Am 19.09.2018 um 01:37 schrieb Kenneth:
> clipka <ano### [at] anonymousorg> wrote:
> 
>>
>> For such a complex scene - with self-modifying code and shit - the
>> actual error scenario turned out to be almost disappointingly simple:
>>
>>     #macro Foo()
>>     #end
>>
>>     Foo()
>>          ^---- END OF FILE
>>
>> So invoking a macro immediately at the end of a file is what breaks stuff.
>>
> 
> Wow, that's weird. (the only useful comment I can make!)

Nah, not too weird - the code that switches between input files just
detects immediately upon returning that it can't read any bytes from the
file, and thinks this is a problem.

Well, not anymore it does. Fixed in the `refactor/tokenizer` branch, and
plan to do a tag and build a new Windows release soon-ish.

Post a reply to this message