[Info-vax] 1 year.

[johnwallace4 at yahoo.co.uk]

On Saturday, 8 August 2015 20:54:37 UTC+1, already... at yahoo.com  wrote:
> On Saturday, August 8, 2015 at 10:34:59 PM UTC+3, Stephen Hoffman wrote:
> > On 2015-08-08 18:47:08 +0000, already5chosen at yahoo.com said:
> > 
> > > On Saturday, August 8, 2015 at 6:11:55 PM UTC+3, Stephen Hoffman wrote:
> > >> On 2015-08-07 14:23:35 +0000, already5chosen at yahoo.com said:
> > >> 
> > >>> Something similar to IA-32 Execution Layer would be an ideal solution 
> > >>> for your case.
> > >> 
> > >> Until somebody sorts out the revenues and particularly the profits, 
> > >> discussions of developing new technologies -- particularly given  
> > >> DECmigrate has been around for many years -- or plans to port to a  
> > >> platform that does not yet exist... seems moot.
> > > 
> > > I meant Itanium host, not x86. So, platform exists.
> > 
> > Here's a more detailed description of the IA-32 Execution Layer that 
> > already5chosen is referring to:
> > <http://www.microarch.org/micro36/html/pdf/goldenberg-IA32ExecutionLayer.pdf>
> > 
> > Which in aggregate, looks rather like what the "DECmigrate" application 
> > migration tools can already provide:
> > <http://h71000.www7.hp.com/openvms/products/omsva/omsais.html>
> > 
> 
> Yes, more or less. Except that AEST appears to be a static translator, so performance and coverage are unlikely to be as good as dynamic triple-phase binary translation of IA32EL. In particular, I don't understand how AEST will handle applications that generate executable code on the fly.
> 
> > Quick DECmigrate overview:
> > <http://labs.hoffmanlabs.com/node/641>
> > 
> > 
> 
> Sounds like you are not particularly fond of AEST. Is it because of bad 1st-hand experience, or just out of principles?
> 
> Anyway, the situation, described by IanD, asks for "last resort" solutions.
> If I was him? I'd definitely try AEST.
> 
> > 
> > 
> > -- 
> > Pure Personal Opinion | HoffmanLabs LLC

If you want an example of what DEC historically had done with on-the-fly
binary translation go read more about FX!32, which was a rather handy
tool for running Win32/x86 applications on an NT/Alpha box, by a mixture
of static and dynamic translation. It's actually mentioned in the IA32EL
paper, though not in much detail. Wikipedia has an FX!32 article, and
more importantly, has references.