Planet Larry

Something is very wrong here. Right after starting a KDE session everything looks normal.

But after running for a day we have a different story. I’m assuming this isn’t the expected behavior (if so I didn’t expect it).

This time I specifically took a screenshot to prove it, but I’ve seen it eat up as much as 1.3gb of my memory, which is rather unnerving.

kwin-kde4        4:4.0.4-0ubuntu1

Bug report.

I was in a project-y mood today and since work was going good all day, I started some work on a little pet-project called "Thinkwhack".

It's somewhat inspired by smack.py but I wanna make it somewhat more flexible and useful.

What it will do is to allow you to have certain action triggered when "whacking" the laptop (if your laptop has a hdaps module which only Thinkpads have I think). Right now it's pretty simple and just plays a sound but it already has a proper systray icon so things should be easy to add there.

I play to give the user the option to play sounds or trigger other options when whacking the machine (maybe tilting also): Hide all the windows on your desktop for example, switch workspaces, run a random script.

The next step will be to get a proper motion detection algorithm in there cause right now it's not that smart, then some proper configuration.

Here is some video footage:


It's not really all that useful I admit but still might turn out to be useful later

What this definitely will spawn are a few python howtos so it's not in vain.

For the last two children, I've been relatively relaxed with the rest of my life, so I was eagerly awaiting and up-to-date with the happenings of our pregnancy.

This time around, I'm quite busy otherwise, and this ultrasound crept up on me.

I'm also quite amazed at the technology which wraps around the medical field.  At the last portion of the video, you can see our little boy's profile quite clearly.  It's amazing.

Anyway...  I was really excited to hear the heartbeat and see our little guy this time, as I wasn't really thinking much about it until it happened.  Kind of a spiritual occasion.

The video was quite big this time, and I haven't got around to editing videos yet, so the above is the complete ultrasound.

Thanks for watching!

Emacs has a, well, "unique" undo system. It only has undo, no redo. When you undo something, the act of undoing is added as itself onto a stack of undo actions. When you've un-done enough things, you do "something, like move the cursor, and that breaks the chain. From there if you undo again, you will traverse back over the undo actions you just did.

This is supposedly powerful. It does help with the following situation:

1. Type something.
2. Type something #2.
3. Type something #3.
4. Undo undo undo.
5. Type something #4.

In most programs once you reach step 4, you can redo to get back to the text you just undid But once you reach step 5, the first three things you typed are lost forever. You've gone back in time and changed history, eradicating the old future and replacing it with a new one. You can never get back to the old future. Emacs undo, on the other hand, where undo actions are just like any other actions and pushed onto a stack of actions, does let you undo back first three things you typed.

However in practice this doesn't work so well. This site has a nice quote:

“By [undoing] repeatedly, you can gradually work your way back to a point before your mistake. This is convenient if you’ve made a mistake four or five commands back. It is marginally useful if you’ve made a mistake twenty or thirty characters back. And it is completely useless if your mistake is ancient history.” - Learning GNU Emacs (page 42)

The problem being, supposing you undo 20 times, and break the chain (by moving the cursor for example), if you then decide to undo one step FURTHER back, you have to undo all 20 of your previous undos, undo 20 more times, then undo once more. Eventually you end up feeling like you're going up and down a roller coaster of undos.

If you hate this, which you probably do, you could use redo mode, which gimps up Emacs undo/redo to be like any other program's, i.e. you get the same behavior as Microsoft Notepad. (Although when I tried it, it was buggy as heck, failing to undo my actions properly, mangling text from different lines together and whatnot.)

Vim's undo system on the other hand is far better and equally powerful. You have a standard undo / redo option via u and CTRL-R. You also have a second completely different way to undo: you can "go back in time". In the above example, Vim will create two undo "branches" and you can jump from one to the other even if you undo and "break the chain" by typing something new.

Doing :undol lists the branches, in a somewhat confusing format. But you can just pound g- and g+ to go to older / newer text states, or use :earlier with a human-readable time (say, 10s or 5m) and it will take you to that point. These will get you all the power of Emacs' undo stack, with none of the pain or confusion. See also :h undo-two-ways.

This is one of many instances where Vim wins, hands-down. Vim's undo system isn't as reprogrammable as Emacs, but it's so powerful and so perfectly what you'd want that it doesn't matter. This is beautifully typical of Vim. I don't have a year to figure out all the nooks and crannies and edge cases and idiosyncrasies of Emacs undo system, let alone the time it'd take to write a custom, crusty elisp script to buggily re-implement it.

Being an "extensible text editor" doesn't help much when such basic functionality is so broken. Unless you want to play your undo history back like a movie, in rainbow colors, which I don't. I want undo/redo that works.

I have a friend and fellow member of the Python West Midlands group. Whenever, someone mentions Django, he asks the person "but is it stable?". This has been repeated so much that is has become a local in-joke. However, lets take the question seriously.

To explore this further, we need to ask what does stable mean? I.e. can we replace the word "stable" with something else to provide some more meaningful questions:

• Can Django handle traffic loads?
• Is Django actively maintained, i.e. are bugs being fixed?
• Will the Django API evolve in the most backwards compatible way possible?

Lets take these one at a time.

Traffic loads

Django's frequently-asked-questions says:

Is Django stable?

Yes. World Online has been using Django for more than three years. Sites built on Django have weathered traffic spikes of over one million hits an hour and a number of Slashdottings. Yes, it's quite stable.

