On Mon, Jan 12, 2015 at 11:19 PM, Niall Douglas wrote: [8<]

I think the quality of your GSoC last year qualifies you as competent in C++, so you don't need to take any programming competency test this year to be preferentially ranked. I think Vicente would concur with me that you're considerably above the average GSoC candidate in programming ability.

It's very kind of you.

All you need is a mentor. Or, if you feel like mentoring another student in Pipeline this summer ...?

I'd rather like to participate as a student, it's my last chance.

I would though Benedek try to be flexible about your choice of project if you want to be sure to get a mentor (Vicente is already very overloaded every year). For example, a very attractive intermediate GSoC might be working on Boost.Fiber, specifically I am thinking some form of awareness of i/o in Fiber maybe. And Fiber ought to be up for its second review shortly, after which it enters Boost. From my short review of Fiber this week, it has very considerably improved over the review last year.

Actually, since Pipeline would be a fiber user, it's a great idea. Inter-thread fiber migration is something a good scheduler would need. Probably we can came up with something with Oliver when the time comes. Benedek

You may be surprised at how often you return to university study in

the future. I have two undergrads, a Masters and a PGCert, and I'm taking a third undergrad in maths by distance. Right now I'd qualify as a student for GSoC myself :)

​Great ! Do you want to implement numerical solvers for Boost.uBLAS ?​ ​I'll mentor you.​ :-D

On 28 Jan 2015 at 9:35, Kenneth Adam Miller wrote:

I would like to know what the qualification requirements are for becoming a GSoC mentor.

Writing one or preferably more project ideas for https://svn.boost.org/trac/boost/wiki/SoC2015 pretty much earns you the right. We can usually tell from how well the project ideas are written as to your depth of understanding and skill level in your domain, and if we see a problem in your project idea proposals we'll contact you privately about those to either tidy them up or in the worst case scenario, to ask you to withdraw (note this has never happened to my knowledge yet, most mentors realise when writing the ideas if they have what it takes or not and don't submit their idea if they know they aren't mentor material). Past that, it helps to like working with students, but to be firm enough as a personality to know how and when to be mean and how and when to be nice. If you get a good student, mentoring is *extremely* rewarding personally. Most of our mentors are serial repeats. Do try to design your ideas to be attractive to students, so what *they* think is a cool project rather than you (trawling past student submissions is an excellent way of figuring out their preferences). If no student picks your idea, we may try to match you to a similar idea, but more often than not we don't get enough slots from Google and you don't mentor that year if no student picks your idea. In that sense mentors compete for students as much as students compete for mentors. And do ask any questions here if you encounter them, in particular if my crappy instructions on https://svn.boost.org/trac/boost/wiki/SoC2015 don't make sense. Our thanks in advance for your interest in mentoring! Niall --- Boost C++ Libraries Google Summer of Code 2014 admin https://svn.boost.org/trac/boost/wiki/SoC2014

On 12 Jan 2015 at 13:32, Robert Ramey wrote:

Niall Douglas wrote

...

over $50,000 of Google funding brought to Boost in 2014

This is quite a surprise to me. I would be curious to know more about it. Could you point to a more detailed breakdown on this number?

Robert, You made an excellent catch. This morning, before coffee, I multiplied US$6000 by 8 in my head and somehow came up with $54,000, rather than $48,000 which is the correct amount. I should have used Calc, I knew it then but I was in a rush. I'll correct the value next GSoC reminder email. To break that down as requested, last year $5,500 went to each student, and $500 to each mentor, making $6000 per slot awarded, and we received eight slots in 2014. Apologies to everyone for the bad mental arithmetic. Niall -- ned Productions Limited Consulting http://www.nedproductions.biz/ http://ie.linkedin.com/in/nialldouglas/

On 9 Feb 2015 at 8:41, Robert Ramey wrote:

Damian Vicino consulted with me about proposing the following GSOC project:

Add safe<float> functionality to safe numerics http://rrsd.com/blincubator.com/bi_library/safe-numerics/?gform_post_id=426

This is an extremely difficult task. But my conversations with him have convinced me that he understands the issue involved. This would be a great addition to the safe numerics library which is still under development.

He has asked me to act as a mentor for this project and I have agreed to do so. I would expect to see him submit a proposal for this project soon if he hasn't done so already.

Firstly thank you hugely Robert for stepping forward to volunteer as mentor. Secondly, if he's the same Damian Vicino as was on this list a few months ago then I agree he's well capable. He looks to be an outstanding student. Indeed this year 2015 is looking to be a bumper year for GSoC, we may well have four prearranged students and mentors before GSoC has even begun. Historically those are shoo-ins for the rankings as they tend to be head and shoulders above the candidates who apply later, but equally we won't count our ducks before they have hatched. However, there is a process here, and his part doesn't start until after we get approved as an org this year, so first off we need to achieve that. To that end can you or him add the proposed project to the list of project ideas at https://svn.boost.org/trac/boost/wiki/SoC2015 please? That's the page Google reviews as part of approving us or not. A lack of ideas on that page caused our rejection in 2012, so we need that page long and detailed. Also, can Damian supply a link to code he has written so we can dispense with you having to set him a programming competency exam? The ideas page needs to be ready before the 20th Feb. Thanks, Niall --- Boost C++ Libraries Google Summer of Code 2015 admin https://svn.boost.org/trac/boost/wiki/SoC2015

On 10 Feb 2015 at 3:39, Damian Vicino wrote:

I sent you a link with the code I submitted to the incubator for library simulation about a week ago. From an email you sent me on Feb 06 I understood it was enough, but I can push more if needed. Do you need me to sent more code for the competency test?

