Archive for the 'FeedBlendr' Category

Heavy Load

Sunday, April 2nd, 2006

FeedBlendr has been experiencing a bit of popularity growth, and so I’m handling more feeds/blends now. This has started raising some eyebrows over at DreamHost (where this site is hosted), and fair enough – looks like I’m using a lot more processor time than is fair on a shared server.

I’m making some modifications to ensure that caching is working locally every time, and trying to avoid remote calls when possible because they seem to be causing the problem, so please bear with me in the meantime if there are any problems. If you’d like to do your part, then please reassess how often you’re requesting your blend – you really don’t need to be hitting them every minute (which it appears some users are doing!). In the future I may have to throttle certain blends if they are requested too often from the same IP and just block some requests.

On a brighter note, if you love FeedBlendr and want to support its future development (bandwidth costs, development time etc), then please jump over here and Make A Donation!

Please let me know via comments or email if something is broken from the changes that I’ve made, or the ones I make in the near future.

UPDATE 2006-04-12: Previous changes didn’thave enough of an effect, so I’ve made some more caching changes and will wait to see what effect they have on my processor load problems.

FeedBlendr Publicity

Tuesday, March 7th, 2006

FeedBlendr has been making the rounds recently and getting some hot linkage action from so great sites. One in particular that I noticed was this great little demo from Max Kiesler of DesignDemo, showing how simple it is to use FeedBlendr to blend together a couple of feeds.

Check out the demo (requires Flash).

New FeedBlendr Features

Saturday, February 18th, 2006

Well, being an overcast, kinda dull Saturday here in San Francisco, and being a little whacked out on cough medicine as I am, I decided it was a perfect time to roll out a few features that either people had asked for, or just that I had wanted to include. Here’s what’s just gone live:

  • Author elements of the output feeds contain the original author information from the source feed, and I make an extra effort to find an author’s name regardless of where it was stored in that feed.
  • The source feed details are now appended to the actual content of the post. This is intended to make it more prominent, and to make sure that it doesn’t overwrite author information in the case where there are multiple unique authors per blog.
  • I’m now storing the title of feeds, so that I can start building a public index of feeds involved in FeedBlendr’s engine (coming soon). I’m also using this in the source attribution process.
  • I’m removing (or at least attempting to remove) duplicate entries from blends now. As an example, if you blend 2 similar tags on Flickr, you should now only ever see each image once, even if it would normally appear under each different tag.
  • Updated the subscribe page to include information about the individual feeds included in a blend.

I have a few more features coming before I start serious work on the next one of my feed manipulation tools, but it shouldn’t be too far off now. Coming up soon are:

  • HTML (probably through JavaScript) output of a blended feed.
  • Public index of blends/feeds in the FeedBlendr index so you can find interesting feeds to check out.

Design Theft

Thursday, February 16th, 2006

Wow – that didn’t take long. The wonderful people over at appear to have liked the design that Ray did for this site so much that they stole it outright and changed a few colors before using it on their own site. Wonderful. Here’s the message I sent to them through their contact form, I hope they read it;

Hi, just wondering what you were hoping to achieve by completely ripping off the design of my website, which was created by my good friend Ray Hernandez. I mean you changed the colors a little and all, but seriously – give me a break.

You didn’t even bother changing any of the id or class names in the CSS for god’s sake, there’s still a “blendr” id in there.

Hope you feel great about stealing someone else’s hard work without any attribution or even permission.

Get a life.


A quick whois says that the domain ‘’ (which, incidentally, appears to be connected to another very dodgy-looking domain ‘’) is registered as follows;

Daniel Nerezov
4/38 Brae st. Bronte
Sydney, New South Wales 2024

Registered through:
Created on: 12-Feb-06
Expires on: 13-Feb-07
Last Updated on: 14-Feb-06

Administrative Contact:
Nerezov, Daniel
4/38 Brae st. Bronte
Sydney, New South Wales 2024

Technical Contact:
Nerezov, Daniel
4/38 Brae st. Bronte
Sydney, New South Wales 2024

Domain servers in listed order:

Auto-discovery of Feeds

Wednesday, January 25th, 2006

I almost forgot to mention another spiffy new feature that I added with the latest update to FeedBlendr. Now, if you enter the URL to a blog’s homepage and it has auto-discoverable feeds linked from it, then I’ll find them automatically, and use them to blend with.

