Etherplex

I started using Linux back in 1997 when a buddy and I put Slackware, floppy by floppy, onto a Pentium 133.  I didn’t start using it full time until a couple years later, when Microsoft started requiring online registration with product keys.  Since then, I’ve used Red Hat, Gentoo, Debian, Arch and Ubuntu.  I abandoned Red Hat because of “dependency hell”, which the other distributions avoided entirely.  So, it’s been about 10 years, and as I’ve tracked Red Hat’s dismissal, and subsequent rewarming to, the desktop market, I have noticed that Red Hat, Fedora Core, and finally Fedora have learned from Ubuntu and made a better experience for the casual user.

So I installed the just-released Constantine to see how it measured up to my beloved Debian-derivatives I’ve been using the past few years.  I installed it on a netbook that I love: the Asus EEE 1002HA.  The verdict?  Yeah, it just works.  It has a smooth boot up, detected screen resolution well, has good sound support, allowed me to customize tap-to-click and scrolling on the touchpad, and allowed me to set up dvorak and rebind caps-lock to control simply and easily.  What didn’t work so well?  The Atheros AR928x chipset, which is why I’m posting this.

Short version: the Atheros AR928x is supported by the open-source code contributed by Atheros to MadWifi under the driver name “ath9k”, which is the kernel module Fedora 12 loads to support this hardware.  Unfortunately, the version included is a bit behind the times, and causes odd intermittent connectivity issues when associating with access points, which effectively makes the wireless on the machine next to useless.  The good news is that it is easy to fix.  Here we go:

1. Update.  Get root on the command line and type ‘yum update‘.
2. Install your development tools so you can compile things.  As root: ‘yum groupinstall “Development Tools” “Legacy Software Development”‘.
3. Head over to the download site for compat-wireless and download the latest version of the ath9k driver.  Uncompress it.  Navigate to the directory it is in on the command line.
4. Select the ath9k driver: ‘./scripts/driver-select ath9k‘
5. Build: ‘make‘
6. Install: ‘make install‘
7. Unload your current ath9k driver: ‘make unload‘
8. Load the new driver: ‘modprobe ath9k‘ (or simply reboot)

There were lots of discussions across the web about various hacks and methods to get ath9k to play nicely with the AR928x chipset, but this seemed to be the most straightforward approach to me, and it seemed to work really well, so I thought I’d post it here in hopes it would be useful to someone.  Happy hacking!

Though perputually on the verge of total disorganization in the physical realm, I try to keep my digital world very organized.  Because of this, I habitually try to move more and more of my life into a digital representation, which allows easy backup, searching and reorganization.  Although I have tried many free and proprietary software products to act as a central hub for storing and retrieving random bits of useful information, even the very best (Emacs Org Mode) has failed to gain traction in my life on a long term basis because it is not immediately accessible to me when its underlying framework is not present (Emacs).  Of course, Org Mode files are just text, but much of the file’s appeal is lost in a pure text mode: Emacs is required to really wrangle all the data present in the text file.  I have used other tools, like DevonThink on OS X, Google Notebook, Google Calendar, iCal, various GTD systems, several Emacs-based packages, and Evernote.  All either store my data remotely, aren’t workable on Windows, OS X and GNU/Linux, or require a sizeable framework to be present whereever I want to view/alter my data.  Each of these shortcomings causes me to shy away from really “investing” in the system.

In the past few months, however, I have rediscovered TiddlyWiki, which, despite its name, is a very serious piece of software.  Though written only in Javascript and HTML (to run in a browser), it is a serious software engineering effort, and has a large community doing active development of many different variants.  The purpose of this post is to provide a look at what TiddlyWiki has to offer, so as to assist potential new users (you!) in determining if it might be useful to organize information in their lives.

TiddlyWiki is a Wiki

This means in is accessed in a web browser that you already have on your computer.  It works best in Firefox, but can be used in Safari, Internet Explorer, and Opera as well.  TiddlyWiki works an whatever operating sytem you have.

This means it has “pages”, as seen in other wikis (like Wikipedia), except instead of entire webpages, TiddlyWiki uses Tiddlers: small (or not-so-small) chunks of formatted text about a particular topic.

This also means it uses hypertext, so you get both internal (links between tiddlers) and external (links to other websites) linking, as well as formatted text: bold, underline, italic, monospace fonts, numbered and bulleted lists.  The linking feature is critical, because it allows you to link tiddlers to each other in the same way the internet links webpages.  You can use this mechanism to organize your data as you see fit – whether that be in a hierarchy (like the files on your computer) or in some web that more closely resembles Wikipedia.  You are in complete control.

TiddlyWiki is Just a File

This means it resides on your hard drive or USB stick just like any other file.  You can copy it, give it to others, post it online, or carry it around with you just like any other file.  To open it, you just double cilck on it and it will open in your browser.

TiddlyWiki Supports Metadata

All I mean by this is that it allows you to tag tiddlers (pages) with their topic.  If you’re a programmer, you might tag tiddlers with the language they relate to.  If you’re a journalist, you can tag a tiddler with the article it relates to, or the source it came from.  Tiddlers can have as many tags as you like.

