flat assembler

Quote:

The purpose of easm is to be a simple-to-use assembly language which is to specialise in producing the smallest possible executable which executes in the smallest amount of time.

i doubt it can accomplish this by introducing higher-level abstractions (string, HLL procedures), and with "one pass" assembling

Quote:

The assembler is being implemented in C++ using the Microsoft Visual C++ 6 compiler.

wonder how much portable the sources are. Is it ANSI C, or Microsofty C?

IMHO it's bad design with the "one pass", author is just making it easier for himself, by limiting possibilities. In fact he has hit this problem himself, and make it two pass assembler already. See this

Quote:

dword dwSomeDword = deadbeefh

wonder how he differentiates numeric constant "deadbeefh" (0xDEADBEEF) and variable called "deadbeefh".

Quote:

A typical easm source code file is organised into 'sections'. A section is designed to contain a certain type of information.

Quote:

The order of sections in an easm source file is significant. The reason that sections must be ordered specifically is due to the fact that easm makes just one pass of the source file. This means that everything that must be done in order to create the final executable must be done by the time the end of the source code is reached by the parser. Consequently, data and function declarations must be located by the parser before any reference to them is made. If the sections containing these symbols are located after the sections that make reference to them, the assembler will determine that the symbols are undefined

veeeeeery bad (or lazy) design. This means assembler is completely unusable for anything larger. Imagine you want to split surce to two files, how would you do it? in FASM it's possible either by compiling to object format, or when compiling directly to EXE by forward referencing and idata/udata macros.

Quote:

Variables occupy physcial space in the executable and as such, the more variables declared in an application, the larger (in size) the application will be. Put another way, if a particular easm application creates a 1024 byte (1KB) string variable, 1KB of data will be reserved in the executable and the final executable will be 1KB larger in size.

WTF?

sorry for being harsh, but this assembler is really bad/lazy designed, and almost unusuable for real things.