The first sentence is a testimony, useful but not a direct answer. In the second sentence, 'stable' is used as in 'strong table', i.e. Django can handle a heavy load, (i.e traffic rather than physical objects).

It goes on to explain that Django has a "shared-nothing" approach, i.e. you can throw more servers directly at whatever bottleneck you have. If the database is the database, then you can add more hardware to the databases, if it is images the are the problem, you can add more hardware to the media servers, and so on.

Is Django maintained?

The next question is whether Django is actively maintained. One simplistic measure is to look at the bug database and see what is going on. In what follows I use "ticket" in the broadest sense, i.e. not just a confirmed code error, but also enhancement requests, invalid bugs and so on.

At time of writing, Django has 1092 open tickets, out of which, 311 are new and unreviewed, I would guess that half of these are valid problems, and half are not.

Meaning the other 781 open tickets have reviewed by someone at least once. Some have been triaged and are waiting to be worked on, some are already being worked on, some will have been closed wrongly and reopened.

In the last three years, 6047 Django tickets have been closed, we can break these down further:

Number	Closed As
3348	"fixed"
867	"invalid"
809	"duplicates"
797	"wontfix"
226	"worksforme"

A ticket being fixed is very useful, but a ticket being found to be not a problem or not Django's problem is still useful information, marginally useful perhaps, but still useful if you have that issue.

So Django is three years old, which is pretty new. In that time, they have closed 85% of tickets, while 10.6% of tickets are open but have been read, 4.4% are open and have never been read.

This seems competitive with similar open-source web frameworks:

Framework	Years	Tickets	% closed
Django	3	7139	82
Pylons	2	444	90
Symphony	3	3612	82
Turbogears	3	1840	85
Zope2	7	2352	81
Zope3	6	886	76

The 'years' column represents how far the ticket tracker data goes back for, not necessarily the age of the project. There may also be sampling errors caused by differences in the ticket tracker software, or a project might have had a clear-out of closed tickets.

On the whole, it is all impressive stuff, all six open-source projects seem to be on top of things. One nice thing about Pylons is that, at the time of writing, it appears that all of the open tickets have been reviewed to some extent. Some other projects such as Django would benefit from more a few more triagers to review new tickets.

API

This is perhaps what my friend meant, i.e. if one makes a web application using Django in 2008, how much pain will it be for it to still work in 2013?

There is of course a balance to be struck between backwards compatibility on the one hand, and keeping the framework modern enough for meet the needs of today's web applications. It is a difficult balance.

As Joel Spolsky points out in How-Microsoft-Lost-the-API-War, if you change your API such that users of it have to learn new things or change software in any significant way, the risk is that the users of the API will instead decide this moment to jump to the next big thing.

On the other side, changing too slow is also a risk. Apart from people like me when I am feeling retro, does any one really create web applications of any complexity using only the CGI support in the Python standard library? The API has not changed much in the last half-decade which is good, but the standard Python CGI is lacking in any pre-built features so you have to do all the repetitive boilerplate yourself. Also CGI generally has no persistence or caching meaning that it is far more processor intensive.

I don't want to be mean, but all the cool things that people are actually using to make web applications in 2008, such as SQLAlchemy, Django and Pylons are outside the standard library and have very little chance of merging with the standard library. I am not going to mention Python Server Pages, (oh I did - d'oh).

So the balance between backwards compatibility and providing modern features that people want is the life-or-death question for a larger toolkit such as a web framework.

Django provide their answer in a webpage called API-stability, in this document, they explain what seems a reasonable compromise between backwards compatibility, and changes that are required to keep the framework both modern and secure.

Of course, the document is mere policy/propaganda and the real issue is whether they stick to it. However, Django is free software/open source and it is being used by a large number of people. If the Django developers change the API unnecessarily, people will vote with their feet and not upgrade and/or fork the API.

Django Development

So is Django stable? Well given what we looked at above, we can at least say that it does not seem any less stable than any other similar toolkit.

The philosophy of a traditional software application, often goes like this: there is a cycle, starting with the development window, followed by a code freeze, followed by stabilisation, followed by a release, followed by a new cycle.

The current fashion in the Django community is to develop sites using the SVN version of Django. This philosophy is that the mainline branch should be usable at all times, and the latest SVN revision is likely to be less buggy and more secure than previous revisions.

This approach is somewhat all or nothing; you decide yourself to become a web developer using Django and so commit yourself with keeping up with the state-of-the-art. Perhaps this is the only way for a toolkit of such complexity. This investment is perhaps less of an issue if Django becomes the mainstream way to make a web application in Python.

The Django developers do however plan on an eventual 1.0 release, their plans are nicely summarised on a wiki page called VersionOneFeatures. At Version 1.0, the toolkit will (in theory) have a fixed point. After that they plan to do the Pythonic thing of not breaking the API in one step. So if they want to introduce an incompatible change that requires code changes to the user's web application, they will provide a warning in the next release, and then make the incompatible change in the release after.

However, I imagine most people will continue tracking SVN. Since a person needs a certain level of technical skills to use Django at all, then they probably have enough skill to handle some minor API changes.

Stability is a process. Complex software is never finished, it just runs of money or becomes obsolete.

Your turn. What do you think? I would love to know your opinions?

