Post Reply 
New Version: 2018.07.06 2.0.0.13865
07-28-2018, 07:18 AM
Post: #41
RE: New Version: 2018.07.06 2.0.0.13865
Actually x=1 is allowed, because the fraction simplifies (or if you prefer the limit at x=1 of your expression is 0, therefore x=1 fullfills the condition).
Find all posts by this user
Quote this message in a reply
07-28-2018, 09:50 PM (This post was last modified: 07-28-2018 09:51 PM by John P.)
Post: #42
RE: New Version: 2018.07.06 2.0.0.13865
Hello Mr. Parisse,
Since you are here I would like to ask you couple of questions.
1. Would it be possible to add to TYPE command to recognize functions not only as type 8, but also distinguish between trig fns., exp fns. log fns. etc with type for example 81, 82, 83 etc.?
2. I think there is a bug somewhere in the HP Prime CAS that crashes the calc. when function is not the right type. HP Prime should not crash, but but gently return some kind of message that the argument for such and such fn. is not the correct type.
Thank you for your time.
Find all posts by this user
Quote this message in a reply
07-29-2018, 11:15 AM
Post: #43
RE: New Version: 2018.07.06 2.0.0.13865
type is a command to distinguish object types, all builtin commands are of the same type. You can test for specific functions with ==, for example if f=='sin' || f=='cos' then ...

If you have found a command that crashes, please give the corresponding commandline, otherwise I obvioulsy can't check and can not fix.
Find all posts by this user
Quote this message in a reply
07-29-2018, 05:00 PM
Post: #44
RE: New Version: 2018.07.06 2.0.0.13865
(07-29-2018 11:15 AM)parisse Wrote:  type is a command to distinguish object types, all builtin commands are of the same type. You can test for specific functions with ==, for example if f=='sin' || f=='cos' then ...

If you have found a command that crashes, please give the corresponding commandline, otherwise I obvioulsy can't check and can not fix.

Hello Mr. Parisse,

The problem I had is described in the thread:
http://www.hpmuseum.org/forum/thread-11125.html
"Bug(s) in the New Version: 2018.07.06 2.0.0.13865"

I was just trying the new release of the firmware and with the inputs described in that tread I had a lot of hang ups of the prime. One thing that reduced the numbs. of hang ups was deselecting the 'Complex' in Home Settings. Some crashes still occur when I do some lists simplifications with trig functions. When I do for example 'trigtan' on those lists I did not have crashes. I did not had crashes when using the Virtual Calc, looks strange because VC and physical HP Prime are suppose to behave in the same way. So, I thought having 'types' for trig, log and exp function could prevent those crashes by applying the appropriate command(s). I could be missing something, but the crashes looks strange and I thing they are related to not using the right command on some lists with functions as elements.
Thank you for your time.

John P
Find all posts by this user
Quote this message in a reply
07-29-2018, 07:59 PM
Post: #45
RE: New Version: 2018.07.06 2.0.0.13865
What I need is a list of commands without any user interaction like say
Code:

l:=[30,45,60,70,80,85,90]*pi/180;
sin(l).^2 .* cos(l).^2;
In other words, something I can just copy/paste to the emulator and see a bug happen.
Find all posts by this user
Quote this message in a reply
07-29-2018, 09:25 PM
Post: #46
RE: New Version: 2018.07.06 2.0.0.13865
(07-29-2018 07:59 PM)parisse Wrote:  What I need is a list of commands without any user interaction like say
Code:

l:=[30,45,60,70,80,85,90]*pi/180;
sin(l).^2 .* cos(l).^2;
In other words, something I can just copy/paste to the emulator and see a bug happen.

No, in the emulator I did not have any crashes at all only on the physical HP Prime. sin(l).^2 .* cos(l).^2; works OK, but if I do simplify(sin(l).^2 .* cos(l).^2); or simplify on the result of sin(l).^2 .* cos(l).^2 physical HP Prime crashes. If I do trigsin, trigcos, trigtan on sin(l).^2 .* cos(l).^2 works OK. All of the above works OK on the VC with no problems whatsoever.

