owatagal

I think the problem is in the backslashes--since the code is adapted from something where I was doing the reverse (stripping the www off the URL), the first backslash in the second line was there only to escape the period. See what happens if you take that backslash (and period) out: >RewriteEngine on RewriteCond %{HTTP_HOST} ^example\.com$ [NC] RewriteRule ^(.*)$ http://www.example.com/$1 [R=301,L] I tried this on a site that doesn't have any other redirects, and it worked ok. I hope it fixes the problem for you.

Which version of the rewrites rules in this post did you try? And do you have other rewrite rules in your .htaccess that might interfere with this? And is the rewrite rule itself not working (i.e. the site doesn't direct to www.example.com) or does the rewrite work but the login doesn't?

I find Page Rank useful because it gives me an overall idea of how my incoming links are doing. When I do a link search in google itself, half the links I know about don't show up--I heard somewhere (a man in a pub, right?) that google doesn't show all site links on purpose. So when my page rank goes up, I have a general sense that incoming links have also gone up, and I can comb through my stats a little more thoroughly to see where they're coming from. As a general sort of link meter, it's not completely useless.

That's unfortunate, although I suppose it's the tradeoff for having CPanel at all. Thanks for the info!

I use my computer's client to check email, and when I need to check online (for whatever reason), I'm happy going through CPanel and the excess logins--doesn't happen enough to irritate me. Is there a way to disable all the shortcuts, like www.****/sqmail or http://your-tch-domain.com:2082/horde/index.php'>http://your-tch-domain.com:2082/horde/index.php ? Since I'm not using them, I'd rather not have excess login boxes lying around. If I have to, I suppose I can put redirects in .htaccess for these pages --would that work? It's not as good as getting rid of these pages altogether, but it would be better than leaving them in the open. How do I find out all the possible options for these webmail shortcuts? (like that you can access Horde via http://your-tch-domain.com:2082/horde/index.php -- I didn't know that until today). Can you even do .htaccess redirects when a port is involved? Is it possible to go one step further and delete Squirrel Mail and Neo Mail altogether? I'm guessing not, since they're bundled in CPanel, but I might as well ask. Then I'd only have to worry about the Horde short cuts.

Both posts help. If I'm reading the headers right, they just confirm that the email comes from Zap Zone. The good news is he's definitely breaking their TOS, so maybe they will do something.

It's recently come out that there is a plagiarist making the rounds on some online journals. Editors and authors are working together to get his work removed, and that all seems to be going well--generally everyone's working together to get this sorted out. I'd like to take it a step furhter and try to get in touch with his ISP and see if they will shut down his account. Granted, he isn't spamming, so it might be hard, but he's definitely plagiarizing work. Would an ISP even care about that? His email is not from his ISP--it's through Zap Zone (zzn.com) and is spoofing a legitimate organizaton. We'll get in touch wtih Zap Zone as well, and the organization that's being spoofed. Zap Zone might not do anything, but the organization might put up a notice. My main problem is I don't know what I'm looking for in the emails to determine his ISP. How do I figure that out? Any help would be much appreciated--I know there's no guaruntee we could actually get his account shut down, and he could definitely come back on any other account/email address he likes, but we don't have to make it easy for him.

Also, on the red-X for image problem, IE doesn't support PNG files. Transparent .gifs or .jpgs with the right color background would fix those problems. The extra white space under the heading is a probably from one of many known white-space bugs in IE. Since everything is within the container div, you could try floating stuff to see if that closes up the white space -- float the banner; if that doesn't work, float the nav. I didn't look at the CSS, but assuming they both take up 100% of the container width, that should close up the vertical white space. The extra white space to the left and right of everything probably has to do with IE's box model problem--they figure out width differently than everyone else. If you search google for "box model hack", you should get some quick fixes you can put in your CSS to take care of that. I agree, it sucks designing for IE. One day, people will stop using it. Ha. One day, I'll have a job, too. Actually, overall, I think the design looks decent in IE--at least it's still readable!

