next up previous
Next: Discussion and Conclusion Up: Sessions 7 & 8: Previous: Rebutia

Pereskia

The final group, Pereskia, had Jeff Chase as its spokesperson. Noting that almost all his points were trademarked statements of others, Chase listed the following problems:
1.
Performance is paramount.

2.
Nothing has changed: everything is broken

3.
The Internet is the Computer

4.
The Whole is Greater than the Sum of the Parts

5.
Naming

Proceeding to explain these sound-bites, Chase noted that performance continues to dominate attention, but needs to be defined more broadly. All other problems would be trivial if we didn't care about it. Other problems, characterized as ``*ility'' for ``reliability'', ``manageability'', ``usability'' and so on, can be addressed by the budget surplus of raw performance that systems now have. On the second problem, Chase said that this was not necessarily a research problem but we did seem to be getting dragged into the same rat holes time after time -- making systems work smoothly is critical. Turning to the third problem, Chase noted that scaling local systems up to a worldwide scale remained a challenge. On the fourth problem, he said that parts are cheap but integrating them into a whole is expensive. To be able to deploy power wherever we need it, incremental scalability and backwards/forwards compatibility are needed. Finally, Chase reminded the audience that naming still remains a poorly solved issue, especially at the intergalactic scale.


next up previous
Next: Discussion and Conclusion Up: Sessions 7 & 8: Previous: Rebutia
Peter Druschel
1999-07-28