the great failure of the hacker halfway house

1
the great failure of the hacker halfway house

• seth hardy
• notacon 3
• april 8, 2006

2
building communities in self-destructive
environments

• seth hardy
• notacon 3
• april 8, 2006

3
what this talk is not
how the hacker halfway house failed
4
what this talk is not
shmoocon is teh sukc hahaha lol
5
what this talk is not
dont be such an immature shitbag
but seriously, please dont be such an immature
shitbag thanks
6
what this talk is not
if you focus your energy like a LASER, you can do
anything!
7
what this talk is

• observations on communities
• a discussion about new methods for solving these
  problems
• helping others avoid mistakes already made
• asking the simple question
• why do some people shit where they eat?

8
irc.2600.net
lets start at the bottom of the barrel irc.
9
2600net a timeline

• years ago network completely unmaintained
• more than a years downtime due to massive ddos
  attacks
• ted stole our passwords!
• today a (mostly) functional irc network

10
2600net historical problems

• lack of involvement
• admins hated the users
• overwhelming bureaucracy
• full consensus needed to make minor changes
• decentralized command structure
• admins were played against eachother
• freedom of speech issues
• ccna is the proper nomenclature

11
ptolomea.2600.net / ptolomea.aculei.net

• one of the main leaf servers for 2600net
• hosted on a SDSL line at the hhh
• hit by distributed synfloods
• ...20Mbits of synflooding
• ...which caused 50 packet loss at our ISP
• why would someone do this?

12
ptolomea.2600.net / ptolomea.aculei.net

• one of the main leaf servers for 2600net
• hosted on a SDSL line at the hhh
• hit by distributed synfloods
• ...20Mbits of synflooding
• ...which caused 50 packet loss at our ISP
• why would someone do this?

haha i can fuck up ur shit lol
(paraphrased)
13
what to do?

• clear organizational structure
• large ddos attacks will always break things
• make the network more resistant, not
  attack-proof
• this is the job of a politician
• its how you respond to problems that matters

14
hackerhalfwayhouse
15
hackerhalfwayhouse

• the short, short history
• some people moved in to a loft in brooklyn
• they did some stuff with computers
• they threw a bunch of parties
• people fucked shit up
• the 2600 afterparties stopped

16
respect issues

• unlike many other hacker spaces, people lived
  in the hhh...
• would you ash your cigarette on someones living
  room floor?
• ...in front of them?
• ...and then get offended when they ask you wtf
  youre thinking?

17
more respect issues

• theft
• multiple laptops stolen
• usually property of guests
• vandalism
• sharpies on walls is ok, right? youll repaint
  them eventually.
• stickers peel off so theyre ok, right?
• how about a can of latex wall paint on someones
  brand new car?
• brooklyn now has indigenous metal-eating
  chipmunks. no, really.

18
respect and the community

• monthly nyc parties
• these are our friends!
• issues, issues, more issues
• 5hope con party
• many hundreds of strangers came through over the
  course of the night
• zero problems. none. nothing.

19
the solution

• close the doors, end the parties
• fairly split on this solution
• nobody liked this way out
• where is the line between keeping an eye on
  guests and policing them?
• there really is no simple answer here.

20
aculei.net

• started six years ago
• shell account box on home adsl connection, added
  a user here, a user there...
• shells, email, web hosting, dns, irc, network
  backup, etc.
• now has almost 70 users, hosted in a colo in
  boston

21
as always, there are problems

• denial of service attacks
• target for hack attempts due to being high
  profile
• abuse of resources
• hardware failure, limited resources

22
but what are the differences?

• denial of service attacks
• people look out for their own
• hack attempts
• no bureaucracy to delay security process
• misuse of resources
• nobody wants to lose their access
• hardware failure, resource limitation
• people would donate what they could!

23
different kind of community

• invite-only instead of we need more people,
  lets invite the world
• self-policing because people use it as their
  primary network services
• more personal, trust is easier
• strong involvement by the person in charge
• time and financial commitment

24
observations

• not all communities can be invite-only
• where is the acceptable compromise?
• not all communities can be financed by an angel
• usually not needed, though
• people often contribute only when its
  fashionable
• nobody wants to do the shit work
• one person pulling out can cause a cascading
  failure

25
observations

• respect breeds respect
• ...usually
• lack of action causes people to push the line
• successful projects are hard work
• nothing comes overnight
• you need to do the dirty jobs before you can
  celebrate your success

26
the real question

• is it worth it to you?

27
questions?