Hi Al,

when GitHub (a big open source repository using, of course, Git) was offered free access to the Rackspace Cloud, they turned it down, and asked for (and got) a managed server. Why? Because the latency for data access against a server in the cloud would hurt their user experience. There are a lot of similar commentaries on the Net if you go looking for them (which I did after reading about the GitHub experience).

Hank

>>Al, servers are getting bigger and bolder, but so are local PCs which is the point I made earlier. And still you can hobble a $250K SS2008 server with heavy munging unless you use resource governor (something that presumably was added because people apart from me have experienced resource overloading ;-) ) in which case the customer has to learn to wait when they never had to before. Why? Because it's better to do it on the server?
>>
>>Of course there are other examples where doing everything on the server is a better idea. Nobody has ever disagreed with that.
>
>I was mainly pointing out that things have changed, a lot, since a few years ago. Taking that theme a little further, the near future offers cloud computing, which you can consider the ultimate back end. For all but the largest organizations, server scaling is essentially unlimited if a suitable cloud back end is available.
>
>Sure, it's possible to devise scenarios where doing everything on the backend, or on workstations, is dumb. In a perfect world, a smart network would automatically load balance and make best use of all resources (both server and workstation) but AFAIK we're not there yet.
>
>ISTR at one point you were spreading the gospel of the workstation being dead, and the future being everyone using smartphones attached to sessions in a cloud somewhere. Has anything happened to give you new respect for workstations (like, say, an overloaded server? ;-))