What is group w are? Soft w are sp e cic al ly designed - - PDF document

what is group w are soft w are sp e ci c al ly designed
SMART_READER_LITE
LIVE PREVIEW

What is group w are? Soft w are sp e cic al ly designed - - PDF document

What is group w are? Soft w are sp e cic al ly designed to supp ort group w orking with co op erativ e requiremen ts in mind NOT just to ols for comm unication Group w are can b e classied b


slide-1
SLIDE 1 Human{Computer In teraction, Pren tice Hall A. Dix, J. Finla y , G. Ab
  • wd
and R. Beale c 1993 Group w are Chapter 13 (1) What is group w are? Soft w are sp e cic al ly designed
  • to
supp
  • rt
group w
  • rking
  • with
co
  • p
erativ e requiremen ts in mind NOT just to
  • ls
for comm unication Group w are can b e classied b y
  • when
and wher e the participan ts are w
  • rking
  • the
function it p erforms for co
  • p
erativ e w
  • rk
Sp ecic and dicult problems with group w are implemen tion
slide-2
SLIDE 2 Human{Computer In teraction, Pren tice Hall A. Dix, J. Finla y , G. Ab
  • wd
and R. Beale c 1993 Group w are Chapter 13 (2) The Time/Space Matrix Classify group w are b y: when the participan ts are w
  • rking,
at the same time
  • r
not where the participan ts are w
  • rking,
at the same plac e
  • r
not

same place different place same time different time face-to-face conversation telephone post-it note letter

Common names for axes: time: sync hronous/async hronous place: co-lo cated/remote
slide-3
SLIDE 3 Human{Computer In teraction, Pren tice Hall A. Dix, J. Finla y , G. Ab
  • wd
and R. Beale c 1993 Group w are Chapter 13 (3) Classication b y F unction Co
  • p
erativ e w
  • rk
in v
  • lv
es: P articipan ts who are w
  • rking
Artefacts up
  • n
whic h they w
  • rk

understanding participants artefacts of work

P P A

direct communication control and feedback

What in teractions do es a to
  • l
supp
  • rt?
computer-mediated comm unication direct comm unication b et w een participan ts meeting and decision supp
  • rt
systems common understanding shared applications and artefacts con trol and feedbac k with shared w
  • rk
  • b
jects
slide-4
SLIDE 4 Human{Computer In teraction, Pren tice Hall A. Dix, J. Finla y , G. Ab
  • wd
and R. Beale c 1993 Group w are Chapter 13 (4) Email and bulletin b
  • ards
asynchr
  • nous/r
emote familiar and most successful group w are Recipien ts
  • f
email: dir e ct in To: eld c
  • pies
in Cc: eld deliv ery iden tical | dierence is so cial purp
  • se
dierences b et w een email and BBs fan
  • ut
  • ne-to-one
| email, direct comm unication
  • ne-to-man
y | email, distribution lists BBs, broadcast distribution con trol sender | email, priv ate distribution list administrator | email, shared distribution list recipien t | BBs, subscription to topics
slide-5
SLIDE 5 Human{Computer In teraction, Pren tice Hall A. Dix, J. Finla y , G. Ab
  • wd
and R. Beale c 1993 Group w are Chapter 13 (5) Structured message systems asynchr
  • nous/r
emote
  • `sup
er' email | cross b et w een email and a database
  • sender
lls in sp ecial elds
  • recipien
t lters and sorts incoming mail based
  • n
eld con ten ts Type: Seminar announcemen t To: all From: Alan Dix Subject: departmen tal seminar Time: 2:15 W ednesda y Place: D014 Speaker: W.T. P
  • h
Title: The Honey P
  • t
Text: Recen t researc h
  • n
so cially constructed meaning has fo cused
  • n
the image
  • f
the Honey P
  • t
and its dialectic in terpretation within an encultured hermeneutic. This talk : : : but, w
  • rk
b y the sender : : : b enet for the recipien t conict glob al structuring b y designer vs. lo c al structuring b y participan ts
slide-6
SLIDE 6 Human{Computer In teraction, Pren tice Hall A. Dix, J. Finla y , G. Ab
  • wd
and R. Beale c 1993 Group w are Chapter 13 (6) Video conferences and comm unication synchr
  • nous/r
emote T ec hnology emerging: ISDN + video compression ma jor uses:
  • video
conferences
  • p
erv asiv e video for so cial con tact
  • in
tegration with
  • ther
applications
  • ften
c heap er than face-to-face meetings (telecom m unications costs vs. air igh ts) but not a substitute:
  • small
eld
  • f
view
  • lac
k
  • f
recipro cit y
  • p
  • r
ey e con tact One solution for lac k
  • f
ey e con tact | the video-tunnel

