Purple Numbers: Optimized for Synthesis

Posted on Published by Eugene Eric Kim

Chris Dent has been having some good exchanges about Purple Numbers with Adina Levin and Phil Jones. I don’t have much to add, as I think Chris is spot on. Two comments struck me, though.    (IQV)

First, Phil claims that Purple Numbers are optimized for reading at the expense of writing. His point is that Purple Numbers, as currently implemented, add overhead to the writing process, whereas the pay-off comes for the reader. I emphasize as currently implemented, because we just haven’t gotten around to making them mostly transparent in the writing process. Hacking one of the WYSIWYG JavaScript text editors to support Purple Numbers should do the trick.    (IQW)

However, I really liked Chris’s response:    (IQX)

Yes, purple numbers do try to favor the reader and the act of reading, but not just for reading. They favor the reader so the reader may more easily do more writing. The whole point is for purple numbers and tools like it to be a generative force in the synthesis of new understandings.    (IQY)

Phil can write all he wants, and I can read all I want, but until I write down something that builds on what Phil says, while making chains of reference back through the many layers of context, there’s been no synthesis, at least not any that is available outside the confines of my own mind.    (IQZ)

Granular Addressability enables synthesis. Wanna know what makes blogs conversational? Permalinks, which are a form of Granular Addressability.    (IR0)

A lot of people don’t get this. I read The Sports Guy over at ESPN.com all the time (despite the fact that I hate all Boston sports teams with a passion), and at the past two Super Bowls, he wrote what ESPN.com called a “blog.” It sure looked like a blog, but in reality, it was just one-way publishing. Folks couldn’t comment on his entries, because they couldn’t link to any of them. I see this all the time with other major media outlets trying to jump on the blog bandwagon.    (IR1)

Which brings me to the second comment that jumped out at me. In his response to Adina, Chris wrote:    (IR2)

Clearly I am in far too deep with purple stuff: I need a translator. The above can be so much meaningless noise and I find little time to make things cogent.    (IR3)

I’m not the best proselytizer of Purple Numbers, not because I’m not proud of them (I am), not because I don’t value them (I do), and not because I can’t explain their value clearly (I can). There’s a tremendous amount of deep thinking underlying these little purple critters, and the implications are fascinating. But before you can understand any of this, you’ve got to care. And unless you’re one of those strange individuals who just gets it, you’ve got to try them before you’ll buy them.    (IR4)

A lot of folks think I invented Purple Numbers. Not true at all. I was one of those folks who didn’t care, one of the first in fact. Purple Numbers are an HTML manifestation of Augment’s granular addressability scheme, invented by Doug Engelbart and made purple years later by his daughter, Christina Engelbart. When I first started working with Doug, he kept insisting that all of our knowledge products on the Web have Purple Numbers. I didn’t think it was a priority, but I knew I could easily whip up a tool to generate them, so I wrote Purple to humor him. Then a funny thing happened. Once I had them, I used them, simply because they were there. Then I started missing them when they weren’t there. Then it dawned on me: These little purple thingies sure were darn useful. And I started thinking about why.    (IR5)

The point of my story is this: I’m perfectly happy to have a deep, convoluted discussion about some esoteric aspect of Purple Numbers. If you don’t believe me, try me. Or read my blog entries on the matter. But if you really want to understand why they’re so important, just give them a try for a month, then try living without them.    (IR6)

Stable URLs in Blosxom

Posted on Published by Eugene Eric Kim

One of the problems with blosxom is that it does not implement truly stable URLs. (Of course, one could argue that no URLs on the Web are truly stable, but that’s a topic for another day.)    (2D)

Bloxom uses the filesystem for data storage and categories. This is one of its distinguishing features. It makes it easy to set up and install, it makes it easy to edit entries (use the editor of your choice, or build your own Web entry system, like wikieditish), and you can easily preprocess entries either statically or dynamically. My blosxom_purple plugin takes advantage of this latter fact, as do many other plugins.    (2E)

Its dependence on the filesystem also has its drawbacks. It’s hard to keep the date of each entry stable, because blosxom uses the timestamp of the file for the date. If you modify the file at all, the date changes. (Plugins like entries_index fix this problem.) Multiple categories are possible via symbolic links, but this is hackish at best. And worst of all, it prevents you from having truly stable URLs.    (2F)

The blog term for such an animal is Perma Link, and the question of the moment is, what’s the proper Perma Link for blosxom? The default flavour suggests that the link should be the date and an anchor link to the entry itself (in case of multiple entries per date). This only works if you don’t modify any of your blog entries, in which case the date will change. If you rename the file, the anchor name will change.    (2G)

I use the category followed by the filename as the Perma Link, because I only wanted a single entry per Perma Link page. Here, the problem is if I recategorize the page (by moving it into a different directory), once again, the Perma Link breaks.    (2H)

These issues will not be a big deal for most folks, but it’s a big deal to me. It means that I have to be extremely careful about my categorization scheme, and I have limited flexibility in changing categories.    (2I)

I want Permalinks that are less fragile. The solution is to separate blosxom’s date and category system from the filesystem itself. This has already been done for dates, and could easily be done for categories as well. The problem is, now you’re getting away from the spirit of blosxom, one of the very features that makes it unique and desirable.    (2J)