February 2009 - cabal-devel

[Hackage] #485: preference on base version should only be applied for open constraints
by Hackage 18 Dec '09

18 Dec '09

#485: preference on base version should only be applied for open constraints ---------------------------------+------------------------------------------ Reporter: duncan | Owner: Type: enhancement | Status: new Priority: normal | Milestone: Component: cabal-install tool | Version: 1.6.0.1 Severity: normal | Keywords: Difficulty: unknown | Ghcversion: Platform: | ---------------------------------+------------------------------------------ Currently we apply the preference on the base version. That means that if you have base 3 and 4 available then for dependencies like: {{{ build-depends: base >= 2 }}} or {{{ build-depends: base >= 3 && < 5 }}} Then we end up picking base 3. That makes perfect sense in the first case. In the first case the developer clearly did not think about base 4. However in the second case they clearly have thought about it and yet we end up picking base 3. On balance it is probably better to pick base 4 in the second case even though both are allowed. One suggestion for how to handle soft constraints on the version of base is to say that we only apply the preference if the version is not bounded above. If it is bounded above then we use no preference and end up picking the highest version available. This policy cannot be expressed in the simple preferred versions syntax. It's probably ok to hard code it for base for now. This will require generalising the interface to the constraint solver to supply preferences in a slightly different form. It may need to be as a function from a list of possible versions to the selected version. The function would also be supplied with the package it's for, the dependency and the list of installed versions since it may want to take any of these into account and it is needed to express the above policy. This is related to the splitting of preferences into real preferences and hacks (see #483). -- Ticket URL: <http://hackage.haskell.org/trac/hackage/ticket/485> Hackage <http://haskell.org/cabal/> Hackage: Cabal and related projects

1 5

[Hackage] #385: emgm package incorrectly shown as "program" but is "library and program"
by Hackage 01 Dec '09

01 Dec '09

#385: emgm package incorrectly shown as "program" but is "library and program" --------------------------------+------------------------------------------- Reporter: guest | Owner: Type: defect | Status: new Priority: normal | Milestone: Component: hackageDB website | Version: Severity: minor | Keywords: Difficulty: normal | Ghcversion: 6.8.3 Platform: | --------------------------------+------------------------------------------- The summary says it all. In the HackageDB package overview, it is shown whether packages contain libraries and/or programs, but this does not seem to work correctly in all cases, as emgm demonstrates. kosmikus -- Ticket URL: <http://hackage.haskell.org/trac/hackage/ticket/385> Hackage <http://haskell.org/cabal/> Hackage: Cabal and related projects

1 4

[Hackage] #423: add option to fetch a single package without dependencies
by Hackage 27 Nov '09

27 Nov '09

#423: add option to fetch a single package without dependencies ----------------------------+----------------------------------------------- Reporter: guest | Owner: Type: enhancement | Status: new Priority: normal | Milestone: Component: Cabal library | Version: 1.2.3.0 Severity: normal | Keywords: Difficulty: normal | Ghcversion: 6.8.3 Platform: | ----------------------------+----------------------------------------------- Right now, {{{cabal fetch}}} fetches all missing dependencies of a package along with its source. In some cases, in particular when studying code, those dependencies are not useful. It would be nice to have an option, say {{{--no-deps}}} to fetch just the specified packages. (This is related to {{{cabal unpack}}} -- see #390.) -- Ticket URL: <http://hackage.haskell.org/trac/hackage/ticket/423> Hackage <http://haskell.org/cabal/> Hackage: Cabal and related projects

1 2

[Hackage] #403: Bad error message when running ./configure fails to find sh
by Hackage 10 Nov '09

10 Nov '09

#403: Bad error message when running ./configure fails to find sh ----------------------------------+----------------------------------------- Reporter: duncan | Owner: Type: defect | Status: new Priority: normal | Milestone: Component: Cabal library | Version: HEAD Severity: minor | Keywords: Difficulty: very easy (<1 hour) | Ghcversion: 6.10.1 Platform: Windows | ----------------------------------+----------------------------------------- For `build-type: Configure` we have to run `./configure`. On windows this will very often fail of course because they usually lack the `sh` program. Currently the error message users get is not great: {{{ sh: runGenProcess: does not exist (No such file or directory) }}} It would be nicer if they got a better message. Perhaps the way to do that is to catch the `DoesNotExist IOException` and re-throw a better message. Alternatively we could explicitly check for `sh` before invoking it. If it's not found on Windows then we could report that the package needs a unix-style command shell program eg from `mingw` to build. -- Ticket URL: <http://hackage.haskell.org/trac/hackage/ticket/403> Hackage <http://haskell.org/cabal/> Hackage: Cabal and related projects

