Re: increasing GSoC visibility within the ASF
On 11 April 2013 23:30, Suresh Marru wrote: > On Apr 11, 2013, at 12:07 PM, Luciano Resende > wrote: > > One of the issues is that we keep using GSoC + year as tag, so we pretty > > much discard all the previous years project ideas that are still open. I > > think using something generic like GSoC would make our project idea look > > bigger without much overhead on mentors to prepare year over year. > > I see that in 2013 ideas, there are JIRA's created all the way from 2005 > through 2013. I worry if you automatically carry over the issues to next > year, we might end up with some unattended projects. What if the mentor > (who proposed it previous year) does not have time or interest or the > project is no longer relevant? I think it is not too much to expect from > the mentor to go to the previous year JIRA's and change the tag from > gsoc2012 to gsoc2013. If students are commenting on zombie ideas and do not > get a response, that might not send a good signal. If a mentor is not > having 30 minutes to create or modify a JIRA should we expect them to > mentor a student, review and accept code and provide feedback? > > Another good point (and now you mention it it's why we chose year tags originally). Perhaps the thing to do is add some documentation to the initial email to PMCs saying "why not search for GSOC2012 and, if the item is still relevant, change it to GSOC 2013 - if it's not then maybe close it as a but of spring cleaning" Ross > > -- > Ross Gardler (@rgardler) > Programme Leader (Open Development) > OpenDirective http://opendirective.com > >
Re: increasing GSoC visibility within the ASF
On Apr 11, 2013, at 12:07 PM, Luciano Resende wrote: > One of the issues is that we keep using GSoC + year as tag, so we pretty > much discard all the previous years project ideas that are still open. I > think using something generic like GSoC would make our project idea look > bigger without much overhead on mentors to prepare year over year. I see that in 2013 ideas, there are JIRA's created all the way from 2005 through 2013. I worry if you automatically carry over the issues to next year, we might end up with some unattended projects. What if the mentor (who proposed it previous year) does not have time or interest or the project is no longer relevant? I think it is not too much to expect from the mentor to go to the previous year JIRA's and change the tag from gsoc2012 to gsoc2013. If students are commenting on zombie ideas and do not get a response, that might not send a good signal. If a mentor is not having 30 minutes to create or modify a JIRA should we expect them to mentor a student, review and accept code and provide feedback? Suresh
Re: increasing GSoC visibility within the ASF
On Thu, Apr 11, 2013 at 2:12 PM, Ross Gardler wrote: > Good point Luciano - we need to move to GSOC only - or better still > something not linked to GSoC, e.g. mentored. A search and a batch update > will do the job nicely. We can adjust our search to include GSoC, GSoC2013 > and mentored to catch those who didn't do the update. > > Ross > > I think we have used mentor in the past (instead of mentored)... We could do the batch update, but I'm afraid new ideas might come with the gsoc2013 and get ignored, so we need to monitor it if the search don't include that. -- Luciano Resende http://people.apache.org/~lresende http://twitter.com/lresende1975 http://lresende.blogspot.com/
Re: GSoC informational talk this thursday in Berlin
Great job Uli Ross On 11 April 2013 19:05, Ulrich Stärk wrote: > Short summary: > > about 30 students attended despite the short notice, some want to apply > now. The greatest hurdles to > overcome seem to be fears of not being good enough and questions regarding > their student status that > is tightly coupled to their income (it's complicated here in Germany). > > More details soon, I'm exhausted. > > Uli > > On 09.04.2013 21:49, Ulrich Stärk wrote: > > Hey folks, > > > > just a short notice: I'm holding an infosession on Google Summer of Code > 2013 this Thursday > > (2013-04-11) 18:00 at the computer science institute of Freie > Universität Berlin, Takustr 9, 14195 > > Berlin. If you know of any local student (computer science, mathematics, > physics, whatever) who > > might want to participate in GSoC this year, please send them ;) > > > > Cheers, > > > > Uli > > > -- Ross Gardler (@rgardler) Programme Leader (Open Development) OpenDirective http://opendirective.com
