Igor van den Hoven

Fixing it like that can result in screen flicker when a lot of data is received. I updated the beta with another fix that checks for FLAG_READMUD in split.c.

Thanks for checking it out. :) As for bit operations, I decided not to use them because the performance gain was minimal or nonexistent while readability suffered. Good call about...

There's probably not much meat left on the bone, but I'm open to suggestions. I'll try to run some benchmarks in the next few days to check if a similar...

The cmp() call is needed however for string comparison support and qsort() compatibility.

Quadsort changed quite a bit since this discussion. Anyone interested in micro optimizations might want to have a look at: https://github.com/scandum/tinysort

Thanks, I'll incorporate this up stream.

What would be the big5 equivalent?

So I can't really come up with a good way to fix this, assuming it needs fixing, suggestions are welcome.

Not sure if I understand what you mean. When you get ``` utf8_to_big5: did not find unicode index '0xfffd' ``` You can use: `#showme \ufffd` to print the unicode character.

You should be able to substitute the message: ``` #substitute {^utf8_to_big5: did not find unicode index '0x%1'} {utf8_to_big5: did not find unicode character '\u%1' with index '0x%1'} ``` And you...