[Info-vax] RMS internals?

[Arne Vajhøj]

Richard B. Gilbert wrote:
> Arne Vajhøj wrote:
>> Bob Koehler wrote:
>>> In article <howard-E4439F.13574308082009 at news.newsguy.com>, Howard S 
>>> Shubs <howard at shubs.net> writes:
>>>> In article <0076b30e$0$7155$c3e8da3 at news.astraweb.com>,
>>>>  JF Mezei <jfmezei.spamnot at vaxination.ca> wrote:
>>>>> If a VMS native file didn't have CRLF at the end, but Windows added
>>>>> them, then it is likely the original VMS->Windows transfer mucked 
>>>>> the file.
>>>> Original was Windows->VMS, as the file was created on Windows.
>>>>
>>>    Sounds like it simply should have been transfered as a text file.
>>>    Are you trying to compensate for all the mistakes the user could
>>>    make?  It isn't really an RMS function to fix those, you're kind of
>>>    rolling on your own.
>>
>> He is trying to solve the problem that most non-VMS files has valid
>> text files with rec len > 32K, but VMS does not.
> 
> Just because other systems allow records greater than 32K does not mean 
> that files have to be written with records that big.

No. But a lot of apps does. And if those files need to be moved onto a
VMS system, then we have the problem.

> The only format that is close to being universal is fixed length 80 byte 
> records.  Many systems can block and/or deblock those fixed length records.
> 
> Either the tapes must be written with record lengths < 32K or they must 
> be read on another system!!

HTTP or FTP is more likely than tape these days.

Arne