Sof

twar
eEngi
neer
ingandPr
ojectManagement
Modul
e–1
1.Def
ineSof
twar
epr
ocess.Expl
aingener
icsof
twar
epr
ocessf
ramewor
k.
(
June2024)
asoft
wareprocessasaf r
amewor
kfort
heact
ivi
ti
es,act
ions,andt
askst
hatar
e
r
equi
redtobui
ldhi
gh-qual
i
tysof
tware.

A generi
c process f
ramewor kf orsoftware engineeri
ng defi
nes f
ivef r
amewor k
acti
vit
iescommuni cati
on,pl
anning,modeling,constr
uction,anddeployment.Inaddit
ion,a
setofumbr el
laact
ivit
ies—pr
ojecttr
ackingandcont r
ol,r
iskmanagement ,quali
tyassur
ance,
confi
gurati
on management ,technicalrevi
ews,and ot hers—are appli
ed thr
oughoutt he
process.

Communi
cation:Befor
eanyt echni
calworkcancommence,iti
scrit
ical
l
yimpor
tantto
communi
cateandcol l
aborat
ewi ththecust
omer,t
ounder
standobj
ect
ivesoft
heproj
ect
andt
ogatherrequi
rements.

Planning:Asof t
war eprojectisacompl i
catedjourney,andtheplanningact
ivi
tycreatesa
“map”t hathelpsgui det heteam asitmakest hej ourney.Themap—cal l
edasof t
ware
projectplan—definesthesoftwareengi
neeri
ngwor kbydescr ibi
ngthetechni
calt
askst obe
conduct ed,t
herisksthatarelikel
y,t
heresourcesthatwillberequi
red,
theworkproductsto
bepr oduced,andawor kschedule.

Model
i
ng:Creat
ingmodelstobet
terunder
standsof
twar
erequi
rement
sandt
hedesi
gnt
hat
wi
ll
achiev
ethoserequi
rement
s.

Const
ruct
ion:Thi
sactivi
tycombi
nescodegener
ati
onandt
het
est
ingt
hati
srequi
redt
o
uncov
ererror
sinthecode.

Depl
oyment:Thesoftwareisdel
i
veredtothecustomerwhoev
aluat
est
hedel
i
ver
edpr
oduct
andprov
idesfeedbackbasedontheevaluat
ion.

Sof
twar
eEngi
neer
ing&Pr
ojectManagement
Thesef iv
egener i
cf r
amewor kactiv
iti
escanbeuseddur i
ngt hedev el
opmentofsmal l
,
si
mpl eprograms,t hecreat
ionoflargeWebappl icat
ions,andfort heengineeri
ngofl arge,
compl excomput er-
basedsystems.
.
Alinearprocessf l
owexecuteseachoft hefi
veframewor kacti
vit
iesinsequence, begi
nning
wit
hcommuni cati
onandcul minat
ingwi t
hdeployment( Fi
gure2. .Ani
2a) ter
ati
vepr ocess
fl
owr epeatsoneormor eoftheact
ivit
iesbef
oreproceedingtothenext(Figure2.2b).

2.Def
inepr
ocesspat
ter
ns.Expl
aint
hemeansofdescr
ibi
ngt
hepat
ter
ns.
(
June2024)
Everysoftwar eteam encount erspr obl
emsasi tmov est hrought hesof twareprocess.
Itwouldbeusef ulifprovensol uti
onst ot heseproblemswer er eadilyav ai
labletot heteam
so thatt he problems coul d be addr essed and r esolved qui ckly.A pr ocess pattern1
descri
besapr ocess-rel
atedpr oblem thati sencountereddur i
ngsof twareengi neeri
ngwor k,
i
dentif
iest heenv i
ronmenti nwhi cht heproblem hasbeenencount ered,andsuggest soneor
mor eprov ensol ut
ionst othepr oblem.St atedinmor egener alterms,apr ocesspat tern
provi
desy ouwi that empl at
ea consi stentmet hodfordescr ibingpr oblem sol ut
ionswi t
hin
thecontextoft hesof twarepr ocess.Bycombi ni
ngpat t
er ns,asof twar eteam cansol ve
probl
emsandconst ructapr ocesst hatbestmeet stheneedsofapr oject.

Pat
ter
nsDescr
ipt
ion:
PatternName:Thepat ter nisgiv enameani ngfulnamedescr ibi
ngitwi thi
nt hecontextof
thesof twar eprocess( e.g. ,Technical Reviews).
Forces: Theenv i
ronmenti nwhicht hepat ternisencount eredandt heissuest hatmakethe
problem v isi
bleandmayaf f
ectitssol ut
ion.
Type: Thepat t
ernt ypeisspeci f
ied.Ambl er[Amb98]suggest sthreetypes:
1.St agepat tern2.Taskpat tern3.Phasepat t
ern
Ini
ti
al cont ext:Describest hecondi t
ionsunderwhi cht hepat t
ernappli
es.Pr i
ort othe
i
niti
ationoft hepat t
er n
Problem: Thespeci ficpr oblem tobesol vedbyt hepat tern.
ResultingCont ext:Descr ibesthecondi t
ionst hatwillresultoncethepatternhasbeen
successf ullyimplement ed.
RelatedPat ter
ns: Provideal i
stofal lprocesspat ternsthatar edir
ect
lyrelatedtot hi
sone.
KnownUsesandExampl es: I
ndicatet hespeci f
icinstancesi nwhichthepat ternis

Sof
twar
eEngi
neer
ing&Pr
ojectManagement
appl
i
cabl
e.

