Well I am sitting here on Thanksgiving with not much to do because Paige is hogging the PS3 playing Resonance of Fate. So I figured, why not check out KVM, I did download the Linux Mint 12 rc cd image to install on Paige’s Pangolin.

After loading the kvm modules and creating an image use qemu-img I am now most of the way through the install process and I must say, KVM is sweet! It seems much lighter than VirtualBox and I love how it is geared towards the command line.

I plan on testing this out for other things.

Hooray! Gnome3 is in the standard Arch repositories! Hooray! I install it! Boo! It says my grahpics card is not capable of runnin Gnome-Shell so it drops into fallback mode. What the heck?

I had the same problem with Unity on my Ubuntu system at work. Now granted both of these machines are at least four years old and running an fairly run of the mill Nvidia chipset for their time. The laptop has an Nvidia GeForce Go 7300 and the workstation at work is sporting a GeForce 5300. Now the issue with Ubuntu is that none of the Nvidia drivers will work with the card as it has been deprecated by Nvidia. I did not think the same thing would be an issue with the System76 I have, but the odd thing is that the Nvidia drivers work just fine and I do have 3d acceleration. So why won’t Gnome3 run?

I even went as far as to use the Nouvueau driver and that did not improve things at all. I’m stuck with fall back and I am stuck with Gnome2 when it comes to Unity at work. But you know what? That is fine by me because I use Fluxbox anyway so bully!

Still, this brings up an interesting problem with regards to both Unity and Gnome Shell being so dependent on 3d Acceleration. Having to deal with Nvidia’s closed driver leaves a good portion of users out in the cold. Do ATI cards fare any better? I’ve only heard bad things regarding that hardware for some time now. How about Intel? That seems to be the way to go. But alas, there is no way I can swap out the chipset in a laptop. And the cost of an AGP card that would work in my workstation at work far exceeds the price I am willing to pay just to give Unity a spin

Nope, looks like I am relegated to Fluxbox for the time being and that is fine by me.

After episode 142 of the Linux Action Show where the hosts exclaimed strongly against Google dropping h264 in favor of WebM largely because WebM is an “Open Standard” I was prompted to discuss this further in their forums and came to the the realization that I may have been completely misled as to what the term Open Standard truly means.

I subscribed to the definition that most fits Bruce Peren’s concept of Open Standard defining it, roughly, as a standard that is publicly available and can be implemented without restriction. Now he goes into more detail with six tenants and these are further expanded upon by a wonderful paper titled The Meaning of Open Standards.

Review the Wikipedia definition of Open Standard and it becomes apparent that there is no single consensus on what an Open Standard is. Most subscribe to something akin to Peren’s Six principles but what about ITU-T? Their definition states that an Open Standard is a standard that is collaboratively developed, balanced, publicly available and implementable via royalty free licenses or on reasonable terms and conditions. The last point contrasts Peren’s Point 3 – Royalty Free implementation and most other defintions. So which definition is right? They cannot all be? Yet it would appear that OpenStandards.net would have us believe they can all exist under the same umbrella.

On the Jupitercolony forums a contributer, ShawJGroff, replied to my post on this topic with the following statement:

“Open standard” describes the standard as an entity by itself: it does not describe the thing the standard is defining.

Now this is not a definition I have ever heard put forward for the term Open Standard. So if I re-iterate what my post was about: “Is h264 truly an Open Standard.” By his definition what is meant by saying h264 is an Open Standard is that the published standard for h264 is openly available to all and the term Open Standard in no way should be applied to the implementation of the technology, h264, detailed in the standard.

This is the the single most salient point I think that all definitions put forth for Open Standard agree upon. That the standard is open to all to read. There is no mention of implementation, privileges or restrictions put forth in this definition. So is this the true historical meaning of the term Open Standard? Has the term Open Standard become polluted over time particularly by members of the Free and Open Source movement to apply it more towards the tenants of Free and Open Software?

When Steve Jobs stands up and says h264 is an Open Standard are we bucking up against sociological, political and generational corruption of what he truly means? Is he saying that and Open Standard refers to the availability of the standard itself and has nothing to do with the implementation, restrictions or privileges defined by the standard and licenses applied to the standard? Is there a purposeful play on words here to describe a technology like h264 as an Open Standard thus playing to the popular “buzzword” term of the day ascribing Open Standard to be akin to Peren’s definition when in truth it is merely the publication of the h264 standard itself and nothing more?

