Well with the alternate download it worked fine now. It's with the TI-Planet version (without the vertical tab things to the left) that I had problems.

At one point it might be good if I or someone else made a topic in this section listing the essential tools and doc to get started on HP Prime programming.

For example they are there with Time units but not with Volume or Distance

Small bug but annoying all the same.

The issue here is basically that the grid chooser shortcuts are supposed to follow a physical layout. However, the keys just run out if you have more then 16 (the grid between the 7 and + keys).

Three options we available in my mind:
1. Just do a normal "numbering" like the list choosers where it does 1-9, A-P, and not worry about matching the physical arrangement like seen on the screen.
2. Do the physical layout but be limited to that 4*4 grid. If more, don't have them.
3. Do #2 but then do A-P for others.

I chose to implement #2 after investiagating the others and testing. It was the one that most people seemed to agree seemed the best. Would be happy to revisit #3 (or another variant) if enough people seem to think it is a good idea.

Comments?

I finally got my HP Prime (with current firmware)!

Anyway now for new bug/issues/etc reports, unless I am really doing something wrong:

1) (Emulator) Sending a program that is too large (for example 23 KB large) freezes the emulator most of the time and in all cases it will never actually send. The only way to send programs this large is to open them in the connectivity kit then copy the code in a new program on the calc, but that can be annoying since it takes a long while to add the program to the calc.

2) (Emulator) Often, the emulator will stop detecting the calc randomly, requiring you to unplug then plug back the calc every few second or so in order to send programs.

3) (Emulator) Not really a bug, but the reset option is very misleading there: On the real calc, a reset will just reboot the machine. However, on the emulator it really wipes out the entire memory. It might be better to add a warning asking you to confirm if you really want to lose all progress.

4) (Connectivity kit) When saving large programs, it can take multiple tries until it finally gets updated on the calc or appear there. Also it takes a very long while before the calc content appears in the list.

5) (Calculator)  BLIT_P doesn't work properly on calc in some cases. In my Tunnel program, for example, in the intro that stretches the screen up, even if I tell it to update the entire screen it only updates parts of it most of the time. After running other programs, however, sometimes it works fine, but then you run the Tunnel game again and the glitch comes back. ALso, after a fresh reset the stretching effect won't even work at all. Or maybe it's an issue with DIMGROB?-EDIT: Figured out why I had this happening. Solved.



6) (Calculator) In certain cases, the Tunnel game and pretty much every other program will stop working past the stretching animation at the beginning, giving an ERR: Argument or Invalid Input errors. I don't know why this happens yet, though, but I had this happen with another program before.

7) (Calculator) In certain cases, the program catalog screen will remain frozen and accessing the APPs menu will be impossible, so that you no longer know what you are selecting (although the navigation at the bottom of the screen will still update depending of what you selected). I believe both issues #6 and #7 are more common when exiting programs with ON, and in both cases, you need to reset the calc with ON+SYMB (or using the reset button on the back).

(Cosmetic issues) The calculator slidecase is a little bit too loose, so the calc can move slightly in it. On the other hand it holds the calc pretty well, so no risks of falling off it.

9) (Cosmetic issues) It is very hard, sometimes impossible to read the yellow text on the calculator keys under certain type of lightning

10) (Cosmetic issues) There's no key to access semicolons, which is a big problem considering almost every line of code ends with a semicolon. Having to go to the Charmap everytime is annoying.

11) (Cosmetic issues) The reset hole is too small. It should be made so at least a pen can fit. That said, a nice workaround is the ON+SYMB combo, but that can be a bit problematic if someone is at school and forgets what the key combo was. I had plans to mod my calc case with a knife to stretch up the hole, but since there is ON+SYMB I'll probably not bother.

Oh wait didn't see that one lol. Since there is no Alpha key for ./= and Space/_, I totally missed the lonely semicolon to the right.

I'll post a video of #5.

EDIT: Here is it (however in the video I removed the wait command in the tunnel code so it's as fast as in Critor's vid):



EDIT 2: Source code:

EXPORT Tunnel()
BEGIN
DIMGROB(G1,320,240);
BLIT_P(G1,0,0,320,240,G0);
FOR Z FROM 2 TO 239 STEP 3 DO
BLIT_P(G0,0,0,320,240,G1,0,0,320,240-Z);
WAIT(0.01);
END;
1▶C;
0▶P;
100▶Y;
20▶T;
128▶L;
0▶L1(512);
 
FOR Z FROM 0 TO 16 DO
INT(RANDOM(100))▶L1(2*Z+1);
60-INT(RANDOM(4))▶L1(2*Z+2);
END;
RECT_P(0,0,319,219,RGB(30,100,255),RGB(30,100,255));
RECT_P(0,35,319,219,RGB(0,100,0),RGB(0,100,0));

