- Glasgow-haskell-users

Deriving regression or bad code?
by Ron Alford 12 Apr '10

12 Apr '10

I've attached the simplest example of my code that used to compile in GHC 6.10 now gives the error in GHC 6.12.1: baddata.hs:33:14: No instances for (Data Const, Data Var) arising from the 'deriving' clause of a data type declaration at baddata.hs:33:14-17 Possible fix: add an instance declaration for (Data Const, Data Var) or use a standalone 'deriving instance' declaration instead, so you can specify the instance context yourself When deriving the instance for (Data (Domain e g)) If I replace: > data Domain e g = Domain > (Expr (Const :+: Var)) > deriving (Data, Typeable) with: > data Domain e g = Domain > (Expr (Const)) > deriving (Data, Typeable) then everything compiles. If nothing else, the error message is misleading. -Ron

2 5

GHC internal error
by Philip Weaver 06 Apr '10

06 Apr '10

Hi all, A program that I built with GHC is crashing at runtime with the following error: internal error: eval_thunk_selector: strange selectee 12 I have not found much via Google, just an old post regarding ghc 6.0 that recommends cleaning and rebuilding. I have cleaned the build directory and re-built the executable, and I still get the error. The program is a compiler, and this error began showing up after I added support for memoization of results using the ST Monad and STRefs. For example, the following function creates a reference to a thunk: newMutableRef m = do ref <- newSTRef $ error "access ref too soon!" let m' = do r <- m writeSTRef ref (return r) return r writeSTRef ref m' return $ ref Here's some information about my system: - GHC 6.12.1 - Mac OS X 10.6.3 (Snow Leopard) So, what could cause this error? Any suggestions that can help me narrow down the problem would be greatly appreciated. If this is in fact a GHC bug, I'd like to find a workaround for now and also reproduce it in a small example so I can file a bug report. - Philip

3 3

ClockTime closure linking issue
by Gracjan Polak 05 Apr '10

05 Apr '10

Hi all, Probable bug in GHC, I want to inquire before I report it proper. Did anybody see something like this: c:\Sources\happstack\happstack\templates\project>cabal build [1 of 1] Compiling Main ( Setup.hs, dist\setup\Main.o ) Linking .\dist\setup\setup.exe ... Preprocessing executables for guestbook-1.0... Building guestbook-1.0... [11 of 11] Compiling Main ( src\Main.hs, dist\build\guestbook-server \guestbook-server-tmp\Main.o ) Linking dist\build\guestbook-server\guestbook-server.exe ... dist\build\guestbook-server\guestbook-server-tmp\GuestBook\State2.o:fake:(.text+ 0x6d89): undefined reference to `happstackzm0zi4zi3_HappstackziStateziClockTime_ constrZMabmKZN_closure' dist\build\guestbook-server\guestbook-server-tmp\GuestBook\State2.o:fake:(.text+ 0x6dd1): undefined reference to `happstackzm0zi4zi3_HappstackziStateziClockTime_ dataTypeZMabmJZN_closure' [...lots of same messages skipped...] collect2: ld returned 1 exit status Application is from http://patch-tag.com/r/mae/happstack/snapshot/current/content/pretty/happst… The Glorious Glasgow Haskell Compilation System, version 6.12.1 System Windows Vista 32bit. Help! -- Gracjan

2 2

Different behavior of GHC 6.10.1 and Hugs (Sep 2006)
by Vladimir Reshetnikov 03 Apr '10

03 Apr '10

Hi list, GHC 6.10.1: Prelude> :t let f x y = return x == return y in f let f x y = return x == return y in f :: (Eq (m a), Monad m) => a -> a -> Bool Hugs (Sep 2006): Hugs> :t let f x y = return x == return y in f ERROR - Ambiguous type signature in inferred type *** ambiguous type : (Eq (a b), Monad a) => b -> b -> Bool *** assigned to : f Who is right? -- Thanks Vladimir

2 1

Profiling bug in 6.10.4
by Gregory Wright 01 Apr '10

01 Apr '10

Hi, I have a program (attached) that is relatively simple, but numerically intensive. It computes the abundances of the chemical elements generated by big-bang nucleosynthesis. At the moment, the executable takes no command line arguments, it simply runs the standard model. When I build the program without profiling, it produces the right answer. When I build it with profiling (runhaskell configure --enable-executable-profiling) it runs without error, but gives completely incorrect numerical results. I'm using ghc 6.10.4 built from source using MacPorts. The program is mostly self contained, but uses the hmatrix package to solve a linear system. Note also that the "Vector" type is hmatrix's Data.Packed.Vector and not the one from the more familiar "vector" package. I was wondering whether this profiling problem is known. Best Wishes, Greg

4 8

Plans for GHC 6.12.2
by Ian Lynagh 01 Apr '10

01 Apr '10

Hi all, This is a summary of our plans for GHC 6.12.2. We plan to put out a release candidate 6.12.2-rc1 shortly and, assuming there are no serious regressions found, 6.12.2 will follow soon after. The bugs that we're planning to fix for 6.12.2 are the high priority tickets in the 6.12.2 milestone: http://hackage.haskell.org/trac/ghc/query?status=new&status=assigned&status… If there is a bug not in that list that is important to you, please let us know. Thanks Ian

4 7

incorrect deprecation warning
by Peter Hercek 30 Mar '10

30 Mar '10

Hi, I got this deprecation warning: Warning: Module `GHC.Handle' is deprecated: use GHC.IO.Handle.Base instead But I think it should have been: Warning: Module `GHC.Handle' is deprecated: use GHC.IO.Handle instead There is no GHC.IO.Handle.BAse in my installation of GHC 6.12.1. The problem (it it was not intended) is there in the current head too: ~/haskell/ghc.head/libraries/base 836> darcs what -u hunk ./GHC/Handle.hs 18 -- #hide -module GHC.Handle {-# DEPRECATED "use GHC.IO.Handle.Base instead" #-} ( +module GHC.Handle {-# DEPRECATED "use GHC.IO.Handle instead" #-} ( withHandle, withHandle', withHandle_, wantWritableHandle, wantReadableHandle, wantSeekableHandle, ~/haskell/ghc.head/libraries/base 837>

