[FoRK] What do you think of my http://techcrunch.com/2013/07/24/google-chromecast/

Stephen Williams sdw at lig.net
Thu Jul 25 11:33:03 PDT 2013

On 7/25/13 9:57 AM, Ken Meltsner wrote:
> On Thu, Jul 25, 2013 at 11:37 AM, Stephen Williams <sdw at lig.net> wrote:
>> On 7/24/13 11:05 PM, Joseph S. Barrera III wrote:
>>> On 7/24/2013 10:54 PM, Ken Meltsner wrote:
>>>> didn't miss the boat.  Everything was a byte-stream file so you didn't
>>>> need
>>>> to force extensions.
>>> Well then, they should have enforced a app mapping based on the first
>>> four bytes. Shebang was a weak implementation of that.
>>> Not casting blame -- hindsight is very cheap. But if only...
>> I used VaxOS, some embedded DEC OS (RSTos etc.?), CPM and MPM8/16, DOS,
>> Win1, and Version 7 Unix all at once.  (Nobody from that era seems to exist
>> anymore, or they are hiding from the Internet.  It is very strange.  It was
>> only 29 years ago.)  Then a short time later SysV5.2.  The VAX was awful in
>> just about every way.  Super slow to start a process, terrible scripting
>> language (terrible subshell support, etc.), the file
>> format/typing/conversion thing was a pain all the time, etc.  The verbose
>> command arguments were incomplete, not well thought out, and a terrible
>> pain.
>> Unix was immediately recognizable as a consistent, simple, powerful, fast,
>> fun environment.  There was no comparison.  Everything was instant.  You
>> could figure things out by experimenting, looking around, and reading a few
>> online man pages.  Pretty much the definition of an ideal environment, even
>> then.
> Roughly that vintage myself.  VMS, Ultrix, Tops-20, CP/M.  All of the cool
> kids at MIT wanted to use LispMs, though -- VMS was for the folks down in
> Mechanical Engineering who wanted to compute thing, and no one wanted toys
> like CP/M.
> And there were random rumors about NU machines -- Network Units? -- that
> were being cloned by some intrepid engineers at Stanford U (Stanford Univ.
> NU or Sun?).  And further afield, so not worthy of attention were the dandy
> Perq workstations from a CMU spinoff.  But LispMs were the *only* systems
> for serious CS types.
> Amazing how much of the microcomputer revolution we managed to miss at MIT.

That my friend is the curse of too much money.  Innovation often happens at the edges, back country, and highly constrained 
fringes. The fat and esoteric solutions don't have the immune system to compete.  But it is not about the size or resources of 
the companies or organizations, but what their outlook, strategy, and culture is. It is a smart company that harnesses that and 
the dumb ones that are too snobbish / NIH to embrace it.

> Ken


More information about the FoRK mailing list