Greetings, and thanks for reading the seventh issue of HWN, a weekly newsletter for the Haskell community. Each Tuesday, new editions will be posted (as text) to the Haskell mailing list and (as HTML) to The Haskell Sequence.
Why is HWN a day late this week? Your HWN editor was stuck in some large airports that had a surprising lack of Wifi. Sigh.
Binary parser combinators. Einar Karttunen asked about a binary parser combinator interface for network protocol parsing. Malcolm Wallac pointed out that nhc98 has a Binary library with a "<<" operator that could be useful.
Windows programming in Haskell. Brian McQueen asked about Windows programming in Haskell, including access to the Windows registry, APIs, and communicating with other Windows apps. Several suggestions relating to Hugs were offered, including .NET support and some libraries.
Functional vs. Imperative. Dhaemon began an interesting discussion by asking for some help understanding functional vs. imperative approaches. Several people commented on the IO monad, and how it is still a functional interface even though it may appear imperative at first glance.
Mixing monadic and non-monadic functions. A long thread on this subject appeared in the Haskell list this week. Rather too long to summarize here -- take a look at the link.
Language workbenches. Yoel Jacobsen wrote about an article on language workbenches, in which configuration files are actually valid code in a general-purpose language. Yoel went on to ask about doing this in Haskell. Some suggestions, such as hs-plugins, were offered.
Types in Template Haskell. Gracjan Polak posted about some trouble with typing in Template Haskell. Several responses regarding quoting types were posted, including a reference to Simon Marlow's "update" paper.
Web applications. Gary began a large discussion by asking about writing Web applications. Several options were mentioned, including Wash and HAppS. S. Alexander Jacobsen mentioned that he will be launching a commercial chat service using Haskell and AJAX with HAppS as the underlying core.
Calling Haskell from C++. Felix Breuer wrote about some trouble calling into Haskell from C++ programs. Several suggestions were provided, mostly relating to C++ name mangling.
What gets profiled? Niels began a discussion on the use of profiling features by commenting that profiling didn't seem to show the problem in his own code. Several suggestions regarding memory use and possible reasons that profilers might miss things were provided.
GHC 6.4.1. Simon Marlow posted an update on GHC 6.4.1. Though more bug reports have been rolling in while he was away, only a few are blockers for 6.4.1. The tentative release date is September 19.
Data.Monoid. Ross Paterson proposed replacing an instance of Data.Monoid. There was some discussion about whether the old or new instance was better.
Want to continue reading HWN? Please help us create new editions of this newsletter. Please see the contributing information, or send stories to hwn -at- complete -dot- org. There is also a Darcs repository available.