stillhq.com : Mikal, a geek from Canberra living in Silicon Valley

Customisation of visible posts

Sat, 05 Jul 2008 22:42:00 GMT

It occurred to me today that its not too hard to hide categories of posts on the HTML version of my site as well. The site has now been changed to have a link at the top of each post which offers to hide that category in the future. If that category is already hidden, then it offers instead to unhide it. The preference information is stored in a series of cookies, so is of course specific to a browser on a given machine. Because its all done in JavaScript, I have no logging of who is hiding what or anything like that. The cookies will expire after 30 days of non-use.

Posts are never completely hidden however -- you will still see the title. This seemed a reasonable compromise and left me with somewhere to display the link offering to unhide the content. It also gives people a very brief summary of what they are missing.

One wart is that when you change your hiding preferences the page is reloaded. This is because I had trouble finding a way to hide divs which worked on plenty of browsers and didn't require unique div ids for each post, which I would need to them iterate through for all posts in a given category. I'd love it if someone with higher JavaScript foo could tell me how to avoid the page reload, as I am sure it will be annoying for some.

So there you go.

Tags for this post: site() Comment

Site janitorial stuff

Thu, 03 Jul 2008 17:02:00 GMT

So, this whole kerfuffle about microblogging prompted me to hack the perl I use to generate this site to provide a blather-free version of my RSS feed. As a side effect I fixed a bug which was generating too many versions of the various RSS files I have on the site, thus removing 11,000 unneeded RSS files. That seemed like a lot to me (although not a lot of disk, just a lot of small annoying files).

There is of course the possibility that I removed too many, or broke something else in the process. Let me know if you notice anything.

Tags for this post: site() Comment

Historical revisionism

Fri, 04 Jan 2008 12:26:00 GMT

Hi. I am dropping the technorati and icerocket tags from all blog posts here. Both sites seem largely irrelevant now, and I'm not so fond of how these forms of tagging give those two sites lots of incoming links. This change will affect all posts both past and future, because of the way this site is generated. You should notice the tags go away over the next hour or so.

Tags for this post: site() Comment

I do delete comments, but only some

Sun, 08 May 2005 10:06:00 GMT

Jeremy says he deletes some blog comments (spam and so forth), while Scoble says he deletes none but has anti spam technology. I delete spam (all comments go to me for approval first), but that is all I remove. I haven't had to deal with hate speech yet, so I don't really know what I'll do about it if it happens.

Technorati tags for this post: site blog comment policy Comment

Sorry for futzing with the RSS feed so much recently

Sat, 02 Apr 2005 08:04:00 GMT

I want to get my RSS feed just right for some stuff I have lined up for a couple of weeks from now. To do that I needed RSS 2.0, enclosures, commenting, and more posts in my feed. All of that is done now and I think I'm pretty much ready to go... Sorry for any inconvenience.

Technorati tags for this post: site rss feed Comment

Comments turned on!

Mon, 28 Mar 2005 12:40:00 GMT

Well, it's more that I've written code to implement them now. After listening to Robert Scoble's couch, I've finally gotten around to writing the code to implement comments here. I wrote custom code because the site is generated statically to help handle the load, and it was easier this way.

Technorati tags for this post: site blog comments Comment

New site look and feel

Sat, 11 Sep 2004 09:34:00 GMT

So, welcome to the new site look and feel. I've transitioned to generating the entire site via Blosxom as a bit of an experiment. We'll see how we go. There are quite possibly broken links still -- I'm running a link validator as we speak which will help me fix those... Comment

Panda 0.5.4 released

Thu, 15 Apr 2004 12:04:00 GMT

(Evan has taken over maintaining Panda, although I will be dropping in from time to time to add code still)

Subject: [panda-development] Panda-0.5.4 Released
Date: Monday 12 April 2004 11:28 pm
From: Evan Nemerson 
To: panda-development@lists.sourceforge.net, panda-users@lists.sourceforge.net

Hi everyone,

I've just released Panda-0.5.4 via Savannah and SourceForge. I will make an
announcement via Freshmeat as soon as the files are available via stillhq.com

The release can currently be downloaded from:
	https://savannah.nongnu.org/projects/panda
	http://www.sourceforge.net/projects/panda

As you may already be aware, the most important change in 0.5.4 is that the
library is now available under the terms of either the GPL or LGPL (your
choice). In order to accomplish this, the TDB code was removed, and either
Enlightenment DB or Berkeley DB (version 4) must be linked to in its place.

