So, last week Adobe moved Alchemy onto the labs website.

I have been interested (and genuinely excited) in this project from the first murmurs of it.
C/C++ to AVM2 byte-code. Sounds great for our development team. We have loads of cross-platform C/C++ skills and very good knowledge of the Flash Player AVM [pdf].

So I have been racking my brain on some really cool uses for it.

The most obvious benefit of Alchemy is that of re-using the vast amounts of already written C/C++ code out in the wild and what we also have in-house.

This however, is not without issues…

To make use of existing C/C++ code it must firstly be platform independent (or at least have very little platform dependence) and secondly be portable to the Alchemy compiler.
The next issue is that of relevance. Is it really worth the effort porting existing C/C++ code to compile against the Alchemy compiler (and I can assure you this is not going to be trivial in 9/10 cases) when you may as well do a language port – e.g. re-write in AS3?

At the moment, there are possibly a few cases where using Alchemy will yield a speed increase over compiled AS3.
This is because the current Alchemy compiled byte-code makes use of a couple of undocumented AVM2 op-codes. However, I assume these speed improvements will filter down to the actionscript 3 compiler at some point.

So whats the future for Alchemy?
Lets assume that these undocumented op-codes get used by the AS3 compiler so that there is no speed advantage of Alchemy and that porting existing C/C++ code to compile with Alchemy takes significant time.
This leaves no advantage at all.
However, lets assume Adobe choose to keep Alchemy with speed advantages over the AS3 compiler. They would then have a commercial product offering for specialized use cases.