November 25, 2008

real programmers don't write documentation

Posted by hufey at 04:26 PM | Comments (0)

September 16, 2008

It's not a || it's an && operation.

Jared Hulbert to Tomas, linux-kernel
show details 11:42 PM (10 hours ago)

Reply

> I read somewhere that AXFS is going to be a part of Linux kernel 2.6.28
> ( http://www.linuxdevices.com/news/NS6179099416.html?kc=rss )
>
> What are the advantages of AXFS compared to squashfs?

Depends on what you are doing and what features you care about. You
can read some of the threads starting ~Aug 20th for the details.

> Why not to include squashfs instead?

To be clear, the submission of AXFS has nothing at all to do with
squashfs. It's not a || it's an && operation.

But yeah, Christoph is right. 2.6.28 is iffy for AXFS anyway.

针对AXFS对squashfs的疑问,这是一个很有技巧的回答。
当然AXFS在performance的提升上,我在OLS2008的会议现场印象深刻!

Posted by hufey at 09:56 AM | Comments (0)

August 23, 2008

比尔盖茨的审查

比尔盖茨的审查

757个读者 翻译: wilddog 昨天 23:42:44 原文 引用 双语对照及眉批

简介

Joel Spolsky 技术博客中的超级牛人,原来是微软EXCEL项目中的一个程序经理,本文回忆了比尔.盖茨第一次审查他的文档的故事(当小牛遇到大牛),非常有趣,

Excel曾自带了一个糟糕透了的开发语言,当时这个语言还没有名字。我们称其为"Excel Macros".它没有变量、局部变量、子函数这些概念,几乎完全不可用。虽然它有是还一些高级功能,比如“Goto”。
­
它存在的唯一理由是:它看上去要比Lotus的宏要合理一些。Lotus的宏不过就是记录了一组键盘敲击的顺序,用来在单元格中输入长字符串。
­
1991年6月17日,我开始为微软的Excel团队工作。我的头衔是“程序经理”(相当于写需求分析的),我的任务是为Excel Macro 找到一个合适的改进方案。有人建议我可以用Basic来做这个方案。
­
Basic?!
­
我花了些时间和不同的开发团队商量。那时候Visual Basic 1.0 才刚刚推出,它简直是酷毙了!但在微软内部当时还有另外一个面临失败的项目"MacroMan". 他们试图开发另一种面向对象的Basic。这个团队好不容易才被告知他们的产品终于有了一个客户,就是我们Excel组. 当时他们的市场经理是Bob Wyman,是的,就是那个Bob Wyman(是谁?很有名吗)。可怜的Bob Wyan,我是唯一一个可以让他发挥推销才能的人。
­
如我上述,MacroMan快失败了,虽然它有一些支持者,但最后还是面临被停掉。Excel团队确信他们这个Basic团队正是我们开发Visual Basic For Excel 所需要的。我在这个项目的基础上试图再加入4个我喜欢的功能。Variant--一个通用的数据类型,可以随时保存其他类型的值。因为如果不这样的话,你必须先要定义一个转换声明才能保存单元格的值。后期绑定,也就是后来赫赫有名的IDispatch.另外两个特性是For Each语句(从csh中偷的)和With语句(从Pascal中偷的)
­
接下来我开始写Excel Basic的规格说明,它真是他妈的工程浩大,我想完成时大概有500页吧。(“瀑布模型?”(传统软件开发模式),也许你正在窃笑。是的,但是请你闭嘴)
­
那时我们常常请比尔.盖茨作些审查。基本上每个重要的功能比尔都得过一遍。我被告知把我的功能说明送到他办公室去,以便让他在审查会前做好准备。这个说明用了整整500页打印纸。
­
我把说明打出来送到了他的办公室后,正好当天还有一点时间,于是我试图评估一下Basic的日期和时间函数是否能处理所有Excel的功能。
在大多数现代程序开发环境中,日期被存为实数。整数部分是从过去某天算起的天数。比如说Excel中,以1900年1月1日为1算起,今天2006年6月16日就是38884。
­
我开始把Visual Basic 中的日期和时间函数和Excel中的都过一遍,看看是不是都能对应。我注意到Visual Basic文档中有一个很奇怪的声明:Basic 使用1899年12月31日而非1900年1月1日作为日期的原点,但不知什么原因,Excel中今天的日期值居然和Basic的一摸一样。
­
出问题了!
­
我跑到Excel资深程序员Ed Fries的面前.我想他应该知道为什么。
“哦”,他告诉我:“看看1900年2月28日”
“59”,我说
“现在再看3月1号”
“61”。
“60哪里去了”Ed问道
“2月29?1900是闰年!它可以被4整除”
“很好的猜想,但不对”,Ed边说边走,留下我一个人在那里深思。
我又做了一些研究。发现可被100整除的年份不一定是闰年,除非它能被400整除。
原来1900不是闰年。
“找到了,这是Excel的Bug”.我欢呼起来。
“错”,Ed说,"我们这样做是为了能导入Lotus 123(另外一种电子表格)的数据"。
“那它是Lotus 123的Bug?”
"正确,但很有可能是故意的。Lotus必须适应640K的内存。但当时并没有那么多的内存,如果你忽略1900,你就可以根据一年的尾数是不是00来判断那年是不是闰年。这样做非常简捷方便。Lotus的小伙子们可能认为历史上两个月误差1天并不是很严重。而Basic团队看上去想给这两个月擦屁股,因此他们又把时间原点调回来了"
­
“啊!”我恍然大悟,然后开始继续研究为什么那里还有一个设置1904为时间原点的复选框。
­
比尔做review的那天来临了。
­
那是1992年6月30日。
­
那时候微软还没有现在这么官僚化。我对Mike Conte,Mike Conte对Chris Graham,Chris Graham对Pet Higgins,Pet Higgins对Mike Maples,Mie Maples最终对比尔.自底向上大约有6级,而非现在的11级或12级。我们和使用8层管理结构通用汽车一样,是个充满乐趣的公司。
在比尔的review中,每一层负责报告的人都出席了,顺便带来他们的堂兄堂弟,表姊表妹,大姨小姑,还有一个专门记录比尔会说多少个“靠!”的人。当然“靠!”越少,结果就越好。
­
比尔进来了。
­
他居然是两手两脚,一个脑袋.....看上去还是象人类。
­
他手里拿着我的说明。
­
他手里拿着我的说明!
­
他坐下来和旁边的人开了个对我而言毫无意义的玩笑,有几个人在窃笑。
­
比尔转向我了。
­
我注意到在我的说明空白处写着一些注解。他读过第一页!(激动)
­
他已经读过第一页并写下了他的笔记!(更激动)
­
因为从提交说明到他手上不过一天时间,他一定在昨晚之前就读过。
­
我们开始一问一答。问题相当简单,但现在我已经记不清了,因为当时我一直在注意着他翻我的文稿...
­
他在翻我的文稿!!!(冷静,冷静,你可不是小姑娘了)
­
...我看到每页空白都写满了注解,他妈的,他居然全部读完了,还写了注解。
­
他居然全部读完了!
­
问题开始变得越来越难,越来越细。
­
提问有些随机。过去我常常认为比尔是我的大爷,事实上他是个好小伙!他读了我的说明,他的问题可能正记在说明的空白处。以后我一定得把这些问题都在Bug跟踪系统中标记出来。
­
终于最难的问题来了。
­
“我不知道,小伙子们”,比尔说,“有人研究过这个细节没有,比如说这些日期和时间函数。Excel有很多日期时间函数,在Basic中是不是能一一对应,他们工作方式是不是都一样的?”
­
“当然”,我说,“除了1900年的1月和2月”。
­
沉默。
­
“靠”的记录人员和我老板惊讶的对视一眼。这家伙怎么知道的?1月,2月是什么东西?
­
“OK,很好”,比尔拿起我的说明。
­
(....等等!我想知道......)
­
他们离开了。
­
“4个”,“靠!”的记录者宣布,"这是我有生以来记得最少的一次。比尔步入中年了"。如你所知,比尔当年36。
­
后来我对这次审查终于有了一个解释,"比尔并没想真的想要审查我的说明文档,他只是想确定一切是否在我控制之中。他的标准方式是问得越来越难,越来越难,直到你承认你不知道,然后,他就会提醒你你尚未准备好。如果我答起了他最刁钻的问题,没有人会确定会发生什么,因为这以前就没发生过。"
­
“如果是Jim Manzi,他会怎么表现?”有人问道,“Manzi会反问,‘什么是日期函数???’”(Jim Manzi就是把Lotus带垮的MBA)
­
正确。
­
比尔. 盖茨是个令人吃惊的技术狂热者,他理解Variants,Com对象,和IDispatch,知道为什么自动化和虚拟函数表不一样,也知道这个区别会导致双接口。他担心日期函数,如果不是他信任的人在开发,他会难以入眠。你可以对此嗤之以鼻,但他确实是一个程序员,一个真实的,地地道道的程序员。不是程序员的人去运作一个软件公司就像不会滑板的人跑去冲浪一样。“得了,我有非常棒的顾问在岸边指导我”,他们会一边说一边一次次从滑板上掉下去。那些标准 MBA就是相信管理是通用函数。鲍尔默会成为另外一个John Sculley吗?John Sculley曾差点把苹果公司搞垮,只因为董事会相信他卖百事的经历会对运作计算机公司有所帮助。MBA的迷信者通常相信这一点:你可以让组织做你完全不懂的事。
­
过去数年,微软变大了,比尔的管理范围过度扩展,他不得不投入更多精力去和US政府打交道。鲍尔默坐上了理论上的CEO这个位置,这可以让比尔花更多的时间做他最擅长的--领导软件开发,但这并不能克服11级管理层带来的问题。它导致无休无止的会议,顽固地不管如何也要创建每种产品的倾向,(微软在研发、官司、公司声誉上损失了多少亿美元,只因为决定开发一个Web浏览器,还让它免费。),还有就是过滥的招聘让微软正逐渐失去中层人员中的精英。(Douglas Coupland在Microserf*中写道:”在1992一年中他们雇佣3100人,你知道并非所有人都是珍珠“)
­
对了,派对已经搬地方了。Excel Basic变成了Microsoft Visual Basic for Appliation for Microsoft Excel,这个名字里面的注册商标符号太多,以至于我都不知道应该把他们放在哪里。我于1994年离开了微软,想来比尔已经完全忘记了我。直到有一天我读到华尔街日报对比尔的一次采访,他提到在过去,象招募一个Excel的程序经理的事情是多么困难,他们并非象苹果长在树上那样垂手可得。
­
他是谈我吗?或许应该是别人了吧。

Posted by hufey at 10:34 AM | Comments (2)

August 20, 2008

iphone browser's marketing share

这个数并不出意料,但是如果是文中说的“Windows Mobile 浏览器市场份额的4 倍,大约是 Linux 平台浏览器市场份额的40%”,那还是挺让人吃惊的。

Posted by hufey at 04:44 PM | Comments (0)

July 16, 2008

From 2.4 to 2.6 to 2.7?