TiddlyWiki Can Search

As you grow the content in your TiddlyWiki (by adding new Tiddlers), you may start to worry that while it is really easy to add information to the TiddlyWiki, retrieving it may be difficult.  Where is that darn tiddler that contained the dentist’s address and office hours?  Even if you can’t remember what you called it (was it under “Dentist” or “Medical”?) so long as you used the word “dentist” anywhere in the tiddler, you can simply type “dentist” into the global search box, and, as you type, TiddlyWiki will search every word you have stored in it and show you all the tiddlers that contain that keyword.  If all you can remember is that you had the words “office hours” in the tiddler, just type that and you’ll track your data down in no time.

You can try this on the TiddlyWiki homepage (which is itself a TiddlyWiki!).  Upon visiting the page, you have no idea if there is any information on whether TiddlyWiki works on the browser available for the Nintendo Wii.  Simply type in the keyword “wii” in the search box and you can find out if there is any information on the Wii anywhere in the TiddlyWiki.

TiddlyWiki is Organized

In addition to full text search of all your data, TiddlyWiki can also list all the tags you’ve ever created and allow you to view lists of all tiddlers that match those tags.  You can then open the tiddlers you want from that list, or open all tiddlers that match a given tag.  At home, I work on schoolwork for my current class (csc576 – Connection Oriented Neworks), so to bring up my notes for that class, I just go to the tag I made called “csc576″ and request all tiddlers matching that tag be opened.  Similarly, at work, if I’m working on a project relating to JBoss, I can open all my notes on JBoss by selecting the tag I created called “jboss”.

In addition to tagging, TiddlyWiki offers a series of tiddler lists that are organized in various useful ways.  In addition to the tag list view (mentioned above), TiddlyWiki also offers a timeline of tiddler edits, allowing you to see what has been changing in your TiddlyWiki at a glance.  Can’t remember the name of the tiddler you were editting on Monday?  Just look at the timeline and it will tell you what changed that day.  In addition to the timeline, you may also find that you create tiddlers that are not referenced from anywhere else.  They may contain useful information (the VIN number on your old 1991 Toyota Corolla), but you won’t find them just by “surfing” through the TiddlyWiki.  Well, luckily, TiddlyWiki keeps track of all these “orphaned” tiddlers and can give you a list, so that you won’t lose them.  Conversely, sometimes a tiddler has a link to it, but it doesn’t yet exist.  TiddlyWiki also tracks these cases, and provides a list if needed.

Conclusion

Overall, I have found that TiddlyWiki has provided a sort of “sweet spot” in the field of tools to organize your data.  It is extremely easy to pick up and start working with, but has an extensive feature set that gives advanced users  a lot to work with.  It is open source and extensible, allowing sties that collect extensions for TiddlyWiki to spring up all over the web.  If you’d like to keep your own TiddlyWiki on your local machine or USB stick to carry with you, you can head over to the official site and download it.  If you’d prefer to keep your data online so you don’t have to carry it with you, a site called TiddlySpot specializes in providing free online TiddlyWikis (public or private!).  Just supply a site ID and password and you’re on your way.

In the last post, I use the excellent SuperDuper! tool to make a
bootable clone of my hard drive. There are some (more advanced)
techniques that can be used that allow bootable clones to be created
on the command line using free tools. I had some trouble getting
clones to work on the command line, but finally worked through the
problems using the technique described below. Most all of this
information is in Bombich Software’s guide.

• Use Disk Utility to set the drive to use a GUID partition table

Macs can boot without this being the case, but since the internal disk
uses this kind of a partition table, it’s good to have your clone use
the same setup. I’m pretty sure this isn’t necessary, but it worked for
me as I was trying to solve some issues with the clones working.

• Ensure the volume’s “Ignore Ownership on this volume” is *not* checked

By default, external drives will set this option when OS X formats
them. This will ruin the clone, as the system will not boot if files
are set to a UID of 99. So, immediately after formatting your drive,
right click on it and open the “Get Info” option. At the bottom of
that panel, *uncheck* the option that says “Ignore Ownership on this
volume”. This will allow the copy you are going to make to boot.

• Clone with a command like asr, ditto, rsync, etc.

I used ditto successfully:

sudo ditto -X / /Volumes/Clone

For incremental backups, rsync can be used as is described on Bombich Software’s guide to backups:

sudo rsync -xrlptgoEv –progress –delete / /Volumes/Clone

• When booting, get diagnostic information with Command-V

During the first boot from the clone, you can hold down Command-V as
the system boots to get more information about the boot process. If
something goes wrong, at least you’ll have some error messages you can
use as a basis for a search on the internet.

The Macbook I bought back in May of 2007 has been getting a bit long in the tooth. Shortly after I bought it, I upgraded the RAM to 2 GB, so the memory and the processor have been holding up just fine for me, as I mainly program and use VMWare (not much gaming). But, after a while, the 70 GB hard drive starts to feel a bit cramped, as the Macbook is my main machine. I’ve mitigated the problem by buying various external drives, but nothing can replace a nicely sized internal disk.

So it was time for a disk upgrade. It turns out that upgrading the disk with command line tools (as was my original intention) like asr, rsync, or ditto copies the files intact, but won’t make the disk bootable unless the root file system is mounted read-only. I can get my Mac into single-user mode, but, coming from Linux, I’m not sure how to remount the root filesystem read-only on the BSD-based OS X. Plus, I can’t use my screen capture utility in single-user mode.

So, I ended up researching other tools, like CCC and SuperDuper! SuperDuper! seemed to fit the bill, so that is what I made the video with, but I’d be very interested to see what people’s experiences have been with CCC and/or command line tools making bootable clones for use in an upgrade. Anyway, the video is now posted on YouTube:

Macbook Hard Disk Clone & Upgrade