[GSoC 2012] Off-Screen Particles project - continuation

[Assaf Raman]

I will be happy if the following will be done this week:
0. Fill this form and send it to the e-mail at the bottom.
1. Clean all the code you wrote, turn it into a new OGRE sample (like the rest of the samples in the sample browser).
2 . finish the sample to include the following controls - an on\off button to the effect, a "off screen texture size" combo, connect and test that it works.
3. Make sure the sample works on both DirectX and OpenGL render systems.
4. Make sure it works correctly with cmake - meaning adding the sample name and such to the right place.
5. Create a new fork of the trunk (as your current one is one big mess), and commit only that sample and relevant resource. Make sure to use the hgext.win32text ext with hg - so you will have the right EOL when committing.

This is the minimal result I expect.

[Assaf Raman]

Did you read my last two posts?

[Assaf Raman]

No answer.

[Assaf Raman]

Alive?

[Assaf Raman]

Hello?

[Assaf Raman]

I guess everybody will understand me - if I will fail the student for lack of communication, right?

[PhilipLB]

I'd say, wait for the 27th of August (the official end of this GSoC) and then decide. Until then, just see if and what happens here. If he comes arround with something finished until then, great. If not, hm...

[Assaf Raman]

No, working with the community is a big part of the summer project, not only doing the work.

[TheSHEEEP]

But most of the work is actual doing the work, not the community interaction.
The coding and testing surely takes much more time than is spent on the forums.

So, I agree with Philip. If this ends up with a working project, fine.
Sure, the lack of communication is astonishing, but I would not let a project fail if the result is running fine.

If you out-source something, you may end up with a working product, and had a terrible communication all the time.
So, next time, you will not hire the same guys again, but you still have what you paid for.
And I see GSoC projects as a very good kind of out-sourcing, as it is a win-win for everyone... in most cases, at least

All of that said, with a lack of communication very often comes a lack of work, so I'm not too optimistic here.

[Assaf Raman]

We defined a daily update, it should be one hell of a result for me to approve it after ignoring any communication from me for a week, and not updating anything, anywhere.

[Assaf Raman]

See his fork: https://bitbucket.org/KarolBadowski/ogr ... nparticles
See for yourself when was the last update.
He didn't work.
I don't see how I can give this a passing grade.
We wanted so much more - but because he didn't work - I have cut down the project to less and less, now it is something I can write in one day, this is not a joke, we selected this student over other that did want to work.
We will not make a joke of this whole thing.
If you don't work, you don't pass, that is it.
You are getting paid to work for two months, not to complete an "out sourced project".
I am spending my time for free, just so we may get in the end - new community members for this open source project.
It seems that this student think he can do whatever and he will pass in any case.
This is not the case.
Sorry, I need to draw a line in the sand here.

[TheSHEEEP]

I guess everyone agrees.

It's just that I don't think you should declare it as failed before the official deadline.
As unlikely as it is that he will appear on the 27th with something that blows our minds

It's of course your decision and everyone will likely understand whatever way you choose

In general, I've been working with a lot of students (which is kinda funny given that I myself never set foot in a university other than for meeting friends), and the diversity of talent & motivation is vast.
In many cases, students have not yet met the "real world" where you will be fired for behaviour like this, and other negative impacts on one's life. And because they've never experienced it first hand, the motivation to stick to the rules is not enough.
Of course, for every such student, there is one that has a very superior working attitude, like the other GSoC students, and many other community members, and many team members of course.

What I'm trying to say is, and please, I do not want to insult anyone: Working on a professional level (and I consider the GSoC projects to be more professional than some things in the industry) with students is always risky.
And having 3/4 GSoC projects running very good is something we can be really proud of
So.. uhm... don't let this drag you down too much.

[drwbns]

It's a bummer to have watched this thread. It's like a " Is he going to do it? " thread....I agree with Assaf that talks have been minimal and poor. The project sounded good but maybe it was out of reach for him and he didn't quite understand the scope of what he was proposing, although I think that if you have a proposal, you most likely understand the scope of it minus the bumps and debugs along the way. This may have been a bit too overwhelming especially if he had many other plans besides the project. I think it may have also been poor judgement to propose this and take up a slot without knowing how much time you were actually going to spend on this which tells me he had very little intentions on sticking to the deadline or an unknown schedule for work time. This isn't real-world work but a small hit to moving Ogre forward. +1 for Assaf for working with him anyways.

[bstone]

As unlikely as it is that he will appear on the 27th with something that blows our minds

I'm afraid that's unrealistic given the time needed for code review and testing by the mentor, then bug-fixing and working on remarks.

Unfortunately I must say that it was really obvious that this project was steadily moving toward the failure. Too much time wasted on figuring out how things work in Ogre by the student without actually grasping much of that. Then the serious lack of communication. Compared to other GSoC participants this one looked seriously underpowered.

I think there should be some pre-GSoC evaluation next time in order to avoid students with near-zero understanding of Ogre. At least that's relatively easy to carry out. But judging the dedication and responsibility will always be a problem.

[Assaf Raman]

What do you have in mind for the "pre-GSoC evaluation"?

[CABAListic]

Essentially I think that to participate in the GSoC, knowing your way around Ogre is a must. Doesn't mean you have to be an expert on every aspect of the engine (hell, I know I'm not), but you need to have the appropriate knowledge for your particular topic. I'm not convinced we need a specific evaluation procedure, but I think with candidates whose Ogre background we don't know we will have to question harder on the proposal until we are sufficiently sure the candidate is up to the task.

[drwbns]

Maybe a "Do you have the appropriate amount of time to complete your project question"

[Assaf Raman]

Maybe a "Do you have the appropriate amount of time to complete your project question"

They will always say yes.

[drwbns]

Why does this project proposal seem so much lighter than the rest of the projects? Did he not post all of it?

[Assaf Raman]

I guess because it is such a clear and simple project - we assumed we don't need to focus on that .

[duststorm]

Some projects require their GSoC candidates to at least do one commit to the project source before they apply. This can be something small, like picking a bug from the tracker and committing a fix. That shows that the student is capable of working with versioning tools and at least knows a little of the project's internals.

Whether something similar is possible for Ogre remains a question, though. This would only work if there are enough outstanding issues that don't require extensive advance knowledge.

Another idea is to post some small challenges from the community as an admission test, and let the students pick one of those. Of course the same problem applies: you need to be able to find useful features with small enough scope. The results of those can be used to pick the most suitable candidates.

[Assaf Raman]

I did asked to make sure they can get the trunk and compile OGRE, but we may need more then that.

[bstone]

What do you have in mind for the "pre-GSoC evaluation"?

What I had in mind is very close to what duststorm said. I would ask to complete a small task related to the GSoC proposal - something that would require 2 days of dedicated work. I think that's short enough to engage if you're serious about your GSoC proposal yet with the real dedication 2 days is plenty of time. The task itself will have to cover a few areas of Ogre working together (which is the case most of the time anyway), something a mentor could do in about 8 hours. It won't have to be a flashing success but whatever get's done will tell a lot about the candidate. Furthermore, since that will be in the context to the proposal it won't be unjustified effort either.

[Assaf Raman]

Different task for each student?

[bstone]

Yes. Handing them one and the same assignment has the following cons: they know that their evaluation work will not be of any use most likely (i.e. with 4 GSoC candidates there's 75% probability, not a hot incentive), it's unrelated to their proposal and might be in the field that is not particularly interesting for the student in question.