Monthly Archives: March 2005

Open Source and Money (Shame on WordPress and PhotoMatt)

Posted on by
1

One of the most difficult aspects of open source projects is money. Often times work is done on free time and project resources are paid for out of pocket. If you are really lucky you have some one who is paid by their company to contribute part of their time to project. Either way, the issue of money rarely goes away. It looks like Matt and WordPress are about to experience this in a very serious way.

Just a few days ago Matt announced the formation of WordPress Inc.. Virtually no details have be given out besides noting that Jonas Luster is the first employee of this new company. Naturally this raises several questions:

  • What is going to be the official relationship between the WP project and WP Inc?
  • How will the services and products that WP Inc. offer impact or relate to WP?
  • Will WP Inc. be making a closed source version of WP?
  • Are donations that were made to WP being used in any way to start WP Inc., or in any way

For the most part I’ve been perfectly happy not asking these questions, Matt seems like a decent guy and figured he’d try to more or less do the right thing. That situation seems to have changed. Over at Waxy.org there is an article about the WordPress site hosting search engine spam. In brief, there are entries at wordpress.org/articles/ that are nothing but pure search engine spam. And there are a lot of them, more than 100,000. Matt didn’t disclose this fact to the community, but did comment on it when it was brought up in the WP forum. Basically Matt was having a hard time keeping up with the money needed to keep things going and the donations weren’t covering enough of it. The forum thread is no closed, so everyone will have to voice their opinions on this somewhere else.

I’m disappointed that someone who should know better didn’t something so ugly on the web. I don’t like spam, I don’t want spam, and I sure don’t want more web sites spreading more spam. It is even more disappointing to see this happen in an open source project because the hope is that such projects have great deal of transparency. The last part of Matt’s comments also concern me:

… The money is used just like donations but more specifically it’s been going to the business/trademark expenses so it’s not entirely out of my pocket anymore.

It isn’t completely clear if donations are now being used to start WP Inc. or if only money from the spam is going to WP Inc.. Making money by spamming the web is wrong, it doesn’t matter if it is going to web the WP project or WP Inc., it is just wrong. If donations to the WP project are being used to start/fund WP Inc., well that doesn’t seem quite right either. Now there is a possible exception to this, if people who donated to the WP project are getting shares of WP Inc. then perhaps this would make things not look so bad.

I’m hoping this is something that Matt did in a moment of weakness (financial or otherwise) and moves quickly to fix the situation. Too bad he’s out of the country right now.

UPDATE 3:10pm 30 Mar 2005: Ug, it looks like things are even worse that then appeared originally, links to the spam articles are being cloaked on wordpress.org. I’m starting to feel dirty by the fact that I’m still using WP.

UPDATE 11:00pm 30 Mar 2005: There’s a response from WP Inc. employee #1, Jonas Luster. He doesn’t like everyone calling this spam, I think that is a mistake. You’ll have a hard time convincing most people that this isn’t spam, don’t even bother with that argument. He does touch on one aspect of the problem, Matt did this in such a way that it was hidden from WP users. I certainly would have not known about it waxy.org hadn’t mentioned it. Jonas’s claim that this isn’t spam because it doesn’t involve a third party is completely wrong, it does. It involves many Google users and everyone who has linked to wordpress.org, there by giving Matt the ability to do this in the first place.

It has been asked if this is worse than what Six Apart did when they changed their licensing. In some ways I think it is worse, at least Six Apart didn’t try to hide the fact that their licensing. On the other hand, WP is still open source and free. Difficult to call, but I bet there will be several comparisons and questions like this in the future.

UPDATE 11:30pm 30 Mar 2005: Something strange just popped into my head, would it be ironic if this issue got even more attention by being covered at CNET News.com, which is owned/operated by Matt’s employer, CNET Networks?

UPDATE 5:20pm 31 Mar 2005: New developments, Yahoo and Google have pulled the spam articles out of their index. The page rank Google gives to WP has dropped to almost nothing. This is also being covered by Slashdot (for better or for worse). Jonas also has another follow up entry about this.

Those who are complaining about the negative response seem to be focusing on the calls for transparency and that this wasn’t really spam (and therefore there wasn’t anything wrong with this). My complaints about the lack of transparency come from Matt’s attempt to do this without WP users knowing about it. Instead of making a big public call for support and help in keeping his projects going and listing search engine spam as one solution he could have saved himself some of the backlash that is coming from this move. As for not being spam, give me a break, that is all this is and that is why someone was willing to pay for these web pages. The folks who paid Matt for these articles are the same type of folks who try to post comment and trackback spam to my blog and send me hundreds of email spam everyday. It was spam, trying to avoid this term by calling it an “experiment” is an attempt to spin an obviously poor choice into to something less foolish.

Bloglines Troubles

Posted on by
Reply

As I’ve mentioned before, I’m pretty happy with Bloglines. It does (pretty much) one thing and does a good job of it. Nothing is perfect though. For several days I haven’t been able to read one of my feeds (a PubSub subscription for everything that contains the word PostgreSQL). Initially I figured I would just wait it out for a day or two, by then they would probably have caught it one their own and then everything would be fine. When that didn’t happen I used their contact form to make them aware of what was going on. I got an email back within a day indicating that they would “reset that feed” and hopefully everything would be fine within a day or two. I’m not sure what was meant by reseting that feed since the number of unread messages continued to climb.

