Hello,
I was wondering about the best metaphor for croquet spaces referencing. There are several possibilities for presenting the information: text-based (name/tag lookup service), 2D (map), 3D (globe? galaxy?). For instance, why not imagining a google earth croquet layer containing kmzs (3D model+location) for spaces? The result would be life-like buildings, linking towards croquet spaces. But the real question is: would the geographical metaphor be interesting for croquet ? Another possibility is to link all croquet spaces from world to world; for instance, a dedicated index island, with lots of portals to all islands. Or an island presenting categorized portals towards subworlds (leading to these worlds). Please tell me what you think of an interesting way of presenting the index information. Florent |
Florent THIERY wrote: > Hello, > > I was wondering about the best metaphor for croquet spaces > referencing. Yes. The ideas of referencing and librarianship also fascinate me. I irritated my brother (not as thrilled I suppose as I with librarianship, I suppose) at my astonishment at cryptic signs with only the words "The Algorithm Killed Jeeves". He took off in another direction as I searched vainly for fine print telling what it was about . Yes. Inquiring minds want to know how references might be handled . Google, guys asking from Jersey, or squeak doing knowledge navigator algorithms ? |
On Fri, 2007-05-04 at 16:13 -0500, Paul Sheldon wrote:
> > Florent THIERY wrote: > > > Hello, > > > > I was wondering about the best metaphor for croquet spaces > > referencing. > > Yes. The ideas of referencing and librarianship also fascinate me. > > I irritated my brother (not as thrilled I suppose as I with > librarianship, I suppose) at my astonishment at cryptic signs with only > the words "The Algorithm Killed Jeeves". > > He took off in another direction as I searched vainly for fine print > telling what it was about . > > Yes. Inquiring minds want to know how references might be handled . > > Google, guys asking from Jersey, or squeak doing knowledge navigator > algorithms ? A typical topological map doesn't offer help, does it? The confluence of the "worlds" paradigm with the data representations being graphical means that typical text indexing probably won't offer much help either. The ability to teleport from place to place also removes the physical link means of location. Perhaps we need to think literally more "universally" rather than "worldly" or mapping. I was thinking that perhaps some form of "travelers guide" perhaps with some co-ordinate system might apply, something like the Stargate used in the television series. A guide with hyperlinking could possibly show snapshots of the entry point, and perhaps some verbal description of the location and what it offers. A bit of script could link to the teleportation code to take one there by simply touching the correct entry in the guide. The guide could be either system resident or could be a bit of hypertext displayed on a terminal at the teleportation point. Also an industry could start up offering more in depth information on the locations available to the traveler. This has considerable depth and potential to it. The location of points, though should be standardized so that all publications could use the same addressing methodology. Any other thoughts? Regards, Les H |
In reply to this post by Paul Sheldon-2
> Google, guys asking from Jersey, or squeak doing knowledge navigator
> algorithms ? > Maybe having a way to take profit out of these hudreds opened web apis (at least 426 as of today cf: http://programmableweb.com/apis) would allow experimenting a different type of webbrowsing.I believe freebase will be one of these very interesting APIs (http://radar.oreilly.com/archives/2007/03/freebase_will_p_1.html), as well as Amazon S3 & EC2 services, Last.fm, and online chatbots (http://www.pandorabots.com/botmaster/en/~19bd35e0e3bdec1489ad9aa1~/mostactive) too :) If intereseted, here follows an fake conversation with a talking robot, which is in fact a quite exhaustive resource about talking search engines: http://www.readwriteweb.com/archives/the_talking_search_engines_hal.php On squeaksource there are some squeak projects that are related: xml projects (SOAPClient, DomView, X Files)... The combination of a hypothetical google 3D warehouse web api metadata with something like freebase could be interesting... Florent |
In reply to this post by Florent THIERY-2
On May 2, 2007, at 1:33 AM, Florent THIERY wrote: > Hello, > > I was wondering about the best metaphor for croquet spaces > referencing. There are several possibilities for presenting the > information: text-based (name/tag lookup service), 2D (map), 3D > (globe? galaxy?). Good ideas. Also, cone trees, and these elastic-node-Web2.0-thingies that I've been seeing. > > For instance, why not imagining a google earth croquet layer > containing kmzs (3D model+location) for spaces? People have been thinking along this lines. For example, look for "intermetaverse" both in the archives of this list and on Google. > The result would be > life-like buildings, linking towards croquet spaces. But the real > question is: would the geographical metaphor be interesting for > croquet ? Sure, especially if the linked Croquet space had some semantic relationship with the geographical location of the link. Of course, not every entrance into Croquet should correspond to a geographic location. > > Another possibility is to link all croquet spaces from world to world; > for instance, a dedicated index island, with lots of portals to all > islands. Or an island presenting categorized portals towards subworlds > (leading to these worlds). Yes, sounds great. Josh > > Please tell me what you think of an interesting way of presenting the > index information. > > Florent |
