My BMX2 IDE, (codename CK IDE) is progressing very nicely. This week I’ve managed to join the dedicated language parser, and the whole IDE GUI in a single project. I’ve been working also with Andrey, from QWhale, in some internal aspects of the lexical parser. Also I’ve been making the first benchmarks in therms of document loading time and resources usage, and CK IDE, in its current status, beats any other IDE I’ve used before, included those based on native compilations. Obviously, as the parsers mature, and perform more complex analytics, this speed is going to be reduced, but in general therms, I think we’re in a very good starting point, we’ll see when we fully support data type inference, as the new bmx2 compiler does (this was mentioned on Mark blog, so I’m not revealing any dev-team secret here…) .

There are still lots of small issues related to Folding and document structures, but all in all, the GDI-less implementation of the lexical and syntactical parsers is really introducing a sort of ‘next level’ or performance in the IDE. If BLIde can hold properly projects with hundreds of files in a single project, our new CK IDE, will be able to handle thousands in an even more resource-free way.

However, there are lots of things that are missing in the IDE, so not really even near a complete and decent development environment, we do not have any solution system yet, no intelliprompt, etc. but things are progressing very nicely.

More info to come soon!

Leave a Reply