DoctorMo's Blog

Since this blog was founded with the simple aims of getting my ideas and thoughts out of my head and onto something more solid, so I could get on with other things. It’s also had the basic aim of improving my writing and presentation skills.

But since the Ubunchu Manga was released the stats for my blog have gone crazy. First it was only getting the a few hundred hits from Planet Ubuntu, but then I started getting a pile of link backs from lots of different places, some manga sites and some ubuntu blogs.

Today was the strongest showing by far, a blog on wired and a position on boingboing seemed to double the total from the average of about 3,000 to 6,731.

But that doesn’t mean any of these visitors are downloading the Manga, and they’re probably all Ubuntu users anyway right? Well I checked the server logs for the pdf and source tar. The pdf has been downloaded a massive 85,244 times since it’s publication a few weeks ago. 22,000 of those were in the last 4 days alone.

For the operating system stats, 54k of those downloads were using some form of windows. 2 of them even used Windows ME. Linux accounted for only 20k and MacOSX for 3k. Does this mean that there may be more windows users now interested in Ubuntu than before? I don’t know, I hope so. (The rest of the hits were unknown or small amounts from various things like BeOS, BSD and SunOS). Although only 25% of the hits come from Microsoft IE, most of them coming from Firefox, Konquorer, wget and Opera users.

These might not seem a lot compared to the number of Ubuntu downloads or the number of Firefox downloads. But it’s way beyond what my blog and my wife’s website is used to. This one PDF had chewed up 170GB of bandwidth, 8GB a day (average).

Anyway this has been a fascinating ride into the online blog-o-square and getting a blog entry that is “sticky” as my friends MenTaLGuy says.

Remember that post I made a few days about Dell support? Well it appears that others have struck up a general mood that Dell is under serving their Ubuntu customers with support.

What is most interesting is that some people decided to use my blog to post unenlightened, highly critical attacks against Ubuntu (Gnu/Linux and FOSS in general). For instance this vitriolic comment by DamnRight on this blog entry, I’m not going to quote it directly, because I feel it would breech my personal taste guidelines, but I’m going to try and provide some useful arguments here:

Normal people (who ever they are) may not want Ubuntu specifically; but to trying to solve a political problem with a purely technical solution kind of misses the point a bit. Users may not care about FOSS technologies, but they know none-FOSS is rotten. Just because “normal” people don’t understand the solution, doesn’t mean they don’t feel the pain brought by the closed nature of technology. We’re trying to provide solutions to other people, because it’s the right thing to do and I can’t see how we could possibly stop trying.

People who I talk to, who are not technical, get the implications of the political imbalance (re Microsoft) much more so then the technical merits of open source. The woman I talked about in my support breakdown post, I called her last night to make sure things were going well and she was happy to report that she had been telling others in the hospital where she works about the volunteer efforts trying to bring choice to computer software. Even one so burned by Ubuntu on a Dell machine was willing to see the reasons for what we do in the community.

I only mentioned it because she was very curious as to why I would volunteer my time for such a thing.

It’s not really insane to believe in freedom of speech. Such a bloody hard won ideal which has given us the sciences, arts and progressive philosophies since the enlightenment should not be sneered at. I can’t really agree with those who are against the Freedom in FOSS, that includes getting basic technologies out from under the hegemony of Microsoft.

I’m sure that to Dell, Ubuntu is an experiment into expanding their offerings as well as giving them a way to leverage Microsoft better. I don’t deny that it’ll cost Dell to get all pieces in place and get all the processes smoothed out. They may end up giving up, but I still commend them from trying. It’s worth learning from and to keep on trying, even when processes break down and you have to find another way of working it.

Now worse is people who actively support Microsoft, not because their business depends on it or because they’re invested into it, but because for them it’s tribal. They’ve placed themselves into a purple sash and look upon anyone outside of their tribe as being dangerous and different. Perhaps that’s why some people post comments full of spite, bile and hate. It’s a kind of fearmongering, trying to demoralise the ‘other’ tribe with the beating of the drums.

On saying that as Jared Spurbeck says in the comments below, it may just be a kind of insecurity, an attempt to undermine the “competition” in order to strengthen ones side. Even so, that insecurity sounds like picking sides to me.

There has been some worrying news that the Obama administration, whilst dealing with many issues in a positive and progressive way, has begun to hire on RIAA lawyers to take key copyright positions in the administrations.

