<?xml version="1.0" encoding="UTF-8"?><rss version="2.0"
	xmlns:content="http://purl.org/rss/1.0/modules/content/"
	xmlns:dc="http://purl.org/dc/elements/1.1/"
	xmlns:atom="http://www.w3.org/2005/Atom"
	>
<channel>
	<title>Comments on: The Reality of Semantic Desktops: Death To Tags, Labels and Folders</title>
	<atom:link href="http://kubasik.net/blog/2008/05/21/the-reality-of-semantic-desktops-death-to-tags-labels-and-folders/feed/" rel="self" type="application/rss+xml" />
	<link>http://kubasik.net/blog/2008/05/21/the-reality-of-semantic-desktops-death-to-tags-labels-and-folders/</link>
	<description>A Place For My Mind to Wander</description>
	<pubDate>Mon, 06 Oct 2008 11:18:28 +0000</pubDate>
	<generator>http://wordpress.org/?v=2.6.2</generator>
		<item>
		<title>By: Rouslan</title>
		<link>http://kubasik.net/blog/2008/05/21/the-reality-of-semantic-desktops-death-to-tags-labels-and-folders/#comment-33730</link>
		<dc:creator>Rouslan</dc:creator>
		<pubDate>Thu, 22 May 2008 02:37:48 +0000</pubDate>
		<guid isPermaLink="false">http://kubasik.net/blog/?p=141#comment-33730</guid>
		<description>Hey. great ideas! Geoffrey Hinton recently improved upon neural networks efficiency by removing the need for labels, and we can improve user experience and efficiency by removing tags from files and deducing meaning from the content. Let's go even further and get rid of files names. Raskin comes to mind: "Let the content be content." He is a big supporter of search too.

So good luck! I hope I can help in any way possible.</description>
		<content:encoded><![CDATA[<p>Hey. great ideas! Geoffrey Hinton recently improved upon neural networks efficiency by removing the need for labels, and we can improve user experience and efficiency by removing tags from files and deducing meaning from the content. Let&#8217;s go even further and get rid of files names. Raskin comes to mind: &#8220;Let the content be content.&#8221; He is a big supporter of search too.</p>
<p>So good luck! I hope I can help in any way possible.</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: Books and Magazines Blog &#187; Archive &#187; Kevin Kubasik: The Reality of Semantic Desktops: Death To Tags, Labels and Folders</title>
		<link>http://kubasik.net/blog/2008/05/21/the-reality-of-semantic-desktops-death-to-tags-labels-and-folders/#comment-33550</link>
		<dc:creator>Books and Magazines Blog &#187; Archive &#187; Kevin Kubasik: The Reality of Semantic Desktops: Death To Tags, Labels and Folders</dc:creator>
		<pubDate>Wed, 21 May 2008 19:38:55 +0000</pubDate>
		<guid isPermaLink="false">http://kubasik.net/blog/?p=141#comment-33550</guid>
		<description>[...] Kevin Kubasik: The Reality of Semantic Desktops: Death To Tags, Labels and Folders So, I recently saw some more updates on the Gnome Live wiki regarding the evolution of a Ã¢Â€Â˜Semantic DesktopÃ¢Â€Â™. I have some bad news people: Its not going to happen. Now before everyone spends 20 minutes explaining all the ways it could, let me clarify my point. ItÃ¢Â€Â™s a largely unattainable goal, &#8230; [...]</description>
		<content:encoded><![CDATA[<p>[...] Kevin Kubasik: The Reality of Semantic Desktops: Death To Tags, Labels and Folders So, I recently saw some more updates on the Gnome Live wiki regarding the evolution of a Ã¢Â€Â˜Semantic DesktopÃ¢Â€Â™. I have some bad news people: Its not going to happen. Now before everyone spends 20 minutes explaining all the ways it could, let me clarify my point. ItÃ¢Â€Â™s a largely unattainable goal, &#8230; [...]</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: Kevin Kubasik</title>
		<link>http://kubasik.net/blog/2008/05/21/the-reality-of-semantic-desktops-death-to-tags-labels-and-folders/#comment-33537</link>
		<dc:creator>Kevin Kubasik</dc:creator>
		<pubDate>Wed, 21 May 2008 18:39:29 +0000</pubDate>
		<guid isPermaLink="false">http://kubasik.net/blog/?p=141#comment-33537</guid>
		<description>@Benedikt Thanks! 

