Hi Rob,
everything you mention is true - and in my mind everything was solved already
:-)
The plan is this:
You order what you need by definingconst LARGE_ARRAY_SIZE = 600 const
LARGE_ARRAY_VARIABLE_SIZE = 2
or
const LARGE_ARRAY_1_SIZE = 300 const LARGE_ARRAY_1_VARIABLE_SIZE = 4
const LARGE_ARRAY_2_SIZE = 600 const LARGE_ARRAY_2_VARIABLE_SIZE = 1
...
Then, you only include the one library, which, depending on the defined
constants, gives you up to four arrays as specified.
The library itself would then check if it was already included, so that a
program would compile, even if the lib is included more than once.
That way, the new lib is fully compatible with the existing solution, the only
change that should be made in an existing program is to replace
include large_array_1include large_array_2
include large_array_3include large_array_4
by
include large_array
but even that change is not compulsory, I would like to issue a warning,
though.
Suggestions 1),2) and 3) is what I meant with "the comments aren't nice". Those
comments are duplicated in several places in the current script. I did not
bother to change them, as the python version will get a different structure
where there's only one place to change.
I think I will need several days to complete the python script, after that I
would highly appreciate if you would upload it all :-)
Greets,Kiste
Am Dienstag, 15. Dezember 2020, 11:57:07 MEZ hat Rob CJ
<[email protected]> Folgendes geschrieben:
Hi Kiste,
I ran your script to see the results I now I understand why Matt suggested to
replace the libraries. I agree with that but I would keep the four libraries.
It can be that you need two independent large arrays of different sizes and
then it would be handy if you could select from more large array libraries.
Some suggestions to change:1) Make the last copyright year 2020 (or 2021 if you
are done next year).2) Add your name to the revision of all created files
(since you changed them all)3) Change the compiler version to v25r4.
I can run the script for you and upload all versions to GitHub if you want. For
the time being we should also upload your script to GitHub.
Kind regards,
Rob
Van: [email protected] <[email protected]> namens Rob CJ
<[email protected]>
Verzonden: dinsdag 15 december 2020 09:59
Aan: [email protected] <[email protected]>
Onderwerp: Re: [jallib] Large_array update Hi Kiste,
Yes, that would be a good solution.
Met vriendelijke groet,Rob Jansen From: 'Oliver Seitz' via jallib
<[email protected]>
Sent: Tuesday, December 15, 2020 9:55:45 AM
To: [email protected] <[email protected]>
Subject: Re: [jallib] Large_array update Hi Rob,
good point. How about this:
large_array.jallarge_array_1.jal...large_array_4.jal
where large_array.jal does all the work, while the numbered libs only (issue a
deprecated warning and) include the general lib.
Greets,Kiste
Am Dienstag, 15. Dezember 2020, 09:42:03 MEZ hat Rob CJ <[email protected]>
Folgendes geschrieben:
Hi Matt, Kiste,
Why not just add a new large array without a number and keep the old ones? All
other large arrays carry a number.
So just large_array.jal
Reason is that we are not sure if other JAL users have uses large_array_x.jal
(x = 1..4) in one of there programs and we remain backwards compatible if we
keep them.
I know I used large_array_1.jal in one of my esp8266 sample programs. I have no
problems with fixing sample files, I am only concerned about backwards
compatibility.
Kind regards,
Rob
Van: [email protected] <[email protected]> namens Matt Schinkel
<[email protected]>
Verzonden: maandag 14 december 2020 18:43
Aan: [email protected] <[email protected]>
Onderwerp: Re: [jallib] Large_array update Hi, you may replace my libraries if
you wish.
Since the code is generated, I'd prefer one library. Several samples will need
to be updated.
Matt.
From: 'Oliver Seitz' via jallib <[email protected]>
Sent: December 14, 2020 8:44 AM
To: [email protected] <[email protected]>
Subject: Re: [jallib] Large_array update Hi all!
