Episode: 520
Title: HPR0520:  Selecting Talks for PyCon 2010
Source: https://hub.hackerpublicradio.org/ccdn.php?filename=/eps/hpr0520/hpr0520.mp3
Transcribed: 2025-10-07 22:21:48

---

So
Welcome to a little bit of Python episode 2 with Andrew Kushling, Jesse Noller, right
Canton and Michael Ford and Steve Holden.
Jesse, I'd like you to start on this one because you, God bless you, had the horrendous
task of trying to, I mean, normally for choosing Python talks, it's sorting the wheat from
the chat, but as far as I could tell, it was almost all wheat this year and there was
a huge amount more of it.
For those of you who are listening to this, that don't know, I decided that because I
saw a bunch of people working on Python, they said, hey, listen, we need a program committee
chair and program committee chair is, I guess, a shorthand for a webbing boy.
And so basically, I took over as the Python program committee chair from Ivan Kristich,
I believe.
Ah, so that's the correct pronunciation.
So in any case, so I took it over and I said, you know, how hard is it going to be?
We're going to review, we're going to review a bunch of talks, we're going to discuss
them, we're going to pick the best ones.
Well, very hard indeed.
Well, so if we had gotten the volume of talks that we have gotten in the past, I don't
think it would have been nearly as painful, but instead this year, for Python 2010, we
got a grand total of 178 talk proposals, which is important.
No, compared to last year, I believe we had 118, something like 120, yeah, around there.
So, yeah, office many again, 50 percent increase.
Yeah, 178 talk proposals.
We were lucky.
We had over 50 reviewers sign up.
Unfortunately, many of those reviewers volunteered before they could realize that real life tends
to get in the way.
So we actually had a relatively, we had an increased reviewer pool, but a significantly increased
number of talks.
So what the program committee basically does is we turn around and we look at all the talks,
we review them and we do this online via a web interface.
We vote, we use identify the champion, and we plus one, we plus zero, we minus one, minus
zero, and we kind of vote on these.
So we get towards the end and all of a sudden turns into marathon IRC meetings where we,
we being the committee, we being the chair, sit down, we individually discuss and vote
on each one of the talks.
I don't remember the exact number, but typically what we do is we start with the only positive
vote talks and these are talks that are all plus one, plus zero votes.
We had well over 100 of those out of 178 talks, the bulk of them had gotten positive feedback
from the reviewers.
This has serious consequences, at least for the review committee, because this means that
we have to iteratively review these talks, we have to go through and say, okay, this
one's good, this one's not good, this one's good, this one's not good, this one's good.
We have to do that on the same batch of talks over and over and over until we get down
to a relatively decently sized pool.
I think, Josie, you really should emphasize here.
As Josie said, we had, how many was it over 100 all positive votes?
We don't have 100 slots for PICOM.
I mean, originally for last year, 2009, we only had just under 70 slots.
So we were well over positive talks compared to the number of slots.
So this is not just, oh, we took just acceptable talks.
These were good talks.
Yeah, I mean, the evidence is clear that there are a huge number of good talks that have
had to be rejected simply because there were other talks which were marginally better.
And the case of quality, if we have a lot of talks on one particular topic, we're going
to have to reject more of those.
Yeah, so you have to try and offer a broad interest in the conference.
Yeah, and I mean, obviously giving current hype and trends, there were a ton of web talks.
There were a ton of asynchronous programming talks.
I mean, you've got things in the news like tornado and twisted and everything else like
that.
So it's even talks, like there were a few talks that were very well positive, very positively
voted.
And they actually survived some of the final rounds.
But at the end of the day, it's like, okay, we can only have so many talks of this type
before we think about 20 to be sure some tornado in the end, didn't we?
So your favorite so-so in the just again, just to give people some idea of the likely
quality.
I've seen in the review process, I've seen talks where they were reviewed by four different
reviewers, one of whom gave the proposal a plus zero and the other three gave it a plus
one.
And that one's not made it in now quite apart from the fact that this means that we're
going to have what I think will be an exceptionally high quality program this year.
There are also indications that the people whose talks didn't make it into the formal program
will be organizing open space sessions and those open space sessions are also going
to be absolutely amazing.
It's one of those things where as the chair, I really tried to steer the committee towards
making as objective decisions as possible.
And it's one of those things where we all have kind of our pet subjects.
And it's funny kind of watching the process because a lot of people kind of left their
comfort zones to argue for a given subject or even a subject that was very controversial.
They actually argued for it and things like this makes it in.
And I think that overall, the past has shown us controversial subjects on things that
maybe aren't necessarily the tip of the height machine makes the best conference because
these are the types of talks people go in and they're floored and they're talking about
from months afterwards.
I really do think that this is going to be probably one of the best, hopefully the best.
Of course, that's me talking selfishly, Picon's we've had.
And one of the things that is out there that I don't think people fully realize is we actually
added an additional track of talks.
I mean, we typically historically, Picon has four tracks.
For semi-themed, you know, maybe beginner, intermediate, web, advanced.
