On 2017-04-04, Ben Finney <ben+pyt...@benfinney.id.au> wrote:
> Dave <dbola...@fastmail.fm> writes:
>
>> I don't care for the idea of replacing the data file for every save.
>
> This is the simplest implementation. It works. Can you say why you don't
> care for it?
After the data is read, it requires reloading the ASR33 with a fresh
spool of unpunched paper tape.
>> My preference would to append to the existing data file - makes
>> more sense.
>
> Can you expand on that? You haven't defined how that would satisfy the
> requirement, so it isn't clear to me why a more complex method “makes
> more sense” than the simple way which works now.
A big advantage to reading the old file and generating a complete new
file is that you can read the existing file, create a new temporary
file, and then 'rename(2)' the temporary file to replace the old one.
At least on Unix, rename is an filesystem-atomic operation and
provides a far smaller window of vulerability for r/w race conditions,
crash-induced problems, etc. I assume there's a similar way to do
that under Windows...
--
Grant Edwards grant.b.edwards Yow! Mary Tyler Moore's
at SEVENTH HUSBAND is wearing
gmail.com my DACRON TANK TOP in a
cheap hotel in HONOLULU!
--
https://mail.python.org/mailman/listinfo/python-list
