3 ms·
Claiming this is dense, OK. That's just math. But how can you claim this is "speedy" when you have an unpolished Python implementation in hand? You need to benc
by jerf 16y ago
Claiming this is dense, OK. That's just math. But how can you claim this is "speedy" when you have an unpolished Python implementation in hand? You need to bench against best-of-breed Base64 implementations, not some pure-Python comparison you may write up in the future, and for that you're going to need to go at least C, and I wouldn't be surprised down to the assembler. Just stick with "dense" and claim "speedy" when you've got some numbers.
(I suspect you won't beat or even tie Base64, but I'm prepared to accept that you only got within X% of base64 but that X% is worth it in some scenario. And maybe if you get clever enough you can prove me wrong.)
- DenisM 16y agoIt's speedy compared to base62. See, the problem is that Base64 is linear is speed, but has large (non-alphanumeric) charset, while base62 has small (alphanumeric) charset but is quadratic in speed (at least in its straightforward implementation, the only one I was able to find). Maze62 is the encoding that is both constrained in charset and liner in speed. Hence, the title. Does it make sense?
- jerf 16y agoOK, that's fair. Thanks.