So I waited for another couple of days, but still nothing. I should describe in a bit more detail what is happening. Normally when you click on a feed subscription in Bloglines it calls some JavaScript and displays the unread entries in the larger right hand frame. It then considers those entries read unless indicate otherwise. In this case what happens is when I click on my troubled feed the right hand frame gets an HTTP error 500 (an internal server error). So their web server got my request, tried to process it and ran into an error while doing so. The result is that I’m unable to read any entries for this feed and the number of unread entries continues to increase.

In an effort to be a little more helpful in explaining what is happening, I dissected some of their JavaScript to determine the exact URL that is being requested. I emailed them back indicating the feed was still not viewable and provided them with the exact URL to use (when logged in as me) to reproduce it. I sent that last night and I still haven’t heard back from them. I can sympathize that such a popular service is growing like mad and being recently purchased by a new company will cause a lot stress and difficulty on a company. But this shouldn’t be rocket science. I’d expect that they’d be monitoring their systems so as to be notified as soon as possible when one of then servers falls over or starts generating errors.

Please Bloglines, don’t let me down.

Oh, and I still haven’t received my I Love Bloglines shirt yet, even though it was supposed to be mailed out on the 14th of March. If I don’t see it by the end of the week I’ll send another email asking if it ever got sent. Can’t complain too much, the shirt is free after all :-)

Begone Nofollow

Posted on by
Reply

I’ve ripped out all of the nofollow tagging that WordPress was doing to the links in the comments. So feel free to comment away with the confidence that you’ll still get your Google Juice. Why did I bother doing this? Because it doesn’t work, I was still receiving hundreds and hundreds of spam comments after my upgrade to WP 1.5. So leaving it in there was only hurting those legit comments. I’ll say it one more time, if your solution doesn’t PREVENT the problem, then it isn’t really a solution.

A Practical Use For AJAX

Posted on by
Reply

The new Google Maps is cool, and so are many of the other sites out there venturing into more client side work (AJAX, if you must, although calling it new is a bit of stretch). But lets face it, how many of them are really practical? One could argue that Gmail is using this for some real down to earth features, but I think you’ll find that in the minority instead of the majority. What I’d like to see this technology being used to replicate more the “desktop software” features in web applications. The example that came to mind is writing blog entries.

I’m using WordPress to manage my blog. It does a nice enough job for most things, although I still have my gripes. When I first start an entry in WP I will generally save as a draft after the first few lines to make sure that if something happens I can at least go back to that point. And I’ll often hit the “Save and Continue Editing” button while composing the rest of the message, for the same reason. If something goes toes up while I’m writing my entry (like my network connection), then I want to keep the loss to a minimum. Hopefully you can already see where I’m going with this.

With using these client side requests (AJAX), why can’t we have an autosave feature like with commonly do with word processors? That should reduce the risk of data loss while editing things online quite a bit (depending on how often you set the system to autosave). If an error occurred when autosave was doing its thing, then it could pop up a little javascript alert window and let you know that there was a problem.

I haven’t fully flushed out this idea yet, let alone done anything that would resemble a proof of concept, but it seems like it should be doable. Who knows, maybe Google will build this into the compose message part of Gmail.

Don't Mess With My Xserve G5

Posted on by
Reply

I’ve been looking at a couple of different possibilities for expanding our file storage capacity at work, specifically I wanted to see if I’d run into any problems using FreeBSD 5.x with an Apple Xserve RAID using the Apple Fibre Channel card. So back on Thursday (17 Mar 2005) I took the spare Fibre Channel card and put it in a test box and broke the mirror on our current Xserve RAID system (connected to an Xserve G5) and plugged it in a tried it out. The good news is that everything seems to work just fine. FreeBSD detected the card using the mpt(4) driver and found the RAID array without any problems.

The Apple Xserve G5 was up and live the whole time, so it started beeping and sending out notifications when we unplugged one portion of the mirror that I used to test FreeBSD with. So after I started running my tests I walked over to server room again to push the system indicator light/button on the front of the Xserve G5 (just to the left of the USB port on the front panel). So I walk in the room, hold the button down for about a second and all of the sudden the lights in the room turned off and all of the UPSs started beeping madly. After about of second of the power being off the building generator kicked on and we had emergency power running (nice to know the generator works), so all of the UPSs went silent again.

I freaked! I thought maybe I had some how killed a circuit by pushing this button on the Xserve G5. I walked out of the server room and discovered that the whole building was out. A minute later we discovered that the whole campus and surrounding area lost power, so it wasn’t anything that I did. I’m not sure if someone took out a power pole or what, but the power came back on after twenty minutes or so. Just to make sure I went back and pressed the system indicator light/button to see if the power would stay on. Fortunately nothing happened, so it was just coincidence that I happened to have pushed that button at the some moment that good sized power outage happened.

Of course it doesn’t make any sense that pressing that button would cause a power outage, but I was sure spooked there for a minute. Don’t mess with my Xserve G5 or I’ll turn the power off again :-)

Open Source Code From Google

Posted on by
Reply

Just in case you haven’t been keeping up with all of the Google announcements lately, here’s another code.google.com: new open source. They’ve released four projects so far at code.google.com, that apparently are used internally at Google and are still in active development. This first batch of projects are all being released under a BSD license, nice to see. A nice start, although it seems like the Yahoo Developer Network is a little more well rounded, even if it aiming for a slightly different target. I’d love to see Yahoo and Google duke it out to see who can best support developers.