RE: [delicious-discuss] feature suggestion/request for discussion

2006-02-06 Thread Larson, Timothy E.
John Sullivan wrote:
 I really don't want every page I tag that happens to be hosted on a
 university's server to be tagged with the name of the university even
 though it has nothing to do with the university. Neither do I want
 every geocities page or blogger page to be tagged with geocities or
 blogger. This would add more noise to my tags. It would make it
 harder for me to find, for example, articles that are _about_
 geocities or blogger, where I wanted to use those tags as indicative
 of the subject matter. It also waters down the effectiveness of small
 features like tag completion.

I've always imagined, and thought I'd stated (but regretably not in the
original post), that this feature would be at the users' OPTION.  If I
hadn't made that clear, I apologize.  My bad.

If you don't want domain tag spam, you won't get it.  Those that would
find it useful could have it without it bothering you at all.  I see it
as a system: type tag that wouldn't exist in your regular tag space at
all - no noise whatsoever.

 I already choose to tag with the domain name when I feel it's useful.
 
 Surely at least part of the point of tagging is the fact that everyone
 can have a schema that works well for them. The job of the system
 should be to get out of the way, to enable that creativity. Over and
 over again we see people writing to the list with the schema that they
 personally use, and wanting to make it the system default.

Absolutely not.  But when a feature can benefit a large segment of the
userbase without impacting anybody else, why not add the capability?
Autotagging is one idea that could enhance the productivity/usefulness
of del.icio.us for those that CHOOSE to use it.

 del.icio.us's policy does not rely on thinking they know better than
 me what I want. Instead, it facilitates my ability to do what I want.
 You are asking them to implement something that would actually shrink
 the sphere of creativity in order to fit with an approach that works
 for you.

Absolutely not.  Del.icio.us could facilitate my ability to do what I
want (tag every entry with its domain) too, without bothering those
(like yourself) who prefer otherwise.  This is not a shrinking of the
creative sphere, it is an expansion of it.

This is what I thought was great about the five suggestions I posted.
Either the impact of the enhancement was completely optional, or
completely transparent to those that chose not to make use of them.  All
five could be added today and you could keep using del.icio.us exactly
as you did yesterday.  BUT if sorting differently would help you, you
could do that; if autotagging author/domain would be helpful, it would
be there; if you wanted to incorporate the title in a comment, you'd
be able to.


Tim
-- 
Tim Larson
a Contract Staffing Specialists consultant with
West Corporation, Interactive TeleServices
___
discuss mailing list
discuss@del.icio.us
http://lists.del.icio.us/cgi-bin/mailman/listinfo/discuss


RE: [delicious-discuss] feature suggestion/request for discussion

2006-02-06 Thread Larson, Timothy E.
Chris Lott wrote:
 I'm not into autotagging either, but the function of seeing all links
 I have from a domain could be interesting. Where in the UI it would
 belong, or if it could just be a search function or maybe a
 greasemonkey extension... I don't know.   

If I have dozens or hundreds of bookmarks in a tag and I'm looking for
one I vaguely recall, that I happen to remember I found on xyz.com, it
would be nice to narrow to the intersection of tag and domain.  

All domain autotagging is, at a fundamental level, is a way to make this
possible.  It doesn't have to be implemented as a literal tag!  You've
already got the domain (in the URL), it should be possible to take the
intersection of some tag(s) with that domain.  That's all.

It would be interesting to see which domains of my collection (as well
as across del.icio.us in general) are bookmarked most heavily.  But that
would be a by-product in my opinion, not the primary use.

Tim
-- 
Tim Larson
a Contract Staffing Specialists consultant with
West Corporation, Interactive TeleServices
___
discuss mailing list
discuss@del.icio.us
http://lists.del.icio.us/cgi-bin/mailman/listinfo/discuss


RE: [delicious-discuss] feature suggestion/request for discussion

2006-02-06 Thread Larson, Timothy E.
Chris Lott wrote:
 but it seems much better handled by search (efficient) than paging
 through a list of links to get to the middle of the alphabet! 

_If_ you happen to be one of those people that intuitively uses
searches.  Not everyone is.  Some people see a list, that's supposed to
contain what they're looking for, and will begin to scan it.