Cheers
John P
Find all posts by this user
Quote this message in a reply
07-29-2018, 09:39 PM (This post was last modified: 07-29-2018 09:42 PM by ijabbott.)
Post: #47
RE: New Version: 2018.07.06 2.0.0.13865
(07-29-2018 07:59 PM)parisse Wrote:  What I need is a list of commands without any user interaction like say
Code:

l:=[30,45,60,70,80,85,90]*pi/180;
sin(l).^2 .* cos(l).^2;
In other words, something I can just copy/paste to the emulator and see a bug happen.

This hangs on the physical calculator in CAS in exact mode with angle set to Radians:

Code:
l:={85,85}*π/180;
sin(l).^2 .* cos(l).^2;
simplify(Ans);

Actually, it may come back eventually, but it takes a long time.

— Ian Abbott
Find all posts by this user
Quote this message in a reply
07-30-2018, 05:49 AM
Post: #48
RE: New Version: 2018.07.06 2.0.0.13865
If it works on the emulator I can't debug myself. I made a few checks, and it happens sometimes that the simplify command takes a few seconds on the emulator (which would mean the physical calc seems to hang). But I can not reproduce in a deterministic manner and that never happened when I did
time(simplify(Ans)). Maybe it's a GUI slowness, I don't know.
Find all posts by this user
Quote this message in a reply
07-30-2018, 10:31 AM
Post: #49
RE: New Version: 2018.07.06 2.0.0.13865
(07-30-2018 05:49 AM)parisse Wrote:  If it works on the emulator I can't debug myself. I made a few checks, and it happens sometimes that the simplify command takes a few seconds on the emulator (which would mean the physical calc seems to hang). But I can not reproduce in a deterministic manner and that never happened when I did
time(simplify(Ans)). Maybe it's a GUI slowness, I don't know.

It's certainly slow - it takes long enough for the watchdog (or something else - perhaps a stack crash detector?) to reset the calculator.

— Ian Abbott
Find all posts by this user
Quote this message in a reply
07-30-2018, 12:42 PM
Post: #50
RE: New Version: 2018.07.06 2.0.0.13865
Can you reproduce the crash if you run time(simplify(Ans)) instead of just simplify?
Find all posts by this user
Quote this message in a reply
07-30-2018, 01:51 PM
Post: #51
RE: New Version: 2018.07.06 2.0.0.13865
(07-30-2018 12:42 PM)parisse Wrote:  Can you reproduce the crash if you run time(simplify(Ans)) instead of just simplify?

Hello,

The best I could do.
1. Evaluate sin(l)./tan(l) went OK
2. The first time using simplify on the soft menu went OK.
3. The second time using the same evaluated expression took very long time more than 10s
I pressed ON+SYMB to brake the operation.
4. Using time(simplify()), used simplify on the soft menu and pasted the same evaluated expression from history gave me "Warning, ^ is ambiguous on non square matrices. Use .^ to apply ^ element by element." repeated many times to fill the whole screen. After waiting long time 1 min or so and pressing ENTER the time was 6.736.
5. Not using soft menu for simplify() and typing time(simplify()) and pasting from history the same expression the first time it finished in 0.012, but the second time doing the same thing repeated what I described in pt. 4 and after couple of minuts I pressed Enter and nothing happened. Later on the HP Prime rebooted or restarted by itself.

I hope this helps.

Cheers
John P
Find all posts by this user
Quote this message in a reply
07-30-2018, 05:49 PM
Post: #52
RE: New Version: 2018.07.06 2.0.0.13865
Does not help me at all, but perhaps Tim or Cyrille will have an idea.
Find all posts by this user
Quote this message in a reply
07-31-2018, 12:27 AM
Post: #53
RE: New Version: 2018.07.06 2.0.0.13865
(07-30-2018 12:42 PM)parisse Wrote:  Can you reproduce the crash if you run time(simplify(Ans)) instead of just simplify?

