Mastering Time-to-Market with Haskell
fpcomplete.com
Mastering Time-to-Market with Haskell
1–10 of 120 posts
Re: Mastering Time-to-Market with Haskell
#2> In summary we've seen that: Haskell decreases development time...
Have we actually seen that or have you just asserted that? Is this really true, and if it is, by how much? Haskell has been around for a couple of decades now, and has had least two hype cycles (I remember that when I was in university in the late '90s, Haskell was the next big thing). It does not seem to expand significantly even within organizations that have tried it (and that's a very negative signal), with at least one notable case where the language has been abandoned by a company that was among the flagship adopters.
In general, we know that often linguistic abstractions that seem like a good idea in theory -- or even seem to work nicely in small programs -- don't end up having a significant effect on the bottom line when larger software is concerned. People say that scientific evidence of actual contribution is hard to collect, but we don't even have well-researched anecdotes. Not only do we not have strong evidence in favor of this hypothesis, but there aren't even promising hints. All we do have is people who really like Haskell based on its aesthetics and really wish that the the nice theoretical arguments translated to significant bottom-line gains.
This blog post by Dan Ghica, a PL researcher, really addresses this point: there is nothing to suggest that aesthetically nice theory translates to actual software development gains, and wishful thinking (or personal affinity) simply cannot replace gathering of data: http://danghica.blogspot.com/2016/09/what-else-are-we-gettin...
Re: Mastering Time-to-Market with Haskell
#3> In summary we've seen that: Haskell decreases development time... Have we actually seen that or have you just asserted that? Is this really true, and if it is, by how much? Haskell has been around for a couple of decades now, and has had least two hype cycles (I remember that when I was in university in the late '90s, Haskell was the next big thing). It does not seem to expand significantly even within organization…
Not the author and I don't know much besides a little dabbling in Haskell, but I'm sure that quote holds true, but only if the engineer already is an intermediate-advanced Haskeller, which probably translates to at least a couple years of daily Haskell work.
Re: Mastering Time-to-Market with Haskell
#4> In summary we've seen that: Haskell decreases development time... Have we actually seen that or have you just asserted that? Is this really true, and if it is, by how much? Haskell has been around for a couple of decades now, and has had least two hype cycles (I remember that when I was in university in the late '90s, Haskell was the next big thing). It does not seem to expand significantly even within organization…
Not the author and I don't know much besides a little dabbling in Haskell, but I'm sure that quote holds true, but only if the engineer already is an intermediate-advanced Haskeller, which probably translates to at least a couple years of daily Haskell work.
> but I'm sure that quote holds true
How can you be sure? Have you seen Haskell shops churn out software significantly faster than others?
Re: Mastering Time-to-Market with Haskell
#5Although very fresh into the Haskell world myself, I tend to agree with the author that, when I know what I am doing, my Haskell code is usually written in less time and has less bugs.
Having said that, Time-to-Market is only partially influenced by my-code, the biggest part is the code that I don't have to write, i.e. third-party libraries.
In my Haskell adventures I am having trouble finding third-party libraries for even the most popular things, e.g. Cassandra. As far as I can tell there are two libraries, the `cassandra-cql` and the `cql-io`, the first hasn't been updated for a year now, and the second has only 3 stars, which makes me uneasy.
So, although I can see where the author is coming from, I don't think you can beat Java, Ruby, JS or Python in that sense. Unless of course your code/project doesn't have a lot of dependencies.
Re: Mastering Time-to-Market with Haskell
#6> In summary we've seen that: Haskell decreases development time... Have we actually seen that or have you just asserted that? Is this really true, and if it is, by how much? Haskell has been around for a couple of decades now, and has had least two hype cycles (I remember that when I was in university in the late '90s, Haskell was the next big thing). It does not seem to expand significantly even within organization…
It's hard to measure these types of things. But if you are interested, there is a related paper regarding that: http://haskell.cs.yale.edu/wp-content/uploads/2011/03/Haskel...
Although the study in the paper isn't very practical, it's still an interesting experiment.
Re: Mastering Time-to-Market with Haskell
#7Why comparisons to C#/Java are mentioned so many times but not a single mention of F#?
Maybe with F# author would see decrease of development time compared to Haskell?
Re: Mastering Time-to-Market with Haskell
#8Earlier quoted context omitted.
Not the author and I don't know much besides a little dabbling in Haskell, but I'm sure that quote holds true, but only if the engineer already is an intermediate-advanced Haskeller, which probably translates to at least a couple years of daily Haskell work.
> but I'm sure that quote holds true How can you be sure? Have you seen Haskell shops churn out software significantly faster than others?
We use Haskell for "rapid application development" in the front office at Standard Chartered bank. We have excellent typed APIs over most of the banks systems and can assemble applications in hours or days, rather than weeks or months as is typical. The language is as succinct as python, but thanks to types, purity and many standardised abstractions, we get better code reuse and maintenance.
Re: Mastering Time-to-Market with Haskell
#9Although very fresh into the Haskell world myself, I tend to agree with the author that, when I know what I am doing, my Haskell code is usually written in less time and has less bugs. Having said that, Time-to-Market is only partially influenced by my-code, the biggest part is the code that I don't have to write, i.e. third-party libraries. In my Haskell adventures I am having trouble finding third-party libraries f…
I've found Scala to be a really good compromise–it has most of the functional features of Haskell, a wide array of existing libraries, and access to Java libraries when you need them. Plus implicits/typeclasses make it easy to extend otherwise annoying Java libraries to make them easier to use.
Re: Mastering Time-to-Market with Haskell
#10> In summary we've seen that: Haskell decreases development time... Have we actually seen that or have you just asserted that? Is this really true, and if it is, by how much? Haskell has been around for a couple of decades now, and has had least two hype cycles (I remember that when I was in university in the late '90s, Haskell was the next big thing). It does not seem to expand significantly even within organization…
It's hard to measure these types of things. But if you are interested, there is a related paper regarding that: http://haskell.cs.yale.edu/wp-content/uploads/2011/03/Haskel... Although the study in the paper isn't very practical, it's still an interesting experiment.
> It's hard to measure these types of things.
I'll settle for well-researched case studies.
> there is a related paper regarding that
That's a step in the right direction, but the paper doesn't discuss software development, but prototyping. We know that "theoretically aesthetic" languages do well in specification and prototyping.