Discuss this post - Leave a comment

Say you're at a party and the only internet is a wireless router and for whatever reason, no one can plug into it. Wireless internet only. You with your fancy laptop are sitting pretty and this is just fine. They your friend shows up lugging his clunky old desktop that only has an ethernet port for internet connectivity. Is he out of luck? Turns out not, because you an come to the rescue!

It's really easy, especially with Ubuntu.

To start with, you need to be a router, so you need firewall software that can do NAT (network address translation). This is part kernel side (NAT and iptables options enabled and modules loaded) and part user space side, in the form of the program 'iptables', so make sure it is installed, which it is by default on Ubuntu (I think).

Now all you really need to do is add two routing rules, one says anything coming on the ethernet port should go through the NAT procedure, which basically means it's IP headers are tweaked to make it look like they originated from your computer and then you send them along to the internet. The second rule helps facilitate this (I'm a little less sure what it does, but it's needed).

In this case we are assuming the interface eth0 is the wired network and eth1 is the wireless. Change as required.

iptables --table nat --append POSTROUTING --out-interface eth1 -j MASQUERADE
iptables --append FORWARD --in-interface eth0 -j ACCEPT

Next you just need to tell the kernel port forwarding should be turned on, and you can do this through the wonderful /proc filesystem.

echo 1  > /proc/sys/net/ipv4/ip_forward

Now you're pretty much ready to go. Connect to the wireless, presumably though NetworkManager or your wireless toll of choice. Then enable the wired network manually.

ifconfig eth0 up 192.168.1.1

This turns on eth0 with a local network address of 192.168.1.1. Now plug your friend's computer into the ehternet port with a crossover ethernet cable or into a hub and then the hub into you with regular ethernet cable and have them manually pick an address on 192.168.1.* (or whatever local network you chose, it doesn't matter) and set you (192.168.1.1) as the gateway router.

If this is a little much for them or they are running an OS that makes this non trivial, than its really another easy step for you to set up a DHCP server and do all the configuration for them :).

So install dhcpd, on Ubuntu it's 'dhcp3-server', on Gentoo I think it's just 'dhcpd'.

Now we have to configure it. We have to tell it the gateway router, which is us (192.168.1.1), the nameservers (the servers in /etc/resolv.conf) and the pool of IPs to use and what interface/network to listen on.

Open the config file, on Ubuntu '/etc/dhcp3dhcpd.conf'.

The relevant parts are as follows

...
# servers in /etc/resolv.conf
option domain-name-servers 192.168.0.1;

...

# the local network you created
subnet 192.168.1.0 netmask 255.255.255.0 {
        #IPs free to assign
        range 192.168.1.100 192.168.1.200;
        #your computer, the router
        option routers 192.168.1.1;
}

And that's it. (Re)Start the server

/etc/init.d/dhcp3-server restart

And you are now serving all the information your friend's computer will need to automatically connect properly.

They should now be online once they restart their internet connection.

One annoying thing about Ubuntu vs Gentoo is that on Ubuntu, the init system is a bit more kludgy and old fashioned. Any server software installed is automatically configured to start at boot time, forever, which in this case isn't what you want. You only want the dhcpd server to run very rarely, at parties, the rest of the time it's a waste. So we need to turn it's auto starting off. Apparently the /ubuntu init system barely supports this, we have to force it.

update-rc.d -f dhcp3-server remove

Now just turn it on when you need with its init.d file.

I have a friend and fellow member of the Python West Midlands group. Whenever, someone mentions Django, he asks the person "but is it stable?". This has been repeated so much that is has become a local in-joke. However, lets take the question seriously.

To explore this further, we need to ask what does stable mean? I.e. can we replace the word "stable" with something else to provide some more meaningful questions:

• Can Django handle traffic loads?
• Is Django actively maintained, i.e. are bugs being fixed?
• Will the Django API evolve in the most backwards compatible way possible?

Lets take these one at a time.

Traffic loads

Django's frequently-asked-questions says:

Is Django stable?

Yes. World Online has been using Django for more than three years. Sites built on Django have weathered traffic spikes of over one million hits an hour and a number of Slashdottings. Yes, it's quite stable.

The first sentence is a testimony, useful but not a direct answer. In the second sentence, 'stable' is used as in 'strong table', i.e. Django can handle a heavy load, (i.e traffic rather than physical objects).

It goes on to explain that Django has a "shared-nothing" approach, i.e. you can throw more servers directly at whatever bottleneck you have. If the database is the database, then you can add more hardware to the databases, if it is images the are the problem, you can add more hardware to the media servers, and so on.

Is Django maintained?

The next question is whether Django is actively maintained. One simplistic measure is to look at the bug database and see what is going on. In what follows I use "ticket" in the broadest sense, i.e. not just a confirmed code error, but also enhancement requests, invalid bugs and so on.

At time of writing, Django has 1092 open tickets, out of which, 311 are new and unreviewed, I would guess that half of these are valid problems, and half are not.

Meaning the other 781 open tickets have reviewed by someone at least once. Some have been triaged and are waiting to be worked on, some are already being worked on, some will have been closed wrongly and reopened.

In the last three years, 6047 Django tickets have been closed, we can break these down further:

Number	Closed As
3348	"fixed"
867	"invalid"
809	"duplicates"
797	"wontfix"
226	"worksforme"

A ticket being fixed is very useful, but a ticket being found to be not a problem or not Django's problem is still useful information, marginally useful perhaps, but still useful if you have that issue.

