Solving replays are autom. saved without finishing level

[Forestidia86]

Maybe this is even desired:
But if you fulfil the save requirement and just exit the level with Esc/Menu then the solving replay is automatically saved although you haven't finished the level properly.
This means that infinite replays can be automatically saved under solved.
Simon indicated that infinite solving replays are no problem for him, so maybe a non-issue. But I find it incomplete to provide such a replay and it accidentally happened to me once.

[ccexplore]

It took me a while to better understand the implication of what you are saying.  I think the scenario may be that the user fast-forwarded through a large amount of game time in order to reach the point where the save requirement is fulfilled, and then exit the level via ESC.  Since (I'm assuming) neither the use nor the timing of the ESC is recorded in the replay (to help inform how far the replay should execute to reach the user's desired end state), it is possible for a replay checker to be unable to properly verify the resulting replay, because it took longer than the timeout used by the checker before the save requirement has been fulfilled in execution.

At least I feel that's probably the primary problem you may be referring to, let us know if there are other aspects to address as well.

[Forestidia86]

Since (I'm assuming) neither the use nor the timing of the ESC is recorded in the replay (to help inform how far the replay should execute to reach the user's desired end state), it is possible for a replay checker to be unable to properly verify the resulting replay, because it took longer than the timeout used by the checker before the save requirement has been fulfilled in execution.

There is a timeout that can lead to a fail of a solving replay but that was actually another issue and is rather unrelated to pressing Esc I think or only in very specific corner cases. Inifinite solving replays as such are no issue for the replay checker because it breaks up after the save requirement is fulfilled and gives an OK. 
It's more about the interactive mode that the replay just goes on without coming to the end screen.
Yeah, maybe I only have issues with it. It just feels incomplete when you provide a solution that if you look at it never leads to an end. But thanks, for dealing with the issue.

[Simon]

The replay checker doesn't care, but it's nice to end solving replays with a nuke.

I could add the nuke at end of replay when somebody exits a game, has not nuked, but has lixes left alive.

I've wanted to do this in multiplayer anyway. You can quit multiplayer games without nuking or lose connection, and then, ideally, this should generate a nuke replay action.

-- Simon

[ccexplore]

Ok, that make sense.  It was tripping me up a bit when pressing ESC is described as "haven't finished the level properly", since a good number of players do use that to finish levels in lieu of wasting more time with a nuke.  It wasn't clear until you clarify that the "incompleteness" is in the replay not stopping on its own.

Like Simon said, the program could potentially auto-stop such replays (I've yet to think of a case where the user somehow cannot abort and take over the replay until after waiting for the save requirement is reached, instead of before, maybe except for some very unusual and very specific challenge conditions) and it would address your concern about replay not stopping on viewing.  The replay does solve the level even though ESC is used and thus most definitely should be saved as such, or it will be pretty annoying to the players who regularly ESC-solves.

[Forestidia86]

Ok, that make sense.  It was tripping me up a bit when pressing ESC is described as "haven't finished the level properly", since a good number of players do use that to finish levels in lieu of wasting more time with a nuke.

Okay, sorry, I didn't know that. Thanks for the clarification.
I personally have just the intuition that in a game a level is only properly solved if you get the ending screen or at least the game finishes, that you have some kind of endpoint.
It furthermore feels just like a proof that is just missing one trivial step and is therefore not valid. Proofs are usually finite entities and the replay then just isn't.
But I can fully accept that it is seen otherwise. It's just my personal feeling about it. That's why I suspected that it is rather a non-issue.