On 17.10.2021 17:00, William Unruh wrote:
> On 2021-10-16, Janis Papanagnou <
janis_pa...@hotmail.com> wrote:
>> On 15.10.2021 15:59, Spiros Bousbouras wrote:
>>> [ Fortran and COBOL ]
>>
>> [...] Not sure what
>> the reference to these two languages helps for the argument. They have
>> at best a relevance in legacy systems (that no one dares to touch) and
>> long grown libraries (that are primarily just called from other modern
>
> Fortran is stillthe language of choice for people doing numerical work
Really? Still? Where? Any why?
> It is not legacy.
We spoke about two languages and I suggested two categories.
COBOL has "relevance in legacy systems". Having it still running
on old mainframes in the financial sector, and being continuously
standardized, doesn't make these _systems_ non-legacy. It's only
practically impossible to replace them, maintenance is expensive,
people with COBOL competence rare. Companies in that sector tried
to replace these systems for long but the long time grown, often
not sufficiently documented software is critical yet to be touched.
Fortran has "relevance in long grown libraries", the second category
I mentioned. There are long time established libraries specifically
for numerical computation since decades and still in active use. Yes.
Or did you mean that for some people Fortran is also the language of
choice to do the programming around these libraries? Well, if so - as
someone with CS background and who has programmed in Fortran and knows
that language -, I can just shake my head. Why would one want to code
in Fortran, nowadays? Seriously asked.
But, frankly, I don't know how big the user community these days is.
Do you know it?
Though, in that formulated generality it's anyway not true. My children,
for example, are working in geophysics, do heavy numerical simulation
work and also occasionally (where necessary) use Fortran libraries, but
their preference to do the programming around the libraries are clearly
other languages, ranging from C++ to Python. (Note: that's of course not
their personal preference - they are no CS professionals - but that just
reflects the situation at the two large universities hereabouts.) That's
why I am astonished about your statement.
Janis