We have been wanting to solve this mystery for years, and the other day, we finally did, so I thought I’d pass on the solution.
It started one day when Steve and I were sending emails to our buddy J. Steve was using Thunderbird, and J and I were using Outlook. For some reason, I looked down into the thread, and saw that in Steve’s messages, there were these weird strings of characters like “ï¿½”. They usually showed up in place of the apostrophe or at the ends of sentences. I asked him what the heck, and he said “I haven’t a clue,” and he thought I was nuts because in his sent mail, they looked just fine. We were confused, but we didn’t really search for an answer. We also noticed that it only happened if we used his Rogers address, not his VC one.
Then I started exchanging emails with my French teacher, also using Thunderbird, and she commented on them as well. She said they were replacing French accents. Since I remembered that, for Steve at least, it was a Rogers-specific problem, I emailed her from my Gmail (still using Thunderbird,) and the symbols vanished! But I didn’t want to not be able to use my Rogers email for sending French messages, so I got mad and started to search.
The solution was really easy to find. I just typed “ï¿½ in my emails”into Google, and Voilà! It seems that the problem isn’t on Thunderbird’s end, it’s something that some ISP’s are doing to emails that are using a certain character set. Non-breaking spaces, apostrophes and special characters get turned into a big pile of broken. This explains why the email looks just fine before it leaves, but goes splat on its way to its destination.
I haven’t sent any French messages lately, but I did send a test message to our buddy J, and he said it looks fine, so I’m assuming I’m all good. Steve also sent me some messages back and forth to my work address, which was fond of turning his mail into splat, and nothing went crash, so I think the solution worked.
Here’s the short version of the solution, copied from the link above.
The “ï¿½” is inserted when there are two or more consecutive spaces. It is trying to convert a space to a non-breaking space, but is using the wrong character encoding. Avoid putting two spaces after a sentence to avoid the problem. Here is a test message I sent to myself that shows the problem:
Test again.ï¿½ Test. test.
no periodï¿½ no period no period
three spacesï¿½ï¿½ two spacesï¿½ one space x
The problem occurs regardless of whether the checkbox is checked or not, and it occurs when the outbound encoding is UTF-8 or ISO 8859-1.
the characters are hex codes EF, BF, BD, which in UTF-8 happens to be the Unicode “replacement” character to be used when the receiver does not understand the encoding.
I have had this problem for a couple of weeks. Finally found the right combination of settings:
Set Unicode on both Outgoing and Incoming
Check box “When possible, use the default text encoding in replies”
Hope this helps others.
I wish it was as simple as cutting down on spaces. Another message says it eats special punctuation, and I second that.
We just had to check the default box and it seemed to resolve, but there’s how to find the settings area where you’ll have to go to fix it if necessary. TBird, why do you make your settings so…so…labyrinthine?
Hopefully this helps someone else. It sure made me happy.