Every allocation request is rounded up to a power of two in size.
Garbage collection is automatically deallocating memory on the heap when it is no longer referred to. (When there are no remaining pointers which refer to an allocated block, the system deallocates it, without requiring the program to explicitly 'free' (or 'delete', in C++) the block.). It is supported by some programming langauges, but many languages (like C and C++) make GC impossible, or very difficult.The JavaOS only runs programs written for the Java virtual machine (which is garbage-collected), while UNIX is a general-purpose operating system, designed to run programs written in any programming language.
UNIX in fact provides no operating system calls for allocating and deallocating heap memory. Processes may use system calls to grow and shrink their heap segments, but management of memory on the heap is up to individual processes, not UNIX.
...In the sense that a memory reference has two 'coordinates': the segment number, and the position within the segment.
Creation, deletion, open, close, read, write, seek. (Possibly: append, 'query current position'.)
In a flat filesystem, all the files are in the same directory, and every user can see everybody else's files. There would be a problem if, for example, two users wanted to use the same file name for different files. In a 'two-level' filesystem, each user has their own directory.
Advantages: directories are a lot smaller (faster to enumerate files); a single file may be refered to by several names (from several directories); files are more self-contained (which is more robust in case of anything going wrong).
Disadvantages: some operations may be slower (since at least two accesses are needed to fetch a file's inode); less space efficient (since each inode takes up a complete disk block).
There may be one or more platters, each of which has one or two surfaces (upper and lower).
Each surface has many concentric, circular tracks. (The same track on all the surfaces is sometimes known as a cylinder.)
Each track is divided up into fixed-size sectors. (Each sector holds the same number of bytes.)
(Logically, the disk is divided into fixed-size, sequentially-numbered 'blocks'. The block size is often the same as the sector size, but need not be.)
Same problems as with partitioned memory management, really: external fragmentation, difficult to 'grow' files.
It might be used when the size of a file is fixed, and when files are never deleted... on read-only media like CD-ROMs, for example.