home bbs files messages ]

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

   sci.space.tech      Technical and general issues related to      3,113 messages   

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

   Message 1,467 of 3,113   
   Kevin Willoughby to All   
   Re: Rover brains?   
   04 Feb 04 19:08:34   
   
   From: KevinWilloughby@acm.org.invalid.retro.com   
      
   In article , lex@cc.gatech.edu   
   says...   
   > Kevin Willoughby  writes:   
   > > A lot of research has been done on garbage collection. It is possible to   
   > > limit the CPU time consumed in garbage collection, allowing a real time   
   > > system (at the expense of perhaps requiring a bit more memory).   
   >   
   > To be pedantic, RT garbage collectors cost *time*, not memory.   
      
   Not necessarily. Long experience has taught us that if you have the   
   absolute minimum amount of memory, you spend a lot of time doing garbage   
   collection. Have a bit of extra memory means the garbage collector   
   doesn't have to work as hard. More important, it allows keeping some   
   extra free memory available under all conditions, insuring that the main   
   processing never has to wait for memory. Real-time programmers get   
   unhappy if their code has to endure unexpected waits. ("Fire the retro   
   rockets 15 seconds ago" isn't an acceptable result.)   
      
   It is possible to collect garbage as a background process. This isn't   
   new -- Dijkstra worked out the details for one approach back in the mid   
   1970s. If you have either excess CPU time or can afford a second CPU to   
   run concurrently, garbage collection need not take much extra time,   
   perhaps not any at all.   
      
      
   > Also,   
   > they often require a awkward parameters to be set regarding the ratio   
   > of various activities such as reading, writing, allocating new memory,   
   > and so on, and if you get these parameters wrong then the garbage   
   > collector loses its timing guarantee.   
      
   Yep. As with any bit of engineering, if you misunderstand the   
   environment, you your designs might not work well.   
      
      
   > A valid reason might be that most real time software is very simple,   
   > and would not benefit from a garbage collector.  That may well be   
   > possible; I don't know.   
      
   I get the impression this was once true but as time goes on the software   
   gets much more complicated. (Just like non-realtime software!)   
   --   
   Kevin Willoughby         kevinwilloughby@acm.orgNoSpam.invalid   
      
   Imagine that, a FROG ON-OFF switch, hardly the work   
   for test pilots. -- Mike Collins   
      
   --- 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