On Tuesday 06 October 2015 16:25:27 Mike Bird wrote:
> On Tue October 6 2015 05:23:45 Lisi Reisz wrote:
> > Mirrors in Germany (Nik) and Belgium would be great!
>
> Hi Lisi,
>
> How many TDE mirrors are needed in Germany?

How many TDE mirrors are needed in the USA?  How many are needed in the UK? 
(Also bad on infrastructure by international, developed and developing, world 
standards, though not as bad as the USA.)  Let's get rid of the inefficient 
ones in favour of the efficient.

> Belgium?
>
> Luxembourg?
>
> China, India, Brazil, Russia, South Africa, and Australia?

China and Russia are just stupid suggestions.  How many people in India, 
Brazil, South Africa and Australia use TDE?  And what is the infrastructure 
like in those countries?

> How much does each mirror cost in terms of IPv4 addresses,
> global warming, person hours, and crucial Tim hours?

How much longer will we be using IPv4?  Many people already don't, 
particularly techie type firms with forward thinking ISPs.  If you have no 
IPv6 in the States - that is something I don't know - then that is another 
reason for moving to other countries.  If you have, why are you taking IPv4 
into account?

https://www.google.com/intl/en/ipv6/statistics.html
https://en.wikipedia.org/wiki/IPv6_deployment

Global warming - not much, since the servers have already been manufactured 
and are running already.  It is only marginal cost that we need to worry 
about.  In Germany, the marginal cost in global warming is zero.  They have 
an oversupply of renewable energy.  Another reason for using a German mirror 
if offered.

We need to organise help for Tim so that fewer of his crucial hours are used.  
If he has to do everything, then indeed the project can't survive.

> How much load does each new mirror impose on other parts of
> TDE infrastructure?

Then replace expensive mirrors, in terms of Tim's time and global warming, 
with cheaper ones in better locations.  Why on earth have we got two in one 
place in Kent????  Incidentally, all those queriable ping tests have merely 
confirmed my prior experience of Kent University's mirrors, and I shall not 
for now be changing to them.  They have had time to improve since I last used 
them, but it doesn't look as though they have.

> What is the optimum number of mirrors for TDE - now and
> for the next year or two?

You tell me.  But lets have better and more reliable ones if we can.

You seem to resent any offers of help from anyone, not just me.  Why?

Lisi