TEXTOUT_P("2013 DJ Omnimaga  (www.omnimaga.org)",122,25,1,RGB(255,255,255));
FOR Z FROM 0 TO 10 DO
TEXTOUT_P("Score:",Z,0,7,RGB(25*Z,255-(25*Z),0),250);
TEXTOUT_P("τ υ η η ε ι  ρ ɾ ί м ɛ",122+Z,0,7,RGB(0,25*Z,0),250);
END;
FOR Z FROM 0 TO 4 DO
LINE_P(0,Z+35,319,Z+35,RGB(0,20*Z,0));
END;
WHILE Y>0 DO
FOR Z FROM 0 TO 15 DO
C+1▶C;
IF C>16 THEN
1▶C;
END;
L1(2*C+2)▶O;
L1(2*C+1)▶Q;
RECT_P(20*Z,40,20*Z+19,O,RGB(0,50+Q,0),RGB(0,50+Q,0));
RECT_P(20*Z,O,20*Z+19,O+L,RGB(100+Q,130,50),RGB(100+Q,130,50));
RECT_P(20*Z,O+L,20*Z+19,246,RGB(0,50+Q,0),RGB(0,50+Q,0));
IF Z=0 THEN
RECT_P(0,Y,8,Y+12,RGB(0,0,0),RGB(150,150,255));
RECT_P(2,Y+4,14,Y+8,RGB(0,0,0),RGB(255,0,0));
IF Y<O OR Y+12>O+L THEN
−9▶Y;
END;
END;
END;
TEXTOUT_P(P,70,2,6,RGB(0,0,0),200,RGB(30,100,255));
TEXTOUT_P(P,72,2,6,RGB(5,255,255));
P+1▶P;
L1(2*C+2)▶S;
INT(RANDOM(100))▶U;
S+INT(RANDOM(30)-15)▶Ans;
IF Ans<45 THEN
45▶Ans;
ELSE
IF Ans>213-L THEN
213-L▶Ans;
END;
END;
C+1▶C;
IF C>16 THEN
1▶C;
L-1▶L;
END;
Ans▶L1(2*C+2);
U▶L1(2*C+1);
Y-(5*ISKEYDOWN(2))+(5*ISKEYDOWN(12))▶Y;
WAIT(.01666);
END;
FOR Z FROM 0 TO 7 DO
INVERT_P();
WAIT(0.05);
END;
DIMGROB_P(G1,122,30);
BLIT_P(G1,0,0,122,30,G0,0,0,122,30);
RECT_P(RGB(30,100,255));
REPEAT
GETKEY▶Z;
BLIT_P(G0,INT(RANDOM(10)),INT(RANDOM(10)),320-INT(RANDOM(10)),240-INT(RANDOM(10)),G1,0,0,122,30);
WAIT(.02);
UNTIL ISKEYDOWN(4)=1 OR ISKEYDOWN(30)=1;
END;
END;

UPDATE!

I managed to solve my issue with BLIT_P and DIMGROB_P. It was a stupid mistake from my part because I forgot the extra P at the end of the DIMGROB command. My bad. So I added strikethrough on issue #5, which you can disregard for now.

Let's see what Tim has to say (assuming he's allowed to say anything), but while it would be great to see the bugs fixed very quickly, it's not very likely that (in the absence of extremely severe bugs) a new OS release is going to be published next week, "only" two months after the previous release. The end of 2013 or beginning of 2014 would be a more realistic timeframe.

In general, large pieces of software need some QA (beta-testing, etc.), often several weeks, before release, even with an automated Continuous Integration system which helps prevent regressions (in the parts covered by the test suite). TI is known to perform weeks of beta-testing (timestamps embedded into a new OS upgrade are usually 3-4 weeks old at the time the OS is published), though I find it dubious whether they perform automated testing (or at least, their test suite is persistently inadequate).

Also, keep in mind that besides performing bugfixes, Tim and other HP calc developers have to work on new features: for instance, at some point, official 3D graphing will certainly become an option. Maybe not now, but hopefully far quicker than it did for the Nspire series, which lacked that capability for 4 years.

UPDATE:

1) When you create a new program in the connectivity kit with an invalid name, which then changes to new_program by default, then try to save it, the connectivity kit crashes with a "The program has encountered a problem and must close" message.


2) Another issue is that when you place the cursor with the touch screen in the program editor to edit text there, it often moves to a different location by itself after blinking, usually one or two chars away, resulting in accidentally inserting text at the wrong location.


3) A considerable slowdown occurs when holding down keys. For example, a program that runs at 15 images per second with no key pressed will run at 12 FPS with keys pressed. My Tunnel, with all timers removed, drops from 60 FPS to 40 when keys are pressed. This might cause issues if someone decides to create a game involving timing.

Heya and welcome here. Glad to see another HP Prime user/owner.

Unfortunately I can't help you, since I haven't used the custom/user-key features, but hopefully if Tim Wessman passes by, he can help you.