2.21 and 2.22 are far from stable. You shouldn't use them, ever. They have sequence graphing bugs (patched with http://brandonw.net/calcstuff/seqfix.zip if memory serves).

Is that why they redid the graphing in 2.30 and onwards?

Hm, can anyone give more details about the "sequence graphing bug"?
I'll admit that I hardly ever graphed sequence equations before.
I just tried it and didn't notice anything special even when tracing it which seems that the problem was from in the readme.

That being said, I still won't upgrade from OS 2.22. I don't exactly know what is meant with unstable. Does it mean random RAM clears? If so, I really didn't notice unless I used buggy programs.

Really now, why should I trade in my fast graphing speed for an USB activity hook, TI peripheral support, PTT, some new features (which I don't use) and mathprint.
The first three are utterly useless to someone who uses his graphing calculator for what it is, a graphing calculator.

There are only three reasons I can think of for why you should upgrade as a "casual user":


-Noteworthy "new" features from http://wikiti.brandonw.net/index.php?title=83Plus:OS:VersionDifferences
#  "manual-line fit"
# improved graphing of functions with discontinuities when Xres=1 (the reason of the graphing slowdown)
# new statistics features:

    * chi-square goodness of fit test
    * confidence interval for slope
    * inverse student T
(excluding features from OS 2.53)


-Stability
I'll admit that it's certainly possible that OS 2.21/2.21 is more unstable than I make it seem. To be honest I really don't have any problems with it.
Maybe there is stuff that's just buggy to use but if it is, I didn't use it up until this point. Simple as that.


-OS 2.53
http://wikiti.brandonw.net/index.php?title=83Plus:OS:OS_2.53MP_Changes
This is by far the most visible change that TI has done. You hate it or love it.
I loved the way that you can have a visual representation of the function "fnint(" for one.
I hated the nagscreen after every ram clear and the extreme slowdown when using mathprint.
This was the change I actually considered. Until I wrote a program that does exactly the same thing as mathprint does with fnint(.

Personally I upgrade to the latest version when possible, but in the 84+ case, I would not go further than 2.43, the latest most stable OS. I do not like MP very much and due to bad experiences I had with early OSes for a new calc, I would be scared of using OS 2.21 or the like on the 84+. When the 83+SE came out, I had OS 1.13 on it and it was by far the most unstable TI-OS I ever used. (e.g: RAM Clears when scrolling the PRGM list from the BASIC editor when you have over 50 or so). However it is your choice to not upgrade. If you never use your graphing calc to program or make games, it may not matter much to you as much as for us, calc game programmers.

Mostly it's the homescreen that got slowed down, since the OS has to update whatever you type in into its "mathprint". Also, is seems to crash randomly after I type certain sequences of tokens (in other words, instead of the classic ERR:SYNTAX, some combos (I don't remember which) just caused a RAM clear ).

I have 2.53 MP, and I'd like to mention that I haven't really experienced any slowdown, probably because I run mostly only my programs, which aren't usually reliant on speed for effective gameplay.  Could anybody provide an example of a game that runs noticeably slower on 2.53 MP?

Besides, doesn't it act like the older versions if you change it to classic mode (goes faster, e.g.)?

The nag-screen is pretty annoying, though. 

Try going to the equations screen (Y=) with mathprint enabled.
I don't think that the ability to turn it off salvages it. From what I remember the shortcut menus are generally unusable when mathprint is disabled. The shortcut menus are pretty much the only reason why I'd upgrade to OS 2.53 in the first place. It seems to add unneeded complexity when you need to press a switch before you can go from a shortcutmenu to the equations screen. The fact that the ability to enable/disable mathprint is hidden deep in the mode screen doesn't help either.

Quote from: Michael_Lee on September 11, 2010, 07:22:41 pm

I have 2.53 MP, and I'd like to mention that I haven't really experienced any slowdown, probably because I run mostly only my programs, which aren't usually reliant on speed for effective gameplay.  Could anybody provide an example of a game that runs noticeably slower on 2.53 MP?

Besides, doesn't it act like the older versions if you change it to classic mode (goes faster, e.g.)?

The nag-screen is pretty annoying, though. 

Try going to the equations screen (Y=) with mathprint enabled.
I don't think that the ability to turn it off salvages it. From what I remember the shortcut menus are generally unusable when mathprint is disabled. The shortcut menus are pretty much the only reason why I'd upgrade to OS 2.53 in the first place. It seems to add unneeded complexity when you need to press a switch before you can go from a shortcutmenu to the equations screen. The fact that the ability to enable/disable mathprint is hidden deep in the mode screen doesn't help either.

I notice no slow down when mathprint is disabled, personally, but with Mathprint ON it is terrible. Also I heard there were bugs in the OS anyway so unless you absolutely need it, I really recommend sticking to another OS (preferably 2.43, or if you absolutely need faster graphing, 2.21 or something)

The only advantage to "MathPrint" is that it's editable in real time and in said form. With PrettyPrint you have to edit linearly and can't evaluate the expressions.
Not that that's very important PrettyPrint also has copy/paste abilities

Yeah it stopped working completly I think. Celtic III worked, but some xLIB functions in Celtic III has minor compatibility issues. Doors CS solve the issue, though.