3.Expl
aint
hedi
ff
erentt
ypesofev
olut
ionar
ypr
ocessmodel
s.(
June2024)
Evoluti
onarymodelsar
eiterat
ive.Theyarechar
acter
izedinamannerthatenablesyouto
developincreasi
ngl
ymorecompl etever
sionsofthesoft
ware.I
ntheparagraphsthatf
oll
ow,
Ipresenttwocommonev oluti
onaryprocessmodels.

• Protot
ypeisdefi
nedasf i
rstorprel
iminaryfor m usi
ngwhi chotherf
ormsar
ecopi
ed
orderiv
ed.
• Protot
ypemodel i
sasetofgener al
objectiv
esf orsoftware.
• Itdoesnotident
if
yt her
equirementsli
kedet ail
edinput,
out put
.
• Itissoftwar
eworkingmodel ofl
imitedfunctional
ity
.
• Inthismodel,wor
kingprogramsarequicklypr oduced.

Adv
ant
agesofPr
otot
ypi
ngModel
• Prototypemodel neednotknowt hedetail
edi nput,out
put,processes,adapt
abi
lit
y
ofoper atingsystem andf ul
lmachi nei
nteraction.
• Inthedev el
opmentpr ocessoft hismodel usersareactiv
elyinvolved.
• Thedev elopmentpr ocessisthebestplatf
or mt ounderstandthesy st
em bytheuser.
• Err
or sar edetectedmuchear l
ier
.
• Givesqui ckuserf eedbackforbet t
ersolut
ions.
• Itidentifiesthemi ssing f
unctional
it
yeasi ly.Italso i
dent i
fi
est heconfusing or
di
ff
icultfunctions.

Di
sadv
ant
agesofPrototypi
ngModel :
• Thecli
entinv
olvementismor eandi ti
snotalwaysconsi
der
edbythedev
eloper
.
• Iti
saslowpr ocessbecauseittakesmor eti
mefordevel
opment.
• Manychangescandi stur
bt her
hy thm oft
hedevel
opmentteam.
• Iti
sathrownawaypr ototy
pewhent heusersar
econfusedwit
hit
.

Sof
twar
eEngi
neer
ing&Pr
ojectManagement
 • Spir
almodel isar iskdriv
enpr ocessmodel.
• Iti
susedf orgener ati
ngt hesof twareproj
ects.
• Inspir
almodel ,
anal t
ernat esoluti
onisprovi
dedi ft
heriskisfoundintheri
sk
analy
sis,t
henal ternatesol ut
ionsaresuggest edandimpl emented.
• Iti
sacombi nationofpr ot otypeandsequentialmodelorwat erf
allmodel.
• Inoneiter
ational l
activi
ti
esar edone,forl
argeproject
st heoutputi
ssmal l
.

Sof
twar
eEngi
neer
ing&Pr
ojectManagement
AdvantagesofSpiralModel
•I
treduceshighamountofr isk.
•I
tisgoodf orl
argeandcr i
ti
cal pr
oject
s.
•I
tgivesstrongapprovalanddocument ati
oncontr
ol.
•I
nspiralmodel,thesoft
warei sproducedear
lyinthel
if
ecy
clepr
ocess.

Di
sadvant
agesofSpi
ralModel
•I
tcanbecostl
ytodevelopasoft
war
emodel
.
•I
tisnotusedf
orsmallproj
ect
s.

4.Expl
ainWat
erf
all
model
.(June2024)
 Ther
eareti
meswhentherequi
rementsf
orapr
oblem ar
ewellunderstood—whenwor
kfl
ows
fr
om communi
cat
iont
hroughdeploy
mentinar
easonabl
yli
nearfashion.

 The wat
erf
allmodel
,somet
imes cal
l
ed t
he cl
assi
cli
fe cy
cle,suggest
s a sy
stemat
ic,
sequenti
alappr
oachtosoft
war edevelopmentthatbeginswit
hcustomerspeci
fi
cati
onof
requi
rementsandprogr
essest hr
oughpl anni
ng,modeli
ng,const
ruct
ion,anddepl
oyment
,
culmi
natingi
nongoi
ngsupportofthecompl et
edsof
tware.