Re: increasing GSoC visibility within the ASF
Good point Luciano - we need to move to GSOC only - or better still something not linked to GSoC, e.g. mentored. A search and a batch update will do the job nicely. We can adjust our search to include GSoC, GSoC2013 and mentored to catch those who didn't do the update. Ross On 11 April 2013 17:07, Luciano Resende wrote: > On Thu, Apr 11, 2013 at 7:38 AM, Ross Gardler >wrote: > > > On 11 April 2013 14:23, Rich Bowen wrote: > > > > > > > > On Apr 10, 2013, at 1:39 PM, Mark Struberg wrote: > > > > > > > + it's also pretty hard to spin of small enough tasks which can be > done > > > in a GSoC project. Most of our projects need some really in-depth > > knowledge > > > prior to hacking a smallish task :/ > > > > > > > > > > > > Looking forward to next year, it would be good for us to encourage > > > projects to have a GSoC tag in their ticket tracker to identify issues > > that > > > would be good candidates for students. The folks at OpenHatch recommend > > > clearly identifying tickets as entry-level, and then, rather than > always > > > cleaning up the so-called "low-hanging fruit", leaving them for > > entry-level > > > people. Such an approach might be employed, say, 3-6 months out from > > GSoC, > > > to start to identify good student projects. > > > > > > > We already do that. We could be more proactive about reminding projects > to > > mark issues throughout the year. > > > > > > > > > > Granted, this leaves things undone, but that's probably ok, if our > focus > > > is indeed community over code. > > > > > > > No problem if an issue is tagged GSoC but gets closed before GSoC comes > > along. > > > > Ross > > > > > One of the issues is that we keep using GSoC + year as tag, so we pretty > much discard all the previous years project ideas that are still open. I > think using something generic like GSoC would make our project idea look > bigger without much overhead on mentors to prepare year over year. > > -- > Luciano Resende > http://people.apache.org/~lresende > http://twitter.com/lresende1975 > http://lresende.blogspot.com/ > -- Ross Gardler (@rgardler) Programme Leader (Open Development) OpenDirective http://opendirective.com
Re: increasing GSoC visibility within the ASF
I think this is a good idea if you have the time. I'm certain the OpenMeetings project will be very happy to provide a teleconferencing solution. Ross On 11 April 2013 15:54, Suresh Marru wrote: > Lots of good thoughts already, and I agree with Ross it is the quality and > not quantity. > > But I think not all projects have the awareness and what they can get out > of GSoC. There is a lot of good documentation as Uli and others have > pointed out. PMC's have to be motivated to look at them. Email is good for > most of the asynchronous communication needs, but my 2 cents will be to > conduct an open forum webeminar and reach it out to all committers and > pmcs. My experience if some one can ask questions directly it will > bootstrap better interest than emails notifications. If such thing will be > useful, I am willing to organize and coordinate. > > Suresh > > On Apr 10, 2013, at 1:26 PM, Ulrich Stärk wrote: > > > Folks, > > > > I perceive a low interest of our projects in GSoC. The list of projects > that submitted project ideas > > that I compiled for Sally contained 33 entries of which some are > subprojects I believe. With 138 > > PMCs plus 35 podlings, this is less than one fifth of our projects. > > > > We only had 34 ideas one week before our application was due. > > > > I run into committers that are not members of their projects PMCs who > are eager to be mentors but > > have no clue about what's going on because nobody from the PMC forwarded > my emails to their dev lists. > > > > So the problem seems twofold: no interest and not reaching the right > people. The latter could be > > improved by simply sending to committers@ instead of pmcs@ but it's the > first that worries me. > > > > I believe GSoC and every other opportunity to attract new contributors > to our projects should be a > > key priority of our PMCs. Apparently it's not, for whatever reasons. I > could think of missing > > cycles, indifference, and wrong priorities. > > > > So what could we do to increase awareness for the opportunities GSoC > offers and that this program is > > important to the foundation? Write more emails? Ask the board to mandate > a section in board reports > > detailing the project's GSoC endeavors? Any ideas? > > > > Uli > > -- Ross Gardler (@rgardler) Programme Leader (Open Development) OpenDirective http://opendirective.com
