Purple Placement
June 01, 2004
In Purple Pilcrows Tim Bray explains how he's adapted his purpling of posts to use an exposed on mouseover Pilcrow which merges several ideas from this weekend's purple breakthrough. It's great how that works. (854)
I like the pilcrow and the mouseover trick. When developing PurpleWiki and related tools one of the reasons we didn't do any hiding of the link was so that people would have the cognitive breakdown that makes them wonder what's going on and do a bit of investigating. For a first time visitor to Tim's posting there's no way of knowing the pilcrows are there without a mouseover. Once you are in the know, though, you're good to go. (855)
This is the delicate balance of present-to-hand and ready-to-hand in tool design. (856)
While I like Tim's building on shoulders solution very much I do have one quibble: (857)
Also, as several others have noted, it’s not obvious that you need an anchor on every paragraph. (tb) (858)
If the author of a document is in charge of where the anchors go, then the author of the document is in charge of what a reader can reference (and later, perhaps, transclude). This sets a precedent for authorial, uh, authority that I find distasteful. The linkability of content is a way of opening up the system to allow greater review, commentary and reuse. Tim's solution is reasonable (adding the links when there are sufficient granules in the document to warrant granular addressing) but leaves room for a later adopter to be less open. (859)
This is why comments about granular linking reeking of ego struck me as way off base. PurpleNumbers and similar systems aren't there for the author. They are there for the reader, the responder, the critic. (85A)
The mouseover solution makes it possible to id everything with abandon, so why not? (85B)
Meanwhile, I'm going to keep mine as visible numbers because I like to transclude (the T marks the transclusion and links to the original source): (85C)
Comments
You know, after reading the "comments about granular linking reeking of ego" link, I just might be inspired enough to starting purpling things. (85G)
Dammit. (85H)
Chris, you use transclusion on your web site backend and on wikis, right? (85J)
Would you agree that for quoting across different web sites, transclusion cannot replace blockquote? (85K)
I have to say, I find the purple numbers themselves incredibly distracting (from a visual standpoint). However, I *love* the idea of granular linking... (86X)
I like Tim's simplification with the Pilcrow, but that makes all of the anchors the same, and I think there is value to a unique reference number, although I think the number scheme for Purple Numbers leaves a lot to be desired. I guess the human reader in me would like a reference that I can glance at and know what it means... a common citation format of some kind. But I know I'm dreamin' here for the time being. Unless, I'm missing something about Purple, which could be... it's late, I'm sleep deprived. (86Y)
-Dave! (86Z)
Dave! Long time no talk. How's the big city? (87I)
There's a long history of debate over why the numbers used to create the links ought to be meaningless. The summary is that it is necessary to make them truly persistent and mobile. (87J)
One of the reasons for the curiosity about XRI is because it might provide a way to get what you want, nice labels, while still maintaining the robust identifiers. (87K)