EmperorArthur wrote:The most annoying thing is that every toolkit implements their own version of the STL containers. They all work the same way as pure C++, until they don't. Seriously, there's std::string, ogre::string, QString, and whatever boost adds to the mix.
Yep, that is one of the major problems I have with Qt. Irritating as hell. The Ogre string type is actually harmless and I don't see boost causing much problem in this area.
The other issue is MOC. Having an additional preprocessor stage complicates things somewhat and is also a bit limiting. More importantly that thing just doesn't work properly with some C++ features (namely namespaces and templates). Furthermore the signal system is a bit lackluster and the multithreading support is annoying. On top of that Qt encourages a bad C++ coding style (with all the new and such).
Don't get me wrong. I like Qt. It is currently my preferred GUI toolkit. But it has substantial issues.
Are profiling and performance improvements going to happen sometime in the next few versions? Performance on (low) settings is my biggest issue with OpenMW right now.
Greendogo wrote:Ogre 2.0 is going to get rid of DX9 support, is that going to be a problem later?
If their OpenGL support gets upgraded/maintained, then not necessarily (OpenGL is slower, at least right now, sure, but it's also universal). DX9 should be less of a universal requirement for things since XP is no longer supported. Computers still running XP will likely be largely business/non-gamer systems, and any relatively modern gaming system should be on Windows 7 or 8. Again, though, with OpenGL support it should be moot.
Even though 0.30 was released less than a month ago, with 128 closed issues 0.31 is already 50% larger than the currently largest releases, 0.28 and 0.30 (84 closed issues each). In addition, 0.31 marks a major milestone in that it's the first version of OpenMW with which Morrowind can actually be finished.
In light of this, might it not be time to start thinking about a release? At this rate, if the release were to wait until late July the changelog would be too long to fit on the front page and WeirdSexy's video would be long enough to qualify as a feature film.
Only if you go by number of closed issues, which is not a very good metric. If you go by number of commits or lines changed, we have already had bigger releases. I'm ok with releasing soon-ish, but I want to get AI spellcasting in first.
K0kt409P wrote:In light of this, might it not be time to start thinking about a release? At this rate, if the release were to wait until late July the changelog would be too long to fit on the front page and WeirdSexy's video would be long enough to qualify as a feature film.
I've been really impressed by this release so far too, I check to activity/roadmap every couple days, for some reason I really enjoy tracking this project as a huge Morrowind fan. Regardless of how many commits or actual changes were made it's nice to see a mountain of bugs squashed and a few new features to boot. Nice work guys.