How the POPFile project runs...
on
Funding Open Source?
·
· Score: 5, Interesting
Just my little experience from the POPFile project.
1. Money
All the money I get from POPFile is through donations made through Amazon.com or PayPal. This brings in enough money to keep me *very* interested in the project, although it does not cover the rent at this point. However as POPFile's user base grew I saw donations grow with I would estimate around 10% of users donating an average of $20.
2. Marketing
I spend no money on marketing, but I am *very* nice to any press that want information on POPFile. They are your friend since they will advertise your product for free if you can get them to write about it. The key to getting them to write about it is to think of the "hook" that they will use. All writers have a "hook" or key idea in the story that they are writing about. If you can relate your product to a hook then you can get them to write about it. In the case of POPFile the hook is spam. Although POPFile is designed for generic email processing it's good at fighting spam too and so I work with writers who deal with the spam problem and they in turn mention POPFile.
On a related note I'd say that the free (as in beer) nature of free (as in speech) software is also a big plus for journalists. There's nothing like recommending a product to their readers that is free.
3. Be Nice To People
Word of mouth is very important to any product (commercial or not) and that means answering every single email you get. I read every message in the POPFile forums and answer every email sent directly to me. This is vital because people then realize that the community around POPFile is welcoming and they feel more comfortable using the tool.
As ActiveState is a sponsor of the Open Source Awards their Active Awards will disappear and ActiveState will be giving the money towards an Open Source Award for a programming language.
It's worth knowing that the people on the committee are the group that responded fast enough to invitations to join. The full list of people was long and you can expect to see the committee grow.
Nominations for the awards are going to be from the public and come through me via the OSI web site. I don't have any specific tie to an OSS project (other than my own... POPFile) nor do I favor strongly any OS or language.
The idea behind the committee was to have people who've been around a long time in OSS. The more people like that the merrier.
Like I give a shit about the karma. And your joke about my last name is *so* funny, I've never heard thast before.
I cut the press release up into pieces so that (a) you didn't read all the press release BS and (b) so that it was digestable by the Slashdot masses who rarely read articles.
That's a good point and there's nothing in the Open Source Awards that says that/. couldn't get an award. In fact the Special Awards (Silver) are designed for whatever purpose the electors want. So they could give an award to/. or some other entity that has helped a lot. It's really a catch all.
Jeremy Allison, one of the lead developers on the Samba Team, a group of programmers developing an open source Windows(tm) compatible file and print server product for UNIX systems. Allison handles the release engineering and the co-ordination of Samba development efforts worldwide and acts as a corporate liaison to companies using the Samba code commercially.
Larry Augustin, a venture partner at Azure Capital Partners where he specializes in software, systems, and related IT infrastructure technologies. He currently serves on the boards of directors of VA Software Corporation (as chairman), the Open Source Development Lab, Linux International, and the Free Standards Group. Previously he was conference chairman for LinuxWorld Conference and Expo, and served on the conference advisory board. Augustin has appeared as a regular columnist in Linux Magazine, has written numerous articles, and is the author of "Hardware Design and Simulation in VAL/VHDL," published by Kluwer Academic Publishers.
Jim Gettys, a member of HP Labs' Cambridge Research Lab, currently working on making open source systems safe on handheld computers. He helped found the handhelds.org community. In 1984, Gettys started the X Window System that forms the base technology of the Linux and UNIX desktops, on which Gnome and KDE are based. Gettys worked at W3C on loan from Compaq Computer Corporation's Industry Standards and Consortia group from 1995-1999. He is the editor of the HTTP/1.1 specification (now an IETF Draft Standard).
Dr. Marshall Kirk McKusick, author, consultant, and professor on UNIX- and BSD-related subjects. While at the University of California at Berkeley, he implemented the 4.2BSD fast file system and was the research computer scientist at the Berkeley Computer Systems Research Group (CSRG), overseeing the development and release of 4.3BSD and 4.4BSD. He has been a strong advocate for the open-source movement since its inception in the mid 1980s.
Keith Packard, developer of open source software since 1986. Packard has focused on the X Window System since 1987, designing and executing large parts of the current implementation. He is currently employed by HP as a member of the Cambridge Research Laboratory working on pervasive and mobile computing. In 1999, he received a Usenix Lifetime Achievement award for his work on the X Window System.
Eric S. Raymond, observer-participant anthropologist in the Internet hacker culture. His research has helped explain the decentralized open-source model of software development that has proven so effective in the evolution of the Internet. His own software projects include one of the Internet's most widely-used email transport programs. Raymond is the co-founder of the Open Source Awards.
