ÿWPC ¶Oªž6ÉkcÐþ¬Æ>Ƥ%tWÍ.‡Ð5ûǬ¿wÀ†ò½‰<[þDtÈTÍßã6kV8'§gxÞ» P¨ÇµÔMˆ†Ö¢Ë °iì’~#ÆܵÓ?#±•*/YÕj6•öÑt(Di7œꄾ\õ€8q ?Ã8#£vØÊå4Ï懙j:òÛàùÊl0ä øµ‹S Ö=\ÿñÂ4¨dvwÍ/>ºÃÂ?4™ãÒ½ó|Ù4¨üi¦e¥jÒŽdÖµKŒ w{ýö#í…¤ÕÞäuDöتÓÔÏ)ñ : ™è:§SŸ]9û–ÆsÙôÃÓž*£ÌieHžüÐtÒ dõ”‹Dí5¾ül–ÎE¡‹Fgþv\‘x$a`dˆaú“€S]n“ƒ2¦… ïèɶdHçßVþ®¾  ç´”(|Ëoªò`XÒY— ÝœÎé'ÇÇ[“OæùT£D,¥=Ž¹Œáªº2‹ E¹†â8F "Ü"•5ø{p?ñÏióFÔ8"„wVô½ƒ+^D³2QÄ5ULæ %2 0Ó8 0î LùEnMULP eœ « w@H 1mˆ Šõ 0ï 1mn Û«áULŒ4Ø„ìpULULË ULULiUVµ   ULULcUV¯2öàp`4New Century Schoolbookô(«ÕD$¡¡ÑÀXÑÑÀXÑÒ(°ÒÒ(°ÒÓ  ÓÔ€ô¾×ôôôÔÔ€ô¾×ôôô¾×ÔÓ3óÓÔ  ÔÑ  Ñ(Æ2Ã1$§§Ý ƒÕD!ÝÑÀ°ÑÑÀ°ÑÒ(°ÒÒ(°ÒÓ  ÓÔ€ô¾×ôôôÔÔ€ô¾×ôôô¾×ÔÓ3óÓÔ  ÔÑ  ÑÝ  Ýd:\office\wpwin\template\x3j3-rpt.wpt«<þ6X9`(CourierXÛxþ6X@ÉüKX@<þ6X9`(CourierôŒdþ6X@ÉüKô@+˜à`4New Century SchoolbookôÓe˜à Pâ’'CôP+˜à`4New Century Schoolbook ¡Ý ƒÃ1'ÝÒ(°ÒÒ(°ÒÓ  ÓÔ€ô¾×ôôôÔÔ€ô¾×ôôô¾×ÔÓ3óÓÔ  ÔÝ  Ýò òà€,°"àISO/IEC€€JTC1/SC22/WG5-òòN12›__ó óóóˆÐ À ÐÓ °°§,°œXÓà€œœ=°"àÔ‡»J»ôô¾×Ôò òX3J3€/€97-116Ô#†ô¾×ô»»J!#Ôó óˆÐ Ì  ÐX3J3€ANSI€Fortran€Standards€Committeeà€mm=°"àCraig€T.€DedoˆÐ È ÐGeneral€Planning€Considerations€for€Fortran€2000à€]]:°"àJanuary€24,€1997ˆÐ …Å Ðà€JJJ°"àPage€Ú  Ú1Ú  Ú€of€4Ô‡ô¾×ôôô¾×Ô ÷EøùHúûUüÊýeþÿÊÊ' e37=CIQYag³³1.a.i.(1)(a)(i)1)a) €Ó °°§,°œXÓòòÔ.<ƒ ,D:\X3J3\140\97-116.WPD<ÔD:\X3J3\140\97-116.WPDÔ/  Ôóó(Ç34q$©©Ý ƒÕD!ÝÑÀ°ÑÑÀ°ÑÒ(°ÒÒ(°ÒÓ  ÓÔ€ô¾×ôôôÔÔ€ô¾×ôôô¾×ÔÓ3óÓÔ  ÔÑ  ÑÝ  Ý# e37=CIQYag­­1.a.i.(1)(a)(i)1)a)È ¡Ý ƒÃ1'ÝÒ(°ÒÒ(°ÒÓ  ÓÔ€ô¾×ôôôÔÔ€ô¾×ôôô¾×ÔÓ3óÓÔ  ÔÝ  Ýò òà€,°"àISO/IEC€€JTC1/SC22/WG5-òòN12›__ó óóóˆÐ À ÐÓ °°§,°œXÓà€œœ=°"àÔ‡»J»ôô¾×Ôò òX3J3€/€97-116Ô#†ô¾×ô»»J!#Ôó óˆÐ Ì  ÐX3J3€ANSI€Fortran€Standards€Committeeà€mm=°"àCraig€T.€DedoˆÐ È ÐGeneral€Planning€Considerations€for€Fortran€2000à€]]:°"àJanuary€24,€1997ˆÐ …Å Ðà€JJJ°"àPage€Ú  Ú1Ú  Ú€of€4Ô‡ô¾×ôôô¾×Ô2öàp`4New Century Schoolbook³'" Ú1ÚGeneral Planning Considerations for Fortran 2000ReportÍ,' ' X3J3/97-10c$Craig T. Dedo$Craig T. Dedo .   /4 Fortran 2000 Standard+None     "1 302Original 3|x2öàp`4New Century Schoolbook+˜à`4New Century SchoolbookÈ2öàp`4New Century Schoolbook+˜à`4New Century Schoolbook9LÔXp>New Century Schoolbook BoldÈH2öàp`4New Century Schoolbook+˜à`4New Century Schoolbook9LÔXp>New Century Schoolbook BoldÝ ƒÕD!ÝÑÀ°ÑÑÀ°ÑÒ(°ÒÒ(°ÒÓ  ÓÔ€ô¾×ôôôÔÔ€ô¾×ôôô¾×ÔÓ3óÓÔ  ÔÑ  ÑÝ  Ýñýñ›ñýññýñ›ñýñÓC€ e'#CÓÖ€ ÿÿÖñýñœñýññýñœÖ€ ÿÿÖ›ñýññýñÖ€ÿÿÖñýññüñ›ñüñÓ  ÓÓr^°°Ü4` Œ ¸ ä <h”ÀìDpœÈô Lx¤Ð ü!(#T$€%¬&°œXrӜԇô¾×ôôô¾×ÔñüñÔ% € Ôñüññýñ›ñýññññÿñÔ% € ÔñÿñññññÔ% € Ôññ›à@ìàò òòòGeneral€Planning€Considerations€for€Fortran€2000óóˆÐ Ñ Ðó óÌà@SS"ìàby€Craig€T.€DedoˆÌà@CC"ìàJanuary€24,€1997ˆÌÌà T àÔ#†ô¾×ôôô¾× #ÔThis€paper€contains€several€loosely€related€opinions€and€concerns€which€are€related€to€theÐ Ñ  Ðplanning€process€for€Fortran€2000.€€I€delivered€some€of€these€remarks€at€the€May€1996€X3J3Ïmeeting.€€Other€parts€are€more€recent€concerns€which€I€have.ÌÌÔ& ô Ôò òÔ2 eÔÚ  Ú1Ú  Ú.Ô3  Ôà0 T àParadigm€Shifts€/€Basic€Assumptionsó óÐÅô T°"T°" Ðà T àAs€we€prepare€to€develop€plans€for€Fortran€2000,€we€should€keep€in€mind€the€concept€of€theÏparadigm€shift.