Rename reader to follower

[EncodePanda]

Fixes #2833

[coot]

The changes one the network side looks good to me; I wonder if the consensus report needs to be update as well.

[mrBliss]

As discussed in the Meet, the alignment still needs to be updated 🙂

[mrBliss]

I wonder if the consensus report needs to be update as well.

Apparently, the current version doesn't contain any references to readers 🙂. I have actually written the section on readers in the last week, it will appear shortly in #2853 (and I'll rename them to followers first).

[nfrisby]

Just sharing an observation. Alignment breaks like this one are the one reason we chose to not align -> with :: in the ouroboros-consensus/doc/StyleGuide.md. (Note that we also chose to keep :: on the first line for other reasons.)

So, as your judgement call, if you feel like you're fixing up this kind of alignment for almost all of the signatures in some module, feel free to also change all those signatures in such a module to adhere to the StyleGuide.md. EG in this case (depending how many of the other signatures in this file you're already touching), it'd change to:

checkIfFollowerExists ::
     CPS.FollowerId
  -> Model blk
  -> a
  -> Either (ChainDbError blk) a

(I'm suggesting preserving this style despite the StyleGuide.md because we don't want to inflate the diff too much with signatures that would otherwise be unchanged by this PR. Also, the "existing style" of a module is generally to be preserved, despite the StyleGuide.md, unless we have good reason.)

Somewhat separately: in the case of a rename, I would be tempted to have separate commits for the rename and for the alignment-fixup. Not crucial, but I think that would make it easier for me personally to review, for example.

[nfrisby]

I see some strays left. Here is an example from each module in which I'm getting hits.

ouroboros-consensus-test/test-storage/Test/Ouroboros/Storage/ChainDB/Model.hs:    = m { cps = CPS.deleteFollower rdrId (cps m) }
ouroboros-consensus-test/test-storage/Test/Ouroboros/Storage/ChainDB/StateMachine.hs:data Cmd blk it rdr
ouroboros-consensus/src/Ouroboros/Consensus/Network/NodeToClient.hs:          (\rdr -> runPeer
ouroboros-consensus/src/Ouroboros/Consensus/Network/NodeToNode.hs:          (\rdr -> runPeerWithLimits

My command was git grep -Ei -e '\brdr' -e 'rdr\b' -e '\breader[^T]' -e 'reader\b' -- 'ouroboros*hs'

So close!

[nfrisby]

Wow! That was a lot of lines! I found a few issues. You were very thorough! Again, I really don't think Thomas and I realized how big this would be. Thanks for persevering.

It's also possible that some of my comments were off base -- I wasn't looking at the context in the module, just the diff on GitHub. So feel free to shut me down in some of these comment threads. EG Maybe I applied the Consensus StyleGuide.md to a module that doesn't deserve it.

FYI, I can re-review the fix ups on this first thing tomorrow -- ie by 3 or 4pm Central European.

[nfrisby]

The last step will be to squash the commits. Like we chatted about, one big commit is fine for this 👍

[EncodePanda]

I see some strays left. Here is an example from each module in which I'm getting hits.

ouroboros-consensus-test/test-storage/Test/Ouroboros/Storage/ChainDB/Model.hs:    = m { cps = CPS.deleteFollower rdrId (cps m) }
ouroboros-consensus-test/test-storage/Test/Ouroboros/Storage/ChainDB/StateMachine.hs:data Cmd blk it rdr
ouroboros-consensus/src/Ouroboros/Consensus/Network/NodeToClient.hs:          (\rdr -> runPeer
ouroboros-consensus/src/Ouroboros/Consensus/Network/NodeToNode.hs:          (\rdr -> runPeerWithLimits

My command was git grep -Ei -e '\brdr' -e 'rdr\b' -e '\breader[^T]' -e 'reader\b' -- 'ouroboros*hs'

So close!

Ha! I know how they got here. After recent patch to memleak some of those variables were introduced and I did not notice that while rebasing. Thanks for the good catch!

[EncodePanda]

Ha! I know how they got here. After recent patch to memleak some of those variables were introduced and I did not notice that while rebasing. Thanks for the good catch!

but I also see there were other few that I've missed
my grep-fu not strong enough

[EncodePanda]

@nfrisby I've added few commits to this PR. Please review them. I will squash everything once I get the final 👍 on this.
Thanks for the detailed and close look on this one.

[nfrisby]

I rebased onto master (no conflicts, it was automatic) and fixed up a couple more alignment hiccups. Some of those hiccups you and I had fixed on our call yesterday, so I'm guessing we accidentally lost them somehow. We probably jumped to the squash without committing first our staged changes?

Anyway, my fixup commit is entirely whitespace. If I show it with the whitespace diff suppressed, it's empty:

$ git show -w 3cefd20c8
commit 3cefd20c8585131235a325df1d2ba9d0028cdfe4 (HEAD -> EncodePanda/rename-reader-to-follower)
Author: Nicolas Frisby <nick.frisby@iohk.io>
Date:   Thu Jan 21 15:51:38 2021 -0800

    TOSQUASH last alignment fixups
$

I'm going to squash it now and then merge! 🙌

Thomas has departed.

[nfrisby]

bors r+

[iohk-bors]

Build succeeded:

• ci/hydra:Cardano:ouroboros-network:required
• buildkite/ouroboros-network