camera monitor mirror half-silvered mirror

slide-7
SLIDE 7 Human{Computer In teraction, Pren tice Hall A. Dix, J. Finla y , G. Ab
  • wd
and R. Beale c 1993 Group w are Chapter 13 (7) Meeting and decision supp
  • rt
systems In design, managemen t and researc h, w e w an t to:
  • generate
ideas
  • dev
elop ideas
  • record
ideas primary emphasis | common understanding Three t yp es
  • f
system: argumen tation to
  • ls
asynchr
  • nous
c
  • -lo
c ate d recording the argumen ts for design decisions meeting ro
  • ms
synchr
  • nous
c
  • -lo
c ate d elec tronic supp
  • rt
for face-to-face mee tings shared dra wing surfaces synchr
  • nous
r emote shared dra wing b
  • ard
at a distance
slide-8
SLIDE 8 Human{Computer In teraction, Pren tice Hall A. Dix, J. Finla y , G. Ab
  • wd
and R. Beale c 1993 Group w are Chapter 13 (8) argumen tation to
  • ls
asynchr
  • nous
c
  • -lo
c ate d h yp ertext lik e to
  • ls
to record design r ationale Tw
  • purp
  • ses:
  • remining
the designers
  • f
the reasons for decisons
  • comm
unicating rationale b et w een design teams Mo de
  • f
collab
  • ration:
  • v
ery long term
  • sometim
e s sync hronous use also Example: gIBIS (issue based information system ) v arious no de t yp es including: issues e.g., `n um b er
  • f
mouse buttons' p
  • sitions
e.g., `only
  • ne
button' argumen ts e.g., `easy for no vice' link ed b y relationships suc h as: argumen t supp
  • rts
p
  • sition
e.g., `easy for no vice' supp
  • rts
`only
  • ne
button'
slide-9
SLIDE 9 Human{Computer In teraction, Pren tice Hall A. Dix, J. Finla y , G. Ab
  • wd
and R. Beale c 1993 Group w are Chapter 13 (9) Meeting ro
  • ms
synchr
  • nous
c
  • -lo
c ate d electronic supp
  • rt
for face-to-face meetings
  • individual
terminals (often recessed)
  • large
shared screen (electronic whiteb
  • ard)
  • sp
ecial soft w are
  • U
  • r
C shap ed seating around screen V arious mo des: brainstorming, priv ate use, WYSIWIS WYSIWIS | `what y
  • u
see is what I see' all screens sho w same image an y participan t can write/dra w to screen
slide-10
SLIDE 10 Human{Computer In teraction, Pren tice Hall A. Dix, J. Finla y , G. Ab
  • wd
and R. Beale c 1993 Group w are Chapter 13 (10) T ypical meeting ro
  • m

