signed/unsigned integer conversion for right shift seems

[Peter T. Breuer]

"Also sprach Jonathan Wakely:"
> 
> On 6 February 2018 at 18:37, Tadeus Prastowo wrote:
> > @Peter Breuer: it seems that you are not very knowledgeable about the
> > internals of the C language.  Hence, I would like to invite you to

Amusing.  Byt as the infamous author of a much used higher order
compiler compiler in C, a linear logic model checker for C, about 4
different user-driven languages that I can recall written in C, ranging
from a persistent higher order lazy functional language to a decompiler
compiler, and a mathematical logician well known as a programming
languages semanticist who has given denotational, operational and
logical semantics to many programming, machine and specification
languages, etc etc etc, "I don't think so".

Try sf.net/p/obfusc for the current snapshots of the encrypting,
ofuscating C compiler.

> > study this very good resource on the internals of the C language:
> > http://publications.gbdirect.co.uk/c_book/.
> 
> And maybe find somewhere else to discuss it. "I don't understand the C
> standard" is not a GCC problem, so doesn't belong on this mailing

Please desist from ad hominen attacks.

If you have a problem with my reasoning in any particlar instance,
please state it with specificity. 

 From the lack of valid argument, I guess the problem seems to be that
the spec has boobooed in this instance.  I would be grateful if you do
manage to find somewhere in the spec that allows the conversion thang
not to be done for >>.  

I now don't know for sure what needs to be done here.

Regards

PTB