Devon Young’s tech blog is gone (Been 86’d—sur­ely a phrase, but I don’t know its mea­ning). This is sad, because there was a lot of inter­es­ting things there. Now, the...

Devon Young’s tech blog is gone (Been 86’d—sur­ely a phrase, but I don’t know its mea­ning). This is sad, because there was a lot of inter­es­ting things there.

Now, the good news: thank to Opera’s fan­tas­tic caching sys­tem, I can still read it: Opera does not refresh a page when re-opening a ses­sion and this is GREAT (he is the only to do that, as far as I know). This is the second time that this fea­ture helps me—it even doesn’t refresh if you close a tab and unclose it .

Now, thanks to Opera and Devon Young’s appro­val, I deci­ded to post some posts.

Remove 75% of the spam on Word­Press 2.1

ori­gi­nal post

[…] The spam­bots weren’t using the form to post any­thing, they were just directly acces­sing the wp-comments-post.php file. Which sur­pri­sed me. It didn’t sur­prise me that they were doing it. It sur­pri­sed me that Word­Press was allo­wing that. So I rea­li­zed I had to create a small change in the script, that should’ve already been coded in by the Word­Press coders before release. So at the start of the file, before any of the code that’s already there, I added this line:

if(strstr($_SERVER['HTTP_REFERER'],"devonyoung.com")) {

Then after all the code, I added these 4 lines:

} else { 	echo "You tried to bypass the form."; }

These 4 lines of code, redu­ced spam by about 75%. Yet, some spam­bots are some­how still post­ing com­ments. So I che­cked really quick and dis­co­ve­red that the ones still get­ting their posts through, aren’t using the com­ment form or the wp-comments-post.php. Huh? Now that sur­prises me. Anyone know how they can manage that? I know a lit­tle about Word­Press, but not enough to know how this works. Likely, the spam­bots are exploi­ting a secu­rity hole or a bug in the sys­tem. I’m using Word­Press 2.1.

Spam­proof email address, with CSS

I think this is a case of paral­lel inven­tion, since Devon Young doesn’t men­tion any ins­pi­ra­tion, but this had already been done by the Polish Lite­rary Moose (inter­view).

Any­way, here it goes:

p a:before { 	content: "soy"; } p a:after { 	content: "devonyoung.com"; }

This address is selec­table with Opera (and Opera only; others will use their eyes). It is the only one I know to do this. That might be worth inclu­ding in any Opera evangelism.

A Google limitation

I copy the whole of it, that’s better.

As if I wasn’t having enough pro­blem with their Feed­fet­cher bot. Appa­rently Google thinks www.devonyoung.com is equi­va­lent to devonyoung.com, because it applied my www’s robots.txt to my entire domain.

I used Google’s auto­ma­tic URL remo­val sys­tem to remove all refe­rences to my out­da­ted www. sub­do­main from the Google database.

I cli­cked the link tit­led “Remove pages, sub­di­rec­to­ries or images using a robots.txt file” and typed http://www.devonyoung.com/robots.txt in the appro­priate text­box. That robots.txt has disallow: / in it, so no craw­lers will bother with the defunct sub­do­main. Kno­wing that robots.txt lists what you want a spi­der to ignore, I expec­ted that what was lis­ted, would be what would be remo­ved. What else could it mean?

This mor­ning, I wake up to find that my entire devonyoung.com domain is mis­sing from Google. Try it, search for it. It’s gone. You can’t even google my name to get it. So appa­rently Google can­not dis­tin­guish bet­ween a www. and the main domain itself.

That was irri­ta­ting, but I thought that fixing this would be rela­ti­vely easy. I just go in again and type http://devonyoung.com/robots.txt (without www.) in the appro­priate text­box, right? No. This cau­sed Google to give me a notice that every single thing at my main domain will fur­ther be gone (again?), and not just the things I list in the robots.txt as ignorable.

What makes this even more pain­ful and annoying is this note at the URL Remo­val System:

all pages sub­mit­ted via the auto­ma­tic URL remo­val sys­tem will be remo­ved from the Google index tem­po­ra­rily for six months

This is more than fairly irri­ta­ting. For 6 months, my entire web­site won’t be lis­ted in Google. I only wan­ted the out­da­ted and defunct stuff remo­ved, and I get burned.

Oh, one more thought on this sub­ject. They never confir­med with me that I had any autho­rity to delete all refe­rences to the domain from their data­base. That just sounds like a huge secu­rity risk. I ima­gine I could sign up using an ano­ny­mous hot­mail address and remove all refe­rences to micro­soft domains from Google’s data­base. I bet Google would hear about that.

Sorry for the rant, but I thought you might all want to know about this for SEO reasons.

News­Fire limitation

I used it for a while (now I use Net­news­Wire). In case you don’t know, it is a cool-looking RSS rea­der for OS X (it is made by David Wata­nabe. Nuff said).

Bot­tom line: News­Fire does not res­pect 301 noti­fi­ca­tion (301 sta­tus code means file moved, so stop loo­king for it here. It is there now.).

Oh, by the way: Google’s RSS bot (Feed­fet­cher) don’t give a shit nei­ther. Yep, Google sucks on this, he ignores my Redirect permanent.

Devon has a nice name for these rude bots: Deaf user agents.