Jump to content

Number Six

Homecoming Team
  • Content Count

    241
  • Joined

  • Last visited

Community Reputation

520 Excellent

About Number Six

  • Birthday 01/01/1004
  1. Just a heads-up, one of the things about the new launcher we're building is that it's designed with native support for add-ons in mind. While the first release won't have that support enabled, the framework for it is there and built into the architecture. Now, I don't know how long after the release of that launcher we'll be looking at for finishing that support and building the UI for it. So I certainly don't want to discourage the work being done here which will undoubtedly be able to come up with something quicker. Whatever you all come up with, once we're past the initial launch and thinking about addon support, it's something I'll want to engage with the community on to get ideas and make sure that what we do meshes well with what the mod authors have been doing. But so you can be thinking about a potential path there; here's what we have in mind for addons: The addon itself is a single file, probably a pigg but possibly a zip file. Addons can use the same package format that we'll be moving to for the game itself - a signed json file - so that mod authors can securely distribute updates that are verified to come from the author. This format already allows for things like specifying package dependencies for if one addon builds on another -- though we may want to think about adding some sort of priority system for situations where they need to be applied in a particular order but it can't be easily expressed in terms of dependencies. The user experience for installation will likely be a simple "do you trust this author?" message, possibly with a user-visible capabilities list that defines what types of files the addon package is allowed to contain -- texture overrides, popmenus, etc. Kind of like "If you hit OK, this addon will be allowed to update textures but not touch anything else" permissions model similar to mobile apps. Instead of extracting the packages into data/, the launcher will instruct the game to insert the addon package into its VFS (above the standard piggs). The layering of the VFS can be crafted to ensure that overrides happen in the proper order, and this makes things a lot simpler and faster than wrangling extracted files. This also makes it very simple for users to toggle addons on and off.
  2. That looks like something completely different.
  3. We'll be talking more about it fairly soon as we get closer to public testing and release, but I can confirm that security is one of the two key design components that led to designing a new launcher from scratch. The other was alluded to earlier and is that it's built without needing any outside libraries or dependencies that the game itself doesn't have, so that if you have something that is capable of running COH (i.e. wine on linux/mac), it should be able to run the launcher. It also has a lot of streamlining of the new player experience to make it easier to get up and running, and several commonly requested features like moving architecture (32/64) out of the launch list and to a persistent setting, ability to download or not download beta updates independently, reordering the launch list, etc. On the security front, while nothing that downloads files from the internet can ever be 100% bulletproof, it's about a hundred times better than Tequila. While there isn't really a "manifest" anymore, all of the components that make up the closest equivalents are cryptographically signed to prevent a malicious third party from modifying them - regardless if it's in flight or on the server directly. There's a full chain of trust all the way from the initial bootstrap that tells the launcher how to install/configure itself down to the package files that contain the sha256 hashes used to verify downloads before saving them to disk. If that somehow were to fail, all of the file operations happen inside a VFS layer that is rooted at the install directory, so it can't be escaped with ".." or absolute paths. There's a lot more that makes it much more robust than the existing methods, but we'll be covering that later, likely in a community update along with some other related items. In general, it should no longer be possible to accidentally launch with half-updated files, and verification is something that will become a last resort to detect disk corruption rather than at the first sign of trouble.
  4. Yes, works out of the box with stock wine on Linux, and we've been working directly with the Island Rum developer to make sure it works on mac (it does, now anyway, macs are a lot more finicky). Should be a much easier setup than Tequila since there's no .NET involved.
  5. Shark Week has been a running gag during development of the launcher and testing of the news feed. I couldn't resist sharing a little bit of that. 😄
  6. Thanks for posting! Tequila is basically dead at this point, but server status is something we already have in our upcoming replacement for it. You should be hearing more about it Soon(TM).
  7. Right here, Patch notes were ready to go but hidden until the patch was actually deployed. It was on staging briefly so we could double check it didn't affect the recent manifest updates to clear out bad mirrors. Only people using the beta manifest would have seen it.
  8. Goldside is amazing on a team. All the XP comes right to you!
  9. To be completely blunt, Rosetta 2 is probably the best bet for a while -- other than just not buying an ARM Mac while they are still making the Intel ones -- but I wouldn't count on much beyond that unless Apple gets a ton of backlash from users and it takes a bite out of their sales. While I wouldn't put an ARM port of COH beyond the realm of possibility, Wine is probably dead outside of Rosetta since there's not really such a thing as a Windows ARM executable anyway. A native neo-Mac port would require porting to ARM, and Xcode, and a proprietary graphics API that isn't used anywhere else.* That's... approaching a full rewrite from the ground up. A native Linux port is more likely, since at least the graphics engine doesn't have to be completely re-engineered, and some of the backend work we've been doing has been moving things in a more cross-platform direction. * barring things like MoltenGL becoming mature enough, but most of those focus on restricted profiles like OpenGL ES or are non-free and expensive to license
  10. AFAIK they are all fixed. Powerhouse did the fix but I glanced at it and it appeared to be correct. Have not heard any reports about any that are still having an issue, but there are a lot of Interface powers.
  11. Oh, I have Corva's icons installed! They were part of my standard 'data' folder full of stuff from live that I copied over. Not sure I have the original distribution for it but I could zip them up easily enough.
  12. Yes, it is intended to be an influence sink. By being more convenient, players who have inf to burn are encouraged to upgrade more often than strictly necessary in order to keep the enhancements at peak efficiency. Players who don't have inf to burn can still make use if it by waiting until they are even level or yellow/red. Or buy them the old fashioned way to pinch pennies and pick and choose specific enhancements. Alternatively they can buy level 30 generic IOs which will save more in the long run but perhaps not always be as effective (but will be consistent!). Not everyone wants to deal with the AH though, even for something relatively easy, and don't care if they're paying more for the convenience. There are pros and cons to both approaches.
  13. Thanks, I'll put it on the list to take a look.
  14. Questioning is healthy, that's why we put these things out for public comment.
  15. But we do have SOs at low level. DfB drops them (and has since it was introduced). As far as the exemplar enhancement scaling, that's something that may be worth looking at, but I wouldn't expect big changes there. IIRC a big part of the scaling was to account for higher-level characters having many more slots available, not just enhancement values.
×
×
  • Create New...