STANDARDS IN COMPUTER GRAPHICS INFORMATICA 2/88 UDK 681.3.083.7 Niko Guid, Borut Žalik Tehnical Faculty Maribor Abstract: In this paper standards in computer graphics are described. At -first the reasons -for- evolution these standards sre given and then the ways o-f accepting the international standards are presented. A-f terwards the evolution phases af the graphical standards under- ISO and ANSI are interpreted and current stage o-f particular standards are given. In proceeding the place o-f graphical standards and standard proposals in a graphical system are shown. FinalXy, the position and role o-f the graphical standards in a modern CAD system is presented. Povzetek: V dlanku podamo pregled standardov v računalniSki gra-fiki. Najprej opižemo vzroke za razvoj teh standardov, nato pa prikaieno poti, preko katerih nek predlog lahko postane mednarodni standard. Zatem predstavimo rasvojne -faze ISO in ANSI standardov ter podamo trenutne rasvojne stopnje posameznih standardov za računalniSko gra-fiko. V nadaljevanju opiSemo mesto gra-fičnih standardov oziroma predlogov standardov v gra-fičnem sistemu. Nasadnje podamo mesto in vlogo gra-fičnih standardov v modernem CAD sistemu. 1. THE BESININGS 0F STANDARDS DEVELOPMENT Today, a large number o-f di-f-ferent graphical hardware and even more dif-ferent graphical softwar-e exist. A big part o-f this graphical software is device—dependent. The consequences sr s: 1. It is impossible to exchange graphical software between di+-ferent graphical systems 2. There &re problems by i nstal 1 ation o-f old programs on new graphical equipment, although it has been suplied by the same producer, etc. Because o-f these pr-oblems an idea has been appeared to make a device—independent graphical packet. Advantageso-f this device-independent graphical packet are: 1. It could serve di-f-ferent device generations. 2. Programs could work on di-f-ferent graphical systems. 3. Programmers could immediately di-fferent graphical systems. work 4. Graphical systems are distinguished only by quality, price, and e-f-f iciency. 0-f course, this device—independent graphical packets have also some weaknesses. They are slOMer than device-dependent and more memory space is needed. Because the power o-f computers is rapidly increasing and their prices are decreasing, advantages o-f standardization are going over its weaknesses. Peoplei who are opposite to standards in computer graphics, a-f-firm that standarde are against inovations. It ie cleai-, when a standard is accepted, it could not be changed immediately. Portabi 1 ity o-f aplication programs could be achieved in some di-f-ferent ways /ENDE84/: — with development o-f computer languages, — with extension o-f existing program languages with graphical -features or — wi th libraries of graphical subi~outines which could be linked into application program. Expert.s -fr-om the -field o-f computer graphics have chosen the last possibility by the construction o-f internati onal device-independent graphical standard. However, it is least elegant o-f al 1 but it is the best way to awoid confusing in structures o-f program languages. The place o-f the device-independent graphical standard in a graphic system is shown by figure 1. 19 APPUCATION PROORAM OEVICE- INDEPENDENT ORAPHICAL PACKAGE ORAPHICAL DEVICE 1 X3HJ grtphleal itandirdi DEVICE DRIVER ORAPHICAL OEVICE N Figure 1. The placE! o-f graphical standard in graphical system The development o-f graphical standards began in the year 1974, when the Graphics Standard Planing Committee (BSPC) was -found by ACM SIBGRAPH < "Association -for Computing Mashinery Special Interest Broup on Graphics"). This committee met with other international memtaer-s, involved in computer graphics in Seillac (France) in 1976. This meeting had a great in-fluence on -first dra-ft standard cal led' Core System. It was introduced on SIG6RAPH 1977. Two years later, on SI6BRAPH in 1979, an improved version o-f Core was appeared. Soon