Hi Daniel,
what exactly _are_ file1 and file2? You are calling them files,
but are you sure they're files? As I see it, the are not. Lessee
what happens here ...
REBOL/Core 2.3.0.4.2 24-Jun-2000
Copyright 2000 REBOL Technologies. All rights reserved.
>> difference file1 file2
** Script Error: file1 has no value.
** Where: difference file1 file2
>> difference %file1 %file2
== "12"
>> difference "file1" "file2"
== "12"
<Doing some magic>
>> difference file1 file2
== {:1235!"%?RUGHCFT89}
So, 'file1 'file2 are normally just words, that can't be
used for difference, %file1 %file2 work just like strings,
as expected, cause %file1 is a series.
And now, about my "magic", it was just:
>> file1: {:12374576$!"-&)(/%&)%%%} file2: {)?&/$&()RUGHCFTR(/87694}
And this may be, what you've done:
file1: read %somefile
file2: read %some_other_file
And then difference worked on the _contents_ of the files,
which were read as series, and referenced as 'file1 / 2
regards,
Ingo
Once upon a time [EMAIL PROTECTED] spoketh thus:
>
> What I'm saying is that it is not consistent with the philosophy in
> Rebol. Datasets are not a datatype, it is a human convention. Take
> note of the help message : arg must be a series. Files are read as
> series. But yet difference won't work on them.
> Inconsistent.
>
Quoted from an earlier mail:
> Difference is a high-level function that returns the differences
> between two datasets. But datasets do not exist in Rebol
> (they're not a datatype), so difference accepts any series including
> strings. However, when you use it on the content of two files (read
> %file), you get a strange result :
> >> difference file1 file2
> == {:"39%.@{}!(^^4567)~}