Thomas Cameron from Red Hat talks about Spacewalk although he slides refer to Red Hat Satellite which is the downstream project:
Kir Kolyshkin from the OpenVZ Project talks about Linux Containers:
Thomas Cameron from Red Hat talks about Red Hat and the Open Source Community:
Thomas Cameron from Red Hat talks about GlusterFS:
John Sullivan from the Free Software Foundation speaks on Freedom.
BTW, the upcoming cert he mentions in the talk happened recently and it was the Atheros-Based ThinkPenguin 802.11 N USB Adapter.
Lance Albertson from OSUOSL talked about Ganeti:
I keep up with Fedora releases. Fedora 19 was due for an "Alpha" release last Tuesday but they had to delay. As most everyone already knows, delays in Fedora are to be expected. Anyway, I thought I'd check out their Alpha test builds. They actually seem to be working quite well. I did several installs from the "Desktop" media which is GNOME 3.8.x-based. The installs I did were inside of KVM-based virtual machines. Then I added on all of the other desktop environments and tried them out. Even in this early stage, it seems to be quite usable and stable. I obviously did NOT run into any of the "blocker" bugs that were the cause of the Alpha release delay. I think most of those were EFI related.
Remixing from Alpha
I thought I'd try doing a MontanaLinux remix from the development repositories. For those not familiar with MontanaLinux, it is basically the vast majority of desktop environments and desktop managers and a lot of desktop software rolled into a 2+ GB live .iso. It includes packages from rpmfusion (codecs, gnome-mplayer, vlc, etc), Google (google-chrome-stable and google-talkplugin) and Adobe (flash-plugin).
So from my Fedora 19 pre-alpha VM, I installed the various kickstart packages, extracted out the KDE related kickstart (.ks) files, and then melded them into a single file, added the packages I wanted to the %packages section, and then did a tiny bit of customization in the %post and %post --nochroot sections of the kickstart. With a MontanaLinux-F19.ks file done, I proceeded with the build.
It built. I discovered (with help from nirik in #fedora-devel on freenode) that Fedora no longer looks at /etc/sysconfig/desktop for the default desktop environment and display manager. That is done with systemd's systemctl.
The Bug Fairy Always Visits
There are a few glitches here and there but that is to be expected. lightdm was messing me up... so I had to add an exclude in the %packages section. I'm still in the process of refining the kickstart but it seems to work well enough.
For some reason, when I boot the .iso in a KVM VM on a Fedora 18 host I can't use the combination of SPICE/QXL. If I change it to VNC/any, SPICE/VGA, or SPICE/VMVGA it works fine.
I made sure to add in the GNOME 3.8.x Classic extensions so the GNOME Classic mode shows up in the display manager Session options.
I've only been working on this for a few hours so I'm sure I've got a bit of learning left to do. Fedora has since released a number of Fedora 19 updates and I haven't tried those yet. More later.
Update: April 22 - There was an additional flood of updates over the weekend. I guess the current build from Fedora went gold for the Alpha release and they had stockpiled some updates until after. I put in a few more hours on my configs and got the 32 and 64-bit versions built. They are working well and the SPICE/QXL combination now works fine. Updates included KDE 4.10.2 and MATE 1.60 among others.
If anyone wants to try it out, feel free to email me (email@example.com) and I'll reply with a URL.
Update: April 24 - Fedora DID release Fedora 19 Alpha yesterday. I've done yet another rebuild and created a screencast video (no audio) showing the Live DVD iso booting inside of a KVM Virtual Machine connected to with the SPICE remoting protocol. I didn't do any fancy editing of the video so there are long boring parts where you stare at a blank screen as it boots or loads. In the 25 minute video I boot, do an install, reboot and then do a quick survey of all the desktop environments, some apps as well as how I like to personalize KDE. The pre-release Fedora 19 base has a debug kernel and I'm sure the installer is doing a lot of extra logging... so the installer and the boot is a lot slower than the final product will be but that is to be expected. Everything seems to be working nicely except for Cinnamon.
Anyone who would like to watch it can do so with the link below. Right-click to download or play in your webm compatible browser. It is about 66MB in size... which is about 3 times the size of my first hard drive back in 1986. :)
Want to know what's going on in kernel development? Then check out this video from The Linux Foundation's Collaboration Summit 2013: Linux Weather Forecast By Jon Corbet
Here's a playlist for all of the videos from that conference.
Canonical has announced quite a few things over the past couple of days, weeks and months. Many of the announcements have been quite exciting in a good way (Ubuntu Phone and Ubuntu Tablet) and some of them seem to be a little shocking... that have some in the Ubuntu Community feeling betrayed, ignored or worse.
Just to review, I've not really been an Ubuntu fan. I'm a Red Hat and Fedora fanboi. I've often been critical of Canonical although not really of the volunteer community that supports Ubuntu. You know the same old stuff about how Canonical doesn't work with upstream, they don't contribute back much, most of the work that is outwardly visible is on their proprietary stuff... they seem to get way more credit than they deserve... and they still, so far as I know, haven't figured out a way to be profitable... which I think is very important for something so many people depend on. You've heard all of that before many times from many people. Nothing new here.
