So I signed up for my two week trial with the intention of downloading a few of the Wier songs and then cancelling. But I still had about 20 downloads left. So I downloaded the entire EP “The Tain,” by the Decemberists, but instead of the five parts being individual songs, the entire EP is offered as one 18 minute song. So I downloaded some other random songs. While the selection is far from limitless, it’s certainly very deep and incredibly varied.

So, 24 songs later, despite my intention to rip off eMusic, I decided that they have earned my $9.99. The downloads go through the eMusic manager, which is really fast and really easy to use, and it’s cross platform. Also, did I mention the downloads are non-DRM, fully-portable, already-tagged MP3 files?

It’s easy to use services like mp3sparks and the like to get songs for cheap, but I don’t mind supporting my favorite artists when the price is right and I get to own a copy of the music that doesn’t impose random limits on me.

So, eMusic earned my $9.99 for what will eventually be 55 songs; I firmly believe that $9.99 for 30 songs is a reasonable price. So, if you’re so inclined, I encourage you to support eMusic. They are approaching things in a way that is actually right for the consumer.

So when I wrote OSNews, I respected Eugenia’s original decision to go flat. But the 15 comments per page is obviously driving by one thing and one thing only – increasing page views. Believe it or not, it’s much easier on the database to present all comments at once. For OSNews version 4, if you choose flat mode, it will be all comments at once on one page, period.

I added threading for two reasons: first and foremost, we had SO MANY users asking for it. I have the emails archived to prove it. It was the #1 most requested feature (along with comment editing, which I also implemented).

The second reason, to be honest, was because *I* wanted it. You see, readers tend to view things in the order they are presented, and they often respond without quoting a comment, so you tend to get comments like “Yes” or “I agree” or “Because the…” With threading, you can easily following multiple conversations. To me, threading makes much more sense than a flat view, because often times there are several sub-discussions in any given discussion.

So, I just completed a new section of OSNews, and for this section, I was completely unfettered. So I did some things:

• All comments are displayed in one page, period.
• Threading is mandatory
• However, threading can only be one level deep. In other words, you can reply to a new comment, but not to replies. To reply to a reply, you must reply to the original comment and discussion are tracked that way.

A lot of the crazy code that threading generates comes from supporting infinite replies. This will change in OSN4; not the structure of the database, just in the presentation (or lack thereof) of the reply/quote functions.

Also, for this new OSNews section, there is no registration. Any personal setting are done via cookies, nothing is kept on the server. This makes it easier to partake without deciding to pass it over because of registration. Of course, the issue here is that anyone can use any name and therefore impersonate someone else, but this is a small issue for a low traffic site. What I’m getting at is that this is a standard setting and not customizable like the views on OSNews.

So I remain mixed on threading. I understand that a flat view is generally easier to follow for the average Joe, but I simply don’t think that you should always cater to the dumbest possible user and build from there. There are those who do cater to the least technical user, there are those whose work suffers because they end up with a bland compromise, and there are those who simply man up and make the decision. Threading makes things easier to follow for *most* people with web experience, and that’s why I endorse it. Because the heaviest participation on websites is from those who are comfortable online, and therefore, will probably appreciate the threads.

We know that often times, once they make the front page, stories continue to receive diggs even if the site is unresponsive. Why? How can this be? This exposes a major flaw in the system – if the title and description are good enough, people will digg it, no matter what is on the other side. And that suggests we need two things.

First and foremost, we need bookmark functionality. Make my digg history almost worthless by allowing me to arbitrarily bookmark posts – seriously. This way, I can digg posts that I think should be in the front page without worrying that my history will be diluted. Maybe a bookmark (but not a digg) counts as a quarter-digg in the algorithm (or something like that). This way, I can bookmark a post to revisit it later without digging it now. I can choose to digg it later or not.

The idea of “digg” I think has always been to play the word “dig,” such as “Yeah, man, I really like this site. Ya dig?” How can you digg something you haven’t seen?