shared screen 1 2 3 6 5 4

slide-11
SLIDE 11 Human{Computer In teraction, Pren tice Hall A. Dix, J. Finla y , G. Ab
  • wd
and R. Beale c 1993 Group w are Chapter 13 (11) Issues for co
  • p
eration Argumen tation to
  • ls
concurrency con trol t w
  • p
eople access the same no de
  • ne
solution is no de lo cking notication mec hanisms kno wing ab
  • ut
  • thers'
c hanges Meeting ro
  • ms
  • r
holders
  • ne
  • r
man y?
  • r
con trol p
  • licies
who can write and when? solution: lo cking + so cial proto col group p
  • in
ter for deictic reference (this and that)
slide-12
SLIDE 12 Human{Computer In teraction, Pren tice Hall A. Dix, J. Finla y , G. Ab
  • wd
and R. Beale c 1993 Group w are Chapter 13 (12) Shared w
  • rk
surfaces synchr
  • nous
r emote A t simplest, meeting ro
  • ms
at a distance, but : : :
  • additional
audio/video essen tial for so cial pr
  • to
c
  • ls
and discussion
  • net
w
  • rk
dela ys can b e ma jor problem Additional sp ecial eects:
  • participan
ts write
  • n
to large video screen problems with p ar al lax
  • shado
w
  • f
  • ther
participan t's hands app ears
  • n
screen
  • electronic
image in tegrated with video and pap er images Example: T eamW
  • rkStation
remote teac hing
  • f
Japanese calligraph y studen t's strok es
  • n
pap er
  • v
erlaid with video
  • f
instructor's strok es
slide-13
SLIDE 13 Human{Computer In teraction, Pren tice Hall A. Dix, J. Finla y , G. Ab
  • wd
and R. Beale c 1993 Group w are Chapter 13 (13) Shared Applications and Artefacts Compare purp
  • se
  • f
co
  • p
eration: meeting ro
  • ms
and decison supp
  • rt
systems | dev elop shared understanding shared applications and artefacts | w
  • rk
  • n
the same
  • b
jects tec hnology similar but primary purp
  • se
dieren t man y dieren t mo dalities (time/space matrix) shared windo ws | synchr
  • nous
r emote/c
  • -lo
c ate d shared editors | synchr
  • nous
r emote/c
  • -lo
c ate d co-authoring systems | largely asynchr
  • nous
shared diaries | largely asynchr
  • nous
r emote shared information | an y , but largely asynchr
  • nous
sync hronous remote applicati
  • ns
usually require additional audio/video c hannel
slide-14
SLIDE 14 Human{Computer In teraction, Pren tice Hall A. Dix, J. Finla y , G. Ab
  • wd
and R. Beale c 1993 Group w are Chapter 13 (14) Similar { but dieren t Shared PCs and shared windo w systems
  • Multiplex
k eyb
  • ard
and screen
  • Individual
applicatio ns not c
  • l
lab
  • r
ation awar e
  • Flo
  • r
con trol problems: user A t yp es: `interle ave the' user B t yp es: `keystr
  • ke
s ' result: `inkey terslt r e aok e ve tshe' Shared editors
  • An
editor whic h is c
  • l
lab
  • r
ation awar e
  • One
do cumen t | sev eral users
  • Similar
to shared screen in meeting ro
  • m
: : : : : : with similar
  • r
con trol problems!
  • Additional
problem | m ultiple views
slide-15
SLIDE 15 Human{Computer In teraction, Pren tice Hall A. Dix, J. Finla y , G. Ab
  • wd
and R. Beale c 1993 Group w are Chapter 13 (15) Shared editors | m ultiple views

your screen your colleague’s screen We will look at some of the

  • ptions and how they affect

the style of cooperation. Thinking about the shared view vs. different view

  • ptions, it at firstQ

P R Hseems

  • bvious that we should allow

people to edit different parts of a document. This is certainly true while they are working effectively independently. More adaptable systems are needed to allow for the wide variation betweenQ P R Hgroups, and within the same group

  • ver time.

