14:29:45 #startmeeting 14:29:45 Meeting started Wed Aug 12 14:29:45 2015 UTC. The chair is mizmo. Information about MeetBot at http://wiki.debian.org/MeetBot. 14:29:45 Useful Commands: #action #agreed #halp #info #idea #link #topic. 14:29:51 #topic 10 ideas that are killing open source 14:30:00 i took notes on the actual preso we're in question mode now 14:30:23 "I think it's great.... theres a lot of technical reasons to not floss. what i don't see a lot of emphasis on is the importance of community in this process. For example look at the issues OpenSSL has.... bash... gits" 14:32:07 sorry kernel crash 14:32:23 langdon: do you know any examples of companies that are successfully doing the n+1 release model? 14:32:33 pingou: vitiozo 14:32:48 they have a proprietary project and a floss project... the proprietary project has more features 14:32:50 langdon - what license? 14:32:54 pingou - probably dual license 14:32:58 langdon - something custom? 14:33:09 audience member - ghostscript used to be like that - latest vesrion was proprietary, n-1 was free 14:33:13 remy - mongo? 14:33:27 another audience member - vesrion +1 thing, then gpl 14:33:38 if people sending you gpl, you can't put them to the proprietary version so its a little dnagerous 14:33:47 pingou - not sure which license they use 14:34:03 langdon - i have a friend whose company woudl go with this model but i didn't know examples of other cos doing it and what licenses they used 14:34:13 gnokii - also possibility to give source code later to public, red hat used that a long time ago 14:34:19 pingou - which android is using now 14:34:39 langdon - yeh its not so much to prove the success, just to show them a model that works so they see something that works 14:34:48 gnokii - i think owncloud does this too, have some closed parts just for customers 14:34:52 pingou: thank you everyone for coming 14:34:59 okay the notes from the beginning of the session i have 14:35:02 i will dump here :) 14:38:03 ok large dump pending 14:38:04 Why people don't open source things 14:38:12 Why people don't open source things 14:38:18 • (0) i do not own the code (not much we can do) 14:38:18 ∘ try to work something out with owner if you can 14:38:26 category 1 14:38:26 • (1) i do not know how 14:38:26 ∘ resources available - books, websites, ppl, conferences, community mgr / teams 14:38:26 ∘ start from the start - make the sources public 14:38:26 ∘ producingoss.com, website and free book 14:38:32 category 2 14:38:32 not correct arguments... fake reasons. 14:38:39 • (2) it is not worth it (small project, small idea, anybody can do it) 14:38:41 ∘ how do you know? 14:38:43 ∘ maybe for you anybody can do it 14:38:47 • (3) there is no point (from an infra pov... nobody can run it, so specific to our infra nobody can replicate) 14:38:50 ∘ good example: COPR - need a complete cloud infra 14:38:52 ∘ bad example: transifex... moved to a closed model bc nobody was running it, so what's the point of making open 14:38:55 ∘ if you don't run the app, it does not mean it prevents contribution... eg unit tests.... 14:38:57 • (4) too much work 14:38:59 ∘ not necessarily - up to you 14:39:01 ∘ how far you want to go is up to you - can just dump a tarball somewhere, can blog it, or can go far and build a community around it 14:39:04 category 3 14:39:06 hubris reasons.... larry wall, author of perl, three great virtues of programmer - laziness, impatience, and hubris 14:39:09 laziness - computer does for it 14:39:11 impatience - make it fast 14:39:13 hubris - makes you write and maintain things people won't say bad things about 14:39:17 pingou's alternative: hubris - the quality that makes youw rite and maintain a welcoming community around your project 14:39:20 cogito ergo sum - i think therefore i am 14:39:22 but we say - i do therefore i am (people being judged by what they do) 14:39:24 eg you're a great programmer not bc of how you solved the problem, but judged by how the code looks like. but the code is the technique, whats important is the idea 14:39:27 kaplan moss pycon 2015 montreal... bit.ly/jkmpykey15 (youtube) 14:39:29 two kinds of programmers in IT.... the very very bad, and the very very good 14:39:31 -- shouldn't be, should follow the normal distribution... so the very very bad vs very very good is perception not reality 14:39:34 how you think about this is important 14:39:36 you are just as good as the next guy just as bad as the last one 14:39:38 just open source the code - then rewrite it, then build the community 14:39:40 ppl afraid if they publish bad code and recruiter sees they feel like they would hurt their chances... but if you can explain to recruiter the history of the work can be an advantage if you point out weaknesses and ways needed to improve it so you demonstrate how to solve problem - writing good code is omsething you can learn just technique. its the good thinking that's important that you can demonstrate with an open project 14:39:47 • (5) no docs / tests 14:39:49 • (6) i want to clean te code first 14:39:51 • (7) i just need this one more thing 14:39:53 category 4 14:39:55 business reasons 14:39:57 • (8) someone will steal my idea 14:39:59 ∘ license! gpl, agpl 14:40:01 ∘ tldrlegal.com 14:40:03 ‣ can cannot must listed out for each license 14:40:05 • (9) what is my business model then 14:40:07 ∘ FOSS != free beer 14:40:09 ‣ Android, RHEL, emacs, Transifex (before) 14:40:11 ∘ experts 14:40:13 ∘ support/teaching 14:40:17 ∘ seven open source biz strategies john koening bit.ly/ossbizmodels 14:40:19 • (10) it is business critical 14:40:21 ∘ is all of it biz critical, or can you open source the parts that are not biz critical- will eventually be less work for your devs to open some stuffe 14:40:24 ∘ is "secret sauce" a myth? 14:40:26 you can keep the secret sauce but open source the components around it 14:40:28 okay taht's what i got :) 14:40:30 #endmeeting