Is this a true statement?

[David C. Ullrich]

On Mon, 25 Jun 2001 15:33:11 GMT, grante at visi.com (Grant Edwards)
wrote:

>In article <3b3745bd.2217895 at nntp.sprynet.com>, David C. Ullrich wrote:
>
>>>>> He's making a very finicky, nitpicking, and frankly silly point [...] 
>>>>
>>>>I don't think the point was entirely silly! It was a valid and
>>>>interesting point (despite being finicky and nitpicking!)
>>>
>>>Yes, it's trivially and obviously valid: a driver is a bit
>>>pattern, and you can write a bit pattern to a file with Python.
>>>
>>>If you show an example of how this is a useful and productive
>>>way to generate a device driver, _then_ it will be interesting.
>>>Until then, it's just somebody chattering to hear themselves
>>>talk.
>>
>>That's a truly remarkable attitude. Nothing is worth knowing
>>unless it solves a practical problem right now.
>
>Not true -- but using a Python program to write into a file a
>bit pattern that is known a-priori to be a device driver for a
>particular OS just isn't interesting (IMO).  If you used a
>Python program to _generate_ the bit pattern from some higher
>level source, that would be interesting.

Right. Now show exactly where I said anything that would
indicate that the second is not what I had in mind.
It was someone else who said sure, you could just copy
a device driver. It was also someone else who said ok,
you could write a device driver with a hex editor as
well - my reply to that said that using Python would
be easier than using a hex editor.

>>I actually learned something about device drivers from all
>>this. Not from something that anyone actually said - I learned
>>something about device drivers by assuming that people _must_
>>be making much more sense than they _appeared_ to be making,
>>and trying to figure out what I could be misconceiving that
>>would make all those comments that must actually be sensible
>>appear like nonsense to me.
>>
>>"Chattering to hear themselves talk" indeed.
>
>I apologize -- I really thought that you knew what a device
>driver was and you were being intentionally obtuse.
>
>>Figured out the answers for myself last night, then this
>>morning the first thing I said was "no, of course in that sense
>>you cannot write device drvers in Python".
>
>Not using currently available tools.  I do maintain that it
>wouldn't be terribly difficult to do, but for performance and
>resource reasons it would be largely an academic exercise.

The original topic was "True or False: Python can do
anything C can do, just slower." How can the answer
to _that_ question be anything _but_ purely academic?
 
>Grant Edwards                   grante             Yow!  I put aside my
>                                  at               copy of "BOWLING WORLD"
>                               visi.com            and think about GUN
>                                                   CONTROLlegislation...



David C. Ullrich
*********************
"Sometimes you can have access violations all the 
time and the program still works." (Michael Caracena, 
comp.lang.pascal.delphi.misc 5/1/01)