So Django is three years old, which is pretty new. In that time, they have closed 85% of tickets, while 10.6% of tickets are open but have been read, 4.4% are open and have never been read.

This seems competitive with similar open-source web frameworks:

Framework	Years	Tickets	% closed
Django	3	7139	82
Pylons	2	444	90
Symphony	3	3612	82
Turbogears	3	1840	85
Zope2	7	2352	81
Zope3	6	886	76

The 'years' column represents how far the ticket tracker data goes back for, not necessarily the age of the project. There may also be sampling errors caused by differences in the ticket tracker software, or a project might have had a clear-out of closed tickets.

On the whole, it is all impressive stuff. One nice thing about Pylons is that is seems that at the time of writing, it appears that all of the open tickets have been reviewed to some extent. Some other projects such as Django would benefit from more a few more triagers.

API

This is perhaps what my friend meant, i.e. if one makes a web application using Django in 2008, how much pain will it be for it to still work in 2013?

There is of course a balance to be struck between backwards compatibility on the one hand, and keeping the framework modern enough for meet the needs of today's web applications. It is a difficult balance.

As Joel Spolsky points out in How-Microsoft-Lost-the-API-War, if you change your API such that users of it have to learn new things or change software in any significant way, the risk is that the users of the API will instead decide this moment to jump to the next big thing.

On the other side, changing too slow is also a risk. Apart from people like me when I am feeling retro, does any one really create web applications of any complexity using only the CGI support in the Python standard library? The API has not changed much in the last half-decade which is good, but the standard Python CGI is lacking in any pre-built features so you have to do all the repetitive boilerplate yourself. Also CGI generally has no persistence or caching meaning that it is far more processor intensive.

I don't want to be mean, but all the cool things that people are actually using to make web applications in 2008, such as SQLAlchemy, Django and Pylons are outside the standard library and have very little chance of merging with the standard library. I am not going to mention Python Server Pages, (oh I did - d'oh).

So the balance between backwards compatibility and providing modern features that people want is the life-or-death question for a larger toolkit such as a web framework.

Django do themselves provide their answer in a webpage called API-stability, in this document, they explain what seems a reasonable compromise between backwards compatibility, and changes that are required to keep the framework both modern and secure.

Of course, the document is mere policy/propaganda and the real issue is whether they stick to it. However, Django is free software/open source and it is being used by a large number of people. If the Django developers change the API unnecessarily, people will vote with their feet and not upgrade and/or fork the API.

Django Development

So is Django stable? Well given what we looked at above, we can at least say that it does not seem any less stable than any other similar toolkit.

The philosophy of a traditional software application, often goes like this: there is a cycle, starting with the development window, followed by a code freeze, followed by stabilisation, followed by a release, followed by a new cycle.

The current fashion in the Django community is to develop sites using the SVN version of Django. This philosophy is that the mainline branch should be usable at all times, and the latest branch is likely to be less buggy and more secure than previous revisions.

This approach is somewhat all or nothing; you decide yourself to become a web developer using Django and so commit yourself with keeping up with the state-of-the-art. Perhaps this is the only way for a toolkit of such complexity. This investment is perhaps less of an issue if Django becomes the mainstream way to make a web application in Python.

The Django developers do however plan on an eventual 1.0 release, their plans are nicely summarised on a wiki page called VersionOneFeatures. At Version 1.0, the toolkit will (in theory) have a fixed point. After that they plan to do the Pythonic thing of not breaking the API in one step. So if they want to introduce an incompatible change that requires code changes to the user's web application, they will provide a warning in the next release, and then make the break the code in the release after.

However, I imagine most people will continue tracking SVN. Since a person needs a certain level of technical skills to use Django at all, then they probably have enough skill to handle some minor API changes.

Stability is a process. Complex software is never finished, it just runs of money or becomes obsolete.

Your turn. What do you think? I would love to know your opinions?

Discuss this post - Leave a comment

Since I moved my laptop to the new OpenRC init framework, my GPRS connection stopped working. At first I didn’t know what needed to be changed, but after having a careful look at /usr/share/doc/openrc/net.example and adjusting the proposed ppp chat script, I got it up and running again. Here is my configuration:

### GPRS ###

config_ppp0=( “ppp” )

link_ppp0=”/dev/rfcomm0″

pppd_ppp0=(
“noauth” # Do not require the peer to authenticate itself
“debug”
“local” # Ignore carrier detect signal from the modem

“defaultroute” # Make this PPP interface the default route
“usepeerdns” # Use the DNS settings provided by PPP

“lcp-echo-interval 15″ # Send a LCP echo every 15 seconds
“lcp-echo-failure 3″ # Make peer dead after 3 consective
# echo-requests

“lock” # Lock serial port
“115200″ # Set the serial port baud rate
“crtscts” # Enable hardware flow control
)

chat_ppp0=”
ABORT BUSY
ABORT ERROR
ABORT ‘NO ANSWER’
ABORT ‘NO CARRIER’
ABORT ‘NO DIALTONE’
ABORT ‘Invalid Login’
ABORT ‘Login incorrect’
” AT
TIMEOUT 5
OK ‘ATH’
OK ‘ATE1′
OK ‘AT+CGDCONT=1,\”IP\”,\”cmnet\”‘
OK ‘ATD*99#’
TIMEOUT 60
CONNECT ”
TIMEOUT 5
~– ”
“

Not much more to say, other than what was in the video.  I'm trying now to work out the best bitrate to video quality trade-off with FLV files.

