SCFMpattern paragraph style default tab maximum page size arrow size slide show script template name circle old arc rectangle old text block 3D viewer polygon old group rounded rect new page image text block arc note page tab count token table current item numberaudio clip old templates encapsulated PS current page numbertemplates version string page image back pointer font name !inventor 3D object "live video frame #Japanese text block$version code %group &starting script 'ending script (I18N text block )page table *linked text table +item id table $ (This file created by IRIS Showcase 3.4dA A`AABeB^BeBeB^BeAABeBh`BeB8B8B<B<B\B\B]PB\ BBL!BDBD"B8B8$BBLp*BBD,-/B0B00A BBo2BB ,Helvetica-BoldHelvetica-Bold $ Helvetica Helvetica $ ZapfDingbatsSymbol < Helvetica-BoldObliqueHelvetica-BoldOblique\_Q dkT BBBBD: B@BB?×?C?D: B?C?×?+D,PAD,NAF@D9A@@o?x??D,NAF@?  Page   %[BDBDD;D?YBhDBhDCD?_8Ymno nameҨYmHYXYCY.YYXXXXXXXqX\XGX2XXWWWWWWW]W VVVHUSRHQONMKJQHHHGG\G FF~FAFEEEBDDDQDCCaCBBkBAArA @@{@)???.>>>*==X<<}<;;-::H99M877+66 55j544j33i22%0/.-,+*)('&%t$"!!W! h x8f!Mt'P L?o&<j0b3SmXC.Y_YJY5Y Y XXXXXXXxXcXNX9X$XWWWWWWWmW$VVV\UeT'RQiPNMLJIbHH;GGwG&FFFTFEEE\EDDmDCC}C+BBB4AAA<@@@E???J>>>G=={=<<3;;S::o99{8847n66<5556444332Z1=0/. -+*)('&%$#;"!s!# 0I ~8d@g"`S=Uz@o> e.t_J5  YfYQY<Y'YXXXXXXXXjXUX@X+XXWWWWWWxW?VVVqUTS>QPOOMLK,IHH_GGG?FFFgF+EEEoE%DDD3CCCCBBBMAAATA@@]@ ??c?>>c>==1<33=210w/n.g-`,W+N*G)D(9'6&%$#"]!!; K\My3Y |8w3f'T nTzOx@ {fQ<'a\rbW_isgj$$ !#%$%%&%'&.5=C-2,**5fNNNNNNLTRTXhhZ   oillntts%$" !!"""###$'15<),*)+=p][[]][Y]WYYmi\   ttsv~|')(#&''((())**2<BM3:(,,,Qx`^^^^^^^ZZdzv%  ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~c~2c#~1b'p~1ac~0a#~/`'~/_c~._#~.^'~-]̂ c~,]# :~,\' 7~+[̨ ct~+[## t~*Z'' [~*Y c~)Y## ~(X'' ~(W c~'VU ~'VH ~&U c~&T# ~%T'pp p~%Stc~&RUU~%RHHp~%ic~&Q:~%Qp7~%kct~&k~&j~%lcc~'k~'k~&l~)k~(k~(~~~~^ ~)U⨨c~(SU::UU~&RϞH77HH~%N!c~"LU!#t~ KpH!'[~It,c~Ht.U~G[.H~Etc c~Dt :~C[ 7~At cct~@# :UU#~@' 7HppH'p~>c'c~>: U$t:t~=7 H"Ϟ[7[~;c0t c~;:U.U:~:7H-H7~ 9ct; c~ 8tt8t~ 7[[7[ p~ 6cA ct~ 5#:?t#~ 5'7>[ '~4cctH c~3Ft #~3ppD[ '~1cN c~1U:Lt ~0H7K[ ~/ttS c~/U#Q# ~.H'Q' ~-tXt c~-UW# ~,HpV' N/+̂]t cM>.+#\U M/*'pZH L.)̨b cK%.)#` #KC.('`p 'Jc.'f ctI.':#e tH44.&7'dp [H[[-~t cG[-%ttit Fu44-%[[h[ F[[-~ cE-~: E4S-~7 C[m-~ ctC>,~ B444-~ B[[[,~ cA,~ :ACuC4u,~ 7@cc[,~tc?%,~t ?444,~[ >[[[+~ c=>+~ U=uS44+~ H=m[[+~tc<+~t <4c4C+~[ :[u[c*~ct:[*~ 9C44*~ p9c[[*~ tc8[*~t #8SC4c*~[ '8mc[u*~!c6>)~! 7444*~ 5[[[)~"c5%%)~": 544)~"74[[)~$tct4)~# 44u44)~# 2[[[(~%ct2[[(~% 1u44(~$1[[(~&ct1 (~&#1S S4S(~%'0m m[m(~(ct/ ['~'0S 44(~'p.m [['~)tct. '~(. 44u'~(- [['~*ct- '~*-C 44'~)+c [[&~+c [ >&~+! 44&~* ի [[&~,c/&~,/S4C&~+ m[c#~-cz>#~-#/44#~-'[["~.c[z"~.#44S"~.'˫[[m"~9c[%"~/ u44u"~/ [["~9c>"~9uCc44"~9cu[[!~9c!~9C4S44!