Re: [PATCH] bindat strz fixes

[Richard Hansen]

On 6/1/22 16:23, Richard Hansen wrote:
> Anyone that needs the current behavior should be using `str N`, not 
> `strz N`. If they're using `strz N`, then I would consider that to be 
> a bug in their code. If this change breaks someone, they can fix it 
> easily: just change `strz N` to `str N`. I understand that we should 
> endeavor to maintain compatibility, but keeping the current behavior 
> would be intentionally preserving a bug to accommodate other bugs. I 
> don't think that's a good trade-off.

Actually I'm wrong; there is a difference between `str N` and `strz N`. With 
`str N`, all zero bytes in the input string are included in the unpacked value 
(including trailing null bytes). With `strz N`, the unpacked string ends just 
before the first zero byte.

I'll remove the `strz N` change from the patch series so that we can get the 
rest merged.