It is confusing that usage elements are displayed in XML as unadorned hexadecimal

[mhagger]

I must have stared at

<usage_minimum>01</usage_minimum>
<usage_maximum>12</usage_maximum>
<report_count>12</report_count>

a dozen (decimal; i.e, 0x0c) times before realizing, from looking at the C code equivalent, that it is broken. Does it really make sense for usage elements to be represented in hexadecimal in XML? If so, could they be written as 0x12 or 12h to make it clear to human readers that the values are hexadecimal? Or could there be a comment noting that fact?

For that matter, when reading XML, could numbers be self-describing? I.e., could 0x0c, 0Ch, and 12 all be considered equivalent representations of the same number? (Perhaps also 0b00001100.)

(I realize that it would be tricky to do this in a fully backwards-compatible way.)

[spbnick]

It was quite a while a go, but I guess I thought the usage page and usage IDs are hexadecimal in the spec and it would be good to match them, I don't remember what I though about having them unmarked. Perhaps I thought the spec and the schema would be enough. Plus you don't often see usage pages/IDs as numbers.

I don't see a problem making the schema and the code accept the optional 0x prefix or an h suffix, and add it on output, in a new version. I'm not sure accepting several formats would warrant the code complexity. Perhaps just one would be enough, and as a C programmer I would prefer 0x naturally.

[spbnick]

OTOH, the specification format uses the h suffix, so it might be a better match after all.

[spbnick]

And, yeah, I agree it is confusing.