purple-include: Granular Transclusions for the Common Man

Many thanks to Jonathan Cheyer, Craig Latta, and Kaliya Hamlin for coming to the HyperScope sprint this past weekend, and special thanks to Christina Engelbart for hosting. Also thanks to Thom Cherryhomes and others who hung out with us on IRC. The notes from the day are up on the Wiki, and I put up some pictures as well.    (MEL)

https://i2.wp.com/farm2.static.flickr.com/1134/681861752_857ef74d28_m.jpg?w=700    (MEM)

The big news, though, is that Brad, Jonathan, and I wrote a cool hack called purple-include, based on Mark Nottingham‘s most excellent hinclude. It lets you transclude granular chunks of content from any web site by using an img-like tag. Check out the examples. I think this will go a long way in making Transclusions more common on the Web.    (MEN)

You address granular content either by using a fragment identifier that the document author provides (such as a Purple Number) or by using an XPath expression. Thanks to Tony Chang for his cool interactive XPath tester.    (MEO)

The planned next step is to create a Firefox plugin that adds a “Transclude” option when you right click inside of a browser text widget. This will allow you to transclude copied content, rather than paste it. Don’t know whether any of us will get to this soon, so we encourage the lazy web and all you Firefox hackers to beat us to the punch.    (MEP)

This was my first non-trivial foray into JavaScript, and I was disturbed by what I saw. The language itself is not horrible, although its object system makes Perl 5 look like Smalltalk. What’s shameful is its API support. We had to use a very ugly, although apparently common hack to get a DOM of external web pages. This is pure silliness. The browser is already doing the hard work of parsing broken HTML and XML and turning it into a DOM. Why not easily expose that functionality to the developer?    (MEQ)

As Brad dryly noted, “Welcome to my world.”    (MER)

Spreadsheets 2.0 and Transclusions

A few weeks ago, I had dinner with my old HyperScope buddies, Brad Neuberg and Jonathan Cheyer. We talked a bit about this Office 2.0 madness, and how a lot of these Web-based applications were disappointly uninteresting. Don’t get me wrong. There’s a lot of really nifty hacking going on behind the scenes to make this all work. But in the end, all you have is a Web-based office application. Most of these applications do little to take advantage of the network paradigm.    (M2P)

A simple and extremely cool way for Web-based spreadsheets to exploit the medium would be to support Transclusions across multiple web sites. As I’ve observed before, spreadsheets were the first applications to popularize the notion of a Transclusion, even though they didn’t call them that. When I type =E27 in a cell, it displays the content of cell E27. This, in a nutshell, is a Transclusion, and oh, is it useful.    (M2Q)

With Web-based spreadsheets, if you made cell addresses universally resolvable, you could easily support Transclusions across web sites. In other words, I could transclude the content of cell =E27 from a spreadsheet hosted on my web site into a cell on a spreadsheet hosted on another web site.    (M2R)

Why would this be useful? Well, why is it useful to link to other web sites? Today’s Web-based spreadsheets are no more collaborative than desktop spreadsheets. In theory, they’re more convenient than emailing spreadsheets back-and-forth, but they’re no different in capability. Cross-spreadsheet Transclusions would break down silos and encourage collaboration.    (M2S)

I would start with spreadsheet-to-spreadsheet Transclusions with an eye toward supporting Transclusion of non-spreadsheet content using Purple Numbers or something similar. The main technical barrier is coming up with the right addressability scheme. Seems to me that the Simplest Thing That Works would be to use fragment identifiers (which is what we did for the HyperScope). In other words, cell =E27 on a spreadsheet at http://foo/bar would have the address:    (M2T)

  http://foo/bar#E27    (M2U)

Eventually, you’d want persistent, non-URL-based identifiers, but first things first.    (M2V)

News Flash: Transclusions Already Ubiquitous!

No, this is not an advertisement for PurpleWiki (although PurpleWiki does support Transclusions). This is a wakeup call. Transclusions already exist and have existed for a long time. No, I’m not talking about Project Xanadu. I’m talking about the World Wide Web and spreadsheets, among others.    (IDZ)

First things first. What’s a Transclusion? A transclusion is a link where the content of the link is displayed inline. For example:    (IE0)

is a link. This:    (IE2)

   (IE3)

is the content of that link displayed inline. Which, of course, is example number one. Images on the Web are transclusions. When I include a URL in <img> tags, the content of that URL is displayed.    (IE4)

We use Transclusions all the time in spreadsheets. When I write =E27 in a cell, the spreadsheet displays the content of cell E27.    (IE5)

Transclusions are useful, and they’re ubiquitous, but not necessarily as “transclusions.” They’re not yet part of a shared conceptual framework for collaborative tools. Once we explicitly acknowledge their existence and their utility, we can think about implementing them across different applications in an interoperable way.    (IE6)

