Mathias LANG
Mathias LANG
When querying the API, the response is minimalized and hard to read for human. Certain server/API offer an option to write "pretty" JSON - JSON that is human readable. While...
With MainNet around the corner, we will sometimes need to update sebak all nodes at once. We should make sure that nodes do not loose their pending txs on shutdown.
Currently this test is just 3 nodes sitting idle. We should do a few transactions to make sure there's not something *deeply* wrong.
Currently what is stored in the LevelDB tables is not documented. It makes diving into the code, and sometimes reasoning about it, harder than should be.
At the moment, after a transaction is successfully POSTed, we just echo it back [with the hash](https://github.com/bosnet/sebak/blob/aa4ff033fed5c18a2e36982b8b7f93c3e7553030/lib/node/runner/api/transaction_post.go#L50-L52). This is not actually useful to anyone: the caller can just save the...
When #307 is done, we need a mean to delay a state change. Basically, when a block is externalized, we should have a check for actions that have been delayed...
The current integration test framework has a few limitation: - It uses a polling approach - It requires putting a network up to update - It only test the APIs...
When an operation fails but the signature was legit, e.g. an attempt to overdraft, we should increase the sequence ID, otherwise replay attacks are made possible.
Currently sending a transaction to a node is a fire-and-forget operation. Even if the transaction is very obviously wrong (missing field, wrong signature / hash, etc...) the node just doesn't...
Today @soonkuk approached me with an issue regarding [a WIP commit](https://github.com/soonkuk/sebak/commit/5b1c22ae32c2f581651ef4566fb8af415e1a46c1) The integration tests were failing for no apparent reason. From the code it was obvious this came from [adding...