A tiny bit of history - One of the things in the beginning, in my opinion, that got Canonical and Ubuntu so popular so fast was that there were a lot of end users of Red Hat Linux that were upset with a few things. The first was that Red Hat started a pay support service for Red Hat Linux where users would pay $5/month ($60 a year) for faster download speeds of updates and isos. Then Red Hat created the pay-only Red Hat Enterprise Linux... and seemed to put way less effort into Red Hat Linux 8 and 9. And of course they said flat out that they didn't think that Linux on the desktop was a viable / profitable option and they were going to put all of their efforts into Linux for servers. It took a while for the Fedora Project to be born and to actually get to a point where they were something that resembled a real community project rather than this awkward thing that Red Hat did to appease the mobs. As us Fedora folks know, the Fedora Project some time ago got to the point where it was on par with Debian (or pretty close to it) with regards to being self sustainable and having a nice set of ethics they operate by. Granted Fedora doesn't support anywhere near as many architectures as Debian does, but you get my point. Fedora (and Red Hat) do a lot of stuff and it's all based on free (as in speech) software. Anyway, I don't want to get too far off on a Fedora tangent... because I don't have much new information to offer.
My point is (I think that) Mark Shuttleworth saw the turmoil in the Red Hat community and as a result he tried to capitalize on it by saying early on that Ubuntu would always be free (as in beer) and that they were going to concentrate on Desktop Linux.
Since I've been through some turmoil with Red Hat and Fedora... it pains me to see the Ubuntu Community in the situation it is in now. The advice I'd like to convey is... relax... don't jump to conclusions... don't let your feelings get the best of you... be logical... keep doing what you've been doing... and as time passes... a lot of the confusion caused by uncertainty will clear up... and things will get way better... and you'll be happy again.
Facts for Ubuntu Developers (a different FUD) - I can understand that the non-Unity spins of Ubuntu are scared about Mir... but how is that different from the turmoil a switch to Wayland would have caused anyway? Regarding the Rolling Release move, Mr. Shuttleworth seems to be against the idea after all, so do you think that is going to happen?
Yes, Canonical is moving in some new and different directions and their vision doesn't seem to match as well as it has in the past with much of the volunteer Ubuntu community... but so what? As a community you can still do what you want to. The vast majority of the software is FLOSS and you can continue to do with it as you wish. You may have to muster more resources that were previously provided by Canonical... so you may have to work harder to move in your own direction... but don't worry... it'll be worth it so hang in there. Don't quit. Don't give up.
I could go on and on with specific examples but I think I'd only bore people and hopefully I've gotten my point across already.
Motivation, Smotivation - Why am I being supportive of Canonical? Well, I'm not really. I'm just trying to be supportive of the volunteer Ubuntu community. Ok, maybe I am trying to be a little supportive of Canonical. Being a Fedora fanboi why would I want to do that? The answer is simple really. I think there is a big enough pie for a dozen Linux and FLOSS companies. Why should Red Hat remain the beacon of success... that seems to prove to be the exception to the rule rather than the rule. Red Hat didn't want the desktop market. They made that very clear... and they picked the server market and have executed and delivered quite well in that space. Who else is trying to be a commercial success in the Linux desktop and mobile space? Do we really want a Google Everything future? Do we want Android to be the "future of Linux"? If you didn't already guess my answers to those questions... it is a strong NO. I've been hoping that Canonical would find some way to make a good profit in an ethical and community friendly way as yet another example of business success with FLOSS... and maybe they'd spark some interest in Red Hat to move into the Desktop and mobile market.
I kind of think Mr. Shuttleworth handicapped himself with the "it'll always be free" comment at the beginning. I mean... ok, one or more forms can remain free but can't they also come up with some way to make a pay version too? That would be a more direct way to be profitable rather than trying to gain a massive userbase where only a small percentage of users are paying for cloud services or purchasing things where Canonical gets a small cut of the revenue... but who am I to question a millionaire about the best ways to make money?
Worst case scenerio... the bulk of the people that are the Ubuntu community now... fork off and become a renamed community... more able to focus on the goals they think are important... without needing or wanting the approval of Canonical nor Mark Shuttleworth. Would that be a lot of work? Heck yeah. Are they anywhere near that point yet? Not even close. Ideally I envision a sort of relationship similar to what Red Hat has with Fedora... between Canonical and a refocused Ubuntu Community. It depends on what Canonical really thinks about their community. I know what they say in public about them... but I don't think that is necessarily what they 100% believe.
In Conclusion - So, my advice for now... to the Ubuntu community folks is... just relax... don't get overly excited... keep doing what you have enjoyed doing... and let some time pass... and it'll probably just get all better by itself. So, boiling that down to two words, "Don't panic!". Even if your worst fears came true, and I don't think they will, you have some positive, viable paths of action.
OpenVZ Booth - Sunday was a much slower day. Activity in the larger exhibit room seemed to be much higher than it was in the Pacific room where the OpenVZ booth was. I spent quite a bit of time twittling my thumbs.
Daniel Robbins stopped by to see Kir again but he was temporarily away from the booth. When he got back I sent him over to talk to Daniel.
