Omnimaga
Omnimaga => News => Topic started by: JosJuice on February 01, 2011, 01:41:15 am
-
Casio has released version 1.02 of the Prizm OS on their website, along with trial software for the Prizm. According to Casio it adds "new 10 eActivities!!".
However, it appears that this OS is a different version of the 01.02.0200 OS that was pre-loaded on certain calculators, because these calculators did not have the Locate command glitch with non-ASCII characters, but this copy of OS 01.02.0200 re-introduces it, so do NOT update if you already have 01.02.0200 installed!!!
https://edu.casio.com/download_service/download/category.php (https://edu.casio.com/download_service/download/category.php)
The update is only compatible with the fx-CG10. Nothing is known about when the fx-CG20 will be updated. The update file is a zipped exe (about 11 MB).
-
Casio did a great job hiding the OS update behind the installer. It's quite clever, actually. I expect it'll take me an extra hour or so to circumnavigate it.
-
11 megabyte installer ??? The OS update it self is probaly no more than a few dozen kilobytes. I hope the installer doesn't use something like rsa keys and extra checksums to update the OS. Also I hope the OS isn't encryted inside the installer. Why can't Casio be like TI and send in a .8xu format. Or to follow the format of other Prizm files .g3x format (if that existed) :P
-
Woah it was about time! It was already available on some calcs for a month, yet they wait a month before releasing that stuff. Oh well, at least I'M glad it's out. I'm gonna upgrade.
Casio did a great job hiding the OS update behind the installer. It's quite clever, actually. I expect it'll take me an extra hour or so to circumnavigate it.
Yeah it was the same when I upgraded my FX-9860G to OS 2.0 (which pretty much turns the calc into a FX-9860GII without a backlight).
-
Wow seems like they actually care about fixing bugs O.O
-
Indeed. Of course I am certain there are still many, knowing them, but at least the fact they fix them is a start. On the 83+ we have been stuck with ERR:VERSION for 12 years. :/
-
Well well well. :thumbsup: to Casio (sorta). I wanted to see if there was a difference between 1.01 and 1.02 in the way of speed and yes there was. The program I ran was a for loop from 0 to 1000 and used the locate command to output A.
In OS 1.01 the times I got were:
42.61
42.60
42.60
Now in OS 1.02 I got:
42.33
42.27
42.26
So I speed increase. Yes. A big one. No. But it might help speed up a few basic programs a little bit.
-
That's good, however I got some pretty bad news: Casio did not fix the Locate glitch :( :( :(
Try running the following code:
For 1->A to 7
For 1->B to 21
Locate B,A,"€"
Next
Next
The bug still happens: the last columns of text remains missing until you hit AC/ON to break execution. :(
So I guess finally this OS adds little changes and no bug fixes. :(
-
@DJ I tried your program, but I did not get any glitches. Did you mean to close the quotes because if I leave them open I get a syntax error after the calc draws the second character
-
DAMNIT! Then it means my calc is defective and was since the very start :'(
I'Ll have to wait until they get in store and get another one...
You're sure you typed a character like the Euro symbol, @, theta, omega, etc? And yes I closed the quotes. Could you attach a copy of your program? I want to see what you did to not get the glitch.
EDIT: Are you sure you did what I said? I just tried my program in the emulator and still had the glitch happening. Here's a screenshot to prove it (first with a regular character, where the glitch won't occur, then I change it to a non-ASCII char, and it happens):
-
I had the program typed exactly the way you had it with the euro character. Try asking qwerty and FinaleTI if they have the same glitch you have. Also I have ran some weird apps (including one that almost bricked my calc) on my Prizm before. I wonder if that could in any way affect the glitches.
Edit: Did the original version of your code read For 1->B to 2. Because that is what a typed in the first time. I tried it with 21 now and I got the glitch where the calc draws the first half of the screen then updates the second half
-
Check the post above. I added a screenshot. The glitch occurs on the emulator too. Possible scenarios:
(http://www.omnimaga.org/index.php?action=dlattach;topic=6428.0;attach=5905;image)
1) Casio released the wrong version of the OS and misnamed 01.01.0200 as 01.02.0200
2) Casio released two different versions of OS 01.02.0200 and calcs that came preloaded with it got a version without the glitch (your calc being an example)
3) The apps you ran magically fixed the glitch
4) Your calc is special and the glitch doesn't occur on them.
If the latter is the case, then a crapload of people will get ripped!
-
Yeah, it happened with my Prizm too (although it didn't before) :(
Of course, I've messed with my Prizm probably as much as you have, z80.
-
Did you have 1.02 preloaded on it?
@Z80man No I had 21. It was just a typo in my post above.
-
Yeah. Originally the calc was immune to it, but now it displays the glitch.
-
When did it start displaying it? Did you load the new OS update on it? It seems like the OS might be different than what you had... (even if the numbers are the same)
-
Okay so it looks like everyone has the glitch. Weird though that the emulator has it too because its not even a real emulator.
-
I did load the new OS to test it and I got a "this OS version is the same as what you have" message.
-
I did load the new OS to test it and I got a "this OS version is the same as what you have" message.
Did it install it anyway or did it just refuse to start the install? Did you see any progress bar? If so, then it might still have updated it anyway.
-
Yeah, I got the full progress bar.
-
So Casio fixed the glitch and then they decided to unfix it for the OS release. Way to go Casio at least this OS is like .1% faster
-
Well, basically they must have updated the OS but they updated from 01.01.0200.
