☆─────────────────────────────────────☆
creation (yy) 于 (Wed Oct 13 22:15:44 2004) 提到:
? in distributed algorithms ??
looks like a cool guy, but not many people seem to understand his work
even many journal reviewers and famous people in the field
☆─────────────────────────────────────☆
wildThing (呼吁美国人每周少吃一包薯条) 于 (Thu Oct 14 02:58:59 2004) 提到:
what about his work in distributed algorithms?
☆─────────────────────────────────────☆
cdr (可读可写) 于 (Thu Oct 14 15:23:49 2004) 提到:
What made you say that?
He is one of the most famous guy in distributed algorithms
Many researchers work around ideas inspired by his work, and
his classic papers.
☆─────────────────────────────────────☆
borg (.)(.) 于 (Thu Oct 14 18:42:43 2004) 提到:
我估计他会拿Turing award
他是分布式算法的祖师爷
很多impossibility result都是他搞的
☆─────────────────────────────────────☆
shuke (小白) 于 (Thu Oct 14 18:46:59 2004) 提到:
谁给推荐几篇?有空也瞻仰瞻仰
☆─────────────────────────────────────☆
creation (yy) 于 (Thu Oct 14 22:34:28 2004) 提到:
他肯定是拿不料了
按理说他比knuth 晚10年吧,
水平就他在distributed system 和knuth 在algorithm/programming language
得位置讲,都差不多, 要拿早该拿了,
这个人太牛比, 你看他自己主页就知道了
他也说不上什么祖师爷吧,虽然他是最早formalize 这个咚咚得,
但这个领域本来就很小, nancy lynch 还有 connell 得一帮人及其徒子徒孙,
还有lamport,再牛人就没什么了吧
当然lamport还是超级牛得, 否则俺也不会去看他得paper
☆─────────────────────────────────────☆
jiangch (呵呵) 于 (Fri Oct 15 14:58:35 2004) 提到:
What amazed me is his broad interests. He made several profound contributions
to many different fields. Also he is even the guy who implemented latex. IMHO,
he is even overqualified for turing award. Some turing award laureates'
contribution is not so impressive anymore comparing with that of lamport's.
☆─────────────────────────────────────☆
creation (yy) 于 (Fri Oct 15 15:06:35 2004) 提到:
没有吧,如果你不是某人得field, 你好象不容易理解其重要性啊,
除了谁要是搞个P=NP 之类得
his interests are mostly on argument of distributed protocols,
from this , going to the theoretical side, becomes temporal logic
to the application side, distributed algorithms and some OS stuff.
early in his career he did some math stuff because he was math major
but seems he was later distained by his academic siblings (who all pursued
math careers ) for his somehow non-pure-maths stuff ------ see his webpage
☆─────────────────────────────────────☆
jiangch (呵呵) 于 (Fri Oct 15 15:21:36 2004) 提到:
I am taking a compiler course, it is not very related to distributed system.
In which I learned that lamport is one of first to formalize data dependence
and its application to vectorization and parallelization compiler. He first
had the concept of the well-known "true dependence", "output dependence",
"antidependece". even though these three are later characterized by Kuck
and credited to Kuck.
☆─────────────────────────────────────☆
jiangch (呵呵) 于 (Fri Oct 15 15:43:14 2004) 提到:
His "byzantine general problem" is one interesting paper to read, which models
the fault-tolerence problem in distributed system as several generals, some of
which might be unloyal, besieging a city. :)
☆─────────────────────────────────────☆
cdr (可读可写) 于 (Fri Oct 15 18:31:05 2004) 提到:
More interesting is his paper on Paxos algo, "The Part-Time Parliament",
which describes a parliament on an ancient Greek island.
The paper was submitted in 1990, and published on ACM TOCS in 1998.
Take a look at the annotations by Marzullo
☆─────────────────────────────────────☆
cdr (可读可写) 于 (Fri Oct 15 18:42:26 2004) 提到:
The problem was sometimes called "the Chinese Generals Problem" before.
Lamport wanted to assign the generals a nationality that would not offend
any readers. He once picked Albanian, but later decided on Byzantine.