~9c[m[[!~9c!~9444!~9[[[!~9c!~9444u!~9[[[!'rUUUrr%:Ur r%rr c!'x[ [[xx.C[x x.x x 444u!&ը 䨁[ ՚ Ŏg[t 䨁[ [0Dg ը [[[!$r U : r %: r % : r U r rc%%:!$x [ C x . C x . C x[ x x44..C4u!#D 0t0:[g0Ot:0Ogt 0[D0[:0:[[D0gg0Dt0D[!#%r %r %r r %r r %r c!# .x .x  .x x .x  x .x 444u!":0Og 0:g [0Og:0[ 0:g O0O :0Dg [[000[!" U: r crrr!" [ C x 44xxx4u!!:0O 0[ O0: :0t0[ O0 0[[[000[!! U  r r :crrr!! [  x x C44xxx4u!!D0[ 0 [0[ [00 0 0t[[000[!!% r  r rcrrr!!. x   x  x44xxx4u!!0g 0 0[ [0O 0 0 0[[000[!!r  U  crrr!!x  [    44xxx4u!!0 0 0 [0[ 0 0 0[[000[!!r r : rcrrr!!x  x C  x 44xxx4u! 0 0 0 0t 0 0 0[[000[!!  r  rcrrr!    x  x 44xxx4u! 0 0 0 0 0 0 0[[000[!! r r r crrr!    x x x 44xxx4u! 0 0 0 0 0 0 0[[000[!! rr r r rcrrr!  x  x x x x44xxx4u! 0 0 0 0 0 0 0[[000[!! rr r r rcrrr!  x  x x x x44xxx4u! 0 0 0 0 0 0 0[[000[!! rr r % rcrrr!  x  x x . x44xxx4u! 0 0 0 0 0 g0 0[[000[!! rr r  rcrrr!  x  x x  x44xxx4u! 0 0 0 0 0 [0: 0[[000[! :rr:r%rrr:r:rr%rcrrr!  Cx x Cx .x xx Cx Cxx .x44xxx4u! 0Ot 00Ot 0Og 00D[00Ot [0Ot0Og[[000[!!)rScrrr! )x>44xxx4u! ŶՎŨ[[000[!~9crrr!~944xxx4u!~9[[000[!~9cr!~944x4u!~9[[000[!~9c%!~944 .4u!~9[[:0Og0:[0:[!~9c!~9444u!~9[[[!~9c!~9444u!~9[[[!~9c!~9444u!~9[[[!~9c!~9444u!~9[[[!~9c!~9444u!~9[[[!@z[%>[Gc!Dȁuc4cuK444u!?Ձu[uՁF[[[!2z[[>9[=c!5ȱuC74cu>444u!1Մc7[u<[[[!'[[ [z[>[5c!)Ȋuu 4cuȁuC4cu7444u!&ս [uՁՁc[u5[[[![ [9[%%[/c!Ȋuu 4cuu>ȞuC 4Cu0444u! [u7˫c [c/[[[!%[[zN[[>+c!C4cuuQuc 4c+444u!Օc[uLիu [u*[[[![`[ [&c!cu 4Cu'444u!^ս [c%[[[!~z% %"c!~ߞS 4c"444u!~ⴕm [u![[[!~ [ c!~ u 4C444u!~ ˫ [c[[[!~z% [c!~ߞS 4c444u!~ⴕm [u[[[!~% %c!~C4C444u!~c[c[[[!~> %c!~c4C444u!~u[c[[[!~ >c!~4S444u!~[m[[[!~z [c!~c4S444u!~u[m[[[!~!z % c!~!S4 444u!~ m[ [[[!~$> cc!~$C4c 444u!~#c[u [[[!~& > t!~&c4C t444u!~%u[c [[[[!~(%%&!~)4 t444u!~(Օ[[[[[!~+[%$!~+C4%444u!~*c[#[[[!~-[%"!~-C4#444u!~,c[![[[!~/[> !~/C4!444u!~.c[[[[!~1[!~1C4C444u!~0c[c[[[!~3[!~34u444u!~2[[[[!~5>!~5u4444u!~4[[[[!~6!~7S4C444u!~6m[c[[[!~8[!~84u444u!~7[[[[!~:>!~:u4444u!~9[[[[!~;!~[c[[[!~@!~AS4444u!~@m[[[[!~B!~B4C444u!~A[c[[[!~C> !~DC4444u!~Cc[ [[[!~E !~Eu4S 444u!~D[m [[[!~Fz[ !~F4 444u!~F[ [[[!~H !~Hc4c 444u!~Gu[u[[[!~I>z!~I4C444u!~I[c[[[!~J!~KS4uC44u!~Jm[c[[!~L!~L4CC4u!~K[cc[!~M[!~NC44u!~Mc[[!~O!~Oc4c4u!~Nu[u[!~P>!~P4C4u!~P[c[!~Q>!~RC44u!~Qc[[!~S!~Su4c4u!~R[u[!~T[!~T4C4u!~T[c[!~2%!~2C44u![~020c[[!~2 !~2 u4c4u![~020[u[!~2 z!~2 4C4u![~020 [c[! s, ! r+ S44u![0 [:q0D[[D+0 m[ [! %orr+  ! .oxx+ 4S4u![0 go0+0 [m [! orr+z ! oxx+C4 4u![0 o0+0c[ [! r : r  rr r ! x C x  xx xu4 4u![00D0 000:t0 0 0 0D0 00:[0[ 0[ [! r  r  rrr [! x   x  xxx4C 4u![000 0 O000OD0 0 0 00O0 000[Ń0[c[!rr rUr    rrUrr[!xx x[x  xx[xxc4C44u![00 00D[000 0 0 0O0 00D0[00u[c[[!   rr>%!    