Tim Bray on Purple Numbers

I’ve long been a fan of Tim Bray‘s work, and although we had never crossed paths before, I had always assumed it would be instigated by me making some comment about something he had done or written about. (A few months back, I emailed him about ballparks in the Bay Area, because we seem to share a love for the sport, but that doesn’t count.) So I was surprised and pleased to see that Tim had made first contact and had (temporarily, as it turned out) implemented Purple Numbers on his blog. Not surprisingly, this is starting to generate some talk in the blogosphere. In particular, see:    (1G9)

Chris has done an excellent job of quickly addressing many of these issues. I’ll just toss in a few thoughts and references here and will look forward to more feedback.    (1GE)

Stable, Immutable IDs    (1GF)

Several people have pointed out that the IDs need to be stable. In other words, as the paragraphs move around or as new ones are added or deleted, the IDs stick to their original paragraphs. This was a fundamental motivation for Engelbart’s statement identifiers (SIDs), which we have renamed node identifiers (NID).    (1GG)

Both Purple (which needs updating) and PurpleWiki handle this correctly. You’ll notice that the addresses are stable on my blog and on Chris’s, as well as on all of the PurpleWiki installations (e.g. Collab:HomePage, PurpleWiki:HomePage). It’s because we use plugins for our respective blog software that call PurpleWiki‘s parser, which manages NIDs for us.    (1GH)

By using PurpleWiki to handle the purpling, blog content has NIDs unique to both the blog and the Wiki, which allows us to do fun stuff like transclude between the two. We get a similar effect with perplog, the excellent IRC logger written by Paul Visscher and other members of The Canonical Hackers. For example, check out Planetwork:Main Page and the chat logs over there.    (1GI)

Mark Nottingham noted that even if the NIDs are stable in the sense that they are attached to certain paragraphs, the paragraphs themselves are not semantically stable. This is a point on which I’ve ruminated in the past, and we still don’t have a good solution to it.    (1GJ)

History and Other Worthy Projects    (1GK)

I’ve written up my own brief history of Purple Numbers, which fills in some holes in Chris’s account. In it, I mention Murray Altheim‘s plink. Plink is no longer available on the net, because it has been subsumed into his latest project, Ceryle. Everything that I’ve seen about Ceryle so far kicks butt, so if you’d like to see it too, please drop Murray an email and encourage him to hurry up and finish his thesis so that he can make Ceryle available! You can tell him I sent you, and you can forward him the link to this paragraph; he’ll understand.    (1GL)

Matt Schneider is the creator of the PurpleSlurple purple numbering proxy, which will add Purple Numbers to any (well, most) documents on the Web. PurpleSlurple deserves a lot of credit for spreading the meme.    (1GM)

Mike Mell implemented Purple Numbers in ZWiki for last year’s Planetwork Conference (see ZWiki:ZwikiAndPurpleNumbers), which in turn influenced the evolution of Purple Numbers in PurpleWiki. Mike and Matt have both experimented with JavaScript for making the numbers less intrusive.    (1GN)

The latest version of the Compendium Dialogue Mapping tool exports HTML maps with Purple Numbers.    (1GO)

In addition to Murray, Matt, and Mike, several other members of the Blue Oxen Associates Collaboration Collaboratory have contributed to the evolution of Purple Numbers, especially Peter Jones, Jack Park, and Bill Seitz. In addition to his contributions to the technical and philosophical discussions, Peter wrote the hilarious Hymn of the Church Of Purple and excerpts of the Book of the Church Of Purple.    (1GP)

Finally, many good folks in the blogosphere have helped spread the meme in many subtle ways, particularly those noted connectors Seb Paquet and Clay Shirky.    (1GQ)

Evangelism and The Big Picture    (1GR)

Okay, that last section was starting to sound like an award acceptance speech, and although none of us have won any awards, one thing is clear: The contributions of many have vastly improved this simple, but valuable tool. I’m hoping that momentum picks up even more with these recent perturbances. I’m especially heartened to see experiments for improving the look-and-feel.    (1GS)

I want to quibble with one thing that Chris Dent said in his most recent account. (When we have distributed Purple Numbers, I’ll be able to transclude it, but for now, you’ll just have to live with the cut-and-paste):    (1GT)

When he [Eugene] and I got together to do PurpleWiki, we were primarily shooting for granular addressability. Once we got that working, I started getting all jazzed about somehow, maybe, someday, being able to do Transclusion. Eugene was into the idea but I felt somehow that he didn’t quite get it. Since then we’ve implemented Transclusion and new people have come along with ideas of things to do that I’m sure I don’t quite get, but are probably a next step that will be great.    (1GU)

