“- About the linking experiment: it won’t work for just any blogger — imagine the spam if it did.”

I was never suggesting to post content that was spam. I was suggesting to watch for something interesting to blog about — or create it yourself — and then watch what happens. Big difference there. And I do realize that not everybody can be a seed in your program which is part of what makes it fascinating to observe and study.

Think you might have misunderstood my intentions there. Bloggers are always looking to write about something that interests them or create content completely anew with no source links, that’s what they do It wasn’t an exercise in trying to game your system, it was an exercise in watching the flow from an original seed and seeing what happens.

As for the date/time thing not being possible?

1) Don’t most the blogs in your system report date/timestamps in GMT? I know this date/time thing is an issue with some RSS feeds because they don’t report this data in their feeds, but you could default to the bot found date in that circumstance. Yes, it can be manipulated, but it would be quite useful to have. I respectfully submit that “when your bot first discovers it” is really, really off in some cases. Perhaps this is something being tweaked in 2006?

2) “one article may provide reported facts” — this is the source I’m talking about. “relates those facts to some rumors about future moves” — this is still referenced from the source, an additional node which your program seems to branch as necessary. while another entry describes why this confirms their image of the company, etc.

Rather than talk hypothetically, let’s take an example right now:

http://tech.memeorandum.com/051222/p15#a051222p15
This is the whole Flock response thread (which I haven’t blogged, but a few others have). You have the Flock response and then beneath that the following:

Steve Rubel’s post: 2005-12-20T14:43:32-05:00 GMT
Roland Tanglao’s post: 2005-12-21 23:33 PST

Branch #2 to Paul Kedrosky: 2005-12-22 01:08 PST
#2a Somewhat Frank blog: 2005-12-21T07:57:00-06:00 GMT
#2b Digital Common Sense: 2005-12-21 20:58:27 +0000 GMT

Branch #3 to Techcrunch: 22 Dec 2005 07:09:56 +0000 GMT (2005-12-21 11:09:56 PST)
#3a Mathew Ingram: 2005-12-22 11:52:00 EST
#3b Broken Kode: 2005-12-20 21:44:15 +0000 GMT

##
I’ll stop there as you should get the idea. The RSS feed for each of those contained the date/timestamp and thus an option could be made that would sort based under the master seed by date, which was my recommendation. Yes, there would be an offset for timezone to get back to GMT per the specs and accuracy/reliability wouldn’t be 100% but neither is the current system.

3) Agree with you that this matters very little to non-blogger readers. Perhaps open up an API and let others like me with the skills hook in and develop this functionality? If I can track down the date/time information manually, a program can do it and resort the data

Just suggestions and things to make you go hmm (hopefully), do with them what you will

Take care, Gabe.

I see problems though with the time stamping/sorting idea:
1. I can’t even determine when an item is posted (publisher clocks may be wrong, or deliberately set early). I can only tell when my bot first discovered it, and that’s different.
2. More often than not, entries and articles offer different facets to a story. E.g., one article may provide reported facts, while a blog entry relates those facts to some rumors about future moves, while another entry describes why this confirms their image of the company, etc. What relevance does “posted first” have here?
3. Non-blogging readers care much less about this.

About the human editor idea: it’s something I may want to do. However, it will require a bunch of new technological features, a paid editor, and the need to communicate the change to readers and publishers. So it’s a big deal.

A few other random facts:
- Most “me toos” don’t become full headlines, they’re just “Discussion” links, which in the default view (without changing “Preferences”) is mostly hidden. So typical readers are already insulated from them. It’s mostly the info junkies bloggers that expand all discussions by default who run into them regularly.
- About the linking experiment: it won’t work for just any blogger — imagine the spam if it did.
- You mistyped the link to tech.memeorandum! (Putting you in good company, e.g. the WSJ!)

Anyway, thanks again, and hope you’ll see some improvements by Christmas…2006!