2 1

trouble with Trac wiki
by nr＠cs.tufts.edu 28 Mar '10

28 Mar '10

I'm trying to log into the trac wiki, and I'm logging in OK, but it says my privileges are restricted until I verify my email address. However, the promised opportunity to verify never arrives. I've checked the spam folder and various other places---not finding it. I could be doing something stupid, but I think I need help. Whom should I ask for help with my trac wiki credentials? Norman

2 1

dblatex fails to escape backslashes
by Thanos Tsouanas 27 Mar '10

27 Mar '10

Hello list. I'm trying to compile ghc 6.12.1, but the documentation part fails because dblatex can't escape all the backslashes that it should. (I've tried dblatex-2.10, -2.11 and -2.12.) A similar problem has been discussed earlier [1], but even with Christiaan's suggested solution it fails at Cabal: "/home/thanos/local/bin/dblatex" libraries/Cabal/doc/Cabal.xml --ps -o libraries/Cabal/doc/Cabal.ps /home/thanos/local/lib/python2.6/site-packages/dbtexmf/dblatex/grubber/util.py:8: DeprecationWarning: the md5 module is deprecated; use hashlib instead import md5 Build the book set list... Build the listings... XSLT stylesheets DocBook - LaTeX 2e (0.2.12) =================================================== No template matches subtitle in article. Build Cabal.ps This is pdfTeX, Version 3.1415926-1.40.10 (PuffTeX-2009) entering extended mode latex failed Cabal.tex:1747: Undefined control sequence \Program. Cabal.tex:1747: leading text: \nolinkurl{C:\Program} Cabal.tex:1978: Undefined control sequence \Program. Cabal.tex:1978: leading text: } Cabal.tex:1978: Undefined control sequence \Haskell. Cabal.tex:1978: leading text: } Cabal.tex:1986: Undefined control sequence \Documents. Cabal.tex:1986: leading text: } Cabal.tex:1986: Undefined control sequence \user. Cabal.tex:1986: leading text: } Cabal.tex:1986: Undefined control sequence \Application. Cabal.tex:1986: leading text: } Cabal.tex:1986: Undefined control sequence \cabal. Cabal.tex:1986: leading text: } Cabal.tex:2035: Undefined control sequence \Program. Cabal.tex:2035: leading text: } Cabal.tex:2035: Undefined control sequence \Haskell. Cabal.tex:2035: leading text: } Unexpected error occured What version of dblatex and what tweeks have you people used to actually build the documentation successfully? (Please note, that the error is not breaking the process of make. make continues it's job ignoring this error.) [1] http://www.mail-archive.com/glasgow-haskell-users@haskell.org/msg17198.html -- Thanos Tsouanas http://mpla.math.uoa.gr/~thanos/

3 2

Re: Libraries Digest, Vol 79, Issue 31
by Louis Wasserman 22 Mar '10

22 Mar '10

Yo, > * I'm not comfortable with having two redundant modules, one for Min- and > one for MaxQueue. It is possible to implement both in the same > data structure and still get type errors when e.g. 'union' on a Min- and > a MaxQueue (I did that in [1], please have a look at the latest version). My > solution needs type families, which would not be good for containers, but > I'm pretty sure there's a containers-compatible solution. I'm pretty sure there won't be a containers-compatible solution, certainly not a solution compatible with the style of containers as it's currently written, because your solution depends on exporting functionality with type classes. If we're going to change that style in containers, it should be done all at once, in a complete rewrite. Whether or not priority queues get added to containers, I'm at least going to keep the style *consistent* with containers. With regards to benchmarks, we examined a comparison with leftist heaps earlier. The biggest complaint is the O(log n) insert time, which is really kind of icky. Here was Milan's conclusion: LeftistByMilan is actually pretty good for small inputs (up to ~ 216), and > the implementation is super-trivial compared to Binomial. But when the list > gets bigger, the complexity gets worse. I think this is beause of O(log N) > inserts -- the Binomial implementation with O(1) amortized inserts allocates > much less memory and does not need to run a GC. I vote for current (v5) > Binomial implementation, even if the O(1) amortized inserts works only > non-persistently (ie. under heavy persistent usage, leftist heaps might be a > _little_ better). Furthermore, I added your implementation to the benchmarks that we'd been using previously, which was essentially a full heapsort, rather than just searching for the five smallest elements as in your quickie benchmark. By this comparison, your implementation is moderately slower<http://code.haskell.org/containers-pqueue/bench-chart.pdf>. (The source code and data for this benchmark version are now in the code.haskell.org/containers-pqueue directory.) Though your implementation might be faster for some particular cases of problems, I think this is a fair comparison overall, and I'm definitely not willing to accept leftist heaps' O(log n) insert time given the slower overall performance for heapsort. This is what the Haskell Platform is for. No real need to add more > stuff to containers; if anything, the libraries stored there should be > decoupled so that e.g. the recent Data.Map proposals wouldn't disturb > programs not using Data.Map. > Again, my preference is to add pqueues to containers, and then if containers gets decoupled, then so will priority queues. If I'm going to split off a separate package for priority queues, it'll be because I've been convinced that it ought to be separated from containers, period -- not just because people think containers should be broken up, and this is a good place to start. Meh. Louis Wasserman wasserman.louis(a)gmail.com http://profiles.google.com/wasserman.louis

2 2