Hi,
another possibility might be to navigate not by where, but by what happens in which space. Like in the university, if you look for courses: where can I do this ? where can I learn that (where means "in which course"). Then croquet beams you directly to the stuff of your interest. That means the space metapher may be used only for local neighbourhood (depends on the issues), but in the large, such thing like "location in euclidean space" may be irrelevant. The other thing is, metric (and so distance) must not necessariliy defined by coordinates. Perhaps it might be semantic distance, or level of detail, or..... That is which I find fantastic at Croquet: the ability to try out different meanings of "space". Regards Hans Am 05.05.2007 um 08:43 schrieb Joshua Gargus: > > On May 2, 2007, at 1:33 AM, Florent THIERY wrote: > >> Hello, >> >> I was wondering about the best metaphor for croquet spaces >> referencing. There are several possibilities for presenting the >> information: text-based (name/tag lookup service), 2D (map), 3D >> (globe? galaxy?). > > Good ideas. Also, cone trees, and these elastic-node-Web2.0- > thingies that I've been seeing. >> >> For instance, why not imagining a google earth croquet layer >> containing kmzs (3D model+location) for spaces? > > People have been thinking along this lines. For example, look for > "intermetaverse" both in the archives of this list and on Google. > >> The result would be >> life-like buildings, linking towards croquet spaces. But the real >> question is: would the geographical metaphor be interesting for >> croquet ? > > Sure, especially if the linked Croquet space had some semantic > relationship with the geographical location of the link. Of > course, not every entrance into Croquet should correspond to a > geographic location. >> >> Another possibility is to link all croquet spaces from world to >> world; >> for instance, a dedicated index island, with lots of portals to all >> islands. Or an island presenting categorized portals towards >> subworlds >> (leading to these worlds). > > Yes, sounds great. > > Josh > >> >> Please tell me what you think of an interesting way of presenting the >> index information. >> >> Florent > |
You could think about the links to websites returned from a Google
search as a 1-dimensional space with the distance between "worlds" based on the relevance to your search criteria. -Peter On May 9, 2007, at 12:46 PM, Hans N Beck wrote: > Hi, > > another possibility might be to navigate not by where, but by what > happens in which space. Like in the university, if you look for > courses: where can I do this ? where can I learn that (where means > "in which course"). Then croquet beams you directly to the stuff of > your interest. That means the space metapher may be used only for > local neighbourhood (depends on the issues), but in the large, such > thing like "location in euclidean space" may be irrelevant. > > The other thing is, metric (and so distance) must not necessariliy > defined by coordinates. Perhaps it might be semantic distance, or > level of detail, or..... That is which I find fantastic at Croquet: > the ability to try out different meanings of "space". > > Regards > > Hans > > Am 05.05.2007 um 08:43 schrieb Joshua Gargus: > >> >> On May 2, 2007, at 1:33 AM, Florent THIERY wrote: >> >>> Hello, >>> >>> I was wondering about the best metaphor for croquet spaces >>> referencing. There are several possibilities for presenting the >>> information: text-based (name/tag lookup service), 2D (map), 3D >>> (globe? galaxy?). >> >> Good ideas. Also, cone trees, and these elastic-node-Web2.0- >> thingies that I've been seeing. >>> >>> For instance, why not imagining a google earth croquet layer >>> containing kmzs (3D model+location) for spaces? >> >> People have been thinking along this lines. For example, look for >> "intermetaverse" both in the archives of this list and on Google. >> >>> The result would be >>> life-like buildings, linking towards croquet spaces. But the real >>> question is: would the geographical metaphor be interesting for >>> croquet ? >> >> Sure, especially if the linked Croquet space had some semantic >> relationship with the geographical location of the link. Of >> course, not every entrance into Croquet should correspond to a >> geographic location. >>> >>> Another possibility is to link all croquet spaces from world to >>> world; >>> for instance, a dedicated index island, with lots of portals to all >>> islands. Or an island presenting categorized portals towards >>> subworlds >>> (leading to these worlds). >> >> Yes, sounds great. >> >> Josh >> >>> >>> Please tell me what you think of an interesting way of presenting >>> the >>> index information. >>> >>> Florent >> > > |
In reply to this post by Hans N Beck-2
Considering that this Google Mail thing I'm using doesn't even believe in file folders anymore (the "one email one folder" concept replaced by "one email many tags"), the concept that a croquet space that is only accessible from a single location is not exactly relevant. Why not have it linked by both where and what? This is after all the web we're talking about.