Re: GSoC informational talk this thursday in Berlin
Short summary: about 30 students attended despite the short notice, some want to apply now. The greatest hurdles to overcome seem to be fears of not being good enough and questions regarding their student status that is tightly coupled to their income (it's complicated here in Germany). More details soon, I'm exhausted. Uli On 09.04.2013 21:49, Ulrich Stärk wrote: > Hey folks, > > just a short notice: I'm holding an infosession on Google Summer of Code 2013 > this Thursday > (2013-04-11) 18:00 at the computer science institute of Freie Universität > Berlin, Takustr 9, 14195 > Berlin. If you know of any local student (computer science, mathematics, > physics, whatever) who > might want to participate in GSoC this year, please send them ;) > > Cheers, > > Uli >
Re: increasing GSoC visibility within the ASF
On Apr 11, 2013, at 12:07 PM, Luciano Resende wrote: > One of the issues is that we keep using GSoC + year as tag, so we pretty > much discard all the previous years project ideas that are still open. I > think using something generic like GSoC would make our project idea look > bigger without much overhead on mentors to prepare year over year. Or, both: 'GSoC' = This would be a great idea for GSoC, some day 'GSoC2013' = This is actually going to be worked on as part of GSoC 2013 -- Rich Bowen rbo...@rcbowen.com Shosholoza
Re: increasing GSoC visibility within the ASF
On Thu, Apr 11, 2013 at 7:38 AM, Ross Gardler wrote: > On 11 April 2013 14:23, Rich Bowen wrote: > > > > > On Apr 10, 2013, at 1:39 PM, Mark Struberg wrote: > > > > > + it's also pretty hard to spin of small enough tasks which can be done > > in a GSoC project. Most of our projects need some really in-depth > knowledge > > prior to hacking a smallish task :/ > > > > > > > > Looking forward to next year, it would be good for us to encourage > > projects to have a GSoC tag in their ticket tracker to identify issues > that > > would be good candidates for students. The folks at OpenHatch recommend > > clearly identifying tickets as entry-level, and then, rather than always > > cleaning up the so-called "low-hanging fruit", leaving them for > entry-level > > people. Such an approach might be employed, say, 3-6 months out from > GSoC, > > to start to identify good student projects. > > > > We already do that. We could be more proactive about reminding projects to > mark issues throughout the year. > > > > > > Granted, this leaves things undone, but that's probably ok, if our focus > > is indeed community over code. > > > > No problem if an issue is tagged GSoC but gets closed before GSoC comes > along. > > Ross > > One of the issues is that we keep using GSoC + year as tag, so we pretty much discard all the previous years project ideas that are still open. I think using something generic like GSoC would make our project idea look bigger without much overhead on mentors to prepare year over year. -- Luciano Resende http://people.apache.org/~lresende http://twitter.com/lresende1975 http://lresende.blogspot.com/
Re: increasing GSoC visibility within the ASF
On Apr 11, 2013, at 11:14 AM, Ulrich Stärk wrote: > On 11.04.2013 16:41, Ross Gardler wrote: >> On 11 April 2013 14:26, Ulrich Stärk wrote: >> >>> On 11.04.2013 15:17, Rich Bowen wrote: On Apr 10, 2013, at 1:39 PM, Mark Struberg wrote: > In this light, 33 is not that bad a number. How does that compare to past years? >>> >>> 27 in 2012, so about the same (the 33 also contain some sub-projects). But >>> at the same time the >>> number of our TLPs and podlings grew so the ratio of participating >>> projects is lower than last year. >>> >> >> As I find myself saying every year. It's about quality not quantity. > > It's both. I find it an alarming