Re: [PATCH][gdb/python] Add interface to access minimal_symbols

[Tom Tromey <tom at tromey dot com>]

>>>>> "Simon" == Simon Marchi <simon.marchi@ericsson.com> writes:

Simon> I think I have read (probably from Phil or Tom Tromey) that the intention was to
Simon> expose minsyms and full symbols using the same Symbol class, to avoid exposing
Simon> the fact that GDB represents symbols in different ways internally.  I don't know
Simon> if there was some concrete plans for that or if it was just at the idea stage.

I think various ideas were kicked around but nothing was really
committed to.  And, to my knowledge, nobody has really done any work in
this area.

The biggest fear, IMO, was that exposing either symbols or minimal
symbols to Python would lock us in to some undesirable API in the
future.  The basis for this fear is the (common?) belief that gdb's
internal symbol tables are not very good.  I suppose I've been hoping
for years that either somebody would finally clean things up, or that
I'd somehow get around to it.

Nowadays I tend to think we should push forward even with something that
isn't perfect.  While the worst case is pretty bad -- having to delete
some Python API and cause some chaos for users -- I imagine it won't
come to that.  And anyway Python is dynamic enough that one can usually
work around the problems.

Tom