€€Paradigm€shifts€are€central€to€strategic€planning€in€all€kinds€of€organizations.€ÏÔ'ôÅhÔ"Paradigm€shift"€means€that€the€mental€models€for€understanding€the€business€environment€whichÏused€to€work€no€longer€work,€due€to€technology€advances,€changes€in€customer€preferences,€andÏother€changes€in€the€environment.€€New€paradigms€are€necessary.€€Making€the€transition€to€newÏparadigms€is€always€a€difficult€and€uncomfortable€process.ÌÌà T àWe€may€be€experiencing€a€paradigm€shift€and€should€keep€this€in€mind€when€we€designÏFortran€2000.€€I€am€specifically€thinking€of€two€assumptions€which€may€no€longer€be€valid.€€Ìà T à1.à0 € àWe€can€safely€ignore€interactions€with€the€operating€system.Ð'V€°"€°" Ðà T à2.à0 € àWe€do€not€need€to€specify€anything€about€processor€character€sets.€€Ð䀰"€°" ÐIn€both€cases,€our€previous€assumptions€may€no€longer€be€valid.€€ÌÌà T àThere€may€be€other€assumptions€which€are€no€longer€valid.€€We€need€to€critically€examine€theÏcore€assumptions€we€have€about€how€the€world€works€and€what€the€market€values.€€ÌÌÔ& ô Ôò òÔ2 eÔÚ  Ú2Ú  Ú.Ô3  Ôà0 T àPlanning€Horizonó óÐRT°"T°" Ðà T àPlease€keep€in€mind€that€we€are€designing€a€language€for€the€years€2001-2012.€€We€need€toÏanticipate€the€environment€that€Fortran€will€be€in€and€the€needs€that€users€of€Fortran€will€haveÏÔ'ôRè Ôduring€that€time.€ÌÌò òÔ2 eÔÚ  Ú3Ú  Ú.Ô3  Ôà0 T àStrategic€Direction€/€Marketing€Considerationsó óÐ 2T°"T°" Ðà T àAlthough€some€persons€may€disagree,€we€need€to€be€aware€of€and€respond€to€marketing€andÏstrategic€direction€considerations€as€well€as€technical€issues.€€The€often„expressed€attitude€of,€"WeÏare€a€technical€committee,€not€a€marketing€committee"€simply€will€not€work€in€any€organizationÏwhich€has€strategic€direction€responsibilities.ÌÌà T àI€am€very€concerned€about€the€image€which€Fortran€has€and€the€continuing€lack€of€popularityÏof€Fortran,€particularly€in€the€academic€community.€€Six€years€after€the€release€of€the€Fortran€90Ïstandard,€many€persons€still€consider€Fortran€to€be€an€archaic€dinosaur€with€little€relevance€to€theÏworld€of€modern€computing.€€In€many€places€in€the€academic€world€the€teachers€still€say,€"FortranÏis€outdated.€€You€don't€need€to€learn€it."€€While€there€may€be€little€we€can€do€directly€to€combatÏsuch€backward€attitudes,€there€is€much€that€we€can€do€to€avoid€aggravating€the€situation.€€ÌÌà T àWe€especially€need€to€be€open€to€modern€programming€concepts€and€features€which€haveÏproven€their€worth.€€We€also€should€strenuously€avoid€mistakes€like€those€made€with€Fortran€77,Ïwhere€the€designers€refused€to€admit€into€the€language€many€modern€features€such€as€recursion,Ïderived€types,€long€variable€names,€a€complete€set€of€control€constructs,€and€pointers.Ìâ âÐ ,¿#0 Ðà T àIn€summary,€we€need€to€constantly€ask€ourselves€where€we€want€Fortran€to€go€in€the€next€fewÏâ âdecades.ÌÌò òÔ2 eÔÚ  Ú4Ú  Ú.