I plan on having a bit of fun with this new camera in the ensuing weeks.

I'm amazed at the technology with which these new cameras / camcorders are built.

The X-Server will be getting multipointer support soon (I guess in Xorg 7.5). That looks like it might allow some awesome stuff. Using more that one mouse for example might change the way that we work with 3D design tools or graphics software. Looking forward to playing with it!

At work, we've setup the Netlog Developer Pages

It is the place where you can/will find all information around our OpenSocial implementation, our own API, skin development, sample code and so on.
We've also launched a group where you can communicate with fellow developers and Netlog employees.
The page also features a blog where you can follow what is going on in the Netlog Tech team.

read more

I've been pondering on how I am to write about this without seeming haughty or prideful, because quite honestly, I feel nothing more than blessed beyond words.

When I was offered this job, it just felt right.  Six months previous to my being offered this job, I was offered another job in SLC which seemed right in every aspect but the fact that it was in SLC.  I decided I'd take it, but after making the decision, it did not feel right.  It just didn't jive with my mojo, ya know?

Well...  I've been at Nature's Way now for almost a full year.

We've got a new server, a new website, a new sales reporting system, and I'm working on getting the website admin system up.  Life is good.

If you didn't get it in the video, might I say that today two of the three owners of this  multi-million (possibly billion) dollar company came to my office to tell me what a great idea it was to hire me.

First, the CEO came in smiling and said something to the effect of, "In the executive meeting today, lots of good words were said about you."

Then, just a couple of minute ago, the CFO came in and he, being a bit more of a relaxed guy said, "Scott (the President of Marketing) said, 'Hiring that guy was the best decision....  He has saved us so much money.'"

What wonderful words to hear in this time of economic trouble in the USA.

Anyway...  I just thought I'd write this down.  I almost wanna celebrate.

Also, for those of you who don't know, Nature's Way was a 100% microsoft shop.  The website was served on an IIS server running Windows Server 2003, using ASP and MSSQL as the scripting and database technology.

It now uses Gentoo Linux as the OS, running Apache as the server.  The scripting language is PHP and it's connected to a Postgres database.  All open-source, 100% free software.

Yeah, I'd say we're saving a load of money from the change.

Thank you, open source technology.  You're paying my bills quite well these days.

Maybe I should ask my boss if he could throw in a turbo-charger for my Mustang as a "thank you." *wink wink*

To cheer up all the Gentoo users I give you a like to the result of the Linux Sucks-Rules-O-Meter: Gentoo is by far the one that "rules" the most and most others "suck" more.

It's just the result of a perl script searching altavista for "DistroX sucks" and "DistroX rules" but still it's funny to see the results. Of course Ubuntu gets bad results because of its mindshare and because it's a beginner distribution that has many people trying it out and bashing it when it does not work but I had not thought that SuSe would be that close to Ubuntu. Nothing serious but still fun.

The sourcecode is free and you can run your own Sucks-Rules-O-Meters with it (if you can bear to fiddle around with Perl).

Ich habe nicht die geringste Ahnung, wie oft der Auslöser hier gedrückt werden musste oder wie lange der gesamte Schaffensprozess an sich gedauert haben mag - es ist das Ergebnis, welches mir bekannt ist und sich mit Superlativen gerecht bewerten lässt. “MUTO - A wall-painted Animation” von BLU aus Buenos Aires. Atemberaubend.

Via Hannes Blog.

Extensions are cool

Extension frameworks are a good way to add new functionality to large applications that are messy to alter directly, they also help to quickly add functionality now, without having to wait six months for the next release.

The whole fun of extensions is that you can make your own and download random ones from the Internet, try them out for a bit, remove the ones you don't like, share the ones you do.

This of course implies two things. Firstly, that the extension interface is well thought out so that a badly written extension does not crash or unexpectedly interfere with the main application. Secondly, that the user who installs the application is adept enough to know what they are doing.

The killer feature of Firefox has been its extension framework, having hundreds or even thousands of useful extensions is what has enabled Firefox to break into the IE-dominated browser market where so many have failed.

All your extensions belong to us

Firefox 3 has changed many things compared to previous versions. One unexpected change is that the extension framework has been locked down.

I have mixed feelings about this. I understand there is a balance between on the one hand, fun and spontaneity, and on the other hand, security and protecting clueless people from themselves. However, for me, I hate software designed as a jail for idiots. That is why I refuse to use Windows for anything beyond testing that web applications work on it.

Firefox 3's extensions framework has a new DRM-like security barrier. You have to either submit your extension to be vetted and hosted by Mozilla, or you must use SSL, or use cryptographic keys.

I am all in favour of cryptography (as long as it is open for everyone to play), but I would be worried if this implementation makes it harder for people to write and share updates.

Computer says no

I tried to install an extension the old fashion way, by going to the author's homepage and clicking on link to install the extension.

Firstly, there a warning popped up, the same warning that was also in Firefox 2:

So I clicked "Allow" and then reclicked the install link.

Then, like a naughty child, we have to wait 4 seconds as a cooling off period, before we can click "Install".

Thirdly, even though we clicked "Allow", then were put into time-out like a child, then clicked "Install", it flatly refused to do what I had told it to do:

'Secure' here being defined as "approved by Mozilla", very few Firefox extensions are secure in the formal sense of 'trusted'.

