Korora is to Fedora as Ubuntu is to Debian. Got that?
That means Korora adds on all those helpful bits that a Fedora user just might want. Everything from multimedia codecs to Steam, Adobe Flash to VirtualBox -- you get it all in Korora, though most of it isn't terribly hard to add to "virgin" Fedora.
Just like Debian: There are plenty of things that ship in Ubuntu, but the halfway knowledgeable user with a little time on his/her hands can do most if not all of it on top of Debian.
But just like with Debian and Ubuntu, it's nice to have something like Korora to give us a complete out-of-the-box experience.
The only difference between Korora and Ubuntu? Nobody's ever heard of one of them.
Never mind that. In the next cycle, Korora is upping its game. The Korora 20 Beta builds are now available, and I'm happy to see that Xfce has been added to the list of available desktop ISOs, which already included GNOME, KDE, Cinnamon and MATE.
I'm downloading the Xfce and GNOME ISOs now, though what I'm really looking for is something with a 3.13 Linux kernel so I can put it through its paces on my still-needs-help-suspending AMD-running laptop.
My Fedora system has most of what is in Korora, though not Steam (don't care), Jockey (do care and WANT it) or VirtualBox (could be worth a play). But I've thought for a long time that Fedora needs its own Ubuntu/Mint, and Korora looks to be fulfilling that role very nicely.
MLED, aka the Microlinux Enterprise Desktop, is Frenchman Nicolas Kovacs' attempt to bring together various bits and pieces of the Slackware community, including Slackbuilds, slackpkg+ (which I confess I've never heard of until now) and more to create what he calls a "full-blown production desktop."
Yes, that includes multimedia codecs.
You get MLED by installing Slackware, then importing "tagfiles" (first time I've heard of this concept) and doing more than a bit of configuration, choosing KDE, Xfce or MATE along the way.
It's not as easy as a full-blown ISO but not as hard as finding all the bits on your own.
Kovacs talks about why MLED is based on Slackware here, and I agree with pretty much everything he says.
If you're looking for a long-term-support distribution with extremely conservative underpinnings, Slackware is a compelling choice, and it looks like MLED will get you from zero to desktop that much more quickly than assembling the bits on your own. I'd prefer this to be more automatic, but those are the Slackware-fueled breaks, I guess.
That's great news since Markdown will really help those of us who use WordPress get posts formatted that much more quickly. I hate using the formatting buttons that come with WordPress, and Markdown beats hand-coding HTML any day.
(Note: This is an Ode blog, and it uses Markdown.)
Now all we need is Markdown in self-hosted WordPress.org. Then we'll be cooking with gas. The thread that announced Markdown for .com sites says it will be eventually be part of Jetpack for .org installations.
Until then, WordPress people remind that there are many Markdown plugins available.
I've been using Fedora Linux for the greater part of this year, starting with F18 and upgrading via Fedup to F19. For most of that time, I've used the closed-source AMD Catalyst driver as packaged by RPM Fusion instead of the open Radeon driver that ships by default with Fedora and most every other Linux distribution.
I'm not proud of it. But the differences in performance are too big to ignore.
Things that stink with both drivers: Neither the open- nor the closed-source driver will resume my HP Pavilion g6-2210us laptop after suspend. (The machine uses the AMD A4-4300M APU with AMD Radeon HD 7420G graphics.)
Things that stink with the open driver: Only the Catalyst driver delivers working 3D acceleration, meaning without it I can't run GNOME 3 at all, most games look like hell, and a certain wonkiness crops up here and there on various web pages.
With Catalyst, my glxgears frames per second are 100 times greater than with the open driver. I don't know what glxgears fps numbers really mean, but 5,200 has got to be better than 50.
Things that stink with the closed driver: In Xfce, many application windows have lost the borders on the left and right sides. I can't explain it.
I also cannot successfuly use UEFI secure boot with the Catalyst-enabled kernel, though I can do so without Catalyst installed. It's not Secure Boot itself that is stopping the boot. It just hangs at some point -- after some IP tables lines in the dmesg, I think. The solution is keeping EFI but turning off Secure Boot.
You'd think the solution would be easy and ubiquitous. Here's what I wanted to do: My personal blog run with the Perl-based Ode system. Ode doesn't use a database. Instead it stores its entries as text files in "normal" directories on the server.
I wanted to have exact copies of everything in my Ode documents directory on my local computer and the server. And I wanted the freedom to add to or modify anything in this directory on either side (server or laptop) and have everything track on both machines.
Many of us use Dropbox (or Box, or SpiderOak, or Google Drive, or ...) to both back up some or all of our files and mirror them on other desktops and laptops we happen to use.
But what if you want to keep a filesystem in sync across any number of servers and desktops and laptops without using a third-party service?
My first thought was, "I'll just use Dropbox. Certainly there must be a way to use Dropbox on my server/VPS/shared-hosting. Nope. No. It doesn't work that way.
My second thought was, "Holy shit, Dropbox is missing out on a whole lot of revenue and screwing its users besides."
For one reason or another, I've been thinking about Movable Type. I went to both of the web sites associated with the blogging software -- movabletype.org and movabletype.com and found no mention of the formerly "free," open-source Movable Type software I used for so many years.
Instead, MT 6 is for up to five users and $1,195 for unlimited users. Ouch. There's quite a gap between ode.cgi and $1,195.
