Closed frank-ang closed 1 year ago
We did not plan to support devnet. Other than the epoch issue you pointed out, there is also lack of devnet explorer making it nearly impossible to track deal status in devnet.
@kernelogic
The best thing we can do is probably have a --manual-start-epoch
but we won't be able to track the status of the deal
Disagree. I believe devnet support aligns with sw engr best practices. Devnet support will with high confidence, improve software quality, by unblocking test automation, CI integration tests, CI load tests, CI fuzz tests, etc, to catch classes of errors in devnet before releasing for mainnet clients.
A complete support for devnet is not trivial and the return is not clear. Most users of singularity are using mainnet where we get all the feedback to improve the software quality. The main difference between mainnet and devnet here is, mainnet uses real money but either data preparation or deal replication via this software does not cost money so the users could use mainnet for their test environment.
...
--manual-start-epoch
but we won't be able to track the status of the deal
This is an acceptable partial workaround for me.
Describe the bug Invalid deal end epoch when replication in Devnet
Version 2.0.0
To Reproduce Steps to reproduce the behavior:
Singularity deal rejected.
Note
--start-epoch=2351097
is referencing Mainnet block height. However devnet was 1691 at the time.Devnet miner error.
Lotus devnet is only at epoch 2803.
Expected behavior Devnet deal epoch should reference actual epoch height, including devnet 4k / calib height.
Probable root cause https://github.com/tech-greedy/singularity/blob/3f8c893ad45df7e0eab06d98a377d9280ec8df0a/src/common/ChainHeight.ts#L23