sign that our numbers don't increase with > our size. > >> >> Our pass rate in the ASF is higher than the programme overall. Part of the >> reason for this is that we spend a great deal of time making sure the right >> students get chosen, but also because our mentors are very committed. We >> state that we expect around 5 hours a week from mentors. That is a >> significant commitment but providing that time does give us a higher >> success rate. >> >> Personally, I wouldn't consider 33 projects low. We only get around 35-40 >> slots anyway. > > I need to correct you. We got 55 slots in 2011 of which 15 were returned and > I believe almost the > same number last year and again we had to return a significant number because > we didn't have the > mentors and participating projects to fill them. Hi Uli, I am not trying to be onerous on your to hunt statistics from previous years but thinking out loud. For 2013 there are 173 project ideas, I am assuming there is a many to one mapping of mentors to ideas. Since there is a recommendation to have one to one mapping of mentor to projects (with exceptions allowed under certain conditions), are we running into an issues where multiple projects are getting good proposals and mentors are choosing one and leaving out rest? If so do we want to encourage already committed mentors to recruit others in PMC to become mentors? I know you have repeatedly insisted on this, but wondering if something more should be done in the area? Roughly speaking 33 PMC's (PPMCs included) prosed these ideas. Based on previous years experience, how much of percentage of these ideas are not attracting students? If so do we want to educate on what makes a good gsoc project? Suresh > > Uli
Re: increasing GSoC visibility within the ASF
On 11.04.2013 16:41, Ross Gardler wrote: > On 11 April 2013 14:26, Ulrich Stärk wrote: > >> On 11.04.2013 15:17, Rich Bowen wrote: >>> >>> On Apr 10, 2013, at 1:39 PM, Mark Struberg wrote: >>> In this light, 33 is not that bad a number. >>> >>> >>> How does that compare to past years? >>> >> >> 27 in 2012, so about the same (the 33 also contain some sub-projects). But >> at the same time the >> number of our TLPs and podlings grew so the ratio of participating >> projects is lower than last year. >> > > As I find myself saying every year. It's about quality not quantity. It's both. I find it an alarming sign that our numbers don't increase with our size. > > Our pass rate in the ASF is higher than the programme overall. Part of the > reason for this is that we spend a great deal of time making sure the right > students get chosen, but also because our mentors are very committed. We > state that we expect around 5 hours a week from mentors. That is a > significant commitment but providing that time does give us a higher > success rate. > > Personally, I wouldn't consider 33 projects low. We only get around 35-40 > slots anyway. I need to correct you. We got 55 slots in 2011 of which 15 were returned and I believe almost the same number last year and again we had to return a significant number because we didn't have the mentors and participating projects to fill them. Uli
Re: increasing GSoC visibility within the ASF
Lots of good thoughts already, and I agree with Ross it is the quality and not quantity. But I think not all projects have the awareness and what they can get out of GSoC. There is a lot of good documentation as Uli and others have pointed out. PMC's have to be motivated to look at them. Email is good for most of the asynchronous communication needs, but my 2 cents will be to conduct an open forum webeminar and reach it out to all committers and pmcs. My experience if some one can ask questions directly it will bootstrap better interest than emails notifications. If such thing will be useful, I am willing to organize and coordinate. Suresh On Apr 10, 2013, at 1:26 PM, Ulrich Stärk wrote: > Folks, > > I perceive a low interest of our projects in GSoC. The list of projects that > submitted project ideas > that I compiled for Sally contained 33 entries of which some are subprojects > I believe. With 138 > PMCs plus 35 podlings, this is less than one fifth of our projects. > > We only had 34 ideas one week before our application was due. > > I run into committers that are not