1 1

[Hackage] #261: Marking packages deprecated
by Hackage 05 Nov '09

05 Nov '09

#261: Marking packages deprecated ----------------------------+----------------------------------------------- Reporter: nibro | Owner: Type: enhancement | Status: new Priority: normal | Milestone: Component: Cabal library | Version: 1.2.3.0 Severity: normal | Keywords: Difficulty: normal | Ghcversion: 6.8.2 Platform: | ----------------------------+----------------------------------------------- There should be a way to mark a package as '''deprecated''' so it doesn't show up on the main Hackage listing. If a newer package, with the same name, is uploaded then the older packages will in effect be defunct. But if you don't have a newer package with the same name, there is currently no way to "get rid of" packages that are no longer valid. There may be several reasons for why you would want to do this, likely most often because the contents of the package are superceded by functionality in some other package, with a different name. -- Ticket URL: <http://hackage.haskell.org/trac/hackage/ticket/261> Hackage <http://haskell.org/cabal/> Hackage: Cabal and related projects

1 6

[Hackage] #382: 'cabal ghci' mode
by Hackage 21 Oct '09

21 Oct '09

#382: 'cabal ghci' mode ---------------------------------+------------------------------------------ Reporter: guest | Owner: Type: enhancement | Status: new Priority: normal | Milestone: Component: cabal-install tool | Version: Severity: normal | Keywords: Difficulty: normal | Ghcversion: 6.8.3 Platform: | ---------------------------------+------------------------------------------ See http://www.haskell.org/pipermail/haskell-cafe/2008-October/049830.html and http://www.haskell.org/pipermail/haskell- cafe/2008-October/049913.html. Two things would be nice: "cabal ghci" loads the main-is file in ghci; and "cabal ghci Data.My.Module" loads the listed modules in ghci. Reiner -- Ticket URL: <http://hackage.haskell.org/trac/hackage/ticket/382> Hackage <http://haskell.org/cabal/> Hackage: Cabal and related projects

2 11

[Hackage] #282: profiling versions of libraries not managed well
by Hackage 30 Sep '09

30 Sep '09

#282: profiling versions of libraries not managed well ---------------------------------+------------------------------------------ Reporter: SamB | Owner: Type: defect | Status: new Priority: normal | Milestone: Component: cabal-install tool | Version: 1.2.3.0 Severity: normal | Keywords: profiling Difficulty: normal | Ghcversion: 6.8.2 Platform: | ---------------------------------+------------------------------------------ At the moment, if you want profiling builds of libraries which you have already installed, the only way to get them is to unregister and rebuild with profiling turned on. This is especially inconvenient when you have a program that depends on many different libraries... Ideally, I could just add a flag to rebuild a program with profiling, regardless of whether I'd installed profiling versions or not... Any ideas on how this should be managed? What configurations should be allowed? How do we make sure the same library version is used for the profiling build as was used for the normal build? When do we just give up and make the user figure it out? -- Ticket URL: <http://hackage.haskell.org/trac/hackage/ticket/282> Hackage <http://haskell.org/cabal/> Hackage: Cabal and related projects

1 7

[Hackage] #320: cooperation with Windows Vista's User Access Control
by Hackage 20 Sep '09