Has the term Open Standard become too polluted now? I’d be interested to hear other people’s opinion.

I do not work with XML enough, that is a fact. Even more, like most technologies I come to I spend more time putzing around implementing it than I should researching it. Now XML is fairly straight forward but validating with a schema is not as easily grasped. So I am passing on some bits of info I learned today while pulling my hair out over some XML that was not validating.

First, most validators suck. I went through a handful today, most of which just told me that it could not validate or gave me some error which was not at all helpful. I finally settled on this which provided the most information http://tools.decisionsoft.com/schemaValidate/. Had I started with this tool I could have saved a few hours of work.

There are two element types: Simple and Complex. When validating with a schema be aware that simple elements cannot have attributes. Thus, if you have an element without any children but with attributes you must declare that element as a complex element and use extension:

xs:element name="lemonTunaMan"
xs:complexType
xs:simpleContent
xs:extension base="xs:string"
xs:attribute name="HoagieGuy" type="xs:string" /
/xs:extension
/xs:simpleContent
/xs:complexType
/xs:element

How about that?

When working with child elements it i a good idea to be aware of occurrence and to specify occurrence. The way to do this is to specify minOccurs and/or maxOccurs. By default the values of these are 1. Now most documentation I was reviewing talked about and showed examples of using one or the other but not both which really threw me.

I had a situation where the test instance I was using to pull the XML did not have complete data and therefore some child elements were not present. This kept throwing and error in validation. Now generally these elements would have values but I know in dealing with other people there is a great than average possibility that someone would not have entered in all the data necessary. Therefore, I needed to cover this base and not just ignore the problem.

The value minOccurs specifies the minimum number of times a element can appear and maxOccurs specifies the maximum number of times. Again, left to defaults, not declaring in the schema, the values are 1 meaning that said element must appear once and only once. Now you can set minOccurs to 0 which means that the element can appear 0 or 1 time. NOTE: 0 or 1 time! You can set the values for maxOccurs to equal “unbound” which means that the element will appear 1 or more times. So what if you want to set an element that can appear 0 or more times? Well then you have to do this:

xs:element name="hairShavingsColor" type="xs:string" minOccurs="0" maxOccurs="unbounded" /


You must declare the minOccurs to equal 0 and the maxOccurs to equal “unbounded.” No where did I see this made evident I just happened upon it by deductive reasoning.

Oh the fun! No my XML validates against the schema and all is well in my world until Monday when I add more elements!

Happy days!

Whoohoo! Paige’s new netbook came today – the System76 Starling. She is very happy, that she is. We got it all set up for here and transferred over her files from my System76 and sent her on her way. The Starling is running Ubuntu 10.06 netbook remix edition and she was not liking the interface so we switched her back to standard gnome and she is configuring away out there.

The Starling is a nice little system. I am not too keen on the track pad but the rest of it has me envious. If it were my system I suspect I would become accustomed to the pad, but never-the-less; it’s all hers.\

We spent the New Years holiday down with my Mother-in-Law in Chester, South Carolina. She runs a bed and breakfast out in the middle of no-where so cell phone’s were not working. On top of that the server seemed to drop down again the day after we left and did not come up. Long story short, the server hiccuped by my scripts caught it and for some reason, while the server was fine and running the isp was not reconnecting. I should have called them and will in the future. But….

My M-I-L had a system she never used anymore with a hard drive that had gone south. She was going to get rid of it so I brought it home, slapped in an ancient ide drive (this thing has sata and it pains me, that it does), installed Arch and now I have a back door into my network should the server go down. So if the server goes and my back door is not accessible it is time to call the ISP (Cox, who has been very good to me).

I had to work with the technicians today to get the IP working. I had done this a few months ago on my other server and it worked for a bit but then stopped. I chalked it up to a problem with my routing and since things were running very well for a while did not get around to fixing the problem. Well lo and behold here we are with a new machine and the same issue. I had to work with the Cox technicians to diagnose this and ultimately we could not figure out what the problem was.

The last thing we tried was plugging the modem right into the nic and that worked. Plugged the modem back into the switch and the nic into the switch and all was fine. It survived a reboot too. For some reason before doing this the modem was not even picking up the node connected to it. Weird. They are going to send out a new modem.

Well I am off now to bigger and better techie things. 2011 should be a lot more exciting and geeky for me.