Ô3  Ôà0 T àTarget€Marketsó óÐ 7T°"T°" Ðà T àWhile€I€have€often€stated€that€I€believe€that€Fortran€should€be€a€general€purpose€language€andÏuseful€for€teaching€programming,€I€do€not€believe€that€we€should€abandon€or€neglect€the€needs€ofÏour€core€market,€which€is€high€performance€numeric€computation,€particularly€in€science€andÏengineering.€€One€of€the€basic€principles€of€strategic€planning€is€that€businesses€should€always€actÏto€serve€the€needs€of€their€core€market.€ÌÌà T àWe€should€not€give€up€the€markets€of€general€purpose€application€development€and€use€ofÏFortran€as€a€teaching€language.€€Although€C€and€C++€dominate€these€markets€and€there€is€stiffÏcompetition€from€Microsoft's€Visual€Basic€in€the€Intel„based€PC€market,€I€believe€that€there€is€stillÏopportunity€for€Fortran€to€capture€a€substantial€share€of€these€markets€if€we€make€the€rightÏdecisions€and€provide€users€of€Fortran€with€the€tools€that€they€want€and€need.ÌÌà T àAnother€area€which€needs€attention€is€giving€users€what€they€want.€€Very€often,€users€haveÏrequested€a€standardized€version€of€a€feature€for€which€there€are€differing€and€incompatible€vendorÏextensions.€€In€many€of€these€cases,€a€large€proportion€of€users€consider€such€features€to€be€highlyÏdesirable,€necessary,€or€even€vital.€€Right€now,€a€good€example€is€support€for€standardizing€aÏcommand€line€interface.€€There€are€many€other€features€which€also€fall€into€this€category.€ÏAlthough€many€such€features€involve€close€interaction€with€the€operating€system€or€similarÏdifficulties,€we€really€should€be€in€the€business€of€figuring€out€how€to€satisfy€the€desires€of€ourÏusers,€not€making€excuses€why€it€cannot€or€should€not€be€done.€€As€the€old€Marshall€Fields€sloganÏgoes,€"Give€the€lady€what€she€wants!"ÌÌà T àOne€other€concern€is€human€productivity.€€This€is€simply€how€much€software€functionality,€suchÏas€function€points€or€feature€points,€which€can€be€produced€with€a€given€number€of€direct€laborÏhours.€€I€believe€that€Fortran€has€a€major€advantage€in€this€area,€due€largely€to€its€generallyÏstraightforward€and€easy€to€understand€grammar€and€syntax.€€We€should€build€on€this€advantage.€ÏWe€can€do€this€by€accepting€into€Fortran€2000€powerful€concepts€and€tools€which€allow€theÏapplication€developer€to€more€closely€model€the€real€world,€without€a€lot€of€effort€and€potential€forÏmistakes.€€It€also€is€desirable€to€design€new€features€in€a€way€which€is€easy€to€understand.€€AndÏwe€can€continue€to€eliminate€gratuitous€irregularities.ÌÌà T àTo€summarize,€I€believe€that€the€target€markets€which€Fortran€should€serve€should€be€these:Ìà T à1.à0 € àà0¬€°"€°"àà € àRobust€application€development€in€numeric,€scientific€and€engineering€applications€withÐ e#”$ Ðparticular€emphasis€on€high„performance€numeric€computation.Ð ¬°"¬°" Ðà T à2.à0 € àà0¬€°"€°"àà € àBe€second€to€none€as€the€language€of€choice€for€general€purpose€application€developmentÐ ß$& Ðand€teaching.Ð ¬°"¬°" Ðà T à3.à0 € àà0¬€°"€°"àà € àUser€demand,€i.e.,€give€the€customers€what€they€want.ÐY&ˆ(¬°"¬°" Ðà T à4.