Another issue here has been addressed by others - the previous/next UI
for paging is inadequate.  The system knows what page you're on, you
should be able to directly jump to page 10 of 20 (for example) to get to
the middle of a list.

 Another thought: perhaps it also has to do with bookmark volume?
 Sorting also implies a desire/ability to scan the entire corpus for
 what one is seeking, since alphabetical sort by title (unlike, say, by
 domain name or tag) provides no other useful contextual data. That
 only works for pretty small data sets.   

Just because it works best for sets of  20 or so is a reason not to
provide the ability?

There are always going to be people visiting del.icio.us for the first
time.  They're going to be used to the other way of doing it.  Why not
provide them mechanisms they are familiar with?  Do they _harm_
delicious somehow?  If not, what is the resistance to something so
simple?

Reverse chrono is a completely non-useful organization in the general
application.  (OK, I can think of one specific application in which it
makes sense:  a what's new view.  But then I probably only care about
the 10 most recent, or the past day or week of entries, not the entire
list.)  Alpha has at least a marginal usefulness.

Tim
-- 
Tim Larson
a Contract Staffing Specialists consultant with
West Corporation, Interactive TeleServices
___
discuss mailing list
discuss@del.icio.us
http://lists.del.icio.us/cgi-bin/mailman/listinfo/discuss


Re: [delicious-discuss] feature suggestion/request for discussion

2006-02-05 Thread Scott Willsey


On Feb 4, 2006, at 8:30 PM, Chris Lott wrote:



but it seems much better handled by search (efficient) than paging
through a list of links to get to the middle of the alphabet!


Yes... and when they give us the ability to easily search ONLY within  
the current list of bookmarks in an obvious manner to those new to  
the site, and not the entire volume of links ever posted to  
delicious, then it will certainly be a lot less important.


Right now that doesn't exist. The tags are the only way (and are  
indeed a good way) but going back to the Logitech AV remote example,  
unless I put the word logitech as a tag, they'll have to hunt for  
what it might be unless they are familiar with my tags.



___
discuss mailing list
discuss@del.icio.us
http://lists.del.icio.us/cgi-bin/mailman/listinfo/discuss


Re: [delicious-discuss] feature suggestion/request for discussion

2006-02-04 Thread John Sullivan
Larson, Timothy E. [EMAIL PROTECTED] writes:

 You seem to think you know better than me what I want.  Maybe you don't
 want a feature like this, and maybe you think most people don't, either.
 And maybe you're right about that!  But as for me, including the domain
 and the author names automatically would save a bunch of time in the
 long run, because those are important metadata that I would use to
 organize my bookmarks.  If I didn't think this would be a useful
 feature, I wouldn't mention it.


I support del.icio.us's resistance to autotagging with domain.

I really don't want every page I tag that happens to be hosted on a
university's server to be tagged with the name of the university even though it
has nothing to do with the university. Neither do I want every geocities page
or blogger page to be tagged with geocities or blogger. This would add more
noise to my tags. It would make it harder for me to find, for example, articles
that are _about_ geocities or blogger, where I wanted to use those tags as
indicative of the subject matter. It also waters down the effectiveness of
small features like tag completion.

I already choose to tag with the domain name when I feel it's useful. 

Surely at least part of the point of tagging is the fact that everyone can have
a schema that works well for them. The job of the system should be to get out
of the way, to enable that creativity. Over and over again we see people
writing to the list with the schema that they personally use, and wanting to
make it the system default. 

del.icio.us's policy does not rely on thinking they know better than me what I
want. Instead, it facilitates my ability to do what I want. You are asking
them to implement something that would actually shrink the sphere of creativity
in order to fit with an approach that works for you. 

-- 
-John Sullivan
-http://www.wjsullivan.net/delicious-el.html
-GPG Key: AE8600B6

___
discuss mailing list
discuss@del.icio.us
http://lists.del.icio.us/cgi-bin/mailman/listinfo/discuss


RE: [delicious-discuss] feature suggestion/request for discussion

2006-02-03 Thread Larson, Timothy E.
ivan santiesteban wrote:
 This has been interesting. I like Timothy's push for more automatic
 tags. We already have system:filetype:, so it's not something entirely
 new.  
 
 As microformats grow these things could work really well for specific
 uses. 
 
