Closed CalebQ42 closed 1 year ago
This would allow the program to continue (in a broken state) but doesn't really fix the issue that causes the malfunction.
@CalebQ42 would you agree that it is better to not run at all rather than to run in an unsupportable half-working "random" state?
For me it comes down to 2 main reasons that it would be fine the ignore the error in this instance:
appimaged
, but instead with the external service (zeroconf) which we have minimal control over.appimaged
.Ultimately this is similar to your decision to not add configuration to appimaged
where you end up with multiple different setups that can each have their own quirks and problems. Though I might disagree with it, it is a 100% valid decision that I will happily follow since it is your project.
Sorry if this isn't very elegantly put, I'm fairly sleep deprived ATM, lol.
You are right in that we are currently not yet making use of Zeroconf. It is something I wanted to do but never came around to doing so far. The idea was that appimaged instances on the local network would find each other, and AppImages could be exchanged/updated over the network "automagically".
Instead of leaving this in a half-working state, I think the cleaner option would be to comment out Zeroconf entirely for now, until we really start making use of it.
A simple fix for #251 by replacing
log.Fatal
in a few places withlog.Println
.