Comments on: Conceptualizing Nepomuk http://blog.chatonka.com/2009/09/conceptualizing-nepomuk/ Climb On Wed, 02 Jun 2010 20:45:05 +0000 http://wordpress.org/?v=2.9.2 hourly 1 By: mat69 http://blog.chatonka.com/2009/09/conceptualizing-nepomuk/comment-page-1/#comment-8927 mat69 Thu, 10 Sep 2009 13:08:16 +0000 http://blog.chatonka.com/2009/09/conceptualizing-nepomuk/#comment-8927 Ups forgot the last paragraph. Baseline is that users are lazy and thus most likely will only enter a keyword, the same is true for tagging a lot of users won't use it that much. Imo the only way for Nepomuk to succeed is automatic getting of semantic information (a lot of programs use that already), making it easy for the user to associate information to a file, like associating a place and a contact to a task and then making it easy to find the stuff. Ups forgot the last paragraph.

Baseline is that users are lazy and thus most likely will only enter a keyword, the same is true for tagging a lot of users won’t use it that much.

Imo the only way for Nepomuk to succeed is automatic getting of semantic information (a lot of programs use that already), making it easy for the user to associate information to a file, like associating a place and a contact to a task and then making it easy to find the stuff.

]]> By: mat69 http://blog.chatonka.com/2009/09/conceptualizing-nepomuk/comment-page-1/#comment-8926 mat69 Thu, 10 Sep 2009 12:59:35 +0000 http://blog.chatonka.com/2009/09/conceptualizing-nepomuk/#comment-8926 The user should not choose that they want to search for a contact when entering the contacts name, that would just take too long and would be avoided as often as possible -- so if no pictures are tagged with Sydney it would be avoided. Instead it should work on a trial and error base, like when you explain something to someone, you talk about "Sydney" and later say "no, not this Sydney, that Sydney who knows this person ...": Enter "Sydney", now as you've mentioned you are provided with anything that has "Sydney" anywhere, be it file name, tag etc. Now the search-dialog should provide you with something like a sidebar, where all occurrences of Sydney are mentioned, something like "Contact", "Files", "Tag" ... Choosing "Contact" would provide you with a link to that contact and everything that is associated with that contact, so choosing now "Task" in the sidebar would show you all task that are associated to "Sydney". It could look like similar to the new sorting system that has been created for Amarok with breadcrumbs --> (Picture of person or contact icon) Sydney > (Task Icon) Tasks So baseline is The user should not choose that they want to search for a contact when entering the contacts name, that would just take too long and would be avoided as often as possible — so if no pictures are tagged with Sydney it would be avoided. Instead it should work on a trial and error base, like when you explain something to someone, you talk about “Sydney” and later say “no, not this Sydney, that Sydney who knows this person …”:

Enter “Sydney”, now as you’ve mentioned you are provided with anything that has “Sydney” anywhere, be it file name, tag etc.

Now the search-dialog should provide you with something like a sidebar, where all occurrences of Sydney are mentioned, something like “Contact”, “Files”, “Tag” …
Choosing “Contact” would provide you with a link to that contact and everything that is associated with that contact, so choosing now “Task” in the sidebar would show you all task that are associated to “Sydney”.

It could look like similar to the new sorting system that has been created for Amarok with breadcrumbs –>
(Picture of person or contact icon) Sydney > (Task Icon) Tasks

So baseline is

]]> By: simca http://blog.chatonka.com/2009/09/conceptualizing-nepomuk/comment-page-1/#comment-8918 simca Wed, 09 Sep 2009 15:00:10 +0000 http://blog.chatonka.com/2009/09/conceptualizing-nepomuk/#comment-8918 IMHO nepomuk is a solution searching for problems. On AmigaOS 3.1 (about 12 years ago) almost every application used the filesystem's "comment field" for tagging files. For example, if it's a jpeg attachment saved from a mail, in the comment there was the sender name and email address. If it was a downloaded file, there was the source url where it came from, etc. Ok it wasnt searchable, but there were some really useful information, and this old technology is still nowhere else today. IMHO nepomuk is a solution searching for problems. On AmigaOS 3.1 (about 12 years ago) almost every application used the filesystem’s “comment field” for tagging files. For example, if it’s a jpeg attachment saved from a mail, in the comment there was the sender name and email address. If it was a downloaded file, there was the source url where it came from, etc. Ok it wasnt searchable, but there were some really useful information, and this old technology is still nowhere else today.

]]> By: astromme http://blog.chatonka.com/2009/09/conceptualizing-nepomuk/comment-page-1/#comment-8917 astromme Wed, 09 Sep 2009 14:52:06 +0000 http://blog.chatonka.com/2009/09/conceptualizing-nepomuk/#comment-8917 I don't expect end users to 'learn' or even know what an ontology is. But that doesn't mean that they aren't a useful tool for application developers. You're absolutely right, the applications should be providing nepomuk with as much data as they can. It's not happening (much) now, and I'd love to see that change. It might be a good goal to have for 4.5 where almost every app is plugged into nepomuk giving some data. I don’t expect end users to ‘learn’ or even know what an ontology is. But that doesn’t mean that they aren’t a useful tool for application developers.

You’re absolutely right, the applications should be providing nepomuk with as much data as they can. It’s not happening (much) now, and I’d love to see that change. It might be a good goal to have for 4.5 where almost every app is plugged into nepomuk giving some data.

]]> By: Tom http://blog.chatonka.com/2009/09/conceptualizing-nepomuk/comment-page-1/#comment-8916 Tom Wed, 09 Sep 2009 07:52:27 +0000 http://blog.chatonka.com/2009/09/conceptualizing-nepomuk/#comment-8916 Well I don't learn Ontologies. I want my task manager app to provide nepomuk with data on who is involved and where it happens. I think all KDE4 apps should finally start providing this information. It has been nearly 4 releases and still no app is doing anything useful with nepomuk. You can't rely on the user to do all this complicated stuff. At this rate of adoption Nepomuk won't be useful before KDE 4.16 in 2014. I thought KDE agreed on the pillars, so why is nobody using them? Well I don’t learn Ontologies. I want my task manager app to provide nepomuk with data on who is involved and where it happens.

I think all KDE4 apps should finally start providing this information. It has been nearly 4 releases and still no app is doing anything useful with nepomuk. You can’t rely on the user to do all this complicated stuff.

At this rate of adoption Nepomuk won’t be useful before KDE 4.16 in 2014.

I thought KDE agreed on the pillars, so why is nobody using them?

]]> By: myselfhimself http://blog.chatonka.com/2009/09/conceptualizing-nepomuk/comment-page-1/#comment-8915 myselfhimself Wed, 09 Sep 2009 06:22:33 +0000 http://blog.chatonka.com/2009/09/conceptualizing-nepomuk/#comment-8915 nice post thank you I like that approach of discovering, trying yourself at it, building stuff around... while documenting all along for community benefits. nice post thank you
I like that approach of discovering, trying yourself at it, building stuff around… while documenting all along for community benefits.

]]> By: Socceroos http://blog.chatonka.com/2009/09/conceptualizing-nepomuk/comment-page-1/#comment-8914 Socceroos Wed, 09 Sep 2009 03:42:39 +0000 http://blog.chatonka.com/2009/09/conceptualizing-nepomuk/#comment-8914 Nice post. I'm looking forward to your post about your use of the APIs! Nice post. I’m looking forward to your post about your use of the APIs!

]]>