xx4u4u![0 D0[:00:[0D0[:0 0[:0O 000:0[[0[[! U   rr % %[![   xx . .S4C ![0ń0[0ń[0 0[0 0[0O0 0g[0[0[ g0m[cի!   rr  %)     xx 4)[000ŁO0 0:0ŁO0 0ŁO 0O0 0:[[0[Ń0[(r %r  rrr r r:'x .x xxx x xCC4C([000 0[gD0[[0 0 000 0 0[0tD0c[c'rrrrr Ur%&xx  xxx[x4&[00[D000::00:0[O0[D00[O00D00DŊ00[0D[0[%  r:  r rr  [$  xC   x xx  u4C$[0 0 00:D00tO0 0 00 00 [00:0::0[ :0[c# U r%  r rr r: %[" [ x.  x xx  xC .S4C"[0 䶁:0 00D000Dg00 0 00 00 [0O0t0[ըg0m[c"~2> ~2u4S [~020[m~2"%z~2"4c[~020![u~2$[~2$C4u[~020#c[U~2([ [[ [~2(uc4Cu ~[2['⫁u[c ~{[~}ȁ~z~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ BDBDD;D@BD?×?C?D;D?C?×?+ CD`CDDD`@@???CD?  National Energy Research Scientific Computing CenterAn AcpAn>AvBSAb:?g???An>Av?  file::Talks/<CVA؋CVAPD*A؋?=???CVAP?   T dF.D#AD,NAF@D9A@@O?x??D,NAF@?  /B0A;An>AC A` ?M???An>A?  file::Talks/MAAMWorkshop.showcase`3B gDB gB$jHD97D?G???B gB$jH?    ?A View from the Front Lines:Computational Math in the   ??"Real"   ?World  ==?David B. Serafini  =?=Lawrence Berkeley National LabNERSC Division  =?=:CEACEA0rD(5A ?????CEA0r?  =<CVA؋CVAPD*A؋?;???CVAP?   1 dFID,PAD,NAF@D;A@@+?x??D,NAF@?  Page1  JAn AcpAn>AC A` ?)???An>A?  file::Talks/MAAMWorkshop.showcaseK<CVA؋CVAPD*A؋?'???CVAP?   10SB ZDHB ZC D<DH?#???B ZC ?   ?   ?The Computational Math World   ?pVBRCkBRB$jHD8jCk????BRB$jH?    Academia: mostly research, little development, no productionGovt. labs: some research, some development, little productionIndustry labs: little research, some development, some productionIndustry: very little research, some development, mostly production d FpD,PAD,NAF@D;A@@ ?x??D,NAF@?  Page1  qAn AcpAn>AC A` ????An>A?  file::Talks/MAAMWorkshop.showcaser<CVA؋CVAPD*A؋????CVAP?   1uBRDBRCɂED<D????BRCɂE?      ?Executive Summary   8zACπAB0D:Cπ????AB0?   l  Applied math in the real world   ?=?is   computational math   l  Comp'l math in the real world is a   ?`>?service  industry   l  Knowledge transfer should be bi-directional   l  Interdisciplinary is the name of the game   l  Software is the coin of the realm d F|D,PAD,NAF@D;A@@?x??D,NAF@?  Page1  }An AcpAn>AC A` ????An>A?  file::Talks/MAAMWorkshop.showcase~<CVA؋CVAPD*A؋ ????CVAP?   1BDBCŀD>D????BCŀ?   ?Executive Summary (cont.)"BCBB8DBC????BB8?  Writing software is the dominant part of comp'lmath in the real world.    Writing good software is much harder than grad.school leads you to believe.    The definition of good software is not what most people think it is.    BCnXBCnByCn????BCn?   l  -BxCBxCByC?y???BxC?   l,BxC9BxC'ByC9?w???BxC'?   l, dFbD,PAD,NAF@DAC A`????An>A?  file::Talks/MAAMWorkshop.showcased<CVA؋CVAPD*A؋????CVAP?   1jBRCBRC D<C????BRC ?   ?The    ?=?"Real"   ? WorldDmBbC,BbB$]VD9C,? ???BbB$]V?       l  Long term research is out of fashion       l  Most research jobs in industry are not in  separate research groups      G  look in the applications groups      G  exceptions: Boeing, Bell Labs, IBM       l  Not much   =?