One of the lead devs on the bukkit mod api that we used to use on minecraft tried to shut the project down (and to be fair it's been pretty dead for a while which is why we stopped using it it was really far behind) Turns out he can't as Bukkit was sold to mojang when they hired the original team to write the long long long long long awaited proper mod api for minecraft. They are going to get one of the other devs who now works for them to bridge it till their proper mod api comes up some time around the head death of the universe.
Bit of an odd story really and since they've been promising the mod api since before the official release of 1.0 in 2011. I even saw these guys discuss it a few pax's ago and I guess they've been twiddling their thumbs since then.
Still sounds like minecraft 1.8 is pretty soon and some nice new stuff coming in with it
Comments
I think they promised the mod API without asking anyone how hard it is to make a mod API. When they came to actually code it, they realised that it was a mammoth task and would have to rewrite loads of stuff. On top of that, they needed to get realms sorted out, which needed loads of changes to the code base. Making promises like that is the work of quite a young development team who aren't very good at estimating.
I also think that there is little or no margin to be had with mods. It does extend its playable life but they don't make money directly from it. I think they would do well to host the mods and make them compatible. That's a huge undertaking with enormous cost and as they've found out, they cannot trust the community to do it for them!
1.8 is looking smashing, can't wait.
This is all true but even so it has been 3+ years and in that time as we can see from this debacle that they purchased bukkit and the bukkit dev team so they had several people who knew exactly what it took to develop a mod api as they had already done so. It just seems like they should have gotten something done by now. Their promise of mod support probably goes further back than that notch originally promised it pre 1.0 if they had wanted they could have integrated bukkit as a quick fix and done a proper version later. I remember them talking about their plans at a pax panel just after Jens took over and they had brought on the bukkit team they were talking about how they were "Doing it right" this time but it seems odd that some 3 years on there is nothing to show for it. I guess maybe they got put on other tasks that made them more money like realms.
I think it's unfair to call bukkit a mod API, it's a library hack. That's different because the way you structure a mod API is to minimise changes to the API given updates to the core. Separating what is core and what is moddable through the API is fabulously difficult to do. With the best will in the world, the Mojang team are only professional games programmers because they have worked on Minecraft. If they had a more heavy weight team, they might find that the mod API would appear quicker because they will have hired in knowledge from the outside.
They defo got put onto tasks with higher margins and rightly so.
Bukkit had the external api down in that it was mostly unchanging and standardized mods worked from one version to the next with no alterations they just had the problem of rewiring their interface into the minecraft code each time. With them inside the team that get's a lot easier and had they just ported bukkit to fit in the internals of minecraft I think they would have been done by now.
I'm not sure if that's true that minecraft was their first game project I know most of the core mojang people were game devs long before minecraft they were people notch had worked with in his previous game dev jobs and then poached once he had mad cash. I'm not sure about the bukkit people but even if they were they weren't working in isolation they had access to experienced devs to get the design done and implemented.
Sure they should focus on paying the bills but they should also be honest and not promise stuff they are not going to deliver.
It smells to me that they made massive promises they could not deliver. That's the sort of thing people with little experience do, especially those in complex code bases. I imagine whatever mod API they're building also needs to work with Realms, so you can click deploy to your own map. That all sounds like a massive task me. I've googled all the original Devs, and none of them had much experience (for example, Dinnerbone). I thinking of someone with 10+ years in game programming that has had to go through Mod support nightmares.
I agree that the lack of honesty is a bit off. There are mods I want to build but don't want to start until they sort out the API. :-(
It also seems that they scuttled the only mod system that we did have by essentially ceasing work on bukkit. At this stage Bukkit seems just a waste of resource for them I know they've promised dinnerbone is going to pick it up again but it seems pointless at this stage dump it and focus on the real deal.
I agree, focussing on the real deal is what they should do for us. I think there are a lot of server admins out there who rely absolutely on the mods that bukkit provides - for revenue. I think it's a PR claim to say that bukkit is going to maintained by Mojang because they've already got a shady reputation for nailing some server providers, for setting up in direct competition and for the rather late EULA enforcement (which I think was understandable).
Also I imagine that it's easier to code bukkit if you own both sides of the integration: you know what's coming, when you break bukkit you know where and why. A lot of the thinking about what to change in bukkit will be done at the same time that the code is changed in core. It's probably less work for Dinnerbone to do it than someone trying to work out the changes from outside.
What I am less convinced about is the level of support for the mod builders. Keeping bukkit up to date is one thing but providing them with detailed support is another. Dinnerbone can't do that alone.