We will look at some of the

  • ptions and how they affect

the style of cooperation. Thinking about the shared view vs. different view

  • ptions, it at first seems
  • bvious that we should allow
Options: same view
  • r
dieren t view single
  • r
separate insertion p
  • in
ts Single view = ) scroll w ars Multiple views = ) loss
  • f
con text with indexic als `I don't lik e the line at the top' `but I just wrote that!'
slide-16
SLIDE 16 Human{Computer In teraction, Pren tice Hall A. Dix, J. Finla y , G. Ab
  • wd
and R. Beale c 1993 Group w are Chapter 13 (16) Co-authoring systems Emphasis is
  • n
long term do cumen t pro duction, not editing Tw
  • lev
els
  • f
represen tation
  • the
do cumen t itself
  • annotation
and discussion Often some form
  • f
h yp ertext structure used Similar problems
  • f
c
  • ncurr
ency c
  • ntr
  • l
to argumen tation systems Sometimes include r^
  • les:
author, commen tator, reader, : : : but who decides the r^
  • les?
and ho w exible are they?
slide-17
SLIDE 17 Human{Computer In teraction, Pren tice Hall A. Dix, J. Finla y , G. Ab
  • wd
and R. Beale c 1993 Group w are Chapter 13 (17) Shared diaries Idea:
  • mak
e diaries and calendars more easily shared
  • allo
w automatic meeting sc heduling etc. Issues for co
  • p
eration: priv acy who can see m y diary en tries? con trol who can write in m y diary? Similar to le sharing issues, but need to b e ligh t wigh t Man y systems ha v e failed b ecause they ignored these issues
slide-18
SLIDE 18 Human{Computer In teraction, Pren tice Hall A. Dix, J. Finla y , G. Ab
  • wd
and R. Beale c 1993 Group w are Chapter 13 (18) Comm unication through the artefact When y
  • u
c hange a shared application:
  • y
  • u
can see the eect | fe e db ack
  • y
  • ur
colleages can to
  • |
fe e dthr
  • ugh
feedtrough enables c
  • mmunic
ation thr
  • ugh
the artefact Not just with `real' group w are Shared data is p erv asiv e:
  • shared
les and databases
  • casew
  • rk
les (often non-electronic)
  • passing
electronic copies
  • f
do cumen ts
  • passing
copies
  • f
spreadsheets Often need direct comm unication as w ell, but indirect comm unication thr
  • ugh
the artefact cen tral F ew example s
  • f
explicit design for co
  • p
eration. Livewar e is an exce ption, a database with `merging'
  • f
copies
slide-19
SLIDE 19 Human{Computer In teraction, Pren tice Hall A. Dix, J. Finla y , G. Ab
  • wd
and R. Beale c 1993 Group w are Chapter 13 (19) Time/space matrix revisited

co-located remote synchronous asynchronous

meeting rooms video conferences, video-wall, etc. shared work surfaces and editors shared PCs and windows argumentation tools email and electronic conferences co-authoring systems, shared calendars

slide-20
SLIDE 20 Human{Computer In teraction, Pren tice Hall A. Dix, J. Finla y , G. Ab
  • wd
and R. Beale c 1993 Group w are Chapter 13 (20) Rened time/space matrix

co-located remote (a) concurrent synchronized (a/b) mixed (b) serial (c) unsynchronized

meeting rooms video conferences, video-wall, etc. shared work surfaces and editors shared PCs and windows co-authoring systems, shared calendars argumentation tools email and structured messages electronic conferences

Mobile w
  • rk
ers and home w
  • rk
ers ha v e infrequen t comm unication | they require unsynchr
  • n
ise d group w are F ew `researc h' system s address this area NO curren t system allo ws uid mo v e m e n t b et w een sync hronised/unsync hronised
  • p
eration
slide-21
SLIDE 21 Human{Computer In teraction, Pren tice Hall A. Dix, J. Finla y , G. Ab
  • wd
and R. Beale c 1993 Group w are Chapter 13 (21) Shared information Gran ularit y
  • f
sharing c h unk size small | edit same w
  • rd
  • r
sen tance large | section
  • r
whole do cumen t up date frequency frequen t | ev ery c haracter infrequen t | up
  • n
explicit `send' lev el
  • f
sharing
  • utput:
shared
  • b
ject shared view shared presen tation input: single insertion p
  • in
t
  • shared
virtual k eyb
  • ard
m ultiple insertion p
  • in
ts
  • ther
participan ts visible
  • group
p
  • in
ter
  • no
visibilit y
slide-22
SLIDE 22 Human{Computer In teraction, Pren tice Hall A. Dix, J. Finla y , G. Ab
  • wd
and R. Beale c 1993 Group w are Chapter 13 (22) Lev els
  • f
shared
  • utput
  • bject

view presentation

VILLAGE_STATS village houses population Burton 23 79 Marleigh 339 671 Westfield 7 15 Thornby 51 123 select houses, population from VILLAGE_STATS where population < 200 sort by houses ascending houses population 7 15 23 79 51 123

population houses

100 50 50

slide-23
SLIDE 23 Human{Computer In teraction, Pren tice Hall A. Dix, J. Finla y , G. Ab
  • wd
and R. Beale c 1993 Group w are Chapter 13 (23) In tegrating comm unication and w
  • rk

understanding feedthrough

P P A

direct communication control and feedback deixis

Added: deixis | reference to w
  • rk
  • b
jects fe e dthor
  • ugh
| for comm unication through the artefact Classied group w are b y function it supp
  • rted
Go
  • d
group w are |
  • p
en to all asp ects
  • f
co
  • p
eration e.g., annotation s in co-authoring system s em b edding direct comm unication bar co des | form
  • f
deixis aids diuse large scale co
  • p
eration
slide-24
SLIDE 24 Human{Computer In teraction, Pren tice Hall A. Dix, J. Finla y , G. Ab
  • wd
and R. Beale c 1993 Group w are Chapter 13 (24) Arc hitectures for group w are I Clien t-serv er arc hitecture

user 1 user 2

. . .

