Re: Future architectures


In article <r1c6o5-v6a.ln1@xxxxxxxxxxxxxxxx>,
Bernd Paysan <bernd.paysan@xxxxxx> writes:
|> Joe Pfeiffer wrote:
|> > Comes from the Harvard Mark I, which used a paper tape for instructions
|> > and some sort of electromechanical storage for data.
|>
|> That goes back to the first Zuse machine. The actual advance was von
|> Neumann's, because he found that instructions are just another form of
|> data. However, many computer languages haven't caught up to that idea, and
|> still only manipulate data, but don't provide meta-programming features.

Credit should be given to him for the theoretical advance, but the
first people to actually implement it were at Manchester and Cambridge
(England).

Also, it isn't true that programming languages haven't caught up with
the idea - it has been deliberately excluded from most of them, for
very good reasons. In particular, they have decided that security,
debuggability and efficiency are more important than metaprogramming.
You may disagree with that decision, but that is another matter.

I know of no language that allows general metaprogramming without
a serious degradation of all of those aspects. In theory, the
security aspects are soluble (because the capability systems solved
it), but I know of no language that delivers on that. The other
aspects are probably insoluble, even in theory.


Regards,
Nick Maclaren.
.

Relevant Pages

  • Re: Public disclosure of discovered vulnerabilities
    ... What about your claim that no one knew anything about security before ... Even now that we have higher-level languages like C, ... Why would that be a disaster? ... For every BIND buffer overflow ...
    (sci.crypt)
  • RE: [Full-Disclosure] automated vulnerability testing
    ... If you are truly interested in security, you won't use C as the programming ... It is security unfriendly by design (so it can get closer to the ... Most other languages prevent these as part of their semantics or syntax. ...
    (Full-Disclosure)
  • Re: [Full-Disclosure] DCOM RPC exploit (dcom.c)
    ... Information Security Engineer ... > IMHO it is TIME to sue corporations like microsoft for their stupidity ... they're sending all their programmers ... interpreted languages, meaning .net. ...
    (Full-Disclosure)
  • Re: VB6 versus VBNet Challenge
    ... that the runtime can't verify all the actions, so it disables bounds ... the fact that you can't run unsafe code in certain security contexts. ... Execution, is ... .NET supports a lot of those earlier languages. ...
    (microsoft.public.vb.general.discussion)
  • Re: To RISC or not to RISC
    ... that immediately fell out of favor on the x86 architecture. ... instructions like LOOP and most of the string instructions. ... isn't that useful for C-like languages). ... you *can* allocate a different amount of local space ...
    (alt.lang.asm)