Guido van Rossum, creator of Python, one of the major free scripting languages. He created Python in the early 1990s at the National Research Institute for Mathematics and Computer Science in the Netherlands, and is still actively involved in the development of the language. van Rossum recently accepted a position at Elemental Security, a start-up founded by Dan Farmer.
That would be up to the person who receives the award. I could imagine that on my project if I received a $500 award I would split it with the other people who make the most contribution to the project. OSI itself is unlikely to try to make that determination for a project leader.
John.
That's one reason that we contacted VC firms. They are smart enough to realize that knowing the smart engineers (many of whom are working on OSS projects) is the way that they are going to make future money.
John.
The Grand Master Award: This award will be given to persons with an outstanding record of contributions to the open-source and Internet cultures. Ideal candidates will have a record not only of technical excellence but of community leadership and service. Along with the recognition as Grand Master, the recipient will receive $10,000 and an invitation to serve as an elector on the collegium that issues the awards.
Merit Awards: These awards will be given four times per year for work on specific open-source or network-service projects. Recipients will be recognized at the annual event and will receive a cash award of $500.
The Special Award - These awards may occasionally be conferred at the Awards Committee's discretion as a way of recognizing praiseworthy projects or conduct not covered by the existing regular categories and experimenting with new categories. Recipients will be recognized at the annual event and will receive a cash award of $1500.
You got it. That's the whole point of POPFile...
Now, if Apple would give me a job I'd happily make Mail.app the coolest mail program out there by adding mutli-bucket Bayesian mail sorting.
John.
Do you have some evidence that points to this? I've been tracking statistics from actual POPFile users for sometime and publishing them on the POPFile home page and there is only a *very* slight decrease in accuracy as the number of classification buckets increases.
John.
Do you have some evidence to back up the claim that word pairs are more accurate than individual words? This is commonly quoted as a "better" approach but most of the research on Bayesian text classification shows that for email classification it is very close to the same performance as more complex k-word classifiers and it's faster.
John.
I agree with you and we are planning to get to that ASAP. There's some underlying work we need to do on performance first (that's planned for v0.20.0) and then we'll have the foundation for multiusers, pretty much as you describe.
If anyone out there wants to write an IMAP module (subclass of Proxy::Proxy) then I'd be very happy to accept it.
John.
Slashdot Mirror
Just my little experience from the POPFile project.
1. Money
All the money I get from POPFile is through donations made through Amazon.com or PayPal. This brings in enough money to keep me *very* interested in the project, although it does not cover the rent at this point. However as POPFile's user base grew I saw donations grow with I would estimate around 10% of users donating an average of $20.
2. Marketing
I spend no money on marketing, but I am *very* nice to any press that want information on POPFile. They are your friend since they will advertise your product for free if you can get them to write about it. The key to getting them to write about it is to think of the "hook" that they will use. All writers have a "hook" or key idea in the story that they are writing about. If you can relate your product to a hook then you can get them to write about it. In the case of POPFile the hook is spam. Although POPFile is designed for generic email processing it's good at fighting spam too and so I work with writers who deal with the spam problem and they in turn mention POPFile.
On a related note I'd say that the free (as in beer) nature of free (as in speech) software is also a big plus for journalists. There's nothing like recommending a product to their readers that is free.
3. Be Nice To People
Word of mouth is very important to any product (commercial or not) and that means answering every single email you get. I read every message in the POPFile forums and answer every email sent directly to me. This is vital because people then realize that the community around POPFile is welcoming and they feel more comfortable using the tool.
John.
As ActiveState is a sponsor of the Open Source Awards their Active Awards will disappear and ActiveState will be giving the money towards an Open Source Award for a programming language.
John.
It's worth knowing that the people on the committee are the group that responded fast enough to invitations to join. The full list of people was long and you can expect to see the committee grow.
Nominations for the awards are going to be from the public and come through me via the OSI web site. I don't have any specific tie to an OSS project (other than my own... POPFile) nor do I favor strongly any OS or language.
The idea behind the committee was to have people who've been around a long time in OSS. The more people like that the merrier.
John.
Like I give a shit about the karma. And your joke about my last name is *so* funny, I've never heard thast before.
I cut the press release up into pieces so that (a) you didn't read all the press release BS and (b) so that it was digestable by the Slashdot masses who rarely read articles.
John.
We are going to open up a nominations mailing list through the OSI web site that I'll be moderating and will look for nominations like this.
Thanks,
John.
Just working on an OSS project is a "Survivor-type contest" because you are constantly battling for your own survival.
John.
That's a good point and there's nothing in the Open Source Awards that says that /. couldn't get an award. In fact the Special Awards (Silver) are designed for whatever purpose the electors want. So they could give an award to /. or some other entity that has helped a lot. It's really a catch all.
John.