 I think the only things you have to decide are a) wether you want your
 system to pull this information automatically, to ask the user to
 include a special system tag, or to do something in between; and b)
 how to make this tags visible.

As for actual implementation of autotags, I'm not positive it's actually
possible.  I haven't gotten to researching if javascript can be written
that will grab the correct elements, assuming they exist.  I assume it
would be, though, with modern browsers.

If a user decides he wants to use autotagging, he'd have to
recreate/edit his bookmarklets to have a form that grabbed the extra
stuff.  Or I suppose the bookmarklet could be standard, but the extra
data gets dumped on submission if the user's prefs say he doesn't want
it.  Domain autotagging would be done on the server, since that info is
already passed in the URL.

Tim
-- 
Tim Larson
a Contract Staffing Specialists consultant with
West Corporation, Interactive TeleServices
___
discuss mailing list
discuss@del.icio.us
http://lists.del.icio.us/cgi-bin/mailman/listinfo/discuss


RE: [delicious-discuss] feature suggestion/request for discussion

2006-02-03 Thread Larson, Timothy E.
Vinay Augustine wrote:
 I'm not sure how useful metadata is. I mean, metadata that's important
 to you (like the webpage URL) isn't important to me at all. 
 
 How does Joshua decide which pieces of metadata to include?

Certain metadata are nearly universal of any resource you'd look up.
Just for starters: it probably has a title, it obviously has a location,
it has a creator, it has a creation date, and there's the date you found
it.

 You're also asking for the original title to be kept, but can't
 *you* keep it? maybe the description field something like (original
 title) // (my title)

But then it's not the title.  It doesn't look or act like a title.
Having pre- and post-title fields for title extension/commentary allows
this to be displayed in the title line where it looks and feels like a
title, while still retaining the original title.

I'm willing to concede I'm probably in the minority on the title thing,
but I discovered it to be a much more convenient way for me to store my
bookmarks.  It works the way I think.  Most people are probably happy
with a simple comment field.  I prefer working the title into my
comment, if you will, and using a direct quote where others would put
their comment.  Very small addition, much greater utility for those that
choose to use it, no headache for those that choose not to.

 It's tempting to add metadata for things that one person finds useful,
 but a lot of people find a large number of things useful; I think
 del.icio.us is excellent in this regard because it presents a fairly
 minimal interface that one extend in ways that are personally
 meaningful.

How can a user extend it?  I can't add custom fields to the database to
store extra bits of info.  As mentioned above, overloading one field
with multiple types of data isn't an optimal solution.

 Given that you've given your coworker the title, I imagine that he'd
 just search for it. 

Most people use search as a last resort when there's comprehensible
navigation.  If a user sees a list of links that ought to contain what
he's looking for, he's going to start scanning it.  Alpha order aids in
that.

 I don't agree.  If there are aspects of the page that I'm going to
 use as tags anyway, why not give the option to have it done
 automatically? Autotag it! 
 
 
 There's been discussion on this list about autotagging, and I was
 under the impression that it was fairly taboo -- besides (referencing
 above), what tags should be automatically added? you mention URL
 (which would be easy enough, though I don't see the use -- you can
 even just type cnn or sourceforge as a tag; I imagine the

I know I can enter it myself - my point is that it's already known (in
the URL) and it would save time if I could tell the system to
automatically extract it consistently, so it's not always up to me to
remember to do it.  It's something I'd always want, but might forget to
do.  Same idea with the author - I'd love for all my links to be
searchable by author, especially if that (potentially) were done
automatically.

 tag-recommendation system would help you out there after a while) and
 author. With the latter, how do you extract the author? There's no
 simple, universally used author metadata; what do you do if it's not
 there?

I mentioned the meta name=author tag.  Not widely used, but it's
there, at least for HTML.  If a service like Delicious started to make
use of it, it could drive more people to use it correctly.  All the
people who have been thoughtfully including such metadata for years in
hopes of the semantic web would be vindicated when it finally becomes
useful.

 Yes, but what about all the *other* meta data he would add at people's
 requests? I don't disagree that metadata is useful, but it doesn't
 seem overly onerous to tag those pieces of metadata that you find
 useful, especially given that it might change from URL to URL.   