Linus对未来linux kernel版本号命名的想法

from Stoyan Gaydarov
to linux-kernel@vger.kernel.org
cc torvalds@linux-foundation.org,
Alan Cox ,
gorcunov@gmail.com,
akpm@linux-foundation.org,
mingo@elte.hu
date Tue, Jul 15, 2008 at 10:10 AM
subject From 2.4 to 2.6 to 2.7?
mailing list Filter messages from this mailing list
mailed-by vger.kernel.org

hide details Jul 15 (2 days ago)


Reply


Filter Assistant


First things first, I would like to know what prompted the change from
2.4 to 2.6 kernels. I know that the change had to do with the
development version, the 2.5 tree and the massive amounts of patches
distros had to carry. Aside from this i think it was also the
scheduler changes that prompted the 2.6 version, but I don't know all
that much about it and any other comments about the change would be
great.

Second I wanted to talk about the linux 2.7.x kernel, whats in the
making or maybe even not started, that could prompt a change to a 2.7
version kernel, i know that a lot of good changes are going into the
kernel as part of the rcX kernels in the 2.6 version. Would we
continue to see 2.6 kernels until some big problem shows its head and
we all go "oh sh**" and then change something so massive that it
prompts the change or are we going to continue with the 2.6 tree. I
just want to get some information and peoples opinions on this, just
to see where things are headed.

-Stoyan G
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
Reply

Reply to all

Forward


Linus Torvalds
to Stoyan, linux-kernel, Alan, gorcunov, akpm, mingo

show details Jul 15 (2 days ago)


Reply


Filter Assistant


On Mon, 14 Jul 2008, Stoyan Gaydarov wrote:
>
> Second I wanted to talk about the linux 2.7.x kernel, whats in the
> making or maybe even not started

Nothing.

I'm not going back to the old model. The new model is so much better that
it's not even worth entertaining as a theory to go back.

That said, I _am_ considering changing just the numbering. Not to go back
to the old model, but because a constantly increasing minor number leads
to big numbers. I'm not all that thrilled with "26" as a number: it's hard
to remember.

So I would not dismiss (and have been thinking about starting) talk about
a simple numbering reset (perhaps yearly), but the old model of 3-year
developement trees is simply not coming back as far as I'm concerned.

In fact, I think the time-based releases (ie the "2 weeks of merge window
until -rc1, followed by roughly two months of stabilization") has been so
successful that I'd prefer to skip the version numbering model too. We
don't do releases based on "features" any more, so why should we do
version _numbering_ based on "features"?

For example, I don't see any individual feature that would merit a jump
from 2.x to 3.x or even from 2.6.x to 2.8.x. So maybe those version jumps
should be done by a time-based model too - matching how we actually do
releases anyway.

So if the version were to be date-based, instead of releasing 2.6.26,
maybe we could have 2008.7 instead. Or just increment the major version
every decade, the middle version every year, and the minor version every
time we make a release. Whatever.

But three-year development trees with a concurrent stable tree? Nope. Not
going to happen.

Linus

Posted by hufey at 04:19 PM | Comments (0)

3G iPhone's Battery Life Beats AT&T Rivals--But EvDO BlackBerrys Run Longer

Posted by hufey at 11:31 AM | Comments (0)

April 18, 2008

为什么离开Google

和上个blog 可以对照着看

---
为什么离开Google
matrix 发表于 2008年4月17日 14时15分 星期四 Printer-friendly Email story
来自人各有志部门
Google
DigitalPig 写道 "谷歌治印报道,一个曾经在Google公司工作过得工程师在他的Blog上面写了他为什么离开Google的原因。(摘要了一下:) 在具有这样大小和影响的公司中,Google是一个让人兴奋的公司。有很多咖啡厅(perk?)和餐厅(在Mountain View总部有15个餐厅,免费供应从早餐到晚餐)。而且有很多的讲座。在Google中得到培训的机会也很多。当然以团队合作的名义出去玩的机会也很多。每个新员工都会有一个24‘’或者30‘’的显示器和一台公司配给的笔记本电脑(MacBook或者Thinkpad).对于Google这样的一个公司来说,它的出现改变了互联网以前糟糕透顶的搜索模式,而且它的新产品,比如GMail、Google Reader和Google Calendar一推出就收到广泛的关注而且变得非常的流行。这也让我感到非常的自豪。而且Google内部的邮件列表非常有活力,你可以在上面得到很多的信息,工作的非工作的。这也让我想到了早期的流行而且让人受益的USENET,而不是像现在这样充满白痴。(同意!)
不过虽然是这样,我还是离开了Google。我很享受我在Google的这段工作经历。但是我更加喜欢作为一个公司的创始人而不是一个大公司平庸职工。当然,咖啡厅和其他都吸引人,但是却不能激励我。相反,在一个公司刚开始的时期,每个员工都会对这个公司有着巨大的影响和很大的责任。这才是我追求的。这些年我也发现很多创新都来自刚刚成立的小公司。当然,在小公司工作也有很大的风险,因为大部分的小公司都倒闭了,但是至少还有1/10的公司能生存下来。相比而言,作为一个Google员工,就算Google股票翻一番或者翻两番,我都不能变得更加富有。

而且,Google中我所开发的项目也不和我的口味。虽然过了一两年员工会交换他们的项目,但是对于一个有很庞大代码库,工具和框架的Google来说,要熟悉这些然后再上手并不容易。开始的时候会感觉非常的郁闷。所以我觉得Google招聘人员应该在这方面多加注意。

现在我在一家即将成立的公司里面做社会网络(Social Networking)/移动空间(Mobile Space)开发。我的职位将是一个工程负责人。当然,对于小公司来说,一个职位将会扮演非常多得角色。不过,我能使用自己喜欢的Ruby开发。对于这份工作,我还是觉得很兴奋。

有人喜欢打天下,有人喜欢守天下。"

Posted by hufey at 11:30 AM | Comments (0)

April 17, 2008

逃离怪物公司及其他

在网上看到一个帖子提到某pixar前员工离开这家公司的事,找到了相关链接
逃离怪物公司
pixar的影片几乎我都看过,而且其中很多都买了非常好的DVD版本。其中一个乐趣就是
看DVD附带的制作花絮,包括对pixar的介绍。很羡慕他们的工作环境和工作方式,公司
花很大力气为大伙创造能尽最大可能发挥个人创造力和激情的环境。我对艺术的创造力
接近为0,对艺术的欣赏虽有提高也是从一个很低的起点开始,不过还是“虽不能至,心向往之”吧。
甚至比较起来这两年对google工作环境的夸张宣传,我觉得离pixar还有差距。
Steve Jobs--这个两个我欣赏公司的共同老板和灵魂人物,最近负面报道也很多。其中
有一个是美国评员工最满意的雇主,apple没有进前25。没看到那个表,不知道pixar是不是在里面。
这个帖子其实很简略,也没说出什么特别称得上内幕的东西。既然pixar人才济济,被埋没只怕
也无可避免。
这个故事告诉我们的是:即使是外人看来多么光鲜照人的公司,在里面生存也未必容易。
我说的容易是easy。

Posted by hufey at 02:35 PM | Comments (0)

April 07, 2008

3G专利分布

WCDMA:诺基亚(35%)、爱立信(31%)、高通(6%)、西门子(5%)、华为(5%)、其他(18%)
CDMA2000:诺基亚(22%)、爱立信(6%)、高通(31%)、西门子(5%)、摩托(11%)、其他(25%)
TD:诺基亚(32%)、爱立信(23%)、高通(2%)、西门子(11%)、大唐(7.3%)、其他(24.7%)

不知数据是否权威。记录一下

Posted by hufey at 10:25 PM | Comments (0)

April 03, 2008

Linux Kernel Development (April 2008)



Who is Doing the Work

Who is Sponsoring the Work

Posted by hufey at 06:44 PM | Comments (0)

March 27, 2008

moto的分拆

昨天晚上吃完饭在床上拿iPhone上网看新闻,一下子看到新浪在几乎很靠前的位置的moto分拆的消息
赶紧点开,不幸safari意外退出了。跳下床到书房拿笔记本上网。看到网友评论调侃说拆成“摩托”和“罗拉”两家公司,笑了出来。

