file size for any variable or appvar would be 65535 bytes and more for Flash apps.
So I just sent OS 5.1 to my TI-84 Plus CE and I am impressed at how fast this update went compared to TI-84 Plus C Silver Edition OS 4.2 (less than a minute for the CE and over 6 minutes for CSE)
So yeah, basically OS files for calcs have the following extensions:
82+/83+/83+SE: 8xu
83+USB/84+/84+SE: 8xu
84+CSE: 8cu
84+CE: 8eu
83 Premium CE: 8pu
Interesting that the two CE models have different extensions, when the TI-83+ and 84+ shared the same one.
Quote from: DJ Omnimaga on September 08, 2015, 05:55:17 AMInteresting that the two CE models have different extensions, when the TI-83+ and 84+ shared the same one.
I suppose it was because before MP OSes, it was basically the same, whereas for the CE, a big chunk of math-related things (well, the exact math engine) isn't there, so it's sufficiently different to warrant a different extension, and make it very clear it isn't the same.
Well, if we compare the TI-83+ OSes with MP OSes, there were some drastic additions as well and the 84+ OSes from back then, while they didn'T have a lot of new extras, would not install on a 83+ and vice-versa.
Quote from: DJ Omnimaga on September 08, 2015, 07:13:11 AMWell, if we compare the TI-83+ OSes with MP OSes, there were some drastic additions as well and the 84+ OSes from back then, while they didn'T have a lot of new extras, would not install on a 83+ and vice-versa.
Sure, but even TI wouldn't change an existing file extension :P Especially since they're getting low on 3-letters 8** names now....
They should just use more letters, it's not like it's impossible. Quite the contrary....
Yeah they could use 4 letters if they wanted to. They'll have no choice at some point.
I guess thats true, unless they change names drastically.
Imagine if every single model number (eg TI-82, 82 stats, 76.Fr, etc) used different extensions. It would have become ridiculous O.O (and it wouldn't have been a surprising move from TI, back then.)
lol, and sharing program would have been crazy
Not necessary. I mean, we can even send TI-82 programs to a 84+SE via link port.
So yea. I tried the CE for a day when i borrowed it from my friend. I love the slick design and the almost instant file transferring, but i think the buttons are too blocky, and that the arrow keys are WAY to small. Also, the menus and stuff still seem sluggish like on the SCE.
Quote from: c4ooo on October 30, 2015, 04:44:43 PM
So yea. I tried the CE for a day when i borrowed it from my friend. I love the slick design and the almost instant file transferring, but i think the buttons are too blocky, and that the arrow keys are WAY to small. Also, the menus and stuff still seem sluggish like on the SCE.
D:
Quote from: Araidia on October 30, 2015, 05:21:31 PM
Quote from: c4ooo on October 30, 2015, 04:44:43 PM
So yea. I tried the CE for a day when i borrowed it from my friend. I love the slick design and the almost instant file transferring, but i think the buttons are too blocky, and that the arrow keys are WAY to small. Also, the menus and stuff still seem sluggish like on the SCE.
D:
Well, tbh, it is
much better then the CSE. Also, tbh, i dont realy care about the blocky buttons, (ime just used to rounded ones :P ) Yet i still think the arrow keys are too small.
If some one was just getting into calc programming, i would recommend the ti84+, although my main resone for doing so is becouse the 84+ has more resources and help :P
IIRC PHASM will soon allow jumping to labels, so perhaps that could solve parts of the slow menu issues.
So I was wondering who here on CW had a 84+CE or 83PCE already and who is planning to get one?
And who plans to program on them in the near future, especially now that an online C compiler is arriving?
When there is an emulator available, i will centrainly try my hand at coding :)
Does it mean that if no emulator ever comes out (which is a possibility, considering that's what happened with the Casio fx-CG20), 84+CE development will be almost dead or sparse like HP/Casio dev? It would be a shame if that new calc never took off because of that >.<
Perhaps there could be some collab thing setup for trusted community members where people can submit their programs for testing, and there could be volunteer 84+CE owners to beta-test them to see if bugs are fixed? That wouldn't be ideal but it would be better than nothing. Plus TI-Planet C compiler features group projects.
I'm positive that there shall be at least one emulator for the TI-eZ80 series :)
Technically, the eZ80 isn't that much more complex than the Z80 is, and it's definitely much simpler than most other ISAs from the 2-3 last decades. Likewise, the devices integrated into the ASIC can be emulated, they're simpler than various devices emulated by e.g. QEMU.
