Changes

key for the success - to carefully filter out wannabes and select really good students by giving them small but real tasks (probably some bugs to fix)

 

and second important thing - student should have some simple plan and mentor should ping him periodically

 

ofc, our wallet was quite small compared to Google's one, but it was still useful, and finally we've used about $700 per student

and one more - using spamming on global (Hacker News, reddit, slashdot, some mailing lists, subject mailing lists) and on local (probably on foreign languages) - is a key for a broad attention and paycheck

 

ofc by spamming I mean writing short still useful articles

and using this you probably can cover even more audience than GSoC

as a bonus - you can invite all ppl, not just only students

also even if it fails - it is useful expierence and raises your chances on the next GSoC

* Carefully filter out wannabes and select really good students by giving them small but real tasks (probably some bugs to fix)

* Student should have some simple plan and mentor should ping him periodically

* ofc, our wallet was quite small compared to Google's one, but it was still useful, and finally we've used about $700 per student

* using spamming on global (Hacker News, reddit, slashdot, some mailing lists, subject mailing lists) and on local (probably on foreign languages) - is a key for a broad attention and paycheck

** ofc by spamming I mean writing short still useful articles

* and using this you probably can cover even more audience than GSoC

* as a bonus - you can invite all people, not just only students

* also even if it fails - it is useful experience and raises your chances on the next GSoC