Bump PKGREVISION. While here, set LICENSE.
It turns out there were a lot of these.
Version 0.4: * New features: * Support for image types (color models): GREY, GREYA, RGB, RGBA, CMY, CMYK, CMYKA and custom * Additional builtin energy functions based on luma instead of brightness * Support for custom energy functions * Functions to read out the energy map * Cancel function to interrupt computations * Additional methods for bias and rigidity masks * Use a single progress instead of several ones when the update is performed in steps * Can handle input buffers non-destructively * Performace related: * Faster rescaling thanks to optimisations in the code * Spare memory when not using features (e.g. the bias) * Configuration * The default installation directory is now /usr instead of /usr/local * Compiling under MinGW now correctly produces a DLL with the default settings * Now it also compiles under MS Visual C++ * All public macro names start with LQR_* (old versions are still available but can be disabled at configure time) Version 0.3: * New features: * Enlargement is not limited to 200% any more (it is automatically performed in steps) * Added readout functions for the internal visibility map Version 0.2: * New fetures: * Support for 16-bit, 32-bit and 64-bit per channel images * Rigidity masks * Left-right/top-bottom balance * Performace related: * Improved attached visibility maps management (lower memory requirements) * All internal buffers turned from double to float, improving both speed and memory requirements * Documentation: * Reviewed manual and completed refence * Added man pages
This changes the buildlink3.mk files to use an include guard for the recursive include. The use of BUILDLINK_DEPTH, BUILDLINK_DEPENDS, BUILDLINK_PACKAGES and BUILDLINK_ORDER is handled by a single new variable BUILDLINK_TREE. Each buildlink3.mk file adds a pair of enter/exit marker, which can be used to reconstruct the tree and to determine first level includes. Avoiding := for large variables (BUILDLINK_ORDER) speeds up parse time as += has linear complexity. The include guard reduces system time by avoiding reading files over and over again. For complex packages this reduces both %user and %sys time to half of the former time.