Rick Rashid: Leading Microsoft into the Future with Research

The Discussion

• 

  well, I dont know. I think MSR is falling short in some ways.
   
  VMWare has got the load balancing of their VMs working better than what MS has. Has MSR been doing VM research? 
  
  Was MSR instrumental in  .NET and managed code? The C9 interviews with Anders imply that when he ( Anders ) arrived at the company in the late 1990s, MS had little in terms of knowing what the followup to COM should be.
  
  Computer useability.  That is something MSR should be working on, no? As is it, as someone who is a programmer, I am terrified of having to config active directory, sql server, IIS, network connectivity on my home office network.  Where is the software agent I can dispatch onto my network with instructions to config things so that user X, working from PC Y, can have authority level whatever to the Z database? ( actually, I should be able to write that, so stand down  )
  
  In practical terms, in .NET I can only code in C# or VB. If I like C# but would like to add 10 or so additional features, I have to go back to square one and write an entirely new language.  And then there is all the work for fitting your C-better language into VS and the interactive debugger. Is MSR working on a computer language that can be defined as a base class for a new language? Where the C-better langauge could be written by removing and adding features to the base C# language.
  
  Back to VM, is MSR working on what I think is called grid computing? VM load balancing falls down in that the entire OS and all its running processes have to be moved enmass to the underutilized server.  Better to be able to move a single process to another server. Where the OS is running not just on multiple cores, but on multiple servers.  That would be one sophisticated thread pool! 
  
  Check out this recent post on the framework.aspnet newsgroup: 
  http://groups.google.com/group/microsoft.public.dotnet.framework.aspnet/browse_frm/thread/62463bd993e98333/?hl=en##
  
  esp this message:
  http://groups.google.com/group/microsoft.public.dotnet.framework.aspnet/msg/6452dab213d47e7a?hl=en&&
  
  and the response:
  http://groups.google.com/group/microsoft.public.dotnet.framework.aspnet/msg/de6f096d70c5fba8?hl=en&&
  
  as I understand it, the experts were advising to run IIS on a different server than the domain controller. The OP said it is nuts that to host a secure website from his basement he should have to run at least two servers.  If the answer is multiple VMs running on a dual or quad core server, that addresses the need for excess hardware to run MSFT software.  But where is MSR in its basic OS research making it so that processes can be securely isolated from each other?  The task of configing multiple OSes, each in its own VM, seems more cumbersome than dealing with a single robust OS running all of your servers.
  
  dont get me wrong. I enjoyed the interview. Where is our man Rory?
  
  -Steve Richter
  
  

• 

  Very Cool Interview.
  

• 

  Hopefully you got more videos from there!
  
  Good interview.

• 

  Chadk wrote:

  Hopefully you got more videos from there!
  
  Good interview.

  
  
  Thanks. Much more to come from Cambridge. Check out the latest installment on the home page.
  
  C

• 

  Very clear and informative Rick.  Thank you.

• 

  I do respect the work that Rashid did when he was at the CMU, he is a big expert in OS development, but pretending that he is the main inventor of Mach is really not honest and that magically Apple has used his "work" to build Mac OS X is quite ridiculous. Mac OS X' s father and architect, Avie Tevanian is also the main architect of MACH and is actually known to be the principal coder of it. Just look at the MACH source code used in Mac OS X (Mac OS X kernel is opens source), they all refer to Avie Tevanian as the programmer. It is not also surprising that Avie decided to use Mach as kernel foundation for Mac OS X, he knew where he was going and what he was using. In other words, Rashid deserves little credit to the development of Mac OS X in contradiction to what he is pretending, even if he was involved in Mach, sorry, i think truth has to be said. Yes Rashid is an important person in the development of Mach, but he failed to say that he is far to be the only one, he should have mentioned at least Avie, i mean history has be said properly or not at all........

• 

  Hakime, you might want to edit Mach-related article on Wikipedia (http://en.wikipedia.org/wiki/Mach_kernel), so it would be consistent with your point of view. As of 5 minutes ago it read '..The lead developer on the Mach project, Richard Rashid, [ship]... Another of the original Mach developers, Avie Tevanian, ...'.
  
  Also, are you positive that you are not looking at the name of person who made code merge into the branch that was used to build OSX? 
  

• 

  sokhaty wrote:

  Hakime, you might want to edit Mach-related article on Wikipedia (http://en.wikipedia.org/wiki/Mach_kernel), so it would be consistent with your point of view. As of 5 minutes ago it read '..The lead developer on the Mach project, Richard Rashid, [ship]... Another of the original Mach developers, Avie Tevanian, ...'.
  
  Also, are you positive that you are not looking at the name of person who made code merge into the branch that was used to build OSX? 

  
  
  The name in the code file is not always the guy who had the ideas...
  
  Great interview, btw.