I've found a terrible hack to get around this bug, but I'm pretty tempted to remove the "/flag" syntax and keep only "-this" and "--that" POSIX-ish convention.

Ramack, I would like to know what do you feel about it.

Windows is progressively becoming "more POSIXish", so I suspect that asking Windows developer to write pretty -zen foo.e instead of pretty /zen foo.e is not such a dramatic requirement.

```
$ se pretty /absolute/path/file.e
**** Fatal Error: pretty: unknown flag: "/absolute/path/file.e"


------
```

Because COMMAND_LINE_TOOLS.is_some_flag() is called and returns true for any argument beginning with a slash.

That's really counter-intuitive. Is there some good reason for this behavior?

I do remember command line flags beginning with a slash from MS-DOS, but on unix-like systems programs accepting filesystem paths should not confuse absolute paths with flags. Or, if they for some reason really need to, they should at least respect the -- convention ( "--" meaning end of flags, beginning of arguments, regardless their possible flag-like appearance).