-
If you are reporting a bug, please head over to our Technical Support section for Bannerlord.
So, is a major code refactor a common or normal thing?
- Thread starter stevepine
- Start date
Users who are viewing this thread
Can't say specifically about games, but for software development in general, it is common. In my experience, code refactoring usually happens when you need to change the way things are made so that it is easier to implement new things around it, or when something was developed out of the pattern/architecture you are using. Programming language/platform changes are also something that require code refactoring.
Also keep in mind that it is not simply rewriting, you have to test everything that uses that piece of code in other to make sure it is working correctly.
Also keep in mind that it is not simply rewriting, you have to test everything that uses that piece of code in other to make sure it is working correctly.
Last edited:
Code refactors happen all the time. It's basically inevitable that in a big project, especially one with poor management, that the code drifts away from the intended structure into a haphazard mess where nobody really understands the entire thing.
For instance a bunch of Valve source code got leaked recently, and from the comments alone you can tell that nobody had a clear idea of how it worked and they were just monkeywrenching quick and easy fixes so that they could just get it working. These things pile up and make it harder and harder to work efficiently, culminating in this:
I recently refactored the entirety of my own game for similar reasons (in technical language i switched from a pointer based OOP to an integer array based ECS). It took months and was a massive pain, but now that it's over the code is way clearer and everything is much better organised. I can now add things and bugfix much much faster, and there is less possibility for bugs in the first place.
Thee thing is, the only reason I could even do this is because I know my own code inside out. If I had to explain the previous awful system to 10+ colleages, it would have taken even longer despite the extra manpower, and if someone made a mistake it would be far harder to track.
To answer your question though, having a code refactor once the game is already released is basically unheard of.
For instance a bunch of Valve source code got leaked recently, and from the comments alone you can tell that nobody had a clear idea of how it worked and they were just monkeywrenching quick and easy fixes so that they could just get it working. These things pile up and make it harder and harder to work efficiently, culminating in this:
I recently refactored the entirety of my own game for similar reasons (in technical language i switched from a pointer based OOP to an integer array based ECS). It took months and was a massive pain, but now that it's over the code is way clearer and everything is much better organised. I can now add things and bugfix much much faster, and there is less possibility for bugs in the first place.
Thee thing is, the only reason I could even do this is because I know my own code inside out. If I had to explain the previous awful system to 10+ colleages, it would have taken even longer despite the extra manpower, and if someone made a mistake it would be far harder to track.
To answer your question though, having a code refactor once the game is already released is basically unheard of.
I just hope the game works properly enough after the update, they are apparently doing a lot of playtests to find the new bugs that came with the refactoring, but a lot of it will have to be found by the players through extended gameplay and feedback, which will halt progress even more (more bug fix patches than feature patches)
I also hope mods dont break completely after the patch
I also hope mods dont break completely after the patch
I expect most every mod will be broken.
There is a reason for coding songs like this
99 little bugs in the code
99 little bugs
Fixed one now, patch it around
112 little bugs in the code
or the really geeky one
100 buckets of bits on the bus
100 buckets of bits
Take one down, sort it to ground
FF buckets of bits on the bus
A thing only a coder will understand...almost impossible to explain to someone also since there really isnt anything else like it.
Oh come on, how could the mods not break? They're going to have to update all mods likely after the refactored version is live.
What hurts my fee fees is that I WANT to do that extended deep gameplay and help, but fear it'll just be the same stuff with no gameplay changes or updates so it's going to be frustrating to play a new game for the 30th time with nothing actually new, just seeing if it ****s it's self on day 1000 or when you upgrade a town or what not.
Lesbosisles
Knight at Arms
Nah, all mods will die with the new code because of major incompatibility. That's why we get few to zero new mods these days.
Yep, people know they are making useless adjustments until this drops.
guiskj
Knight
In all honesty, major code refactors are definitely not a "run-of-the-mill" thing in software development.
Having been a lead software developer and lately a software architect consultant with some pretty big projects under my belt, I can confidently say that, in my experience, small refactors are somewhat common, but major refactors are certainly not.
If you need to majorly refactor something, that means changing the architecture around it. And that only happens when an error of design has been made.
If I came to any of my firm's clients and said we would be stopped for 4+ weeks to do a major refactor, we would be told to eat the cost ($$$) and likely would never get a repeat contract/project for that client.
Now, let me add the following:
I am not saying that only bad developers/designers cause the need for major refactors. Pioneering something is a valid excuse to need to rethink how you aproach a problem.
What I am saying, though, is that this is not a common thing in software development.