I am root

I know Firefox are trying to keep us safe, but I have to admit that the interface here gets on my nerves somewhat. There are "Allow" and "Install" buttons, I press them, but Firefox changes its mind and does not do what the button says.

Fortunately, in Firefox, the extension censorship can be overridden in the about:config settings dialog. So I went to Firefox 3's about:config page, and guess what? Yes you guessed it, they have added another new confirmation screen!

Oh well, all's well that ends well. I am sure some of you will have strong views either way on this. Let me know!

Discuss this post - Leave a comment

I have used Ekiga quite a bit. It is one of the main free software voice over IP clients. It is often installed by default on many GNOME-based systems such as many free software distributions, and it is also available on Solaris.

It also now has a Windows port, so I thought I would try it out. Giving any remaining Windows-using relatives Ekiga means they can chat to you online without you having to install proprietary phone clients that do strange things on your network.

To start with, you need a SIP username, I got a free SIP username from Ekiga's online-registration, but you can get one from any SIP provider, or you can be your own provider by running your own SIP server such as Asterisk.

To start with, lets just use Ekiga's free usernames.

So once you have a SIP username, we want to download Ekiga. the Windows binary is available from Ekiga's Windows-Users page.

Download the exe and click on it. This will bring up a little box asking you for your language. My first language, English, was the default so I went with that.

Next comes the Ekiga Setup Wizard:

Press next twice and you get this:

Press finish. I did this on Windows XP, which brought up a warning me that Ekiga is attempting to connect the Internet. Being a program to chat over the Internet, I pressed "Unblock":

Next is the First Time Configuration Assistant.

It will ask you for your SIP username and password (i.e. the username we created at the beginning of this article), and it will try to detect your microphone, speakers and webcam (if you have one).

Press "Forward" a lot and read through each of the screens. when you are done it will summarise your choices:

Press "Apply" if you are happy, or press "Back" to change them.

So now hopefully you are done with all the hard work. You should see the main Ekiga interface like this:

The first icon on the right allows you to open a text chat with the person you are talking to.

So far, I only really use Ekiga to call my friends over the Internet, I have never phoned a landline, therefore I turned the numberpad off, you can do this by clicking the second item on the left, or by going to the "View" menu, and choosing "View Mode" then "Videophone".

Typing usernames into the address bar is a bit dull, so pressing the third icon on the left of the interface brings up the address book. If you add your friends to the address book then you can click on them rather than having to remember their SIP username.

Ekiga has many more features, but you can easily discover most of them yourself.

Now you probably want to test it. Put sip:500@ekiga.net in the address bar and press enter. This will call the Ekiga 500 echo service. It is a robot who will echo back whatever you tell it, this way you can tell whether your connection, microphone and speakers are working properly.

Happy chatting!

Discuss this post - Leave a comment

Read or add comments to this article

Today I was trying to upload a lot of mail (something like 160000 mails stored in a deeply nested folder structure of about 13000 folders) to an IMAP server. Those mails were converted from the Eudora mailbox format to a more sane Eudora mailbox format using “Eudora rescue”, and then imported into Thunderbird’s “Local folders”. From there, so we thought, it would only be a matter of minutes to put those mails onto the newly setup dovecot IMAP server, with one drag-and-drop action.

But…

The upload horribly failed after the 3rd folder or so, with no error message whatsoever. No matter what I tried, it would just fail silently. The debug logging (see Debugging/Thunderbird article in the Dovecot wiki) didn’t help, and neither did dovecot’s log on the server. The problem seems to be connected with creating folders on the server somehow - because uploading a big amount (400 for a small test) mails within one folder wasn’t a problem. I couldn’t find out if Thunderbird or dovecot is to blame here…

I eventually managed to get this big pile of mail to the server by using a perl script that I found on perlmonks.org. Thanks to Davis for providing this! I downloaded it, installed the Mail::MboxParser and Mail::IMAPClient dependencies (with two ‘cpan install …’ calls), changed the configuration in the script to our site’s data and ran it. ~30 minutes later, the mails were on the server. Only thing that could have been better: All mails were ‘marked as unread’ - but that’s easy to fix.

Update:
After looking into the perl code, I found a way to ‘mark as read’ during the transfer. I posted it under the original perlmonks.org script as a comment.

Ekiga is the main voice over IP client on GNOME-based systems such as many free software distributions and Solaris. It also has a Windows port. So I thought I would try it out.

Giving your relatives Ekiga means they can chat to you online without you having to install strange proprietary phone clients that do strange things on your network.

To start with, you need a SIP username, I got a free SIP username from Ekiga's online-registration, but you can get one from any SIP provider, or you can be your own provider by running your own SIP server such as Asterisk.

To start with, lets just use Ekiga's free usernames.

So once you have a SIP username, we want to download Ekiga. the Windows binary is available from Ekiga's Windows-Users page.

Download the exe and click on it. This will bring up a little box asking you for your language. My first language, English, was the default so I went with that.

Next comes the Ekiga Setup Wizard:

Press next twice and you get this:

Press finish. I did this on Windows XP, which brought up a warning that Ekiga attempts to connect the Internet:

Since I want to use Ekiga to make phone calls over the Internet, I pressed "Unblock".

Next is the First Time Configuration Assistant.

It will ask you for your SIP username and password (i.e. the username we created at the beginning of this article), and it will try to detect your microphone, speakers and webcam (if you have one).