If bookmarks were implemented, posts wouldn’t necessarily continue to accumulate diggs even if the page goes down. Loyal readers suckered by a good headline, as I was yesterday when I kept going back to a posts entitled “Uber-Nerd Calls 911” which pointed to this clip from Reno 911, wouldn’t be forced to digg something just to remember to come back to it. The site went down mere seconds after being front paged, and then to rescue to the poor site, the video was temporarily removed and links to other videos were put in its place. Now, sure, there’s DuggMirror, but DuggMirror can be blocked by robots.txt, so it’s not a sure thing, and it requires me to remember or bookmark, rather than simply click a link on digg.com itself and have a pending list of links to revisit – my digg bookmarks.

This is very different from the recently introduced “My #1″ feature, which is mostly useless to me. This allows you to chose a story you’d like to feature, for some unknown reason, but only from the stories you’ve dugg. Furthermore, since I added something to “My #1″ the day it was rolled out mistakenly (due to a *Digg* coding error at that!), it’s stuck there, the first story in my myone section. I can’t remove it, even though I haven’t even dugg it.

Now, this could go the other way. While some may be more reserved in their digging once they can bookmark, it can be argued that adding bookmarks could potentially mean more diggs. Once you can segregate your important stories, a digg is easy to hand out with no limits. So the next step is that instead of the links going right to the site, the links should all be measured. I wouldn’t mind seeing a digg link point to http://digg.com/link/ae59f907b23 and then having my own personal “clicked link” history too.

On top of that, the number of actual clicks could factor into the algorithm. I’m betting that several people digg each front page story without ever having looked at it. Like all trends here, this is good and bad. It’s possible that I may feel as though this piece should be shared, particularly if it’s a headline like “Rumsfeld Quits!” that I know to be true whether I read the article or not. But for most stories, the number of clicks should matter. If the number of diggs paces the number of clicks, for most stories, this is suspicious. And I bet it happens a lot. This goes double for comedy videos or submitters who conjure great headlines.

Yes, I always check out the window.status onmouseover, but this could easily be replaced even if the link doesn’t point to the URL is redirects to. Perhaps this feature could even be toggled for the ultra paranoid who are freaked out about being monitored. I just assume that since HTTP is sniffable by anyone who uses TCP/IP and has access to routers, my click history – at least at digg.com – isn’t that sensitive.

The last big problem I see with Digg is the ever-growing stream of duplicates. Yesterday, someone submitted this genius piece. It was probably the 20th article about the Macbook Core 2 Duo release. Since it was a dupe, the author got around it by pointing to the site: http://www.apple.com/powerbook. Of course, since there is no Powerbook anymore, it redirects you to a meaningful page, but one that has already been submitted. This is silly – the author almost assuredly knew this was a duplicate story and purposely entered this URL to fool the digg engine. 2300+ diggs later, it paid off. Digg needs a better way of monitoring dupes. Wikipedia has an interesting practice – when a potentionally controversial entry is linked from an external source that drives large amounts of traffic in, the administrators lock down that entry for some period. Perhaps digg needs some site admins whose job is solely to monitor the site, a few smart operators to scan the Matrix. When there’s an announcement such as new Macbooks or Vista’s RTM, they look for dupes and simply blow em away. Sure, is is community run. But I’m sure the community agrees that softing through the same stupid announcment 20+ times makes digg.com/view/all/upcoming a real chore.

Digg is a great site, its only shortcoming is that various methods of using the tools provided to us make for results that are inconsistent. Furthermore, bad links can be, and often are, dugg up if the site is down but the title is good enough. A few changes to the core of the site would make the entire thing much more usable and would make the concept of a digg more valuable.

So, in summation:

1. Digg should add a bookmarking system independent from the digg system
2. Outgoing clicks should be measured
3. Hire community moderators to weed out duplicate posts