This is a very strange move unless you think about copyright properties in a wholly economic fashion. I can’t believe that the government would deliberately try to bring people from the most unenlightened and morally dubious legal enterprise in modern time.

Lest the argument that the general public should not be criminally charged for doing what comes naturally: Sharing Stories. We are a species that loves to share interesting stories, music, ideas and it’s in our blood. Now we have a set of society trying desperately to nudge popular opinion into the idea that sharing is bad.

What by all rights should be happening, is a more enlightened and more engaging discussion about how best to redesign and re-engineer the practical economic model for creative industries. Even if we just ended up having a debate about the reasons for criminal copyright enforcement.

But it seems that on this one issue, I must disagree very strongly with the Obama administration. So far, the signals appear to say that they are willing to do what ever it takes to protect the big media companies’ properties, even if it means damaging free speech, surveillance without reasonable suspicion and hurting the prospects of alternative creation models while this protectionist game is played out.

Hopefully I’m wrong.

I was surprised this week by the release of a video detailing the plan to move forward with investments into Hgih Speed Rail. This is an important issue which looked like it had dropped off the presidents rhetoric about improving the USA’s infrastructure.

But now it’s back! And I’m happy to see that there are people in the government who are as concerned as I with this counties failing rail ways. Forget getting on expensive, slow and substandard trains, I’m looking forward to getting on modern, cheap more European rail connections.

Trains aren’t some bridge to no where, they perform a very efficient function in allowing people to live in one place and work in another. Take London as an example, without it’s extensive rail and subway system you’d never be able to get from Brighton to the West End in time for that play,  going by car would be impossible if everyone else was also going by car and there is simple not enough physical space for everyone of London’s workers to drive into the city.

Now to get things cheap. Back home, I could take a train from Liverpool to London Euston (240 miles) in two and a half hours and so long as I booked 2 weeks in advanced it only cost me £23 return (~$30 PPP). What we use in England is a chartered agreement system where even though the train operators are private (much against my personal belief) they at least have to agree to some rules which provide the poor with access to the train network at a price they can afford.

Compare that to taking the National Express coach which takes six to seven hours and would cost £36 return in the same kind of cramp style that Americans would be familiar to in the Greyhound bus services.

No, it’d about time America brought it’s train technology out of the 19th century and into the 21st. I just hope that consideration will be given to allowing the poorest of citizens to access this most important infrastructure.

In one of my previous blog entries about a Dell Support issue some of the comments suggested that the reason we were in this mess was because of the inflexible nature of the Linux kernel, the start of which was this blog entry about how evil the GPL is when it comes to making closed source drivers and why this is stopping hardware manufacturers from contributing to the kernel.

I’ll try and address all of the issues that came up in the discussion, and then if those who’s comments I have removed for being OT from the previous article want to discuss further, they can on this entry.

Hardware Manufacturers Need Trade Secretes for their drivers – False, hardware is not sold because the software is good, it’s sold because the hardware is good. The loss in effectiveness and quality that a manufacturer will suffer from keeping everything in house will be massive as the amount of possible peer review and reuse of other modular parts is small.

The Musings blog above states that “where some zealots believe that people who make money selling software are somehow evil.”: Firstly hardware makers make money selling hardware, not software. Secondly, the GPL has no issues with people making money from selling software, it’s concern is with the reciprocal freedom you must grant to all users. There may be some reorganisation that’s required in order to make money effectively but none of these issues effect hardware manufacturers.

A constantly changing kernel is required to support latest hardware – False, there is no technical reason why the kernel must be changed in order to support hardware that isn’t in the main line. It’s just not something we’ve focused on solving. It’s true that if you have a fix for your hardware as a patch or source that our current packaging systems won’t keep track of it and automatically add in your extra or fixed modules. Even reverse deps for the kernel when it comes to things like virtual box cause functionality to regress when updating. But these are distribution issues (for Ubuntu or Debian to solve) not for the Linux Kernel.

We will never have the year of the Linux desktop until Linux can stabilize their driver model to the point that new drivers can run in 5-year-old kernels – False, if your release cycle is 5 years, you’re doing something wrong. If your computer is running an OS that old, then we’re doing something wrong. A modern OS should be flexible enough to continue to run on old hardware, doing the same things it did before without needing extra resources. Logically only extra functionality requires more resources. Although entropy might be playing a joker card.

