[erlang-questions] dependency on inets for string functions

[Bob Ippolito]

On 3/19/07, Kenneth Lundin <kenneth.lundin@REDACTED> wrote:
> On 3/19/07, Vlad Dumitrescu <vladdu55@REDACTED> wrote:
> > > > What's the status of bit-level binaries and general comprehensions? Is
> > > > that coming in R11B-4 too? That would REALLY be awesome.
> > >
> > > Their status is that these are present in R11B-3 already !
> >
> > Hi,
> >
> > Maybe it would be useful to know that in order to be able to use
> > those, the module has to be compiled with
> > -compile([binary_comprehension, bitlevel_binaries]).
> >
> > I didn't know :-)
>
> The complete support for binary comprehensions and bitlevel binaries will
> not be released in R11B-4.
> With complete I mean e.g. as default in the compiler and more.
> There are still some compatibility issues that we have to think of
> very carefully
> before we turn the support for bitlevel binaries on as default.

Are there any known caveats to using the current R11B-3
implementation, or is it just a matter of the shell not yet supporting
them?

Would it be possible to support both old and new binary syntax in the
same source file? Is it possible to use preprocessor macros to use one
function instead of another? Are there predefined macros that state
what the compiler's capabilities are?

I ask because it would be convenient to keep compatibility with R11B-2
for a little while longer, but I'd like to migrate to using bit level
binaries for clarity and performance reasons over the next release
cycle or two.

-bob