Python 1.0 was directed more towards a programming literacy idea.
Python 2.0 from ~2000 was largely directed at web applications and text processing (a la Perl).
Python 3 (and later versions of Python 2) has been more directed (in the past decade) towards number crunching (numpy) and data analysis and machine learning.
It would be interesting to see a breakdown of perspectives on the language based on the applications it's been used for. Would web app maintainers still feel the love if they're maintaining something originating circa 2005, versus a data analysis app that was started in 2015.
I see nothing in Python 3 itself directed towards number crunching. Its integer performance actually worse than for Python 2 (because integers in 3 are big integers). nympy can be considered a DSL based on Python and it was used with Python 2 too.
I don't understand the point of your comment. I called out numpy for number crunching explicitly along with later versions of Python 2. I didn't say Python was good for number crunching, but that it's being used for number crunching and listed the library that has largely driven that trend (by making the language effective at number crunching).
Your phrasing actually did very much imply that Python the language was intended ("directed") for these tasks.
>Python 3 (and later versions of Python 2) has been more directed (in the past decade) towards number crunching
I don't know where "more directed" means anything but planned/intended e.g. you are saying here that the Python core devs have been building a language as a data analysis tool.
(This is obviously up to some interpretations and it's fair to say some poeple won't read it this way at all).
It sounds like you mean to say the growth/use has been in data analysis due to the Python-based tools that have grown up around the language, the ecosystem. I would agree with you.
Directed is an overloaded term. How people are using it is what the language is directed towards this is the sense I mean.
Directed could also mean intended by the developers of the language itself. This is not what I mean. Numpy is a secondary capability added by others, not core to the language proper.
Could you please elaborate then in which way Python 3 has been more directed towards number crunching?
I see that Fortran and Chapel are designed for number crunching, but Python IMHO used for this because it is a popular language in the first place and people having a computational problem choose the tool they already like.
I'm talking about what people are using it for, where the community (not necessarily the language's core devs) are directing its use. Contrast to the 00s, where it was not (as a general rule) being used for number crunching activities (ML or just regular number crunching a la Fortran or Matlab at the time). In the 10s, Python has been heavily used for applications in machine learning (and obviously number crunching heavy activity) and scientific simulation, data (number heavy) analysis, statistics work.
Prior to the 10s, those domains were more the purview of Fortran (as you list), Matlab, R (statistics), and others.
In the 00s, Python was more used for web applications and text processing (it was a "better" (subjective so in quotes) Perl).
Python 2.0 from ~2000 was largely directed at web applications and text processing (a la Perl).
Python 3 (and later versions of Python 2) has been more directed (in the past decade) towards number crunching (numpy) and data analysis and machine learning.
It would be interesting to see a breakdown of perspectives on the language based on the applications it's been used for. Would web app maintainers still feel the love if they're maintaining something originating circa 2005, versus a data analysis app that was started in 2015.