All Advocates want to see Everyone using their Linux Distro – False, Not all advocates care about market share, or perceived market share, they care about being treated fairly. If removing the freedom aspect of my distro is required to make it popular, then I say that we must find another way or simply admit that the cost is too great. I don’t think this is something that we have to do though.

There is no difference between closed source and open source production, only the people involved – False, open source is not just about where the code is made, it’s about how it’s made. The fact is, FOSS has an efficiency that can speed past closed source, even when more attention is being paid to code quality and a greater number of releases.

None Technical Users don’t need access to the Source Code – False, all users should have access to the source code so that they can hire or in some manner convince a programmer to fix problems and help them effectively. No problems in proprietary software is ever fixed outside of the company that writes it, only workarounds and mitigations are possible.

Politics has no Place in Software – False, whenever you’re talking about more than one person and furture insight more advanced than next week in any human endeavour you will run into politics. Social issues are not simply distractions from rigorous logic. Practicalist extremism is just as unfortunate as political extremism. Idealism is a guiding light, it might not tell you what to do, but it can help explain you the long term technical and social costs of a given solution. If you don’t care for other people in your community it doesn’t mean that those issues are not important, it just means your not able to be compassionate and considerate towards others enough to take effects towards them seriously in discussions.

Linux programmers should be catering to the hardware suppliers’ every whim – False, hardware makers are in business of selling hardware, if they want their hardware to be effective then they will cater themselves by joining the community or pay someone already in the community to make sure their hardware works effectively. Linux does not owe OEMs some social responsibility to cater to their whims, not unless those OEMs are paying for that service.

Linux programmers deliberately make deployment as difficult as possible – False, all the top linux programmers I’ve ever known have been happy to work with and give their time towards hardware manufacturers who want to join in as partners. The Linux Foundation even went out of their way to make sure there were no issues by allowing members to sign NDAs with certain companies to acquire their specifications. Linux programmers are champions of working well with others, despite their rough social standards.

Even when vendors do release specs/source, the result is still sucky – False, most code done for Intel hardware is made by Intel developers, most code for the AMD ATI graphics cards are fairly young and are already making good progress with stability. Most FOSS code is of better quality then closed source, this includes drivers. One of the reasons companies don’t want to open up code is because they’re embarrassed by it’s quality. see bugs per 1,000 lines.

There are things we could do to improve the driver situation, introducing some form of source based package which sits in a special directory where all the custom drivers sit, that are automatically recompiled for each new kernel update. We could further modularise the kernel driver set, so updates can be more modular too. We could do more about making sure that updates are held back when functionality will be lost. But whatever, it’s up to the developers and the packers of the kernel.

Now I’m off for a nice walk, it’s a nice day outside.

Yesterday I got a call from a woman in Massachusetts. She had a Dell Mini 9 which she got on a deal when she bought an XPS Laptop last year. This Mini 9 was running Ubuntu 8.04 with Dell’s modified netbook GUI.

After some automatic updates 6 weeks ago, the computer froze and wouldn’t come back. Turning it off and trying to boot it up failed. It seems the computer failed to update the kernel correctly. Now she was in for the ride of a life going from Dell support to Dell Ubuntu support, and somewhere along the lines being thrown over to our LoCo website.

The Dell Ubuntu support attempted to get this lady to reinstall the OS from the CDDVD (even though the Mini 9 doesn’t have a DVDROM drive) and having the gual to sell her a cheap rickety external DVD drive which doesn’t actually function for DVDs without extra power (cord not included in purchase, so I’ve advised her to send back for a refund). It boggles the mind why they would ship CD drive-less machines with a CD to recover them. Wouldn’t a 1GB USB stick install be a better thing to ship? You could even keep the install updated through some clever means.

Not that I don’t mind helping people, but this is really Dell’s problem and I’m slightly offended that their support is so bad that they need to direct people to local volunteers in order to get computers sorted out. This isn’t the first time this had happened either, another young women bought an XPS machine with Ubuntu and had terrible trouble with Dell Ubuntu support and also ended up finding our LoCo and getting involved to get her issues fixed.

Talking to Canonical support people myself I found out that Dell isn’t playing fair with them either. Dell doesn’t have a contract with Canonical for complex queries (tier 3) calls to go to Canonical (not that this issue really needed tier 3). I’m sure Canonical would love to have such a deal in place, but Dell is unwilling to pay the price for it (although if the price is anything like that charged to end users, I’m not too surprised). So some of their support staff have been directing people to Canonical support, even though Canonical can’t do anything for Dell customers other than sell them end user support. I know this because the XPS owner mentioned above had this happen to her.

