Guy Smith <[log in to unmask]> wrote:
>
>Larry Boyd ([log in to unmask]) wrote:
>
>> I have an additionally suggestion.  What if HP selected a filecode that
>> would define command files?  If the file is ASCII it would still work, and
>> if the file is binary, the code could be checked.
>>
>> The primary reason for the suggestion is that I've always thought that
>> command files and UDCs should have filecodes.  It makes it very easy to
>> look at files with listf or MPEX and "see" if they are UDCs or command
>> files.
>
>I wish to respectfully disagree.  Being the type who never remembers to
>do a chmod on my UNIX scripts, and who then spends several seconds in
>dememtia figuring out why the script won't run, I know I'd make the same
>omission with file codes.
>
>However, if HP defined a file code strictly for binary command files and
>added the appropriate if-logic (i.e., allow any file code for ASCII and
 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
>a specific file code for binary), I would not complain.
 
Oh, Guy, I agree completely!  It must definitely still allow the ASCII
file to work (See 3rd sentence above).  I surely would *not* want to change
*every* one I have or create to make sure it works.  What I was suggesting
was a way to allow for binary files, and to allow for better control of the
files.  I don't think it would be a good idea to *enforce* a filecode.
 
John Dunlop ([log in to unmask]) wrote:
 
>Call me old fashioned, but I cannot see why anyone would want to use a binary
>command file.
> Can anyone take half a minute to enlighten me ? I also agree with Larry Boyd
>that a dedicated
> filecode for these types of files would be very desirable.
 
If you do a lot of file transfers over dialup lines, I have found that you
typically get faster transfers (mainly because of compression) with binary
files.  I have tried binary transfers on ASCII files, and didn't always get
the results I expected.  On the other hand, it doesn't appear to be a problem
with a BINARY file, even if it's only ASCII data.  Since, I have no idea
what goes on at all in these highspeed compressed transfers, I have no idea
why things worked the way they did.  (BTW, before someone tries to explain
it to me, I prefer the "magic" answer, as with disk drives;>)
 
Stan asked about QEDIT files.  Unfornately, these will definitely not work
with the binary check, even if you QEDIFY the CI.  I would suppose that the
type of file would be checked (which Robelle would return as BINARY), and
the CI, while it would now be able to read the file, would say, "Sorry, this
is a binary file."
 
Thanks,
 
Larry Boyd              [log in to unmask]