Emulating the USB controller faithfully enough shall be a departure from past TI-Z80 emulators, but there's no alternative: the TI-eZ80 series doesn't have a legacy I/O port. TI-Z80 emulators would also have to bite the bullet of implementing emulation of the older USB controller, if faithful emulation of the 82A became a goal.
I hope, because people hoped for a FX-CG20 emulator when it came out and 5 years later, still none from the community, and while the ez80 isn't that much complex to emulate than the z80, what about the various protections from TI, such as the RSA keys? Also would the lack of an I/O port prevents such emulator from having the ability to receive files by drag-n-drop for the few years it would take to figure out how the USB controller works?
I don't want to pressurize anyone to make such emu, but from what I could gather on CodeWalrus, an TI-84+CE emulator is a prerequisite to convince our members to adopt the TI-84+CE as a development platform, even in TI-BASIC, and remember what happened with the PRIZM a few years after it became clear that no community emulator would ever come out: Game development for that platform came to a nearly complete halt. Could an existing base be used for the emulator, such as z80e, WabbitEmu or TilEm, for temporary purposes? TI-Smartview isn't an option since it doesn't emulate undocumented stuff, and jsTIfied is closed-source, so only its main author can update it and that's if a CE update is still being considered.
Also, is it even possible right now to dump a ROM from a CE?
Even if an emu doesn't emulate USB at first, I suppose that it could, until it's properly figured out, mess with the RAM directly, but that's quite hacky, so the earliest USB is understood, the better.
That could probably work. I think that'S what WabbitEmu does, right? What about the Flash? IIRC with Wabbitemu the main problem is that sending APPS was a PITA, often requiring restarts because the apps menu would remain empty. But regular files seemed to send fine overall. It would be better than nothing, since there is no community app for the CE anyway.
It's kind of a shame that the future of CodeWalrus, Cemetech and TI-Planet ez80 ASM/BASIC/C programming community depends entirely on the release of an emulator, though (judging by the abysmal amount of CE releases compared to when the CSE came out). I kinda wish that people would just get the calc and test on-calc, like everyone did with the TI-82, 85 and 92 back in the days. I personally made First Fantasy: Mana Force CE entirely on-calc and only used an emu for screenshots.
From https://www.cemetech.net/forum/viewtopic.php?p=242220#242220
It looks like programs that uses xLIBC in half-resolution mode, as well as Celtic 2 CSE, will be supported. Basically, FF:MF CSE, which abused half-res mode, would not be supported, but there is already a CE Textlib version of the game out anyway. But if DCE 9 works out and that Kerm doesn't get hit by roadblocks during development (or real life), then that is great news for xLIBC users, since that would mean that not many modifications have to be made to your games to run on the CE (I assume most modifications will involve slowing down the game :P)
I wonder if this could run fast enough on the TI-84 Plus CE in 8 bits mode (or lower, if available)? http://www.ticalc.org/archives/files/fileinfo/333/33316.html
Here is the source code: http://www.ticalc.org/archives/files/fileinfo/409/40992.html
From what I remember in 2004, the camera capture of the second animated GIF was speed up, while the emulator one was not, and this was before the TI-89 Titanium came out, meaning that the max speed showcased was 12 MHz.
The TI-84 Plus CE can go up to 48 MHz, right? I wonder if this could be ported to the CE and possibly at an higher resolution (or doubled-pixels) without massive slowdowns?
This engine could be handy for a port of Big Rigs: Over the Road Racing :trollface:
Does anyone remember how much RAM does the CE has? I know it has 150 KB of user RAM, but I forgot how much in total it has (eg 256, 512, etc)
The CE has 256Kb of available ram, plus 150Kb of video ram. Hope this helps! :)
Ok thanks for the info, and I didn't know its video memory was separate. :)
Quote from: DJ Omnimaga on February 24, 2016, 12:26:37 AM
Ok thanks for the info, and I didn't know its video memory was separate. :)
It's actually not known if it *is* separate, (I personally believe that they just encoded it on the same die), but you never know :)
Oh ok. TO be honest I was surprised and I think it might be part of the same chip, but still memory-mapped display. Perhaps it's a chip combo?