=traditional  analysis work is done   l  Analysis still needed for risk assessment for medium and long term R&D       G  "analyze first, then develop"     dFD,PAD,NAF@D;A@@G?x??D,NAF@?  Page1  An AcpAn>AC A` ?E???An>A?  file::Talks/MAAMWorkshop.schowcase<CVA؋CVAPD*A؋ ?C???CVAP?   1ADAC؀D<D?A???AC؀?      ?Audience Poll   *BCBBD:@C?)???BB?     l  students/recent students?000   l  in math/applied math depts?000   l  in computer science or engineering depts?000   l  never written a program longer than 500 lines?0 d FD,PAD,NAF`D;A@@?x??D,NAF`?  Page1  An AcpAn>AC A` ????An>A?  file::Talks/MAAMWorkshop.showcase<CVA؋CVAPD*A؋ ????CVAP?   10BD@BC׀D=@D@????BC׀?      ?Some Real World Comp'l Math Problems   ,ACABD:C?'???AB?  0   l  Airline and railroad scheduling00   l  Computational finance0   G  derivative pricing    0   G  risk assessment    0   G  portfolio management  0    G  econometric modeling    00   l  Computational biology00   l  Computational chemistry00   l  Factory automation, manufacturing modeling0 d FD,PAD,NAF`D;A@@?x??D,NAF`?  Page1An AcpAn>AC A` ????An>A?  file::Talks/MAAMWorkshop.showcase<CVA؋CVAPD*A؋ ????CVAP?   1BDBCD=D????BC?   ?Where I'm coming fromBCBB$D:C????BB$?   !l  Engineering and Applied Math degrees!!!   l  10 years in national labs (NASA, DOE)!!!   l  5 years in academia!!!   l  3 years in industry in research, development and ! production!! dFD,PAD,NAF@DAC A` ????An>A?  file::Talks/MAAMWorkshop.showcase<CVA؋CVAPD*A؋ ????CVAP?   1LBD@BC̀D=D@????BC̀?   ?What do computational mathematicians in the real world do?@   ?    ?hBCBBD:C????BB?   !l  Help application specialists solve problems!!!   l  Write code !!   G   find and reuse existing code when possible!!!   l  Run programs!!!   l  Write, write, write.!!   G   reports, proposals, software documentation! dF8D,PAD,NAF@D;A@@?x??D,NAF@?  Page1  9An AcpAn>AC A` ????An>A?  file::Talks/MAAMWorkshop.showcase:<CVA؋CVAPD*A؋ ????CVAP?   1;B DB CՀD=@D????B CՀ?   ?Computational math as a service industry8<BCBBD;C????BB?   !l  "Customers" have problems they need to solve.!!   l  "Customers" are experienced in their problems, ! but lack knowledge in some aspect of solution.!!   l  Identify important and unimportant factors.!!   G   "negotiating the problem"!!   l  Formulate the problem mathematically and ! computationally!!   G   communication between customers, c.m. and! computer scientists, other experts! dFD,PAD,NAF@D;A@@?x??D,NAF@?  Page1  An AcpAn>AC A` ????An>A?  file::Talks/MAAMWorkshop.showcase<CVA؋CVAPD*A؋ ????CVAP?   1BDBCD=D????BC?   ?Personal ExperiencesAC؀AB$D;@C؀????AB$?   l  written   =?