Forgot your password?
OSS Workshop Brainwriting Session
Document Actions
Facilitator: Kevin Crowston
Scribe: Yeliz Eseryel
BRAIN WRITING INSTRUCTIONS
1) Write down an idea for a research project : Title and a sentence or two..
2) Write something additional to the idea you got based… you can write whatever you want..(this is just brain writing)… just do it!!!)
-- And if you cannot think of anything, draw a line and write a new idea. And when you're ready, just pass it along…
3) Swap your document with someone else. Read the new idea, and if you find something that's especially interesting, grab the microphone and tell everyone about it.
INTERESTING IDEAS
1) Developing a tool for Debian (ah, that's too difficult for me to summarize… sorry-- Will share the details later with you)
a. Instead of forcing your research agenda, you can hear what the developers need, and then use it for research.
b. Analyze the use patterns of Debian to understand why people adopted and why they didn't adopt Debian.
c. Any feedback to the Debian community on what part of debian is most interest for specific class of users. They want quantitative feedback from the research community. (Ah, now we know who the idea is from-- Francesco Lovergine)
d. Kevin Crowston: A lot of the attention was focused on successful projects (or packages). You can ask the people who looked at a package but stopped using it, why they stopped using it. (Tell them to fill out a survey) If people know what the data will be used for, they might be willing to fill out the survey.
e. Guoli Zhang: Look at developer as well as user point of view.
2) More research on identifying team dynamics using software analysis (such as coupling, software modularity, etc) to understand the team dynamics in successful FLOSS projects.
a. Alessandro Rossi: Also look at unsuccessful projects using case based approach.
b. Kevin Crowston: One of the problems with unsuccessful projects is that many of them are inactive projects.. Although there are some projects that have actually tried doing something but failed…Maybe have a wiki of "Failed Sourceforge Projects."
c. Also talk about what is a "failed FLOSS project". This is helpful, so that the community understands what it means to be a failing project.
i. Francesco Lovergine: Some projects may not have failed but have frozen..
ii. Filippo Lanubile: Well if projects failed, we don't have info on them?
iii. Barbara Scozzi: For inactive projects don't have any data on bug-tracking systems, then the best way is to interview with people.
iv. Luca Manzari: It would be good if the communities keep their links active even if the project failed
v. Kevin Crowston: We use waybackmachine (www.archive.org) to find archives of information.
vi. Qing Li: Projects fail at different levels, some projects fail at the code level, some project fail at the social level.. So we should consider both aspects of failure.
d. Use FLOSS ideas and apply them to Small Medium Enterprises
i. Massimiliano Gambardella: We tried to do this in Naples (Chamber of Commerce). If you have ideas, I can connect you guys to them.
ii. Kevin Crowston: In industrial district, people have many years of experience and social interaction with each other….
iii. Filippo Lanubile: There's a huge problem of trust because companies are competing.
iv. Massimiliano Gambardella: Use instututional channels.
v. Yeliz Eseryel: Benchmark against other industries.
vi. Francesco Bolici: The trust issue is against the people out of their own network. For example people make verbal agreements in church… they keep their promise, because they know that if they don't then, they will be kicked out of the network.. The issue in FLOSS is similar to this.
vii. Kevin Crowston: The difference is related to risk… If FLOSS teams have issue, they can back up, if SME's fail, they cannot go back.
viii. Kevin Crowston: How would you make this into a research question?
ix. Alessandro Rossi: Find a bunch of SME's who are facing extinction due to competition and then see if you can implement this…
x. Massimiliano Gambardella: In Naples, we're talking about it. The idea is to talk to the firms to understand what they are producing. The firms are so small that they don't have resources to link to each other. And also we don't know what they produce. Also we can talk to an institution to understand what is useful to the them… [Talk to Massimiliano if you are interested in this idea]
xi. Francesco Bolici: Knows an industrial district close to Sora which has these problems. Mirandola is another organization that I studied. They have the problem of internationalizing on their own… and they have a broker, which assures that all the partners in the industrial districts.
xii. Kevin: But there are industries in this case,…
e. Francesco L: Do it yourself information technology in USA: Open source is developed by IT managers not by management. So the management doesn't understand what's going on.