 Theor
igi
nal
wat
erf
all
model
proposedbyWi
nst
onRoy
cemadepr
ovi
sionf
or“
feedback
l
oops,
”thevastmaj
ori
tyofor
gani
zat
ionst
hatappl
ythi
spr
ocessmodel
treati
tasi
fitwer
e
st
ri
ctl
yli
near
.

 Thewat erfallmodeli stheoldestpar adigm f orsoftwar eengineer i

ng.theproblems
thatar esomet imesencount eredwhent hewat er
fall
model i
sappl i
edar e:
 Realpr ojectsr arelyfol
lowthesequent i
alfl
owt hatthemodelpr oposes.Al t
hought he
l
inearmodelcanaccommodat eiterati
on,itdoessoi ndi
rectl
y.Asar esult
,changes
cancauseconf usionastheprojectteam pr oceeds.
 Iti
sof t
endi ffi
cul tforthecustomert ostat
eal lr
equirement sexplici
tly
.Thewat erfal
l
modelr equirest hisandhasdi ffi
cultyaccommodat ingt henat uraluncertai
ntythat
existsatt hebegi nningofmanypr ojects.
 Thecust omermusthav epatience.Awor kingv ersi
onoft hepr ogram(s)wi l
lnotbe
availableunt i
ll ateintheprojecttimespan.Amaj orbl under,i
fundet ecteduntilthe
wor kingpr ogram i srevi
ewed,canbedi sastrous.

Sof
twar
eEngi
neer
ing&Pr
ojectManagement
  Thel i
nearnatureoft heclassi
clifecyclel
eadsto“
blocki
ngstat
es”inwhichsome
proj
ectt eam member smustwai tforothermembersoft heteam to compl
ete
dependentt asks.Inf act
,thet i
mespentwai ti
ngcanexceedt heti
mespenton
producti
vewor k.
 Thewat erfal
lmodelisof t
eninappropri
atet
oday,
assoftwar
eworkisfast
-pacedand
subjectt
oanev er
-endingstr
eam ofchanges.

5.Def
ineSof
twar
eandExpl
aint
heNat
ureofsof
twar
e?
Soft
war ei s:
(1)Instructi
ons( computerprograms)thatwhenexecut edprovi
dedesi r
edf eat
ures,funct
ion,and
perf
ormance;
(2)Dat a str uctures t
hatenablet he programs to adequatel
ymani pulat
ei nfor
mat i
on,and (3)
descri
pt i
v einformationinbothhardcopyandv ir
tualfor
mst hatdescr
ibestheoperationanduseof
theprogr ams
Comput ersoftwarecont
inuestobet hesinglemostimportantt
echnologyont heworldstage.
Today ,
sof twar etakesonadualr ole.I
tisapr oduct,andatt
hesamet ime, t
hev ehi
clefordeli
veri
ng
aproduct .

●asaproduct
,itdel
i
ver
sthecomput
ingpotent
ial
embodiedbycomput
erhar
dwar
eormor
ebr
oadl
y,
byanet
workofcomputerst
hatar
eaccessi
blebylocal
har
dware.

●ast hevehicleusedt odeli

vertheproduct,soft
war eactsast hebasisf ort
hecontrolofthe
computer(operati
ngsystems),t
hecommuni cati
onofi nfor
mat i
on(networ
ks),andthecr
eationand
contr
olofotherprograms.
Sof
twar edeli
ver
sthemosti mport
antproductofourtime—informat
ion.Therol
eofcomputer
soft
warehasunder gonesigni
fi
cantchangeoverthelasthal
f-cent
ury.

6.Expl
ainDi
ff
erentSof
twar
eAppl
i
cat
ionDomai
ns.
Sev
enbr
oadcat
egor
iesofcomput
ersof
twar
epr
esentcont
inui
ngchal
l
engesf
orsof
twar
eengi
neer
s

●Sy stem softwar e--acol l

ectionofpr ogramswr it
tent
oser v
iceotherpr ogr
amsSome
syst
em sof t
war e(e.g.,
compilers,edit
ors,andfil
emanagementut i
li
ti
es)processescomplex,
butdeterminate,
4i nfor
mationst ruct
ures.Othersystemsappl
icat
ions(e.
gOper ati
ngsyst
em
component s,dr iv
ers,networ king software,t el
ecommunicati
ons processors) pr
ocess
l
argelyindeter
mi natedata.

● Appli
cati
on Softwar
e--stand-
aloneprogramst hatsol
v ea speci
fi
cbusi
nessneed.
Appl
icat
ionsint hi
sareaprocessbusinessort echni
caldatainawayt hatfaci
l
itat
es
busi
nessoperat
ionsormanagement/t
echnical
decisi
onmaking.

● Engineer
ing/
sci
entif
ic sof
tware--character
ized by “numbercrunchi
ng”algor
ithms,
Appli
cat
ionsrangefrom astronomytov ol
canology,from aut
omotiv
est r
essanalysi
sto
spaceshutt
leor
bit
aldynamics,andf
rom molecul
arbiologytoautomat
edmanufactur
ing.