其实打去年起就听到很多传言诸如分拆,卖出等等各种消息。等看到了还是唏嘘了一下。

昨天看到澳洲的电信运营商大概是最早运用WiMax的运营商之一抱怨WiMax效果很差,还在想
moto这次是不是又压错宝了。

今天早上听说moto股票因分拆消息暴涨12%,拿出iPhone打算看一下,却发现前两天因为意外
重刷iPhone软件之后并没有把moto再次加到股票列表里。看来moto确实已经不在我的世界里
了。

从engadget借用了个图,这个图的文件名叫unhappy-moto.jpg

另外googlechinablog最近正在写motorola的兴衰史2,还是很好看的

Posted by hufey at 10:56 AM | Comments (0)

March 17, 2008

换个话题, 说说宝马


也是机缘巧合,一辆宝马325由我保管一周。今天上下班开了
50公里。在这写点体会。

从开始学车开始,是普桑,手档。我一直觉得普桑是最适合
学车入门的一个车型。当然社会上的普桑我一直也没开过,
就驾校的普桑来说,每个档位都清晰而且精准的和车速联系
起来,也就是说如果你在该加减档的时候没有加减档,立马
给你诸如挫车等反馈。非常有利于新手知道自己哪里做的不
够好。

拿到车本之后,先后短时间开过金杯,捷达,宝来。其中金
杯是亲戚主动借我练习上路的,除了坐着很不舒服忘了其他
的特点,总之没有优点,最大的缺点就是不舒服。

手档捷达和自动档宝来是陪练公司的,学车时我也算是驾校
的明星学员,等真正上路之前还是有点心虚,在家附近找了
一个陪练公司。第一次是手档捷达,也是当时自己准备买的
候选车型。也许是陪练公司保养的比驾校的好,总之捷达给
我留的印象非常好。同样是1.6的发动机,动力比普桑要好得
多的多,换挡行程也比普桑的少,而且对换挡时机感觉比较
模糊,对新手来说很能鼓舞士气,因为不象普桑在驾校里往
往因为换挡早或晚挨师傅的骂。直到现在我还经常劝预算少
的朋友考虑捷达,当然一个也没劝成功过。我自己老琢磨着
如果自己买轿车捷达是非常有性价比的好的选择,也可以低
调一点嘛。第二次租车是同一家租车公司的自动档宝来1.6。
换自动档是因为自己从来没开过自动档,省着将来有机会开
自动档时候干瞪眼不会开。一上车,陪练师傅大概讲解了
PRND几个档是干什么的,然后就开上路,顿时感慨要是都
学自动档根本不要驾校的58小时,可能有半个小时就行了。另
外感觉宝来也是很不错的车,之前调研都说宝来1.6排量不够
怎么也得上1.8T或者2.0云云,自己开起来油门踩下去却马上
就明白了什么是推背感,总之在那之后对一汽大众的车都有
好感,感觉他们走的都是“驾驶者之车”的路子。

再后来自己买车的时候还是买了6缸切诺基。为的是自己的越
野梦。切诺基这车也是经典了,4.0排量保证在山路上有无穷
动力,平路上也敢和各种轿车飚车。

再后来短暂的开过奥迪A4,马6,Liberty,camry。奥迪的
感觉就像是在开一个大铁陀,重心极低,踩下油门之后仿佛
车停滞了一下然后蹭的蹿了出去,很过瘾。最不适应的是奥
迪的刹车,开始感觉似乎没有,再稍微一踩,车立马站住,
感觉能把车上的人都快甩到前盖上了。马六这两年在国内卖
的极火,我开了觉得不怎么样。加速什么的都没得说,可是
明显飘的很,就像是在开一个纸盒。马六是我第一次亲身体
会到所谓日本车风格。后来开过一次camry也是这种感觉,
车是好车,很精细,就是感觉很单薄。轻松上180,可是就
像一片纸片。

再说Liberty,3.7 V6的发动机,够用或者也可以说的上是不
止够用,反正美国人似乎永远都会给配上一个超出你的需要
的发动机。跟我自己的同一家族的切诺基4.0直六发动机来说
感觉动力略有不足,不过这也许是自动档的原因。说起来这
个自动档,也是我对liberty最不满的地方就是换挡不够平
顺,经常在他换挡的时候吓到同车的mm,老让mm以为是我
车开得不够好。

说了这么多,再说宝马。外观,内饰我都没仔细看,我想这
个档次的车这些就不用看了,我也不怎么关心这些。据说宝
马没有备胎,因为车胎扎了还能跑100公里,足以找到换胎的
地方。当然这点我是存疑的,因为在中国找一个方圆100公里
换不到宝马胎的地方应该还是很容易的。不过在说明书上确
实发现了轮胎胎压不足的时候车内液晶屏上的自动显示。

主要说驾驶感受,这一点宝马可以说是无可挑剔,虽然宝马3
和奥迪A4应该是同一档次的车,可是带来的驾驶感觉完全不
一样,如我所说,奥迪的驾驶感觉是很有“自己个性”的
车,驾驶者和车是有一种疏离感,人要有一个适应的过程。
而宝马在驾驶的过程中,是人车合一的,就是你马上就知道
你方线盘动多少角度的时候车会拐多大的弯儿,而你油门踩
下去的时候车能有多快。在刚坐到车里的时候就让人感觉你
是和“她”早就相识很久的老朋友。这种贴合也会让人“不
适应”,是那种“怎么会这样”的“不适应”。说实话,在
开宝马之前,我对自己的切诺基很有感情,就像自己一个有
若干缺点的老朋友,平常偶尔犯葛,关键时候却能共同依赖
摆脱困境。虽说随着油价的上涨也有想卖车的想法,却对下
一辆的候选车心存不屑,老觉着即使好车固然买不起,可即
便买得起也未必比我的切诺基好到哪去,要是有一辆全新的
原产6缸切诺基只怕给我宝马也不换。可是这一次的宝马给我
的感觉让我觉得完全不同,我甚至认真算了一下自己有没有
买宝马的能力。

早上上班的时候四环堵车堵的厉害,我对自己对宝马的认识
还没有太多的信心,尤其是刹车距离和急加速时的加速度,
开的还很拘谨,有很多车从我左右穿插超车,似乎是对我的
嘲笑。待到车距在我平时能把握的1.5到两倍的时候,我也经
常在车流中钻来钻去,感觉游刃有余。即使是那些因为我的
保守而超过我的车只要有机会我可以立马反超,而且绝对不
需要压到对方踩刹车的程度。走机场高速一段上五环,车流
少的时候油门一踩就轻松上120,因为我早已决定在任何时候
都不超速到了限速就收油保持车速所以没去试车速的极限
值,当然可能也没这个胆量。在五环到奥运场馆的一段连续
很长距离的大弯路,尝试80公里时速过大弯,人几乎贴到车
门的玻璃上,激发了前面一辆凌志350的斗志,加速颠了,
我没稀的追丫。下班走的略晚,四环上已经有足够的空间你
追我赶,在亚运村附近左道一串车开的又慢又不打算让开左
道,我不想晃远光提醒,而是看了后视镜后面没车直接向右
跨两个车道从右侧超越一串车,一瞥车速已经超过100,赶紧
收油回到80左右,心想这车太容易超速了。于是后面一段路
程,始终保持80的时速,也不轻易变换车道了。经常有各种
车辆开到并行然后加速超过去,我看着那些车心说“又一个
超速的”。然后心里比划着他的车速想我可以用宝马如何变
线超过他们,只是再也没有超过一辆,直到到家。

Posted by hufey at 11:33 PM | Comments (0)

March 11, 2008

Idol Jobs


Posted by hufey at 10:58 AM | Comments (0)

February 14, 2008

smartphone 用户满意度调查


PALM居然是最低,真没想到。。。
MOTO倒数第二。。。

Posted by hufey at 10:56 AM | Comments (0)

February 10, 2008

build iphone toolchain

今天尝试了一下iphone toolchain的安装
先是装了一个编译好的版本, 声称支持ppc
装的时候很顺利,就是要自己下载一个iphone filesystem
结果运行的时候报错, 说是cpu必须是x86
卸载之
教程下载svn版本
编译安装,结果在./configure的时候失败,看log,发现原来是运行/usr/bin/ld的时候出错
file /usr/bin/ld,原来也是x86的版本,file /usr/bin/gcc发现还是universal binary format
看/usr/bin/ld的时间, 推算起来就是装那个toolchain的时间,看来那个包居然把系统的ld都改成自己的了
想了想似乎只能重装xcode才能恢复原来的文件, mount xcode.dmg
看到xcode里其实有分开的多个pkg文件,不过不知道是哪个包含ld
懒得挨个去试, 于是安装xcode 同时不断运行file /usr/bin/ld,最后发现是DeveloperToolsCLI这个包包含的ld. 完成之后再看ld已然是universal binary格式,对apple这个同时支持x86和ppc的技术很好奇, 有空研究研究, 目前先存疑
再去按教程编译, 在指定SDK的时候发现自己的leopard既有10.4的也有10.5的, 想了想,也许教程写的早了些没眼正过10.5的吧, 于是指定10.5的安装头文件, 前面出了一些错误滚动过去了, 结果在编译gcc的时候头文件又出错, 找不到arm下的OSByteOrder.h. 看来还要用10.4的SDK.
再重复上次动作, 换10.4 的SDK, 再编译gcc, 正常过去了
编译helloworld验证, 改arm-apple-darwin-cc为arm-apple-darwin-gcc, 编译找不到crt1.10.5.so, 发现上边漏了软连接crt1.so一步, 链接之, 再编译, 出undefined symbols错误
_objc_sendMsg, google之, 发现其实就在教程里的comments就有人遇到并解决了这个问题
加-lobjc.
按人家说的修改Makefile编译
上传到iphone上, ssh登录并运行, 没问题, 运行的很好.
下边的问题是我给iphone写个什么程序呢????

