I'm experiencing a weird issue with SEA certify. If I provide the pub of 1 user it works fine (the user is able to write to the authority user's graph). If I provide more than 1 pub, then it does not work at all. Is this a bug with certify or am I using it wrong?
This works:
const cert = await SEA.certify(
['H5LBBxrb0wFukxlI9YAFif9M-LVE6JFtRwvMZCTfe54.CzOB4zTpzR3OPa7Ncf6pfOVmyXiAJp3-XohtcBgvwBw'], // 1 pub provided
{"*": "messages", '+': '*' },
user.is.alias,
null
);
This doesn't:
const cert = await SEA.certify(
['H5LBBxrb0wFukxlI9YAFif9M-LVE6JFtRwvMZCTfe54.CzOB4zTpzR3OPa7Ncf6pfOVmyXiAJp3-XohtcBgvwBw', 'eBv261G2jBJOQzAUEb-jn6XISvRQq3rbPgLaatrC9qQ.ZUXChPYESeYFiGizTuppffBETXaBnOTAYjValT1qpsg'], // 2 pubs provided
{"*": "messages", '+': '*' },
user.is.alias,
null
);
Gun({axe: false
fix this bug, if not, more details? If in browser, check networking tab websocket messages?.put(
when usually you use .set(
for .map()
over. Does this help at all?gunFetch
posted :) ? I didn't see!user.pair()
and see if fixed now. It was always a bug if alias
had something else on it.
Following up on the "performance issues" that I was reporting from before--the issue was that I was maxing out gun performance. I had done map().on() for ~50-60 objects and 3 - 4 additional on()s for each of those as well. gunjs seems to fire quite a lot for map().on() continually (honestly I haven't seen these calls end in my testing) so all of these on()s ended up clogging the system so that it would take 30 - 90 seconds to have a usable interface.
I refactored the whole system to only rely on the one map().on() for data and now things load an order of magnitude faster
// add listener to foo
gun.get('foo').on(callback, true)
// remove listener to foo
gun.get('foo').off()