Newer email programs are specifically designed to STOP the sender knowing if the email was received. That way spammers can't use invisible images in the email to tell which emails they send that have actually been opened. You can't tell if someone has read the email unless they decide to tell you.
(1) It depends on the client
(2) It depends on the user settings within the client
(3) It depends on the server
(4) It depends on the user settings on the server
I did state that the image solution is not reliable but it works - when it does. And if it never didn't, as we're led to believe, then why were these tools developed? And why are linked images still being sent?
Having said all that, one must remember that there are laws against spamming and if that's what the OP intends to do, then good luck...
Of course, if it is spam, the linked images are not likely to reach the intended end-reader and the issue becomes moot anyways.
Just last night - to some textent verifiable because I commented it in one of the threads - I sent out about 50 inivitations for a dinner party - HTML + linked graphics included. I realize that it is a very, very empirical sample but, nonetheless, everyone has received AND opened the mailed invitations, including graphics.
It is in the nature of forum posts to be definite and ultimative in the statements we make - note the "we", no one exempted here - but each PC is different, as is each user, settings vary across the spectrum, and internet knowlege goes from nil to expert.
I find it dangerous to say "it won't work anyways" because someone else just reading may come to believe that it's true and will perhaps be a little less careful based on such well-intended but nonetheless less erroneous statements. We'd like it to be that way, but, unfortunately, it isn't.
"The only way I know of - and it's not very reliable - is to send an HTML mail in which you include something like:
That script should obviously return an image but in the process, you can track the GET user."
All you are doing is adding some sort of identifier to an image in the email. The URL for the email can be to a valid image (like your logo) or to any random webpage. If it is a valid image, the request will not be blocked as often. This requires that you send a slightly different email to every user (the value for user would be unique for each user).
When the user views the email, your webserver records the "hit" of the image. In your web server's log file you will also have the value of "user". Run a report (there are tools available for this). If you do not have access to the log file, then you may need to go the "invisible image" route and make the webpage create its own file that logs the URL parameters.
In gmail, or any email application like Thunderbird, when user has received an email, he has an option to "Show remote content". Only when the user allows remote content to be shown, u will be able to track that activity via false image. There is no other option to achieve this except, maybe http://www.spypig.com/