David Millington

Why should it be a separate list? What's wrong with having nodes that represent comments inside the syntax tree? A comment is syntactically valid - its syntax is such that...

That is a good point - abstract vs concrete. Is it possible - or should it be possible - to reconstruct the original code, exactly as it was, from the...

Roman, all I can think of is that you end up with nodes all over the place: embedded anywhere. And that might not be ideal. On 2 September 2015 at...

That's tricky. I have empirical evidence, which is memory stability in the IDE. I haven't done a full test which would require tracking all FastMM allocations and free areas of...

Yes, I branched from my string cache branch, since it's part 2 of the memory fragmentation code. The string cache is IMO the most important of the two. On 23...

It's not threadsafe, no. Of course, it's not used unless you turn it on - the code, when compiled with the define undefined, is absolutely unchanged from the current code....

Hmm. Well, I could certainly rework this to do that instead. That would be a better option?

No. Why is one typeparams / typeparam / type, and the other typeargs / type? Is there a reason for the different system to contain the same information?

Many of the strings could be replaced completely by enums, especially if there were specialised node types for particular kinds of element. Eg, the method type of "procedure", "function" etc...

> Actually, I was going to make the abstract tree typed Yes, I should have mentioned it was Roman's idea to introduce types (although I'd been thinking about it as...