@khiraly: Yeah, most of those issues are distro/linkage problems, there are tons of problems making sure that gtk/nautilus get built against the proper API's. Things like that should work in Beagle through the main interface. My thought is more towards an augments ranking system to make desktop search much more accurate, and the means to know what desktop objects are important to others (ie. Path-Based browsing, not unlike how dasher works, but cooler and more awesome-er). I want to see more natural language parsing on a massive scale. Imagine not opening an application to add something to the calendar or add a contact, but simply typing a sentance like:
'I have a meeting on Saturday with Joe'
or 
'I need to fix bug #555555 by Friday'
Google Calendar has an implementation of this within a narrow scope, but I would like to try and interact on a more natural level. In the above mentioned scenario, if I proceed to ask 3 people about bug #555555 then it has an inferred or suggested priority (overwritable). I'll have some code to demo soonish.</description>
		<content:encoded><![CDATA[<p>@Benedikt Thanks! </p>
<p>@khiraly: Yeah, most of those issues are distro/linkage problems, there are tons of problems making sure that gtk/nautilus get built against the proper API&#8217;s. Things like that should work in Beagle through the main interface. My thought is more towards an augments ranking system to make desktop search much more accurate, and the means to know what desktop objects are important to others (ie. Path-Based browsing, not unlike how dasher works, but cooler and more awesome-er). I want to see more natural language parsing on a massive scale. Imagine not opening an application to add something to the calendar or add a contact, but simply typing a sentance like:<br />
&#8216;I have a meeting on Saturday with Joe&#8217;<br />
or<br />
&#8216;I need to fix bug #555555 by Friday&#8217;<br />
Google Calendar has an implementation of this within a narrow scope, but I would like to try and interact on a more natural level. In the above mentioned scenario, if I proceed to ask 3 people about bug #555555 then it has an inferred or suggested priority (overwritable). I&#8217;ll have some code to demo soonish.</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: khiraly</title>
		<link>http://kubasik.net/blog/2008/05/21/the-reality-of-semantic-desktops-death-to-tags-labels-and-folders/#comment-33536</link>
		<dc:creator>khiraly</dc:creator>
		<pubDate>Wed, 21 May 2008 18:00:21 +0000</pubDate>
		<guid isPermaLink="false">http://kubasik.net/blog/?p=141#comment-33536</guid>
		<description>I would be more than happy if nautilus' search functionality would work.
If Im looking for something I always use the good old boy: midnight commander's search functionality.
 
For example, if I search in nautilus for: *.avi
It doesnt return with all of them, just some or none.
I often looking for '*something*.pdf'.
So I know two things: 
* the extension (the type of the document)
* part of the filename

Would be good, if the base will work before to do some exotic things.. (I could written buzzwords)

If somebody keep their filesystem clean and organize stuff cleanly, for him the tags are not helping, f-spot/picasa photo organizing is just confusing (like the internet explorer's auto-guess folder views under windows). All these softwares are just trying to be more clever as I am, and keep always reorganizing (my already organized)  stuff.