So, what did I do. I got the Dell Mini 9 owner to drop it off at my home and I reinstalled it with Ubuntu 8.10. Gave her a few pointers and she paid me for my time. She’s now happy to have her working computer back. Hopefully, I’ve tempered her opinion that Ubuntu is a crack-pot cheap operating system that no one should use in place of Windows.

I am genuinely concerned about the cost to the Ubuntu trademark if Dell persists in under supporting it’s customers. While I don’t mind the extra money; it’s going to make advocating Ubuntu much harder if people associate it with crappy, always breaking, Dell computers that their friends bought.

What I’d like to see is Dell support training in some way in the community, building relationships between their support people and our support volunteers, I’d also like to see some of us hired to do more education and improve their abilities to support this platform. Although that’s remain a fantasy I think.

Download via PPA Here

Last year I created a tool one weekend that could randomly choose a wallpaper from a source, download it (if required) and display it as the gnome background.

It worked great for Hardy, but broke horribly for Intrepid. I was updating gconf with the correct details and making sure the backgrounds selection was all handled correctly. I was doing things the right way. But gconf can no longer be modified from a user’s crontab and my program died.

I was not prepared to create a crontab like daemon just for the task of changing the desktop wallpaper, it falls foul of the DRYs (Don’t Repeat Yourself) principle. We shouldn’t need to replicate functionally when a perfectly good system infrastructural component will do it and save the user running resources (especially when my replacement for crontab would have been written in python).

So, I’ve managed to get around the problem by doing it the wrong way, instead of updating gconf the program will now just replace a backgrounds file in .gnome2/ settings folder and make sure that an entry exists in your appearance options. So long as that entry is selected, your background will be random.

Do report bugs in the launchpad project if you try it out and find some. I’d like to get this into the main repositories at some point.

The wall changer project that I did quickly last year, to control gnome wallpapers via crontab. Was a pretty cool tool, and lots of people said so.

But with intrepid it broke, because of some new security feature where gnome settings (gconf) can’t be accessed from crontab.

So I think there may be a way around it, gnome will automatically refresh a wallpaper when the file for that wallpaper changes. If I can hook in a single file and regardless of the options, update that one file, then it should be possible to update the entry without gconf.

I’ll post an update it it works.

Sometimes I like to relax, and nothing is more relaxing then a game of Savage2.

It’s not open source, but it does have a linux client/server that is fairly easy to install.

The history of this game and it’s company S2 is most interesting. The first game “Savage1″ is now FOSS and was created by the main people at S2, they decided at some point to incorporate into the group they are now and produce a second version of the game.

It’s unique mix of straedgy and combat make it an excelent balancing game. You can’t just go all guns blazing (mostly because ranged weapons don’t do that much damage) and what do you do with you’ve got a Behemoth stomping towards your garrison, shoot it with an arrow? I think not.

I miss the Balister from the first version of the game, but the hellbourn units make up for this easily. Nothing is better than killing 20 players in their main base whilst protecting your seige units.

So is anyone else still playing?

OK so after my previous blog entry about the LoCoRST project, we wanted to allow users the chance to help make it happen.

And I believe this project is very useful for all Ubuntu users, those that support users and those that receive such support. For home users that want to feel safe that they have a way to get help, to LoCos that want to feel safe that the install they do for someone won’t be riddled with problems that they are unable to fix the next week.

So I’ve already requested funding and programming assistance from the LoCos who will most likely find this tool very useful for their support of remote home users. Now I want to invite users on the planet and other Ubuntu users who have to support other Ubuntu users remotely.

So we’ve set up this cofundos project to collect money, this project has 3 very clear requirements so no one should be afraid that they’ll be investing their money into the void. Support the project here:

http://www.cofundos.org/project.php?id=168

Please visit here if you think you can help us raise the money, or if you think you can program python and xmpp and clear up the requirements. We want to be able to pay someone a fair price for working on this, so I’ve put decent money in myself.

Let’s not be the community of cheapskates, lets be the community of wise investors that use their money to create new things. Even if it’s just the odd €5-€10 towards a worthwhile and useful project. It’d also be good if we got familiar with websites like cofundos that allow us to pool our money for these types of tasks.

Visit my previous blog entry to see what the support tool looks like:

http://doctormo.wordpress.com/2009/04/09/gui-walk-through-of-the-loco-remote-support-tool/