=lots  of software (i.e. development) $   G  most of it written quickly, used a few times$$   G  a few took years of effort, used for years$$   G  some took moderate effort, used by many$$   G  most involved modifying or resuing existing code$$$   l  solved many problems (i.e. applications)$$   G  some "real" problems, some "test" problems$$   G  usually with some software I wrote$ dFD,PAD,NAF@D;A@@7?x??D,NAF@?  Page1  An AcpAn>AC A` ?5???An>A?  file::Talks/MAAMWorkshop.showcase<CVA؋CVAPD*A؋ ?3???CVAP?   1BDBC؀D=D?1???BC؀?   ?Experiences (cont.)0 BCBBD;C?/???BB?   l  performance analysis $   G  system benchmarking$$   G  code tuning$$ $   l  current position at NERSC:$$   G  1/2 time Applied Numerical Algorithms$$   G  1/2 time Future Technology$ dFD,PAD,NAF@D;A@@?x??D,NAF@?  Page1  An AcpAn>AC A` ????An>A?  file::Talks/MAAMWorkshop.showcase<CVA؋CVAPD*A؋ ????CVAP?   1BDBC؀D<D????BC؀?   ?ObservationsBCBBD;@C????BB?  *   l    In the real world, applied math == comp'l math.**   G  software, not papers, is the main result    ***   l  Current research funding trends provide little demand* for theoreticians outside of academia.    ***   l  The main task of comp'l math's is to write software.**   G  research => unsolved problem* => no code exists*  *  * dFD,PAD,NAF@D;A@@?x??D,NAF@?  Page1  An AcpAn>AC A` ????An>A?  file::Talks/MAAMWorkshop.showcase<CVA؋CVAPD*A؋ ????CVAP?   1BDBC؀D=D????BC؀?   ?Observations (cont.)B CڀB B,D:Cڀ????B B,?  *   l    Good software skills are more important than good* theorem-proving skills.    ***   l  The    only really meaningful measure of successful * software is usage.***   l  Success and quality are not equivalent.*  *  * dF0D,PAD,NAF@D;A@@%?x??D,NAF@?  Page1  1An AcpAn>ACA` ?#???An>A?  file::Talks/MAAMWjorkshop.showcase2<CVA؋CVAPD*A؋ ?!???CVAP?   1d3B CB BpD7C????B Bp?  *   l  Good quality software is:**   G  easy to use,   ?`>?maintain   and   =?=extend  **   G  robust (flexible and fault tolerant)**   G  solves a problem in an   ?`>?acceptable  amount of * time and resources**   l  Good interface design is a very important factor* in software success (i.e. usage)    **   G  user interfaces**   G  programmer interfaces*4ADACހD<@D????ACހ?   ?Software quality issues dFD,PAD,NAF@D;A@@?x??D,NAF@?  Page1  An AcpAn>AC A` ????An>A?  file::Talks/MAAMWorkshop.showcase<CVA؋CVAPD*A؋ ????CVAP?   1BD@BC׀D<D@????BC׀?   ?BCـBB(D;Cـ????BB(?  *  ?`>?Q: Why not hire professional programmers to write code?  **A: Programmers are more expensive than comp'l* mathematicians.    **A: Programmers usually work from precise specifications* -- in comp'l math problems, algorithms and  *  implementations often developed concurrently    **A: Programmers with adequate math background are * not common    * dFD,PAD,NAF@D;A@@u?x??D,NAF@?  