There are many things I don’t quite get, but Transclusions are not one of them, at least at the level we first implemented them. That’s okay, though. Ted Nelson felt the same way about my understanding of Transclusions as Chris did, although for different reasons. (And, I suspect that Ted would have had the same opinions of Chris.) I mention this here not so much because I want to correct the record, but because it gives me an excuse to tell some anecdotes and to reveal a bit about myself.    (1GV)

Anyone who knows Doug Engelbart knows that he complains a lot. The beauty of being an acknowledged pioneer and visionary is that people pay attention, even if they they don’t think much of those complaints. When I first began working with Doug in early 2000, I would occasionally write up small papers and put them on the Web. Doug would complain that they didn’t have Purple Numbers. At the time, I recognized the value of granular addresses, but didn’t think they were worth the trouble to add them to my documents. I also didn’t think they were the “right” solution. Nevertheless, because Doug was Doug, I decided to throw him a bone. So I spent a few hours writing Purple (most of which was spent learning XSLT), and started posting documents with Purple Numbers.    (1GW)

Then, a funny thing happened. I got used to them. I got so used to them, I wanted them everywhere.    (1GX)

A few people just get Purple Numbers right away. Murray was probably the first of those not originally in the Engelbart crowd to do so; Chris followed soon thereafter, as did Matt. The vast majority of folks get the concept, but don’t really find them important until they start using them. Then, like me, they want them everywhere. Getting people past that first step is crucial.    (1GY)

A few nights ago, I had a late night conversation with Gabe Wachob (chair of the OASIS XRI committee) on IRC. (This eventually led to a conversation between Chris and me, which led to Chris’s blog entry, which led to Tim discovering Purple Numbers, which led to this entry. Think Out Loud is an amazing thing.) Gabe knew what Purple Numbers were, but hadn’t thought twice about them. I had wanted to ask him some questions about using XRI addresses as identifiers, and in order to do so, I gave him a quick demonstration of Transclusions. The light bulb went off; all of a sudden, he really, truly got it.    (1GZ)

Richard Gabriel, one of our advisors, is well known for his Worse Is Better essays (among other things). I think Purple Numbers are an outstanding example of Worse Is Better. They fulfill an immediate need, and they cause us to think more deeply about some of the underlying issues. I’d like to see Purple Numbers all over the place, but I’d also like to see a group of deep thinkers and tinkerers consider and evolve the concept. It’s part of a larger philosophy that I like to call The Blue Oxen Way.    (1H0)

This last point is extremely important. Chris has thankfully been a much more enthusiastic evangelist of Purple Numbers than I have, and in the past he’s called me “ambivalent” about Purple Numbers. That’s not so far from the truth. It’s not that I’m any less enthusiastic about Purple Numbers themselves — I am a card-carrying member of the Church Of Purple, and the current attention and potential for wider usage thrill me. However, I’m cautious about evangelizing Purple Numbers, because I don’t want people to get too caught up in the tool itself and forget about the bigger picture. It’s the reason I didn’t mention Purple Numbers at all in my manifesto.    (1H1)

At the Planetwork forum two weeks ago, Fen Labalme, Victor Grey, and I gave the first public demo of a working Identity Commons Single Sign-On system. We were tickled pink by the demo, which to everyone else looked just like any other login system. The reason we were so excited was that we knew the system used an underlying infrastructure that would eventually enable much greater things. The demo itself, unfortunately, didn’t convey that to anyone who didn’t already understand this.    (1H2)

I’m probably a bit oversensitive about this sort of thing, and I’m constantly seeking better balance. But it’s always in the back of my mind. When I talk to people about Blue Oxen Associates, I usually spend more time talking about the sociological aspect of collaboration rather than the tools, even though I have plenty to say about the latter. Can Purple Numbers make the world a better place? I truly, honestly, believe that they can. (This is a topic for another day.) But when I see groups that excel in collaboration (or conversely, those that stink at it), Purple Numbers are usually the furthest thing from my mind. Much more important is the need to identify and understand these patterns of collaboration (of which tool usage is an important part).    (1H3)

Transclusions, Path-Based Addressing, and Version Control

The PurpleWiki community has been rumbling recently, thanks largely to the contributions of two members of the Canonical Hackers, Jason Cook and Matthew O’Connor. Jason wrote perplog, an IRC logger that supports Purple Numbers and Transclusions. Matthew hacked a PurpleWiki node manager, then started adding and fixing other stuff, including an XML-RPC interface. Additionally, John Sechrest developed an experimental email interface to PurpleWiki. Lots of great stuff. It’s forcing us to get off our butts and make some long-promised changes and explain some long-undocumented things. Open Source is a wonderful thing.    (117)

