Job van der Zwan

> As far as releasing is concerned, I would be a bit more cautious than you here. People have gazillions of data snippets compressed in browsers all over the world,...

> since the dictionary entries just before splitting symbol are never used there is no need to add them to the dictionary. So, say that I have: AAA,AAA,AAA,AAA,AAA,AAA,AAA,AAA,AAA Are you...

Ah right, it's been a while so I forgot about the trick :) Also, if we're breaking anyway, we could prevent duplicate entries being added to the dictionary with an...

> Why your code is written in es3? This is a really old library, still being used in really ancient contexts. We'd like to keep that backwards compatibility working, which...

We really appreciate the effort! I'm sure it's OK (not enough TS experience to judge) and in the worst case you'll have created a good starting point! :) (reminds me:...

I just updated the tests with a dictionary decode as well, [since that is what you use elsewhere too](https://github.com/pieroxy/lz-string/blob/master/libs/lz-string.js#[L14-L26]). It's almost 2.5x as fast on my desktop browsers, and in...

@pieroxy Sure, I'll make a branch and try. (Aside, anyone else annoyed by how both jsperf and jsbench repaint the DOM while doing pure-JS benchmarks? I had it lag on...

Ok, before I do this I have some more testing to do. You see, it occurred to me that we can also try pairs of characters. In that case, both...

At home on sick leave, finally have some time to look into this again. And it's such a small change too... Hope it actually improves performance a bit! I'm not...

I think @ipekarik is trying to use LZString from within [Google App Script](https://developers.google.com/apps-script/)? I can reproduce it. [Here is a link to my example AppScript](https://script.google.com/d/1v28c_hkgiP2T-_NSZ3q7t-EPfqpS89qwc_WxkM6WVqBAmxALoq6Gx3VH/edit?usp=sharing). Note that I have both...