Posted by hufey at 11:46 PM | Comments (0)

February 09, 2008

《出非洲记》读书笔记

春节长假看了一堆人类学书籍,前几本看的很轻松,睡眼蒙胧的就看完了,没想到最薄的《出非洲记》确是最艰深看的最痛苦的。相信作者已经尽力写的通俗易懂了,不过关于如何用DNA甄别人类迁徙分布进化的一段还是看的很懵懂,想当年我还是高中生物课代表,70分满分的生物能打67分呢。看到书中一个Y染色体谱系分布图很感兴趣。google了一下,找到的是wikipedia上的,被咱们ZF伟大光荣正确的GFW给挡住了,只好带套浏览。顺便把两张图片存下来贴自己的blog里。

人类Y染色体DNA单倍型类群
[编辑首段]维基百科,自由的百科全书
跳转到: 导航, 搜索

人类Y染色体DNA单倍型类群(Human Y-chromosome DNA haplogroups),是利用Y染色体遺傳變異特性進行人類學研究的一門科學,主要用于研究人類的“非洲起源論”及以後的種群分布的遺傳學證據。
目录
[隐藏]

* 1 原理
o 1.1 当前统计数据
o 1.2 说明
* 2 走出非洲实例分析
o 2.1 推测棕色人种C=M130的迁徙路线
* 3 注释
* 4 外部鏈接

[编辑] 原理

人類有23對46条染色体,其中22對44条为常染色体,另外一對為性染色体,XY组合的为男性,XX组合的为女性。Y染色体只能父子相传,所以研究Y染色体,可以发现人群在父系關系上的迁徙和发展。
Y染色体谱系树FamilyTreeDNA的更詳細的谱系树Genome.org的更詳細的谱系树
Y染色体谱系树
Y_Haplogroups_Tree.jpg
FamilyTreeDNA的更詳細的谱系树
Genome.org的更詳細的谱系树

上面的人类Y染色体谱系树是根据Y染色体单倍型类群的不同把全部现代智人分为18个类型,用从A到R的十八个字母作为索引。谱系树中的父节点代表的对应基因突变是所有子节点共有的,但反之不然。

[编辑] 当前统计数据

公元1500年世界人群分布的Y染色体单倍型类群圖

在「Y-chromosome SNP testing at DNA Heritage」這個鏈接里,如果你的瀏覽器支持Flashplayer,則可把鼠標移到Y染色体谱系树的各個索引字母上看各個单倍型類群的獨立分布。

[编辑] 说明

按此谱系树,现今所有人类的Y染色体单倍型类群的根都能在非洲找到。A=M91和B=M60也是非洲居民的特征。而出走到非洲以外的居民后裔,包括棕色人种、蒙古人种(黄色人种)和高加索人种(白色人种),其Y染色体上都带有M168的突变点。

此后在M168突变的基础上,又分别产生了C=M130、DE=YAP和F=M89三个子类型,其中DE=YAP分为D=M174和E=M96两种;DE=YAP又被称为小黑矮人(Pigmy Negroid)基因,和C=M130几乎同时走出非洲。现在DE=YAP主要分布在印度安达曼群岛的安达曼人(达100%)、藏族(58%)、土家族、彝族、瑶族、日本(34.7%)、朝鲜、满族、缅甸人、克钦人。在汉族人中,D=YAP出现的频率小于1%。F=M89则是所有其他人群共有的,现今大概全世界80%以上的人都有这个变异点。

在F=M89的基础上,又产生G=M201、H=M52、I=M170、J=M304和K=M9=K*几个子类型,其中K=M9是最重要的一个子类型,亚欧大陆上除西亚和西伯利亚地区,其绝大部分民族都共有这个变异点,现今中国汉族中,96%的人都是K=M9类型。

在K=M9的基础上,又产生了K1=M177、K2=M70、K3=M147、K4=M230、L=M11、M=M4、NO=M214、P=M45=P*几个子类型,其中NO=M214又分为N=M231和O=M175两个子类,N=M231主要分布在乌拉尔山两侧、北欧北部、东欧北部和北极圈内的爱斯基摩人,而O=M175主要分布在东亚、东南亚、北亚东部、部分太平洋岛屿,现今中国汉族的主要类型就是O=M175。

在P=M45的基础上,又产生了Q=P36和R=M207两个子类,其中R=M207分为R1a=M17和R1b=M173两个子类。Q=P36包括美洲土著印地安人;R1a=M17包括西亚的伊朗、南亚的印度;而整个R=M173类型被认为古雅利安人的基因,分布在欧洲大部和西亚、南亚。

W-MAP

一个很cool的flash

Posted by hufey at 03:23 AM | Comments (0)

February 07, 2008

worldwide smartphone marketing share q4 2007 vs q4 2006


RIM更疯狂

Posted by hufey at 10:21 AM | Comments (0)

January 31, 2008

Moto said, "We were over Qt, anyway."

我预料的应验了。第一个就是MOTO。不过这个对MOTO的打击可是够大的。切换到另一个平台不是说换就换的。从打击对手角度来看,NOKIA的这次收购太成功了。

news from linuxdevices.com
--
Motorola responds to Nokia's Trolltech buy Jan. 30, 2008 Motorola has responded to the news earlier this week that rival Nokia plans to purchase Trolltech, long-time supplier of the graphical development framework used in Motorola's Linux phones. In a nutshell, the response boils down to, "We were over Qt, anyway." Christy Wyatt, who heads up Motorola's software platforms and ecosystems group, told LinuxDevices, "A year ago, we announced that we had founded LiMo [the Linux Mobile Foundation], along with five other companies, to collaborate on mobile Linux [story]. Since then, many other companies have joined. As part of its first specification, LiMo specifies GTK, and we intend to comply." Wyatt admitted that Motorola has no definitive "cut-over date" yet planned for switching from Qt to GTK. She said, "We did evaluate qt4. What we're using is an older version. Limo hasn't published that part of the framework yet, and there's really been no reason for us to rush ahead of the Foundation in moving to GTK." Wyatt said that in the near term, Motorola is working to document the places in its MotoDev tools where the Qt API is exposed. She said, "We're documenting them so developers will know when they're touching a part of the API that might go away." Asked about Motorola's dependence on Trolltech for development tools, or for other parts of Motorola's Linux phone stack, such as webkit integration, Wyatt said there was no reason for concern. "MotoDev Studio is in no way, shape, or form built around Qt tools," she said, adding, "We did actually buy out a part of our license earlier. And our web UI framework, that's a Motorola implementation. We've had it working for a long time." Wyatt concluded, "In the mobile Linux space, Motorola believes in open standards and open source. We encourage Nokia and others to participate in the open development model [exemplified by] LiMo." Asked specifically if she thought Motorola's five-year long embrace of Linux had hurt or helped it, Wyatt laughed and replied, "It's hard to draw a direct connection between one program and a company the size of Motorola. Adopting Linux lets us participate in a thriving open source ecosystem. Linux is a long-term bet for any company. The investiment -- our investment -- is toward leveraging ecosystem, working with partners in the supply ecosystem. We're very happy with our Linux products."

Posted by hufey at 11:07 AM | Comments (0)

January 29, 2008

nokia 收购 trolltech


半夜失眠,起床上网,没想到看到的是这样的消息。
虽说被收购不能说就丧失了独立性--Trolltech的公开信也说了会继续以GPL开放QT的源代码--,但至少业界的nokia的对手们是不是还会继续使用qt作为framework都会考虑再三了。现在是互相进入对方市场的混战时代,即便是google也不会变相给nokia送钱的。
说不定这反倒是gtk的一个机会。

Posted by hufey at 01:55 AM | Comments (0)

January 16, 2008

US Smartphone marketing share

BB居然有39%那么多???真没想到!!!

Posted by hufey at 01:47 PM | Comments (0)

November 08, 2007

终于用上了leopard

过程是如此的艰辛
IT人的历史往往伴随着追捧软件最新版本的历史
当年大家都有MSDOS 3.3的软盘,最好是3.31的
除了经典的3.3,还要有高版本MSDOS,从5.0到6.0,到6.2,到6.22
后来是Windows 3.1,PWindows 3.2。记得Windows 95出来的时候,我在上海郊区,
虽然是上海,但却是郊区,信息不够发达,连盗版盘都不知道去哪买。哈尔滨的同学特快专递给我
一张Windows 95的光盘。
再后来是Windows 98,Windows ME。Windows 2000从Beta版开始用来架Web Server。
迷上Linux后就是RH 5.2,RH 6.0,RH 6.4,RH 7.0,RH 9.0,Fedora,Mandrake,SUSE,GENTOO,Debian,Ubuntu,一路走来
现在是OSX。
迷上MAC比较晚,用上已经是Tiger了。然后就盼着花豹。
买正版是不太可能了,一直等着下载。
等BT下载了几天,Azureus报告磁盘读写错误。
重新下载。
极慢。
看到有D5的下载,赶紧放弃D9的。
D5的下完,是spaseimage格式。GOOGLE之。先转成DMG格式,再转成ISO格式。
烧盘。启动。启不来。
上论坛问别人,没有答案。
再烧一张,还是如此。
直接上淘宝买一张D9,安装。
耗时一晚上,终于安装成功。
继续安装Optional Installation,失败,失败,失败。Installation Disk Damage。
气愤!!!

