We've also started off a new Telegram group. Feel free to join and share with folks! We're going to try to keep the Gitter focused on development and algorithms, so the telegraph might be a good venue for more general discussion
Is anyone else getting the error message "one of the params is missing or invalid" when trying to join the Telegram?
Yes, me 2.
Hmm, not sure what's up. New to Telegram myself :). Will figure this out and update here
@rbharath just got through the paper -- great read! How are you guys planning to address dilution of token economies vulnerability. I'd imagine that most nested (esp. deeply nested) substructures would have pretty small economies making cost of attack pretty low.
@daanishkhazi Good question. This is definitely a hard challenge. We have a couple preliminary thoughts we're exploring right now. Will address in a future version of the paper once we have more clarity
chung chan hai
hello, do the project issue token? how can I find infor about token metric?
@NhungDoan10_twitter This chatroom is for developer and research discussions. We don't have any information about tokens at the present time
@rbharath Hello! How can we discuss the possibilities of a strategic partnership between you and the fund that I represent - Gt Blockchain Investments?
@3loyWiking This chatroom is for developer and research discussions. We currently don't have any information about tokens at this time. In the future, we will make available more information for investors
@rbharath, I've got a quick question about section 4.6 in your paper. More specifically, it wasn't immediately obvious to me how, given a tokenized dataset TD, I would be disincentivized to take the data and sell at a lower price. Is this what you meant when you wrote about "Leakage Resistance"?
@athappa Yes, that's correct. Leakage resistance measures the intrinsic resistance a dataset has to being leaked. For example, some datasets might be important to buy first hand, so would be very leakage resistant. Others are much less so and counterfeiting (copying and selling at a lower price) is a serious problem. We're exploring ways to deal with counterfeiting right now. Some type of digital watermarking scheme is looking interesting
Got it. To your point, the way I was thinking about it is the more staleness of the data matters, the less likely it is to be leaked. But it seems unlikely that there’s some kind of a priori test for whether a dataset is leakage resistant or not.
Staleness is a good way of thinking about it. Time series datasets are likely to be highly leakage resistant; leaking last week's stock prices doesn't matter that much when people are looking for a feed of future prices
Genomic/medical data might be quite leakage resistant due to legal constructions. If one dataset is HIPAA compliant (for those in the US), and is run by the NIH, most purchasers would be strongly disincentivized against buying from an alternative source due to potential legal issues
Hey just read about blog posts, I have few questions and suggestions.
When can I expect to read about the white or yellow paper?
Also most of the tech stack for the computable protocol is written on top of ethereum(?) Using JS. Any particular reason for going with JS instead of Golang other than the fact it's easy to prototype with JS and easy to find devs related to it.
You guys are doing amazing work.kudos and keep it going. I got to know about computable labs via bharath on Twitter I mostly do deep learning with NLP. So I'm only good with python. Nothing with JS/Golang/protocol research.