After a short diversion into other projects, I've resumed work on OGL. Here's a bit of an overview of what I have. Again I changed the orthography a bit so I won't have to mess with diacritics or special characters as much as this is primarily a written language for data storage purposes. I'm still thinking about making <š> into <x>, and it's looking like I may not need the phoneme /ʒ/ at all. I'm also looking to see if /ɨ/ can be easily removed, then I can reassign <y> to /ə/. With those changes the only non-ASCII characters will be edh and thorn. They too could possibly be eliminated by merging with /d/ and /t/. < a b c d ð e ə f g h i j k l m n o p q r s š t þ u v w y z ž > / a b ʧ d ð e ə f g h i j k l m n o p ʤ ɹ s ʃ t θ u v w ɨ z ʒ / I'm going to stick with the idea of using English mnemonics to label roots but the CVC and CCV shapes weren't working out too well so I now have all four-letter roots in the forms CCVC and CVCC, except for a few V words used to swap argument ordering. I'm still not sure I'll need the markers to change arguments but I'll leave them for now. As a result of the new four letter system, I've had to stretch out some words to make them fit (ex: þrij=three, bluw=blue) and there are cases where I've added an <h> (which has  as an allophone) to some (tuwh=two, ðyhs=this). So far the lexicon is made up of the English mnemonics attached to Lojban gismu but I've purged out a few unnecessary gismu, mainly those having to do with nationalities and such. In a few cases the gismu will be compounds in OGL like "huwm'fehm" (human-female) for "woman/girl" and "huwm'mejl" (human-male) for "man/boy".