[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

RE: "semantically identical" requirements text

To: <zhigang.liu@nokia.com>, <rohc@cdt.luth.se>
Subject: RE: "semantically identical" requirements text
From: "Paul Francis" <francis@tahoenetworks.com>
Date: Tue, 24 Apr 2001 17:02:49 -0700
Cc: <Erik.Nordmark@eng.sun.com>, <micke@cs.arizona.edu>
Thread-Index: AcDHaS6APQs5LJydT0Ozn4pzcx4a3QFlK2kwAAKbh+AAALcogAADntuQ
Thread-Topic: "semantically identical" requirements text



> 
> That does not change the root of the problem: the semantic
> of a field (use your example: whether it is for monitoring 
> purpose or something else) is still determined by the end 
> receiver.
> 
> I'm glad that you agree on this.
>  

Thinking about this more, I agree in theory, I'm not sure I agree in
practice.  In practice, speaking of the seq and timestamp fields of RTP,
it seems that the source knows when the receiver is treating one these
fields as pliant or brittle.  I.e., if its SRTP, then the fields become
brittle, but the source knows that it is SRTP.  I can't think of a
practical case where the source doesn't know the field is brittle, but
the receiver is able to treat it as brittle.  

Now, you might agree that today there is no case where the sender
doesn't automatically know what the receiver needs, but still argue that
there might be some future case.  My response would be that if we are
going to choose between a real need today and an unknown future need, we
should err in the direction of today's needs.  

> What I'm trying to say is that in practice, a requirement
> already involves a first look into the problem, identifying

sure.

> reasonable assumptions (and those are not), and then set the 
> bar. In this case, I'm afraid that the requirement itself 
> introduces too many assumptions and consequently additional 
> complexity, which may not be needed otherwise.
> 

I'm not sure we'll have additional complexity.  Gehco sure is simple.
If we decide to split gehco between two boxes (the app and the
compressor/router), then we get into some more complexity.  But this may
be well worth it if it gives us better performance.  (Compression, after
all, is fundamentally about a complexity/performance trade-off).

I will go off and fine tune the draft requirements based on these
discussions.

