- Official Post
I'm going to bump this because my points around the network and similar, really do need a decision making on. videogamesm12
How we can all improve TF
Please Note: The TotalFreedom Forum has now been put into a read-only mode. Total Freedom has now closed down and will not be returning in any way, shape or form. It has been a pleasure to lead this community and I wish you all the best for your futures.
-
-
I'd say let's not rush to decisions we might regret later.
As long as we have at least 2 playable game modes, I don't see why the hub system can't stay.
My only concern is that Ryan is still the only one able to do some stuff network-wise, which makes him a single point of failure.
In the next paragraph I will lay out the procedure for getting a change to TFM pushed to the server:
- You make a branch off the development branch in the TFM repo.
- You commit your changes to that branch.
- You open a pull request for the changes to be merged into development.
- You wait for another developer to review your changes.
- If necessary, you make the requested changes and then go back to step 4.
- If your changes are approved, you or another developer merge it into the development branch.
- The changes on the development branch are tested until they are deemed ready for the main branch.
- Successful testing on the development branch means that it will be merged into main.
- A release will be created from the main branch.
- That release will be put on the server.
We can simplify the process. What about skipping the whole development branch and opening pull request directly to main? Releases can then be created monthly (if there are changes) or when a critical update is to be deployed.
-
We can simplify the process. What about skipping the whole development branch and opening pull request directly to main? Releases can then be created monthly (if there are changes) or when a critical update is to be deployed.
How would this simplify development? You're just moving the point of failure from dev branch to main branch..
would have been a complete fucking disaster at the companies i worked for kekw
-
We can simplify the process. What about skipping the whole development branch and opening pull request directly to main? Releases can then be created monthly (if there are changes) or when a critical update is to be deployed.
How would this simplify development? You're just moving the point of failure from dev branch to main branch..
would have been a complete fucking disaster at the companies i worked for kekw
Devs say the process is too complicated. This is the only simplification I can think of. If it's unsafe, then what's the point of complaining?
-
- Official Post
How would this simplify development? You're just moving the point of failure from dev branch to main branch..
would have been a complete fucking disaster at the companies i worked for kekw
Devs say the process is too complicated. This is the only simplification I can think of. If it's unsafe, then what's the point of complaining?
The reason for the dev branch was to ensure we had a route to apply hot fixes or bug patches to the current release.
This process is not complex or confusing, this was the bare minimum you'd see anywhere that has a focus on code quality.
Any further simplified process will be ultimately at the risk of unknown or unstable code making it to prod.
-
- Official Post
My only concern is that Ryan is still the only one able to do some stuff network-wise, which makes him a single point of failure.
There's very little network wide currently that only I can do, and honestly the majority of the things that fall into this category will already break if I die / go awol as things like the domain will go bang anyway.
-
We can simplify the process. What about skipping the whole development branch and opening pull request directly to main? Releases can then be created monthly (if there are changes) or when a critical update is to be deployed.
Yeah, just have a main branch and then create branches off of it to add features & fix bugs, and then open those pull requests. There's no need to have a "buffer" branch like the purpose of the development branch, since as we've seen it just gets in the way of deploying critical security fixes in the plugins we maintain & convergent histories between the development branch and main branch (i.e. the main branch getting merged into the development branch, which ideally shouldn't happen with that system. Ever.)
Any further simplified process will be ultimately at the risk of unknown or unstable code making it to prod.
This is what pull request reviews & testing are supposed to prevent. Which would still happen with the simplified system that Tizz kindly suggested.
-
- Official Post
There's very little network wide currently that only I can do
You mean aside from the plan to update to 1.19.4 that we are supposed to and need to carry out as soon as possible which requires you to create another server to add to the network, which requires you yourself to do in our current situation?
-
- Official Post
There's very little network wide currently that only I can do
You mean aside from the plan to update to 1.19.4 that we are supposed to and need to carry out as soon as possible which requires you to create another server to add to the network, which requires you yourself to do in our current situation?
it doesn't require me to do it. Rylie for one is able to do so but again due to 1.20 and the length of time it's taken so far I'd asked you to do something on teams prior to us progressing in any case.
-
- Official Post
it doesn't require me to do it. Rylie for one is able to do so
Would it not require you, the only person on this server with access to NetworkManager, to add it as a server to the network?
-
- Official Post
it doesn't require me to do it. Rylie for one is able to do so
Would it not require you, the only person on this server with access to NetworkManager, to add it as a server to the network?
Yes, though the server doesn't need to be joined to the network to start getting things working. And again, we have 2 test servers, neither of which are staged ready to be "Promoted" to a prod server.
I'll also say, while we decide if TF will remain a network or not, I don't have any intention of creating more servers and using up more capacity, as those servers would cease to be accessible once the network is terminated if we un-do the network.
Right now there isn't a server to add to NM and there may never be one following this thread and the push to de-network and "simplify" things back to how they were prior to me taking over
-
while we decide if TF will remain a network or not
I disagree on removing the current network. Where are we supposed to put the plot server when the replacement for the freedom server is deployed? Do we really want to throw away more than a year of planning new stuff to open new opportunitites and make the place less boring for the few still playing? Do I want to go through the hassle to maintain separate server listings after efforts to regain control of the forgotten ones and to keep all of them in sync? I don't.
