You are viewing a single comment's thread from:
RE: SteemWorld ~ Newsletter #4 / March, 2018
Thanks! Most of the data comes directly from the steem nodes via the 'Steem JS API'. When my server is ready I will change some requests to work with my own webservice so that the other nodes are no longer being required and I'm not so much dependent on external data sources.
Thank you for your great work.
Another question, does that mean you are trying to set up a steem full node and create a webservice based on that? Based on my experience, a full node need at least 200GB memory.
Wonder how you can implement that. I also have a similar requirement for that, still don't have a good way to achieve that.
A full node requires way too much RAM in my eyes. I'm building an own solution with PHP and SQLite on SSD. Important to know is that you can't use all the API functions with a simple node but it is still possible to receive blocks.
I'm going through all these blocks and put the relevant data in indexed form into my own SQLite DBs. From there it is available for super fast access by the requests from the clients. It has been a lot of work in the past weeks for me but it works really fast and enables me to do much more than the default node APIs provides to us.
Edit:
An example of what I can show only by parsing the whole chain is the real post count. The field 'post count' on SteemWorld will display the real number of posts and comments (separated). That's even not being shown on Steemit because it is not possible with a full node.
Thanks for your full explanation. It sounds similar to https://github.com/steemit/sbds
Yes, that goes into that direction. I've tried to get sbds to run but no chance... :) I now prefer to stick to my own service so that I don't loose much time on finding out how external developments work.