LP问我,新版本有什么新功能?
我答曰:你看,多了好多新壁纸耶。

Posted by hufey at 04:25 PM | Comments (0)

November 01, 2007

来自Google Analytics对rovert love blog的统计

来自Robert Love的Blog
os-stats-20071030.jpg

And Google Analytics does operating system statistics, too:

* Windows 57%
* Linux 32%
* Mac OS X 11%

browser-stats-20071030.jpg

# Firefox 61%
# Internet Explorer 25%
# Mozilla SeaMonkey 5%
# Safari 3%
# Opera 2%

看来hacker和关注hacker的人中用linux和firefox的人还是很多

Posted by hufey at 11:39 AM | Comments (0)

October 27, 2007

git pull error

好多网上的文档都过期了

git pull http://xxx.xxx/xxx.git出现
error: pick-rref: HEAD not found
的错误,没有一个网页是有价值的
其实只要在该目录中直接git pull就可以了

Posted by hufey at 11:23 AM | Comments (0)

October 25, 2007

montavista 和 access的合作

对双方对业界都是一个好事。对montavista,争取一个大客户而且该大客户本来就是一个非常有实力的竞争对手。对access来说,节约了成本,专注在自己擅长的方向。对业界来说,access出货变快了可以使linux占领更大的市场。尤其是在moto开始三心二意的当下。

Posted by hufey at 11:04 AM | Comments (0)

October 22, 2007

ubuntu 7.10 升级后中文方块问题的解决

除了需要fontconfig-voodoo -f -s zh_CN之外,还要
sudo fc-cache -f -v

大概是某些config文件没有Yes覆盖而是保留原来的结果

Posted by hufey at 03:58 PM | Comments (0)

October 16, 2007

Which companies are helping developing the kernel

Re: Which companies are helping developing the kernel
Previous message: [thread] [date] [author]
Next message: [thread] [date] [author]
From: Greg KH
To: Alistair John Strachan
Cc: Stefan Heinrichsen ,
Subject: Re: Which companies are helping developing the kernel
Date: Sunday, October 14, 2007 - 5:59 pm

On Sun, Oct 14, 2007 at 11:28:46PM +0100, Alistair John Strachan wrote:
quoted text
> On Sunday 14 October 2007 23:06:22 Stefan Heinrichsen wrote: > > Hello, > > > > I posted this question at comp.linux.misc and where told this would be a > > better place therefore. I would like to do a internship in the field of the > > Linux kernel. > > Can someone tell me where to find a list of companies (don't matter in > > which country) that employ kernel developers? > > I think Greg wrote a paper on this subject, so I've added him to CC in case he > has the link handy.

Yeah, but my paper didn't really track companies very well. The lwn.net
article is the best, and below is my version of who did things in
2.6.23. Note, the lack of a company is not an indicator that they did
nothing, just that I could not easily determine someone worked for them.
I'll try to send out my "who are you working for" emails in a week or so
to see if I can further categorize the "unknowns".

thanks,

greg k-h
-------------------------

Processed 7075 csets from 992 developers
126 employers found

Top changeset contributors by employer
(Unknown) 1116 (15.8%)
(None) 843 (11.9%)
Red Hat 827 (11.7%)
IBM 557 (7.9%)
Linux Foundation 528 (7.5%)
Novell 449 (6.3%)
Intel 242 (3.4%)
Oracle 158 (2.2%)
MIPS Technologies 143 (2.0%)
Nokia 133 (1.9%)
NetApp 119 (1.7%)
NTT 99 (1.4%)
Astaro 97 (1.4%)
MontaVista 90 (1.3%)
(Consultant) 86 (1.2%)
SGI 84 (1.2%)
Qumranet 74 (1.0%)
QLogic 70 (1.0%)
(Academia) 70 (1.0%)
SWsoft 64 (0.9%)
Analog Devices 61 (0.9%)
HP 60 (0.8%)
Sony 59 (0.8%)
rPath 56 (0.8%)
XenSource 53 (0.7%)
CERN 49 (0.7%)
CC Computer Consultants 48 (0.7%)
Freescale 47 (0.7%)
Fujitsu 47 (0.7%)
Tripeaks 46 (0.7%)
linutronix 44 (0.6%)
Snapgear 39 (0.6%)
Simtec 34 (0.5%)
Atmel 28 (0.4%)
Google 28 (0.4%)
Cisco 27 (0.4%)
Toshiba 25 (0.4%)
Broadcom 25 (0.4%)
SteelEye 24 (0.3%)
Renesas Technology 23 (0.3%)
Mellanox 21 (0.3%)
LSI Logic 17 (0.2%)
Adaptec 16 (0.2%)
Wipro 15 (0.2%)
Marvell 14 (0.2%)
Miracle Linux 14 (0.2%)
Solid Boot Ltd. 14 (0.2%)
AMD 12 (0.2%)
Hitachi 11 (0.2%)
ARM 11 (0.2%)
Canonical 10 (0.1%)
XIV Information Systems 9 (0.1%)
OpenedHand 9 (0.1%)
Open Grid Computing 9 (0.1%)
Veritas 8 (0.1%)
Secretlab 7 (0.1%)
Neterion 7 (0.1%)
Katalix Systems 7 (0.1%)
SANPeople 7 (0.1%)
Digi International 7 (0.1%)
Znyx Networks 6 (0.1%)
Wind River 6 (0.1%)
NEC 6 (0.1%)
Wolfson Microelectronics 6 (0.1%)
SUNY Computer Science 6 (0.1%)
NetXen 6 (0.1%)
NVidia 6 (0.1%)
Myricom 6 (0.1%)
Chelsio 5 (0.1%)
Realtek 5 (0.1%)
IPUnity-Glenayre 5 (0.1%)
Linux Networx 5 (0.1%)
Barco 4 (0.1%)
SIOS Technology 4 (0.1%)
MIPS 4 (0.1%)
University of Aberdeen 4 (0.1%)
PiKRON s.r.o 4 (0.1%)
Pardus 4 (0.1%)
Crash Barrier 4 (0.1%)
Tresys 4 (0.1%)
ClusterFS 4 (0.1%)
Macq Electronique 3 (0.0%)
OLPC 3 (0.0%)
CompuLab 3 (0.0%)
DENX Software Engineering 3 (0.0%)
Embedded Alley Solutions 3 (0.0%)
Twin Sun 3 (0.0%)
Transmode Systems 3 (0.0%)
Cosmosbay~Vectis 3 (0.0%)
Pengutronix 2 (0.0%)
Sierra Wireless 2 (0.0%)
Real-Time Remedies 2 (0.0%)
Bull 2 (0.0%)
ScaleMP Inc. 2 (0.0%)
Samsung 2 (0.0%)
MicroGate Systems 2 (0.0%)
Atomide 2 (0.0%)
US National Security Agency 2 (0.0%)
VMWare 2 (0.0%)
Verismo 2 (0.0%)
Akamai Technologies 2 (0.0%)
CE Linux Forum 2 (0.0%)
Siemens 1 (0.0%)
ONELAN 1 (0.0%)
Xorcom 1 (0.0%)
EXOSEC 1 (0.0%)
LWN.net 1 (0.0%)
Texas Instruments 1 (0.0%)
Promise Technology 1 (0.0%)
VA Linux Systems Japan 1 (0.0%)
University of Sevilla 1 (0.0%)
Motorola 1 (0.0%)
HES-SO Valais Wallis 1 (0.0%)
SYSGO 1 (0.0%)
Via 1 (0.0%)
Telecom-Service 1 (0.0%)
Hauppauge 1 (0.0%)
DSA 1 (0.0%)
Vivecode 1 (0.0%)
Mandriva 1 (0.0%)
ATRON electronic 1 (0.0%)
Cytronics & Melware 1 (0.0%)
M&N Solutions 1 (0.0%)
Rockwell 1 (0.0%)
8D Technologies 1 (0.0%)
University of Cambridge 1 (0.0%)

Posted by hufey at 05:33 PM | Comments (0)

September 26, 2007

Who writes the Linux kernel?

没听说过以下公司:
Tripeaks, Emulex, rPath, Solid Boot, PMC-Sierra, Astaro

Posted by hufey at 04:00 PM | Comments (0)

September 16, 2007

SCO申请破产保护

王安是我知道的第一个电脑公司的名字,好像是八十年代初CCTV搞的智力竞赛,赵老师忠祥主持,王安电脑公司作为华人在美国成功的代表,组队参赛,好像拿了个第二(?)。我当时还不知道电脑是个啥东西。
苹果大概是我知道的第二个电脑公司的名字,因为风传所在的中学花巨资买了两台“苹果二”电脑,放在专用的机房里当宝贝供着,只有某专职老师和校长公子可以进去使用。我甚至都没机会见到。
四通联想浪潮是在电视上看到的广告。四通的比较科幻,霹雳贝贝风格,看完了以为四通的人都是外星人派来的。联想好像是“人类没有联想,世界将会怎样”,比较傻。浪潮是一个大浪打过来,然后是“浪潮0520,如何如何”,一直也不明白0520是什么意思。
然后就是SCO了,电视台有个SCO UNIX的讲座,我那会对跟电脑搭边的东西都特别感兴趣,苦于没有机会接触电脑。于是盯着这个讲座连续学了好几期。记得第一讲是讲“ls”的用法,讲了几十分钟。想来这个讲座也是针对企业中使用SCO UNIX的用户,因为从那之后的十几年里我没用过SCO UNIX,甚至认识的人里也没有用过SCO UNIX。所谓学习,就是把电视上讲的内容,记到纸上,想象如果我坐到电脑前,应该怎么使用这些命令。
对SCO一直有莫名的好感,因为那些记在纸上学到的命令和概念在我几年之后在大学里接触学习SUN III工作站的时候上手比别人快了一些。
这种好感一直延续到SCO和Linux打官司之前。
现在可倒好,官司失败了,SCO也流落到破产保护的境地。

Posted by hufey at 10:28 AM | Comments (0)

September 08, 2007

不带这么拿人打嚓的

有一哥们一直用root登录使用fedora,但是auto login不能以root自动登录,这哥们居然给linux kernel mailinglist发了一个邮件如下:

Dear whoever is in charge of writing fedora core 7,

The auto login is a wonderful feature. However, there is a HUGE flaw in
it at the moment. It won't let me log in a root that way. That is MY
decision, not YOUR decision. Please fix it in an update!!! I'm sick of
having to log in as root every single time I use my computer. I
understand that there are situation where people have other people use
their computer and don't want them to be root. However, I never want to
be logged in as anything other than root, and I suspect everyone who has
half a brain is the exact same way. Honestly, have you ever been logged
in as someone other than root? You absolutely can do nothing on the
machine. For those who make the argument that someone logged in as root
might break the machine by doing something they shouldn't, I would say
it's already broken without being root. You cannot mount drives, edit
anything about how the /etc directory is setup, or do many other things
that everyone I know does on a regular basis. IN general, there are
times when you are logged in as a non-root user when it will ask you for
a password. That is so inconvenient when you are the only one who uses
your own computer. PLEASE get root auto login as an option. No one but
me has to set it up this way, but please give ME the option.

