It's awesome, but since my calc got reseted at my exam I haven't yet put it back on. (The farthest I got was level 8 or 9 iirc).

I hope you put it back on your calc. By the way... I will be releasing a new demo within a week. It will have at least 24 levels divided up into 2 or 3 level packs.

I hope you pick up Cyanoid again and find those pesky bugs. Looks like a great game!

Good news, i ordered a silverlink and amazon says it will arrive on Wednesday, so i'll be able to try it out!

EDIT: Oh, and did you program it in asm or TIGCC?

Awesome!

It is coded in C and assembly. I use GCC4TI as my development environment.

He usually programs in GCC4TI

GCC4TI rocks!!!

Why do you prefer GCC4TI over TIGCC? (and is there a IDE for GCC4TI?)

erm, where is it available? And does it run on linux?

Right, https://github.com/debrouxl/gcc4ti is the location where the most recent source code is.

GCC4TI contains an eclectic range of:
* bugfixes: for instance, several crasher bugs, examples which have failed to compile for years in TIGCC, a compiler bug on 64-bit hosts, a bsearch routine which has never worked correctly since before 2001;
* additions: for instance, a subset of the new functionality contributed to TIGCC in 2002, 2003 and 2005. Yes, that's how contributions from multiple persons have been treated in TIGCC, leaving users out in the cold for functionality added up to 10 years ago;
* speed and/or size improvements in the static library linked with programs, for instance -16 bytes on the SAVE_SCREEN code, basically the most frequently used piece of startup code, making the Sprite* routines smaller and faster while adding the RPLC drawing mode, or using a better algorithm for qsort and rewriting to assembly for a speed gain;
* maintainability improvements: build automation in GCC4TI is much better than in TIGCC, whose scripts require a fair amount of hand holding. The GCC4TI build scripts supports cross-compilation of Windows executables;
* portability improvements: care has been taken to make the GCC4TI build scripts work on FreeBSD, MacOS X, OpenSolaris, while the TIGCC build scripts only care about Linux (and badly so, see below);
* other user-visible improvements, such as, but not limited to, not replacing the host system's gcc / as if $TIGCC/bin is put into the PATH before the rest of the path, or explicitly aborting the build upon the first error instead of silently leaving users with broken, incomplete builds, or even supporting building right out of SVN/Git (as any well-behaved open source project does). TIGCC does all three blunders and more.

You'll definitely want to build from the sources available in Git, after checking out the "experimental" branch (whose name sucks - it was experimental for several days, and I never renamed it). GCC4TI 0.96 Beta 10 is pretty outdated by now, though nowhere near as much as TIGCC 0.96 Beta 8.
No release of GCC4TI was produced since I became the maintainer of libti*/gfm/tilp & skinedit/tiemu: tilp clearly has more users than GCC4TI, so I've spent more time on TILP because it makes more sense for the community of users

The Git repository contains suitable tarballs in /pool

Looks awesome as always.

"Git repo", not "Git repro"
https://github.com/debrouxl/gcc4ti.git , git://github.com/debrouxl/gcc4ti.git .