For some time now my virtual server has been going up and done with some frequency but I cannot find anything other than a cryptic error message in the vmware logs that states I should send it to vmware for review if I have a support contract which I don’t. It only seems to affect my main server, not the asterisk or streaming workstation images. So I suspect there is a problem in the image itself, a bad block or something.

Saturday was the last straw so I decided to move the server images off the main disk onto a second disk as I heard this will also improve vmware performance. Having the hosts images on a different disk than the server seems to be the best idea. I also too the time to clean out the dust balls in there. Boy was it a mess in the vents. This cleared out any disk errors related to overheating I believe. And while it seemed to perform a lot better, the main server image was still going up and down.

For some time now Vmware web access has all but stopped working with Firefox 3.0.x. I wanted this mainly for remote console but I cannot even get in to the damn thing with firefox anymore without borking the management of the server side. It baffles me. Sometimes Chromium or Konqueror works but I cannot run the remote console.

I found some references on the web on how to run remote console outside the browser: http://www.geeklab.info/2010/02/running-vmware-remote-console-outside-the-browser/

I found the file I was looking for at: /usr/lib/vmware/webAcces/tomcat/apache-tomcat-6.0.16/webapps/ui/plugin

I unzipped it on my workstation and fired it up with this command line:

./vmware-vmrc -h myipaddy:8333 and it worked like a charm. I was able to get in there to my main workstation, boot into single user mode and check the partitions. I hope this rectifies my problems. I think it may. I noticed the first time I ran fsck on my one partition it dumped the host instance immediately. Subsequent runs did not produce the same effect. We shall see (I keep my fingers crossed).

I am in no way a big fan of ink jet printers. I think they are cheap and the ink costs way to much. Not only that, but the ink invariably mucks up the print head mechanism rending the printer a paper weight over time. This was a huge problem with Epson printers (I so hated trying to clean those damn things, took hours and wasted so much ink) but not so much HP’s who put the print head mechanism on the cartridge. Still, ink jets are a far cry from laser printers. I have one of those, a trusty HP LJ 4p which has served me well since I picked it up used over 7 years ago; and it was in service for many more before that. This thing has chugged along with the same toner cart that came with it and I have a spare. What I don’t have is a parallel port on any system since my workstation took a dump. Thus it serves me no good right now.

Well this does not make the wife and kids happy that there is no printer to use, especially for homework assignments so it was with an upturned nose I met my wife’s suggestion that we pick up a cheap printer on Black Friday. She had seen a wireless HP OJ 6000 in some add for $60, about half the price. I was skeptical, said sure, and left it at that. So Friday rolled around and she was like let’s go. When I got there I read the box and there was no indication that it was compatible with Linux. Only Windows and OS X on the box, but this is to be expected. So I did a quick search on the web and found a compatibility chart on HP’s support site listing all the distros with support. Of course Ubuntu was at the top of the list stating it could print via usb and network but that setup of the printer, especially for network may not work at all. All the other distros either had printing support or no support listed. Or course my distro of choice Arch was not listed at all. Still, I knew, that if the printer worked with Ubuntu, Fedora, Slackware, and all those other distros listed it sure as heck was going to work with any other distro released at or around the same period as the distro versions compatible. I know this, but perhaps the new Linux user would. Never-the-less, I decided to go with it. I was not overly concerned about getting it to print wirelessly or through the network as I could just pop it on my System76 and share it out through cups.

Well I finally got around to setting this thing up today and man, was there a lot to put together physically. Getting all the tape and plastic off then putting in the 4 different ink cartridges took a hell of a lot longer to do then getting this working with Linux.

I fired up my web browser, typed in localhost:631 and set about working magic with the wonderful program CUPS. I selected to add a new printer and sure enough, a few seconds later it detected the HP which I had plugged in to the usb port. But wait, it was not the usb it detected, it was the newtork printer. Whoa! I could see the IP address which meant maybe there was a web interface. I popped the IP into Chromium and sure enough web administration tools at my fingertips. In a few minutes I had it set up with a static IP through the network and the wireless perfectly configured. Went back to CUPS, set it all up there and was printing out a test page in less than 2 minutes all totaled.

I went to my Meso, fired up the cups interface and was completely configured in less than 60 seconds.

HP what the hell? Setting this printer up under Linux was a hell of a lot easier than any of the directions listed in the manual that came with it and I did not need to install any extra software. None at all! No crapware mucking up my system. I want to print and that is it. I don’t need to monitor every aspect of the printer through some software system. No, I can just go to the web page of the printer and get all the info I need. So why is this not in the manual at all? Why is this hidden from the average user?

