Timothy Guan-tin Chien

@graingert I wonder what "support NPM directly" means? Your wrapper package will be still very useful in terms of download the specific version of XULRunner/Firefox for SlimerJS. It's not that...

@graingert You are right. I was just trying to figure out the desired outcome. Yet, I failed to see the reason of porting it to JavaScript as a requirement; there...

> Also a port of the python/shell code to JavaScript @graingert I don't understand this part and you never replied my comment above. I would doubt this is a requirement...

> I don't think it should download Firefox, just put a package.json in the > root of this repo It would be an issue for locking down the right Firefox...

@KevinGrandon Upon seeing your patch, I was going to say download the Firefox binary itself should be a package of it's own, e.g. [mozilla-download](https://www.npmjs.com/package/mozilla-download). But it pull from the Mozilla...

Thanks for the quick reply. Is it possible to set multiple codepoints to the same glyph? That feels like a better approach than duplicating the glyphs. It makes sense for...

> We use ICU for handling the codepage conversions (or iconv, but our official ports all use ICU). Because the missing Glyphs are reported in the PUA already it seems...

Fixing this and #25 probably means implementing the properly B-tree data structure...

The current approach can be found in `no-linear-search` branch. The commits removes the linear for loop when search each of the block (i.e. the `_getIncompleteMatchedSingleCodesInBlock` function stated on the title...

This is related to OpenVanilla/McBopomofo#94; we should have getting all the valid sounds on the 0th table but the current dataset does not give us this promise. Maybe not checking...