waitForBlock is incorrect

[jeapostrophe]

waitForBlock does not wait for the given block, but waits for the next block:

https://github.com/algorand/go-algorand/blob/master/daemon/algod/api/server/v2/handlers.go#L328
https://github.com/algorand/go-algorand/blob/master/daemon/algod/api/server/v1/handlers/handlers.go#L629

Even though the end-point is named "wait-for-block-after", the docs say --- https://developer.algorand.org/docs/reference/rest-apis/algod/v2/#get-v2statuswait-for-block-afterround --- "Gets the node status after waiting for the given round." meaning "/v2/status/wait-for-block/after/5" means "Gets the node status after waiting for round 5" (and that's the way the Indexer uses it)

This is easy to fix, just remove the + 1 on those two lines.

[jeapostrophe]

Discovered with devMode --- #2334 (comment) --- at the beginning of July

[brianolson]

There's existing software relying on the current behavior, so the fix will probably be to the documentation.

[jeapostrophe]

The incorrect behavior breaks the Indexer on the DevMode

[jeapostrophe]

(By breaks, I mean that it makes it so the "fast" DevMode is actually slower because every round takes 1 minute.)

[brianolson]

Ah, you mean that 'wait-for-block on devmode is different than wait for block in normal algod operation' ? Wait-for-block in regular mode has been a little confusing sometimes, sorry if I jumped to the wrong interpretation.

[jeapostrophe]

No, it's not different. It's just that on devMode, the next round isn't going to come, so the indexer will ALWAYS wait 1 minute between updates. So if you have an app that does "post txn, look for it on the indexer" then it will take 1 minute per loop, because the indexer uses waitForBlock to get the next one.

[winder]

I think dev mode should be changed to generate a block every 4.5 seconds as usual OR when a tx enters the tx pool. Created #2770 with this feature request.

[gmalouf]

Fixed in #4910