●Embeddedsoftwar
e--r
esideswit
hinapr
oductorsy
stem andi
susedt oi
mpl
ementand
cont
rol
feat
uresandf
unct
ionsfort
heenduserandf
orthesy
stem i
tsel
f.

Sof
twar
eEngi
neer
ing&Pr
ojectManagement
● Pr oduct-l
inesof t
war e--designed to provi
deaspeci f
iccapabi
li
tyforusebymany
dif
ferentcust omers.Product-
li
nesoftwarecanf ocusonali
mitedandesoter
icmarket
place
(e.
g.,inv ent
or y contr
olpr oducts) or address mass consumer market
s( e.
g.,wor d
processing,spreadsheets,computergraphi
cset c)

● Web Appl i
cati
ons --his network-cent
ri
c soft
war e cat
egoryspans a wide arr
ayof
appl
icati
ons.WebAppsar eev olvi
ngi nt
osophisti
catedcomput i
ngenv ir
onment
st hatnot
onl
ypr ovi
destand-
alonefeatures,comput i
ngfuncti
ons,andcont enttotheenduser,but
al
soar ei
ntegr
atedwithcor
por atedatabasesandbusinessappli
cations

●Art
ifi
cialI
ntel
li
gencesoftwar
e- -makesuseofnonnumericalal
gori
thmstosolv
ecomplex
pr
oblemst hatarenotamenabl etocomput ati
onorst
rai
ghtfor
wardanalysi
s.Appl
i
cati
ons
wi
thi
nt hisareaincl
uder obot
ics,expertsyst
ems,pat
ter
nr ecogni
ti
on(imageandv oi
ce)
,
ar
ti
fi
cialneural
networks,
theor
em pr ov
ing,andgamepl
ayi
ng.

7.EXPLAI NTHEUNI QUENATUREOFWEBAPPS

Int heearlyday soft heWor l
dWi deWeb,websi teswerejustasetofl i
nked
hypert
ext f il
es whi ch present ed informat i
on using t ext and li
mited graphi
cs.The
augment ationofHTMLbydev elopmentt oolsli
keJav a,XMLenabl edwebengi neerst o
provi
decomput ingcapabi l
i
tyal ongwi thi nformati
onalcont ent.
Web- basedsy st
emsandappl ications(WebApps)ar esophi
sti
catedtoolsthatnot
only present st and-alone inf ormation but al so integr at
e databases and business
appli
cations.Web- based sy stems and appl i
cati
ons “inv ol
ve a mixtur
e between print
publi
shingandsof twaredev elopment ,betweenmar keti
ngandcomput i
ng,bet
weeninternal
communi cati
onsandext er
nal relati
ons, andbet weenartandt echnol
ogy.”

Thef
oll
owi
ngar
ethecommonat
tri
but
esf
orWebApps:

 Networkint
ensi
veness:AWebAppresi
desonanetwor
k(I
nter
netorI
ntr
anet
)and
mustser
vetheneedsofadi
versecommuni
tyofcl
i
ent
s.

 Concur
rency
:Al
argenumberofuser
smayaccesst
heWebAppatonet
ime

 Unpr
edict
abl
eload:Thenumberofuser
soft
heWebAppmayv
arybyor
der
sof
magnit
udefr
om dayt
oday.

 Perf
ormance:I
faWebAppusermustwai
ttool
ong,heorshemaydeci
det
ogo
el
sewher
e.

 Av ai
labi
l
ity
:Alt
houghtheexpect
ati
onof100percentavail
abi
l
ityi
sunr
easonabl
e,
usersofpopul
arWebAppsoft
endemandaccessona24/7/365basi
s.

 Dat
adriv
en: Thepri
maryf
uncti
onofmanyWebAppsistousehy
per
medi
atopr
esent
t
ext
,gr
aphics,audi
o,andv
ideocont
entt
otheenduser

Sof
twar
eEngi
neer
ing&Pr
ojectManagement
  Contentsensi
ti
ve:Thequal
it
yandaest
het
icnat
ureofcont
entr
emai
nsani
mpor
tant
det
erminantoft
hequali
tyofaWebApp

 Continuousevolut
ion:Unl
ikeconvent
ionalappl
i
cationsof
twar
ethatevolvesovera
seri
es of planned,chronologi
cal
ly spaced r
eleases,Web appl
icat
ions ev
olve
conti
nuousl
y.

 Immedi
acy:WebAppsof
tenexhi
bitat
ime-
to-
mar
kett
hatcanbeamat
terofaf
ew
day
sorweeks.

 Securi
ty:BecauseWebAppsar
eavail
abl
ev i
anet
wor kaccesssensi
ti
vecont
entmust
beprot
ectedandsecuremodesofdatat
ransmi
ssionmustbepr ovi
ded.