Of course people are going to have different opinions.  But I think
there can be a sensible middle ground.  Currently, Delicious provides
the bare minimum for a bookmarking system, plus a comment field.  I
think it could do much more with little effort, and be a better utility
for so doing.

Tim
-- 
Tim Larson
a Contract Staffing Specialists consultant with
West Corporation, Interactive TeleServices
___
discuss mailing list
discuss@del.icio.us
http://lists.del.icio.us/cgi-bin/mailman/listinfo/discuss


[delicious-discuss] feature suggestion/request for discussion

2006-02-03 Thread Hamish MacEwan
On 2/3/06, Larson, Timothy E. [EMAIL PROTECTED] wrote:

 Why not take advantage of some of the things the producer has already
 included in the content when you bookmark it?  There must be _some_
 overlap between what the producer and consumer think is noteworthy about
 the content, right?  That is to say, certain tags the consumer will use
 are, or could be, inherent in the page.

I think all your suggestions have merit and agree with many of
Joshua's push-backs.

In the end, I would personally prefer del.icio.us to remain simple,
and your ideas and others implemented, as best they can (which may be
better), either via posting scripts that use the API or extensions on
the client like Greasemonkey scripts.

Light-weight cores (preferably open) win after all:

this is why a heavy-duty core will always lose...by definition, it must
offer services which are of interest to only a subset of its users and
yet all users are impacted by them...  /mtr

Marshal T Rose on the X.400 mailing list (which actually used SMTP)
explaining the demise of X.400, but I find it has much wider
application.


Hamish.
--
http://del.icio.us/Hamish.MacEwan
___
discuss mailing list
discuss@del.icio.us
http://lists.del.icio.us/cgi-bin/mailman/listinfo/discuss


RE: [delicious-discuss] feature suggestion/request for discussion

2006-02-03 Thread Larson, Timothy E.
Larson, Timothy E. wrote:
 I mentioned the meta name=author tag.  Not widely used, but it's
 there, at least for HTML.  If a service like Delicious started to make
 use of it, it could drive more people to use it correctly.  All the
 people who have been thoughtfully including such metadata for years in
 hopes of the semantic web would be vindicated when it finally becomes
 useful. 

Author autotagging for Firefox 1.5, using the META tag above:

javascript:(function(){if(document.getElementsByName){var authors=;var
metaArray=document.getElementsByName(author);for(var
i=0;imetaArray.length;i++){var
author=metaArray[i].content;if(authors.length0) authors+= ;authors +=
author.replace(/
/g,);}}location.href='http://del.icio.us/ChristTrekker?v=3url='+encod
eURIComponent(location.href)+'title='+encodeURIComponent(document.title
)+'tags='+encodeURIComponent(authors)})()

Delicious even recognizes the tags parameter and drops the values
right into place.

Tim
-- 
Tim Larson
a Contract Staffing Specialists consultant with
West Corporation, Interactive TeleServices
___
discuss mailing list
discuss@del.icio.us
http://lists.del.icio.us/cgi-bin/mailman/listinfo/discuss


RE: [delicious-discuss] feature suggestion/request for discussion

2006-02-02 Thread Larson, Timothy E.
Joshua Schachter wrote:
 1. Title extension.  In my own link collection, I like to expand upon
 the title a bit as a memory-jogger for myself, perhaps incorporating
 
 
 This is why the title is editable.

But then the original title is not preserved.  Just because the reader
thinks he can improve the title does not mean the author's title for the
page is not important as well.

 2. Sorting.  Having the display sortable only reverse chronologically
 by date added is horrible.  There should be options for alpha by
title
 and popularity (num users linking to page), at least.  
 
 Does alpha ordering provide anything that search does not? This keeps
 coming up and I'm not entirely sure it's all that sueful. 

A search is a filter.  I don't always want to filter.  I just prefer an
alternate presentation of the same information.  When scanning a list,
alphabetical order is a more natural style for many people.  Maybe
it's just conditioning because so many lists are alpha sorted, but if it
works for people, why not have it?  Since it keeps coming up, maybe it
is a hint that many people would find it useful even if you personally
do not.

 3. Display.  I like to get more on a page, preferably without
 scrolling. 
 How about an option to not display descriptions - rather, put them in
 the title attribute of the link so they appear when on mouse hover?
 
 Interesting idea.
 