Thank You,

几个人就他的错误行为给出了简短的回复,例如给fedora提bug之类的。

也有人比较刻薄,回复如下:

Hint 1: If you're not smart enough to figure out how to fix your system so that
it does what you want, you're not smart enough to run as root all the time.

Hint 2: You're posting to a kernel list about a non-kernel problem. See hint 1.

结果另一人跟着回复如下:

>Hint 1: If you're not smart enough to figure out how to fix your system so that
>it does what you want, you're not smart enough to run as root all the time.

yum install smart

Posted by hufey at 06:31 PM | Comments (0)

August 08, 2007

垃圾邮件长本事了

居然会自动生成跟邮件列表名称相关的subject和附件。虽然内容还是垃圾

[Ltp-coverage] Technical_Analysis-ltp-coverage Inbox ltp-coverage

Reply
Reply to all Reply to allForward Forward Print Add rhadona to Contacts list Delete this message Report phishing Show original Message text garbled?
rhadona Renfroe
to ltp-coverage

show details
7:32 am (3 hours ago)

-------------------------------------------------------------------------
This SF.net email is sponsored by: Splunk Inc.
Still grepping through log files to find problems? Stop.
Now Search log events and configuration files using AJAX and a browser.
Download your FREE copy of Splunk now >> http://get.splunk.com/
_______________________________________________
Ltp-coverage mailing list
Ltp-coverage@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/ltp-coverage


Technical_Analysis-ltp-coverage.pdf
13K View as HTML Download

Posted by hufey at 10:30 AM | Comments (0)

June 13, 2007

用Tor突破GFW

听说很久了,一直没试过,因为那些反动的带色的网站看够了,曾经为wiki和google cache烦恼过,时间一长居然习惯了。
不过最近被GFW封的网站越来越多,甚至flickr.com。简直没有天理。
看到洪文《互动:一个网络警察的自白-》让我想起还有Tor可用。
装之,Tor支持Windows, MAC OSX还有linux,我就挨个装个遍。
再上wikipedia, flickr,都没问题了,就是慢点。上网慢了点,机器速度好像也拖慢了不少。

Posted by hufey at 11:33 PM | Comments (0)

June 10, 2007

装嫩

看最近twitter类服务火爆,也去注册了twitter和饭否
贴在blog的边上
感觉这都是80后,90后甚至00后才喜欢的咚咚吧
虽然自己也不时改改msn的名字,好像自己到不感兴趣别人是不是对我的改名感兴趣
不过不妨尝试一下流行的东西,省得自己老的太快

Posted by hufey at 10:16 AM | Comments (0)

June 07, 2007

QT的Java API

上边的Application应该为Java Application才对,native application肯定还是直接和QT API打交道

Posted by hufey at 10:52 AM | Comments (0)

June 05, 2007

Poll Results: Will you line up for an iPhone?

TUAW的大饼,居然有更多的人认为No, I have better things to do



Will you line up for an iPhone?








Selection 
Votes
No, I have better things to do 45%792
I'm considering it 17%290
If I have the free time, yes 9%152
OMG ABSOLUTELY!! 29%513
1,747 votes total

Poll results are subject to error. Pollhost does not pre-screen the content of polls created by Pollhost customers.







Posted by hufey at 10:33 PM | Comments (0)

May 08, 2007

robert love去了google

一声叹息

Prologue: Google!

I am jubilant to announce that I have accepted an offer of employment from Google, a public corporation specializing in search, advertising, and unabated innovation.

Joining the staff of the Open Source Program Office, purveyors of such wonderful offerings as the Summer of Code, I shall remain in Boston.

I will leverage a few weeks of anything-but-relaxing unemployment to finish my latest book, move across the river, and advise political campaigns, then I will hit the ground running.

Posted by hufey at 12:08 AM | Comments (0)

May 01, 2007

Why device developers prefer Debian


by Henry Kingman (Apr. 30, 2007)

LinuxDevices.com's survey results consistently show Debian to be the most popular distribution among device developers. For example, our 2007 survey indicated that Debian was used in device-related projects by 13 percent of the survey's 932 participants, roughly double the score of MontaVista, the most popular strictly-embedded distribution.


What sources of Linux were/will be in your (company's) embedded designs?
(Click to enlarge)


Spread the word:
digg this story
In addition to Debian's 13 percent score, Ubuntu, which is based on Debian packages, jumped to 6 percent this year, its first year in our survey. In contrast, Red Hat, achieved a 5 percent score and Fedora came in at 6 percent, while SUSE scored just 2 percent. The complete results and analysis are here.

Why do device developers prefer Debian?

Here are what I believe are the top reasons:

1. All the latest packages --Ironically, Debian is sometimes faulted for infrequent stable releases. But developers don't care so much about the kind of "stability" that the stable branch buys you -- desktop integration and UI fit and finish. They all run the "unstable" branch, and generally apt-get important packages weeks, days, or hours after they are released upstream -- and months before they appear in packaged distributions.

2. It's comprehensive -- Last time I checked, Debian had more than 14,000 packages. It's rare to find an open source software package that is not maintained in Debian. The average maintainer's age is about 23. Learning to build and package software is a great introduction to open source programming.

3. It's never going out of business -- Debian is not in any way affiliated with any commercial venture, and thus is pretty much not going to be the next BeOS.

4. It's convenient -- Hm, I wonder if there's an open source software package that does xyz. Apt-cache search xyz. Dpkg -l '*xyz*'. Apt-cache show xyz-utils. Apt-get install xyz-utils. It's as simple as that. And probably even simpler with aptitude, which is what we're all supposed to be using by now, you know.

5. 11 architectures -- Actually, I think there are more architectures than this, but only 11 appear to have installation manuals (those incredibly prolix things! Someone should take a machete to them). Developers (especially device developers) like lots of architectures, since they all have their strengths (x86 for complex multi-tasking UIs, ARM for low power, MIPS and PowerPC for extensibility, etc).

6. It's infinitely malleable -- When you install Debian, you can opt for one of the "tasksel" packages, such as "desktop" or "print server," etc., etc. However, you don't have to. You can still very easily install only the most minimal of components, and then simply add what you need, as you need it. After a week or so, you've got everything you need, and nothing you don't -- a great strategy for increasing security and getting the most from your system.

7. You're not stuck inside Debian -- If you have to install RPMs, you can convert them into DEBs using the alien scripts. If you want to build stuff from source, you can "hold" packages so your custom-built stuff won't get clobbered by the next system-wide upgrade.

8. Easy system maintenance, even between major releases -- With each new release, instead of downloading a CD and wiping out your root filesystem, just apt-get dist-upgrade. And, you can sync up to the latest and greatest versions of everything as often as you like.

9. Support of other Debian users -- Debian users tend toward omniscience, or at least fiercely competitive peer review, even of support advice. The #debian channel on freenode is a great place to observe this in action, and get help solving any question you like (as long as it's actually a Debian question).

为什么编辑忽视几点:
1,最高的是kernel.org
2,debian比例在下降
3,比例上升的几个幅度很大
4,最commecial的是MontaVista
5,Ubuntu是更好的debian

Posted by hufey at 08:58 AM | Comments (0)

April 19, 2007

he wrote the 100K patch in 62 hours

Linux: The Completely Fair Scheduler
April 18, 2007 - 5:51am
Submitted by Jeremy on April 18, 2007 - 5:51am.
Linux news

Ingo Molnar [interview] released a new patchset titled the "Modular Scheduler Core and Completely Fair Scheduler". He explained, "this project is a complete rewrite of the Linux task scheduler. My goal is to address various feature requests and to fix deficiencies in the vanilla scheduler that were suggested/found in the past few years, both for desktop scheduling and for server scheduling workloads." The patchset introduces Scheduling Classes, "an extensible hierarchy of scheduler modules. These modules encapsulate scheduling policy details and are handled by the scheduler core without the core code assuming about them too much." It also includes sched_fair.c with an implementation of the CFS desktop scheduler, "a replacement for the vanilla scheduler's SCHED_OTHER interactivity code," about which Ingo noted, "I'd like to give credit to Con Kolivas [interview] for the general approach here: he has proven via RSDL/SD that 'fair scheduling' is possible and that it results in better desktop scheduling. Kudos Con!"