Some registrars will let you "backorder" a domain name for a certain fee (at GoDaddy it's $18.95), and if the name expires they try to grab it for you. There's no guaruntee you'll get it even if the current owner doesn't renew the name, and your money won't be refunded under any circumstances-- the owner renewing, the owner not renewing but someone else getting it, whatever. I know GoDaddy will let you reapply the $19 to another backorder name, to try to grab it instead, but they won't outright refund it to you. Other registrars might be different. So yes, it's possible, but you have to pay more for it and there's no guaruntee. I could speculate on the domain transfer lock, but I'll leave it for someone who actually knows the right answer.

Bruce-- Good idea-- I found a couple trends I hadn't thought about. I might have been thinking backwards in terms of how to promote this site, given its focus/audience. (of course, for every relevant search term, there's someone who reaches my site by looking for "drunk mattress" or "painting people drinking tables," so... that just proves people have weirder topics on their mind thant I do) Jim-- thanks for the link. I'll check it out -- sounds very useful!

Suckerfish Dropdowns This is sort of a drastic solution, since you'll end up rebuilding your menus on all your pages, but the code is good, it's a popular method so support is widespread, browser quirks are documented (and generally with workarounds). It's also easy to implement, and after your first page you can just copy/paste code anyway.

So, after a couple months of tweaking, I've pulled a top ten placement in several search engines (and currently top 1 in google--whatever they did to their algorithms, I love it) for a reasonably prolific term (400,000-500,000 results). I'm happy about that, obviously. It's also a term that I thought was a no-brainer for searches; it's one of the first I would use to find my sort of site. As it turns out, what I would use to find sites is not what everyone else is using, because I get very few hits on this search term. This may be a quirk of the industry; by and large, news of sites spreads by word of mouth and link swaps more than new hits from search engines. I'd still like to pull more visitors in from search engines. That's so silly of me, I know--next I'll be saying I want them to visit more than one page while they're on my site. So how do I find out what people are actually searching for? I could use the other search terms I might use to find similar sites, but usually I just... go by word of mouth and link swaps. And I wasn't very good at guessing popular terms the first time, obviously. I could use meta tags on my competitors' sites, but they would suggest, among other things, the term I already have. Any ideas?

For a ready-made program, take a look at Expression Engine. I've never actually used it--just looked at the trial version--but from what I remember it was more customizable than the others I looked at. I think with any ready-made CMS you're going to have to deal with some compromises--having 20,000 users sign up for the CMS and/or having them email the form to you and not directly to the database (unless you hack a form into it, where they can submit directly to the database), but at least EE lets you set up custom fields--it might make editing things after they get in the database easier. Take a look, anyway. They offer a free trial version, so you can set it up and see if it would work or not before you decide to buy it.

I've seen the same bottom-scrollbar problem in IE when using div tags that had scrolling content (I don't use iframes... I don't even know what they are or why they'd be used.) My solution with div boxes, though, is to just add about 20px of padding on the right-hand side. I'd assume you could do this with an iframe, as well. This allows the vertical scrollbar to slide into the box in that padding space, and lets the content display normally in IE (without a bottom scrollbar). At least it does for me. Of course, depending on your layout and how exact you have everything designed, this will screw up your pages in every browser except IE, thanks to the way IE handles box sizes. You'll either need to apply a box model hack to the CSS, to set a different width on the iframe for IE browsers, or you'll need to play with the iframe width in the CSS until you get it looking ok in all the other browsers. Neither solution is ideal, but there's not a lot you can do, because IE simply does not build boxes like every other browser. Hopefully it'll help that you know it's the width of the iframe you need to adjust, however, if you do find the 20px padding screws up your layout.

Right--which is why I'd mung the unlinked addresses. As an aside, is munging a common term for that? I picked it up somewhere, but I never really thought about whether that was the actual term or just slang.

In my experience, any address a visitor can see on the screen can be picked up by spam bots. This includes any that are supposed to be hidden by character encoding and javascript magic. Basically, if a browser can figure it out, so can a halfway decent spam bot writer. I'd second using a form instead--instant, easy way for people to send email. And if you want to give them the addresses as well, in case they want to write it down and mail you later, then provide them with unlinked munged addresses. Without the "mailto" in the code to alert the spam bot, they should pass without notice.

