Feat: Autoconnect and Retry

[MrAdder]

Refactored my Auto Connect and Retry feature following the guidance from #12

Possibility to Fix #7

I have not tested this, so I apologise if this does not work first try.

Compiled test version for those who can not compile is linked below; however, it does not include #18

https://ci.appveyor.com/project/MrAdder/vsmr/builds/52123465

For a build that contains #18 as well, use

https://ci.appveyor.com/project/MrAdder/vsmr/builds/52123426

[MrAdder]

@19wintersp your insight into this is welcome
as well as @BenWalker01

[19wintersp]

How does this address #2?

[MrAdder]

Forgot to commit that part its comming soon™

[19wintersp]

Please consider when re-formatting entire files that it can cause merge conflicts and makes reviewing harder. In future, if you do choose to re-format a file, please do it in a separate commit, apply it to all files, and include the formatting rules used (via .clang-format or similar).

There are quite a few instances of implicit casts which should be avoided, but there are worse things already in the codebase. Please do check the compiler output though, as I'm fairly sure at least some of them will be causing warnings to be produced—any such warnings should be addressed before merge.

[19wintersp]

Forgot to commit that part its comming soon™

Please do so in a separate PR if so.

I still believe some more consideration of the best course of action wrt #7 is warranted; I'll put some thoughts in the issue thread and would welcome others' input.

[19wintersp]

I implore you to please test your changes before submitting PRs, or at least mark it as a draft if you do not expect reviews yet.

[19wintersp]

Copying comments from #12:

I'm also not totally convinced of the utility of continued attempts at connection when doing so has failed already, since this is particularly liable to unnecessarily pestering the server with requests it has already deemed invalid (in the likely scenarios of invalid authentication or an existing conflicting connection), or shouting into the void (if it's a network problem).

As before, I think classifying the cause of failure to connect should be a prerequisite for any attempt at reconnection. What exactly is the intended scenario for which automatic retry on connection failure is to be used?

[MrAdder]

I have no issues with reviews at this stage, as ultimately it won't get merged until everyone is happy.

• Tested my side. Empty logonCode
• Max Retries
• login loop

Still seeing if I can get any errors from Hoppies for an inactive Hoppies code

[BenWalker01]

@MrAdder Been busy, sorry. Can't remember where this was up to. Has the auto connect been removed?

[MrAdder]

Will try to remove that in the morning before work; however, sleep is needed

[MrAdder]

@BenWalker01 Done, sorry work kicked up due to weather

[19wintersp]

My previous comment regarding reconnection still stands; I think at minimum an exponential backoff should be implemented for this to be considered at all

[BenWalker01]

Please consider when re-formatting entire files that it can cause merge conflicts and makes reviewing harder. In future, if you do choose to re-format a file, please do it in a separate commit, apply it to all files, and include the formatting rules used (via .clang-format or similar).

There are quite a few instances of implicit casts which should be avoided, but there are worse things already in the codebase. Please do check the compiler output though, as I'm fairly sure at least some of them will be causing warnings to be produced—any such warnings should be addressed before merge.

Coming back to this, as a few people are working on the file, please keep your changes specific to the part you are working on. Please revert any formatting changes, not relevant to the content you are adding.

For @19wintersp you are correct re the compiler output. I'll fix it eventually (probably).