home bbs files messages ]

Forums before death by AOL, social media and spammers... "We can't have nice things"

   alt.os.development      Operating system development chatter      4,255 messages   

[   << oldest   |   < older   |   list   |   newer >   |   newest >>   ]

   Message 3,684 of 4,255   
   Dan Cross to Scott Lurndal   
   Re: This newsgroup.   
   22 Mar 23 12:18:39   
   
   From: cross@spitfire.i.gajendra.net   
      
   In article ,   
   Scott Lurndal  wrote:   
   >cross@spitfire.i.gajendra.net (Dan Cross) writes:   
   >>So, is anyone here actually working on developing real, novel   
   >>operating systems on modern hardware here?   
   >   
   >Been there, done that, moved on.  Microkernels, distributed   
   >operating systems, massively parallel systems, MCP, Unix,   
   >Linux.   All have common features, all have unique implementations.   
   >   
   >I've sometimes considered that the golden age of operating system   
   >design is behind us.   
      
   I'm not sure I agree.  We're largely stuck with system   
   interfaces that were first described in the late 1960s through   
   the early 1980s, but hardware has changed dramatically since   
   then.  I don't think we've really stopped to ask ourselves   
   whether the systems we seem to have gravitated towards are still   
   a good fit for either the hardware we have, or the sorts of   
   programs we're running.   
      
   In particular, one sees lots of inate parallelism and   
   distribution across CPU, RAM, and IO fabrics at the hardware   
   level, implying asynchrony to use effectively.  And one also   
   sees lots of asynchrony at the application layer, often embedded   
   in language runtimes and the like.  But most systems provide a   
   highly synchronous system interface: so the asynchronous and   
   parallel application talks to the asychronous and parallel   
   hardware through this soda-straw of an interface designed to   
   make a multiprocessor machine with gigabytes of main memory,   
   a deep memory cache hierarchy, and an inherently asynchronous,   
   parallel IO architecture look like a PDP-11.   
      
   Interfaces like Linux's ioring implementation try to bridge this   
   gap within the context of an existing system, but are not   
   particularly orthogonal to the rest of the OS.  One wonders if   
   we can do better, but it will require a shift in the accepted   
   wisdom about the system's overall architecture.   
      
   	- Dan C.   
      
   --- SoupGate-Win32 v1.05   
    * Origin: you cannot sedate... all the things you hate (1:229/2)   

[   << oldest   |   < older   |   list   |   newer >   |   newest >>   ]


(c) 1994,  bbs@darkrealms.ca