And yet, I don't automatically have it at my fingertips just by virtue of having the software — which was my point in the first place.
Barring licensing requirements to the contrary, software is usually distributed to end users in executable format, and only in source form if specifically desired. When an executable format is binary or bytecode, that's not human-readable. Sometimes, such passing around of software ends up happening in a manner in which the source is not "trivially accessible".
I like the fact that, if I grab a Perl 5.x program just to use it without any specific desire to have the source, then three years later think I'd like to see the source code, I can. I don't have to go get the "trivially accessible" source, whether it's actually accessible in this case or not, because the executable code is the source. It's really convenient that way.
I'm not saying there shouldn't be capability for bytecode compilation, by the way. I just worry that the trend will be to distribute in bytecode form all the time, rather than continuing to distribute in executable human-readable source code form, which is (to me) the preferred format — mostly due to potential accessibility issues.
It's even possible to simply lose the source for something compiled to a persistent, non-human-readable form. That would be awfully inconvenient.
|
- apotheon
CopyWrite Chad Perrin |
In reply to Re^9: Perl 6... do we need another vm?
by apotheon
in thread Perl 6... do we need another vm?
by jfroebe
| For: | Use: | ||
| & | & | ||
| < | < | ||
| > | > | ||
| [ | [ | ||
| ] | ] |