 Aest
het
ics:
Anundeni
abl
epar
toft
heappeal
ofaWebAppi
sit
slookandf
eel
.

8.Def
ineSof
twar
eEngi
neer
ingandexpl
aini
tfeat
uresi
ndet
ail
.
Sof
twar
eEngineer
ingDefi
nit
ion:Theappl
i
cat
ionofasy st
emati
c,di
sci
pli
ned,
quant
if
iabl
eappr
oachtot
hedevelopment,
operati
on,
andmaint
enanceofsof
tware

Feat
uresSof
twar
eEngi
neer
ing:
I
nor
dert
obui l
dsoft
waret
hatisreadytomeett
hechal
l
engesoft
het
went
y-f
ir
stcent
ury
,
y
oumustrecogni
zeafewsimplereal
it
ies:

 I
tfol
lowst
hataconcer
tedeff
ortshoul
dbemadet
ounder
standt
hepr
obl
em
bef
oreasoft
war
esolut
ionisdevel
oped.
 I
tfol
l
owst
hatdesi
gnbecomesapi
vot
alact
ivi
ty.
 I
tfol
l
owst
hatsof
twar
eshoul
dexhi
bithi
ghqual
i
ty.
 I
tfol
l
owst
hatsof
twar
eshoul
dbemai
ntai
nabl
e.

Thesesimpler
eal
iti
esleadtooneconcl on:sof
usi twar
ei nal
lofi
tsf
ormsandacr
ossal
lof
i
tsappli
cati
ondomai nsshouldbeengineer
ed.Andt hatl
eadsustothetopi
cofthis
book—soft
wareengineer
ing.

9.Expl
aint
heEssenceofSof
twar
eEngi
neer
ingPr
act
ice

Under
standtheproblem (
communicat
ionandanaly
sis).
• Whohasast akeinthesolut
iont
otheprobl
em?Thati s,
whoarethestakehol
ders?
• Whataretheunknowns?Whatdat a,f
uncti
ons,andfeatur
esar
er equi
redtoproper
ly
sol
vetheprobl
em?
• Can the probl
em be compar tment
ali
zed? I
si tpossibletor epr
esentsmal l
er

Sof
twar
eEngi
neer
ing&Pr
ojectManagement
 pr
obl
emsthatmaybeeasiert
ounderst
and?
• Canthepr
oblem ber
epr
esentedgr
aphi
call
y?Canananal
ysi
smodel
becr
eat
ed?

Pl
anasol ut i
on(modeli
ngandsof twaredesign).
• Havey ouseensi mil
arproblemsbef ore?
• Hasasi milarprobl
em beensol ved?Ifso,ar
eel ementsofthesol
uti
onreusabl
e?
• Cansubpr obl
emsbedef i
ned?I fso,ar esol ut
ionsreadi
lyappar
entforthe sub
pr
oblems?
• Cany ourepr esentasolutioninamannert hatleadstoeffect
ivei
mplementat
ion?
Canadesi gnmodel becr eated?

Carr
youtthepl
an(codegenerat
ion) .
• Doesthesol
uti
onconformtothepl an?I
ssourcecodetr
aceabl
etothedesi
gnmodel
?
• I
seachcomponentpar toft
hesol uti
onprov
ablycorr
ect?Havethedesi
gnandcode
beenrev
iewed

Exami net
her esul t
.
• Isitpossi
blet ot esteachcomponentpartoft
hesol
uti
on?Hasar easonabletesting
strat
egybeeni mpl emented?
• Doest hesolutionpr oduceresul
tst
hatconfor
mtothedata,f
uncti
ons,andf eatures
thatar erequi red? Has the soft
ware been v
ali
dat
ed againstallst akeholder
requir
ements?

10. Explai
nDavidHookersevenpr
inci
plest
hatf
ocusonsof
twar
e
engi
neer
ingpract
iceasawhole

1.TheFi rstPr i
ncipl e:TheReasonI tAl lExi stsAsof twar esy stem exi stsforoner eason:to
providev aluet oitsuser s.Alldecisionsshoul dbemadewi tht hisinmi nd.
2.TheSecondPr incipl e:KISS( KeepI tSi mpl e,St upid!)Al ldesi gnshoul dbeassi mpleas
possibl e,butno si mpl er.Thisf acil
itat eshav ing amor eeasi lyunder stood and easily
maint ainedsy stem.
3.TheThi rdPr inciple:Mai ntaint heVi si onAcl earv i
sioni sessent i
altot hesuccessofa
softwar epr oj
ect .Wi t
houtone,apr ojectal mostunf aili
ngl yendsupbei ng“ oftwo[ ormor e]
minds”abouti t
self.
4.TheFour t
hPr i
nci ple:WhatYouPr oduce,Ot her sWi l
lConsumeal way sspecify,design,
andi mpl ementknowi ngsomeoneel sewi llhav et ounder standwhaty ouar edoi ng.The
audiencef oranypr oductofsof twar edev el opmenti spot entiall
ylarge.
5.TheFi ft
hPr inciple:BeOpent ot heFut ureNev erdesi gny our selfintoacor ner
.Al waysask
“whati f
,”andpr epar ef oral lpossi bleanswer sbycr eatingsy stemst hatsol vethegener al
problem,notj ustt hespeci ficone.6.TheSi xthPr inciple:Pl anAheadf orReusePl anni
ng
aheadf orr euser educest hecostandi ncr easest hev alueofbot hther eusablecomponent s
andt hesy stemsi ntowhi chtheyar ei ncor por ated.
7.TheSev enthpr i
nci ple:Thi nk!Pl aci ngcl ear,compl etet houghtbef oreactional most
alway spr oducesbet terr esult
s.Wheny out hinkaboutsomet hing,y ouar emor elikelytodo
i
tr i
ght .Youal sogai nknowl edgeabouthowt odoi trightagai n.

