Post Reply 
newRPL on HP 39gs
05-03-2017, 06:26 AM
Post: #1
newRPL on HP 39gs
(05-02-2017 08:16 PM)Claudio L. Wrote:  The changes to the code on github don't seem enough to make newRPL run on a 39gs, do you have more changes that you didn't commit? I'd like to see them all, perhaps I can put the 39gs/40gs as a secondary target. I'd have to recompute and recompress all those tables with less precision, though, and I don't know how small I can get them to be but I could give it a shot.

Hi, Claudio!

That would be great to have newRPL on HP 39gs.
Thanks you!

Repository contain the following important changes regarding original newRPL (September 2016):
1. Changed firmware identifier to "KINPOHP39G+IMAGE".
2. Changed display size to match HP 39gs.
3. Removed SD card and filesystem subsystem.

If also shrink tables size (I don't commit this changes), this will be enough to get working firmware for real calculator.

I can update repository to synchronize with actual newRPL changes, if it helps.
Find all posts by this user
Quote this message in a reply
05-03-2017, 05:42 PM
Post: #2
RE: newRPL on HP 39gs
(05-03-2017 06:26 AM)stored Wrote:  Hi, Claudio!

That would be great to have newRPL on HP 39gs.
Thanks you!

Repository contain the following important changes regarding original newRPL (September 2016):
1. Changed firmware identifier to "KINPOHP39G+IMAGE".
2. Changed display size to match HP 39gs.
3. Removed SD card and filesystem subsystem.

If also shrink tables size (I don't commit this changes), this will be enough to get working firmware for real calculator.

I can update repository to synchronize with actual newRPL changes, if it helps.

I saw those changes, but the firmware does not autodetect the 512 kb vs 256 kb of RAM, I thought you had made those changes as well.
I guess in your tests you never filled up more than half the memory, otherwise it would've crashed.

No need to merge the latest, I can see the changes you made on github and are quite simple.
I actually want to create a proper folder target-39g (like there is for the 50g), define the constant TARGET_39G, and I'm debating whether I should include stubs for the file system functions rather than remove them completely (to keep the commands visible, so RPL source code can be exchanged between 39g and 50g targets, all SDxxx commands will simply work as if no card is inserted, ever. The same stubs will be needed when newRPL is ported to the Prime too.
I did find the RPL code to generate the constants (yes, the tables were generated using newRPL itself), and the paper I took the idea from, although I don't recall much of the math involved to actually successfully modify the tables, so don't expect this to be working tomorrow.
What I don't have is a 39g for testing, I can test on my 50g but I'm going to need at least a couple of people to test the ROM thoroughly on real hardware once I release it.
Find all posts by this user
Quote this message in a reply
05-03-2017, 07:56 PM
Post: #3
RE: newRPL on HP 39gs
(05-03-2017 05:42 PM)Claudio L. Wrote:  I saw those changes, but the firmware does not autodetect the 512 kb vs 256 kb of RAM, I thought you had made those changes as well.
I guess in your tests you never filled up more than half the memory, otherwise it would've crashed.

Oh. Unfortunately, I missed this moment.
And in some conditions it crashed, you're right.

I just commit my changes with shrinking tables, but it seems it does not help much.
Sorry.
Find all posts by this user
Quote this message in a reply
05-03-2017, 07:58 PM
Post: #4
RE: newRPL on HP 39gs
(05-03-2017 05:42 PM)Claudio L. Wrote:  What I don't have is a 39g for testing,

A 39G, G+, gs or gII? :D

Greetings,
    Massimo

-+×÷ ↔ left is right and right is wrong
Visit this user's website Find all posts by this user
Quote this message in a reply
05-03-2017, 09:34 PM
Post: #5
RE: newRPL on HP 39gs
(05-03-2017 07:58 PM)Massimo Gnerucci Wrote:  
(05-03-2017 05:42 PM)Claudio L. Wrote:  What I don't have is a 39g for testing,

A 39G, G+, gs or gII? Big Grin

Well, since I don't have any of them, I had to do my homework first, but now I can answer more informed.
newRPL will target 39g+, 39gs and 40gs, which are practically the same machine as the 50g (the 40gs has 256kb RAM and 2MB flash, so it can run the "full" newRPL, the 39g+ and 39gs both have only 1 MB of flash, hence we are talking about cutting down on ROM space).
Find all posts by this user
Quote this message in a reply
05-11-2017, 01:29 AM
Post: #6
RE: newRPL on HP 39gs
Status update:
I created almost everything needed to release 3 different targets simultaneously: 40gs, 39gs and the 50g. Almost because I still need to recreate the transcendental tables for the smaller footprint of the 39gs.
Since that takes some work and I decided to do it, I'm also reworking the tables to switch from decimal CORDIC to binary. Not using decimal in the loop should increase speed substantially, although a conversion to decimal will be needed at the end. We'll see how much faster it is (if any).
And the tables required have a reduction in size due to the more compact storage (going from 8 digits in 32-bits to 9 and a fraction), plus the fact only 4 tables are needed instead of 8 (but those 4 are larger than the other 8, still a reduction in size).
Find all posts by this user
Quote this message in a reply
05-23-2017, 11:39 AM
Post: #7
RE: newRPL on HP 39gs
I've been testing NewRPL on 39gs with stored's binary release, and so far, it's working fine for me in basic calculations.
One thing that is bugging me it that keyboard overlay of 39gs differs greatly from 50g. I have memorized some of the differences (especially # and h keys, I'm using my calculators mostly for hexadecimal calculations), but on the long term, I think key remapping or new key overlay is a must on this hardware.

I'm really sorry for my bad English. English is not my first language.
Find all posts by this user
Quote this message in a reply
05-23-2017, 07:31 PM (This post was last modified: 05-23-2017 07:32 PM by Luigi Vampa.)
Post: #8
RE: newRPL on HP 39gs
[OFF TOPIC]
(05-23-2017 11:39 AM)Minori Hiraoka Wrote:  I'm really sorry for my bad English. English is not my first language.
Welcome to the club of non-native speakers.
You don't need to apologise, just try your best, and don't forget 'Google Translate' is our best friend in case of doubt :0)

Saludos Saluti Cordialement Cumprimentos MfG BR + + + + +
Luigi Vampa +
Free42 BlackviewA7 '<3' I + + +
Find all posts by this user
Quote this message in a reply
05-25-2017, 01:54 AM
Post: #9
RE: newRPL on HP 39gs
(05-23-2017 11:39 AM)Minori Hiraoka Wrote:  I've been testing NewRPL on 39gs with stored's binary release, and so far, it's working fine for me in basic calculations.
One thing that is bugging me it that keyboard overlay of 39gs differs greatly from 50g. I have memorized some of the differences (especially # and h keys, I'm using my calculators mostly for hexadecimal calculations), but on the long term, I think key remapping or new key overlay is a must on this hardware.

I'm really sorry for my bad English. English is not my first language.

Eventually, the keyboard mappings can be changed to make them more suitable to each hardware target, that's relatively easy to do. Getting trig to work is really hard due to limited space, I'm still working on that to see if we can compact the ROM some more for the 39 target.
Find all posts by this user
Quote this message in a reply
Post Reply 




User(s) browsing this thread: 1 Guest(s)