Re: To: Dave, Derek, Dan, Rob and all who are upgrading EU
Gene Mannel wrote:
> But Im concerned that if you go the road of
> letting the interperter decide how a sequence,
> atom, object etc. should be compared in rational
> statements and ETC, you are going to slow down
> EU to a crawl. Im sure that is why Rob made it the
> way it is to help keep the processing speed up.
When I first began coding in Euphoria, that was my assumption as well.
But it's not true. Euphoria *already* knows the types involved in an
operation, so it can easily decide to generate integer/atom/sequence
optimized code. Don't take my word for it - look at the code generated by
the Euphoria to C translator. It generates different code for adding
integers than it does for adding atoms and sequences.
> It seems to me if you try to make EU look and run like,
> Ahem!, Microsoft basic then we might as well go back
> to it. NOT!!!
Why choose to say 'Microsoft' Basic, other to stir up anti-Microsoft
sentiment? And why choose Basic, when there are a host of other well
established languages that have similar functionality? Shouldn't features be
considered for their face value, not on their supposed parentage?
The irony is that, during the short-circuit discussion, the complaint was
that people were "trying to turn Euphoria into C".
> It looks to me like that people who want to write
> fast programs would be willing to sacrifice a moment
> of time to learn to use sequences, atoms and the works!
Programming languages exists for coders, not the other way around.
> There is no way you can make rational statements which
> decide how to act on a variable without slowing down the
> process.
There's no reason to think that adding this functionality would result in
slower code. This isn't just speculation, since Robert has, in previous
posts, noted that Euphoria can and does generate different internal code at
"compile" time, based on the type of the variable.
-- David Cuny
|
Not Categorized, Please Help
|
|