PF



begin 666 winmail.dat
M>)\^(C,``0:0" `$```````!``$``0>0!@`(````Y 0```````#H``$(@ <`
M& ```$E032Y-:6-R;W-O9G0@36%I;"Y.;W1E`#$(`0V ! `"`````@`"``$$
M@ $`+P```%)%.B B<V5M86YT:6-A;&QY(&ED96YT:6-A;"(@<F5Q=6ER96UE
M;G1S('1E>'0`,1$!!8 #``X```#1!P0`& `1``(`,0`"`#H!`2" `P`.````
MT0<$`!@`$0`"`#(``@`[`0$)@ $`(0```#)$,C<Q1C8X0S8U-T1!-$(X,D(W
M,$)$,S$T148V,#E&`$4'`0.0!@!("P``,@````L``@`!`````P`F```````#
M`"X```````,`-@``````0 `Y`'K.I1 ;S< !'@`]``$````%````4D4Z( ``
M```"`4<``0```#<```!C/75S.V$](#MP/51A:&]E3F5T=V]R:W,[;#U43D58
M5E,P,BTP,3 T,C4P,# R-#E:+3$Q,3 ``!X`< `!````*P```")S96UA;G1I
M8V%L;'D@:61E;G1I8V%L(B!R97%U:7)E;65N=',@=&5X= ```@%Q``$````J
M`````<#':2Z /0LY+)R=3T.SGXIS<QX:W0%E*VDP``*;A^ ``+<H@ `#GMN0
M```>`!H,`0````T```!0875L($9R86YC:7,`````'@`=#@$````K````(G-E
M;6%N=&EC86QL>2!I9&5N=&EC86PB(')E<75I<F5M96YT<R!T97AT```"`0D0
M`0```$ %```\!0``V @``$Q:1G5Q=*/[`P`*`')C<&<Q,C7B,@-#=&5X!4$!
M`P'W_PJ `J0#Y <3`H /\P!0!%8_"%4'LA$E#E$#`0(`8VCA"L!S970R!@`&
MPQ$E]C,$1A.W,!(L$3,([PGWMCL8'PXP-1$B#&!C`%#S"PD!9#,V%E +I@KC
M"H1M"H ^'.4=L%03X 5 9#)O!Y%N;P5 $]%N9W!E('1H'X #8!\!;QIF'Y-P
M`V "8&5M.G<?DQ00`X%T#> =5B Q800@9@B0;&0@*'4I%! @>0AA( [ 86UU
M"U!E(1!W'[ ?H07 :;\%0 0`(K %L01@`P!T!;"5"X!G'<=P"'!P;R-!?06Q
M<P-P)'$EXB+@%! I=23R<R' ; ,@`0`.L'+&;0N "8 @8GD?DPGPAR,`'588
M(&-E:78$D((N'5Y))VT@9PM@_R,`'Z ><2-Q(I )PB @`Z"])Y%S*T<<ZQY0
M"X!K)>*_`: (8 5 +@(E81@@+"Q [RUE"X ?D@6P>3$A+&$>\GYS"' ?@#%)
M(* `T"'!91HN+L!),]@Q('-P9>YA+^,@-100<2*0*A$AP#\'@0&0(_ BM 0@
M(#%25/Y0,2 DT100(/ $("SC(3/3"&$JX"!K'O!W!" D4?<QTRJW)/)T&" >
M<#7#*5#W'Y(C03>5800@"U ', (PO2<28@40`D @X#1R+C1@_SAA($ DT 9!
M.#,?H3JD-Y7>8@61!X ^)C$@8C!C.8V_+.,DU#^B-'(?( !P)S!S_R_0(F0S
M]0= 1($C0211,Q'_0DD>HD2R.A) 2"3R07\["O<!H"#@'Y!O.[0DPCUA/C@]
M'/I..C Q("TR*3!G:.],`2V#.10$<&$IHS,1)0'?'O!&/RH!$H%'IF$P8 -Q
M_47C;"F@2",D4#DU.P<I4-T)@'-!Y"B$"L!G"E Y!_\S$4Y$00`G0R*P,& S
M`D9"_31Q32F@&" U< (@1F((8.\B\5<A0U0@0'<?@ K 'X#\9V\EXDMA$] ?
M\"-!00#^=%I0.I$BH#O1`R!4HD]%^S:R`Y%U+] Z(0.@5Z54HOLQ(%I1<UM@
M66($D"2Q.J1\9&DJP2' +=$@,D]B)^\>T52R3,T=L%<>8BQ2.\#Z>5KE<T^!
M.Y)#8S2Z7$+\<75@@0> `C =5@= .]'R9&1Q;G8&\"L0/4$BL?<4``5 %[!O
M13 +@$MB('CW.&$!`"&Q9F/2'/HR\BM%_RI(/6 "($L3/6 R\"/P8,)]!" H
M-K-;8"-!6H(>\2F_9<%N8SJ1%!$?DQU68@K WS1P-*$PDT9",E1A`U +</\L
MQ3K$9B<_4B+@($ =5FD1_0-@9!K0!Y$EL$MP`X%=0?]MF3:R!:!9`68@9H%3
M`2RPOV!P8,)&$1U602$D`7@DT/<R0210#>!H=B$IH![R5R'_5*(I81\`))$#