a-fter that a new group has been -found by German Institute -for Standards DIN which has been worked on a new graphical standard basis. The group has been directed by Jose Encarnacao and it prepared in 1977 a dra-ft standard called BKS (Graphical Kernel System). Two propositions were apeared by ISO in 1979: Core and GKS. Ulorking Group WG2 by ISO decidE?d that only ef-forts on GKS continued. GKS was much more simple, it was 2D, and it was intended -for r-aster devices. On other side Core was 3D and destined -for vector devices. The first dra-ft proposal o-f GKS was made by IBO in 1982. BKS wae accepted as an ISO standard in 1985. In 1981 SIGBRAPH GSPC committee was disbanded and passed over to the ANSI X3H3 committee, which was -founded in 1979. GS - (Ff,',"".' > CGI/CGM (Llltgmgi GKS (Wl«l)ow.) Spiciriollon) BlnSlngi) Figure 2. X3H3 Tehnical Committee and its subcommi ttees It is similar -for International Organization •for Standardization (ISO). ANSI is only a secretariate in IBO's Technical Committee TC97. Ulorking Group WG2 in subcommi ttee SC21 is responsible -for graphical standards. Its sign is IB0/TC97/SC21/W62. ' Some standards which are set up by ISO or ANSI1, are e-ffective, but others are even ignored. From this point o-f view standards could be consider-ed as de -facto and -formal standards /STRA86/. For eKample, IBM's Color Sraphics Adapter (CGA) is a de -facto display standard -for PCs, just as the IBM PC is a de -facto standard for- personal computers. Neither CBA neither IBM PC was -formal standards, but market -factors has adopted them as standards. Among -formal successful and •formal standard, RS-232-C. On about ANS X3.23 This standard -facto standards, layout and later standards we distinguish unsuccess-ful ones. A case o-f which has been widely used, is the other hand, who has heard standard for keyboard layout. has been totally eclipsed by de •first by the IBM Selectric by PC and AT layouts. 3. DEVELOPMENT PROCESS OF STANDARDS GKS has become a basis iar many other proposals o-f standards including PHIGS, C6M, and CGI. IGES, as a standard for transferring CAD/CAM data bases, has been develop£?d in completely another way than Core and GKS (thr-ough others ANSI committees). IGES was accepted as an ANSI standard in 1981. WHO SETS UP THE BTANDARDS? That a proposal becomes a standard, there sre several phases it must go through. It takes rauch more time ior standardisation process in computer graphics than -for standards -from other areas, because the projects are very large and completly new. Evolution process o-f an ISO standard Evolution phases o-f an ISO standards /B0N085, B0N086/: American National Standard Institute (ANSI) does not set up the standards, but it only whaches over the process, through which the standards are accepted. ANSI has to notice i -f a standard dra-ft is acceptable by most wide par-t of industry. Only such standard could be adopted and used in industry and other institutions. ANSI adopts a standard as a national standard when it is acceptable by most companies and organisations. ANSI consists of by several committees. So the ANSI X3 is the standards development committee •for in-formation processing and has about 3i3 commi ttees, each wit)^ about 15 tq 80 members /B0N086/. One of them is X3H3 tehnical committee, which is responsible for computar graphics standards. X3H3 committee consists o-f 6 subcommi ttees, which is showed by -figure 2 /STRA86/. More than 1130 part icipants , represent i ng about 8(3 companies (CalComp, Control Data, DEC, HP, Honeywel 1 , IBM, Intel , Tektronix, TI , etc), attend X3H3 meetings. - the -first: New Work Item proposal (NUII); discussion about new project is started, when subcommittee (like SC21) or a neniber body (like ANSI) makes a propasal. Representati ves o-f di-f-ferent countries decide i -f they accept the de-finition o-f the work item and i -f the1 nork is continuing on this pt-oposal . This stage can take 5 to 8 months. the second: Working Dra-ft (WD) ; . document could be in this stage 6 to 18 rnonths; the third: Draft Proposed (DP); this stage can take 12 to 14 months; the -fourth: Draft International Standard (DI5) ; document could take placs; in this stage -for 9 to 12 months; the final: International Standard (IS). 