
Geo-counter Frequently Asked Questions...
FAQ #17
THE TOP FREQUENTLY ASKED QUESTION OF ALL TIME
... it puts ALL the rest of the FAQs to shame!!! FAQ-wannabes!!! N[early]F[requently]A[sked]Q[uestions]!!!
I tried to register for a counter - I completed the registration form
correctly (no corrections were suggested), but the message returned was
"no more counters can be issued"
Answer 1:Go to
http://www.ualberta.ca/htbin/whoami
Rebuttal 1:But I am a U of A [staff, student] - I'm just trying to do this from my computer at [work, my friend's house, Texas, insert some foreign machine ISP here, etc, etc, ...]
Answer 2:What did http://www.ualberta.ca/htbin/whoami return?If it did not return "ualberta", you can't use the counter program.
Go to a machine that authenticates as "ualberta". [example: one of the LABS]OR, set up a proxy server to authenticate your desktop as a UofA client: http://www.ualberta.ca/CNS/PROXY/
If you are using an on-campus machine that does not return "ualberta" (maybe only numbers), then it has been set up wrong! -- see your local area administrator.
Rebuttal 2:Geez, Geo. You should have made that more clear.But I want to STAY AT HOME and do this. Now what?
Answer 3:Go to http://www.ualberta.ca/CNS/PROXY/, which will explain how you can set your browser so that you are identified as an on-campus machine.
Rebuttal 3:Oh.
Here are good alternatives that you can install:
The file that keeps the URL references gets cleaned out with a probability of 1 chance in 200 hits. That is, it resets itself.
You can reset the URL list and several other things (including the count) if you are using a registered counter (then "inspect" the counter, then "Register the changes", which tells you that is will also reset the URL list).
As for "moving your counter around" to another site. Go ahead. Essentially you are "stealing" your own counter. Just do it, OK?! Once a counter exists, anyone can attach to it - including YOU, from somewhere else! The only control you have over it is the stuff in housekeeping, namely a "reset".
It is YOUR responsibility to email a "thief" and politely request that he remove "your" counter from his page. Only after he does not comply, should you give me a history of your email correspondence, and I'll blacklist him.
The "magic" in this counter program is that it updates itself in "real time". It does this by relying on your web browser going back over the network to find a graphic. -- in this case an XBM graphic.
Browsers don't go back over the network to find additional text, unless the web page is on a computer that has allowed for "server-side-includes", which compromises security. And if your server allows "server-side-includes", then there's no need to use my counter program -- you can see how that's done at http://melmac.corp.harris.com/access_counts.html
When the browser calls my server to ask for the graphic to be sent, I do a bunch of code instead, and then only when the code is finished, do I sent back what the browser is expecting - an XBM graphic. This is possible since unix makes no differentiation between a file (such as a graphic) and a program. That's how I can handle this non-obtrusive "trojan horse".
Eventually, I'll add an attribute for "time zone offset" into the Registration program.
You can see your environment by trying http://www.ualberta.ca/WebSupport/DumpEnvironment.html
You will notice that there's no way to show the User's name.
Actually, for some desktop unix users, it is very possible for me to trap the name of the user via other techniques, but it is not universal, and certainly not possible for the majority of browsers, which are on Macs or Windows machines. When Browsers start telling the Web servers via the Environment variables, I'll report the user name.
That is, the only counters that you can "control" (resets, etc) are the ones you build using the "add a new counter" button from within the housekeeping program.
Your counter is a generic counter that was built WITHOUT the aid of the housekeeping program.
No new counters will be registered during the
UofA's prime time. (That's 8am-5pm, Edmonton time.)
or the GIF that is sent when you try!! And if you're getting the XBM, then you're doing OTHER things wrong, too, like not suffixing your counter with ".gif".
I haven't been keeping exact logs of why sites have been blacklisted, because I originally didn't expect that I would be blacklisting that many sites. I was wrong - there are lots. And because of that, I can't remember exactly why one particular site has been put on the list. That also means there's NO WAY I'll re-release a site.
Sorry about that.
A 9: Because of two things:
A 10: NO.
#!/usr/local/bin/perl
# Add the following lines anywhere to
# any perl counting script, and you too can
# track "last N users". All you'll have to
# do is use a little imagination on how to
# keep the log files organized, and how to issue
# a system("tail -n/tmp/somelogfile") command
# sometime later.
#
#
open(SOMELOG,">>/tmp/somelogfile"); # Append to end
# of existing log file
print SOMELOG "$ENV{'REMOTE_HOST'} ".
" $ENV{'HTTP_USER_AGENT'}\n";
close(SOMELOG);
#
exit;
If you would rather I add it to someone's code - fine. Send it to me. (As well as a cheque for $150/hr.)
Or, Here -- write your own from scratch.
... Also, why isn't the password blanked out?
A 11:
Pick a password, any password. It has nothing to do with any log-in of yours (nor should it).
If you don't want to build a password, you don't have to use the registration program.
Although it is trivial to blank out a password on a form, I thought leaving it visible would cut down the "duh, I forgot my password" email. It will also encourage people to NOT pick the password to match their standard machine or log-in passwords, which I'd rather not travel across the internet world.
As mentioned above, the "password" is an extra level of authentication to stop other internet users from stealing your counters - I doubt if any of THEM will be looking over your shoulder as you type.
A 12:
Quote from REVISIONS, Sep 25:
A 13:
I haven't built a "wildcard" structure into the &ignore feature yet. I suppose (via housekeeping) you could enumerate ALL possibilities -- in fact, some people have done this.
However, here's the dilemma... By ignoring a range of addresses that several clients other than yourself may be using, you ignore *their* hits, as *well* as your own.
I'll bet that you have an intuitive idea of how many times you have accessed your own page. I'd also submit that you'd like an idea of how many on-campus (via this dynamic addressing scheme) accesses have been made.
Bottom line, you can kludge an &ignore to everyone, but do some soul-searching about the value of the results if you do.
A 14:
By allowing you to arbitrarily set the number of accesses, I had two choices - change the "create date" to this value, or leave the "create date" as its true original.
The math for references per day is:
total count / number of days since create date.
If I keep the create date as its original, then people with
busy counters can circumvent my upper limit (500 rpd) by making the total count arbitrarily low.
If I move the create date every time you make a change to the offset, then I get people setting their number to huge (like you), making it to the "top 30 list", and breaking past the 500 rpd (as you have). Note: Since UofA counters are exempt (ie, no upper limit), this feature WORKS!! --- That's why it even exists.
Set the number back to zero, and let it count up naturally. I'm working on a new scheme, based on watching the "daily counts", rather than a nightly calculation of "references per day".
A 15:
It's likely you have never been able to do housekeeping on your counter. (resets, re-sync, higher limits, etc)
Unfortunately, because I'm not letting any more off-campus people build new counters, registering your name will do nothing for you.
Eventually, I hope to have a "claim" function added, which will let users turn un-registered counters into registered ones, but I don't have it yet, and probably won't for a while.
A 16:
Those (and several other packages) have a really nasty habit of wrecking <IMG SRC= tags when you pull the page back off of a server to make that baby, tiny, insignificant wording change.
The result is that your <IMG SRC= call to Geo-counter was probably changed from:
<IMG SRC="http://www.ualberta.ca/htbin/Geo-counter.gif?yoursite.yourfile.gif">
to something resembling:
<IMG SRC="Geo-counter.gif">
... the result is either a statically attached number that Netscape pulled down for you, or a broken icon.
... and you didn't even know it. Because you were working on a title change three screens back, and would NEVER have assumed that your trustworthy program would "optimize" your code for you.
See also: This description from the main Geo-counter page.
