With the 508 beta feeling a lot more stable now, I'm in good shape to really hunker down on getting the webclient DSified. The issue I'm facing with that is, it's a rather big change and I don't think it can all be done incrementally, so I'm gonna try to do as much as I can to prep for it first, holding back the biggest changes till it's further along. I've already started on this process by getting new structures and messages ready. I'm hoping to completely finish it within the 508 series, but I'm not wedded to that so if there's a compelling reason to bump to 509 (like more new language features), I'm open to it.

To better explain what I mean about DSification, from the start the webclient was building the entire map itself, which isn't even close to how Dream Daemon and Dream Seeker handle this; Dream Seeker handles most of the load. It behaved pretty well in a lot of games, but at higher player counts and especially higher frame rates, it gets bad. Currently, the webclient is in sort of a hybrid state where it's still building the lists of icons and deciding how to position them, but the webclient is handling the sorting of those icons. It's not ideal but it's generated huge improvements from the old way. The end goal I have is to make the webclient behave much more like Dream Seeker, so the server does very minimal work--and therefore can handle more players.

With the webclent DSification, one of the things I'm hoping will shake out of it is multiple map control support. It should be fairly trivial at that point.

On the Dream Seeker front, DarkNinjaNaut provided an example of a project with huge icons that Dream Seeker has trouble with. I'm planning to improve the garbage collection as a response, and have a tenative solution working already for the next release.

Speaking of big icons, there's also another longstanding issue with them (server-side) where in some cases they'll just flat-out disappear when you move. I looked into this a while back and found the cause, but it's one of those bugs that doesn't have a trivial fix. I do have a fix in mind however that I think will take care of it (though it will not respect client.dir; the complexity of that is just more than I think the server should deal with). It's not coming just yet, but it is something on my horizon that I've pegged as fairly high-priority. I wanted to put it out there that I have not forgotten about this issue, even though I don't expect to get to it just yet.

There are some minor features on my radar for Dream Maker involving pixel vs. step offsets. 200% zoom is something I'd like for it, but there's a time investment involved so I really don't want to promise any kind of ETA on it yet.

Have an idea for Under the Hood that you'd like to see me write up? Let me know! And as always, thanks to all Members and donaters for their support.