R devel - Sep 2012 - French Toulouse CRAN mirror always down

Hi!

One of the three French CRAN mirrors, cran.cict.fr (Toulouse) seems to
be down most of the time, and at least really not reliable. It has
currently been down for 28 days (!) [1], and I know that over two years
I've rarely been able to use it (no response, slow, hangs...).

I'm sure the maintainers of this mirror mean well, but the result is
really negative for French R users. This week I observed one user that
failed to install a package because she selected that mirror and it was
not available. Such a minor detail can give a very poor impression to
beginners.

I suggest that either the maintainers of this mirror explain why it does
not work and provide solutions to improve the situation, or it should be
removed from the list until it works reliably. Not as a punishment of
course, but for practical considerations.

At the very least this mirror should be moved down the list: at the
moment it's the first of the three French mirrors, while the two others
work very well. If the mirrors were sorted by the alphabetical order of
their full names, it would be the last one, and it would be bite French
users less often, while still acting as a fallback solution.


Thanks for taking care of this list



1: http://cran.r-project.org/mirmon_report.html#fr

>>>>> On Fri, 28 Sep 2012 09:37:39 +0200,
>>>>> Milan Bouchet-Valat (MB) wrote:
  > Hi!
  > One of the three French CRAN mirrors, cran.cict.fr (Toulouse) seems to
  > be down most of the time, and at least really not reliable. It has
  > currently been down for 28 days (!) [1], and I know that over two years
  > I've rarely been able to use it (no response, slow, hangs...).

  > I'm sure the maintainers of this mirror mean well, but the result is
  > really negative for French R users. This week I observed one user that
  > failed to install a package because she selected that mirror and it was
  > not available. Such a minor detail can give a very poor impression to
  > beginners.

  > I suggest that either the maintainers of this mirror explain why it does
  > not work and provide solutions to improve the situation, or it should be
  > removed from the list until it works reliably. Not as a punishment of
  > course, but for practical considerations.

  > At the very least this mirror should be moved down the list: at the
  > moment it's the first of the three French mirrors, while the two
others
  > work very well. If the mirrors were sorted by the alphabetical order of
  > their full names, it would be the last one, and it would be bite French
  > users less often, while still acting as a fallback solution.

It has already been removed a couple of days ago, have a look e.g. at

http://cran.r-project.org/mirrors.html

will need some time to propagate to all places.

Best,
Fritz