The initial sequence:

l:={85,85}*π/180;
((sin(l)).^2).*((cos(l)).^2);
[OFF]
[ON]
time(simplify(Ans));

gives an answer between 40 and 41 seconds. Subsequently:

((sin(l)).^2).*((cos(l)).^2);
time(simplify(Ans));

gives an answer between 0.8 and 1 second.

— Ian Abbott
Find all posts by this user
Quote this message in a reply
07-31-2018, 07:03 AM
Post: #54
RE: New Version: 2018.07.06 2.0.0.13865
And what happens if you run directly time(simplify( sin(l).^2 .* cos(l).^2)) ?
Find all posts by this user
Quote this message in a reply
07-31-2018, 07:47 AM
Post: #55
RE: New Version: 2018.07.06 2.0.0.13865
(07-31-2018 07:03 AM)parisse Wrote:  And what happens if you run directly time(simplify( sin(l).^2 .* cos(l).^2)) ?

It remains busy for a couple of minutes or so and then resets the calculator. (Same as without the time() function.)

— Ian Abbott
Find all posts by this user
Quote this message in a reply
08-03-2018, 10:02 PM
Post: #56
RE: New Version: 2018.07.06 2.0.0.13865
(07-31-2018 07:03 AM)parisse Wrote:  And what happens if you run directly time(simplify( sin(l).^2 .* cos(l).^2)) ?

Hello,
No news from Tim or Cyrill. I hope the bug described above does not become some kid of internal pingpong match at HP Calculator department.

Cherrs
John P
Find all posts by this user
Quote this message in a reply
08-03-2018, 10:11 PM
Post: #57
RE: New Version: 2018.07.06 2.0.0.13865
I think after discussing it in this forum, you have to report the problems at calcbeta@hp.com
Find all posts by this user
Quote this message in a reply
08-04-2018, 03:08 AM
Post: #58
RE: New Version: 2018.07.06 2.0.0.13865
(08-03-2018 10:11 PM)compsystems Wrote:  I think after discussing it in this forum, you have to report the problems at calcbeta@hp.com

Hello,
Do they have web page for 'calcbeta' or I just send them an email for 'calcbeta@hp.com'. If they have a web page could you post the link to it.
Thank you.

John P
Find all posts by this user
Quote this message in a reply
08-16-2018, 03:16 AM
Post: #59
RE: New Version: 2018.07.06 2.0.0.13865
(07-20-2018 09:16 AM)scorpio72 Wrote:  Does it mean we have 3 HP Prime versions?

The 1: A
NW280AA

The 2: C
G8X92AA

The 3: G2
2AP18AA

What processor is used in the G2 ?

The G2 shows as Hardware Version D on the about screen.

Tom L
Cui bono?
Find all posts by this user
Quote this message in a reply
12-12-2018, 12:53 PM
Post: #60
RE: New Version: 2018.07.06 2.0.0.13865
I realize the thread is a bit stale...as of 2018-12-12, still no hint of the G2 in the US, though I see I can buy it now from the UK. Wonder what the story is here in the US? Not gonna release it until all the old stock is sold? Tariffs? I asked HP twice, and they have no clue.

One of the things I miss with my Prime that I really liked on the 50g is that I was able to make a fairly detailed soft menu of the math features that I use often, and those included swapping among XYZ and polar and polar 3D, PLUS all the vector functions, such as dropping the stack into a 2D 3D vector, changing from XYZ to polar, etc. But I don't know how to do that easily on the Prime.

I actually tried to go back to the 50g for these features, and then run it with the Prime so that I'd have quicker access to those softkeys and math functions, but I could not get it to work in Windows 10, so supporting the 50g became much harder.

I'm heading back to grad school (physics) and was hoping to have the G2. I have used only HP calculators in school, from HS on. I suppose that I could buy from the UK...

~R~
Find all posts by this user
Quote this message in a reply
Post Reply 




User(s) browsing this thread: 2 Guest(s)