I would prefer not to change names, just replace the already existing
libraries. I'm sure it is compatible. The code generated is in all cases
smaller using the new lib, and executes faster in most cases. Only for 16bit on
a PIC14 core I've seen the program running slower by about 2%, in all other
cases it was faster by up to 33%. Speeding up the 16bit/PIC14 case would cost
around 20 code words. Would still be over 200 code words smaller than the
current lib, though.
What I did not yet investigate: While usually needing the same or up to 7 bytes
less, the new lib uses more RAM in two variants, 1 additional byte at
16bit/PIC14 and 4 additional bytes at 32bit/PIC16. That might be an issue, if
an existing program is designed to use almost all RAM.
The current address range of the PIC16 cores can only address 16kbyte. If
there's a chip with more than 14kbyte GPR, the compiler has to be extended. If
that happens, rebuilding the libs with the then-maximum amount of GPR would be
the least problem. So it wouldn't be difficult to maintain compatibility within
each package: As long as the compiler doesn't support more than 14k, the
library doesn't need to. And if one is changed, the other is done easily.
What gets onto my mind thinking about names: The current lib is designed to
make use of up to four large arrays. The library is split up into 28 source
files. When I'm already rewriting the lib, shouldn't I combine it into one, or
at most four, source files?
The script produces working libraries, but the comments arten't as nice as I
wanted them ;-)
Greets,Kiste
Am Montag, 14. Dezember 2020, 10:07:35 MEZ hat Rob CJ <[email protected]>
Folgendes geschrieben:
Hi Kiste (and others),
Can you send your latest version of your script?
I propose to add a library called large_array_13k.jal or should we give it
another name like super_array.jal? I think the size must be in the filename in
case MicroChip releases a chip with 32k.
@Others: Suggestions for a library name are welcome.
Python seems to be the most popular language to use. I could give it a try to
translate your script but currently I am investigating a compiler issue that
takes up my time.
Thanks.
Kind regards,
Rob
--
You received this message because you are subscribed to the Google Groups
"jallib" group.
To unsubscribe from this group and stop receiving emails from it, send an email
[email protected].
To view this discussion on the web visit
https://groups.google.com/d/msgid/jallib/1684278583.441369.1607953459506%40mail.yahoo.com.
--
You received this message because you are subscribed to the Google Groups
"jallib" group.
To unsubscribe from this group and stop receiving emails from it, send an email
[email protected].
To view this discussion on the web visit
https://groups.google.com/d/msgid/jallib/SN6PR02MB49768E1B4F7B82A1F1992173DEC70%40SN6PR02MB4976.namprd02.prod.outlook.com.
--
You received this message because you are subscribed to the Google Groups
"jallib" group.
To unsubscribe from this group and stop receiving emails from it, send an email
[email protected].
To view this discussion on the web visit
https://groups.google.com/d/msgid/jallib/AM0PR07MB6241DD383872E10648F4FA19E6C60%40AM0PR07MB6241.eurprd07.prod.outlook.com.
--
You received this message because you are subscribed to the Google Groups
"jallib" group.
To unsubscribe from this group and stop receiving emails from it, send an email
[email protected].
To view this discussion on the web visit
https://groups.google.com/d/msgid/jallib/1529225854.799613.1608022545690%40mail.yahoo.com.
--
You received this message because you are subscribed to the Google Groups
"jallib" group.
To unsubscribe from this group and stop receiving emails from it, send an email
[email protected].
To view this discussion on the web visit
https://groups.google.com/d/msgid/jallib/AM0PR07MB62414D5AE839B49039E7BC46E6C60%40AM0PR07MB6241.eurprd07.prod.outlook.com.
--
You received this message because you are subscribed to the Google Groups
"jallib" group.
To unsubscribe from this group and stop receiving emails from it, send an email
to [email protected].
To view this discussion on the web visit
https://groups.google.com/d/msgid/jallib/AM0PR07MB62419BEF22493F250907586AE6C60%40AM0PR07MB6241.eurprd07.prod.outlook.com.
--
You received this message because you are subscribed to the Google Groups
"jallib" group.
To unsubscribe from this group and stop receiving emails from it, send an email
to [email protected].
To view this discussion on the web visit
https://groups.google.com/d/msgid/jallib/951323379.870773.1608036437984%40mail.yahoo.com.
