I started out [Thursday](http://linux.conf.au/programme/thursday) by going to
the keynote by [Stormy Peters](http://www.stormyscorner.com/). It was a little
less crowded getting seated this time around. A good keynote overall on what
motivates people to participate in open source projects and how that
interacts with companies paying them.
First main talk I went to was about “Memory-Efficient and Fast Websites — Pick Two!” which
by Malcolm Tredinnick was a very interesting talk although I did know a good
percentage of what he said already. However I picked up some links to tools
I didn’t know about already and some good tips on best practice.
One big problem was that the talk was very crowded, part of the problem with
this was that LCA this year didn’t have enough large rooms to go around. So while
the main Copland theatre could hold 493 people the others maxed out at 100-150
each. This meant that it took less than 25 percent of the attendees to fill
up one of these rooms ( from just 5 streams ). I guess that the organisers
just have to go with what rooms they have but future organisers should probably
try to have at least 3 rooms holding 200-250 people for a 600 person conference (with
one of these holding 400 plus ).
The other part of the problem was that [Rusty](http://ozlabs.org/~rusty/) got
the largest room and then spent 3 days scaring away as many attendees as
possible by telling them how much expertise was required. I suspect the organisers
had a “deer in headlights” moment or two with room allocations for this section
and decided that Rusty would probably be the most popular. However as I
stated above the big disparity in room sizes put them in this tough position.
The other thing I got slightly annoyed about in the talk was the couple who
re-arranged several other people in order to sit next to each other during
the talk. I really think that you could have survived being a couple of
metres apart from an hour or two ( especially since you didn’t talk or touch
during the talk except during the break ) rather than shuffling other
people ( not me ) around. You really should remember that you are ( I hope )
independent thinking entities not some creepy hive mind.
Anyway enough blog bitchiness. Next I went to a talk about Speech and Braille
computer outputs by Jason White. A pretty good talk with some interesting bits,
in a survey of the current state of the technology. I think many people
were surprised that a [Refreshable Braille display](http://en.wikipedia.org/wiki/Refreshable_Braille_display) costs
some $8000 dollars. I think there are a couple of efforts starting to
happen to produce lower cost alternatives.
[Val Henson](http://www.valhenson.org/) next did a talk on [chunkfs](http://www.valhenson.org/chunkfs/)
which I liked. Not too technical for me to follow ( Everything I know about
the Kernel is from the Linux Weekly News ) but it seemed to cover the details
in the serious Kernel people were interested in.
Afterwards I went to a talk about [netconf](http://netconf.alioth.debian.org/)
by [Martin Krafft](http://en.wikipedia.org/wiki/Martin_Krafft) . The netconf
project is intended to replace ifupdown with something more event driven
to handle the more dynamic environment systems (especially laptops) live in
these days with networks coming and going dynamically and everything being
While most people at the talk seemed to agree that something had to be done
there seemed to be many that felt that [NetworkManager](http://fedoraproject.org/wiki/Tools/NetworkManager)
might be a better tool rather than Debian developing one of their own. Personally
I said that creating a whole event-driven system seemed to be a duplication
of what [upstart](http://en.wikipedia.org/wiki/Upstart) is doing. Martin replied
that he didn’t think upstart was getting into Debian any time soon which I
assume is some heavy political issue since Martin maintains the package for Debian.
Last up was [David Jones](http://www.codemonkey.org.uk) talking about maintaining
the fedora kernel. Generally quite interesting and entertaining with minimal flashes
For dinner I went to the Professional Delegates Networking Session ( confusing
referred to as the [PDNS](http://www.powerdns.com) ) at the [Melbourne Museum](http://museumvictoria.com.au/MelbourneMuseum/)’s [CSIRAC](http://en.wikipedia.org/wiki/CSIRAC) exhibition.
We got buses there but it turned out that the Museum is right next to
[Lygon Street](http://www.melbourne.com.au/lygon.htm) it it wasn’t really
that far away. The sessions was the usual sort of thing with nibble and drinks
( not much coke though ) and finished around 9:30 or 10 (forget which). There
was a bit of confusion with some people waiting for buses back ( there were none
but nobody had been told ) before everybody either walked back of to the
Google Student party. I ended up walking back ( ice cream on the way ) with
a couple of others