user n client 1 client 2

. . .

client n

server

F eedbac k and net w
  • rk
dela ys

screen feedback user types local machine client remote machine server application ➀ ➁ ➂ ➃ ➄ ➅ ➆ ➇ ➈

A t least 2 net w
  • rk
messages + four con text switc hes With proto cols 4
  • r
more net w
  • rk
messages
slide-25
SLIDE 25 Human{Computer In teraction, Pren tice Hall A. Dix, J. Finla y , G. Ab
  • wd
and R. Beale c 1993 Group w are Chapter 13 (25) Arc hitectures for group w are I I Dieren t arc hitectures: cen tralised | single cop y
  • f
application and data client-server | simplest case N.B.
  • pp
  • site
  • f
X windo ws clien t/se rv e r master-slave sp ecial case
  • f
clien t-serv er N.B. serv er merged with
  • ne
clien t replicated | cop y
  • n
eac h w
  • rkstation
also called p e er-p e er + lo cal feedbac k
  • race
conditions Often `half w a y' arc h tectures:
  • lo
cal cop y
  • f
applicati
  • n
  • cen
tral database
  • lo
cal cac he
  • f
data for feedbac k
  • some
hidden lo c king
slide-26
SLIDE 26 Human{Computer In teraction, Pren tice Hall A. Dix, J. Finla y , G. Ab
  • wd
and R. Beale c 1993 Group w are Chapter 13 (26) Shared windo w arc hitecture
  • Non-collab
  • ration
a w are applications = ) clien t/serv e r approac h corresp
  • nding
feedbac k problems
  • no
`functionalit y' | in the application but m ust handle
  • r
con trol

user 1 user 2

. . .

user n

X X X

user stub 1 user stub 2

. . .

user stub n

application stub application

Xevents Xlib calls

Xevents Xlib calls

slide-27
SLIDE 27 Human{Computer In teraction, Pren tice Hall A. Dix, J. Finla y , G. Ab
  • wd
and R. Beale c 1993 Group w are Chapter 13 (27) F eedthrough Need to inform all
  • ther
clien ts
  • f
c hanges F ew net w
  • rks
supp
  • rt
br
  • adc
a st mes s ages , so : : : n participan ts = ) n
  • 1
net w
  • rk
mes s age s ! Solution: increase gran ularit y reduce frequency
  • f
feedbac k but : : : p
  • r
feedthrough = ) loss
  • f
shared con text T radeo: timelines s vs. net w
  • rk
trac Graphical to
  • lkits
Designed for single user in teraction Problems for group w are include
  • pre-emptiv
e widgets (e.g., p
  • p-up
men us)
  • v
er-pac k aged text (single cursor, p
  • r
view con trol) notic atio n based to
  • lkits
with c al lb acks help (see Ch. 10)
slide-28
SLIDE 28 Human{Computer In teraction, Pren tice Hall A. Dix, J. Finla y , G. Ab
  • wd
and R. Beale c 1993 Group w are Chapter 13 (28) Robustness and scaleabilit y crash in single-user in terface |
  • ne
sad user crash in group w are | disaster ! but, group w are complex: net w
  • rks,
graphics etc.
  • net
w
  • rk
  • r
serv e r fails | standard solutions
  • clien
t fails | three `R's for serv er: robust | serv er should surviv e clien t crash recongure | detect and resp
  • nd
to failure resync hronise | catc h up when clien t restarts
  • errors
in programming defensiv e programming simple algorithms formal metho ds
  • unforeseen
sequenc e s
  • f
ev en ts de ad lo ck | nev er use blo c king I/O nev er assume particular
  • rders
net w
  • rk
pac k et 6= logical mes s age Scaling up to large n um b ers
  • f
users? T esting and debugging: hard!