--------------------------------------------------------------------- DDD TODO-LIST 971008 kb created Apart from many TODOs spread out in the source code there are some more features which are still missing. --------------------------------------------------------------------- --------------------------------------------------------------------- 1. Dynamic ddd_CplTable. ddd_CplTable, ddd_NCplTable and ddd_ObjTable are allocated dynamically since 1.8.10. However, the strategy is double-on-overflow, which temporarily needs more memory than with double-sized tables. There should be an option (at compile-time) to switch the strategy to kind-of segmented, i.e., segments of fixed size are allocated and freed on demand. Then, accessing the tables will be more time-expensive, but we could handle more delicate low-memory situations. --------------------------------------------------------------------- 2. DDD_NCopiesPrio(). Implement a function returning the number of object copies with a given priority. --------------------------------------------------------------------- 3. XferCopyObjX and additional bytes. XferCopyObjX sends objects which are bigger than their declared DDD_TYPE. There should be a way to specify whether the additional data bytes are LDATA or GDATA or a mixture of both. --------------------------------------------------------------------- 4. IF-communication and DDD_ATTR. In all IF-communication functions there could be an additional function with masked DDD_ATTR, i.e. a value and a mask. With this feature one could communicate over a set of DDD_ATTR-subinterfaces. --------------------------------------------------------------------- 5. Integration check via DDD_HEADER checksums in debug-mode. There should be a debug-mode which allows to detect memory errors (bad pointers). In this debug-mode, each DDD_HEADER contains a checksum variable, which is a parity-like encoding of the data in the DDD_HEADER. Before each time the HEADER data is accessed (i.e., read _and_ write access), the checksum is computed and compared with the stored one. Errors according to bad pointers and memory faults could be found more easily. --------------------------------------------------------------------- 6. Communicators. MPI-like communicators into DDD. This should be supported already by PPIF. (i.e. groups of processors execute certain DDD functions, like Interface-comm, Xfer, Identify). --------------------------------------------------------------------- 7. Make Notify-module faster. Perhaps it is possible to schedule the small messages in Notify not only through the PPIF-tree, but also to immediate processor neighbours, where this is possible. This possibly can eliminate bottlenecks on the master processor. --------------------------------------------------------------------- 8. F_FRONTEND: ARCH_FLINK. Introduce ARCH_FLINK definition in all arch subdirectories. This could be necessary for the F_FRONTEND production version. --------------------------------------------------------------------- 9. AddData without previous XferCopyObj. It should be possible to start the Xfer-communication directly with calls to DDD_XferAddData (with previous specification of a destination processor). This would enable simple Xfers based on objects w/o DDD_HEADER. ---------------------------------------------------------------------