[Haskell] More work to be done re: haskell.org move?

[Claus Reinke]

> Everyone keeps saying "the move is complete" but I'm not sure it's
> really true.  I understand in theory that people were given ample time
> and warning to migrate content, but in practice it has not worked.
> .. Are there any
> plans to copy over things like this?  Relying on "maintiners" to do it
> has obviously not worked.  I stand ready to help but am not quite sure
> what needs to happen from a technical point of view.

Thanks! It would be good to get an idea of the amount of damage,
in terms of differences between old and new server, before the old
server goes offline (has the offline backup been verified, btw? and
does it include all non-wiki content as well, such as accounts to be 
deleted, and tools pages?):

- for static content, an rsync *dry-run* on the server directories
    would give an idea of differences, provided those directories
    are somehow identifiable (not spread over user accounts)

- for dynamic content, web-crawling both servers and comparing
    the availabe URLs might be necessary (the problem here would
    be to identify all roots for crawling - with multiple haskell.org
    servers, not everything on www.haskell.org might be linked 
    from somewhere on www.haskell.org; eg, non-wiki pages for
    hackage packages or Haskell tools..)

- checking the server logs for the new server will only give a
    partial picture, but using the server logs for the old server,
    and checking whether all successful URL requests also work
    on the new server might help to spot gaps (build a Map of
    new-server URLs, check old-server log URLs for presence
    in that Map)

Then posting the list of missing URLs here might alert those
who care about them (or not, but at least fewer things would
disappear silently). It would at least give some concrete idea
of how complete the move really is.

Claus