Sof
twar
eEngi
neer
ing&Pr
ojectManagement
 11. Expl
ainDi
ff
erentSof
twar
eMYTHS.
• Softwar emy ths:erroneousbel iefsaboutsoftwar eandtheprocessthatisusedt o
buil
di t
Wecat
egorizemy thsf r
om t hreedif
ferentperspecti
ves.
Managementmy t
hs
• My h:Weal
t readyhav eabookt hat’
sfullofstandardsandproceduresforbuildi
ng
softwar e.Won’ tthatprov i
demypeopl ewitheverythi
ngtheyneedtoknow?
• Realit
y:Thebookofst andardsmayv er
ywel lexist
,butisitused?Ar esof t
ware
practi
tionersawar eofi tsexistence?Doesi treflectmodernsoftwareengineeri
ng
practi
ce?I si tcompl ete?Isitadapt abl
e?Isitstreamli
nedtoimprovetime-
to-del
iver
y
whilest il
lmai ntaini
ngaf ocusonqual i
ty?Inmanycases,t heanswertoallofthese
questionsi s“ no.”

• My t
h:I fwe getbehi nd schedul e,we can add mor e programmer s and cat ch
up(somet imescal ledt he“ Mongol i
anhor de”concept )
.
• Reality
:Sof twar edev elopmenti snotamechani st i
cprocessl i
kemanuf acturing.I n
thewor dsofBr ooks[ Bro95]:“addingpeopl etoal atesoft
war eprojectmakesi tl ater.
”
Atf ir
st,thisst atementmayseem count eri
nt ui
t i
ve.Howev er,asnew peopl ear e
added,peopl ewhower ewor kingmustspendt imeeducat ingthenewcomer s,t here
byreducingt heamountoft i
mespentonpr oduct ivedevel
opmentef fort
.Peopl ecan
beaddedbutonl yinapl annedandwel lcoordinated manner .
• My h:
t I
fIdeci det oout sourcet hesoftwarepr ojectt oathir
dpar ty,
Icanj ustrelaxand
l
etthatf i
rm bui ldit.
• Reality
:Ifanor gani zationdoesnotunder st
andhowt omanageandcont r
olsof twar e
projectsinternally,itwillinvariabl
ystrugglewheni toutsourcessoftwar eprojects.

Customermy ths
• My h:Agener
t alst atementofobj ectiv
esi ssuf f
ici
entt obeginwrit
ingprograms—we
canfil
lint hedet ail
slater
• Reali
ty:Al thoughacompr ehensiveandst ablestat ementofr equi
rement sisnot
al
way spossi ble,anambi guous“ statementof obj ectives”i
sar eci
pef ordisaster.
Unambi guousr equirement s ar edev elopedonl ythr ougheffect
iveandcont inuous
communi cationbet weencust omeranddev eloper
• My t
h:Sof twar er equirement s cont i
nually change,but change can be easi l
y
accommodat edbecausesof twarei sflexi
ble.
• Reali
ty:Itistruet hatsof t
war er equirement schange, butt hei
mpactofchangev ari
es
withthet imeatwhi chiti sint r
oduced.Whenr equirement schangesarer equested
earl
ythecosti mpacti sr elati
v el
ysmal l.
16Howev er
, astimepasses,thecostimpact
growsr apidly

• Pr
act
it
ioner
’smy
ths
• My
th:
Oncewewr
it
ethepr
ogr
am andgeti
ttowor
k,ourj
obi
sdone.

