The sites that keep on giving

Recovered from the Wayback Machine.

I used to have a server and hosted several friends, each of whom contributed to the upkeep. It was fun having the server, but a lot of work, too, so eventually I moved back to my current host, Hosting Matters.

I still hosted most of the folk through a re-seller account, but my finances can be iffy at times, and didn’t want people dependent on me when I may not be that accessible. So I helped the folks move — either to other resellers at Hosting Matters, or accounts elsewhere. One person was going to move to a Blogspot position, but she never did return–she was getting tired of the weblogging scene towards the end. I miss her.

I still have one person hosted at; normally a quiet person who didn’t take up too many resources. He’s had situations at home that have required all of his attention so his posting has been infrequent, but that’s okay. After all, we don’t need to monitor a dormant site.

Well, we never used to have to monitor a dormant site.

During my recent cleanup, I checked my bandwidth use and was astonished to see how much bandwidth had been used this month. Further checking showed that my friend’s site was hammered, badly, about a week ago and the high bandwidth use was still ongoing.

Well, after considerable research I think I found the reasons: hotlinking and a malicious attack. At least I have to assume it’s a malicious attack, associated with, but may not be arising from, forums at I say may not be arising from, because anyone can fake a referer from any location. And I say malicious, because I’ve never seen a referer spammer that obsessive with one site, to hit continuously.

I’ve since put a ban on anything from the one site, as well as a ban on hotlinking. I’m now running an active monitor on the logfiles for the site to see if the problem re-occurs.

It used to be when we’d take a break from our spaces that, as long as comments were turned off there wouldn’t be much of a concern about the site just sitting. Unfortunately, those days are over. Hotlinking is becoming more and more of an issue, as more and more people join forums, each wanting their own cute icon and, thanks to search engines, finding and picking images from your site. Then there are those from, who hotlink with the best of them, as Scott from Random Choas recently discovered.

More than hotlinking, though, referer spam is really becoming an issue. It wouldn’t seem like anything other than a nuisance, but as the experience with my friend’s site demonstrates, out of control spam applications or malicous behavior can quickly push a site’s bandwidth use above even that generated by a direct Slashdot link.

So what’s a person to do if they host their own pages but want to, or have to, take a break? How do you counter increasing abuse of our systems by outsiders who are ignorant, rude, greedy, or outright malicious, yet still keep the pages open for those who actually want to access our pages to read them?

I don’t know, but I imagine I’ll learn more about the problem as I continue to monitor log files.

(PS: if you haven’t looked at your bandwidth use recently, I suggest doing so on a daily basis now.)

Technology Weblogging

Bug reports

I had thought I’d configured SourceForge to email me with bugs, but evidentially it didn’t take. I just now noticed the bug reports out at SourceForge and am busily going through them.

Thanks for submitting these. And I’m also trying to get the patch system to accept patches.

Technology Weblogging

Wordform status

Thanks to some folks I’ve been able to identify out of synch stuff with the installation of Wordform, and a few other issues. I hadn’t even noticed bug reports at the SourceForge site, as I had thought it was configure to email me automatically with bug reports. No such luck, but better late than never.

I won’t be doing a second alpha release, though. I’m finishing up the functionality for this release, adding in the reported bug fixes, and then releasing the program as beta. At that time I’ll then be going through the WordPress bug lists to see which of the bugs and fixes also apply to Wordform–they do still share a significant code base. Once this happens, though, I think this will be the last of the shared effort between the two products, as they are diverging too much.

The beta test should have the install kinks worked out, so this is open to non-techs. However, I do not recommend this for production use. My plan is to have the first release of Wordform at SourceForge by month end.

One major change I will most likely need to make is to change the license. Currently much of the software I’m using is licensed GPL, and I wanted to license LGPL. The main difference is that I would allow those parts of Wordform I created to be used with closed source applications as well as open source. To be honest, I didn’t really care who used the software, as long as they found it helpful.

However, my license contradicts the other software holder’s license, and while normally this shouldn’t be an issue — each individual still maintains their licensing control over their specific pieces of software, and I had planned on detailing this in the license to accompany the beta software–the mix of code is diffuse enough to generate problems and conflicts. So as soon as I figure out how to change this in SourceForge, I will be switching Wordform to the GPL license.

Anyone have an idea how I change this in SourceForge?


Wordform: Alterations

Recovered from the Wayback Machine.

I am in the process of establishing flags and options for several hard coded references in the code. For instance, right now the code automatically closes all comments older than ten days, and I need to make this into a user controlled option.

I’m also making use of the /tmp space, for some of the metadata work as well as creating the static copies of the pages (the ’static’ mode in WordPress doesn’t create truly static pages; Wordform does). I need to make this configurable, and allow a person to designate a write directory…or /tmp if they prefer. Gallery does this, as well as most other tools that support writes of this nature.

For the metadata, I’ll have finished the API interface layer to RAP (RDF API for PHP), which can then be coded to a different RDF backend.

A last major job is to finish the data isolation layer. Working on this gets tedious at times so I only clean a block of code at a time. I’m working on my last block now.

Clean-up and metadata, and the release probably won’t be out until towards the end of this weekend.

In the meantime, if you have installed Wordform alpha-01 and have played with it, I would appreciate hearing from you. Especially if you have any feedback other than what I’ve received already. For those who have been kind enough to provide feedback, my appreciations.

Must add option to turn on, or off, enclosures in syndication feeds.