Other changes include the addition of link annotations (internal and
external), pkg-config support, and an RPM. The main point of this release,
however, was the licensing issue.

The time has begun to begin thinking about the next release of Panda, and
though I do not have a time frame, I do have a few features I'd like to
implement:

*	Improved font handling, using FreeType. I would like to be able to embed
 and use arbitrary fonts, supporting line and page wrapping of text boxes.
 *	Support XMP metadata, noteably Creative Commons licensing data.
*	Bindings for different languages, such as PHP, Python, etc.
*	Comprehensive documentation, using DocBook XML
*	Stronger Windows support

If you have any other ideas you would like to see integrated within the
library, please send a message to panda-development@lists.sourceforge.net

The files for this release can be found at:

Comment

Announcing the first version of mcachefs

Wed, 17 Mar 2004 21:03:00 GMT

mcachefs is a simple caching filesystem for Linux using FUSE. It works by copying the file that you asked for when the file is opened, and then using that copy for all subsequent requests for the file. This is really a fairly naive approach to caching, and will be improved in the future. It's good enough for now though.

mcachefs needs two filesystems to operate. The target filesystem is the slow filesystem you want to cache accesses to, and the backing filesystem is where mcachefs can stash stuff which it has copied. The backing filesystem should therefore be on local disc, or the whole point of the exercise is gone.

It is also assumed that the target filesystem is read only for now.

Installing:

You'll need FUSE, so your best bet is to go up a directory level, and then type:

  ./configure
  make
  make install

(That last step needs to be done as root of course).

Using:

mcachefs looks for a config file to determine where the target and backing filesystems are located, as well as a debugging verbosity level. mcachefs looks in the following locations for it's config file (in this order):

/etc/mcachefs
~/.mcachefs
`pwd`/mcachefs.cfg

The format of the config file is something along the lines of:

/cache/backing  /backing
/cache/target   /
/cache/verbose  99

Note that the two columns must be separated by tabs for the config file to parse properly. This config section above says that we have a cached filesystem to mount at /cache. The backing store is /backing, the target store is /, and we want to be _very_ verbose. Lower numbers are less verbose. Please note that being very verbose might help with debugging mcachefs, but _will_ slow down the filesystem.

You can now mount the cached filesystem but doing a:

	mcachefs /cache

This program wont terminate... If you kill it, the filesystem will unmount. Don't do that. Use umount instead. I have making mcachefs daemonise on the todo list, so don't bother reporting it as a bug please.

Known bugs:

The only known bug at the moment is that if the program which is opening a large file terminateds before the cached copy has been generated, then a bus error occurs in mcachefs. I'm looking into that now.

So, I came up with this great patch:

Email me at mikal@stillhq.com

New code?

You can find the latest version of the code at http://www.stillhq.com/extracted/fuse, and http://www.stillhq.com/extracted/fuse.tgz Comment

aus-dotnet list archives

Wed, 17 Mar 2004 07:03:00 GMT

So, there's an Australian list for dotNet developers, but I couldn't find an archive of it. So I made one. The archive isn't updated in real time at the moment, but it's good enough for searching and stuff like that. Comment

Read about my Christmas lights

Sun, 07 Dec 2003 22:00:00 GMT

Which this year include a Tux made out of fairy lights. Comment

Panda is now available under the GPL and the LGPL

Mon, 01 Sep 2003 00:00:00 GMT

The next release of Panda, which is imminent, is released under both the GPL, and the LGPL, which will make it much easier for some users to use the code... Comment

Mandocs for linux 2.6.0-test3-bk1

Thu, 14 Aug 2003 00:00:00 GMT

I have submitted a patch which changes the man page output for the mandocs target in the kernel. Sample output is here. Comment

Mandocs included in linux kernel 2.6.0-test3

Mon, 11 Aug 2003 00:00:00 GMT

The mandocs patch from myself has been included in kernel 2.6.0-test3. Sample man pages are here. Comment

Command line image procesing article Slashdotted

Sun, 20 Jul 2003 00:00:00 GMT

Graphics Tricks from the Command Line

Posted by michael on Thursday July 17, @11:36AM
from the thumbnails-galore dept.
An anonymous reader writes "There's nothing quite like command-line tools for handling large batches of tasks, and image manipulations are no exception. Web developers and administrators will appreciate the ability to handle large numbers of files easily, either at the command line or in scripts. This article presents the ImageMagick suite, a Linux toolkit for sizing, rotating, converting, and otherwise manipulating images, in a huge number of formats, whether one or a hundred at a time."