Regarding the actual implementation, Ingo explained, "CFS's design is quite radical: it does not use runqueues, it uses a time-ordered rbtree to build a 'timeline' of future task execution, and thus has no 'array switch' artifacts (by which both the vanilla scheduler and RSDL/SD are affected). CFS uses nanosecond granularity accounting and does not rely on any jiffies or other HZ detail. Thus the CFS scheduler has no notion of 'timeslices' and has no heuristics whatsoever. There is only one central tunable, /proc/sys/kernel/sched_granularity_ns, which can be used to tune the scheduler from 'desktop' (low latencies) to 'server' (good batching) workloads." He went on to note, "due to its design, the CFS scheduler is not prone to any of the 'attacks' that exist today against the heuristics of the stock scheduler".

During the followup discussion, Ingo explained that he wrote the 100K patch in 62 hours. In response to concerns that his efforts had not been discussed first on the Linux Kernel Mailing List, Ingo explained, "I prefer such early releases to lkml _alot_ more than any private review process. I released the CFS code about 6 hours after i thought 'okay, this looks pretty good" and i spent those final 6 hours on testing it (making sure it doesnt blow up on your box, etc.), in the final 2 hours i showed it to two folks i could reach on IRC (Arjan and Thomas) and on various finishing touches." He went on to add, "the 'design consultation' phase you are talking about is _NOW_! :)" Later in the discussion that touched on egos, Linux creator Linus Torvalds noted, "one of the most motivating things there *is* in open source is 'personal pride'," going on to add, "it's a really good thing, and it means that if somebody shows that your code is flawed in some way (by, for example, making a patch that people claim gets better behaviour or numbers), any *good* programmer that actually cares about his code will obviously suddenly be very motivated to out-do the out-doer!"

Posted by hufey at 02:11 PM | Comments (0)

March 15, 2007

在google定制主页里加gtalk插件

点这里安装
效果是这样地

Posted by hufey at 05:55 PM | Comments (0)

Locationbar²

Locationbar²
很不错的firefox extension,记得把这个选项打开

Posted by hufey at 05:35 PM | Comments (0)

January 18, 2007

重新看好Symbian

Embedded OS的战争最后往往是API的战争。
Windows CE有广大的Windows 程序作基础, Linux有Posix API。只有Symbian(以前)是与众不同的。虽然Symbian是smartphone OS的老大,也养活了一堆小公司,但是对于一个第三方公司想为不同的系统写程序去维护几套API是痛不欲生的。Porting现有的程序也很困难。
直到今天看到Symbian的这个举措,我才相信Symbian的未来是可持续发展的。

--------------------------------------------------------
Symbian OS to gain POSIX libraries
Jan. 17, 2007

Symbian plans to beta test POSIX libraries for its Symbian OS this quarter. The PIPS (PIPS is POSIX on Symbian) libraries provide C and C++ APIs (application programming interfaces) in standard "libc," "libm," "lipthread," and "libdl" libraries, with the aim of helping programmers migrate existing middleware and applications to Symbian.

The PIPS libraries were announced Jan. 16 in Beijing, where Symbian opened an office.

Symbian says PIPS will significantly reduce the work of migrating existing desktop and server components to Symbian OS. Typical applications might include web servers and file sharing software, the company says. For instance, mobile phone marketshare leader Nokia, which licenses Symbian for its smartphones, ported Apache to Symbian last summer.

Symbian executive VP Jorgen Behrens stated, "It is now realistic and desirable to migrate desktop and server code onto mobile devices."

Symbian claims that there are about 100 million Symbian-based smartphones in the market, making it the top smartphone OS. However, Symbian's fortunes may be falling, according to research reports last year from ARCchart, ABI, and The Diffusion Group.

Availability

A beta version of PIPS will be available for Symbian OS v9.1 (and above) as a downloadable .SIS file from the Symbian Developer Network by the end of Q1 2007, Symbian says.

Posted by hufey at 11:19 AM | Comments (0)

December 11, 2006

QuickCheck

删垃圾留言删的我头昏眼花
想着要不要自己写一个extension来快速删除
想了想。。。估计有人写过了
于是找到这个:
QuickCheck

Posted by hufey at 08:10 PM | Comments (0)

December 01, 2006

推荐本书


Advanced Programming in the UNIX Environment

一个新入行的MM程序员让我推荐Linux编程方面的书
我说linux下的编程其实都是标准C和posix api,不用另买书看
她问我什么是posix
我说是一套IEEE定的api标准
她有问有没有哪本书讲posix标准的
我想了半天却想起来大概7年前我做的一个项目,大概就是给一个简陋的RTOS做一个文件系统
我非常希望做到posix-compatible,但是没有讲posix细节的文档,于是上网搜,搜出来的
文档都是要花钱去IEEE买的,而那时的公司又不愿意出这笔钱(好像不是个小数目),最后就不了了之了。当然我还是尽我所能去写了那些代码来贴近我理解的posix。
后来落下了个心病,去书店逛的时候希望能找到一点posix的雪泥鸿爪。
直到在(大概是三联)书店发现了这本UNIX环境高级编程。当时还是第一版,其实我到现在也没完整看完一遍,大概看了2/3内容。多数都是遇到问题时去查。

本来是打算给MM发pdf的,不过在joyo买东西打算凑订单,结果偶然发现了此书的第二版。当即拿下。不过要送人,自己还没打算买。原价99元,即便我以VIP价买也要67块,还是挺贵的。

另外,七百多页的书确实有点太厚,家里书架已经没地方了。。。。。。

Posted by hufey at 04:43 PM | Comments (0)

November 04, 2006

web-based image editor

本来只是想upload一个图片到google group member profile
没想到自己常用的图大小不合适,然后发现ibook里居然没有软件可以做图片编辑
装photoshop太痛苦了,想google一个小软件来着
后来冒出来一个想法,既然现在web上连office都有了,说不定也该有一个photoshop了
google "image edit web"结果还出来不少
找到一个www.myimager.com
嘿嘿 功能还真不少

顺便又找了一个web-based diagram editor :D

Posted by hufey at 11:21 PM | Comments (0)

哎呀,好恶心


一贯投机的Novell和一贯恶心的Micro$oft
听起来是好事
但是我怎么就恶心的直起鸡皮疙瘩呢???

Posted by hufey at 05:59 PM | Comments (0)

October 26, 2006

装上了FF2和IE7

FF2之前在公司某台电脑上装过一个Beta版,印象很不错
这次是在自己的电脑上,来正式release的吧
IE7不知为什么Windows update里没有,只好下载再装
比较了一下,IE7的UI设计实在是太丑了,难道微软的人都跑光了?

另外把bloglines的feed都挪到了google reader上
嗯,不是bloglines不好,只是他的FF extension出的太慢
不过据说可以显示google reader的subscribe的extension GMark我还没研究明白。。。

Posted by hufey at 12:56 AM | Comments (0)

August 09, 2006

Andrew Morton去google了

Posted by hufey at 11:41 AM | Comments (0)

May 26, 2006

Linux DDK?


Greg KH要干什么?
下了一份,刻盘之前想了想还是先打开了iso文件看了一下
里面不过是LDD3的电子版和kernel 2.6.18的source code
这些在网上都能免费下到啊!

实在不行难道你不会把linuxjournal的历年文章做成pdf放里面吗?

Posted by hufey at 07:25 PM | Comments (0) | TrackBack

April 29, 2006

Apache vs. IIS

apache调用关系图

IIS调用关系图

Posted by hufey at 10:14 AM | Comments (0) | TrackBack

March 07, 2006

gentoo dev map


gentoo 开发人员在世界分布的地图
悲哀啊,中国一片空白

Posted by hufey at 12:00 PM | Comments (0) | TrackBack

March 01, 2006

freemind顿悟

知道freemind有很长时间了,一直没想明白对自己有什么用。
今天却有了顿悟的感觉
这个东西做项目管理、年度计划之类的再合适不过了。
以前上学的时候,每年做一个计划--当然不是课程方面的。
在小笔记本上--当然不是笔记本电脑--写几页纸。
后来这个习惯在毕业后持续了数年,再后来就忘了,再后来,就是现在又捡起来这个好习惯。
不过现在有电脑了,就想用软件来记。记得office 2003有个什么存笔迹的软件写过几回
发现并不好用,也就放弃了。
今天终于发现freemind太适合整理思路了
现在杂事比较多,正好可以freemind无限制分类。
而且键盘快捷方式也是我喜欢的。
边想编记,慢慢写了一大坨,单位的任务,自己要做的,技术的,非技术的。。。
非常不错。
要说不足就是图标不能自己添加,原配的图标根本不够用的。。。

Posted by hufey at 01:37 AM | Comments (2) | TrackBack

January 17, 2006

难道QT目前不支持GB18030?

Konsole在GB18030下不能用scim输入中文
改locale为gbk就可以了
gnome-terminal都可以,奇怪!