But for the most part, they're kind of a mishmash.
This year I approached Van very early on and I said, listen, if we get a lot of extremely
high quality talks, I want to see if we can burn the resources necessary to add a fifth
track.
And he and I were very positive on it.
Van should mention being Van Lindbergh, this is the next year's conference chairman.
Yes.
So I approached Van very early on to say, hey, listen, do you think we could do this?
Now there's some negative sides of that, which is if you add a fifth track, you actually
increase the volume, you increase the number of talks, which means you have to have more
talks running at once, which means people have to make harder decisions.
But when you've got 178 really good talks, it doesn't make sense to kind of throw the
amount and say, you know, we only have this number of tracks, especially when we have
the availability to add the fifth one, we actually have the room.
We actually have the hotel, you know, saying, we're cool with you using the space.
So we now have a fifth track.
We actually have over 80 talks that are going to happen.
We've got a massive number of, you know, we've got more 45 minute talks than we ever have,
30 minute talks on broader ranges of subject matter.
I mean, we've got twist.
We don't have this year.
Don't we have this new thing that the post-decessions, which is a not a first for PICON?
It is indeed.
Yeah.
That was an idea that came out of last year's PICON and it's something that's popular
at academic conferences.
And it does mean that people get a chance to present their work in a way that allows more
people to see it actually, because the posters are available generally.
They throw out the whole conference and that means that, you know, as people are walking
maybe from the exhibition area to the next session, they can take in a poster session.
And there's lots of interesting stuff that people will be able to get that way too.
Yeah, for those of you who don't know what a poster session is, it's basically if you
think back to grade school, those dioramas, you always have to, not dioramas, but those
posters you had to do for, you know, science projects, you know, this is, you know.
I was going to say the same thing, but it's not like, because I guess that Jesse was basically
great to say for those of you who don't know what a poster session is, it's posters.
Not necessarily.
So, I mean, you always have the, you know, this is a poster, it's got a bunch of planets
on it.
And, you know, somebody says, you know, this is, you know, the way the planets are aligned
and, you know, so on and so forth.
This is the exact same thing, which means that, you know, somebody can sit there and do
a lot of work.
But together, a great poster on a subject that, you know, may not, may not otherwise get
exposure and they can just stand there and they can, you know, evangelize, they can show
it to people.
They can do a lot of work to, you know, squeeze information, which might be lost inside
of a talk into a much different, in a much more succinct and compact format.
So, it's going to be interesting.
It's worked out for the scientific community quite well.
I just think it adds a very interesting aspect to the conference because you have talks
for dissonally information out to a large group.
You have open spaces for much smaller group discussions, usually, the hashed out ideas
and such.
And then you have posters, which allow you to kind of, once again, at a smaller level
compared to talks, broadcast out information by staying there and having one-on-one conversations.
So, we have talking to an audience for talks.
We have talking amongst a small group in open space and then we have a single person
with more one-on-one talks and posters, allowing basically covering every aspect of so many
information, giving everyone a chance to at least somehow get some idea or a piece
softer or whatever they want out.
And it's all been done at Picon in 2010.
Well, and not only that, I mean, just look at the list of invited speakers itself is
awesome.
I won't even describe that.
I won't even give you their titles, although you can find that from the Picon blog that
we've got Ian Bicking, Jeff Rush, Jack Dietrich, Mike Fletcher, Raymond Hattinger,
Bob DePolito, Alex Martelli, Joe Grigorio, Ned Batcholder, Holger Crickle and Ted Leung,
all giving invited talks.
I mean, any, I would be surprised if any of those talks aren't completely amazing.
Yeah, it's just, let's ignore the keynotes, let's ignore, you know, ignore the keynotes.
Well, I mean, let's ignore, let's ignore the, let's look at the core of Picon.
I mean, we've got a killer list of invited talks, we've got poster sessions, we've got,
I saw the tutorial list today in those, we've got a talk on robots in space and another
one on robots underwater, I mean, you can't be bad.
Yeah, we actually, there was a large discussion about whether or not it was actually technically
legal for us to actually get rid of any talks involving robots and Python.
I think we actually decided that there would be a longstanding rule, yes to robots, yes
to Python, go to complete.
But I mean, we've got, we've got an amazing amount of great tutorials.
We've, you know, added a ton of talks, we've got awesome invited speakers.
And from what I can tell, the hotel looks, you know, just pure awesome.
Yeah, I've actually been to the hotel.
I've been to the hotel.
The conference advises that the Python software foundation use seemed to favor the
Mario to cross the road, which was also, I have to say, an awesome hotel, but the
Hyatt is just a fabulous space, quite apart from the, you know, the conference venue that
we'll be using.
It's just a beautiful space.
It's got this huge 20 story atrium with elevators going up and down in it.
And I'm sure it could, it could house a very good conference indeed.
This reminds me, I did want to suggest to everybody that if you're coming to Python, I would
like you to obey what they call in some of the science fiction conventions, the 321 rule.
Have any of you guys ever heard of the 321 rule?
