Development

ZHAO Xin (a colleague of mine) has pointed out that GNOME Simplified Chinese does not did a good work when displaying time. He then filed ports/72722 and ports/72723.

Adam Weinberger (adamw@, one of the major FreeBSD GNOME maintainers) has then pointed out that it might be better to contact the Simplified Chinese Translation team. In a private mail, I have explained that the root cause of the issue, however, is that our libc implementation does not implemented the GNU Extensions, say, the -_^0 extensions of strftime(3). Adam have then kindly approved that the patches should be committed, and we should have a long-run solution then finally remove them.

很多人管Debian叫「大便」，不过这里我还是宁愿称其为Debian。

今天ld说了Debian的包的一个比较有意思的维护模式，我之前也了解一些，那就是它的stable和unstable包。Stable和Unstable两种。

It’s available here, in my home directory at freefall: http://people.freebsd.org/~delphij/mfc.xml.

Please note that this is an (incomplete) MFC list for sys/ only! I baked it up manually so I don’t guarantee thet there is no mistakes, but I really tried my best to prevent so.

Scott Long, the Release Engineering Team Lead, has announced the “Final MFC Call” which means we must MFC all changes in the next 24 hours. A Release Candidate would be tagged soon (in 1 day) by creating RELENG_5_3 branch.

Accroding to the historical release engineering practice, we will release at least two release candadites. However, the release engineering team thought that we have seven (7) beta releases and we may reduce the need of RC’s to only one.

此文章版权归 李鑫 所有。版权声明如下：

版权所有 © 2004 李鑫，保留所有权利。转载必须完整地复制这些文字，不允许修改，并附上其原始连接：

我在许多场合公开地发表过类似的意见，有时言辞比较激烈，有时只是委婉地反对。

这种意见，同意也好，不同意也好，我希望其他人能够把它当作一种意见，而不是某种强加于人的意志。我只是说出自己的看法，并不要求任何人接受——如果他们认为这是对的，愿意接受，我当然很高兴；但如果他们觉得我说的有道理而不愿意接受，那是他们的问题；如果他们觉得我说的没道理，我很希望看到反驳的意见。我始终相信，真理越辩越明。

其实，最关键的问题就在于，我们没有必要重新发明轮子。在阐述这个观点之前，我希望大家能够接受一个密码学界的常识，即安全不能建立在别人不知道的基础上。另一个常识是，安全总是从最薄弱的环节被突破的，不要去尝试解决最强点上的问题——因为那不解决问题。

有人一天改了一万一千多行代码。

有人一天改了一万一千多行VFS代码。

有人一天改了一万一千多行FreeBSD VFS代码。

我靠……

Matthew Dillon总是给人惊奇，这是一个你没办法不佩服的人。也许他的一些话让有的人感觉不爽，也许他被一些人误解，但他是Matthew Dillon。

而不好的代码读起来如同吃进了一只苍蝇，而且苍蝇还在你的嘴里不停地蠕动……恶心死了。

我终于明白了老韩当时说的「代码复审很累的」那句话。

写在这样的一天，有些人的代码看起来是一种享受，能学到很多东西，但并不是所有的人的代码都这么友善。

It seems that sater@ have deployed the necessary resources for ASP.net forum on Frontfree’s internal network. I have not read the code but I wish this a success.

The script I am using for updating my own website(s) is not quite complex, it is here:

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14

#!/bin/sh
# Detect and update the files
#
# $Phantasm$

CVSROOT="/home/ncvs"
SUPFILE="/home/delphij/cvsup/supfile-delphijweb"
LOCKFILE="/home/delphij/cvsup/cvsup-lock"

if [ -f ${CVSROOT}/CVSROOT/commitlogs/have_updates ]
then
	rm ${CVSROOT}/CVSROOT/commitlogs/have_updates
	cvsup -1gL 2 ${SUPFILE}
fi

Of course you need some CVSROOT scripts to get it work, but the idea is not too complex.

Today FreeBSD will have its 5.3-RC1 if there is no other problem occours. I have sent a private mail to Doug Barton about a strange issue I have in BIND9, however, it does seem to be serious enough nor it is clear to determine the impact. I may post it later on a public list.

5.3-RC1 is also the point that I can suggest you to install 5-STABLE instead of 4-STABLE because most of the show-stopping issues are addressed and being corrected.