TI-Nspire OS 3.2 is out!

[Jim Bauwens]

Some of the new bugs have been fixed.

[Lionel Debroux]

From what I read on the nspire teachers site, 3.2 has caused new bugs which is not good news.

It did, indeed. But did anyone seriously expect better from TI, despite the delay of three months ?

[adriweb]

Yep, in 3.2 : new bugs came, some old ones were fixed and some others (but sometimes there are even on the 68k) were not.

[Dingus]

I guess not. 

From what I read on the nspire teachers site, 3.2 has caused new bugs which is not good news.

It did, indeed. But did anyone seriously expect better from TI, despite the delay of three months ?

[ruler501]

What are the bugs?

[adriweb]

Slowness in Basic loops, still some CAS bugs, and still a bug that was in 3.1 about a random reboot with 3D graphs

[ElementCoder]

Has anyone else noticed the dramatic decrease in speed in the notes app and TI-BASIC (and probable some other places too)? It's unbearable. Loops, for example, are very slow with printing results to the screen, really annoying since I was developing something that does that a lot. I read that the slowness in the notes app is caused by bad XML, it's a tree with a leaf for every word (and a word has a lot of params). I think I'll return to OS 3.1 again for now...

This loop finished in +/- 20 sec on OS 3.1 and in approx 2 min. ( ) on OS 3.2

Code: Select All | Copy To Clipboard

1 2 3 4 5

Local i For i,1,100 Disp i," ",(i^2)-1 EndFor

grilled : https://groups.google.com/forum/?fromgroups#!topic/tinspire/_mB8zLLoCLg

I said I read it somewhere.  I'm not claiming I discovered it, I read it, ran the example on my calc and decided to post it, and then forgot where I read it. So if it seems like I claim it as my discovery, sorry for that , I was just pointing to it.

[Jim Bauwens]

Ah, don't worry about it

I have send some benchmarks to TI regarding this bug a couple of days ago, and they are looking into it.

[Levak]

I said I read it somewhere.  I'm not claiming I discovered it, I read it, ran the example on my calc and decided to post it, and then forgot where I read it. So if it seems like I claim it as my discovery, sorry for that , I was just pointing to it.

But in the thread you read, there were two questions. One regarding the slowiness of the Disp function, and another one regarding the Notes app. The XML tree problem is in the Notes app. I have never looked closely how the Calculus app was encoded.

[Sebasu]

wow 3.2 is slowest!!!

Local i
For i,1,100
Disp i," ",(i^2)-1
EndFor

in os 3.1 it takes 14 sec to complete but in os 3.2 it takes 51 sec
it happen too with the 'While' function
what happen to TI?
fortunately OS 3.1 can open 3.2 files, then we can use scriping software 3.2 whith api 1.0 and use this in calculators with OS 3.1
now... os 3.2 has new interesting functions that someone could need then:
depending of what kind of use anyone want, will use 3.1 or 3.2

i will test more... i think that the problem can be the 'Disp' function and not the loops
because, the progams that i did some time ago with no loops, takes some time to display al text that it has to

[cyanophycean314]

That is a pretty big slowdown... 
Basic, right?

[Jim Bauwens]

It is the Disp function that is the problem. How more arguments how slower. TI is aware of the issue and looking into it.

[Sebasu]

Basic, yes
I just try my program in the system 3.1 and the print speed is dramatically faster than in OS 3.2
we'll have to wait for the next update to correct this horrible bug

[Dingus]

If TI tested 3.2, wouldn't it be apparent on some level that it is much slower than 3.1.  And if that is true, does that indicate that TI doesn't do much testing but counts on the customer to find their screw ups for them, for free of course.  That's the only way I can understand why the os updates are constantly flawed, and I wonder how math teachers have time to do things like develope lesson plans when they are constantly reacting to new os bugs and other hardware issues if they use nspires to teach.     

It is the Disp function that is the problem. How more arguments how slower. TI is aware of the issue and looking into it.

[adriweb]

I actually wonder how nobody testing the BASIC stuff saw that o_O