I added a red text in the first post on front page warning people to not upgrade if they already had 01.02.0200 installed.
-
11 megabyte installer ??? The OS update it self is probaly no more than a few dozen kilobytes. I hope the installer doesn't use something like rsa keys and extra checksums to update the OS. Also I hope the OS isn't encryted inside the installer. Why can't Casio be like TI and send in a .8xu format. Or to follow the format of other Prizm files .g3x format (if that existed) :P
Whoa, I didn't think we'd ever say casio should be like TI. We'll try and crack it.
Should we send CASIO about this bug?
-
11 megabyte installer ??? The OS update it self is probaly no more than a few dozen kilobytes.
The Prizm flash chip is 32 MB. 16 MB of this can be used by the user? What are the other 16 MB used for? Most likely the OS. 11 MB seems like a realistic size for the OS if 16 MB is available for it. Of course, the OS might be a bit smaller than the PC installer... 9-10 MB, maybe?
-
11 megabyte installer ??? The OS update it self is probaly no more than a few dozen kilobytes. I hope the installer doesn't use something like rsa keys and extra checksums to update the OS. Also I hope the OS isn't encryted inside the installer. Why can't Casio be like TI and send in a .8xu format. Or to follow the format of other Prizm files .g3x format (if that existed) :P
Whoa, I didn't think we'd ever say casio should be like TI. We'll try and crack it.
Should we send CASIO about this bug?
Done.11 megabyte installer ??? The OS update it self is probaly no more than a few dozen kilobytes.
The Prizm flash chip is 32 MB. 16 MB of this can be used by the user? What are the other 16 MB used for? Most likely the OS. 11 MB seems like a realistic size for the OS if 16 MB is available for it. Of course, the OS might be a bit smaller than the PC installer... 9-10 MB, maybe?
Really? O.O I wonder if the OS really uses all the other 16 MB... I seriously hope the rest of the memory isn't used to store screen data... (that would explain why it's so slow)
Either way if they don't fix the locate glitch or simply cannot replicate it at all or worse: can't figure out what I mean, then I guess bug fixing will be in the community hands or we'll have to wait until someone writes a new language for BASIC programmers.
-
Since this update doesn't appear to fix the bug, I'll stay with 1.01 if you guys want until there's a glitch fix or a dumping program is created.
-
11 megabyte installer ??? The OS update it self is probaly no more than a few dozen kilobytes.
The Prizm flash chip is 32 MB. 16 MB of this can be used by the user? What are the other 16 MB used for? Most likely the OS. 11 MB seems like a realistic size for the OS if 16 MB is available for it. Of course, the OS might be a bit smaller than the PC installer... 9-10 MB, maybe?
Really? O.O I wonder if the OS really uses all the other 16 MB... I seriously hope the rest of the memory isn't used to store screen data... (that would explain why it's so slow)
Based on what I saw in the installer, the OS seems to be about 6 MB, but I'm not sure yet.
-
But then what would be the remaining 10 MB for? Did they just leave 10 MB unused for future additions or something? ???
-
But then what would be the remaining 10 MB for? Did they just leave 10 MB unused for future additions or something? ???
Most likely. It wouldn't surprise me if they did.
-
I guess for now it could be used for some ASM/C programs or a mod could be written that increases the user archive by 10 MB, but then it's not guaranteed to continue working on newer OSes.
-
OS? We can now know more about the prizm disassembling it!!!!
;D
-
This is gonna be hard to look through the 11 MB file though O.O
-
I'm trying to isolate the OS, but that would still leave 6MB to work through. So far my strategy has been to disassemble the Prizm usb drive using Hxd. Then trying to match hex strings between that and the exe.
Yeah 200th post!!!!!!!!!!! :w00t: :hyper: :w00t: :crazy: :hyper: :crazy: :w00t: :hyper:
-
Hmm I see. I hope the OS isn't 11 MB, tho O.O
Also congrats on 200th post! I just reached 23000 so you only have 22800 to go. ;D j/k
-
Z80man, I hate to tell you this, but that strategy will almost certainly fail. The reason is that the USB dump isn't actually a dump of the Prizm's internal memory. It's a dump of the virtual memory, which is entirely overwritten during USB transfer to make room for that folder that appears on your desktop. Take a closer look at the file structure. You'll notice that everything is organized into the directories and the sections are each named with the name of the folder. The files from each folder are likewise under in the same sections.
In other words, it's the hex version of what you see on the screen.
Also, I took the liberty of simply disassembling the exe to look for valid code and there's nothing that I would call valid in the entire file. I suspect compression, but I don't know how that would work.
-
Z80man, I hate to tell you this, but that strategy will almost certainly fail. The reason is that the USB dump isn't actually a dump of the Prizm's internal memory. It's a dump of the virtual memory, which is entirely overwritten during USB transfer to make room for that folder that appears on your desktop. Take a closer look at the file structure. You'll notice that everything is organized into the directories and the sections are each named with the name of the folder. The files from each folder are likewise under in the same sections.
In other words, it's the hex version of what you see on the screen.
Also, I took the liberty of simply disassembling the exe to look for valid code and there's nothing that I would call valid in the entire file. I suspect compression, but I don't know how that would work.
What I was thinking though is that if there is some hex string inside the usb folder then the OS must have written that there earlier. Then that string of data might be in the exe file somewhere. Not in the same location, but that string might exist. And so far I've already found a few 16 byte hex strings that match between the two.