Meeting with Alan 12:30 June 16, 2010 mkljun ZPrivate: Conversations ZPrivate: Minutes This content has restricted access, please type the password below and get access. Tags: building TICemaillinks Next Sort multiple citatations in Latex by citation numbers Previous Finder and its Preview function No Responses Comments3 Pingbacks0 Matjaz says: June 9, 2010 at 10:15 pm Hi Alan, (a) unification of personal information space was nicely divided by Karager in ACM communication in 2006 (attached file – it is also in the PIM book by Jones and Teevan). A summary can be found in my blog: http://pim.famnit.upr.si/blog/index.php?/archives/7-Review-of-Prototype-design-a-peice-that-was-cut-out-the-paper.html – grouping (in a folder or by tags) – cross referencing (urls in files and emails) Attaching files to emails is ‘linking’ but the original file in the file system and attached files loose their connection and attached file also looses it’s semantics with the file system. With email-email we also have ‘links’ based on subject (thread), time (not only sender, folder). And time: ‘linking’ items based on the (aproximatelly) same creation time or sequential creation time mostly helps (if at all) re-finding information. These links are not based on task (or semantics) and I doubt that users do actually ‘link’ this information in a way I understand these ‘links’ (related to task). But they certainly are linked in the creation context. So the creation time sequence is just a context in which information was created. Users might as well remember the weather or the meeting or the on-line auction that happened in the same time period (hour, day, week). They might remember all other information items they created that week. Example: I need some information to complete a task; I know I have stored it but where, when, how; It was the first week in May; what happened to me then; I remember editing a file then for a meeting; try browsing folder; try remembering other clues … But yes, these are the ‘links’ that are possible to create (implicitly or explicitly) with common software. Other groups mentioned by Karger are not durable (based on opened applications on the desktop) or are prototypes (RDF, tags in a file system …) (b) I thought of the “little app” I could do last night and it seams that we are almost thinking simultaneously the same thing. It sounds simple but I’m not sure if the implementation is simple as well :). The idea is to link URLs, files (hierarchy path) and emails based on tags. But it would work only in few directions: – file (or folder) – url: adding an URL as a tag to a file or folder example: dragging URL from the address bar of a browser to a file or folder example: adding a location tag to a file could add URL to google maps – email – file (or folder): adding a file hierarchy path as a tag to the email example: a thunderbird extention could add a hierarchy path to a file as tag – email – url: adding URL to email as tag example: the same as previus one – url – file (or folder): a firefox extention to add a hierarchy path to bookmarked site as a tag example: dragging file over the opened web site in FF a site is bookmarked and a hierarchy path added as a tag – url – email: missing as I can no think of any email URI to be added to URL – file – email: missing -||- The app could have a DB for all stored links and could show the relation file-email as email-file (the same for url-email as email-url). And visualize all relations in a graph like structures. Problem would be if a file is deleted, moved, renamed. The first step is to find if it is possible to add tags to a windows file system in an easy way. (c) Organization/maintenance: it seams I have missed you answer on Skype or you haven’t answered yet. Enjoy Scotland :). lp mk Reply Alan says: June 9, 2010 at 10:17 pm At 16:50 +0100 10/5/10, mkljun wrote: > > (a) unification of personal information space was nicely divided by Karager in ACM communication in 2006 … Although there is such a lot of overlap, the interactions between these trad. PIM research and sem desktop research seems relatively light. But I may not be up-to-date with this, anything you’ve seen that brings these together … and of course sem desktop does not include string links to apps on wider web! > > Attaching files to emails is ‘linking’ but the original file in the file system and attached files loose their connection and attached file also looses it’s semantics with the file system. absolutely, two isseus here: (i) one way linking – can navigate from email to file, but not the other way interestingly also a problem with sem web linked data! (ii) lost semantics – this is an example where the semantics is available at some point in the computer, but either forgotten, or unused > > (b) >> I thought of the “little app” I could do last night and it seams that we are almost thinking simultaneously the same thing. 🙂 > > It sounds simple but I’m not sure if the implementation is simple as well :). 😀 > > The idea is to link URLs, files (hierarchy path) and emails based on tags… > >… > > The app could have a DB for all stored links and could show the relation file-email as email-file (the same for url-email as email-url). And visualize all relations in a graph like structures. this part will be straightforward > > Problem would be if a file is deleted, moved, renamed. structural change is always the most difficult! even harder when files are broken into bits or assembled > > The first step is to find if it is possible to add tags to a windows file system in an easy way. … and I’m glad that is you not me 😉 I need to put you in touch with the people in Athens and Tripolis as they are doing this sort of thing too. Alan Reply Alan says: June 9, 2010 at 10:21 pm At 16:50 +0100 10/5/10, mkljun wrote: > > (c) > > Organization/maintenance: it seams I have missed you answer on Skype or you haven’t answered yet. Here is the stuff form Skype Alan =============================== I recall longish conversatuon about it, but also not all the details!In a sense both are forward looking, why organise [10/05/2010 05:23:36] Alan Dix: However organisation seems to be driven by the current need to ‘do something’ with this file/emai [10/05/2010 05:24:01] Alan Dix: whereas maintenace is much more “better sort thinsg out so I can deal with things better iun future”. [10/05/2010 05:24:41] Alan Dix: As you say future needs of org. more clear as it tends to involve thinsg that are part if an ongoing task (which may itself be long lasting) [10/05/2010 05:25:19] Alan Dix: whereas mainenance more about future, but possibly as yet unconceoved tasks [10/05/2010 05:25:25] Alan Dix: … however … [10/05/2010 05:25:49] Alan Dix: where does gthat put the web page you find by accident while doing something else? [10/05/2010 05:26:02] Alan Dix: the *need* to file it is current [10/05/2010 05:26:44] Alan Dix: but the potential future application not related to present task [10/05/2010 05:27:04] Alan Dix: it seems to me then we have at least three dimensions/properties connecting to organisation: [10/05/2010 05:28:16] Alan Dix: (1) the trigger for doing maintabce/organisation – need to file a thing, thins looking untidy, displacment activity as I should be doing soemtjing else, “I always file on fridays” [10/05/2010 05:28:31] Alan Dix: (2) clarity of future needs [10/05/2010 05:28:47] Alan Dix: (3) relation of future needs to current activity [10/05/2010 05:30:08] Alan Dix: Note (3) influences (2) – tend to have clearer idea of current needs [10/05/2010 05:30:15] Alan Dix: I guess also: [10/05/2010 05:30:46] Alan Dix: (4) how long is the time perspective: hours, days, weeks, months [10/05/2010 05:31:10] Alan Dix: … all for now on this, but separet mail about other things … ===================== Reply Leave a Reply Cancel replyYour email address will not be published. Required fields are marked *Comment Name * Email * Website Save my name, email, and website in this browser for the next time I comment. This site uses Akismet to reduce spam. Learn how your comment data is processed.