Alahup!

There seems to be a new, interesting CMS in the works: Alahup!. As Ölbaum said:

Must keep an eye on this upcoming CMS.

Most definitely, yes.

Bloglines in PulpFiction Newsreader

Update: I made a little PHP script that lets you upload your subscriptions and will return the converted file. It can be found here.

bloglines is (maybe was, after they had been bought by AskJeeves) a great service if you happen to read your subscriptions at two or more places. Newsreaders get support for their service that allows one to synchronize all your feeds via Bloglines.

Unfortunately, my favourite Newsreader PulpFiction does not yet support bloglines. But while having a look at the services description on bloglines, it occured to me that adding support for bloglines is actually fairly easy: The only thing one needs is an OPML file that lists the bloglines getitems URL as the feed URL instead of the original feed’s URL.

So what I did is write an XSL file that will take an OPML file from http://rpc.bloglines.com/listsubs and convert it to an OPML file with bloglines feeds.

You can then load this new OPML file into PulpFiction — et voilà — Bloglines wherever you want it.

This approach, of course, applies to almost any Newsreader that can import OPML files.

Use Instructions

What follows is a step-by-step instruction to get this going. However, I assume that you are familiar with the command line and that you have a running LibXSL, including the libxslt executable, on your system.

1. Download the subscriptions from bloglines and save them as bloglines.opml. You find them at http://rpc.bloglines.com/listsubs and you’ll have to authenticate with your bloglines login credentials.

2. Download the XSL file: bloglines2opml.xsl

3. On the shell, run the following command:

   xsltproc -O bloglines-sub.opml bloglines2opml.xsl bloglines.opml

4. Import the newly created file (bloglines-sub.opml) into PulpFiction (or any other Newsreader that supports OPML)

5. Enjoy

Update: You can alternatively skip steps 2 and 3 and use the online script available here.

Caveats

• Because access to subscriptions is based on HTTP authentication, your Newsreader must support feed authentication. PulpFiction does so, as well as NewsFire.

• In the current form, all unread items per feed will be downloaded and marked as read in bloglines. If this is not what you want, change line 33 of the XSL from this:

  <xsl :value-of select="concat('http://rpc.bloglines.com/getitems?n=1&s=', ./@BloglinesSubId )"/>

  to this:

  <xsl :value-of select="concat('http://rpc.bloglines.com/getitems?n=0&s=', ./@BloglinesSubId )"/>

(The parameter n in the xmlUrl defines if downloaded items are marked as read in bloglines or not. Default is 1 in my XSL).

Knowledge Mining

One of the things that always grabs my interest is all kinds of knowledge tools:

• Desktop Search Engines
• Knowledge Tools like
  • Flow
  • DEVONthink
  • StickyBrain
• WiKi engines
• Outliners, MindMaps and other “List” tools
• … and many more

All these products are fine, fulfil their intended use, even share common traits, but everyone also has its limits:

Desktop Search Engines

Desktop Search Engines simply deliver search results from their indices, not less but not more as well. They help me find content fast, but I must know what I search for — or what exactly I need.

Knowledge Tools

All the Knowledge Tools we have at our fingertips are cool, some have better design patterns than others, but that’s not the point here. But every one of these also requires me to change my habits and workflows when I want to efficiently use them: * I must start to re-index all my mail conversation, if not import all mails from time to time * I must manually add every single document I want to mine for knowledge * I basically must make the software my digital hub

Wikis

Wikis, then, are a great tool, too. But here, adding external resources is mostly a clumsy process: Of course, I can link PDF and Word files (or XML, if you want :)), but the content is simply linked, no additional knowledge is taken from the relationships I define by adding a link to a certain file to a page about a certain subject.

That is, they provide very intelligent and intuitive hyperlinking, but not before I told them to do so.

Outliners

Last but not least, we have all these great outlining apps out there, bringing some kind of structure to information, helping people better achieve their goals by presenting lists of things. It’s no wonder that almost every efficient productivity tool or “theory” falls in this category: Mind Maps, Outliners, GTD, Task lists …

… but

Nevertheless, there’s little to no additional knowledge leveraged in all these tools. They know what you tell them, show them to index or manually link to them. But somehow, every single “productivity” tool out there in the world is, well, an autist tool: Completely immersed in a world of their own, powerful in but limited to a very specific subject (Search, outline, manage data) or simply too limited to find interesting knowledge themselves.

… there’s hope …

A rather big company, called Autonomy has a very interesting product called Active Windows Extensions:

Active Windows Extensions (AWE) automatically links users with relevant information they require, accurately, in context and in real time without the user being needlessly diverted from their work in progress to perform a search or retrieval operation.

The solution adds unobtrusive icons to the title bar of Windows applications. As you write a document, create a presentation, surf the web or write e-mail messages, AWE automatically conducts a real-time analysis of the ideas involved and suggests relevant documents, which you can view by clicking on the icons.

As a user writes a document, creates a presentation or writes e-mail messages, Active Windows Extensions conducts a realtime analysis of the ideas involved and provides links to relevant internal and external information, automatically.

Unobtrusive Knowledge

This is tool certainly handy in a corporate environment, right, but you could equally apply this to a personal workstation: Instead of conducting searches in any of the apps I mentioned above, you simply go and get the relevant content, without even searching …

And I think, that Spotlight would be the ideal base for this:

• It builds a comprehensive, up-to-date index - no spidering is necessary
• It can be extended to include App-specific formats - I can keep working in my preferred apps
• It has the inherent knowledge of every object on my computer - no need to manually add items
• It can return structured information, enriched with metadata - no need to manually assign relationships

If someone thus applied the concept of Active Windows Extensions to a nice little Spotlight plugin, we could see an Knowledge Mining Tool that is finally worth its name. Instant, context-sensitive search results wherever and whenever I need them.

Add some spiffy visualisation layer and you can start analysing your knowledge behaviours, can create topic maps and — finally — use all these spiffy tools that exists to give form to statistics.

I’m waiting …