Home
Feedback
FAQ
Getting Started
Discussions
Site operation discussions
Recent Posts
(new topic)
Departments
Courses
Research Papers
Design Docs
Quotations
Genealogical Diagrams
Archives
This just came up in the discussion group,
The OS team at Linspire, Inc. (formerly Lindows) would like to announce that we are standardizing on Haskell as our preferred language for core OS development
Interesting!
I always thought that Java became widely adopted because a lot of programmers decided that they wanted to learn and use the language in the mid-90s. Only later did the use of Java become corporate policy. I also see this bottom up decision making process in languages like Python, Ruby and PHP. I find it revealing that VB is a mere teenager of 15 years, and a recent LtU post christened it one of the most popular PLs on the planet. And those languages like Java and C++ that have stuck around have undergone multiple radical changes in the same time period.
For those that believe we are stuck with the popular choices of PLs, all I can say is that the PL landscape is much more fluid than many have come to appreciate. This choice of Haskell by a few hardcore faithful may or may not result in yet another revolution, but I do think it indicative of a larger picture where the decisions of individuals to use the PLs of their own choice can make a difference. Of course, the biggest problem with this viewpoint is that it is very difficult to predict which languages will ultimately gain traction.
I agree with everything you said. However, when you say:
Of course, the biggest problem with this viewpoint is that it is very difficult to predict which languages will ultimately gain traction.
I can only just shrug. When does predicting the next big language actually matter? I can only think of two scenarios: (i) you are a language/tool vendor, (ii) you are out of touch with the software development practice and are looking to get hired by some big firm quickly.
If you don't fall into one of those categories, then anticipating the next big language is a waste of time. Simply use the best tools available (or the tools you are most comfortable with) to do the best job you can. Leave marketing to marketers.
Well, I don't see in which of your two categories Linspire would fall, but I do see why it could be of great importance to them whether or not Haskell will gain more traction in the future. Especially when they are stressing that they need enthousiasts from the community. Also, as any distribution vendor (and even more so as a deriviate), they build upon a lot of other tools and depend on good interaction between all parties involved. Imagine their situation in three years if they are the only four guys in the world using Haskell by then.
Of course, you can now argue that the 'use the best tools available' applies to all this, but then there is no discussion.
For those that believe we are stuck with the popular choices of PLs, all I can say is that the PL landscape is much more fluid than many have come to appreciate
Who is we? There's lots of shops where PL choice isn't dictated by a PHB who's breadth of technical knowledge comes out of InfoWeek articles.
This choice of Haskell by a few hardcore faithful may or may not result in yet another revolution, but I do think it indicative of a larger picture where the decisions of individuals to use the PLs of their own choice can make a difference.
How big do you think Linspire is? There's probably 3 guys that are going to be programming in Haskell. It's not really indicative of any "larger picture" than "a few guys at Linspire decided to write some code in Haskell"
... welcome our new Haskell overlords.
Which animal will grace the cover of Haskell in a Nutshell? Is anybody running a sweepstake? My money's on the llama.
The Llama is taken by "Learning Perl".
. . . kind of Australian. Like a duck-billed platypus or a clown fish (the _Finding Nemo_ fish)?
The platypus is taken.
Now, more to the point, is Haskell a good choice for OS development?
Apparently so.
Nice question, but not really fitting here. They create a Linux distribution and seem to be sick of using bash and such for their tools. It does not seem that they replace C parts in the kernel and drivers with Haskell versions.
What about the dugong? I've always had a soft spot for them, and a dugong would look pretty cool on a book cover. Yes. It should be a dugong.
(The llama was a stupid idea, and I formally disavow it.)
I hope that O'Caml doesn't claim it first; Haskell's class system is far superior. Should I set-up an on-line petition and lobby O'Reilly?
...it would SPJ wearing his hair shirt. :-)
Perhaps we could compromise on SPJ wearing a hare shirt
This correspondence is now closed. (I hope.)
Recent comments
27 weeks 2 days ago
27 weeks 2 days ago
27 weeks 2 days ago
49 weeks 3 days ago
1 year 1 week ago
1 year 3 weeks ago
1 year 3 weeks ago
1 year 5 weeks ago
1 year 10 weeks ago
1 year 10 weeks ago