1) Write down an idea for a research project : Title and a sentence or two..
2) Write something additional to the idea you got based… you can write whatever you want..(this is just brain writing)… just do it!!!)
-- And if you cannot think of anything, draw a line and write a new idea. And when you're ready, just pass it along…
3) Swap your document with someone else. Read the new idea, and if you find something that's especially interesting, grab the microphone and tell everyone about it.
INTERESTING IDEAS
1) Developing a tool for Debian (ah, that's too difficult for me to summarize… sorry-- Will share the details later with you)
a. Instead of forcing your research agenda, you can hear what the developers need, and then use it for research.
b. Analyze the use patterns of Debian to understand why people adopted and why they didn't adopt Debian.
c. Any feedback to the Debian community on what part of debian is most interest for specific class of users. They want quantitative feedback from the research community. (Ah, now we know who the idea is from-- Francesco Lovergine)
d. Kevin Crowston: A lot of the attention was focused on successful projects (or packages). You can ask the people who looked at a package but stopped using it, why they stopped using it. (Tell them to fill out a survey) If people know what the data will be used for, they might be willing to fill out the survey.
e. Guoli Zhang: Look at developer as well as user point of view.
2) More research on identifying team dynamics using software analysis (such as coupling, software modularity, etc) to understand the team dynamics in successful FLOSS projects.
a. Alessandro Rossi: Also look at unsuccessful projects using case based approach.
b. Kevin Crowston: One of the problems with unsuccessful projects is that many of them are inactive projects.. Although there are some projects that have actually tried doing something but failed…Maybe have a wiki of "Failed Sourceforge Projects."
c. Also talk about what is a "failed FLOSS project". This is helpful, so that the community understands what it means to be a failing project.
i. Francesco Lovergine: Some projects may not have failed but have frozen..
ii. Filippo Lanubile: Well if projects failed, we don't have info on them?
iii. Barbara Scozzi: For inactive projects don't have any data on bug-tracking systems, then the best way is to interview with people.
iv. Luca Manzari: It would be good if the communities keep their links active even if the project failed
v. Kevin Crowston: We use waybackmachine (www.archive.org) to find archives of information.
vi. Qing Li: Projects fail at different levels, some projects fail at the code level, some project fail at the social level.. So we should consider both aspects of failure.
d. Use FLOSS ideas and apply them to Small Medium Enterprises
i. Massimiliano Gambardella: We tried to do this in Naples (Chamber of Commerce). If you have ideas, I can connect you guys to them.
ii. Kevin Crowston: In industrial district, people have many years of experience and social interaction with each other….
iii. Filippo Lanubile: There's a huge problem of trust because companies are competing.
iv. Massimiliano Gambardella: Use instututional channels.
v. Yeliz Eseryel: Benchmark against other industries.
vi. Francesco Bolici: The trust issue is against the people out of their own network. For example people make verbal agreements in church… they keep their promise, because they know that if they don't then, they will be kicked out of the network.. The issue in FLOSS is similar to this.
vii. Kevin Crowston: The difference is related to risk… If FLOSS teams have issue, they can back up, if SME's fail, they cannot go back.
viii. Kevin Crowston: How would you make this into a research question?
ix. Alessandro Rossi: Find a bunch of SME's who are facing extinction due to competition and then see if you can implement this…
x. Massimiliano Gambardella: In Naples, we're talking about it. The idea is to talk to the firms to understand what they are producing. The firms are so small that they don't have resources to link to each other. And also we don't know what they produce. Also we can talk to an institution to understand what is useful to the them… [Talk to Massimiliano if you are interested in this idea]
xi. Francesco Bolici: Knows an industrial district close to Sora which has these problems. Mirandola is another organization that I studied. They have the problem of internationalizing on their own… and they have a broker, which assures that all the partners in the industrial districts.
xii. Kevin: But there are industries in this case,…
e. Francesco L: Do it yourself information technology in USA: Open source is developed by IT managers not by management. So the management doesn't understand what's going on.