Press "Forward" a lot and read through each of the screens. when you are done it will summarise your choices:

Press "Apply" if you are happy, or press "Back" to change them.

So now hopefully you are done with all the hard work. You should see the main Ekiga interface like this:

The first icon on the right allows you to open a text chat with the person you are talking to.

So far, I only really use Ekiga to call my friends over the Internet, I have never phoned a landline, therefore I turned the numberpad off, you can do this by clicking the second item on the left, or by going to the "View" menu, and choosing "View Mode" then "Videophone".

Typing usernames into the address bar is a bit dull, so pressing the third icon on the left of the interface brings up the address book. If you add your friends to the address book then you can click on them rather than having to remember their SIP username.

Ekiga has many more features, but you can easily discover most of them yourself.

Now you probably want to test it. Put sip:500@ekiga.net in the address bar and press enter. This will call the Ekiga 500 echo service. It is a rebot who will echo back whatever you tell it, this way you can tell whether your connection, microphone and speakers are working properly.

Happy chatting!

Press Next a lot

Discuss this post - Leave a comment

Often when using my computer, I will have a "project", by this I mean a directory containing an arbitrary number of files and directories. This could be a software project, or a website, or an essay, or whatever.

So when going back to a project, I might want a quick overview of what has been going on. This is what my filehistory.py script is for. It recursively orders your project files by last modified date.

The unique feature is that it mixes them all together as one stream. There is of course a tradeoff, the larger the number of files, the longer it takes, however it is quite fine for projects of up to several thousand files.

There are a few command line arguments:

• By default, the output is just the filepaths, if you use -v then it outputs timestamps as well.
• By default, it will output warnings if it finds permission errors or malformed symlinks in your project. If you want to suppress these messages, use -q.
• By default, it will output all the files within your project. If you want to specify the number of lines then use -n followed by the number. So -n 20 will output the most recently edited files.

If you followed my post about making a scripts-directory, then you can just make the Python file executable and drop the Python file into it.

It really comes into its own when you pipe it to other tools. For example, to copy the last 20 modified files to a USB stick mounted at /media/disk :

filehistory.py -q -n 20 | xargs -I x cp x /media/disk/

So feel free to check out my filehistory script. As you know I use Linux, but I think this particular script should work on Windows/Mac as well.

Discuss this post - Leave a comment

Quick reminder: If you use dm-crypt on Gentoo to crypt your harddrives and you have trouble entering the right passphrase it might be that you don't have a US keyboard. This bug is the one responsible: The proper terminal keymap is not set before starting dmcrypt (which means you have US keyboard settings and not your localized ones). Annoying and makes you think you are insane

Since the newest available fotox ebuild I found was for version 17, I decided to change it for version 42 which is currently the latest and greatest. To compile and install it using portage you also need a minor patch I made for it’s build system. fotox-42.ebuild fotox-build.patch

Ubuntu, Canonical's linux distribution "for human beings" has been dominating the linux news and reports for a few years now: At first out of the "look at the freaky African rich guy who tried to build a distro" interest, then because blogs and other media started picking it up. Today Ubuntu is pretty much a synonym for linux for many people. Opensuse? What's that? Fedora? Never heard of it. Debian, Arch, Slackware, Gentoo? Words without meaning.

I'm writing this as I'm wrapping up the Gentoo installation on my thinkpad that replaced the horrible mess that was Ubuntu 8.04 ("Hardy") and it kinda got me thinking about how it came to be that Ubuntu dominated the mindshare.

We don't often see a complete reduction to one product in a certain area (it might be comparable to the Ipod when it comes to MP3 players) and there are some reasons why that usually happens:

• Technical superiority: One product is just that much better than the rest that, while the competition is playing catch-up, the leader improves even further.


• Elegance: One product is just that much prettier or stylisher than the rest. The sold image is just that much more appealing (pretty much the reason why the wannabes all run around with Macbooks).


• Price: The one product might not be the best but it's just so cheap that it would be retarded to buy something more expensive


• Novelity: One product is just new and so different that it becomes the thing to have. It changed the rules in some respect that no one had thought about before.

So what's the reason for Ubuntu's success?

Well we have to make a quick reality check first: It's hard to see how the actual installed base of operating systems is so the fact that there is a lot of buzz about Ubuntu does not mean that the installed base of linux systems is in any way related to that. When we are talking about success here, we are actually just talking about publicity and representation in media (whether blogs or traditional doesn't matter).

Ubuntu is obviously not technically superior to the rest. While Ubuntu has added a few things like graphical dialogs here and there that can simplify things (and often keeps those patches to themselves instead of sending them upstream so everyone can benefit) the base technology is not that different from the rest and in some aspects it's actually behind the competition (pulseaudio integration being one of the examples).

Ubuntu's default theme is actually not bad (I kinda like it) but it's not that much prettier than the competition.

Ubuntu is free of charge but so is the competition so price is not important. And novelity is nothing Ubunut excels in either.

So where does the success stem from?

It comes from two things mainly:

1. Communication/community: Ubuntu has Jono Bacon as community manager. Then there's the forums and wikis. All in all it's like a bee hive of people all generating buzz and noise. Some of their work might be sub-par but still it's a stream of constant buzz about Ubuntu that keeps the word in the public discussion.



2. The slogan: "Linux for Human beings". I have no clue who came up with it, but that person is, in my opinion, responsible for at least 50% of the success of Ubuntu. The slogan is not geeky, it's not technical, it's talking about human beings (which we probably all are [except Harvey the invisible rabbit which is obviously reading this]) and it talks to feelings. It tells you that this is for you. No restriction, no catch.

