StackOverflow wrote:This was originally going to be a response to another thread, but it kind of trailed off a bit so I'm posting it in a separate thread. Besides, you can never have enough 0.75/0.76/OpenSpades/Iceball threads, right?
I tend to avoid these topics because they often go nowhere. Hopefully I can summarize a few things here that could possibly be used for reference.
As a bit of a preface, let me just say this. The notion of "x is better than y because y is different from x" is not a good one. It's very easy to make the assumption that just because something is different from what you're used to, it is inherently better. I mean this mainly in regards to the 0.75 vs 0.76 discussions that occur day after day.
First the 0.x issue...
Anybody who has been in this community for even a short while has seen the amount of hatred towards 0.75. But why is there so much animosity towards it? Well, for starters it has been the primary release since March 2012. It has had long-term exposure and has been ran through nearly every edge case possible.
In contrast, 0.76 has had nowhere near the amount of exposure so we can't say for certain which one is "better" than the other. It's a bit of a catch 22. 0.76 has it's own quirks and just blindly asserting that it's superior to 0.75 is a bit baseless. This isn't to say that we should ignore 0.76. Instead, we should be trying to seamlessly introduce users to 0.76 without much (if any) real work on their part. This is where the idea for the BnS Launcher stemmed from.
In regards to Iceball...
Iceball, in my opinion, is not a viable alternative. It has basically turned into something separate from Ace of Spades. There seems to be no clear sense of direction or design. (Reinventing the wheel doesn't count, nor does delegating over insignificant details/micro-optimizations).
New users are easily discouraged by the intentional roadblocks left in place to keep them out. There is little to no sense of support for said users.
The transition from Voxlap to Iceball is far from seamless. With new protocols/file formats etc, users are playing a new game. This was something I warned about early on and is a big reason why Iceball has not been adopted. Sure, it might seem like a bit of an inhibition at first, but nobody is going to adopt it unless there is a seamless transition.
And OpenSpades...
OpenSpades is what a lot of people have been waiting for. An open-source client implementation that has a clear sense of direction. It adheres to the existing protocols (0.75/0.76) seamlessly and uses file formats that players are used to. Upon initial public release, it was already in a relatively stable state. Plus it looks sexy.
Basically, the entry level is extremely low. The only real trade-off is system requirements. Voxlap requires next to nothing to run, meanwhile OpenSpades' requirements are a bit higher.
While OpenSpades is compatible with existing protocols, it's only natural for people to want to extend that. Users want to see new features added, which would require changes to said protocol. This would in turn break backwards compatibility with Voxlap clients. You would essentially have OpenSpades servers and Voxlap servers, which could be a bit confusing for new users. Fragmentation is hardly ever a good thing, but it's only natural for people to want to extend the game.
While on the subject, something else that should be noted about OpenSpades is the double-edged sword that is open-source. With relative ease, a user could fork OpenSpades, modify it with cheats baked into the client and build it. We've considered this and will probably limit OpenSpades usage to only yvt's builds via the launcher.
So what are we supposed to do?
As stated, OpenSpades is pretty ideal for what we want. The key isn't to instantly jump to a new version/implementation, but try to seamlessly support as much as we possibly while simultaneously avoiding severe fragmentation.
Ideally, there would be something like a separate PySnip fork for OpenSpades servers that wish to extend the protocol. OpenSpades servers should be separated from Voxlap servers on the serverlist, to help avoid confusion.
This is where the master server suggestion comes into play. It's something that has been discussed before and I have no objection to it. The master server is very flexible and I could easily add an extra flag check for OpenSpades servers. This way we could differentiate between the two flavors of servers and go from there.
I know I'm forgetting to mention a few things, I'll probably think of them shortly and will append accordingly.
