Rubenerd

This post originally appeared on the Annexe, which dave sunsetted in 2016.

As I eluded to in a previous post, I’m in the process of moving my self-hosted WordPress blog over to Jekyll. It has over 5,000+ posts written since 2004. It’s been a huge pain, but I’ve got all of them exported and in the Markdown format Jekyll expects. Now I’m just in the process of writing the Liquid themes and RSS feed templates.

And then I read Dave Winer’s posts about Fargo, and tried it out.

Now I find myself writing here, instead of continuing my old blog migration. After a brief session riddled with mistakes, I already find it effortless. It’s an outline, I have a dinousaur looking at me.

Jekyll and Fargo are simple, but couldn’t be more different. I’ll be interested to see how this battle between them plays out.

One thing is for sure: I’m going to use (I hate the word “try”) this for posts on university assignments and work, and see how it goes. I have a month to get my Fargo blog looking spiffy before next semester starts! ☺

This originally appeared on PenguinCoffee, Clara’s and my old shared weblog.

By 青冥童子@お仕事募集中.

This originally appeared on PenguinCoffee, Clara’s and my old shared weblog.

Saber in her legendary, epic purple mizugi! At least, I t-think it’s Saber!

By マルサ on Pixiv.

This post originally appeared on the Annexe, when it was hosted on Fargo.

Part of the reason I’m thankful for Fargo is I’m overhauling my primary blog. Rubenerd.com has been operating since 2005 on WordPress, but lately I’ve grown tired of maintaining all the server side architecture on my little VPS to run it, so I’m moving to Jekyll.

There are few things that frustrate blog readers about blog migrations than the endless posts about how they’re doing it, so I figured I’d explore some of it here. Who knows, perhaps my Fargo blog turns into a metablog about blogs!

Firstly, Jekyll has given me the chance to really think about the DOCTYPE I’m using. I’ve been comfortably on XHTML1.1 for years, but all the cool people my age are on HTML5, so I thought I’d finally give it a try.

Having spent years writing well formed, valid XML, HTML5 is a little scary. Lots of unclosed tags, lack of XML extensions, a spec that keeps changing so often the W3C validator spits out different results each time you use it. Say what you will about specs, but I can finally appreciate why Dave froze RSS 2.0.

I’ve always been fascinated by metadata, so I figured for a static text-based blog I could take the alternative route and try XHTML1.1+RDFa. It’s a W3C spec, and though its not the microdata the cool people are using with HTML5, RDF and I have always gelled (I’ll admit, I probably still prefer the so-called RSS1.0 for this selfish reason).

The problem with this approach is, again, the W3C Validator. While the DTD clearly specifies the use of ARIA “role” (something critical if I’m not going to have HTML5 elements like “nav” and “header”), the validator tersely tells me my sandbox XHTML1.1+RDFa markup is invalid because there is no “role” element. Yikes.

So, I’m back at square one. I really want to provide accessibility, and I really was looking forward to getting stuck into RDFa with my new site design. Oh well.

This post originally appeared on the Annexe, when it was hosted on Fargo.

(A big thanks to everyone in the Fargo community for their comments. If I had a Disqus account, I’d thank you! I’ll try out your suggestions. ☺)

As far as I can tell, no way to disable Disqus comments. For reasons I won’t get into here, I prefer not to use them.

My custom CSS doesn’t seem to be coming through.

No spell checking. I make so many typos this is bound to be an issue.

Each panel in the Settings window has to be updated seperately. If I change something in “Font” then click “Advanced” without clicking OK first, my changes are deleted.

The Fargo outline editor doesn’t respect the [Shift]-[Down] combination to select an entire line, like most contemporary editors.

It’s US centric, or at least the timestamps are. I’m in Sydney, Australia, and these dates are wildly wrong. Wonder if there’s a way to define a +1000 timezone?

The title for my blog here seems to have vanished.

Those issues aside, usability wise I’m liking this a lot. I’m not going to give up my primary self-hosted blog, but this could be used for any number of side projects I’m working on. And I’ve always had a thing for outlines.

Being able to hit [Command]-[R] and [Command]-[L] to indent or move out an outline point is nice.

+1 for RSS 2.0 (It’s Dave, so what would you expect?). The less I have to deal with Atom, the better.

This post originally appeared on the Annexe, when it was hosted on Fargo.

Wow, here I am, and I even got this cool vanity URL! This is too cool. Pity about the Disqus comments, wonder if they can be turned off?

A small item of note as I updated Homebrew yesterday:

% brew update
[..]
=> HEAD is now at 156ce4d4 Add Haswell to CPU tests

Included for your convenience is an Intel press slide about the Haswell CPU. Haswell that ends well.

While unsubscribing from an InfoWorld mailing list, I was presented with this rather fabulous 2-bit graphic above the final unsubscribe button. So many 1990s memories, flooding back…!

For what it’s worth though, the original GIF was only 4KiB, and I converted it to a 1KiB PNG. You could convey a lot with so little data back then.

UPDATE: This issue has since been resolved.

As of last night, something is wrong with the W3C’s HTML5 validator and RDFa prefix syntax (maybe). Previously validating pages that employ either the html prefix tag, or Dublin Core schemas no longer validate.

Broken HTML5 prefix

This is how I currently employ RDFa Dublin Core metadata properties on my pages (and will be on Rubenerd.com soon):

<html prefix="dc: http://purl.org/dc/elements/1.1/">
[..]
<meta property="dc:date" content="[some timestamp]" />

Doing so now generates this error:

“Element meta is missing required attribute name”

Attributes for element meta: name http-equiv content charset

The “property” attribute is now no longer even listed. Huh.

Broken HTML5 schemas

If you replace RDFa syntax with a schema link (such as to Dublin Core), and replace the meta properties with meta names:

<link rel="schema.DC" href="http://purl.org/dc/elements/1.1/" />
<link rel="schema.DCTERMS" href="http://purl.org/dc/terms/" />
[..]
<meta name="DC.date" content="[some timestamp]" />

You get this error:

“Bad value schema.dc for attribute rel on element link: The string schema.dc is not a registered keyword or absolute URL.”

I am unsure how to proceed. If this is an error in the evolving, beta HTML5 validatior, I’ll understand. If instead they’ve decided RDFa-like syntax can’t tango with HTML5 after all, I won’t be a very happy munchkin.

Clara Tse over at Kirinyan.net, saying succinctly what we all think at times:

They control what happens and we can only follow along and take what we are fed, even if we don’t like it [..]