More importantly, why is this not listed anywhere on the HP site with regards to Linux? This printer works 100% with Linux and sets up so easily. I practically had to do nothing! Nothing! It’s a shame, that it is, that such ease of use is hidden from us.

Today at the dinner table my oldest, Paige, blurts out that she cannot use Windows. Not sure I was hearing her correctly I asked her what she said/meant. She indicated that her friend down the street had a computer with Windows 7 and that she, nor her friend, could figure out how to get songs off the web (legally) or off my daughter’s media player (sansa fuze) and onto her friend’s system and media player (not sure what she has). So her solution? Bring it home and do it on one of the many systems here. Right now she is monopolizing my System76 running Arch, her desktop of choice: Gnome. She also uses a eeePC 701 running Mint.

My daughter is obviously very familiar with computers. She has used many different flavors of Linux and many different desktops including KDE, Gnome, Fluxbox and XFCE. She seamlessly goes from her netbook to my laptop to the mythtv box doing everything she wants to do. When she has question she cannot figure out she just asks. Rarely does she ask more than once and more often than not she figures things out on her own.

What does this say to me? Windows 7 obviously was not made for her. Even more, I throw this in the face to all those who say Linux is not for the masses. Bull crap. It’s not that Linux is difficult, it is that it is more often than not different. Windows 7 is not at all difficult it is just different, and given time I am sure my daughter would figure out what to do or just ask for help.

That is another thing that really differentiates adults from children. Children are more willing to seek help before giving up. Adults, they just thrown their hands in the air and give up.

I do hope my daughter does figure out how to do what she wants to do in Windows, I think it would be valuable for her to have experience across all different operating systems. Right now I am just proud that she is using, loving, and very comfortable with GNU/Linux.

At work we are struggling with how to make pages more accessible to users with impaired vision, particularly senior citizens who may find the text too small on the page. Coding a fluid layout is the proper solution but more often than not, no matter how fluid you design the layout there is always potential for elements to fall apart. Top navigation bars are a great example.

Now there is this really, really, really great and easy way to solve a lot of these issues and retain layout, and that is through the browser’s zoom page feature. In all the browsers that I have used zooming the page in and out using ctrl – + or ctrl – - works like a charm. The only problem is that most of the people who would benefit from this probably are unaware of this feature.

So what seems like a perfect solution would be able to make these controls more accessible to the user, have them on the page where they would most likely benefit. A simple + and – for zooming the page, controlled through javascript. Sounds great, right? Well you can do this with IE but not Firefox, Chrome or Safari. Their reasoning is sound though, as they do not want to allow access to the users personal settings via javascript controls. Doing so opens the door to potentially serious abuse where designers are deciding for users how their browsers should be set. I agree whole heartedly with this but damn, it would be a fantastic feature to have full page zoom controls right there on the page.

So now we are back to having to figure out through css how we can control the sizes of the elements on the page in a manner that does not break the layout.

Any thoughts, suggestions, I would love to hear them.

Dave Yates had brought up on TLLTS and then later on LottaLinuxLinks Oggcast the possibility of Microsoft buying Novell now that the latter company is up for sale. He felt, and these are my own words revisiting his, that this would place Microsoft in an interesting position to own the Unix Copyrights and to have their own Linux distrobution. Then he wondered what the downsides for Microsoft might be; and so do I.

Then a thought occurred to me. If Microsoft did purchas Novell and then continue to distribute SuSE would that not invalidate their patent claims against other companies for using/selling GNU/Linux? Wouldn’t the GPL prevent them from distibuting SuSE while they hold patent suits against other companies for the same GPL’s software? I know more previsions were put in regarding this for version 3, but I wonder if even version 2 would uphold their inability to not file patent suite against companies using Linux and force an end to their FUD.

If that is the case, I don’t see how Microsoft would benefit from buying Novell unless they shut SuSE down and flex the Unix Copyrights. But then again, if they did this, I suspect there would be a lot of pressure on them from the likes of Google and IBM. While they may continue to have a strangle hold on the desktop market, I think there is a switch coming to more mobile devices like smartphones and tablets that will eventually topple the Microsoft Monopoly on computing devices and render their flexing of FUD less effectual.

Just some thoughts.