20 Sep '09

#320: cooperation with Windows Vista's User Access Control ----------------------------+----------------------------------------------- Reporter: Toxaris | Owner: Type: enhancement | Status: new Priority: normal | Milestone: Component: Cabal library | Version: 1.4.0.1 Severity: normal | Keywords: Difficulty: normal | Ghcversion: 6.8.2 Platform: Windows | ----------------------------+----------------------------------------------- Installation of a cabal package under Windows Vista with UAC activated may fail even when working as administrator, because cabal doesn't request the necessary rights to write to, e.g., "c:\Program Files\Haskell\<whatever>". A workaround is to start an "Administrator console" to start Cabal from a process which already acquired admin privileges. Cabal should request the necessary rights instead of failing. cabal-install should check in advance whether the target destination is writeable and fail early or request the necessary rights. (currently, cabal-install fails after compiling all dependencies. After starting cabal-install again from an Administrator console, everything has to be build again). {{{ C:\Downloads\Haskell\typeof-0.1.1>runhaskell Setup install Setup: C:\Program Files\Haskell\doc\typeof-0.1.1: copyFile: permission denied (P ermission denied) }}} -- Ticket URL: <http://hackage.haskell.org/trac/hackage/ticket/320> Hackage <http://haskell.org/cabal/> Hackage: Cabal and related projects

1 9

[Hackage] #240: setup install does not preserve MacOS X resource forks
by Hackage 11 Sep '09

11 Sep '09

#240: setup install does not preserve MacOS X resource forks ----------------------------+----------------------------------------------- Reporter: guest | Owner: Type: defect | Status: new Priority: normal | Milestone: Component: Cabal library | Version: 1.2.3.0 Severity: normal | Keywords: Difficulty: normal | Ghcversion: 6.8.2 Platform: | ----------------------------+----------------------------------------------- I use the attached Setup.lhs file to do some Mac-related magic (which I unfortunately do not understand) on files built with wxhaskell on MacOS X. The key point about this magic is that it allows me to run a GUI from the command line. Without doing the magic, the GUI will open, but it will not react to any user interaction. I think the problem is that whatever mechanism we use to copy the binary over (say to /usr/local/bin) does not preserve the magic. That is, if I run it directly after a runhaskell Setup install, I get the non-reactive GUI. If I do a cp geni /usr/local/bin by hand, then the GUI reacts, presumably because on MacOS X cp also copies the magic over. Sorry for being so vague. Let me know if I can help in any way. -- Eric Kow -- Ticket URL: <http://hackage.haskell.org/trac/hackage/ticket/240> Hackage <http://haskell.org/cabal/> Hackage: Cabal and related projects

1 4

[Hackage] #274: Add a 'build from sdist tarball' feature
by Hackage 23 Aug '09

23 Aug '09

#274: Add a 'build from sdist tarball' feature ----------------------------+----------------------------------------------- Reporter: guest | Owner: Type: enhancement | Status: new Priority: low | Milestone: Component: Cabal library | Version: 1.2.3.0 Severity: normal | Keywords: Difficulty: normal | Ghcversion: 6.8.2 Platform: | ----------------------------+----------------------------------------------- I've often seen Darcs repos where the sdist tarball was incomplete, or where a person had a tarball which was incomplete. I think this is because most people just never use the tarball; if they do ever use it, it's because it's what was downloaded from Hackage. The reason is that doing so is much more complex than 'configure; build; install': you need to configure, create an sdist tarball, cd into dist/; do a tar -xvf or whatever of the correct tarball; cd into *that* - and then you can configure/build/install. Assuming it is a working tarball. Now, of course you can automate it. After the first dozen times, I wrote myself a shellscript to do all that for me: function build_sdist { clean_configure && sdist && cd dist/ && untar *.tar.gz && cd `ls -t ./ | grep "/" | head -n 1` && clean_configure && build && hinstall; haddock && hinstall; } (This is pretty long and imperfect, but it would be even longer if all the other aliases and function definitions were included; I think they're pretty straightforwardly named though, so I haven't included them.) It seems to me that this would be, at the very least, a useful adjunct to cabal check - if it makes sense to see whether various fields are filled out sanely, surely it makes sense to see whether needed files are included/listed? It might be even better to make the normal 'runhaskell Setup build' do it from sdist, but this is harder to code (besides the forgoing, you'd need to copy in the configuration from the top-level, from when the user ran 'runhaskell Setup configure' and added whatever options) and possibly more confusing ("But Foo.hs *is* there, right there in src/! Why can't Cabal find it?"). So I'd be satisfied to do something like 'cabal checkdist'. -- gwern -- Ticket URL: <http://hackage.haskell.org/trac/hackage/ticket/274> Hackage <http://haskell.org/cabal/> Hackage: Cabal and related projects

1 5