 4. Domain autotagging.  I think it would be useful if the domain of
 the story were automatically used as a tag, so a user could quickly
 
 If it's automatic, then it is not a tag.  

I don't agree.  If there are aspects of the page that I'm going to use
as tags anyway, why not give the option to have it done automatically?
Autotag it!

Traditional web searches rely completely on what the producer has
supplied - the content.  Bookmarks give a limited ability for the user
to have some input on his browsing experience.  Delicious greatly
enhances this ability by letting the user easily categorize and share
the bookmarks.

I think the ideal bookmarking system as a collaboration between the
content producer and content consumer.  The producer makes something
worth noticing, the consumer takes notice and bookmarks it.  There must
be something inherently interesting/noteworthy in the content, otherwise
the consumer wouldn't bookmark it.  

Why not take advantage of some of the things the producer has already
included in the content when you bookmark it?  There must be _some_
overlap between what the producer and consumer think is noteworthy about
the content, right?  That is to say, certain tags the consumer will use
are, or could be, inherent in the page.

For example, the domain.  People's minds are funny things - sometimes I
recall I read an interesting article last week on cnn.com, but darned if
I can remember what it was about.  Sure, I _could_ put cnn.com as a
tag, but if the system could (optionally) do this automatically for me,
why not?  It sure would be convenient to go to delicious/user/cnn.com to
quickly find the page I remembered.  This would be very useful to me.
In implementation, this autotag would have to be checked on updates to
make sure the domain didn't change (maybe the original link went 404 and
it moved to a new host) but that wouldn't be terribly hard - or maybe
handle it as a virtual tag based on the saved URL, which is obviously
always correct.  It might be useful to flag these as part of a different
tag space so that the user-chosen tag cloud/list doesn't become
overwhelmed, though - easily done if it's a virtual tag.

Another example, the author's name.  I bookmark articles and commentary,
and could put CalThomas or DanKnight as a tag for their pages.  But
if the page already knew this information (and it usually does, just not
often in a standardized way), it would be awfully slick if that tag was
done for me since I'm going to use it anyway.  Many pages have a link
rel=author tag generally used for email addresses.  It seems to have
a partner in meta name=author that is perfect, currently in limited
use.  As popular as delicious is becoming, it could drive the usage of
this meta tag to increase as it would finally do something useful.

Another example, the author-supplied keywords from the meta tag.  Maybe
a good starting point for the consumer's own tags is the producer's
tags.  I probably wouldn't use this myself, but some people might find
this useful in addition to the current recommended tags.

 You can search for domains,
 though. 

How do I search for a domain?  I tried putting a domain I've bookmarked
from (like cnn.com) in the search box, but didn't come up with
anything.  This isn't nearly as nice an interface as just clicking tag
names.  Simplicity is why I use delicious.  If I wanted to search, I'd
go to Google.

 5. Favicons.  If domain autotagging is done, it wouldn't be too hard
 to store the URL to a site's favicon, if it has one, and then display
 
 We don't currently fetch anything on a store. This 

Re: [delicious-discuss] feature suggestion/request for discussion

2006-02-02 Thread Joshua Schachter



But then the original title is not preserved.  Just because the reader
thinks he can improve the title does not mean the author's title for the
page is not important as well.
  
The point is so that you can get back to the original document. The 
titles are very frequently unhelpful on many sites.


The creator owns the site, but the reader owns the bookmark. Please 
explain important in that context.



A search is a filter.  I don't always want to filter.  I just prefer an
alternate presentation of the same information.  When scanning a list,
alphabetical order is a more natural style for many people.  Maybe
it's just conditioning because so many lists are alpha sorted, but if it
works for people, why not have it?  Since it keeps coming up, maybe it
is a hint that many people would find it useful even if you personally
do not.
  
Again, I'm not sure that other things work like this is a good 
explanation - what is the mechanism and

use case in which alpha order helps?


I don't agree.  If there are aspects of the page that I'm going to use
as tags anyway, why not give the option to have it done automatically?
Autotag it!

Traditional web searches rely completely on what the producer has
supplied - the content.  Bookmarks give a limited ability for the user
to have some input on his browsing experience.  Delicious greatly
enhances this ability by letting the user easily categorize and share
the bookmarks.

