[Glass] Possible Bug: String>>#= treats nulls as a terminator

[monty via Glass]

My example and thread title were wrong. It skips null *and* various control chars entirely when comparing:
(0 to: 255) select: [:each |
	(String with: $a with: $b) =
		(String with: $a with: each asCharacter with: $b)]

which yields:
anArray( 0, 1, 2, 3, 4, 5, 6, 7, 8, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 127, 128, 129, 130, 131, 132, 134, 135, 136, 137, 138, 139, 140, 141, 142, 143, 144, 145, 146, 147, 148, 149, 150, 151, 152, 153, 154, 155, 156, 157, 158, 159, 173)

The GS Prog Guide (p. 77) says the ICU lib handles string comparisons internally, and it seems to ignore these characters for the sake of normalization.

But that means it's possible for two Strings to be #= while having different #sizes and indexable characters, and that comparisons between Strings containing binary data aren't reliable, and that other String methods aren't consistent with #=:
| one two |
one := String with: $a with: 0 asCharacter with: $b.
two := String with: $a with: $b.
one = two
	and: [(one at: 1 equals: two) not
		and: [(two at: 1 equals: one) not]]

And since GsFile #next and #contents are character based:
(GsFile open: 'bin.one' mode: 'wb' onClient: false)
	nextPutAll: #[100 25 200];
	close.
(GsFile open: 'bin.two' mode: 'wb' onClient: false)
	nextPutAll: #[100 200];
	close.
(GsFile open: 'bin.one' mode: 'rb' onClient: false) contents =
	(GsFile open: 'bin.two' mode: 'rb' onClient: false) contents.

Consider this more as a "heads-up" for users than a bug report, since this is apparently the intended, documented behavior.

> Sent: Friday, January 26, 2018 at 2:20 AM
> From: "monty via Glass" <glass at lists.gemtalksystems.com>
> To: glass at lists.gemtalksystems.com
> Subject: [Glass] Possible Bug: String>>#= treats nulls as a terminator
>
> Is this correct?
> 
> (String with: 12 asCharacter with: 0 asCharacter) =
>     (String with: 12 asCharacter with: 0 asCharacter with: 32 asCharacter)
> 
> Other string methods, like #copyAfter:, don't treat null the same way.
> _______________________________________________
> Glass mailing list
> Glass at lists.gemtalksystems.com
> http://lists.gemtalksystems.com/mailman/listinfo/glass
>