I talked to a long, blond-haired Google employee for an extended length of time... about fingernails. He had very long fingernails (for a dude) that kind of looked like claws so I asked if he played the guitar. He told me quite a bit about fleet programming and mapping memory to the disk... and mentioned that Google uses Ganeti and Xen for internal virtualization stuff... but not for running their external facing services. I learned more about fingernails than I thought possible.
Two guys came up to the booth to ask if they could borrow my terminal window for a minute. I asked why and one said he wanted to show the other GNU Screen for the first time. I told them not to use screen but to use tmux instead... and then broke into an impromptu demo of tmux. They were amazed. Then a third friend came up who said he had been using screen for years but as a serial communications program (after having previously used minicom)... and that he had no idea that screen had the terminal multiplexing features that 99.9% of other screen users use it for. Showed him tmux too. I referred them to a UTOSC 2012 video on Screen vs. Tmux. They were very happy to learn it and thought it was awesome that it came from an OpenVZ booth and wasn't about OpenVZ. :)
There were about a dozen people who stopped by the booth to ask about OpenVZ. One guy said he had been to Bozeman before because he wife was on a trip skiing in the Bozeman area and broken her leg so they ended up visiting the Bozeman hospital.
Other Booths - I visited the Fedora booth and got the Multi-Desktop Live DVDs for Fedora 17 and 18 along with some stickers. I also got some media from the OpenSUSE booth. I really enjoyed several of the hardware related booths. For example there was a company there named ODROID Hard-Kernel that had a number of tiny Linux machines. The Tiny-Core booth had a few tiny machines as well. The server hardware guys were there with HP, ServerMicro and ServersDirect among others. Inkscape, KDE, LibreOffice and VideoLAN were represented. Several Linux distros had booths including Debian, Gentoo, Arch, Ubuntu, Fedora, OpenSUSE, and Tiny-Core. Did I leave anyone out? Cloud providers were there in numbers... as were storage solutions, databases, and backup solutions. The usual third-parties were there like the FSF, EFF, FreeBSD / PC-BSD, OpenBSD, OLPC, and Haiku. Linux box makers were there like System76 and Zareason. Various area user groups had booths. There were close to 100 booths in total. I believe I got pictures of everything.
OpenShift Presentation - I went to a presentation at 3PM entitled, Build Your Own PaaS using OpenShift Origin given by Red Hat's Steven Citron-Pousty (ODP slides). He was giving away 4GB USB thumb drives that were stainless steel with a bottle opener to those who asked good questions. I asked a good question but I didn't want the swag. OpenShift is available as an online hosted service (public cloud on Amazon EC2 known as OpenShift Online), as a pay service from Red Hat known as OpenShift Enterprise, or as a do-it-yourself upstream project known as OpenShift Origin. The only thing OpenShift requires is either Red Hat Enteprise Linux 6 (or clone) or Fedora. I learned that OpenShift uses SELinux and cgroups to make service gears which are functionally very similar to containers. OpenShift is an elaborately designed system whereby a wide variety of components can be mixed and matched quickly and easily to stand up almost any web-based service. Some of those services include databases, middle-ware, development languages, frameworks, developer tools, and some packaged web applications. Each category is a who's who of big name open source projects. OpenShift grabs what you want it to, creates a gear of out of it, and then auto-magically configures everything with a unique private IP address and uses HA-Proxy to tie it all together. Other, non-web-based, services are on the road map but do not currently have a target date. OpenShift looks very interesting for hosting companies or any mid-to-large company that is constantly deploying a number of new web-services. The fact that they have tied together SELinux and cgroups to make psydo-containers is very interesting. Someone in the audience asked if they considered OpenVZ and Steve said he didn't think so because he was unfamiliar with it. In summary with OpenShift, System Administrators can easily deploy anything that developers want and developers can do what they need to do and very little more.
Post Show - After the show I just went back to the room and watched TV. The Oscars came on. They come on much earlier in L.A. They ended at 8PM. I watched more TV (while uploading pictures I took from the show). Then the local news came on. It had the normal news stuff except they also included who was attending whos after oscars party. Being in L.A. on Oscars night is kind of cool.
Conclusion - This was my first visit to SCaLE and I must say I was impressed. The Hilton was a very fancy place to have it and the accommodations for the presentations were reasonable. SCaLE definitely was not the only stuff the Hilton had going on as I saw a number of other events in a few of the other conference rooms.
Wnen the show was over there was a team of about 10 people from SCaLE breaking down all of the network stuff. This was their 11th year and it went like clockwork. There was a significant amount of Linux luminaries and a wide variety of talks that ranged from beginner to kernel development and everything in-between. There were close to 100 exhibit booths and tons of swag. The exhibit floor was packed for much of Saturday but quite a bit thinner on Sunday. The exhibitors were made up of a good mixture of commercial venders as well as .org projects. While I was stuck in a booth for most of the show I did get to attend a few presentations. Supposedly most of the presentations were video recorded and I hope to see them posted online ASAP so I can check them out. I'd definitely attend a future SCaLE if given the opportunity.