erlangbook ii and stocks?
erlang 的 maillist 里面最近在嗡嗡作响的事是——第二本 erlang 的书，以及一个有趣的新点子“erlang the stock”。
事情缘于 Joel Reymont 最近考虑要写第二本 erlang 的书，暂定名为《Hardcore Erlang》。
实际上，如果按照正式出版的纸版图书来说的话，应该是 erlang 语言的第三本。
[》以下插播 n 多字的八卦]
第一本是 Joe Armstrong 和 Mickaël Rémond 在 2003 年合著的《Erlang programmation》，第二本是 Joe Armstrong 今年早些时候出的《 Programming Erlang: Software for a Concurrent World》。对于 Eyrolles (出版商)来说(向他们的勇气表示敬佩和慰问吧)，这当第一的经验，恐怕是惨痛的。过于超前的推出时机，导致销售情况极不理想，以至于后来成为“票房毒药”的 Joe 想出第二本关于 erlang 的书时变得困难重重，此乃后话。
而时隔4年之后的第二本，是在 Joe Armstrong 的多方奔走之下，加上 intel 多核芯片面世带来明显升温的 erlang 社区，以及日渐汹涌的出书呼声，才由 Pragmatic Bookshelf 满腹狐疑的推出。想必，做出决策的时候，出版商的心里面也是凉凉的，没什么底。而这一本出乎意料的火爆(在 amazon 的排行榜上迅速攀升到了3k多，而且还在迅猛上升中)，这自然引起了书商们出“续集”的强烈渴望。
相比第一本和第二本之间的四年，这第三本仅过了四个月时间就嚷嚷着要出了(当然了，目前还只是在策划阶段)，相同的数字，不同的度量衡单位，让人不得不感叹“彼一时也，此一时也”。
[《好了，八卦插播结束，欢迎回来]
说回 Joel Reymont，此兄何许人也？请看这里这里还有这里这里。(是的，就是这位帅气拉风迷死人不偿命的 Joel 同学了，与 Yariv Sadan 并称为 erlang 两大新晋偶像，江湖地位十分了得。附送玉照一张，粉丝们不要抢，数字版，人人有份)
简单来说，此兄在 05 年就以一篇《Writing Low-Pain Massively Scalable Multiplayer Servers》名动编程界，震得跟帖无数(本站亦有收藏译文在此)。后转为 self employ 过着逍遥快活的日子，最近据说项目收尾了，要找有意思的事情做做。旋即传出要和 Roberto Saccon 一起来做“ActionScript (ECMAScript 4) to Erlang Compiler”的消息(本站亦有报导报导在此)。没想到，时隔不久就又得到了要写书的消息，两件颇不容易的事情能够齐头并进，确实是精力过人。
且来看看，这本《Hardcore Erlang》(暂定名)都有些什么精彩内容。
- The architecture of a poker server from far above
- Thinking processes instead of objects
- Game logic
- Stacking state machines
- Swapping logic
- Storing data in Mnesia
- State machines (gen_fsm)
- OTP behaviours
- Poker bots
- Simulating players
- Scripting
- Designing a network protocol
- Binary parsing
- Pickler combinators
- Automatic clustering
- Fault tolerance and fail-over
- Load balancing
- Testing a network server
- Debugging Erlang software
基本上，就是以 open poker 为基础，讲述 erlang 编程心得与思维方式的一本书。应该说，像这样来自实践的剖析和呈现会是非常吸引人的。更贪心一些，如果作者的语言又活泼风趣，能在轻松愉快的气氛中层层展现出自己清晰明了的思路(又看了一遍那篇《Writing Low-Pain Massively Scalable Multiplayer Servers》，对此我觉得甚为靠谱)，那必将会是一本佳作。配合起 Joe Armstong 的 THE BOOK ，一个讲技术，一个讲实践，必将能够大大拉低“主流” OOP-er 们向 COP 转变的陡峭学习曲线。
而就在今天(呃，我的机器显示过了12点，已经是昨天了)，此兄又冒出一个新点子：
I want to build a stock exchange and show you how to do it.
And of course I want to write a book about it.
The focus of the book is not changing, it’s still fault-tolerance,
scalability, distribution, etc. What’s changing is the software the
book is built around. Reading about how to build a stock exchange sure
as hell beats reading about a poker server.What do you think?
我觉得很有意思！你意下如何？
Comments
善！
把Open Poker的设计展示给大家真是个好主意，也真正是很多人迫切想看到的。
不过关于Stock Exchange项目，我觉得mailing list中的一位仁兄的说法很中肯：
——————————
Message: 8
Date: Mon, 12 Nov 2007 00:33:03 +0100
From: “Michael Regen”
Subject: Re: [erlang-questions] Lets build a stock exchange!
To: erlang-questions@erlang.org
Message-ID:
Content-Type: text/plain; charset=ISO-8859-1
What an interesting topic!
Just my humble two cents:
On Nov 11, 2007 2:58 AM, Joel Reymont wrote:
> I want to build the biggest Erlang cluster in the world and push
> Erlang to its limits.
>
> I want to build a stock exchange and show you how to do it.
This means either a lot of money and/or people are needed. So, how do
we motivate people to spend money or time?
I guess it will be easier to make people spend time than money.
Choosing the right project will affect very much how many people will
join and whether they need to spend money or time.
Building a stock exchange is an incredibly interesting aim from the
technical point of view and could certainly push Erlang to its limits
- if they exist :).
But if you want to build the biggest Erlang cluster than you need
something which is actually useable in real life for at least some
time. That would also be a much bigger benefit for the people to join
compared to just build a showcase example.
And here I’m not sure how big financial and legal hurdles will be to
actually operate a stock exchange. Does anybody have some insight in
this topic?
Just a half-baked idea and my empty brain was not able to figure out a
practical application so far. Anyway, maybe just think about it for a
short while and forget it soon if it leads to nothing: What about some
kind of socially desirable project. E.g. something one or better more
NGOs can use. Something that motivates people to join because they can
feel good if they contribute? Feel better than if they just
contributed code to the public.
Sometimes consulting companies compete hard for poor paid contracts
with NGOs just because it will be great PR for them. So there might
also be a bigger chance to push Erlang.
But I have to agree, this idea is not worth much without an
application. And of course, building something which is usable also
means responsibility…
So I hope this mail only drives people away from the stock exchange
idea if someone finds a good application soon.
Cheers,
Michael
——————————
Write a Comment