Heh. That earlier email of mine be a classic example of what tiredness does to my mind. My sprint to get AFIO v1.3 released has turned into a long distance run, lots of staying up till 3am and then getting up early day after day week after week ... anyway I somehow had it in my head that I had seen your code and knew you could program whilst somehow not attaching your name with the Damian Vicino whose code I only reviewed last week. I am sorry about the confusion there. Yes you need no competency exam.

I will try to submit the proposal tomorrow after checking it with Robert. The harder thing right now is to define a realistic scope before sending it. I don´t want to submit milestones and finding out later they are impossible to reach.

Very wise. Well done on snagging Robert as your mentor. Myself and Robert disagree about most things, but hopefully he won't disagree with my belief that Robert will be an excellent mentor and you couldn't have done much better. Congratulations Damian. Niall --- Boost C++ Libraries Google Summer of Code 2015 admin https://svn.boost.org/trac/boost/wiki/SoC2015

On 10 Feb 2015 at 12:01, Thorsten Ottosen wrote:

I think the old idea of a boost::devector container would be a very use contribution. It's a bit like std::deque, but with much more control over the buffers. It should be described in one of the old GSOC pages.

Sounds like a great student project. However, the format for project ideas has been significantly tightened recently and so therefore the scope for copying old GSoC ideas reduced. In particular, ideas ought to have a programming competency test written with someone willing to mark it (usually the volunteer mentor(s)). We also need someone to step forward and volunteer to mentor such a project. Would you? Niall --- Boost C++ Libraries Google Summer of Code 2015 admin https://svn.boost.org/trac/boost/wiki/SoC2015

On 11 Feb 2015 at 10:25, Thorsten Ottosen wrote:

...

...

I think the old idea of a boost::devector container would be a very use contribution. It's a bit like std::deque, but with much more control over the buffers. It should be described in one of the old GSOC pages.

Sounds like a great student project. However, the format for project ideas has been significantly tightened recently and so therefore the scope for copying old GSoC ideas reduced. In particular, ideas ought to have a programming competency test written with someone willing to mark it (usually the volunteer mentor(s)).

So each proposal needs a seperate competency test?

Not necessarily. If the student has already written a library exceeding 1000 lines then we can examine that in lieu of an exam. The library just needs to be verifiable as not copy and pasted from someone else. Past GSoCs count too of course. Usually for the competency test one asks for some small part of the overall proposal. You're basically just checking if they can figure out on their own how to build Boost, how to use Boost.Test etc. and to get some look at their C++ ability e.g. do they understand exception safety? If they can't manage that on their own, the chances are low they would complete a GSoC anyway.

...

We also need someone to step forward and volunteer to mentor such a project. Would you?

I can do that, although the time difference, if it's a US student, can give problems. It all depends on the student I guess.

That would be fantastic Thorsten. Thank you. BTW a little over half the mentors are European/White Russian usually, I assume because we have sane working hours here. If you include Europeans who have temporarily sold their souls to the American machine, the number approaches two thirds. Americans don't seem keen to mentor, at least for Boost. I know the proportions are similar for KDE, another large GSoC org, but maybe that's just coincidence (e.g. I have no idea about Apache). Also nowadays most students are Asian, then European, and the fewest are American especially Latin American. I think the stipend isn't enough for American students given the opportunity costs. After all any US tech company will pay you per month for a summer what Google pays you for the entire summer. Conversely, out in India the Google stipend is a fortune. Niall --- Boost C++ Libraries Google Summer of Code 2015 admin https://svn.boost.org/trac/boost/wiki/SoC2015

On 12 Feb 2015 at 9:47, Thorsten Ottosen wrote:

...

...

So each proposal needs a seperate competency test?

Not necessarily. If the student has already written a library exceeding 1000 lines then we can examine that in lieu of an exam. The library just needs to be verifiable as not copy and pasted from someone else. Past GSoCs count too of course.

Usually for the competency test one asks for some small part of the overall proposal. You're basically just checking if they can figure out on their own how to build Boost, how to use Boost.Test etc. and to get some look at their C++ ability e.g. do they understand exception safety? If they can't manage that on their own, the chances are low they would complete a GSoC anyway.

I think if they cannot show code that have already made as students, it doesn't make much sense to expose them to a competency test, does it?

The submission of an existing library in lieu of a competency test is only there for the 15% of applicants who have been actively contributing to open source separately and independently from their studies for a period long preceding their application to GSoC (a rule of thumb is 12 months). It would be considered an extraordinary rather than ordinary competency test. Typical coursework would not suffice. We'll even take existing *library* code not written in C++. If, for example, a student have five years worth of verified contributions to the Linux kernel we would not ask for a competency test, even if they are brand new to C++. All we really care about is if you can program, if you have evidence of self motivation, and a work ethic. It also helps if we see proof that you can differentiate between asking for help when you really need it versus using advice as a crutch for laziness. The competency test was brought in to address a mismatch of expectations between students and mentors. This was leading to disappointment for both sides in some cases. It is also in fact optional, though those GSoC applications with a competency test or equivalent will be completely recommended to Google for selection before any applications without. That said, setting a competency exam now rather than at the last minute helps prospective students find out if a Boost GSoC is right for them before application. Historically half to two thirds of applications we receive every year stand no chance of selection, and filtering those out costs us considerable time and effort. Niall --- Boost C++ Libraries Google Summer of Code 2015 admin https://svn.boost.org/trac/boost/wiki/SoC2015