members of their projects PMCs who are > eager to be mentors but > have no clue about what's going on because nobody from the PMC forwarded my > emails to their dev lists. > > So the problem seems twofold: no interest and not reaching the right people. > The latter could be > improved by simply sending to committers@ instead of pmcs@ but it's the first > that worries me. > > I believe GSoC and every other opportunity to attract new contributors to our > projects should be a > key priority of our PMCs. Apparently it's not, for whatever reasons. I could > think of missing > cycles, indifference, and wrong priorities. > > So what could we do to increase awareness for the opportunities GSoC offers > and that this program is > important to the foundation? Write more emails? Ask the board to mandate a > section in board reports > detailing the project's GSoC endeavors? Any ideas? > > Uli
Re: increasing GSoC visibility within the ASF
On 11 April 2013 14:26, Ulrich Stärk wrote: > On 11.04.2013 15:17, Rich Bowen wrote: > > > > On Apr 10, 2013, at 1:39 PM, Mark Struberg wrote: > > > >> In this light, 33 is not that bad a number. > > > > > > How does that compare to past years? > > > > 27 in 2012, so about the same (the 33 also contain some sub-projects). But > at the same time the > number of our TLPs and podlings grew so the ratio of participating > projects is lower than last year. > As I find myself saying every year. It's about quality not quantity. Our pass rate in the ASF is higher than the programme overall. Part of the reason for this is that we spend a great deal of time making sure the right students get chosen, but also because our mentors are very committed. We state that we expect around 5 hours a week from mentors. That is a significant commitment but providing that time does give us a higher success rate. Personally, I wouldn't consider 33 projects low. We only get around 35-40 slots anyway. Ross -- Ross Gardler (@rgardler) Programme Leader (Open Development) OpenDirective http://opendirective.com
Re: increasing GSoC visibility within the ASF
On 11 April 2013 14:23, Rich Bowen wrote: > > On Apr 10, 2013, at 1:39 PM, Mark Struberg wrote: > > > + it's also pretty hard to spin of small enough tasks which can be done > in a GSoC project. Most of our projects need some really in-depth knowledge > prior to hacking a smallish task :/ > > > > Looking forward to next year, it would be good for us to encourage > projects to have a GSoC tag in their ticket tracker to identify issues that > would be good candidates for students. The folks at OpenHatch recommend > clearly identifying tickets as entry-level, and then, rather than always > cleaning up the so-called "low-hanging fruit", leaving them for entry-level > people. Such an approach might be employed, say, 3-6 months out from GSoC, > to start to identify good student projects. > We already do that. We could be more proactive about reminding projects to mark issues throughout the year. > > Granted, this leaves things undone, but that's probably ok, if our focus > is indeed community over code. > No problem if an issue is tagged GSoC but gets closed before GSoC comes along. Ross > > -- > Rich Bowen > rbo...@rcbowen.com :: @rbowen > rbo...@apache.org > > > > > > > -- Ross Gardler (@rgardler) Programme Leader (Open Development) OpenDirective http://opendirective.com
Re: increasing GSoC visibility within the ASF
3 months plus community bonding period, so yes, plenty of time to become familiar with the project and product. Uli On 11.04.2013 15:25, Bertrand Delacretaz wrote: > On Wed, Apr 10, 2013 at 7:39 PM, Mark Struberg wrote: >> ...Most of our projects need some really in-depth knowledge prior to hacking >> a smallish task :/... > > GSoC is 4 months of work IIRC, should be enough for more than a > smallish task given capable students. > > -Bertrand >
Re: increasing GSoC visibility within the ASF