M\6O6';8<]/LR>UI0)RBQ$^ K$'A*>8CQ-'%'96@%H#+D*%('</\D`31R6C,%
M@6HQ9!,]D05 GQ]P@,);QEO@@[%O> >1\B@?HF%P-W!OE7ES6+'Y)U!R+P-@
M,& $D&]A0 -_6E$?<&3R9")!,C#B?\Q"_S!G3X%7(5I0*+%90 `@>H#_)-(_
M0BR *P$$(",P6[(I`=\@D 20)3$#@313* A0AB3_8-%EP0& $H%2X3AA)1%=
MT+]/<&9R4N,P-"*@>8@OC0F3.[$LL&4M(#!F*6OK_S% `_ HL5K ("$B@2H1
M(L#W/&)=T!^$9#0``8!E^D#A_T91>Z$QTR-!8' $\",PCF)3+C8<]%!&'/I]
MFG >`#40`0```$(````\-#$V0C5!1C,V,$1%1#4T,#@X1$%$,T-!.$)&0D5!
M-D4V.#<T0%1.15A64S R+G1A:&]E;F5T=V]R:W,N8V]M/@```!X`0A !````
M-@```#PX-3<R0T8Q13)!.35$,C$Q03$Q.3 P,#A#-T5!03(T-C R,3-"0D9!
M0&1A96ES,#5N;VL^`````P" $/____\?`/,0`0```'(```!2`$4`)0`S`$$`
M( `E`#(`,@!S`&4`;0!A`&X`= !I`&,`80!L`&P`>0`@`&D`9 !E`&X`= !I
M`&,`80!L`"4`,@`R`" `<@!E`'$`=0!I`'(`90!M`&4`;@!T`',`( !T`&4`
M> !T`"X`10!-`$P```````L`]A `````0 `','K.I1 ;S< !0 `(,!E4#!$;
MS< !`P#>/Z]O```#`/$_"00``!X`^#\!````#0```%!A=6P@1G)A;F-I<P``
M```"`?D_`0```&0`````````W*= R,!"$!JTN0@`*R_A@@$`````````+T\]
M5$%(3T5.15173U)+4R]/53U&25)35"!!1$U)3DE35%)!5$E612!'4D]54"]#
M3CU214-)4$E%3E13+T-./49204Y#25,`'@#Z/P$````5````4WES=&5M($%D
M;6EN:7-T<F%T;W(``````@'[/P$````>`````````-RG0,C 0A :M+D(`"LO
MX8(!`````````"X````#`/T_Y 0```,`&4 ``````P`:0 `````>`#! `0``
M``@```!&4D%.0TE3`!X`,4 !````" ```$9204Y#25,`'@`X0 $````(````
M1E)!3D-)4P`>`#E `0````(````N`````P`)60$````+`&:!"" &``````# 
M````````1@`````.A0````````,`?H$(( 8``````, ```````!&`````%*%
M``!];@$`'@!_@0@@!@``````P ```````$8`````5(4```$````$````.2XP
M``,`PH$(( 8``````, ```````!&``````&%````````"P#'@0@@!@``````
MP ```````$8``````X4````````#`,R!"" &``````# ````````1@`````1
MA0````````,`T8$(( 8``````, ```````!&`````!"%`````````P#8@0@@
M!@``````P ```````$8`````&(4````````+``^""" &``````# ````````
M1@`````&A0````````L`*0``````"P`C```````#``80[(+>Q@,`!Q"D!0``
M`P`0$ (````#`!$0`````!X`"! !````90```%1(051$3T533D]40TA!3D=%
M5$A%4D]/5$]&5$A%4%)/0DQ%33I42$5314U!3E1)0T]&049)14Q$*%5315E/
M55)%6$%-4$Q%.E=(151(15))5$E31D]234].251/4DE.1U!54E ``````@%_
M``$```!"````/#0Q-D(U048S-C!$140U-# X.$1!1#-#03A"1D)%039%-C@W
B-$!43D585E,P,BYT86AO96YE='=O<FMS+F-O;3X```"X]P``
`
end

Prev by Date: RE: "semantically identical" requirements text
Next by Date: RE: [rohc] "semantically identical" requirements text
Prev by thread: RE: "semantically identical" requirements text
Next by thread: Re: "semantically identical" requirements text
Index(es):
- Date
- Thread