[tor-dev] PRELIMINARY: [PATCH 3/3] Replace 'TorDNSEL.System.Timeout' with 'System.Timeout'.

[Lunar]

Nikita Karetnikov:
> Well, I agree.  However, even obvious changes might be useful in the
> long term.  The GNU Coding Standards, which I use as a guide, suggest
> the following: "Subsequent maintainers will often search for a function
> name to find all the change log entries that pertain to it..." [1]  (For
> instance, cgit allows that.)  It also helps when you're trying to fix
> bugs.  Moreover, I often spot mistakes when I'm writing such messages.

I both agree, but I also try to keep in mind than the (at least most)
GNU Coding Standards were thought in a time when we did not have version
control solutions as good as git. Using `git blame` or `git log -S` or
`git log -G` will actually work better than a file by file summary. If
you throw in the `-M` option, it'll work accross renames, for examples.

> Attached.  Do you have any other comments?
> […]
> From 87ad0fe6f010b251aed042618dc82651294a477a Mon Sep 17 00:00:00 2001
> From: Nikita Karetnikov <nikita at karetnikov.org>
> Date: Thu, 25 Jul 2013 20:29:08 +0000
> Subject: [PATCH 1/2] Reimport internal GHC functions from new locations.
> 
> ---

Why not put what you wrote in your description email in there? This was
a good explaination of why the change was actually needed! :)

> "GHC was recently changed to not allow you to use newtypes in FFI
> imports unless the constructor of the newtype is in scope." [1]
> 
> [1] http://ghc.haskell.org/trac/ghc/ticket/5610


I am looking forward to the moment you'll have a patch set that builds.

-- 
Lunar                                             <lunar at torproject.org>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 836 bytes
Desc: Digital signature
URL: <http://lists.torproject.org/pipermail/tor-dev/attachments/20130725/9984bc24/attachment.sig>