[seqfan] Re: Program syntax and the OEIS

Charles Greathouse charles.greathouse at case.edu
Sat Jun 25 00:45:27 CEST 2011

Right, there are problems going backward and forward with most
languages.  But I'm not looking for a perfect solution, just a 'good
enough' one.  For example, as Robert Israel points out, Maple V users
are rare, so the best solution is probably to change whatever old code
we have sitting around.  Similarly, it wouldn't be that hard for us as
a group to write working (but not fast) equivalents for the new
Mathematica functions so that the "90%" solution could be "download
and run this compatibility workbook/package/etc." that has most of the
popular functions like NextPrime.  (I assume it's not hard to test if
a function is defined and import a function only if it is not.)

Perhaps one day we'll have preferences for logged-in users like

Mathematica: 7
PARI/GP: 2.6
C: 0x
. . .

which can draw in relevant needed functions on the hope that they will
make the program work.  But this, like many other features, depends on
having versions recorded somehow.

Charles Greathouse
Case Western Reserve University

On Fri, Jun 24, 2011 at 3:41 PM, Harvey P. Dale <hpd1 at nyu.edu> wrote:
> Al writes " I get the feeling Wolfram will always maintain backwards
> compatibility."  This, unfortunately, is not always true.  For example,
> what is now IntegerPartitions used to be Partitions.  Programs written
> under the older version, using Partitions, will not run properly in the
> new version.  The existence of such changes is no doubt what prompted
> Wolfram to provide a Mathematica Version Advisory that pops up
> automatically whenever an older-version program is loaded.
> Harvey
> -----Original Message-----
> From: seqfan-bounces at list.seqfan.eu
> [mailto:seqfan-bounces at list.seqfan.eu] On Behalf Of Alonso Del Arte
> Sent: Friday, June 24, 2011 11:22 AM
> To: Sequence Fanatics Discussion list
> Subject: [seqfan] Re: Program syntax and the OEIS
> With Mathematica, I think the only problem is people still using older
> versions, I get the feeling Wolfram will always maintain backwards
> compatibility.
> For example, Zak Seidov's program for A099349 uses NextPrime, which
> wasn't
> introduced until Mathematica 6.0. I'm sure in Mathematica 10.0 NextPrime
> will work as before, as well as older commands like PrimeQ and Range.
> For
> that sequence, a comment like (* Version 6.0+ *) might be sufficient.
> Those
> using 5.2 and earlier probably have already cobbled together their own
> version of NextPrime.
> Al
> On Fri, Jun 24, 2011 at 1:05 AM, Charles Greathouse <
> charles.greathouse at case.edu> wrote:
>> The recent case of A006506 (where a correct program in Maple 5 no
>> longer gave correct answers in Maple 6+) has made me think about ways
>> to store information on programming language and version into the
>> program, maple, and mathematica fields.  (One day, perhaps, simple
>> conversion could be done between versions, but in the medium term this
>> would at least alert people to the version tested so they could check
>> compatibility.)
>> But in the short term, before any attempt to add hidden markup or
>> simply standardize notation, it might seem worthwhile to search code
>> for particular snippets known to have changed in meaning.  It's hard
>> in this case: the affected operator is "." , which is presumably
>> rather common in other uses.
>> I have recently come across what will be another example.  PARI/GP
>> just released a new stable version 2.5.0, and along with this
>> deprecated the operators & and | (which have been shorthand for the &&
>> and || short-circuit operators).  They continue to work in 2.5 and
>> 2.6, but probably the next version will drop support for them.
>> Russ/David/Neil, is there a way to search for instances of these so
>> they can be corrected?  Regular expression like
>> /(^|\n)\(PARI\).*[^&]&[^&]/s
>> and
>> /(^|\n)\(PARI\).*[^|]\|[^|]/s
>> would give reasonably accurate matches, but I have no way to run those
>> (short of screen-scraping the entire OEIS!).
>> Charles Greathouse
>> Analyst/Programmer
>> Case Western Reserve University
>> _______________________________________________
>> Seqfan Mailing list - http://list.seqfan.eu/
> _______________________________________________
> Seqfan Mailing list - http://list.seqfan.eu/
> _______________________________________________
> Seqfan Mailing list - http://list.seqfan.eu/

More information about the SeqFan mailing list