Hacker News Reader

Comment by bsaul | Hacker News Reader

I've built some client / server syncing code for a while, and never wondered to check for papers on that subject ( since i did it in a wqy specific to my model each time). What those guys are doing seem quite generic. Are there any famous algorithms for multi end-point data syncing ?

I suppose it's related to the ditributed database research field...

View on HN

The last 1/3 of my talk at last year's Strange Loop was on the data structures used in the Couch world for multi master sync. http://www.infoq.com/presentations/sync-mobile-data

Also google CRDT

Paxos is a good place to start, as it's well-regarded and fairly accessible.

http://en.wikipedia.org/wiki/Paxos_%28computer_science%29

Thanks, i already knew about this one. I was looking to something more specific to single server / multiple clients that need to sync. It seems a bit different to the type of algorithm used when trying to reach a concensus between multiple servers upon a read request ( or is it ?)

Edit : when i said related to ditributed database, i related more to things like log shipping.

Most real-time frameworks like Wave and Derby use Operational Transformation. http://www.waveprotocol.org/whitepapers/operational-transfor... http://blog.derbyjs.com/2013/06/04/derby-v0-dot-5-0/

(Edit: Apparently Meteor does not use OT; yet another reason it's behind Derby.)

What are the other reasons it's behind Derby?

Server-side rendering, npm support, etc.

Reply on news.ycombinator.com