Jump to content

pagoda

Members
  • Posts

    205
  • Joined

  • Last visited

Everything posted by pagoda

  1. hilt
  2. Greetings, As mentioned before the search engines don't so much care about much of this. That said, I like to use meta tags to record project and page information for my own purposes. It's pretty useful to put notes, or copyright information in there. I also might use these for client specific information. This stuff could also just as well go in comments, but I find it easier to parse with scripts (like perl) when it's in the meta tags. pagoda
  3. bubbles
  4. Satellite phone- just like the big guy in the red suit!
  5. I've just received notification that NORAD has spotted the Head Guru leaving the TCH Headquarters... He He... so much like Christmas Eve!
  6. I need a calmative! I'm all discombobulated! The suspense here is killing me... <Programs 991 into speed dial just in case...>
  7. nest (nothin' bettern a 4 year ol' thread on a lzay saturday...)
  8. That statement is true as far as I know. It is also true that when someone papers a blog with one of my images I see this trail as a link into my site. Now, that said, I am not certain how the blogs where this has been the case work exactly, and frankly I haven't really cared since it is has helped. My guess though, is that somewhere embedded in the script generated page(s) there is an explicit reference to my website other than in an <img> tag. This discussion has actually got me curious just how this is working in an advantageous manner in some cases. I think I'll look over the logs and see what I can discover... Hmm. This isn't obvious to me why some uses of my images appear in a link search and some do not. I was hoping to have something to say in short order but I'm going to have to look further...
  9. Greetings, A link in does help. As an example, do a Google search, something like: link:www.mysite.com Google is one such search engine where links into a site do affect ratings, so, it's a "Good Thing". Cheers, pagoda Whoops- didn't catch the total gist of your post. Yes, I see what you are pointing out. Yet the fact remains that, for example, on a blog setup where an image is linked to from your site it will be logged as a link as in the search example above. I see this all the time when I search links into my site using various link searches (not just Google). But, as I said, if this is causing you more headaches than it's solving, then it's worth blocking. When I do care about either blocking excessive bandwidth problems, or content I care to retain ownership and control over I use the tools mentioned in the previous posts and/or use watermarks (either obvious ones or steganographic watermarks with a tracking service), but this might be overkill depending on what your site aims to do. Have a great day!
  10. Greetings, Please pardon me if this is a rather contorted way of looking at this problem, but... I own a graphic and web design business. When I first put my site up in 2003 I noticed this same problem. I eventually switched my site from page based to CMS since the amount of content of I was serving started to get out of hand. What does this have to do with people horking your images? As it turns out, those links in were a major boon to my site's SEO. So, I ultimately redesigned the site structure, decided to not mess with the folks who were in effect stealing my copyrighted materials and write it off as an SEO expense. Is this a weird way to look at this? Perhaps. But it has really helped search engine ratings. As a result of these early incidents, I try to sprinkle new images in fixed directories once in a while so people (usually kids and teens wallpapering their blogs) set up these incoming links and ultimately help my SEO efforts. Just something to think about if SEO is on your radar. Of course when this does get out of hand threatening your resources (and your invoice from TCH) then you have to act. But if your site is trying to deliver goods, services or information to the public, then this might be a welcomed benefit. Happy New Year!