Sof
twar
eEngi
neer
ing&Pr
ojectManagement
 • Reali
ty:Someoneoncesaidthat“t
hesoonery oubegi
n‘wri
ti
ngcode,’t
helongeri
t’
ll
takeyout ogetdone.
”Indust
rydataindi
catet
hatbetween60and80per centofall
effor
texpendedonsoftwarewil
lbeexpendedaft
eriti
sdel
iv
eredt
ot hecustomerf
or
thefi
rstti
me.
• My th:Theonl ydel i
v erabl
ewor kproductforasuccessf ulprojectist hewor king
program.
• Real i
ty:Awor kingpr ogram isonlyonepar tofasof twareconf i
gur at
ionthati
ncludes
manyel ement s.Av arietyofwor kproduct
s( e.g.,models,document s,pl
ans)pr ovide
af oundationf orsuccessf ulengineeri
ngand,mor eimpor t
ant,guidanceforsoftwar e
suppor t.
•
My th:Sof tware engi neeri
ng wi l
lmake us cr eatev oluminous and unnecessar y
document at
ion andwi llinvar
iablyslowusdown.
• Real i
ty:Softwar eengineeringisnotaboutcr eati
ngdocument s.Itisaboutcreatinga
qual i
typroduct .Bett
erqual it
yleadstoreducedr ework.Andr educedr eworkresultsin
fasterdeliverytimes.

12. Expl
aini
ndet
ail
Incr
ement
alPr
ocessModel
s
 Theincrementalmodelcombi neselement
sofl i
nearandparall
elprocessfl
ows.
 Theincrementalmodelappl iesli
nearsequencesinast aggeredf ashi
onascalendarti
me
progr
esses.
 Eachlinearsequenceproducesdeliver
abl
e“ i
ncrements”ofthesoftwareinamannerthatis
si
mi l
artotheincrement
spr oducedbyanev ol
uti
onaryprocessflow.
 Whenani ncrement almodelisused,thefir
stincrementi sofenacor
t epr oduct
.That
i
s,basi cr equir
ement s ar
e addressed butmanysuppl ementaryf eatures (some
known,ot hers unknown)r emain undeli
vered.The cor e productis used byt he
customer( orunder goesdetai
ledevaluati
on).Asar esul
tofuseand/ orev al
uation,a
planisdev el
opedf orthenextincr
ement .
 Thei ncr
ement alpr ocessmodelf ocusesont hedel i
veryofanoper at
ionalpr oduct
with each increment .Earl
yi ncr
ement s are str
ipped-down v er
sions oft he fi
nal
product,butt heydopr ovidecapabili
tythatser vest heuserandal sopr ovidea
plat
formf orevaluationbytheuser.

Sof
twar
eEngi
neer
ing&Pr
ojectManagement
14Expl
ainConcur
rentDev
elopmentModel

• Theconcur
rentdev
elopmentmodel
iscal
l
edasconcur
rentmodel
.

Sof
twar
eEngi
neer
ing&Pr
ojectManagement
 • Thecommuni cat
ionactiv
ityhascompl etedinthefirstiter
ati
onandexit
si nthe
awaiti
ngchangesstate.
• Themodel i
ngactiv
itycompleteditsinit
ialcommunicati
onandt hengoest othe
underdevel
opmentstate.
• Ift
hecust omerspecifi
esthechangei ntherequi
rement,thenthemodel
ingacti
vit
y
mov esfr
om theunderdevelopmentstateint
otheawait
ingchangestat
e.
• Theconcur r
entpr
ocessmodel acti
vi
ti
esmov i
ngfr
om onest at
etoanot
herst
ate.

Advantagesoft heconcurrentdevelopmentmodel
•Thi
smodel isappli
cabletoallt
ypesofsof t
waredevel
opmentprocesses.
•I
tiseasyf orunderstandi
nganduse.
•I
tgivesimmedi atefeedbackfrom test
ing.
•I
tprovidesanaccur atepi
ctureofthecurrentst
ateofapr
oject
.

Disadvantagesoft heconcurrentdevelopmentmodel
•I
tneedsbet tercommuni cati
onbet weentheteam member s.Thi
smaynotbe
achievedalltheti
me.
•I
trequiresremember i
ngt hestatusofthedif
ferentact
ivi
ti
es.

15.EXPLAI
NSPECI
ALI
ZEDPROCESSMODELSi
nDETAI
L.
Special
izedpr ocessmodelstakeonmanyoft hecharacter
isti
csofoneormor eoft he
tr
adit
ionalmodel sthesemodelstendt
obeappli
edwhenaspeci al
i
zedornarr
owl
ydefined
soft
wareengi neeri
ngapproachi
schosen
Component -Based Development: component
-based development modelconst
ructs
appl
icationsfrom pr
epackagedsoft
war
ecomponents.

Modeli
ngandconstr
uct
ionact
ivi
ti
esbegi
nwi t
htheidenti
fi
cati
onofcandi
dat
ecomponent
s.
Thesecomponent
scanbedesignedasei
therconv
entionalsof
twar
emodules.

