Another notable change is the re-emergence of Fortran in the index at 20th position, up from 34th spot a year ago. Fortran, which emerged from IBM in the 1950s, remains popular in scientific computing. Its highest ranking on Tiobe’s index was 10th in 2002.
“This dinosaur is back in the top 20 after more than 10 years. Fortran was the first commercial programming language ever, and is gaining popularity thanks to the massive need for (scientific) number crunching. Welcome back Fortran,” says Tiobe.
Thanks to Milan, Ondrej, and others for their efforts. I am pleased to see Fortran’s rise in the TIOBE rankings, although I wonder if its relative popularity really rose so much in 1 year.
Awesome, T-Rex is back , crunching numbers! It’s clearly a consequence of the efforts of this community and its leaders.
Although one may think this is just a ranking full of biases, it is a clear signal that something is happening concerning Fortran. And it will help attracting more people here.
We can now hope that Fortran has a R_0>1 ! (or will soon)
Still, the argument is always valid that there are no good measures to ascertain what are “top” languages.
Regardless, with a few more convenient features in the standard that can enable better and faster library development, Fortran can really play a lead role in scientific and technical computing. To address this gap is what should be focus re: the base language itself.
If you select “IEEE Spectrum” and “Enterprise” as a language type, Fortran is 18th.
And you can create your “custom ranking”: if you click on that button, you can see the criteria and modify their weight. That’s very interesting!
Fortran always needed a user-centric and user-friendly community. This happened last year. More and more people are learning about the fitness of modern Fortran for scientific computing. and such boosts in rankings naturally follow the increased activities. Fortran marches on…
Occasionally, I still meet colleagues who resentfully mention their times with Fortran, and when I get deep I realize they were taught and used FORTRAN77 or at best F90. Anyone who still teaches or limits development to Fortran90 and older standards is still potentially contributing to this feeling of resentment. Multiple compilers support the Full Fortran 2018 standard. No excuse for anyone to teach or use or stick to any lower standards, although I don’t know if there is anyone.
All of these said, IMO, the next standard (which will hopefully significantly expand meta-programming capabilities of the language) is absolutely necessary to resolve the remaining major criticisms of the language.
Perhaps I should have said a well-organized vendor-independent user-centric community. The Intel Fortran forum has been also around for I guess 2 decades if not more, which is still one of the best places to discuss any Fortran-related question, not limited to ifort. The same probably holds for the NVIDIA Fortran forum.
But, for the past 30 years, Fortran’s relative popularity has been also going downward. Now in almost a year, the trend has reversed and Fortran has jumped halfway through the ranking list. The team behind this new Fortran community and all related activities deserves credit and huge applause for their achievements and bravery.
p.s.1 Now that I recheck comp.lang.fortran, I realize that how active it has been and still is. Although I knew of its existence for a long time, I have not been active in the community, mostly because of its mailing list structure. Are there any plans to make this group a standalone organization, separate from Google groups? One big advantage of this new forum is that it is a standalone website with modern features.
p.s.2 I have removed an anecdote in this comment to prevent it from getting picked up by the search engines.
I have been following c.l.f. for decades as well. This forum is very lightly moderated (I don’t remember seeing a post or thread deleted by a moderator), but maybe the fact that a moderator could intervene maintains quality. I remember on c.l.f. that a poster with initials G.T., who was actually knowledgeable about Fortran, became frequently abusive and surely drove away some people. Another still-active poster is always explaining how Fortran is inferior to PL/1. Nowadays c.l.f. is pretty good, but it’s nice to have a forum that is overseen by Fortran enthusiasts.
I have also found many valuable posts at c.l.f, and think it is still the preferred communication avenue of many Fortran experts.
There a few things I don’t like about c.l.f. including:
occasional spam posts which don’t seem to get flagged or deleted
occasional posts in a negative or abusive tone, with some arguments that seem to go far back
lack of support for formatting posts in Markdown (I might be wrong here, and the interface works better via email).
I would also add that there are hardly any non-internet communities built around Fortran, that is apart from the standards committee.
In comparison, Python enthusiasts meet at conferences like PyCon, the Scipy Conference, EuroPython, EuroScipy, PyConDE…; C++ has the well known CppCon conference series, and Julia has been able to grow participation in it’s own JuliaCon conference incredibly fast.
As a further example, the Munich C++ User group (also known as MUC++) counts around 3000 members, and organizes lectures and meetups every month. Before the pandemic, I visited a few of the real-life meetups which were so popular the organizers struggled to find venues which could accommodate all the people that wanted to join (a few times they even had snacks, pizza, and beer).
Even the OpenMP community has been engaging with developers by offering free workshops and organizing hackathlons for students.
In comparison to these young and vibrant communities, the c.l.f group (and especially the J3 mailing list) can sometimes feel more like an old boys’ club.
At the same time I feel very lucky that the Fortran world is so small we get to interact with Fortran authors like Arjen or Milan on a weekly basis, and even Steve Lionel joins our monthly meetings now and then.
Thank you for your encouraging words @shahmoradi. I put incredible amount of effort and time into this and I know others have too, but I think it is worth it. The main benefits will come in the future, but we are already seeing some first signs.
@kargl, our efforts are a lot more than just a communication venue (although that is part of it), we now have a Fortran website that comes up second in all search engines (including Google for me now) after Wikipedia, we have a solid Twitter and GitHub presence and we collaborate as community on many open source projects (fpm, stdlib, the website, … and hopefully a lot more in the future). It might sound like nothing substantial, but it is actually incredibly important for a solid foundation to build upon and very hard to pull off. I tried to create a community like this for 10 years and failed every time until just now. I described my personal journey here:
Several people have objected to the title, so please just imagine it means “Rejuvenating Fortran” which is what I meant by it.
As I describe in the blog post, I don’t think we are done, we only finished the first step. I don’t talk about next steps in the post, but I think some of the next steps that I would like to see is maturing our core tools (fpm, stdlib) as well as getting more open source compilers maturing, good documentation/tutorials and many libraries and package ecosystem maturing; growing our community and seeing a lot more new projects starting in Fortran. The ultimate measure for me personally will be when I see around me that people stop saying that “Fortran is dead or dying” (the anecdote that @shahmoradi described is all too common, I witness it every year from somebody) and they not only stop moving away from it, but will move away from C++ and other languages into Fortran for some projects (not all projects, but some projects). If we achieve that, then we will have succeeded in “rejuvenating” Fortran.
By the way, it’s interesting the article states, “Fortran … Its highest ranking on Tiobe’s index was 10th in 2002”
During the mid to late 1990s even i.e., not much prior to that 10th ranking on that index, many of the engineering programs I was familiar with among the US universities and colleges had stopped using Fortran as part of their curriculum toward courses in numerical methods, etc.
Around 1996 or soon thereafter, the authors of Numerical Recipes - Press et al. - too had announced their future editions will not be based on Fortran.
A lot of such trends had an adverse impact on the adoption of the language.
old enough to remember this perhaps annoying but nostalgic sound: ALL Old Modem Sounds (300 baud to 56K) - YouTube
followed by my parents yelling at me for disconnecting their phone conversation.
, crunching numbers! It’s clearly a consequence of the efforts of this community and its leaders.