Its soo boring.</description>
		<content:encoded><![CDATA[<p>I would be more than happy if nautilus&#8217; search functionality would work.<br />
If Im looking for something I always use the good old boy: midnight commander&#8217;s search functionality.</p>
<p>For example, if I search in nautilus for: *.avi<br />
It doesnt return with all of them, just some or none.<br />
I often looking for &#8216;*something*.pdf&#8217;.<br />
So I know two things:<br />
* the extension (the type of the document)<br />
* part of the filename</p>
<p>Would be good, if the base will work before to do some exotic things.. (I could written buzzwords)</p>
<p>If somebody keep their filesystem clean and organize stuff cleanly, for him the tags are not helping, f-spot/picasa photo organizing is just confusing (like the internet explorer&#8217;s auto-guess folder views under windows). All these softwares are just trying to be more clever as I am, and keep always reorganizing (my already organized)  stuff.</p>
<p>Its soo boring.</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: Benedikt</title>
		<link>http://kubasik.net/blog/2008/05/21/the-reality-of-semantic-desktops-death-to-tags-labels-and-folders/#comment-33535</link>
		<dc:creator>Benedikt</dc:creator>
		<pubDate>Wed, 21 May 2008 17:55:44 +0000</pubDate>
		<guid isPermaLink="false">http://kubasik.net/blog/?p=141#comment-33535</guid>
		<description>Bravo!
That really is all I wanted to say, Kevin. It would be great if your vision became reality.</description>
		<content:encoded><![CDATA[<p>Bravo!<br />
That really is all I wanted to say, Kevin. It would be great if your vision became reality.</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: Kevin Kubasik</title>
		<link>http://kubasik.net/blog/2008/05/21/the-reality-of-semantic-desktops-death-to-tags-labels-and-folders/#comment-33532</link>
		<dc:creator>Kevin Kubasik</dc:creator>
		<pubDate>Wed, 21 May 2008 17:04:58 +0000</pubDate>
		<guid isPermaLink="false">http://kubasik.net/blog/?p=141#comment-33532</guid>
		<description>@oliver Exactly! I have tons of tags in f-spot, which I spend a lot of time attending too, but still, when I'm looking for a photo, I never use them... I think that some intelligent grouping based upon time, and people identified from the photos might be a cooler way.... at least I wouldn't spend 30 minutes tagging every roll I take. ;)</description>
		<content:encoded><![CDATA[<p>@oliver Exactly! I have tons of tags in f-spot, which I spend a lot of time attending too, but still, when I&#8217;m looking for a photo, I never use them&#8230; I think that some intelligent grouping based upon time, and people identified from the photos might be a cooler way&#8230;. at least I wouldn&#8217;t spend 30 minutes tagging every roll I take. <img src='http://kubasik.net/blog/wp-includes/images/smilies/icon_wink.gif' alt=';)' class='wp-smiley' /></p>
]]></content:encoded>
	</item>
	<item>
		<title>By: oliver</title>
		<link>http://kubasik.net/blog/2008/05/21/the-reality-of-semantic-desktops-death-to-tags-labels-and-folders/#comment-33529</link>
		<dc:creator>oliver</dc:creator>
		<pubDate>Wed, 21 May 2008 16:43:39 +0000</pubDate>
		<guid isPermaLink="false">http://kubasik.net/blog/?p=141#comment-33529</guid>
		<description>Seems like a good idea to me. For example, I noticed that in F-Spot, I usually tag a set of photos with something like "Mallorca vacation 2007" or "Christmas 2007" or (if there was no special occasion) just "Family".
But actually I never really look at those tags... It's more my sense for "it _has_ to be ordered somehow or I'll forget whether that photo was from Mallorca or from somewhere else - OMG!".... Thinking of it, if F-Spot could just correlate some loose info (from other sources) to find that I've been to Mallorca during a specific time and that a certain set of photos must hence display Mallorca, that would be cool.

OTOH, it's scary to imagine that suddenly some private info that I stored in a non-public "diary" or whatever is used by some intelligent app and is then leaked onto the net...</description>
		<content:encoded><![CDATA[<p>Seems like a good idea to me. For example, I noticed that in F-Spot, I usually tag a set of photos with something like &#8220;Mallorca vacation 2007&#8243; or &#8220;Christmas 2007&#8243; or (if there was no special occasion) just &#8220;Family&#8221;.<br />
But actually I never really look at those tags&#8230; It&#8217;s more my sense for &#8220;it _has_ to be ordered somehow or I&#8217;ll forget whether that photo was from Mallorca or from somewhere else - OMG!&#8221;&#8230;. Thinking of it, if F-Spot could just correlate some loose info (from other sources) to find that I&#8217;ve been to Mallorca during a specific time and that a certain set of photos must hence display Mallorca, that would be cool.</p>
<p>OTOH, it&#8217;s scary to imagine that suddenly some private info that I stored in a non-public &#8220;diary&#8221; or whatever is used by some intelligent app and is then leaked onto the net&#8230;</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: Dread Knight</title>
		<link>http://kubasik.net/blog/2008/05/21/the-reality-of-semantic-desktops-death-to-tags-labels-and-folders/#comment-33520</link>
		<dc:creator>Dread Knight</dc:creator>
		<pubDate>Wed, 21 May 2008 15:27:19 +0000</pubDate>
		<guid isPermaLink="false">http://kubasik.net/blog/?p=141#comment-33520</guid>
		<description>I agree with you. But then, really, providing the user with more things to do ("power") doesn't mean he has to use all the stuff in excess.

I really see this as the way people use twitter for example. You can micro-blog, but you need to keep a balance (how much 'useful/often/important/etc')

