Mudlet - Forum

The first draft of QRegularExpression didn't return the names for the named capture groups, so I requested that feature and it was finally added. It's PCRE-compliant, too, unlike the old QRegExp, so I definitely recommend it.

You have a good point with the functionality in libzip, but it might be something that could be built as a library for Qt for easier integration and compilation. I find a lot of good examples and tutorials out there, if not a ready-built library. Didn't Mudlet used to use QuaZip?

JSON support has been added/improved. I've found that it can be easily integrated with the Lua engine.

I don't see the external dependencies that we've got right now in need of replacement. If it is a pain point, let us know where, we'll see if we can help with that.

YAJL is very well optimised - we use it for the speed. I wouldn't want to replace it with a Qt implementation without running performance benchmarks on. There is also the community thing, YAJL is very widely used - more likely to have fixes & a wider support base.

We've done away with QuaZip.

Hope that helps.

I still can't get libzip to compile on Windows 7. If anyone else has any success, please let me know.

Yeah that's a fair one. Chris, could you take a look?

I just updated the wiki with instructions from a fresh windows 7 install using Qt 5.3

While your efforts are appreciated, this is easily the most complicated set of instructions I've ever seen for compiling anything.

Try compiling qt statically on Windows and your opinion will change drastically . But in any case, compiling on Windows sucks and to be honest, if those instructions are too complicated for someone to follow they have no business using the experimental builds.

I must disagree on both points. Compiling Qt statically is no small feat, but for very good reasons. Mudlet is a relatively small project, and the dependencies are clumsy, which is what I see as the primary reason for delays between official releases.

That is hardly the reason for delays between official releases. That's entirely because we have a release cycle tied to a single individual. After this is finally put out we're going to adopt a better release cycle from the official github repository.