The component -
based developmentmodelleads t
o soft
warer euse,and r
eusabi
lit
y
prov
ides softwar
e engineers wit
h a numberofmeasurable benef
it
s.Yoursoftware
engi
neeri
ngt eam canachiev
ear educt
ioni
ndevel
opmentcy
cleti
measwel lasareduct
ion
i
nprojectcost

TheFor malMet hodsModel :Thef ormalmethodsmodelencompassesasetofact ivi

ti
es
thatleadstof ormalmat hemat i
calspeci
fi
cati
onofcomput ersoftware.Formalmethods
enableyoutospeci fy
,develop, andveri
fyacomputer-
basedsy st
em byapplyi
ngarigorous,
mathemat i
calnot ati
on.A v ari
ati
on on t hi
s approach,caled cl
l ean-r
oom software
engineeri
ngiscurrentl
yappliedbysomesof t
waredevelopmentorgani
zati
ons.

Aspect-Or
ientedSoftwareDev elopment:t
heyareof t
enrefer
redtoascr osscutt
ingconcerns.
Aspectualrequir
ement sdefi
net hosecrosscut
tingconcer
nst hathaveani mpactacrossthe
soft
war earchit
ecture.Aspect-ori
ent
edsof t
waredev el
opment( AOSD) ,oft
enrefer
redtoas
aspect-
ori
entedpr ogramming( AOP),isarelat
ivel
ynewsof t
war eengineeri
ngparadigm t
hat
provi
desapr ocessandmet hodologi
calapproachf ordefi
ning,specifyi
ng,desi
gning,and
constr
uctingaspects.

Sof
twar
eEngi
neer
ing&Pr
ojectManagement
16.Expl
ainPhasesoft
heUni
fi
edPr
ocess

Thei ncepti
onphaseoft heUPencompassesbot hcustomercommuni cati
onandpl anni
ng
acti
viti
es.
Theel aborati
onphaseencompassest hecommuni cati
onandmodel i
ngact i
vi
ti
esoft he
genericprocessmodel
Theconst ructi
onphaseoft heUPisident
icaltotheconst r
uct
ionacti
vit
ydefinedf ort
he
genericsoftwareprocess
Thet ransi
ti
onphaseoft heUPencompassest helat
terstagesofthegenericconstr
ucti
on
acti
vityandthef i
rstpar
tofthegener
icdepl
oymentact i
vi
ty.
Thepr oducti
onphaseoft heUPcoinci
deswiththedeploymentacti
vit
yofthegeneric
process.

15.EXPLAI
NPERSONALANDTEAM PROCESSMODELS

ThePer sonalSof
twarePr
ocess(PSP)emphasizesper
sonalmeasurementofbot
hthewor
k
productthatispr
oducedandtheresul
tantqual
i
tyoft
heworkproduct.

Pract
iti
onerr
esponsi
blef
orpr
ojectpl
anning andempowerst
hepract
it
ionert
ocontr
olt
he
quali
tyofallsoft
wareworkproduct
st hataredevel
oped.ThePSP modeldefi
nesfi
ve
fr
amewor kact
ivi
ti
es:

• Pl
anning.
• High-
level
desi
gn.
• High-
level
desi
gnrev
iew.
• Devel
opment.

Sof
twar
eEngi
neer
ing&Pr
ojectManagement
 • Post
mor
tem.

PSPst r
essestheneedt oident
if
yerrorsearl
yand,justasimport
ant
,tounder
standthe
ty
pesofer rorsthatyouar eli
kel
yt o make.Thisisaccompli
shedthr
oughar igor
ous
assessmentacti
vi
typer
for
medonal lworkproduct
syouproduce.

PSPr epr
esentsadi scipli
ned,metri
cs-basedapproacht osof
twareengineer
ingthatmay
l
eadt ocult
ureshockf ormanypr acti
ti
oners.However,whenPSPi spr
operlyi
ntroducedto
softwar
e engi neer
s [ Hum96],t he result
ing i
mpr ovement i
n software engineeri
ng
product
ivi
tyandsof t
war equali
tyaresi
gnifi
cant[
Fer97].However
,PSPhasnotbeenwi del
y
adoptedthroughouttheindust
ry.

16.Expl
ainTeam Sof
twar
ePr
ocess(
TSP)
Thegoal
ofTSPi st
obui
l
da“
sel
fdi
rect
ed”pr
ojectt
eam t
hator
gani
zesi
tsel
ftopr
oduce
hi
gh-
qual
itysof
twar
e.

 Buil
dsel f-
dir
ectedt eamst hatpl
anandt r
acktheirwork,establ
i
shgoals,andown
thei
rpr ocessesandpl ans.Thesecanbepur esof t
wareteamsorintegratedproduct
teams( IPTs)of3t oabout20engineers.
• Showmanager showt ocoachandmot iv
atetheirt
eamsandhowt ohelpt hem
sustainpeakper f
ormance.
• Acceleratesof t
war eprocessimprovementbymaki ngCMM23Lev el 5behav i
or
normal andexpect ed.
• Provideimpr ovementgui dancetohi
gh-maturit
yorganizati
ons.
• Facil
itateuniversi
tyteachingofi
ndustri
al-
gradeteam skil
ls.

Sof
twar
eEngi
neer
ing&Pr
ojectManagement