On Wednesday 08 December 2004 04:09, Sean wrote: > HTML is good enough for web pages, its good > enough for email HTML is _the_ internet standard for web pages. It's _not_ a standard for email. There is a very simple rule for asking questions on this list (or any others): Send your email in a format that the highest number of potential respondants will read. I don't recall anyone saying that they _expect_ people to email to them in HTML, but a good number of people say that HTML is undesirable or unacceptable. _I_ require that email is rendered apprpriately by the email client I've chosen for reasons that seem good to me. When I remind people to post no HTML, I do so because I cannot read their mail without skipping over the HTML myself. Your questions are not that important to me. What my email client does with plain text is quite irrelevant: I can customise my GUI browser to display in a font and size that suit me, it recognises imbedded links so I can click on them, but those links are recognised because they look like links, not because some (possibly invalid) markup language says they are links. Here are some potential HTML nasties. I don't pretend that these apply specifically to email sent to this list, but email clients set to render HTML from one source likely do so from all. web bugs. These are little (one-pixel) images included in email. If you open the document, the image automatically loads and the link a) validates your email (spam) and b) indicates you're "read" the email (think estranged party's lawyer). mismatched quotes Quoting characters include "", '' and <>. Typically, a programmer assignes a "big enough" space to store legal strings and maybe some fudge factor to allow for something longer. Then the programmer writes a loop to copy until he finds a matching quote. If the string is "carefully crafted," the program copies more then the programmer allowed for and writes on other stuff. The result is a "buffer overflow," and this is how many viruses can infect your computer. It's not just HTML tags, of course, that can exhibit this kind of problem: email headers can too, and as I recall metamail had just this problem a few years ago (on Linux no less). Corrupted files These can be images (see web bugs) or attachments. Something in the file is invalid, either accidentally or "carefully crafted" with the result that programs that don't validate the data can crash or do worse things. There are many things that can go wrong, and Broken Windows is notoriously prone to these problems, but Linux (and the BSD family) are not immune to these. Turning off HTML rendering if you can is one way to avoid some of these problems. kmail has HTML off by default and I do not intend to change it. -- Cheers John Summerfield tourist pics: http://environmental.disaster.cds.merseine.nu/