I double checked, but there's nothing in the .htaccess related to the protected directories. And if I access the protected directories, I can log in and get to the pages and/or not log in and get the expected 401 error. So it looks like it's just the bots getting 301. I rechecked the logs, and it looks like the 301 error is followed by a request to get 401.shtml. I guess they are getting the 401 error... eventually, anyway. Maybe something to do with the bots being set up to index the "not authorized" page instead of the page they can't access? Intesting.

I noticed that it looks like your meta, break, and image tags aren't closed. They all have to have trailing slashes in them to be valid HTML transitional (I'm pretty sure they do, anyway): ><meta name="" content="" /> <br /> <img href="" width="" height="" alt="" /> Right now yours don't have that closing slash. That might be generating the error. Edit: No, sorry, that's an XHTML, not HTML, requirement. My mistake.

Thanks for checking--I'll tell her to call her ISP and sort it out with them. Good info on the .ico images too--that's easy enough to do! I'd still be interested in the 301/401 error thing if anyone has any ideas. It doesn't affect anything, it's just my curisoity and wanting to understand how things work. Thanks again!

A couple questions-- I have some password protected directories, and bots are trying to index it (I'm working on a robots.txt to keep at least the legitimate bots away from there). When they try accessing, they're getting 301 errors; I would have expected 401. Any ideas why they'd get 301 instead? I'm just using the CPanel directory protection, nothing fancy. Is there a way to fake a favicon.ico file? Not having one is clogging up my error logs, and last time I checked the online ico generators they were down and/or charging. I'm not that interested in an actual ico file, but I do want the errors to disappear. And I have a friend swearing she can't access my site. She didn't say if she was getting error pages or just not connecting, but when I checked the raw logs it looks like her last visit was March 3, and that visit was normal. Since then, nothing for her IP address--so I assume she isn't connecting to the site at all, even to generate error pages. I think it's her ISP, because I can access just fine, but if anyone has a second can you tell me if http://millertarized.com and http://millertarized.com/blog/ are visible to you? Thanks!

All Mac browsers -- In Firefox I see the home/forums/contact links right-aligned, in white boxes, with a grey background when I hover. The login/register is lined up sort of under the big red O (not flush with left margin, not centered in the red link area) and is just black text. In Netscape and Safari the right links look the same but the login/register is floating up and is actually OVER the big red O and some of that header' back ground image. They aren't in the red bar at all. They're also hard to see. In IE, the white boxes on the right menu items only come down about half way, and the register/login links are just gone. I don't know where they went. I'd get you screenshots, but I've forgotten what the shortcut keys are.

Assuming the files aren't CGIs, you can put them in the directory above all your /public_html/ files--create a new folder called /private/ or whatever you like and put the files in there. In file manager, you just create the new folder in the first screen that opens up. If you're FTPing in, make sure your FTP program isn't set to go directly to /public_html/, and then add the new folder at the same level as (not in) /public_html/. The path to these files (I'd assume you have to fill that in somewhere?) is /home/cpanelusername/folderyoucreated/ If they're CGI, I think you have to do .htaccess magic to protect them. At least, schussat was suggesting CGI scripts won't run above the /public_html/ folder response to a question I asked.

The PHP scripts are running fine in the the private folder, so it must be a difference between PHP and whatever runs in CGI (Perl? who knows. I never did CGI stuff). I'm able to do all my user authentication, data checks, database queries, etc, and keep the files for them out of the /public_html/ folder. Which I like. A lot.

Thanks guys! That's good to know!

Does it take any extra server resources to call on files that are above /public_html/? I've moved some password and permissions-based navigation menus there, and I'd like to move all my PHP scripts up--it's not so much that they're sensitive info as I want my scripts in once place. The files call them in as includes anyway, and I generally use a full path -- /home/username/public_html/folder/include.php -- so this wouldn't change anything in terms of the numbers of includes I use or how the scripts actually run. But if for some reason the server can include files faster or with less resources when they're in /public_html/, I can leave the bulk of my scripts there and play with .htaccess or something to block off a folder. I'd just prefer keeping it all together to spreading it across two sections of the site.