Ubuntu has applied the strategies that we know from marketing and advertising better than any other open source project (well the KDE people are quite good with that, too, with, based on the slogans, some people believing that KDE4 was the second coming of christ or the revolution of the desktop).

Ubuntu is not a bad distribution but its representation in our media is not based on merit but on advertising. People believe that it's easy because everybody says so. When it's not easy they look up stuff in the wiki or forums or ask. How is that different from Debian or Fedora? It isn't. But those don't claim aggressively enough that they are easy.

It's, like so often, a story about makebelieve. Make people believe that things are easy and they will claim that they are. Installing software on windows is a pain in the butt: No dependency resolution, hunting down software yourself. But still people claim that it's easy. Because that's what they were told. And that's what they have convinced themselves of.

Let's see how long it takes the distros competing for Ubuntu's user base will need to realize that they have to work like advertising people do. Right now is the right time with the Ubuntu devs dropping the ball on their last release.

Quite a few people have been wondering how they can get RSS feeds which are specific to the individual users of ILoveMyJournal.com.  Until now, they couldn't.

ILoveMyJournal.com now has feeds which will contain public posts from only the user you're interested in!

To get it, you simply click on the RSS feed logo found within the user's blog.

First, click on the public sub-nav on the title bar,  then click on the user's blog you wish to view, then click on the RSS feed logo, and you'll be there!

My first Linux distribution was a very old version of Red Hat with X-Windows as a GUI. I think I was in 5th or 6th grade, and I bought it from a used book store for $10. From there, I went to Mandrake Linux 7, then Fedora, and finally to Gentoo. I was never a fan of RPM installers, as there were always issues with dependencies and uninstalling.

One of my friends, Zack had recently become a fan of Gentoo Linux. I abandoned RPM hell and became a fan of Gentoo. Despite, at the time, the best way to install Gentoo was through a stage 1 (compiling everything from start to finish), I loved it. Even if it took a day to get a usable system on a 400MHz Pentium 2, it was great.

Despite some obvious issues with dependencies in Gentoo, and many claiming Portage is just a slew of unmanageable code, I'm still in love with the speed and configuration of Gentoo. Sure, some things aren't install-and-go, but you learn more when having to configure most programs yourself.

Recently, I've been installing the latest version of Ubuntu on many of the machines I use to benchmark hardware for BIOSLEVEL. I can't say I'm a fan. I'll agree that it's fantastic that Ubuntu developers have put so much time into developing and improving many applications and daemons for desktop use, but it feels slow.

Why run native 586 code on a 686+ CPU? Why install software that doesn't take advantage of a CPU's extensions? To be completely honest, using the same hardware, a Gentoo System with Compiz Fusion running feels more responsive than an Ubuntu system without any desktop effects enabled.

While much of my testing will be done with Ubuntu, my primary Linux environment for regular usage remains Gentoo.

I've been needing to write a bit more about what's going on in my life lately, so... I'll give it a try this evening.

First, you'll notice by the video above that yesterday morning a sad event occurred.

No, my cat didn't die.

No, my computer didn't fry.

My mohawk went bye-bye.

Yup.  I woke up yesterday morning and just couldn't get myself to spend the necessary time to make it look acceptable, so, with the alternative being to wear a hat, or look like a complete idiot, I got out the buzzer and within a couple of minutes, it was gone.

Also, today I had a run-in with a police officer of the law.  hehe....  who just happened to be driving a camaro with some serious power.

He drove past me going about 65, I was doing 60.  I then caught up to him, and as soon as I passed him, he slowed down to about 2 car-lengths behind me, pressed the gas just a bit, and I'll be darned if I didn't hear the unmistakable sound of the supercharger waking up.

It made me smile just knowing that he couldn't let me go without knowing he had some power to punch.  hehe....  Anyway...  I got a pic of his back-end as he pulled off the same exit as I.

Anyway...  Good times today.

Hopefully I'll be chattin' with ya again soon.

Also - if the video at the top of this page didn't load for you, or you had any problems with it, please comment, so I can know what's going on.

Thanks.

Okay...  I've been wondering how to do this for quite some time, and I finally figured it out with the help of this mailing list thread.

I've always thought ffmpeg was a pretty amazing program, but now, I will build a shrine to it and worship it weekly.

Ffmpeg is the program which allows me to create a video with sound in Linux.

I have a Logitch Orbit AF webcam, and with the following command, it records both video and audio.

ffmpeg -f video4linux2 -s 320x240 -i /dev/video0 -f audio_device -ac 2 -i /dev/dsp1 -f mp4 Filename.mp4

and the file "Filename.mp4" will be created and stored in the current working directory.

Great stuff!

After trying to get this video to work with flowplayer (flowplayer.org), it would reproduce the audio, but not the video, so in order to get it to work (albeit pretty low quality), I had to convert the m4v file to an flv file with the following command:

ffmpeg -i input_file_name.mp4 -vcodec flv -ar 22050 output_file_name.flv

That seemed to do the trick.  Here's the first video recorded using this method.

PS - Even though my camera is v4l2, I had to enable the v4l USE flag. Make sure you have the v4l USE flag turned on when you emerge it, or it won't work.

Go gentoo!

I cannot believe its happening to me.