As you enter a URL and jump to the next one, FeedBlendr will scurry off and go looking for feeds (checks if the URL you entered is a feed, if not, then tries to auto-discover one from there), and then if it finds one, updates that URL. Ahh the wonders of AJAXy goodness.

This should make it a little easier to add the feeds you want just by pasting in the address of the blogs you read.


OPML Feed Import

Wednesday, January 25th, 2006

More News! FeedBlendr now supports OPML by allowing you to either upload a file from your computer or specify a URL if you have an OPML file available online somewhere (HINT: If you maintin a blog-roll through then you can grab a URL direct to an OPML version of your blog-roll from their “Get Code” link).

Once you upload or link to your OPML file, FeedBlendr will read it and grab out all the tasty feeds you’ve listed, then go grab them all and blend them up for you. Just like before, you’ll get a single new feed to subscribe to which will now contain everything in your OPML file.

Please let me know using the comments or these contact details what you think of this!

GZipping Feeds

Wednesday, January 25th, 2006

I’m now GZ-compressing all output from FeedBlendr, so please let me know if this causes you any problems by commenting below.

This is an intermediate step in trying to reduce the amount of bandwidth required to serve up blends from this site.

On a related note, FeedBurner turns out to be a no-go as a caching/features partner for now at least, so I’m looking into as an alternative to at least give me some caching, even though they won’t give me any extra cool features like FeedBurner could have 🙁

FeedBlendr to start FeedBurning?

Thursday, January 19th, 2006

So FeedBlendr has only really been publicly operational for a couple of days now, and already I’m transferring around 120Mb of data per day. This is cool, but it’s also a little concerning. I love providing a service, but I don’t want to kill my hosting plan through abusing my bandwidth.

But wait, there’s hope.

The guys over at FeedBurner really know what they’re doing when it comes to managing feeds and providing added value. I got to thinking “gee, it’d be nice if I could manage all my feeds through FeedBurner, which would not only give me a break on my bandwidth, but it’d also allow me to add some cool features to those blended feeds.” Enter the FeedBurner Feed Management API! Genius.

I’m looking into what it’ll take to transfer everything over to automatically publishing all feeds created on FeedBlendr through FeedBurner, and adding some options for cool things like FeedFlare to blended feeds. Updates here as they happen…

UPDATE 2006-01-23: I’ve noticed that according to this post, normal accounts are restricted to 20 burned feeds. Obviously this would be a problem for FeedBlendr (there are already nearly 80 blends running), so I’m contacting the guys at FeedBurner to see if there’s something we can work out between us.

Introducing FeedBlendr!

Thursday, January 12th, 2006

I’m pretty happy to announce that FeedBlendr is now live, and ready for your blending pleasure. You may be wondering what this does, and why you should care? Well let me tell you.

I read a bunch of blogs and news sites via their RSS/Atom feeds, but I found that organizing them and managing individual feeds is a pain. I also realized that I never really went through the process of reading individual feeds, since I just checked all my current, unread items. I’m using the wonderful Vienna News Reader on my PowerBook, and I find that I only ever look in the Smart Folder ‘Unread Articles’.

That got me to thinking – why do I even bother categorizing and grouping all my feeds in separate folders, when I barely ever actually use those folders once I’ve configured them? I started thinking that combining feeds into a single, date-ordered stream would be pretty useful, and from there, I heard people like Dave Winer over at Scripting News talking about the idea of a River of News and that made sense to me. He was (is) talking about making a whole new feed reader to work like that, but I had already started working on bringing it to everyone, without requiring them to change reader.

Thus FeedBlendr is born!

Enter in a bunch of feed URLs, and then click the big ol’ BLEND button, and you’ll get back a single URL. From that new address, you can get a blended version of all of those feeds, ordered (hopefully) by the date of individual posts, so you can have your very own river of news (or tasty blog-smoothie as the case may be).

A simple tool that hopefully scratches a simple itch. Hope you enjoy it.