I think the ideal bookmarking system as a collaboration between the
content producer and content consumer.  The producer makes something
worth noticing, the consumer takes notice and bookmarks it.  There must
be something inherently interesting/noteworthy in the content, otherwise
the consumer wouldn't bookmark it.  


Why not take advantage of some of the things the producer has already
included in the content when you bookmark it?  There must be _some_
overlap between what the producer and consumer think is noteworthy about
the content, right?  That is to say, certain tags the consumer will use
are, or could be, inherent in the page.

For example, the domain.  People's minds are funny things - sometimes I
recall I read an interesting article last week on cnn.com, but darned if
I can remember what it was about.  Sure, I _could_ put cnn.com as a
tag, but if the system could (optionally) do this automatically for me,
why not?  It sure would be convenient to go to delicious/user/cnn.com to
quickly find the page I remembered.  This would be very useful to me.
In implementation, this autotag would have to be checked on updates to
make sure the domain didn't change (maybe the original link went 404 and
it moved to a new host) but that wouldn't be terribly hard - or maybe
handle it as a virtual tag based on the saved URL, which is obviously
always correct.  It might be useful to flag these as part of a different
tag space so that the user-chosen tag cloud/list doesn't become
overwhelmed, though - easily done if it's a virtual tag.

Another example, the author's name.  I bookmark articles and commentary,
and could put CalThomas or DanKnight as a tag for their pages.  But
if the page already knew this information (and it usually does, just not
often in a standardized way), it would be awfully slick if that tag was
done for me since I'm going to use it anyway.  Many pages have a link
rel=author tag generally used for email addresses.  It seems to have
a partner in meta name=author that is perfect, currently in limited
use.  As popular as delicious is becoming, it could drive the usage of
this meta tag to increase as it would finally do something useful.

Another example, the author-supplied keywords from the meta tag.  Maybe
a good starting point for the consumer's own tags is the producer's
tags.  I probably wouldn't use this myself, but some people might find
this useful in addition to the current recommended tags.
  
This is wrong-headed. The domain is metadata. Tags are metadata + 
attention; people tag so as to
identify the things they feel are important, not what the publisher 
feels is important.


I'm fairly sure if I added the meta stuff or the domains or whatever 
to your account, you'd find it

vastly annoying.



How do I search for a domain?  I tried putting a domain I've bookmarked
from (like cnn.com) in the search box, but didn't come up with
anything.  This isn't nearly as nice an interface as just clicking tag
names.  Simplicity is why I use delicious.  If I wanted to search, I'd
go to Google.
  


It should Just Work. Perhaps something is broken.

I personally think a favicon can often be a better representation of a
page than a thumbnail.  Small images are not necessarily iconic.  Road
signs have distinctive colors and shapes to be iconic of their general
meaning - they'd be less useful if all were white rectangles with black
text.  Apple learned this lesson in MacOS X - though minimized items
showed as thumbnails in the Dock, it was hard to distinguish multiple
documents.  Overlaying the 

RE: [delicious-discuss] feature suggestion/request for, discussion

2006-02-02 Thread Angus Fraser
To add to the the discussion on wether search is satisfactory 
alternative to other ways of sorting.


I'm not interested in alpha sorting particularly but do think there is 
something exciting about being able to publish a list of links sorted by 
their importance/value to me rather than just when I added them to 
del.icio.us. If it was possible to arbitrarlily sort my links within a 
tag listing (so that that order is retained when I and others view the 
listing) it would be particularly valuable for me and potentially for 
others and  del.licio.us. It's an addition to del.licio.us' data set: 
not only did lots of people add this link but some of them bumped it up 
to the top of their list of links for this tag  for me I'm now 
publishing my opinion of the relative value of a link not just the fact 
that I discovered it/liked it  thought it was related to a topic/tag.


Of course its not obvious to me how to reconcile this with the 
beauty/simplicity of the current model where new good stuff takes 
priority and can be subscribed to via rss.


Angus

--

e: [EMAIL PROTECTED]

___
discuss mailing list
discuss@del.icio.us
http://lists.del.icio.us/cgi-bin/mailman/listinfo/discuss


Re: [delicious-discuss] feature suggestion/request for, discussion

2006-02-02 Thread Joshua Schachter

Yeah, this is definitely something we are thinking about.