à0 € àà0¬€°"€°"àà € àIncreased€human€productivity,€i.e.,€software€functionality€per€direct€labor€hour.Ð'E)¬°"¬°" ÐÌÔ& ô Ôò òÔ2 eÔÚ  Ú5Ú  Ú.Ô3  Ôà0 T àValue€of€Application€Programmer€Interfaceó óÐ(¿+T°"T°" Ðà T àI€believe€that€as€we€develop€new€features,€we€need€to€keep€in€mind€the€value€of€a€high€qualityÏinterface€for€the€application€programmer.€€This€means€syntax€constructs€which€are€straightforward,ÏÔ'ô($Ôintuitive,€and€easy€to€use.€€Following€is€a€quote€from€the€mathematician€and€philosopher€AlfredÏNorth€Whitehead€on€the€value€of€a€good€notation:Ìâ âÐ A,p#0 Ðà8 T àBy€relieving€the€brain€of€all€unnecessary€work,€a€good€notation€sets€it€free€to€concentrateÏâ âon€more€advanced€problems,€and€in€effect€increases€the€mental€power€of€the€race.€€BeforeÏthe€introduction€of€the€Arabic€notation,€multiplication€was€difficult,€and€the€division€evenÏof€integers€called€into€play€the€highest€mathematical€faculties.€€Probably€nothing€in€theÏmodern€world€would€have€more€astonished€a€Greek€mathematician€than€to€learn€that€...Ïa€large€proportion€of€the€population€of€Western€Europe€could€perform€the€operation€ofÏdivision€for€the€largest€numbers.€€This€fact€would€have€seemed€to€him€a€sheer€impossibilityÏ...€Our€modern€power€of€easy€reckoning€with€decimal€fractions€is€the€almost€miraculousÏresult€of€the€gradual€discovery€of€a€perfect€notation.Ð T„!T„! Ðà€¡¡1°"à--€Alfred€North€WhiteheadˆÌÌà T àThis€principle€applies€to€computer€programming€languages€as€much€as€it€does€to€any€other€areaÏof€human€endeavor.€€In€my€experience,€a€lot€of€time€and€effort€is€wasted€fighting€the€limitationsÏof€the€application's€programming€language€rather€than€solving€the€problems€that€the€applicationÏwas€designed€to€solve.€€The€development€of€user„friendly€syntax€for€new€features€could€help€a€lotÏtoward€making€Fortran€a€far€more€productive€language€than€previously.€€This,€in€turn,€could€helpÏFortran€increase€its€popularity.ÌÌò òÔ2 eÔÚ  Ú6Ú  Ú.Ô3  Ôà0 T àGround€Rules€for€Reconsideration€of€Previously€Rejected€Featuresó óÐJ T°"T°" Ðà T àRecently,€there€has€been€some€controversy€on€what€the€ground€rules€should€be€for€considerationÏof€features€which€were€considered€and€rejected€previously.€€Unfortunately,€there€appears€to€be€aÏsignificant€fraction€of€committee€members€who,€for€whatever€reason,€believe€that€once€a€feature€hasÏbeen€rejected€once,€it€should€never€be€considered€again.ÌÌà T àI€believe€that€such€an€attitude€is€not€in€the€best€interests€of€Fortran,€nor€in€the€best€interestsÏof€any€of€the€Fortran€Standards€Committee's€constituencies.€€A€feature€may€be€rejected€at€anyÏparticular€time,€for€a€wide€variety€of€reasons,€including€a€heavy€schedule€of€more€important€items,Ïdifficulty€of€implementation,€or€lack€of€significant€user€interest.€€However,€in€the€area€ofÏprogramming€languages,€where€there€is€rapid€technological€progress,€conditions€changeÏdramatically€from€one€release€of€the€International€Standard€to€the€next.