It's really related to the 'hammer effect' thingy, when you give somebody a hammer and he starts nailing everything xD</description>
		<content:encoded><![CDATA[<p>I agree with you. But then, really, providing the user with more things to do (&#8221;power&#8221;) doesn&#8217;t mean he has to use all the stuff in excess.</p>
<p>I really see this as the way people use twitter for example. You can micro-blog, but you need to keep a balance (how much &#8216;useful/often/important/etc&#8217;)</p>
<p>It&#8217;s really related to the &#8216;hammer effect&#8217; thingy, when you give somebody a hammer and he starts nailing everything xD</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: gus</title>
		<link>http://kubasik.net/blog/2008/05/21/the-reality-of-semantic-desktops-death-to-tags-labels-and-folders/#comment-33478</link>
		<dc:creator>gus</dc:creator>
		<pubDate>Wed, 21 May 2008 12:24:03 +0000</pubDate>
		<guid isPermaLink="false">http://kubasik.net/blog/?p=141#comment-33478</guid>
		<description>I think you are going the right way. You can not expect that the user start filling all the information you need to make the Semantic Desktop thingy happen. Cause filling metadata sucks and no one that has a life will do it. So you need to start looking at what are the relations that you have now a day in the desktop (contacts in evolution, chat logs, mails, photos). But also add more logic to automatically add metadata to objects in the desktop ( i.e. algorithms for keywords generations there are some very cheap, algorithms for face detection something like http://code.google.com/soc/2007/gnome/appinfo.html?csaid=7CCBE044F1515ED3 ). I think semantic desktop should be one of the main goals for gnome 3.0, but if you think about it as something that the user needs to work on you are totally in the wrong path! Yes you can ask him to tag one face but the next ones will need to be guessed by gnome.

Just my 2 cents, hope not to have spammed a lot....</description>
		<content:encoded><![CDATA[<p>I think you are going the right way. You can not expect that the user start filling all the information you need to make the Semantic Desktop thingy happen. Cause filling metadata sucks and no one that has a life will do it. So you need to start looking at what are the relations that you have now a day in the desktop (contacts in evolution, chat logs, mails, photos). But also add more logic to automatically add metadata to objects in the desktop ( i.e. algorithms for keywords generations there are some very cheap, algorithms for face detection something like <a href="http://code.google.com/soc/2007/gnome/appinfo.html?csaid=7CCBE044F1515ED3" rel="nofollow">http://code.google.com/soc/2007/gnome/appinfo.html?csaid=7CCBE044F1515ED3</a> ). I think semantic desktop should be one of the main goals for gnome 3.0, but if you think about it as something that the user needs to work on you are totally in the wrong path! Yes you can ask him to tag one face but the next ones will need to be guessed by gnome.</p>
<p>Just my 2 cents, hope not to have spammed a lot&#8230;.</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: Mike</title>
		<link>http://kubasik.net/blog/2008/05/21/the-reality-of-semantic-desktops-death-to-tags-labels-and-folders/#comment-33477</link>
		<dc:creator>Mike</dc:creator>
		<pubDate>Wed, 21 May 2008 12:16:13 +0000</pubDate>
		<guid isPermaLink="false">http://kubasik.net/blog/?p=141#comment-33477</guid>
		<description>Isn't the idea behind a semantic desktop to make all this metadata about various things machine understandable, rather than providing yet another way of organizing things. The goal of which being to allow the computer to make inferences based on it, i.e. just know what the user wants.

Or am I confusing it with the Semantic Web?</description>
		<content:encoded><![CDATA[<p>Isn&#8217;t the idea behind a semantic desktop to make all this metadata about various things machine understandable, rather than providing yet another way of organizing things. The goal of which being to allow the computer to make inferences based on it, i.e. just know what the user wants.</p>
<p>Or am I confusing it with the Semantic Web?</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: Rob J. Caskey</title>
		<link>http://kubasik.net/blog/2008/05/21/the-reality-of-semantic-desktops-death-to-tags-labels-and-folders/#comment-33476</link>
		<dc:creator>Rob J. Caskey</dc:creator>
		<pubDate>Wed, 21 May 2008 12:05:44 +0000</pubDate>
		<guid isPermaLink="false">http://kubasik.net/blog/?p=141#comment-33476</guid>
		<description>I'd be a happy boy if I could just get at recently used files. And I mean the files that were _really_ recently used.

Jane Smith edits a file and gives it to me on USB? I put in the device and *BAM* it should be there in a Recent Files place.

I download a file? Where did it get saved, I don't care it's in Recent Files!

That email attachment from this morning? Yup, there too.

Last document I wrote? Recent files.

We get a filterable Recent files places "my files, john's files, other files, system files, etc" and I could eliminate 90% of my file-system chores.</description>
		<content:encoded><![CDATA[<p>I&#8217;d be a happy boy if I could just get at recently used files. And I mean the files that were _really_ recently used.</p>
<p>Jane Smith edits a file and gives it to me on USB? I put in the device and *BAM* it should be there in a Recent Files place.</p>
<p>I download a file? Where did it get saved, I don&#8217;t care it&#8217;s in Recent Files!</p>
<p>That email attachment from this morning? Yup, there too.</p>
<p>Last document I wrote? Recent files.</p>
<p>We get a filterable Recent files places &#8220;my files, john&#8217;s files, other files, system files, etc&#8221; and I could eliminate 90% of my file-system chores.</p>
]]></content:encoded>
	</item>
</channel>
</rss>
