Initializers for multi-dimensional arrays

[Daniel Elbaum]

In <16984 at paris.ics.uci.edu> Doug Schmidt <schmidt at zola.ics.uci.edu> asks
why 


I.
	static bar foo[4][4] =
	{
	  {1,0},  {2,0},  {3,0},  {4,0},  {5,0},  {6,0},  {7,0},  {8,0},
	  {9,0}, {10,0}, {11,0}, {12,0}, {13,0}, {14,0}, {15,0}, {16,0},
	};

isn't the same as


II.
	static bar foo[4][4] =
	{
	  { {1,0},  {2,0},  {3,0},  {4,0},},
	  { {5,0},  {6,0},  {7,0},  {8,0},},
	  { {9,0}, {10,0}, {11,0}, {12,0},},
	  {{13,0}, {14,0}, {15,0}, {16,0},},
	};

according the the latest pANSI draft and traditional UNIX C
compilers.

	dpANSI says that the bracketing of the initialization in I
requires assigning 1 to foo[0][0].i, 0 to foo[0][1].i, 2 to
foo[1][0].i, 0 to foo[1][1].i, etc., since each enclosed left
bracket specifies that the next nested member is to be given the 
following value.  In other words, {2,0} says "give 2 to the first
element of the next foo," which means to assign to the first element
of the next array of bar.

-- 
The workaday world must remain transparent to those who maintain it if
they are to find inspired within them a vision of the history they create.

({uunet,tektronix,reed,sun!nosun,osu-cis,psu-cs}!oresoft!(dan)@oresoft.uu.net)