€€A€feature€which€may€beÏdifficult€or€expensive€to€implement€in€a€particular€year€may€become€much€less€expensive€or€farÏeasier€to€implement€by€the€time€the€next€version€of€the€standard€will€be€ready,€simply€due€toÏtechnological€progress.ÌÌà T àIt€is€generally€the€rule,€rather€than€the€exception,€that€major€progress€in€human€endeavorsÏrequires€several€attempts€before€they€are€successful,€rather€than€just€one€or€two€attempts.€€In€aÏdebate€on€comp.lang.fortran€around€January€20,€1997€on€a€proposal€for€command€line€argumentsÏand€environmental€variables,€I€presented€an€analogy€of€14€events€from€world€history,€from€the€timeÏof€Moses€to€the€present,€where€repeated€attempts€were€required€in€order€to€make€progress.€€EachÏof€these€events€had€a€major€impact€on€the€quality€of€life€today.€€In€responding€to€criticism€of€myÏanalogy,€Glen€Clark€ñýñœñýñ(Glen€Clark,€posting€on€comp.lang.fortran,€dated€22€Jan€1997)€madeñýñ›ñýñ€theÏfollowing€observations:ÌñýñœñýñÌñýñ›ñýñà8 T àIf€we're€going€to€resort€to€straw€dogs,€we're€never€going€to€get€anywhere.€Craig€Dedo€wasÏnot€trying€to€seriously€discuss€biblical€history€and€everyone€who€read€it€knows€that.€It€wasÏan€illustration€called€an€allegory€which€makes€a€point€by€drawing€a€parallel.€€I€thought€heÏmade€his€point€well.Ð T„!T„! ÐÌà8 T àThe€position€of€the€Committee€appears€to€be€that€once€an€issue€has€been€decided€against,Ïthat€is€the€way€it€should€stand€for€eternity.€And€that€anyone€so€brash€as€to€suggest€aÐ þ,-$1 Ðsecond€iteration€of€discussion€is€a€pariah.€Craig's€point€was€that€repeated€attempts€is€notÏthe€exception,€but€rather€is€the€rule€for€many€successful€human€endeavors.Ð T„!T„! ÐÌà8 T àI€think€the€belief€that€many€have€is€that€there€should€be€some€kind€of€feedback€loopÏbetween€the€users€and€the€Committee.€While€being€mindful€that€committee€members€areÏpeople€too€and€that€their€time€is€€not€unlimited€for€the€committee,€there€is€a€sense€thatÏthere€should€€be€some€semblance€of€an€iterative€process,€even€if€the€number€of€iterationsÏis€low.€If€feedback€is€neither€welcome€not€possible,€what€€is€the€purpose€of€a€DRAFTÏSTANDARD?€In€the€absence€of€an€iterative€feedback€process,€the€final€will€not€differ€fromÏthe€draft,€so€why€€call€it€a€draft?Ð T„!T„! ÐÌà8 T àFurther,€there€is€tremendous€legal€liability€in€being€part€of€an€"industry€standards€group".ÏFederal€antitrust€laws€formally€require€a€very€visible€and€responsive€procedure.€For€moreÏinformation,€see€"US€vs€ñýñœñýññýñœñýñHydrolevelñýñ›ñýññýñ›ñýñ".€I€don't€have€a€citation€handy,€but€any€patent€or€IPÏattorney€will€know€it€by€heart€from€the€case€title.Ð T„!T„! Ðñýñœñýñ[End€of€quotation€from€Glen€Clark]Ìñýñ›ñýñÌñýñœñýñà T àIt€is€ñúñ›ñúñmy€hope€that€committee€members€will€take€his€words€to€heart€and€develop€a€sound€ñúñprocessÏñúññùññôñmethodÏñôññùñfor€reconsidering€previously€ñùñ›ñùññöñrejected€œfeaturesñöññþñ€and€develop€closer€communications€with€the€generalÏpublicñþññöñ.›€ñöññ÷ññöñ€We€also€need€to€ñöññ÷ññøñ€ñøññüñ›ñüññûñÌœñûññýñ›ñýññÿññüñÔ% € Ôñüññÿñ›