There is a new BASIC lib that recently came out here: https://tiplanet.org/forum/archives_voir.php?id=510643 (original topic: https://tiplanet.org/forum/viewtopic.php?f=12&t=18327 )
It displays 10x10 sprites (you can store 5 sprites at a time) one by one. I am trying to convince the author to implement the ability to chain sprites (eg display multiple ones in one command, like CE Textlib can do with rectangles and other stuff) and zoomed sprites display (eg 2x, 3x, 4x, etc, via a 4th argument). If that got implemented, then this lib could be a good alternative for hybrid BASIC programmers since DCE9 isn't out yet.
This isn't really big news and perhaps not related to TI-84 Plus CE programming, but TI just posted a picture of TI-84 Plus CE calculators still in production at a factory:
https://twitter.com/AroundTI/status/732597838543216640
(https://pbs.twimg.com/media/Cipl9HYUUAAVTpW.jpg)
I do not know if that is the assembling process or if the flashing process is separate, and I don't know if they flash calculators with OSes in a different building or not, but it's still interesting from them to show this to us :)
https://tiplanet.org/forum/viewtopic.php?f=41&t=18398
Woah, thats kinda cool... That looks like they are testing whether they turn on...
Yeah, basically they use some sort of robot with three "fingers" that presses the MODE+ALPHA+LN combo and stuff to put the calcs in self-test mode.
Interesting, do you know how they check that it works? Do they have some sort of color sensor?
On a side note,
@unknownloner on IRC last night mentioned something about the TI-84 Plus CE having support for 1 hardware sprite that is up to 64x64 pixels large that can contain up to 2 colors. How does it work? Is it a bit like the hardware scrolling on the TI-84 Plus C Silver Edition and can that sprite be moved anywhere on the screen with zero speed drop, regardless of its size? Or does it work like Atari 2600 sprites?
Quote from: DJ Omnimaga on June 20, 2016, 03:32:04 AM
On a side note, @unknownloner on IRC last night mentioned something about the TI-84 Plus CE having support for 1 hardware sprite that is up to 64x64 pixels large that can contain up to 2 colors. How does it work? Is it a bit like the hardware scrolling on the TI-84 Plus C Silver Edition and can that sprite be moved anywhere on the screen with zero speed drop, regardless of its size? Or does it work like Atari 2600 sprites?
This is incorrect. It was originally believed that the lcd hardware cursor could be used for this purpose, but through much research into it for the creation of CEmu it was determined TI messed something up and it does not work at all.
Yeah I think the cursor is what he might have been talking about, or at least I suspected it had something to do with a TI-OS feature. Sad that it doesn't work, though :(
Quote from: MateoConLechuga on June 20, 2016, 03:41:25 AM
Quote from: DJ Omnimaga on June 20, 2016, 03:32:04 AM
On a side note, @unknownloner on IRC last night mentioned something about the TI-84 Plus CE having support for 1 hardware sprite that is up to 64x64 pixels large that can contain up to 2 colors. How does it work? Is it a bit like the hardware scrolling on the TI-84 Plus C Silver Edition and can that sprite be moved anywhere on the screen with zero speed drop, regardless of its size? Or does it work like Atari 2600 sprites?
This is incorrect. It was originally believed that the lcd hardware cursor could be used for this purpose, but through much research into it for the creation of CEmu it was determined TI messed something up and it does not work at all.
Aww that sucks :(
Well, it's still 1024 bytes of free RAM with very low wait state, so... yay?
Could those be used for ultra fast low-quality graphics that need to be refreshed fast?
Question: Does the TI-84 Plus CE-T (the European version of the TI-84 Plus CE released outside France) have ASM/flash app support? I am wondering, since the TI-84 Plus T doesn't support ASM programming: https://tiplanet.org/forum/viewtopic.php?f=41&t=17651
Yes, only the 82A and 84+T are the ones without third-party native programming support.
The 84+CE-T is basically an 83PCE without the exact math engine.
Ah ok.
I wouldn't be surprised if the TI-84+SE eventually got replaced in North America with a model with no exam light, but also no ASM/app support and 1 MB of RAM instead of 1.5... It seems like TI only wants their higher end color ez80 calculator to be able to do advanced programming...
OS 5.2.1 has been released now. It's unclear what changed in it, though: https://tiplanet.org/forum/viewtopic.php?t=19052&p=207656#p207656
Update: What has changed apparently in OS 5.2.1 for the TI-83 Premium CE is that they fixed a major bug with math suites.
Good to know.
I have updated the first post of this topic (which is stickied) with extra programming information. It might be very incomplete but I guess it's a start for those who wants to start programming. It now includes links to wikis and resources for BASIC, ASM, C, ICE and hybrid BASIC. Feel free to suggest additions.