Download - Alitheia Core - University of Oregon · Writes an analysis tool by implementing 3 methods Uses raw data or metadata Installs the plug-in Resolves metric dependencies Schedules Metric

Transcript
Page 1: Alitheia Core - University of Oregon · Writes an analysis tool by implementing 3 methods Uses raw data or metadata Installs the plug-in Resolves metric dependencies Schedules Metric

3Run analysis tools

2Preprocess raw data and extract metadata

1Import the data

CPU

Mail Metadata

Threads

Metric Results

SCM Metadata

1

2Versions

Bug Metadata

Dev Metadata

Mu

ltiple

Ide

ntitie

s

MetricMetrics can use more

than one data source or results from other

metrics

CPU

CPU

Raw Data Mirror

Projects SCM Mails Bugs 3

4Share results and tools

5Leave your comments here

Stage 1: Raw data to metadata

Stage 2: Metadata Relationship Inference

All graphics used are copyright of the Oxygen project, distributed under the CC Atribution ShareAlike 3 licenseProject logos are copyright of each respective project

Researcher Alitheia Core

▪Writes an analysis tool by implementing 3 methods▪Uses raw data or metadata▪ Installs the plug-in

▪Resolves metric dependencies▪Schedules Metric Jobs on CPUs

Don't forget your sticker!

Αlitheia Core is cool!

Yeah, but let's pray it

won't find any bad

smells

Researcher▪Downloads Alitheia Core▪Downloads example raw data or metadata▪Writes a plug-in▪Submits it to sqo-oss.org

Community▪Replicates experiments▪Benefits from each other's work

www.sqo-oss.org

SQO-OSS project▪Maintains common experiment infrastructure▪Develops Alitheia Core▪Shares raw data and metadata with researchers that submit plug-ins

Currently, we have712 projects

4,674,236 revisions822 mailing lists

3,757,367 emails450,210 bug reports

Stage 1 Stage 2

▪Uses SCM log to create avirtual file tree in the DB ▪ Imports mail and bug metadata

▪Reconstructs mailing list threads▪Resolves developer identities across data sources▪Other analyses possible

More than 25 metrics already available

0

0.02

0.04

0.06

0.08

0.1

0.12

0.14

66595

66630

66681

66731

66867

66872

66874

66883

66884

66889

66892

66929

67013

Cont

ribut

ion

poin

ts

Developer Id

Positive contribution per developer broken down per action for project: Tracker.

MSTMFRMCTCNDCNSCADDCCGNCREMCTFCDFCBNCPHBRPBCC

0

20

40

60

80

100

Gnom

e-D

eskto

p

Epip

hany

Gnom

e-S

yste

m-T

ools

Gnom

e-V

FS

AbiW

ord

Gnom

e-U

tils

Gnom

e-G

am

es

Evin

ce

gedit

Lib

Gnom

eP

rint

EyeO

fGN

OM

E

GT

K-D

oc

Tom

boy

Deskbar-A

pple

t

Sabayon

Orb

it2

Gnom

e-T

hem

es

Netw

ork

Manager

Gnom

e-P

ilot

GO

K

Gnom

e-M

ud

Tra

cker

F-S

pot

Meld

Gnum

eric

Cheese

Garn

om

e

Dashboard

gvfs

Saw

fish

Gnom

e-N

etw

ork

Banshee

Gnom

e-S

cre

ensaver

Seahors

e

Bra

sero

Gnom

e-K

eyrin

g

Gnom

e-C

om

mander

Muin

e

MlV

iew

Pla

nner

LS

R

Gnom

e-P

ow

er-M

anager

Lib

soup

Driv

el

Vala

Glo

m

Gnom

e-C

hem

istry

-Utils

Conduit

%

Total contribution percentage for the top 30% of developers

66%

Top-30%Average

Data Warehouse

Alitheia CoreSoftware Engineering Studies in Four Steps

Georgios Gousios & Diomidis Spinelis

{gousiosg,dds}@aueb.grAthens University of Economics and Business

This work is supported by the Greek Secretariat of Research and Technology, throughout the operational program COMPETITIVENESS, measure 8.3.1 (PENED).