Jeremy Allison, one of the lead developers on the Samba Team, a group of programmers developing an open source Windows(tm) compatible file and print server product for UNIX systems. Allison handles the release engineering and the co-ordination of Samba development efforts worldwide and acts as a corporate liaison to companies using the Samba code commercially.
Larry Augustin, a venture partner at Azure Capital Partners where he specializes in software, systems, and related IT infrastructure technologies. He currently serves on the boards of directors of VA Software Corporation (as chairman), the Open Source Development Lab, Linux International, and the Free Standards Group. Previously he was conference chairman for LinuxWorld Conference and Expo, and served on the conference advisory board. Augustin has appeared as a regular columnist in Linux Magazine, has written numerous articles, and is the author of "Hardware Design and Simulation in VAL/VHDL," published by Kluwer Academic Publishers.
Jim Gettys, a member of HP Labs' Cambridge Research Lab, currently working on making open source systems safe on handheld computers. He helped found the handhelds.org community. In 1984, Gettys started the X Window System that forms the base technology of the Linux and UNIX desktops, on which Gnome and KDE are based. Gettys worked at W3C on loan from Compaq Computer Corporation's Industry Standards and Consortia group from 1995-1999. He is the editor of the HTTP/1.1 specification (now an IETF Draft Standard).
Dr. Marshall Kirk McKusick, author, consultant, and professor on UNIX- and BSD-related subjects. While at the University of California at Berkeley, he implemented the 4.2BSD fast file system and was the research computer scientist at the Berkeley Computer Systems Research Group (CSRG), overseeing the development and release of 4.3BSD and 4.4BSD. He has been a strong advocate for the open-source movement since its inception in the mid 1980s.
Keith Packard, developer of open source software since 1986. Packard has focused on the X Window System since 1987, designing and executing large parts of the current implementation. He is currently employed by HP as a member of the Cambridge Research Laboratory working on pervasive and mobile computing. In 1999, he received a Usenix Lifetime Achievement award for his work on the X Window System.
Eric S. Raymond, observer-participant anthropologist in the Internet hacker culture. His research has helped explain the decentralized open-source model of software development that has proven so effective in the evolution of the Internet. His own software projects include one of the Internet's most widely-used email transport programs. Raymond is the co-founder of the Open Source Awards.
Guido van Rossum, creator of Python, one of the major free scripting languages. He created Python in the early 1990s at the National Research Institute for Mathematics and Computer Science in the Netherlands, and is still actively involved in the development of the language. van Rossum recently accepted a position at Elemental Security, a start-up founded by Dan Farmer.
That would be up to the person who receives the award. I could imagine that on my project if I received a $500 award I would split it with the other people who make the most contribution to the project. OSI itself is unlikely to try to make that determination for a project leader. John.
That's one reason that we contacted VC firms. They are smart enough to realize that knowing the smart engineers (many of whom are working on OSS projects) is the way that they are going to make future money. John.
The Open Source Awards categories include:
The Grand Master Award: This award will be given to persons with an outstanding record of contributions to the open-source and Internet cultures. Ideal candidates will have a record not only of technical excellence but of community leadership and service. Along with the recognition as Grand Master, the recipient will receive $10,000 and an invitation to serve as an elector on the collegium that issues the awards.
Merit Awards: These awards will be given four times per year for work on specific open-source or network-service projects. Recipients will be recognized at the annual event and will receive a cash award of $500.
The Special Award - These awards may occasionally be conferred at the Awards Committee's discretion as a way of recognizing praiseworthy projects or conduct not covered by the existing regular categories and experimenting with new categories. Recipients will be recognized at the annual event and will receive a cash award of $1500.
Seems a pity to fork the POPFile code just to get IMAP support when I'm happy to include it. How would you suggest intercepting the mail? John.
You got it. That's the whole point of POPFile... Now, if Apple would give me a job I'd happily make Mail.app the coolest mail program out there by adding mutli-bucket Bayesian mail sorting. John.
Do you have some evidence that points to this? I've been tracking statistics from actual POPFile users for sometime and publishing them on the POPFile home page and there is only a *very* slight decrease in accuracy as the number of classification buckets increases. John.
Do you have some evidence to back up the claim that word pairs are more accurate than individual words? This is commonly quoted as a "better" approach but most of the research on Bayesian text classification shows that for email classification it is very close to the same performance as more complex k-word classifiers and it's faster. John.
I agree with you and we are planning to get to that ASAP. There's some underlying work we need to do on performance first (that's planned for v0.20.0) and then we'll have the foundation for multiusers, pretty much as you describe. If anyone out there wants to write an IMAP module (subclass of Proxy::Proxy) then I'd be very happy to accept it. John.