Something that has just puzzled me: which words, if any, that are now standard abbreviations in texting will one day be accepted by the OED?  And what form of evidence do you think they would use?  A printed email of the text?

Some of the more or less standard text conventions have been features of chatspeak for something like 20 years and logs of that stuff are abundant. But you’ve raised a very interesting point. I’ve noticed that some people compress their text more than others which I suspect is at least partly due to how you are billed. My provider charges 5 cents per message, so as long as I keep it under 160 characters, its al th same & I dnt sqez my txt much. A couple of chatspeak words that have become part of my regular vocabulary, if you will, are nite and prolly, although I’m sure nite has been around a long time.

Yes, I would imagine the more standard texting abbreviations will appear in the OED. The reference sources would most likely be linguistic articles and books about texting and internet language. Although, I believe that printed copies of text messages would be considered acceptable. The key is that it can be verified.

I’d have thought textspeak would become increasingly rare, and eventually vanish altogether, with the rise of predictive text, though there are still people who haven’t got the hang of it. At least, I hope so.

Predictive text on phones has been around for a long time now, though. It seems to me that it’s more likely that text speak is being propagated and popularised more by messageboards - though not this one, of course - and chat rooms, rather than phones, these days. 
My godson, who is 17, tells me that there are now mobile phones that have predictive text speak on them. Dearie dearie me. I hoped he was having me on, but it wouldn’t surprise me if he wasn’t.

Wasn’t there some sort of forereunner to txt spk?  ISTR seeing adverts for courses to teach you Speedwriting or whatever it was called in the Tube in the eighties, which started “If u cn rd ths msg...”