Having been a lead software developer and lately a software architect consultant with some pretty big projects under my belt, I can confidently say that, in my experience, small refactors are somewhat common, but major refactors are certainly not.
If you need to majorly refactor something, that means changing the architecture around it. And that only happens when an error of design has been made.
If I came to any of my firm's clients and said we would be stopped for 4+ weeks to do a major refactor, we would be told to eat the cost ($$$) and likely would never get a repeat contract/project for that client.
Now, let me add the following:
I am not saying that only bad developers/designers cause the need for major refactors. Pioneering something is a valid excuse to need to rethink how you aproach a problem.
What I am saying, though, is that this is not a common thing in software development.
StewVader
Sergeant
I'm just curious why this wasn't mentioned in the half-ass "road-map" post that Tale Worlds put out not too long ago. Could anyone be so incredulous to believe they had no idea they were going to conduct this massive lengthy process? Alas, it's just another failure in the tragedy of failures that is Bannerlord. A long line of missteps, failures to communicate, and piss poor decisions from this company.
After 8 years you do it when you totally screw up.
But no worries, in 6 weeks+ they start fixing known bugs. I mean known bugs that are in the game for 3+ months (formation bug as an example). This "game" is a scam.
TW likes money, they got it. They also got more money because of all those good Steam reviews for a piece of **** like this.
According to their description this game will leave EA like in 8 months, LOL!
They also said that no new features will be added. They just try to get the basic stuff working and let modders finish their game for free.
If we´re lucky sieges will work in like 10 months+
Just never buy any TW game again, maybe they´ll learn then.
Viking Conquest > Warband > This piece of **** (but it looks good, lol)!
But no worries, in 6 weeks+ they start fixing known bugs. I mean known bugs that are in the game for 3+ months (formation bug as an example). This "game" is a scam.
TW likes money, they got it. They also got more money because of all those good Steam reviews for a piece of **** like this.
According to their description this game will leave EA like in 8 months, LOL!
They also said that no new features will be added. They just try to get the basic stuff working and let modders finish their game for free.
If we´re lucky sieges will work in like 10 months+
Just never buy any TW game again, maybe they´ll learn then.
Viking Conquest > Warband > This piece of **** (but it looks good, lol)!
Last edited:
This is great point. Sounds like they just decided out of the blue to do this, not the best confidence for buyers.
Slay the Spire would be an example of a game that had some refactoring done after (Early Access) release, though I don't recall them ever calling it a "refactor." Many of the mechanics were essentially hardcoded such in a way that made them difficult to expand upon, and when a modding scene developed around the game the devs changed/rewrote parts of the code to give them more flexiblility and options for future content as well as to facilitate modding.
It's not clear what "some" refactoring means, and it's possible that the scope of the refactor grew after they put out the list of priorities, however we also don't know the extent of what they've refactored in the past month and a half or so. It actually sounds like the refactor has been done for some time though and they've been bug hunting for the last few weeks.
It was. It's right there near the bottom, and always has been.
It's not clear what "some" refactoring means, and it's possible that the scope of the refactor grew after they put out the list of priorities, however we also don't know the extent of what they've refactored in the past month and a half or so. It actually sounds like the refactor has been done for some time though and they've been bug hunting for the last few weeks.
Last edited:
StewVader
Sergeant
Saying "Refactoring some campaign related classes" is not them saying it. Please try to be less disingenuous.
Saying "we're going to refactor some classes" is different than refactoring some classes? Ok.
StewVader
Sergeant
Yes, because it appears different. Based on the actual Refactor announcement in June, it appears the amount of code involved in this endeavor was quite extensive. You making it sound like they made this plain and known in May in the half-ass "road map" announcement - is disingenuous.
Right, but like I said before, we don't know the extent of what they were planning to refactor vs. what they've actually refactored because the word "some" is ambiguous and we won't know what's changed until they put the patch out. They might've only refactored what they originally had planned, or they might've expanded the scope, but they've almost surely not refactored the entire codebase. Besides, it sounds like the actual refactoring phase didn't take all that long, and the hold-up has been caused by persistent bugs, which are harder to anticipate and plan around. The point is, they stated that they were going to refactor some of the code in the priority list months ago but most people apparently ignored it because it's not that interesting.
Last edited:
I will be dissapointed if we don't get new scenes, quests etc. In this upcoming update. Honestly what have the level designers been doing these last 8 years since not all scenes are done, when they even say themselves it takes 3 weeks to finish their most complicated scenes. (Yes I know they had to create a new engine, but they showcased the level editor many years ago, so it must have been in a somewhat finished state so they could start on the scenes).