A lot of the excitement is because of PurpleWiki‘s support for Transclusions. We had Transclusions in mind when we architected PurpleWiki, but we (or I, at least) didn’t think Transclusions would actually be implemented until much later. However, exactly one year ago today, Chris Dent got the itch and started playing. A few months later, Chris committed some code, and suddenly, we had Transclusions. It was a total hack, but it worked, and it was unexpectedly cool.    (118)

It’s still a hack, and it needs to be cleaned up, but it’s suddenly become a higher priority. First, we have a pretty good idea of how to support Transclusions “correctly.” Second, having had the chance to use Transclusions regularly, we are starting to recognize their utility and want to take greater advantage of that. (See, for example, my early specs for Abelard.) Third, people are starting to get excited about them.    (119)

Transcluding Multiple Chunks    (11A)

Currently, we support Transclusions of individual nodes (paragraphs, headers, list items, etc.) via the following syntax:    (11B)

  [t nid]    (11C)

where nid is the ID of the node you want to transclude. This works fine when you want to transclude small chunks, but at times, it’s useful to be able to transclude multiple chunks on a page. Rather than specify a transclusion for each individual node, it would be nice to have a syntax for specifying a collection of nodes in a single transclusion.    (11D)

Chris proposed the following syntax:    (11E)

  [t nid,nid,nid,...]    (11F)

This is problematic. The current implementation suggests that the transclusion command be replaced by the content identified by the specified NID. This proposal suggests that the command be replaced by both the content and the structure of the content. If you had a document like:    (11G)

  = Plan for World Domination {nid 1} =    (11H)
  # Finish PurpleWiki. {nid 2}    (11I)

and you tried to transclude this content with:    (11J)

  * [t 1, 2]    (11K)

what is the parser supposed to translate this to?    (11L)

A proper solution must be treated as its own structural element within a PurpleWiki document. More importantly, the syntax should capture document-specific context. This contrasts with the current syntax, which ignores document context entirely.    (11M)

Why is document context important? Suppose you have the following task list:    (11N)

  = To Do {nid 3} =    (11O)
  * Buy milk. {nid 4}   * Feed iguana. {nid 5}   * Implement distributed Transclusions. {nid 6}   * Vote in primaries. {nid 7}   * Expose API to Backlinks. {nid 8}    (11P)

Suppose you want to start a PurpleWiki-specific task list, transcluding all of the list items relevant to PurpleWiki (in this case, nodes 6 and 8). The resulting document might look like:    (11Q)

  {title PurpleWikiToDo}    (11R)
  = PurpleWiki To Do {nid 9} =    (11S)
  * [t 6] {nid A}   * [t 8] {nid B}    (11T)

Now, suppose you want to replicate this task list on another page. You could transclude all of the items individually, just as you do on the PurpleWiki To Do page. In this case, a slight variation of Chris’s proposed syntax (a standalone structural element) would simplify that process. (It also raises an interesting question: Which NIDs do you use for the transclusions: 6 and 8, or A and B? Does it make a difference?)    (11U)

However, what I really want to do is say, “Transclude all of the list items on the ‘PurpleWiki To Do’ page.” For this, you want something like XPointer:    (11V)

  [transclude PurpleWikiToDo#xpointer(id("9")/li)]    (11W)

A few observations: First, the command should be on a line by itself, and it should be interpreted as an independent structural element that will be replaced by a set of structural elements and content. I used “transclude” instead of “t” to make the point that these are two different commands. Second, the Transclusion command specifies a range of nodes within a document, as opposed to a document-independent list of nodes. Third, this combines a path-based address with an ID-based address.    (11X)

Fourth (and this is an implementation detail), if we want to support such syntax, it would behoove us to use an XML data model rather than the home grown model we’re currently using. This way, we could easily plug in existing XPointer implementations to do the queries.    (11Y)

Version Control    (11Z)

The fact that Wiki pages are dynamic throws a kink into all of this. The syntax I propose above takes this into account for the most part. Barring major changes to the PurpleWiki To Do page, the transcluded content will include all of the PurpleWiki tasks, even if more items are added later. If you had to list a set of NIDs, then you would have to be diligent about updating that list manually every time the To Do page changed.    (120)

In addition to supporting path-based addressing, we also need to allow people to specify versions in the address. In other words, you may want to transclude a specific version of a node or a set of nodes from a specific version of a document.    (121)

This shouldn’t be too difficult, but there are some complications. The biggie is whether to transclude a node if the node no longer exists in any document. My instinct right now is telling me that yes, it should, but it should make it clear somehow that it’s an orphan node. (See my previous entry on link integrity for more on this.)    (122)