Posted by hufey at 12:48 AM | Comments (0) | TrackBack

December 20, 2005

subversion+apache2安装成功

自己家里也缺一个版本控制环境,正好之前听RedSox说起Subversion的好,也想尝试一下
新东西--目前已经用过VSS,CVS,ClearCase,大概只有svn没有被我临幸过了。

上网找howto,装!--应该说 emerge !
装完,设置,装TortoiseSVN client,browse repos,输入密码
不对!再输,还是不对。没辙上网求救吧,先是在paowang it发帖子,paowang it
虽然帖子不多,但是我深信有很多高手潜水看热闹。
在单位继续google,不得要领,搜到一个subversion.org.cn的网站
下300页pdf手册一本,看!
又注册id到该网站论坛发帖求助如下

=================================
hufey
斑竹


注册时间: 2005-12-20
帖子: 4


帖子发表于: 星期二 十二月 20, 2005 1:01 pm 发表主题: 请教svn和apache2的配置,现在密码访问通不过 引用并回复 编辑/删除帖子
我按照网上文档的步骤在gentoo linux设置好了apache2 mod_dav和svn
看上去似乎是对的
用的Basic AuthType,加用户、密码是htpasswd2生成的
但是当我 import project的时候(在linux里)
总是先问我要root的密码,然后要一个username和passwd
我把我刚设好的给他,要两次之后就fail了
我在另外一台windows机器tortoise连http://ip/svn/repos
也会要用户密码,同样我把刚设好的给他
他会一直不停的要
请问是何道理?
谢谢
返回页首
阅览成员资料 (Profile) 发送私人留言 (PM)
luke
斑竹


注册时间: 2005-11-01
帖子: 12


帖子发表于: 星期二 十二月 20, 2005 4:34 pm 发表主题: 引用并回复
感觉似乎是配置文件有问题,发上来看看。没有用ssl吧,如果最后import失败应该是配置不对,看看Apache日志文件里有什么错误。
返回页首
阅览成员资料 (Profile) 发送私人留言 (PM)
hufey
斑竹


注册时间: 2005-12-20
帖子: 4


帖子发表于: 星期二 十二月 20, 2005 7:19 pm 发表主题: 配置如下 引用并回复 编辑/删除帖子
/etc/conf.d/apache2中
APACHE2_OPTS="-D DEFAULT_VHOST -D SVN -D SVN_AUTHZ -D DAV -D DAV_FS"

/etc/apache2/modules.d/47_mod_dav_svn.conf



LoadModule dav_svn_module modules/mod_dav_svn.so


DAV svn
SVNPath /var/local/repos
AuthType Basic
AuthName "Subversion repository"
AuthUserFile /etc/passwd
Require valid-user



LoadModule authz_svn_module modules/mod_authz_svn.so



返回页首
阅览成员资料 (Profile) 发送私人留言 (PM)
hufey
斑竹


注册时间: 2005-12-20
帖子: 4


帖子发表于: 星期二 十二月 20, 2005 7:24 pm 发表主题: 可以了 引用并回复 编辑/删除帖子 删除这个主题
原来是看不同的文档看到头晕了
写的是/etc/passwd
实际添加用户的在/path/to/repos/passwd里
贴配置时多看了一眼,明白了
改回来就好了
不过/etc/passwd里也有我的用户,密码一样,
连不上是加密方式不一样吗?
=======================================
最后一个帖子已经是回家之后发的,居然是passwd文件写错了。
不管怎样,毕竟能用了。。。
发帖子之前还subscribe了svn user mailing-list,看来可以取消它了

Posted by hufey at 07:36 PM | Comments (0) | TrackBack

December 11, 2005

telnetd和sshd

在gentoo下
emerge xinetd
emerge netkit-telnetd
改/etc/xinet.d/telnetd中disable为no
再启动xinetd
/etc/init.d/xinetd start即可
不过这样下来只能本地登陆,其他电脑还是登不上来地
必须去掉/etc/xinetd.conf中only_from=localhost才行

sshd比较简单,只要emerge sshd后再/etc/init.d/sshd start即可
但是putty登陆是会自动退出地
把window 改成never close会看到
No supported authentication methods left to try!
google了一遍查到据说putty版本问题
可我升级到最新版本也还一样
最后 改了/etc/ssh/sshd_config中的PasswordAuthentication no为yes
才行
不过这样不会有安全性的问题吗???

Posted by hufey at 02:04 AM | Comments (0) | TrackBack

November 30, 2005

Firefox 1.5 新鲜出炉

很少见OOS社区这样期盼一个软件的release
29号N多人期待“今天下午Firefox 1.5就release了”
仿佛中国小孩盼过年一样
我本来也等了等菜单条的蓝色有新update的标志
等了两天还没有,只好手动下载安装
下载的时候只有一个thread被允许,估计也是临时性的限制吧
升级完发现20多个extension有十多个被disable了
因为不兼容于1.5
不过1.5extension 下边提供了一个find update的按钮
以前有么?我没注意过
check update后发现了大部分都有了兼容1.5的升级
现在还有6个不能用的,只有慢慢等了
LiveLines居然也没有,bloglines这就有点不像画了

另外,某些错误提示信息好看了

Posted by hufey at 09:25 PM | Comments (0) | TrackBack

November 20, 2005

有些人是不是真的有病?


而且数量高达9999人。

Posted by hufey at 09:05 AM | Comments (1) | TrackBack

November 13, 2005

vidalinux

在DistroWatch中叫VLOS。怪不得我没找到。
直接装gentoo stage 3的一个distribution。

New Gentoo-based desktop adds system update capabilities
Jul. 22, 2004

Based on Gentoo Linux, the Vidalinux Desktop OS has been announced on the project's website, and the first beta of the software is expected to be available on Gentoo mirrors soon. The goal of the Vidalinux Desktop OS is to make Linux easier to use and administer for daily work for home or office users.

Described as a "metadistribution", the software uses a technology called Portage that is said to allow the Gentoo-based desktop to become a secure server, development workstation, professional desktop, gaming system, embedded solution, among possible uses for the system.

This "near-unlimited" adaptability allows users to increase their desktop productivity by enabling updates and the installation of applications, the project claims. Portage technology keeps systems up-to-date with the most current application and security patches, compiling from source code or downloading precompiled binaries. Portage also resolves dependencies, according to the project's website.

A "Porthole" system (see image on right -- click to enlarge) provides automated software updates and patching, including desktop alerts/notifications and dependency checking/analysis, to ensure that systems are secure and are running the latest software.

Vidalinux Desktop OS includes a GNOME desktop environment, Ximian edition of OpenOffice.org, Mozilla browser, Evolution mail and calendar client, Gentoo Linux system, and portage package manager. The project says that the distribution includes additional multimedia and productivity applications for the home user, including media players, browser plugins for flash, real player, pdf viewer, media, graphics design, and administration tools.

The recommended requirements for running Vidalinux are a Pentium III compatible processor (500 MHz or faster), a 10 GB hard disk, and a minimum of 256MB RAM. Additionally, Vidalinux requires 1024x768 screen resolution or better.

Posted by hufey at 12:10 PM | Comments (0) | TrackBack

October 31, 2005

Poll: What Your Next Cellphone Will Be?

Poll: What Your Next Cellphone Will Be?
Written by Eugenia Loli-Queru on 2005-10-31 09:42:33 UTC
According to recent news, global shipments of smartphones are up 75% in Q3 2005, with PDA-only handheld shipments falling 18% and converged devices more than doubled in volume. So, what are the tech-inclined OSNews readers are opting for? Come in and vote.


Cellphones today usually are categorized in 5 ways. Here are some sample configurations for these categories (configurations that could vary in many ways of course and are given here only as a guideline):
1. Low-end phones. Few LCD colors and very low resolutions (e.g. 96x64), only SMS support, no camera, no java, 500KB to 4 MB internal storage. Example: Sony Ericsson J210i, Nokia 1221.
2. Low-end feature phones. CIF/VGA camera (no flashlight), 128x128 STN LCD or similar, Java, 5-15 MB internal storage. Example: Sony Ericsson K300i, Nokia 6235i.
3. Mid-range feature phones. VGA to 1 MP camera with flashlight, 128x160 or 176x220 TFTs, Java, EMS, 15-35 MBs internal storage, Bluetooth. Example: Sony Ericsson K700i, Motorola ROKR.
4. High-end feature phones. 2MP cameras, QVGA or 176x220 TFTs, Bluetooth, flash storage slot and 35-50 MBs internal storage, high multimedia inclination. Examples: Sony Ericsson K750i and W800i, Nokia 7370.
5. Smartphones. Any Windows, PalmOS, Symbian or Linux powered devices.

Note: This poll requires javascript to view and vote for.



My next cellphone purchase will be:
None, I do not need a cellphone
48 (6 %)
None, my carrier gives one to me for free
27 (3 %)
Just a plain & cheap low-end cellphone
98 (13 %)
A low-end feature phone
57 (7 %)
A mid-range feature phone
102 (13 %)
A high-end feature-phone
112 (14 %)
Windows Mobile smartphone
69 (9 %)
Symbian smartphone
84 (11 %)
PalmOS smartphone
38 (5 %)
Linux smartphone
114 (15 %)
Total votes: 749

Posted by hufey at 10:24 PM | Comments (0) | TrackBack

September 20, 2005

linux kernel 里coding style的讨论

kerneltrap

Linux: Reiser4 and the Mainline Kernel
Posted by Jeremy on Monday, September 19, 2005 - 06:38
Linux news

Hans Reiser [interview] sent an email to the lkml titled, "I request inclusion of reiser4 in the mainline kernel". He provided a list of objections raised earlier, noting that all had been addressed. Among the listed issues, Reiser4 now works with 4k stacks. "There have been no bug reports concerning the new code," Hans added.

The request was followed with some suggestions by Christoph Hellwig, including general comments about the coding style. This was one of many issues that led to debate in which Hans commented, "most of my customers remark that Namesys code is head and shoulders above the rest of the kernel code. So yes, it is different." Alan Cox [interview] replied that while the kernel coding style isn't his own style, he tries to follow it when working on the kernel, "one big reason we jump up and down so much about the coding style is that its the one thing that ensures someone else can maintain and fix code that the author has abandoned, doesn't have time to fix or that needs access to specific hardware the authors may not have." Much of the rest of the thread was less friendly, leaving the question of merging Reiser4 into the mainline kernel still up in the air.

Posted by hufey at 12:04 AM | Comments (1) | TrackBack