earthlink reader update.

Well, I don’t think that Earthlink should say it will take 10-15 minutes when it has been several days now since I last uploaded my opml export from bloglines. I would like all my feeds to be there please. Here’s what I got when I uploaded it again.

“We already know about 42 sources, so we’ve added them to Your Sources. There were 39 sources that were new to us, so we’ve sent our little robots out over the internet to go fetch articles for you. Give them about 10-15 minutes to retrieve everything, and you’ll have all 81 sources from your OPML file available.”

Also I noticed that my blog, managed to now have two ID’s in their system. 2007, and 4039. They aren’t checking for unique sites in their feeds database I guess? Could this be why I have been getting hit twice each time the earthlink robots come to my site?

[tags]freedbacking, earthlink, reader, aggregator, feeds, opml_import[/tags]

Aggregator Showdown.

I started writing this last week, but with a lack of energy, here’s the skinny on what I’ve written up.

Dave Winer reminds us the other day. “2/3/06: “Aggregator developers could sure use some competition!
I like server side aggregators more than desktop. In a multi-user aggregator ideally it should be customizable by the end-user. One that shows you your feeds how you like to view them, whether it be email style, linear, reverse, upside down, reverse print, etc.
It should also remembers what’s new since the last visit.
I decided I’d see just what the opensource world has to offer in web-based single and multi-user aggregators.
I really wasn’t too impressed from a selection point of view with a multi-user installation in mind.
most opensource aggregators aren’t “pretty” or even still in development.
2004 was last real opensource aggregator development time period.
I even tried to get some of the ones listed in google’s directory to work.
here is some of my results in demo form.

DATE NAME Code Users? In Devel? OPML? Working?
2004.02.14 rnews php single no yes yes – with manual fix
2004.11.06 TALAggregator python single no yes no
2004.08.08 planet python single no no not latest dev version
2004.12.16 feedonfeed php single no import yes

Installation notes.
I first tried the Gentoo ebuild version of planet which is quite old but worked. planet has a nightly snapshot of their devel repo, which is missing a python dependancy I can’t figure out.

TALAggregator lacks a logout button, which makes it difficult to create users, and use your own account to browse feeds. Also logging in doesn’t usually work properly. TALAggregator stores it’s password in an encrypted format in the mysql database.

rnews stores the passwords as plaintext in the mysql database and also lacks a logout button/link. It is also having issues when creating it’s tables in the database. Being able to create the user_prefs table, but not the links, and users’ links table. ex. “links”, “px_links”

Once I applied some force and a crowbar, I now have the sql to create the needed table. It’s support with importing of opml files works fine and also adds your categorized material correctly. And of course the interface is a tad bit clunky.

CREATE TABLE `links` (
`id` INT( 16 ) NOT NULL AUTO_INCREMENT PRIMARY KEY ,
`src_id` VARCHAR( 36 ) NOT NULL ,
`link` TEXT NOT NULL ,
`title` TEXT NOT NULL ,
`description` TEXT NOT NULL ,
`state` VARCHAR( 16 ) NOT NULL ,
`pubdate` DATETIME NOT NULL ,
UNIQUE (
`src_id` ,
title( 255 )
)
) ENGINE = MYISAM
CREATE TABLE `px_links` (
`id` INT( 16 ) NOT NULL AUTO_INCREMENT PRIMARY KEY ,
`name` TEXT NOT NULL ,
`main_link` TEXT NOT NULL ,
`rss_link` TEXT NOT NULL ,
`image_url` TEXT NOT NULL ,
`user_order` INT( 8 ) NOT NULL ,
`category` TEXT NOT NULL ,
`last_update` DATETIME NOT NULL
) ENGINE = MYISAM ;