- Hugs-Bugs - Haskell.org

non-bug report
by Doug McIlroy 24 Sep '03

24 Sep '03

I just installed hugs98-Nov2002 on Linux. Absolutely smooth--a great improvement over hugs 1.4, which needed considerable fiddling back in 2001. yours delightedly, Doug McIlroy

1 0

Re: Hugs and osX 10.3?
by Alastair Reid 23 Sep '03

23 Sep '03

[moved from hugs-authors to hugs-bugs (which is where most development-related discussion takes place] > I use Mac OS X, and Apple plans a new release of 10.3 > very soon. Is Hugs still in development (I see the > last version was Nov 2002), and is there a OS X 10.3 > release planned? Hugs is still very much in development and is available for MacOS X 10.2. A new source release of Hugs is due in the next week or so. Binary releases for popular platforms are usually donated by volunteers within a few days of a source release. -- Alastair Reid www.haskell-consulting.com

2 1

qualified import of prelude
by Andrew Frank 22 Sep '03

22 Sep '03

i seem to have a problem with a qualified import of the prelude in hugs nov03, haskell extensions, under windows: Reading file "C:\gisTheoryCodeSept03\lib\Language\Value240.hs": ERROR "C:\gisTheoryCodeSept03\lib\Language\Value240.hs":207 - Ambiguous variable occurrence "quotRem" *** Could refer to: Language.Value240.quotRem Hugs.Prelude.quotRem the relevant code is: instance Ring Float where quotRem a b = (toFloat i, toFloat (b * f)) where ab = a/b i::Int (i,f) = Prelude.properFraction ab rem a b = f where (i,f) = quotRem a b --here is line 240 with the error (*) = (Prelude.*) quot a b = i where (i,f) = quotRem a b div = error "div not implemented for float" mod a b = error "mod not implemented for float" i had a similar error (redefining a class Integral and a complaint in the export list of ambiguit between the local definition and the qualified import). it seems that qualified import of the prelude is not properly handled? unfortunately this is not a small file and i did not manage to isolate the problem quickly in a small example. can you see what is wrong here? the file value240.hs is attached. thank you for your attention andrew Andrew U. Frank Geoinformation E127 phone: +43 1 588 01 12710 TU Vienna secr. +43 1 588 01 12700 Gusshausstrasse 27-29 fax +43 1 588 01 12799 A-1040 Vienna Austria cellular phone +43 676 41925 72 http://www.geoinfo.tuwien.ac.at/persons/frank/frank.html

2 1

-W option
by Paul Hudak 19 Sep '03

19 Sep '03

This isn't really a bug report, but I wasn't sure where else to ask the following question. Hugs used to handle all of my SOE code with no problem. Now, however, I get the error message: ERROR "C:\Program Files\Hugs98\lib\exts\MarshalUtils.hs":2 - Syntax error in module definition (unexpected keyword "with "; possible cause: -W option not used.) Indeed, when I invoke Hugs -W from a command line, everything works fine. So the problem is apparently because Hugs used to allow this language extension as a default, and now doesn't. I would hate to tell everyone using SOE that they have to invoke Hugs from a command prompt, so I'm looking for an alternative suggestion if you have any. Of course, I could rewrite the code, but it's actually Alastair's code, and I'm not sure how big of an effort it is... Thanks, -Paul

5 8

illegal export of lone data constructor
by Andrew Frank 17 Sep '03

17 Sep '03

i am moving fom the nov 99 to the nov 02 hugs.exe (under windows) when recompiling old files i find two problems which i do not understand: i want to export (:) from the prelude because i do not use the standard prelude usually. module XX ( (:), ....) does not work (it did before). now i get the error "illegal export of lone data constructor" i do not see any hint in the report (nov 99) that (:) cannot be exported. (it may be unusual, but not illegal?) the goal is to export certain functions from the prelude (especially (:), which cannot be redefined) but not all of it. how can this be achieved? (minor point: my report says that the prelude module is always available as qualified import (p. 71) the hugs now (2002) seem to require an explicite import qualified (after a import Prelude hiding (stuff)) ) i would appreciate an answer on the first question (export of (:)) andrew Andrew U. Frank Geoinformation E127 phone: +43 1 588 01 12710 TU Vienna secr. +43 1 588 01 12700 Gusshausstrasse 27-29 fax +43 1 588 01 12799 A-1040 Vienna Austria cellular phone +43 676 41925 72 http://www.geoinfo.tuwien.ac.at/persons/frank/frank.html