There's a bunch of odd sub-effects floating around, though. What if 
while looking at tag X, you you put link A below link B. Then look at 
tag Y, which also contains links A and B. What's the expected ordering?


Perhaps I we should just let people star items? It'd certainly be a 
simpler UI... I've always been against it, though; why bookmark a bad 
item? I dunno.


Joshua

Angus Fraser wrote:
To add to the the discussion on wether search is satisfactory 
alternative to other ways of sorting.


I'm not interested in alpha sorting particularly but do think there is 
something exciting about being able to publish a list of links sorted 
by their importance/value to me rather than just when I added them to 
del.icio.us. If it was possible to arbitrarlily sort my links within a 
tag listing (so that that order is retained when I and others view the 
listing) it would be particularly valuable for me and potentially for 
others and  del.licio.us. It's an addition to del.licio.us' data set: 
not only did lots of people add this link but some of them bumped it 
up to the top of their list of links for this tag  for me I'm now 
publishing my opinion of the relative value of a link not just the 
fact that I discovered it/liked it  thought it was related to a 
topic/tag.


Of course its not obvious to me how to reconcile this with the 
beauty/simplicity of the current model where new good stuff takes 
priority and can be subscribed to via rss.


Angus


___
discuss mailing list
discuss@del.icio.us
http://lists.del.icio.us/cgi-bin/mailman/listinfo/discuss


RE: [delicious-discuss] feature suggestion/request for discussion

2006-02-02 Thread Larson, Timothy E.
Joshua Schachter wrote:
 Perhaps I we should just let people star items? It'd certainly be a
 simpler UI... I've always been against it, though; why bookmark a bad
 item? I dunno.  

Not good-vs-bad, but good-vs-great.  

Or a 1-2-3 rating, where 1 is your standard good enough to bookmark, 2
is if you're interested in this topic, definitely take a look at this,
and 3 would be OMG why hasn't everyone already bookmarked this one!?!?

I don't think I'd rate my bookmarks (I'd let popularity serve as a
rating) but some people might.

Tim
-- 
Tim Larson
a Contract Staffing Specialists consultant with
West Corporation, Interactive TeleServices
___
discuss mailing list
discuss@del.icio.us
http://lists.del.icio.us/cgi-bin/mailman/listinfo/discuss


RE: [delicious-discuss] feature suggestion/request for discussion

2006-02-02 Thread Scott Willsey

		I'm just going to jump into the fray here and offer my thoughts on the navigation issues:1. I definitely agree we need to be able to sort the list alphabetically if so desired, as well as chronologically by oldest first and latest first. Not being able to choose the sorting order is, well, silly. This is a very, very basic "given" in pretty much all applications these days. Especially ones where the user is the one entering the data. I mean, it's OUR data, right? Let us look at it however we want.2. The navigation with "earlier" and "later" and no ability to jump to a specific page of our bookmarks is also, well, as someone's kids might say "that's so '90's".  It makes no sense. You know how many pages there are, you show that, give us the ability to jump right to a page.3. The search works great - now let me search only within my own bookmarks. Again, a no-brainer. Search with the ability to narrow the focus is a no-brainer in modern app.Just some thoughts on some features I think this great app is begging for.Scott

___
discuss mailing list
discuss@del.icio.us
http://lists.del.icio.us/cgi-bin/mailman/listinfo/discuss


Re: [delicious-discuss] feature suggestion/request for discussion

2006-02-02 Thread ivan santiesteban
This has been interesting. I like Timothy's push for more automatic
tags. We already have system:filetype:, so it's not something entirely
new.

As microformats grow these things could work really well for specific uses.

I think the only things you have to decide are a) wether you want your
system to pull this information automatically, to ask the user to
include a special system tag, or to do something in between; and b)
how to make this tags visible.
___
discuss mailing list
discuss@del.icio.us
http://lists.del.icio.us/cgi-bin/mailman/listinfo/discuss


[delicious-discuss] feature suggestion/request for discussion

2006-02-01 Thread Larson, Timothy E.
Hi list,

I was almost ready to start writing my own bookmark manager last year
when I was introduced to del.icio.us - how fortunate for me!  It
provides an easier-to-use interface than I'd envisioned, which is a
major point in its favor.