And before you ask, no, I'm not the anonymous reader... Comment

usblogdump Released

Sat, 28 Jun 2003 00:00:00 GMT

In the words of the annoucement email:

To:  linux-usb-devel@lists.sourceforge.net
cc:  linux-kernel@vger.kernel.org
Subject: [Announce] Linux command line Snoopy Pro logfile dumper

I had two maths exams last week. This of course means that I had to find 
something to distract me. That thing was whipping up a SnoopyPro logfile 
dumper for the command line. This was motivated by generalised frustration 
with the SnoopyPro user interface.

For those wondering, SnoopyPro is a Source Force hosted USB traffic dumper
for Windows. It's useful when reverse engineering USB device drivers.

This version of the dumper only implements the URB types which I 
immediately needed. Adding additional URBs isn't hard, but I didn't have 
any samples. Feel free to mail me usblogs, and I'll add them to the 
decoder.

The only really cool feature in this version is that it implements 
"repeated URB sequence suppression", so if the Windows driver says to the 
USB device "hey, you still there" every second for 60 seconds, and there 
is no other traffic between the machine and that device, then the output 
will only show one of those interactions, and let you know it hid 59 more. 
This feature can be turned on and off with the -r command line option.

You can get the GPL'ed CVS version of the source code from:
http://www.stillhq.com/extracted/usblogdump.tgz

There is sample output et cetera at:
http://www.stillhq.com/cgi-bin/getpage?area=usblogdump

The next step is to modify the display of the URBs so that they're closer 
to the Linux data structures.

Cheers,
Mikal

-- 

Michael Still (mikal@stillhq.com) | Stage 1: Steal underpants
http://www.stillhq.com            | Stage 2: ????
UTC + 10                          | Stage 3: Profit

Comment

PandaScript 0.5 Released

Tue, 10 Jun 2003 00:00:00 GMT

PandaScript 0.5 has been released. PandaScript 0.4 was skipped to keep the version numbers inline with Panda. The cool bit about this release is that PandaScript is now automatically generated by Objectify, which means it should always be up to date with the latest Panda version. You can download the latest source from here and a signature from here. Comment

Objectify 0.3

Tue, 10 Jun 2003 00:00:00 GMT

Objectify 0.3 can now generate command line interfaces based on a flex lexer, and a yacc grammar. Sorta cool. You can grab the sourcecode from here and a signature from here. Comment

Panda 0.5.3 Release

Thu, 29 May 2003 00:00:00 GMT

This version has heaps of bug fixes, as well as a compilation fix.

Source (with examples) (signed)
Source (without examples) (signed)
Source (with MS Windows stuff) (signed) Quite big
Comment

mandocs and mandocs_install

Thu, 29 May 2003 00:00:00 GMT

This is the announcement email I sent to LKML:

Hey,

this patch adds two new targets to the 2.5 makefiles -- mandocs, and 
mandocs_install. The targets require two new perl scripts in the scripts/ 
directory, but in return we get a series of man pages for kernel 
functions, which are installed in man section 9. This is a good thing, as 
many programmers expect documentation to be available with man, and 
hunting through various PS or PDF documents to find the documentation for 
the function you want can be quite frustrating.

The man pages are just extracted from the various existing DocBook SGML 
documents, which are generated by kernel-doc. You also need to have 
docbook2man installed on your machine.

Note that the two scripts (makeman and split-man) need executable 
permissions, and I couldn't work out how to make this happen in a mere 
diff. Also please note the formatting is not perfect, but I will tweak 
other stuff later with further patches -- this is just an initial 
implementation.

Sample output (HTMLised) can be found at 
http://www.stillhq.com/linux/mandocs/2.5.70/

Linus, please consider applying.

Thanks,
Mikal

The patch can be found here, and the sample output is here. Comment