42

The Promises and Perils of Mining Github: MSR'2014

Embed Size (px)

Citation preview

Page 1: The Promises and Perils of Mining Github: MSR'2014
Page 2: The Promises and Perils of Mining Github: MSR'2014

A  short  (me  ago,  in  an  office,  far  far  away...        

Page 3: The Promises and Perils of Mining Github: MSR'2014
Page 4: The Promises and Perils of Mining Github: MSR'2014

a  postdoc  states:  

Page 5: The Promises and Perils of Mining Github: MSR'2014
Page 6: The Promises and Perils of Mining Github: MSR'2014
Page 7: The Promises and Perils of Mining Github: MSR'2014
Page 8: The Promises and Perils of Mining Github: MSR'2014
Page 9: The Promises and Perils of Mining Github: MSR'2014

[...]  

Page 10: The Promises and Perils of Mining Github: MSR'2014

A  pull  request  that  is  not  marked  as  merged    is  not  necessarily  rejected!!  

 

markedAsMerged  !=  merged    

!merged  !=  rejected  

Page 11: The Promises and Perils of Mining Github: MSR'2014
Page 12: The Promises and Perils of Mining Github: MSR'2014

But  others  will  not  know  about  this...          

Page 13: The Promises and Perils of Mining Github: MSR'2014

From  proceedings  in    a  close,  close  loca(on...  

Page 14: The Promises and Perils of Mining Github: MSR'2014

Except  from  a  paper  at  ...  

Page 15: The Promises and Perils of Mining Github: MSR'2014

Except  from  a  paper  ...  

Page 16: The Promises and Perils of Mining Github: MSR'2014

A  pull  request  that  is  not  marked  as  merged    is  not  necessarily  rejected!!  

 

markedAsMerged  !=  merged    

!merged  !=  rejected  

Re-­‐emphasizing  

Page 17: The Promises and Perils of Mining Github: MSR'2014

Many  are  already  using  this  data  to  reach  conclusions!    

Page 18: The Promises and Perils of Mining Github: MSR'2014

Open  source  report  card  

Page 19: The Promises and Perils of Mining Github: MSR'2014
Page 20: The Promises and Perils of Mining Github: MSR'2014
Page 21: The Promises and Perils of Mining Github: MSR'2014
Page 22: The Promises and Perils of Mining Github: MSR'2014
Page 23: The Promises and Perils of Mining Github: MSR'2014
Page 24: The Promises and Perils of Mining Github: MSR'2014

Mixed  methods  approach  

Page 25: The Promises and Perils of Mining Github: MSR'2014

Surveys  

Page 26: The Promises and Perils of Mining Github: MSR'2014

Interviews  

Page 27: The Promises and Perils of Mining Github: MSR'2014

Random  Sta>s>cal  Sampling  

Page 28: The Promises and Perils of Mining Github: MSR'2014
Page 29: The Promises and Perils of Mining Github: MSR'2014
Page 30: The Promises and Perils of Mining Github: MSR'2014
Page 31: The Promises and Perils of Mining Github: MSR'2014
Page 32: The Promises and Perils of Mining Github: MSR'2014

"I  store  my  presenta>ons  in  github.    I  don't  need  USB  s>ck  anymore!"  

 github  interviewee  

Page 33: The Promises and Perils of Mining Github: MSR'2014
Page 34: The Promises and Perils of Mining Github: MSR'2014
Page 35: The Promises and Perils of Mining Github: MSR'2014
Page 36: The Promises and Perils of Mining Github: MSR'2014
Page 37: The Promises and Perils of Mining Github: MSR'2014
Page 38: The Promises and Perils of Mining Github: MSR'2014

•  “Any  serious  project  would  have  to  have  some  separate  infrastructure  -­‐  mailing  lists,  forums,  irc  channels  and  their  archives,  build  farms,  etc.  [...]  Thus  while  GitHub  and  all  other  project  hosts  are  used  for  collaboraDon,  they  are  not  and  can  not  be  a  complete  solu5on.”  

Page 39: The Promises and Perils of Mining Github: MSR'2014
Page 40: The Promises and Perils of Mining Github: MSR'2014

•       

•  In  sample  projects:  – 44%  of  pull-­‐requests  marked  as  merged  

– Depending  on  sample  •  Extra  19  to  42%  of  pull-­‐requests  have  been  merged,    but  are  not  marked  as  merged  in  github  

Page 41: The Promises and Perils of Mining Github: MSR'2014

A  New  Hope  

Page 42: The Promises and Perils of Mining Github: MSR'2014