Nowhere on those "official" sites could I find a link to the ode.cgi versions of Movable Type (i.e. everything up to Version 5).
But if you want MT 4.x or 5.x, they are available.
And the software that swallowed Movable Type's user base whole is still available -- and still free.
Movable Type was always a great platform, and it still handles multiple blogs and multiple users better than WordPress in my opinion.
But you really need a full-time hacker on the job if you want to use Movable Type seriously. There never was enough of a community out there with plugins and themes to get you going.
I am now using my updated, responsive CSS at the non-blog root of my domain, stevenrosenberg.net.
I also added to the list of blogs and social networks. (I have too many in both categories.)
Later: The page isn't terribly responsive on mobile. That's a project for the future.
The next day: Problem fixed. Per Hans Fast, I added this code to my HTML:
<meta name="viewport" content="width=device-width" />
I didn't expect the post-WordPress blogging system Ghost to ship with all of its promised features, but it's more basic than I thought it would be. (If you want to read this very entry on my Ghost blog, here it is.)
It's basically entries tagged with Markdown and presented on the page.
As far as I know there are no categories or tags (though I do see them on other Ghost sites), and none of the promised back-end stats. There is no easily-implemented provision for comments,
not even though you can hack in Disqus. Clearly this sort of thing needs to get easier if Ghost has any hope of going beyond the geeky contingent that champions such systems as OctoPress, Pelican and Nikola.
It looks like there is only one user (and one blog) per installation.
In short, while the code that is out now does use Node.js and does use a two-windowed Markdown-on-one-side, styled-text-on-the-other composition screen, and what you write in there appears on your life site in the form of blog entries, that's pretty much it.
So I give the Ghost team this: They have code out in the wild, and it does work. Now they have to build on it and start delivering the features promised on the main Ghost site.
I hope they get there.
And unless you, like me, use a Node.js-friendly service like OpenShift to host your Ghost (I'm sure the AWS Elastic Beanstalk would do just as well) or have access to (or can spin up) a Node-running server and care deeply about running your blog on Node.js as opposed to PHP or Perl (or Ruby or Python for that matter), I'm not yet ready to recommend Ghost just yet. (Note: Ghost's documentation tells of many other ways to run it.)
For me, Ode creator Rob Reed's "Ode means you know how it works" credo is keeping me firmly in the Ode camp. Sure Perl is "old." (Just like PHP, which powers WordPress and Drupal and probably most other Web services.) But Rob has put a lot of thought into the design and subsequent execution of Ode. I'd love to see the Ghost team follow his example and create a system maintainable and hackable by the average human. If you look at a Ghost composition window and Ode's EditEdit side-by-side, you'll find more alike than different.
But at the end of the day, there's more to any blogging/publishing system than the language used on the back end, and Ghost will have to sell itself with features and ease of use, not the tools used to bolt it together.
Later: The Ghost Forums are essential for getting the most out of Ghost.
Aaron Toponce is one of those insightful writers about Linux that I like to follow.
Now he joins those publicly leaving the Ubuntu project after what he refers to as a long line of disappointments in the project and its parent company Canonical, the last of those being the "trademark aggression" exhibited over the Fix Ubuntu site, the heavy-handedness for which SABDFL Mark Shuttleworth has apologized.
SABDFL apology aside, Aaron states many reason for leaving Ubuntu as a contributor and user (he's running Debian on everything, if you want to know). Those reasons include swapping GNOME for Unity, the Unity Lenses and the Amazon shopping "app."
He ends (but please do read the entire post):
I can't be associated with a project like this any longer. Effective immediately, my blog will no longer on the Ubuntu Planet. My Ubuntu Membership will be cancelled. My "UBUNTU" license plates, which have been on my car since August 2006, will be removed, in favor of my Amateur Radio callsign. I wish everyone in the Ubuntu community the best of wishes. I also hope you have the power to change Ubuntu back to what it used to be. I have no ill feelings towards any person in the Ubuntu community. I just wish to now distance myself from Ubuntu, and no longer be associated with the project. Canonical's goals and visions do not align with something I think should be a Unix. Don't worry though -- I'll keep blogging. You can't get that out of my blood. Ubuntu just isn't for me any longer. Goodbye Ubuntu.
I found Aaron's post via Benjamin Kerensa's post on the need to establish a Ubuntu foundation. The idea is intriguing, but I doubt anything will come of it.
As I've been saying lately, there are a few hundred other Linux distributions out there, and even close to home there are a number of fine Ubuntu-affiliated/derived projects like Xubuntu, Ubuntu GNOME, Kubuntu and Lubuntu that offer compelling desktop systems and are run by engaged, growing and inclusive communities. And there's always Mint, Debian, CrunchBang, Slackware and many, many more.
More for technical than philosophical reasons, I'm running Fedora with Xfce. Until my hardware runs better (i.e. suspend/resume works), I need the latest kernels and video drivers, and Fedora offers (in my experience anyway) the easiest, least painful way of getting them. And while Fedora also has a strong corporate parent/overlord in Red Hat, the relationship between company and community is much less frought.
I don't get on Google+ all that often. But I was going through my mail today and got a notice about a post from a few weeks ago. I went to it and made a comment.
Further down in my G+-related mail, Google offered me this personalized URL: https://plus.google.com/+StevenRosenberg. In true land-grab fashion, I took it.
Will this make me more likely to use Google+? It certainly won't make me less likely to do so.