  11. Greetings All, Ok, I've made some minimal progress on this: -Using /usr/sbin/sendmail works, even when I send an email to someone outside of my domain (as I am going for). -Attaching to a socket (port 25) from within PERL does not work, using the same "To:" and "From:". (For those interested- a code snippet is at the end of this e-mail.) So- Am I to assume that the exim version of sendmail wraps my messages somehow and allows them to go through, but that doing the same thing directly through a PERL socket is verboten? Looks like it. I'm still hopeful someone will tell me otherwise- mainly because I am lazy and don't want to change all the socket references to /usr/sbin/sendmail reference (not impossible- just a hassle). pagoda In the example: $WEB_SERVER = "askmegraphics.com"; $SMTP_SERVER = "mail.askmegraphics"; Everything works fine until a "RCPT TO:" is issued to the open socket that is not a "local" address, i.e. not something@askmegraphics.com. > local($SMTP_SERVER_PORT) = 25; local($AF_INET) = ($] > 5 ? AF_INET : 2); local($SOCK_STREAM) = ($] > 5 ? SOCK_STREAM : 1); local(@bad_addresses) = (); $, = ', '; $" = ', '; # Translate hostnames to corresponding addresses and pack local($local_address) = (gethostbyname($WEB_SERVER))[4]; local($local_socket_address) = pack('S n a4 x8', $AF_INET, 0, $local_address ); local($server_address) = (gethostbyname($SMTP_SERVER))[4]; local($server_socket_address) = pack('S n a4 x8', $AF_INET, $SMTP_SERVER_PORT, $server_address); # Translate protocol name to corresponding number local($protocol) = (getprotobyname('tcp'))[2]; # Make the socket filehandle if (!socket(SMTP, $AF_INET, $SOCK_STREAM, $protocol)) { $Error_Message = "Could not make socket filehandle ($!)."; return(1); } # Give the socket an address bind(SMTP, $local_socket_address); # Connect to the server if (!(connect(SMTP, $server_socket_address))) { $Error_Message = "Could not connect to server ($!)."; return(1); } # Set the socket to be line buffered local($old_selected) = select(SMTP); $| = 1; select($old_selected); # Set regex to handle multiple line strings $* = 1; # Read first response from server (wait for .75 seconds first) select(undef, undef, undef, .75); sysread(SMTP, $_, 1024); # Initiate a conversation with the server print SMTP "HELO $WEB_SERVER$CRLF"; sysread(SMTP, $_, 1024); while (/(^|(\r?\n))[^0-9]*((\d\d\d).*)$/g) { $status = $4; $message = $3} if ($status != 250) { $Error_Message = $message; return(3) }
  12. Hi AndyB, This is where the my script is failing: print SMTP "RCPT TO:$address$CRLF"; sysread(SMTP, $_, 1024); /[^0-9]*(\d\d\d)/; if ($1 ne '250') { push(@bad_addresses, $name{$address}, $_) } else { ++$good_addresses } When the value of $address is equal to: -An e-mail with my TCH domain in it (i.e. info@askmegraphics.com), Then everything works as expected. -An e-mail address other than one with ???@askmegraphics.com, then it fail with the following message from sendmail: 550-server10.totalchoicehosting.com (askmegraphics.com) [209.51.134.50] is 550-currently not permitted to relay through this server. Perhaps you have not 550-logged into the pop/imap server in the last 30 minutes or do not have SMTP 550 Authentication turned on in your email client. Note: $address is the e-mail address being sent _to_. It apparently does not matter where I claim to be sending _from_ so long as the _to_ is an address in the askmegraphics.com domain. I can reliably reproduce this result by manually telnetting to the sendmail port and talking to sendmail. So I must be confused about how to send mail _out_ via a script. Note- my mail goes out just fine from my POP account. Also- the first thing I do to identify myself to sendmail is the following: HELO askmegraphics.com (I also tried explicitely saying "HELO server10.totalchoicehosting.com"). I'm sure this is just a simple misunderstanding on my part. Any ideas? Thanks! PS: I'm using mail.askmegraphics.com to create a socket to the sendmail port. pagoda
  13. Greetings, I'm using FormHandler.cgi from the CGI/Perl Cookbook to handle forms on my website. When a user enters information, two e-mails are supposed to be sent, one to my business and one to the user to confirm the e-mail. The e-mail to me arrives just fine, but the e-mail to the user does not . sendmail is saying that I am trying to relay (wise of it to notice!), and obviously this is not allowed. Surely there is a way to accomplish this. Any ideas? Thanks! pagoda
×
×
  • Create New...