.IX XDR futures.LPThe XDR standard lacks representations for bit fields and bitmaps,since the standard is based on bytes.  Also missing are packed (orbinary-coded) decimals..LPThe intent of the XDR standard was not to describe every kind of datathat people have ever sent or will ever want to send from machine tomachine. Rather, it only describes the most commonly used data-typesof high-level languages such as Pascal or C so that applicationswritten in these languages will be able to communicate easily oversome medium..LPOne could imagine extensions to XDR that would let it describe almostany existing protocol, such as TCP.  The minimum necessary for thisare support for different block sizes and byte-orders.  The XDRdiscussed here could then be considered the 4-byte big-endian memberof a larger XDR family..NH 1\&Discussion.sp 2.NH 2\&Why a Language for Describing Data?.IX XDR language.LPThere are many advantages in using a data-description language suchas  XDR  versus using  diagrams.   Languages are  more  formal thandiagrams   and   lead  to less  ambiguous   descriptions  of  data.Languages are also easier  to understand and allow  one to think ofother   issues instead of  the   low-level details of bit-encoding.Also,  there is  a close analogy  between the  types  of XDR and  ahigh-level language   such  as C   or    Pascal.   This makes   theimplementation of XDR encoding and decoding modules an easier task.Finally, the language specification itself  is an ASCII string thatcan be passed from  machine to machine  to perform  on-the-fly datainterpretation..NH 2\&Why Only one Byte-Order for an XDR Unit?.IX XDR "byte order".LPSupporting two byte-orderings requires a higher level protocol fordetermining in which byte-order the data is encoded.  Since XDR isnot a protocol, this can't be done.  The advantage of this, though,is that data in XDR format can be written to a magnetic tape, forexample, and any machine will be able to interpret it, since nohigher level protocol is necessary for determining the byte-order..NH 2\&Why does XDR use Big-Endian Byte-Order?.LPYes, it is unfair, but having only one byte-order means you have tobe unfair to somebody.  Many architectures, such as the Motorola68000 and IBM 370, support the big-endian byte-order..NH 2\&Why is the XDR Unit Four Bytes Wide?.LPThere is a tradeoff in choosing the XDR unit size.  Choosing a smallsize such as two makes the encoded data small, but causes alignmentproblems for machines that aren't aligned on these boundaries.  Alarge size such as eight means the data will be aligned on virtuallyevery machine, but causes the encoded data to grow too big.  We chosefour as a compromise.  Four is big enough to support mostarchitectures efficiently, except for rare machines such as theeight-byte aligned Cray.  Four is also small enough to keep theencoded data restricted to a reasonable size..NH 2\&Why must Variable-Length Data be Padded with Zeros?.IX XDR "variable-length data".LPIt is desirable that the same data encode into the same thing on allmachines, so that encoded data can be meaningfully compared orchecksummed.  Forcing the padded bytes to be zero ensures this..NH 2\&Why is there No Explicit Data-Typing?.LPData-typing has a relatively high cost for what small advantages itmay have.  One cost is the expansion of data due to the inserted typefields.  Another is the added cost of interpreting these type fieldsand acting accordingly.  And most protocols already know what typethey expect, so data-typing supplies only redundant information.However, one can still get the benefits of data-typing using XDR. Oneway is to encode two things: first a string which is the XDR datadescription of the encoded data, and then the encoded data itself.Another way is to assign a value to all the types in XDR, and thendefine a universal type which takes this value as its discriminantand for each value, describes the corresponding data type..NH 1\&The XDR Language Specification.IX XDR language.sp 1.NH 2\&Notational Conventions.IX "XDR language" notation.LPThis specification  uses an extended Backus-Naur Form  notation fordescribing the XDR language.   Here is  a brief description  of thenotation:.IP  1.The characters.I | ,.I ( ,.I ) ,.I [ ,.I ] ,.I " ,and.I * are special..IP  2.Terminal symbols are  strings of any  characters surrounded bydouble quotes..IP  3.Non-terminal symbols are strings of non-special characters..IP  4.Alternative items are separated by a vertical bar ("\fI|\fP")..IP  5.Optional items are enclosed in brackets..IP  6.Items are grouped together by enclosing them in parentheses..IP  7.A.I * following an item means  0 or more  occurrences of that item..LPFor example,  consider  the  following pattern:.DS L"a " "very" (", " " very")* [" cold " "and"]  " rainy " ("day" | "night").DE.LPAn infinite  number of  strings match  this pattern. A few  of themare:.DS"a very rainy day""a very, very rainy day""a very cold and  rainy day""a very, very, very cold and  rainy night".DE.NH 2\&Lexical Notes.IP  1.Comments begin with '/*' and terminate with '*/'..IP  2.White space serves to separate items and is otherwise ignored..IP  3.An identifier is a letter followed by  an optional sequence ofletters, digits or underbar ('_').  