4 3

winhugs
by Richard Lawlor 16 Sep '03

16 Sep '03

1 0

type synonym or fun dep problem
by Iavor Diatchki 16 Sep '03

16 Sep '03

apologies if this gets to the list twice. --- hello, there seems to be something wrong with the implementation of functional dependencies, or perhaps the expansion of type synonyms. i am using hugs (september 2003). the program: > module Test where > > class C a b | a -> b where > mk :: IO a > eq :: a -> a -> IO Bool > > type F = IO > > instance C a x => C [a] (F a) > > test ys = do x <- mk > xs <- mk > eq ys (x:xs) causes the error: ERROR "test.lhs":18 - Constraints are not consistent with functional dependency *** Constraint : C [a] (IO a a) *** And constraint : C [a] (F a) *** For class : C a b *** Break dependency : a -> b if 'type F = IO' is replaced with 'type F a = IO a' one gets *** Constraint : C [a] (IO [a]) if i inline the definition of 'F' things seem to work. if i remove the assumption on the instance: *** Constraint : C [a] (b a) if i remove the assumption on the instance, and eta-expand the type synonym: *** Constraint : C [a] (IO (IO a)) bye iavor -- ================================================== | Iavor S. Diatchki, Ph.D. student | | Department of Computer Science and Engineering | | School of OGI at OHSU | | http://www.cse.ogi.edu/~diatchki | ==================================================

1 0

Hugs home page broken
by Alastair Reid 15 Sep '03

15 Sep '03

While checking the links from my own pages, I noticed that the link: http://www.haskell.org/hugs/ gives a directory listing instead of a home page. The home page is there however. It is listed as index.htm and clicking on it gives the expected page. I suspect the problem can be fixed by either one of: 1) Rename index.htm to index.html 2) Tweak the configurations of haskell.org's web server to treat either index.html _or_ index.htm as the default page in a directory. -- Alastair Reid

1 0

type synonyms/fun deps
by Iavor Diatchki 15 Sep '03

15 Sep '03

hello, there seems to be something wrong with the implementation of functional dependencies, or perhaps the expansion of type synonyms. i am using hugs from july 2003. the program: > module Test where > > class C a b | a -> b where > mk :: IO a > eq :: a -> a -> IO Bool > > type F = IO > > instance C a x => C [a] (F a) > > test ys = do x <- mk > xs <- mk > eq ys (x:xs) causes the error: ERROR "test.lhs":18 - Constraints are not consistent with functional dependency *** Constraint : C [a] (IO a a) *** And constraint : C [a] (F a) *** For class : C a b *** Break dependency : a -> b if 'type F = IO' is replaced with 'type F a = IO a' one gets *** Constraint : C [a] (IO [a]) if i inline the definition of 'F' things seem to work. if i remove the assumption on the instance: *** Constraint : C [a] (b a) if i remove the assumption on the instance, and eta-expand the type synonym: *** Constraint : C [a] (IO (IO a)) bye iavor -- ================================================== | Iavor S. Diatchki, Ph.D. student | | Department of Computer Science and Engineering | | School of OGI at OHSU | | http://www.cse.ogi.edu/~diatchki | ==================================================

1 0

winhugs in the Hugs98-Nov2002
by Paqui Lucio 12 Sep '03

12 Sep '03

Hi, When I have installed the Hugs98-Nov2002, the interface for windows "winhugs" is not installed (as it was in the Dec2001 version). Is that normal? Thanks in advance, Paqui ------------------------------- Paqui Lucio Dpto de LSI Facultad de Informática Paseo Manuel de Lardizabal, 1 20080-San Sebastián SPAIN Tfn: (+34) (9)43 015049 Fax: (+34) (9)43 219306 -------------------------------

1 0