Page 1  An AcpAn>AC A` ?s???An>A?  file::Talks/MAAMWorkshop.showcase<CVA؋CVAPD*A؋ ?q???CVAP?   1HBDBCD=D?o???BC?   ?Recommendations for a computationalmath career@   ?    ?ACABDD;C?k???ABD?  *   l  Learn about computer science & software engineering **     G  system architecture (to understand performance)**   G  programming language issues**   G  data structures and algorithms* (not everything is a matrix)    **   G  software development tools:* language sensitive editors    * syntax checkers    * source management systems    * dFD,PAD,NAF@D;A@@i?x??D,NAF@?  Page1  An AcpAn>AC A` ?g???An>A?  file::Talks/MAAMWorkshop.showcase<CVA؋CVAPD*A؋ ?e???CVAP?   1B DB CՀD=@D?c???B CՀ?   ?Recommendations (cont.)BCBBD<@C?a???BB?  *   l  Learn an application area (or several)**     G  must be able to communicate effectively with* application specialists**   l  Learn where to find software applicable to your * areas of interest:**   G    G  *   G    G  *   G  *   G  sci.math.num-analysis, comp.programming*   G  NAG, IMSL (VNI), SLATEC, Port (Bell Labs)*   G  MathSoft Inc, MathWorks Inc.* dFD,PAD,NAF@D;A@@_?x??D,NAF@?  Page1  An AcpAn>AC A` ?]???An>A?  file::Talks/MAAMWorkshop.showcase<CVA؋CVAPD*A؋ ?[???CVAP?   18BCBB\D<C?W???BB\?  *   l  Always follow the "  ?`>?principle of least astonishment  "**   l  Beware the "  ?`>?not invented here  " syndrome**   l  Work on your writing and speaking skills.**   G  Strunk & White,    The Elements of Style  *   G  N. Higham,    Handbook of Writing for the      * Mathematical Sciences     *   G  take a technical writing class (then take it again)    **   l  Read the documentation!*BDBCՀD=D?O???BCՀ?   ?Recommendations (cont.) dFD,PAD,NAF@D;A@@U?x??D,NAF@?  Page 1  An AcpAn>AC A` ?S???An>A?  file::Talks/MAAMWorkshop.showcase<CVA؋CVAPD*A؋ ?Q???CVAP?   1BDBCՀD<D?M???BCՀ?   ?Other ReferencesBCBD;C?I???B?  D.E.Knuth,    The Art of Computer Programming (>=3 vols)  //   G  how to take ideas and make them into programs//Fred Brooks,    The Mythical Man Month//    G  everyone should read this//David Goldberg, "What every computer scientist should/ know about floating point"/    ACM Computing Surveys,  1991/    / Sun Numerical Computing Guide/  / dFBD,PAD,NAF@D;A@@ ?x??D,NAF@?  Page1  CAn AcpAn>AC A` ? ???An>A?  file::Talks/MAAMWorkshop.showcaseD<CVA؋CVAPD*A؋ ? ???CVAP?   1EBD@BCD=D@????BC?   ?Other ReferencesFADBAC`D6xRC????AC`?  Abelson & Sussman,    Structure and Interpretation of / Computer Programs//    G  intended for CS students, not much math, but / a good approach to writing programs///reference manuals for   ?`>?all  the languages and tools you use/8HBC@BBD=@C@????BB?  newsgroups22the Web2 T)p=v | M5m  }Yܩ% ]