On Apr 11, 2013, at 9:23 AM, Rich Bowen wrote: > Looking forward to next year, it would be good for us to encourage projects > to have a GSoC tag in their ticket tracker to identify issues that would be > good candidates for students. The folks at OpenHatch recommend clearly > identifying tickets as entry-level, and then, rather than always cleaning up > the so-called "low-hanging fruit", leaving them for entry-level people. Such > an approach might be employed, say, 3-6 months out from GSoC, to start to > identify good student projects. > > Granted, this leaves things undone, but that's probably ok, if our focus is > indeed community over code. I see email on February 15th, from Ross and others [1], recommending just this, but it went to comdev, and from there didn't appear to have propagated to the various dev or pmc lists. That's the step we need to be sure to make happen so that we can catch the ear of the people on each project who are anxious to see this happen. Might be useful to identify which projects have representation on the comdev mailing list and which don't. [1] - http://markmail.org/message/tl4do4miegk4v2e7 and thereabouts -- Rich Bowen rbo...@rcbowen.com Shosholoza
Re: increasing GSoC visibility within the ASF
On 11.04.2013 15:17, Rich Bowen wrote: > > On Apr 10, 2013, at 1:39 PM, Mark Struberg wrote: > >> In this light, 33 is not that bad a number. > > > How does that compare to past years? > 27 in 2012, so about the same (the 33 also contain some sub-projects). But at the same time the number of our TLPs and podlings grew so the ratio of participating projects is lower than last year. Uli
Re: increasing GSoC visibility within the ASF
On Wed, Apr 10, 2013 at 7:39 PM, Mark Struberg wrote: > ...Most of our projects need some really in-depth knowledge prior to hacking > a smallish task :/... GSoC is 4 months of work IIRC, should be enough for more than a smallish task given capable students. -Bertrand
Re: increasing GSoC visibility within the ASF
On Apr 10, 2013, at 1:39 PM, Mark Struberg wrote: > + it's also pretty hard to spin of small enough tasks which can be done in a > GSoC project. Most of our projects need some really in-depth knowledge prior > to hacking a smallish task :/ Looking forward to next year, it would be good for us to encourage projects to have a GSoC tag in their ticket tracker to identify issues that would be good candidates for students. The folks at OpenHatch recommend clearly identifying tickets as entry-level, and then, rather than always cleaning up the so-called "low-hanging fruit", leaving them for entry-level people. Such an approach might be employed, say, 3-6 months out from GSoC, to start to identify good student projects. Granted, this leaves things undone, but that's probably ok, if our focus is indeed community over code. -- Rich Bowen rbo...@rcbowen.com :: @rbowen rbo...@apache.org
Re: increasing GSoC visibility within the ASF
On Apr 10, 2013, at 1:39 PM, Mark Struberg wrote: > In this light, 33 is not that bad a number. How does that compare to past years? -- Rich Bowen rbo...@rcbowen.com :: @rbowen rbo...@apache.org
Re: increasing GSoC visibility within the ASF
Good idea. We already have most of it at [1] and [2] but the blog won't hurt. Uli [1] http://community.apache.org/gsoc.html [2] http://community.apache.org/gsoc-admin-tasks.html On 11.04.2013 10:53, Bertrand Delacretaz wrote: > On Wed, Apr 10, 2013 at 7:26 PM, Ulrich Stärk wrote: >> ...what could we do to increase awareness for the opportunities GSoC offers >> and that this program is >> important to the foundation?... > > IMO, most of the communications that you currently send to pmcs@ could > go to http://blogs.apache.org/comdev/ - this would help make things > more visible, and help write the history of GSoC @apache (which is > useful for future GSoC admins and participants). > > If there's stuff that needs to be private, it can go to an svn > repository that only committers or PMC members can read, and/or to the > code-awards list. > > -Bertrand >
Re: increasing GSoC visibility within the ASF
On Wed, Apr 10, 2013 at 7:26 PM, Ulrich Stärk wrote: > ...what could we do to increase awareness for the opportunities GSoC offers > and that this program is > important to the foundation?... IMO, most of the communications that you currently send to pmcs@ could go to http://blogs.apache.org/comdev/ - this would help make things more visible, and help write the history of GSoC @apache (which is useful for future GSoC admins and participants). If there's stuff that needs to be private, it can go to an svn repository that only committers or PMC members can read, and/or to the code-awards list. -Bertrand