On 5/9/07, Hans N Beck <[hidden email]> wrote: Hi, |
Hi,
I have not said it's the only one possibilty, I've said it may be *another* possibility..... :-) Regards Hans Am 09.05.2007 um 20:54 schrieb Erik Anderson: Considering that this Google Mail thing I'm using doesn't even believe in file folders anymore (the "one email one folder" concept replaced by "one email many tags"), the concept that a croquet space that is only accessible from a single location is not exactly relevant. Why not have it linked by both where and what? This is after all the web we're talking about. |
In reply to this post by Erik Anderson-9
Perhaps, then the idea we should all pursue should be not representation, but search. However searches seem counter-intuitive to browsing, or meandering though a space, even a multiple dimensional one. Perhaps spaces rather than space. In other words, by some means you could select the type of representation, for example, by friends links something like LinkedIn, another means might be by topics such as space, software or Aritificial intelligence, where distance might be the inverse of the some relevance factor. Another means might be geographical, for instance a metaphor of the earth, or in the event of a nano technologist perhaps a relavance map to the current chemical structure being browsed. In otherwords, the space would be simply a representation of relevance to the content of the current browsing process, and could be selected by the operator via some front panel control.
I don't quite have an image of it yet, but I think there is some key principal here that would be new, exciting and offer greater opportunity for browsing and learning from each other than anything I see now. Not quite like hypertext, because there is no "map", and relevance is there dictated by the author and editors, which often loses something in relevance to the user (speaking for myself of course). The links I am thinking of would evolve with use, based on current context and recent history, so the mapping would be somewhat dynamic, but with some means of locking a location for future reference. Sort of a static address to a location in dynamic spaces, which could continue to evolve. So when you browse a space, your current context and recent history would affect what is presented to you, but if you choose to record the current location, and perhaps some level of linkages to that location, you could store that locally and return to it later without loosing the full context of your earlier exploration, yet new information could be available to you immediately. Does this make any sense? Regards, Les H On Wed, 2007-05-09 at 11:54 -0700, Erik Anderson wrote: Considering that this Google Mail thing I'm using doesn't even believe in file folders anymore (the "one email one folder" concept replaced by "one email many tags"), the concept that a croquet space that is only accessible from a single location is not exactly relevant. Why not have it linked by both where and what? This is after all the web we're talking about. On 5/9/07, Hans N Beck <[hidden email]> wrote: |
> However searches seem counter-intuitive to browsing
In fact, the very beginning phase is: you HAVE to input a word (i can't imagine how to do it without: choose withing a million-terms html tag cloud / huge virtual 3D library for themes > Does this make any sense? Yes. I 100% agree with: * the search dimension * tags * semantics (RDF) If every object in croquet had metadata, when requests are passed, either answear if you have related answears (lookup in you local metadata), or forward to other nodes. Semantics could improve search results. Moreover, the search element and the representation of the results/browsing can be separate, as long as the lookup service is there, like: * plain text search (bot chat) * local virtual world, with recursive 3D portals (ex: lookup "atom", 2 doors appear: "science" and "rss" ---- ok the example sucks ); what about 3D portal clouds (bigger ones / closer are the most relevant)? There's also the "browse by users" option: what if one could browse the user node's cached world references? I you found a "specialist" node, then you have found a great linking resource. My interest towards a distributed overlay search engine (used as lookup, browsing and as dns replacement) is growing every day; what's good is that it could work for other metaverses (as long as the search engine program has an interface for them), or even regular websites. I hope to join a related project next year (the Cloudstack project http://www.cloudstack.com/mwiki/index.php/Main_Page -- former wiredreach looks like an interesting candidate). Thanks a lot for this topic ! Florent |
In reply to this post by Peter Moore-5
On 5/9/07, Peter Moore <[hidden email]> wrote:
> You could think about the links to websites returned from a Google > search as a 1-dimensional space with the distance between "worlds" > based on the relevance to your search criteria. The semantic similarity may be used as relevance metric http://en.wikipedia.org/wiki/Semantic_similarity |
Free forum by Nabble | Edit this page |