However, it lacks a few key features which I think would be really nice
to have.  If the code were open-source, I'd have added them already, as
they are small compared to the hard work that has already been done.  As
it's not, I'm throwing them out here for discussion.  I couldn't find a
TODO list posted anywhere on the site, so please forgive me if these
have already been debated to death or are slated to be added.

1. Title extension.  In my own link collection, I like to expand upon
the title a bit as a memory-jogger for myself, perhaps incorporating the
title into my one-liner commentary about the page.  Before you suggest
description, I don't use the description for this, because that's where
I like to save a short quote from the page so that in case my link goes
404, I can search for a copy elsewhere.  For example, today's CNN
headline Postal shooter's former neighbor found dead might appear in
my collection as California Postal shooter's former neighbor found dead
next day in home where only the original title is hyperlinked.

2. Sorting.  Having the display sortable only reverse chronologically by
date added is horrible.  There should be options for alpha by title and
popularity (num users linking to page), at least.  If the above idea
were adopted, the sort by title should allow by actual page title (the
link itself) as well as the user-expanded title.

3. Display.  I like to get more on a page, preferably without scrolling.
How about an option to not display descriptions - rather, put them in
the title attribute of the link so they appear when on mouse hover?

4. Domain autotagging.  I think it would be useful if the domain of the
story were automatically used as a tag, so a user could quickly pull up
all links he has from cnn.com, for instance.  You wouldn't be able to
edit these tags, of course, as they are implicit in the URL, but
otherwise should work pretty much the same.

5. Favicons.  If domain autotagging is done, it wouldn't be too hard to
store the URL to a site's favicon, if it has one, and then display the
favicon for the domain in the bookmark entry.  One of the primary uses
of favicons is bookmark lists in the browser because they help with
visual identity; they'd be useful in the same way for bookmark lists on
the web.  Naturally, this should be optional for those who feel it slows
their page loads too much.

OK, that's it.  What do you think?

Thanks,
Tim
-- 
Tim Larson
___
discuss mailing list
discuss@del.icio.us
http://lists.del.icio.us/cgi-bin/mailman/listinfo/discuss


Re: [delicious-discuss] feature suggestion/request for discussion

2006-02-01 Thread Joshua Schachter




1. Title extension.  In my own link collection, I like to expand upon
the title a bit as a memory-jogger for myself, perhaps incorporating the
title into my one-liner commentary about the page.  Before you suggest
description, I don't use the description for this, because that's where
I like to save a short quote from the page so that in case my link goes
404, I can search for a copy elsewhere.  For example, today's CNN
headline Postal shooter's former neighbor found dead might appear in
my collection as California Postal shooter's former neighbor found dead
next day in home where only the original title is hyperlinked.

  

This is why the title is editable.


2. Sorting.  Having the display sortable only reverse chronologically by
date added is horrible.  There should be options for alpha by title and
popularity (num users linking to page), at least.  If the above idea
were adopted, the sort by title should allow by actual page title (the
link itself) as well as the user-expanded title.

  
Does alpha ordering provide anything that search does not? This keeps 
coming up and I'm not entirely sure

it's all that sueful.


3. Display.  I like to get more on a page, preferably without scrolling.
How about an option to not display descriptions - rather, put them in
the title attribute of the link so they appear when on mouse hover?

  

Interesting idea.


4. Domain autotagging.  I think it would be useful if the domain of the
story were automatically used as a tag, so a user could quickly pull up
all links he has from cnn.com, for instance.  You wouldn't be able to
edit these tags, of course, as they are implicit in the URL, but
otherwise should work pretty much the same.

  

If it's automatic, then it is not a tag. You can search for domains, though.


5. Favicons.  If domain autotagging is done, it wouldn't be too hard to
store the URL to a site's favicon, if it has one, and then display the
favicon for the domain in the bookmark entry.  One of the primary uses
of favicons is bookmark lists in the browser because they help with
visual identity; they'd be useful in the same way for bookmark lists on
the web.  Naturally, this should be optional for those who feel it slows
their page loads too much.
  
We don't currently fetch anything on a store. This may change in the 
future. I'm more interested in doing

page thumbnails instead, though.

Joshua
___
discuss mailing list
discuss@del.icio.us
http://lists.del.icio.us/cgi-bin/mailman/listinfo/discuss