(I'm probably getting some details wrong here because this was all nearly 30 years ago, but anyway...)
The realtime hit counter and fascination therewith brought up a memory.
Back in '95 the regional dial-up ISP I used offered personal website hosting on one of their boxes (the standard http://www.isp.com/~username gig). I was very excited to put up a personal website. That was fun for awhile but soon I wanted to do more with it. I liked the look of hit counters, guest books, etc.
I became annoyed because this "CGI" thing I'd read about didn't work. I could FTP up Perl scripts named with ".cgi" filenames, however accessing them only resulted in seeing the Perl code.
I setup a webserver (NCSA httpd) on my Slackware box and started experimenting. I learned about .htaccess files and, more importantly, learned I could override the server's main httpd.conf with directives in my directory's .htaccess file.
A little experimenting with my ISP's web server turned up that, sure enough, they had not configured "AllowOverride None". Lo and behold I could enable server-side includes[0] in my directory-- including executing CGIs via SSIs!
Once I figured out how to chmod +x my scripts thru the FTP server I was in business. I had a hit counter! I had a script to check the referer (sic) and add a link in the footer back to the "main page" if you didn't "come from" there. I wrote out my own log files (because I didn't have access to the server's main log file) so I could see the IPs, user agent strings, and referers of visitors. (I also learned about file locking and parallel execution of the script by multiple requests occurring simultaneously. Fun!)
(The ISP didn't offer shell access but I figured out, pretty quickly, that I could write a rudimentary web shell with my SSI-based CGIs. I didn't do too much with that because I didn't want to get caught and I'd mostly gotten the desire to do "unauthorized pro-bono remote system administration" out of my system by that time...)
This little "site" my friend and I ran got listed on Yahoo (we submitted it... >smile<). I remember seeing a bunch of clients with reverse-lookups of the form "xxx.yahoo.com" that day.
Wired had recently done an article on Yahoo[1]. I saw a client named "srinija.yahoo.com" and remembered the article talked about a "Srinija Srinivasan" being an employee there (working on ontology). It was really exciting to see somebody named in a Wired article accessing my silly little site. (The most memorable client name from Yahoo was "ratbastard.yahoo.com", BTW. I'd love to know the backstory. Guess I already wrote about this once...[2])
Comments
(I'm probably getting some details wrong here because this was all nearly 30 years ago, but anyway...)
The realtime hit counter and fascination therewith brought up a memory.
Back in '95 the regional dial-up ISP I used offered personal website hosting on one of their boxes (the standard http://www.isp.com/~username gig). I was very excited to put up a personal website. That was fun for awhile but soon I wanted to do more with it. I liked the look of hit counters, guest books, etc.
I became annoyed because this "CGI" thing I'd read about didn't work. I could FTP up Perl scripts named with ".cgi" filenames, however accessing them only resulted in seeing the Perl code.
I setup a webserver (NCSA httpd) on my Slackware box and started experimenting. I learned about .htaccess files and, more importantly, learned I could override the server's main httpd.conf with directives in my directory's .htaccess file.
A little experimenting with my ISP's web server turned up that, sure enough, they had not configured "AllowOverride None". Lo and behold I could enable server-side includes[0] in my directory-- including executing CGIs via SSIs!
Once I figured out how to chmod +x my scripts thru the FTP server I was in business. I had a hit counter! I had a script to check the referer (sic) and add a link in the footer back to the "main page" if you didn't "come from" there. I wrote out my own log files (because I didn't have access to the server's main log file) so I could see the IPs, user agent strings, and referers of visitors. (I also learned about file locking and parallel execution of the script by multiple requests occurring simultaneously. Fun!)
(The ISP didn't offer shell access but I figured out, pretty quickly, that I could write a rudimentary web shell with my SSI-based CGIs. I didn't do too much with that because I didn't want to get caught and I'd mostly gotten the desire to do "unauthorized pro-bono remote system administration" out of my system by that time...)
This little "site" my friend and I ran got listed on Yahoo (we submitted it... >smile<). I remember seeing a bunch of clients with reverse-lookups of the form "xxx.yahoo.com" that day.
Wired had recently done an article on Yahoo[1]. I saw a client named "srinija.yahoo.com" and remembered the article talked about a "Srinija Srinivasan" being an employee there (working on ontology). It was really exciting to see somebody named in a Wired article accessing my silly little site. (The most memorable client name from Yahoo was "ratbastard.yahoo.com", BTW. I'd love to know the backstory. Guess I already wrote about this once...[2])
[0] https://www.oreilly.com/openbook/cgi/ch05_01.html
[1] https://www.wired.com/1996/05/indexweb/
[2] https://news.ycombinator.com/item?id=16741431