Apparently the rule is you're supposed to go for three hours sleep a night, two square
meals and one shower.
I need my beauty sleep, man, three hours of sleep.
That's not enough.
I sometimes get three hours of sleep at Python.
The trouble is it's so damn interesting.
You just don't want to go to bed and leave interesting things behind some of the time,
you know?
Well, I like to find people's head-offs if I only get three hours of sleep, so I really
don't think you want that from me.
I like to find interesting things to pass out under.
What do you mean like tables in bars and so on?
Well anyway, I mean, there you go, I think we can wrap it all up by saying that it
looks like the Python, Python 2010 is going to be the most amazing Python yet.
Yep.
And everyone should come.
Jesse, do you just want to outline the, what are the registration fees this year?
Oh, Lord, I actually can't remember.
Oh, of course, you're the program chairman, not the conference chairman, aren't you?
If you want to go to the Python in the US, then the URL is us.pycon.org, which will
actually, currently, we direct you to the Python 2010 conference.
It's February 17th, sorry, the tutorials of February 17th and 18th.
The conference days are Friday through Sunday, February the 19th to the 21st.
And then there is yet again a four day following sprint from February the 22nd to the 25th.
Stay for the sprints.
Yeah, so it's the rates.
So registration this year went up a little bit, but for the most part, the prices, we kept
them in line, not too far off of last year.
So corporate government early bird is 450, hobbyist is 300, student early is 200.
Okay.
For anyone who's listening and wonders what the hobbyist thing is about, clearly there
are some people who can't get their companies to fund a low price conference.
And so for them, the corporate rate would apply.
The hobbyist rate is supposed to apply basically to anybody who wants to come to PyCon.
And either their company won't fund them to come so they have to pay their own way,
or it's to make it easy to get people who are interested in Python to get there.
Now clearly, because the Python software foundation is a non-profit, we want everyone
who can do to come at the corporate rate.
But if you can't come at the corporate rate and you can come at the individual rate,
and we'd much rather you came at the individual rate than didn't come at all, because
PyCon was started with the intention of being as inclusive as possible.
And one of the great parts about PyCon is that you get so many different people there
to contribute their experience and their knowledge.
Tutorials only, which is basically show for tutorials, it's $100 plus the fee for the
tutorial.
Your first tutorial costs $120, and this is all American.
Additional tutorials cost $80 each.
Also remember something that was left out earlier is on the same days that tutorials are
running if you get an invite, the Python language summit, and what's the second one?
The other one is the virtual machine summit.
Yeah, basically these are areas where we get people, the point of the VM summit is to try
to encourage people from other language implementations to come and share their knowledge and to
learn what they can about PyCon's implementations so that we can help to distribute and disseminate
information that will be helpful to all language developers.
The language summit itself is about the development of the language, and I think it was there
last year that we got the announcement about the Unladen Swallow project, for example.
But anyway, bottom line is, even if you're paying the corporate rate, you're getting a
three-day conference for $450, which is pretty good value for money.
Speaking of somebody who, you know, all things being, you know, equals still relatively
new to the community, it's only been like six years now.
It's, I can remember my first PyCon to the state, and they just keep getting better.
I mean, the exposure to not just core developers, but twisted developers to, you know, Google
engineers to just random people, it's fantastic.
And the conversations that you have, you know, for instance at lunch, you know, they can
range from astrophysics to, you know, bike shedding about, you know, the name of a built-in.
I knew that, well, I knew we wouldn't get through this podcast without you uttering the word
bike shedding.
Oh, are you going to be that way?
All right.
And of course, and of course, I should, in modest dimension, that next year, PyCon has
something that so far has only been available at Macworld, I'm like, correct, Jesse?
What's that?
The Steve note.
Oh, yes.
The Steve note.
Yes.
So we gave Steve Holden, you know, current company, a PSF slash state of the Python community
slot.
And it's, that should be cool, because I mean, historically we've had, you know, Guido
stand up and say, this is Python.
And you actually bring that up is actually great, because one of the other things that
we've done is we sat down and, well, I sat down and I said, what if we gave a 15-minute
slot to each of the Python interpreters to give a state of, so we're actually going
to have, you know, Guido is going to give a state of C Python, Colin Winters is going
to give a state of unlearned Swallow, we're going to have a state of iron Python, we're
going to have a state of PyPy.
All these guys are going to stand up and say, here's where the interpreters sit.
And Steve is going to stand up and say, this is where the community isn't probably
talk about, you know, very interesting subjects if you're into the community.
Well, and I hope, I hope to have some interesting things to announce as well.
Yeah.
So I mean, it's, it's going to be a great time.
So let me just say from Andrew Kusling, Brett Cannon, Jesse Nola, Michael Ford and myself
Andrew Kusling, please send your comments and suggestions to the email address all at bitofpython.com.
Our theme is tracked 11 from the Headroom Project's album, Heifa, available on the Magnetune
label.
Thank you for listening to Hack with Public Radio, HPR sponsored by Carol.net, so head
on over to C-A-R-O dot-E-C for all of us here.