Development Projects
This page lists relatively independent Tycoon development projects.
Implementation Projects
These are projects have a clear path to completion and do not have clear research value.
- Connect the bank with P!ayPal or other payment system.
- Write GUI front-end for Tycoon.
- Write test/coverage code for Tycoon.
- Port Tycoon client/server to Debian.
- Port Tycoon client to OS X.
- Port Tycoon client/server to RHEL.
- Port the Tycoon client to Windows.
- Connect Tycoon to PlanetLab.
- Forensic tool to connect traced network traffic to Tycoon user.
- Fix #$@! problem with Xen that does not allow booting in ballooned state.
- Add mechanism for auctioneers to asynchronously notify users.
Research Projects
These projects are more unspecified, but probably have research value.
- Modify sshd so that port-forwarding isn't necessary. Is this even possible?
- Look at OpenSSH privilege separation mode feature vs ssh cascading
- Carefully investigate and evaluate man-in-the-middle vulnerabilities
- Examine different market mechanisms
- Combinatorial v.s. not
- Reservations v.s. prop share
- Market v.s. fair share
- Work-conserving v.s. not
- Utility-conserving v.s. not
- Predictability mechanism v.s. not
- Examine the performance of different applications
- Single v.s. multiple host
- DHT
- Web server
- database
- scientific application
- Examine the impact of different macro-economic policies
- Design distributed banking or micro-payment protocol
- Use distributed resource allocation system, e.g., SWORD
- Design, use multicast RPC mechanism
- Economic security policies (e.g., discriminate type of network traffic)
- Connect Tycoon to power management facilities and adjust performance based on bid price and power price.
- Come up with algorithm to measure memory pressure of Xen domains like Carl Waldspurger's OSDI paper. Can probably do better with Xen because of OS-level information (rather than being restricted to VMM-level information).
[KevinLai]
