[inforoots] Spacecraft Pioneer and it's data on old media problem

[Neal Laurance]

I feel I walked into the middle of a movie and am trying to figure out 
who the characters are.

Somehow, I never saw the original message or document that inspired this 
extended conversation. Could someone please post is (again?). The 
discussions have been very interesting, but I can't quite get the context.

Thanks,

Neal Laurance


Van Snyder wrote:
> =======================================
>
> Posts to inforoots at computerhistory.org is information known to or the opinions of the poster.  All posts to inforoots at computerhistory.org are archived.  By posting to this list you grant a license for use of this material to the Computer History Museum located in Mountain View, California, USA. 
>
> =======================================
> On Sun, 2007-09-30 at 04:45 -0700, Carl Baltrunas wrote:
>   
>> On the PDP-11s, they had a couple of OS's, RT-11, RSX-11, and RSTS.  
>> I've programmed on all three of these (RT and RSX, in both FORTRAN and
>> Assembly). I'm not sure what they had available on the PDP-8 or PDP-12
>> systems as far as OS's.  And I know they had systems with, 8-bit, 12-
>> bit, 15-bit, and 36-bit word sizes, but do not know what would be so
>> special about the file formats other than ascii or bcd encoding in 6
>> or 7 bit bytes.
>>     
>
> It's probably Fortran unformatted files that cause the problem.  These
> aren't ASCII or other text.  The numbers in them are in the internal
> format of the machine.  Frequently, each record of the file has an
> "envelope" that indicates whether it's all or part of a logical record,
> how long it is, how long the logical record is....  Unformatted files
> are intended to be read by programs compiled by the same compiler that
> compiled the program that wrote them.  There are modern portable ways to
> store unformatted data, such as XDR and HDF, but these might result in
> the same bind in 40 years that now exists w.r.t Fortran unformatted
> files.
>
> Fortran 66, per se, shouldn't be a terrific problem.  The only thing in
> Fortran 66 that was absent in Fortran 77 was Hollerith, and one arcane
> incompatibility having to do with a comma after a P desriptor in a
> FORMAT statment (which will be repaired in 2008).  All of Fortran 77 was
> included in Fortran 90.  The only things deleted in Fortran 95 were real
> and double precision DO variables (which were added in Fortran 77 but
> shouldn't have been), branching to an END IF statement from outside the
> block (an accidental addition in 77), ASSIGN, ASSIGNED GOTO, PAUSE, and
> the H edit descriptor.  Compilers snivel about these, but still compile
> them.  The only thing deleted in Fortran 2003 was the concept of
> "printing," using the first character of a formatted record for vertical
> format control.  This was deleted because it's easy to do with a post
> processor, and a program couldn't find out whether it was supported on a
> particular unit, or request it.
>
> Of course, if the programs use a lot of DEC extensions, you might not
> find them in modern compilers.
>
> How much code are we talking about?
>
>