The case of identifiers isnot ignored..IP  4.A  constant is  a  sequence  of  one  or  more decimal digits,optionally preceded by a minus-sign ('-')..NH 2\&Syntax Information.IX "XDR language" syntax.DS.ft CWdeclaration:	type-specifier identifier	| type-specifier identifier "[" value "]"	| type-specifier identifier "<" [ value ] ">"	| "opaque" identifier "[" value "]"	| "opaque" identifier "<" [ value ] ">"	| "string" identifier "<" [ value ] ">"	| type-specifier "*" identifier	| "void".DE.DS.ft CWvalue:	constant	| identifiertype-specifier:	  [ "unsigned" ] "int"	| [ "unsigned" ] "hyper"	| "float"	| "double"	| "bool"	| enum-type-spec	| struct-type-spec	| union-type-spec	| identifier.DE.DS.ft CWenum-type-spec:	"enum" enum-bodyenum-body:	"{"	( identifier "=" value )	( "," identifier "=" value )*	"}".DE.DS.ft CWstruct-type-spec:	"struct" struct-bodystruct-body:	"{"	( declaration ";" )	( declaration ";" )*	"}".DE.DS.ft CWunion-type-spec:	"union" union-bodyunion-body:	"switch" "(" declaration ")" "{"	( "case" value ":" declaration ";" )	( "case" value ":" declaration ";" )*	[ "default" ":" declaration ";" ]	"}"constant-def:	"const" identifier "=" constant ";".DE.DS.ft CWtype-def:	"typedef" declaration ";"	| "enum" identifier enum-body ";"	| "struct" identifier struct-body ";"	| "union" identifier union-body ";"definition:	type-def	| constant-defspecification:	definition *.DE.NH 3\&Syntax Notes.IX "XDR language" syntax.LP.IP  1.The following are keywords and cannot be used as identifiers:"bool", "case", "const", "default", "double", "enum", "float","hyper", "opaque", "string", "struct", "switch", "typedef", "union","unsigned" and "void"..IP  2.Only unsigned constants may be used as size specifications forarrays.  If an identifier is used, it must have been declaredpreviously as an unsigned constant in a "const" definition..IP  3.Constant and type identifiers within the scope of a specificationare in the same name space and must be declared uniquely within thisscope..IP  4.Similarly, variable names must  be unique within  the scope  ofstruct and union declarations. Nested struct and union declarationscreate new scopes..IP  5.The discriminant of a union must be of a type that evaluates toan integer. That is, "int", "unsigned int", "bool", an enumeratedtype or any typedefed type that evaluates to one of these is legal.Also, the case values must be one of the legal values of thediscriminant.  Finally, a case value may not be specified more thanonce within the scope of a union declaration..NH 1\&An Example of an XDR Data Description.LPHere is a short XDR data description of a thing called a "file",which might be used to transfer files from one machine to another..ie t .DS.el .DS L.ft CWconst MAXUSERNAME = 32;     /*\fI max length of a user name \fP*/const MAXFILELEN = 65535;   /*\fI max length of a file      \fP*/const MAXNAMELEN = 255;     /*\fI max length of a file name \fP*/.ft I/* * Types of files: */.ft CWenum filekind {	TEXT = 0,       /*\fI ascii data \fP*/	DATA = 1,       /*\fI raw data   \fP*/	EXEC = 2        /*\fI executable \fP*/};.ft I/* * File information, per kind of file: */.ft CWunion filetype switch (filekind kind) {	case TEXT:		void;                           /*\fI no extra information \fP*/	case DATA:		string creator<MAXNAMELEN>;     /*\fI data creator         \fP*/	case EXEC:		string interpretor<MAXNAMELEN>; /*\fI program interpretor  \fP*/};.ft I/* * A complete file: */.ft CWstruct file {	string filename<MAXNAMELEN>; /*\fI name of file \fP*/	filetype type;               /*\fI info about file \fP*/	string owner<MAXUSERNAME>;   /*\fI owner of file   \fP*/	opaque data<MAXFILELEN>;     /*\fI file data       \fP*/};.DE.LPSuppose now that there is  a user named  "john" who wants to  storehis lisp program "sillyprog" that contains just  the data "(quit)".His file would be encoded as follows:.TSbox tab (&) ;lfI lfI lfI lfIrfL rfL rfL l .Offset&Hex Bytes&ASCII&Description_0&00 00 00 09&....&Length of filename = 94&73 69 6c 6c&sill&Filename characters8&79 70 72 6f&ypro& ... and more characters ...12&67 00 00 00&g...& ... and 3 zero-bytes of fill16&00 00 00 02&....&Filekind is EXEC = 220&00 00 00 04&....&Length of interpretor = 424&6c 69 73 70&lisp&Interpretor characters28&00 00 00 04&....&Length of owner = 432&6a 6f 68 6e&john&Owner characters36&00 00 00 06&....&Length of file data = 640&28 71 75 69&(qui&File data bytes ...44&74 29 00 00&t)..& ... and 2 zero-bytes of fill.TE.NH 1\&References.LP[1]  Brian W. Kernighan & Dennis M. Ritchie, "The C ProgrammingLanguage", Bell Laboratories, Murray Hill, New Jersey, 1978..LP[2]  Danny Cohen, "On Holy Wars and a Plea for Peace", IEEE Computer,October 1981..LP[3]  "IEEE Standard for Binary Floating-Point Arithmetic", ANSI/IEEEStandard 754-1985, Institute of Electrical and ElectronicsEngineers, August 1985..LP[4]  "Courier: The Remote Procedure Call Protocol", XEROXCorporation, XSIS 038112, December 1981.