20 Evolution proceas o-f an ANSI etandard Evolution stages o-f an ANBI standard di-f-fer •from stages o-f an ISO standard and are •f ollowing: the -firsts Standing Document 3 (SD-3) is an initial proposal which can take no less than 6 months; - the second: hlorking Dra-ftsj X3H3 prepares a series o-f working dra-fts that are circulated among X3H3 members. This stage typically takes several years. the third: Dra-ft Proposed American National Standard (dp ANS)j this stage takes 6 to 10 months; the -fourth: Public Review; document could be in this stage 8 months or more, which depends on the number o-f public reviews. At least two public reviews are required by X3H3. the final: months. Final Approval takes 6 to 9 The current stage o-f graphical standards under ISO and ANSI is shown by table 1 /B0N086, SELEB7/. Table 1. The stage o-f graphical standards project Project GKS GKS Fortran GKS Pascal GKS Ada GKS C GKS-3D ISO status IS 7942 published in Avgust, 1985. Known as ISO DIS 8651/1. DIS ballot closed in Avgust, 1986. Known as ISO DIS 8651/2. DIS Ballot closed in August, 1986. Known as ISO DP 8651/3. Second DP ballot closed in Apri 1 , 1986. Not yet an ISO standard language. WD available now (SC21/N669). Known as ISO DP 8805. Second DP ballot closed in March 1986. ANSI status ANS X3.123-1985. Published in October, 1985. ANS X3.124.1-1985. Published in October 1985. ANS X3.124.2-1987. Public review closed in May, 1987. ANS ;<3. 124.3-198x. Public review closed in 1986. ANS X3.124.2-198x. Public revien will be -finished by Octaber,1987. Public review -finished in 1986. 6KS-3D Fortran Known as ISO DP 8806. Public rewiew -finished 1986. GKS-3D Pascal Not yet available. PHIGS WD -finished in 1986. ANS X3.144-198x. Second public review -finished in 19B7. PHIGS Fortran WD available. (SC21/N667) Second public review -finished in 1987. PHIGS Ada WD available. (SC21/N819) Public review -finished in 1986. CGM (•former VDM) IS 8632 published in 1987. ANS X3.122-1986 published in 1986. CGI (•former VDI) DP began in 1986. ANS X3.161-198x. Public review -finished in Juny, 1987, 4. THE PLACE OF THE GRAPHICAL STANDARDS IN THE GRAPHICAL SYSTEM Six known standards (suggested or accepted) could be devided into three chategories /DEUS84/: 1. Core, GKS, and PHISS represent an aplication programming inter-face (API). This API standards are usually implemented as a set o-f the external procedures and an application programmer could link them into his application code. 2. 3. IBES and CGM are used by transfering storing the graphical i n-f ormation. and CGI represent inter-f ace. graphical device 21 APPLICATION PROGRAM LANGUAGE BINDINGS OEVICE AND LANGUAGE INDEPENDENT GRAPHICAL PACKAGE COMPUTER GRAPHICS INTEFFACE OEVICE DRIVER CURRENT GRAPHICAL DEVICE FUTURE GRAPHICAL DEVICE METAFILE READER 7.— METAFILE VVRITER metafile Figure 3. Interfaces o-f the graphical system These tree classes could help us by de-fining common features o-f current and -future standards with regard on per-f ormance, price, and use-fulness o-f graphical software and hardware. A comparison and valuation o-f the graphical standards is not easy, because the majority o-f them are very comple;< and because they are comming -frotn di-f-ferent areas. Figure 3 gives a review a-f the standard graphical i nter-f aces. The most important part o-f each language and device independent graphical system is a member o-f the GKS standards •family. These are GKS (Sraphical Kernel System) or its 3D extension or a standard for dynan>ical manipulati on with graphical data structures PHIGS (Programmer's Hierarchical Inter-act i ve Graphics System) . Functions o-f graphical system are e: