Olin Lathrop wrote: >> So if you know that your email setup is broken. > > It's not. I just doesn't use all the standards you think it should. Your mailer sent out an email in a charset that you say it didn't display correctly. I call that "broken"; broken in the sense that it allowed you to send something that you didn't see how it would look like on a reader that supports this charset. You basically didn't know how it looked what you were sending. If that's not broken... >> You really should take a read on Mime (it's in one of the email RFCs) >> before talking about email standards. It's the only standard for email, >> and the font I used is one of the not so many that are in widespread >> use and well standardized (for a while now). > > MIME is a standard about how to encode things that are not 7 bit ASCII > in 7 bit ASCII. This is a part of it. This is the transfer encoding. > It has a way of specifying which font to use, but it does not itself > specify those fonts (unless I'm mistaken, it's been a while since I read > over that stuff). You're mistaken. Mime doesn't allow you to specify a font; it does, however, allow you to specify a character encoding (the "charset"). And it specifies exactly which charsets are allowed. It's a rather small selection. > In other words, this discussion has nothing to do with MIME. It has a lot to do, as the problem of your reader seems to be that it doesn't handle the "charset" part of the Content-Type field correctly. What this field means is part of the Mime definition. What it has nothing to do with is your complete lack of knowing what you're talking about. (Don't you ever get tired to make wrong statements as if they were the ultimate truth?) >>> I don't know what encoding you used, >> >> Why don't you? I've stated this repeatedly in this thread now, and it's >> the same encoding you used: >> >>>> Mime-Version: 1.0 >>>> Content-Type: text/plain; charset="iso-8859-7" >>>> Content-Transfer-Encoding: base64 > > I didn't "use" that encoding. It came from your reader, and the message definitely had your "flair", so it seemed to me that yes, you were using it :) > I think what happened is that someone somewhere along the chain of > replies changed the encoding to base-64 when it was originally plain > text. This is not exactly what happened. When I used first the omega symbol, my reader knew it had to do /something/, so it switched to the ISO 8859-7 charset (and consequently to base64 transfer encoding) and added the appropriate Mime headers to inform the recipients of how this message is to be encoded. Your reader received that message. It seems that it doesn't know how to decode that charset (since you're using a Windows version, probably Win2k or later, the system is perfectly capable of displaying that symbol the standard fonts that come with it, so it is the reader) and showed you the message in a different charset. When you replied to it, your reader kept the charset -- even though it wasn't the charset you were seeing on the screen. (This is what I called "broken" above. Now you tell me that isn't broken in your book? I'm sure that if you hired me to write some software and it did something like that, you /would/ call it "broken" and refuse to pay until I fixed it :) > Anyway, what I meant with "encoding" was what character code you used for > the omega symbol. I mentioned this now several times over: ISO-8859-7. It's very clearly spelled out in the message headers of all those messages that contain the omega symbol. > I'm pretty sure the normal 8 bit (not ASCII) character set does not > include the omega. There is no "normal 8 bit character set"; there are a few standard ones, a few manufacturer-specific ones, but there is only a small set that is allowed by the Mime spec in the charset field of a text/plain section. You really should read the Mime spec before talking more about it. This is getting tiresome. http://www.ietf.org/rfc/rfc1521.txt FWIW, the 8 bit charset my reader was using (ISO-8859-7) does contain the whole Greek alphabet. > It does include a whole pile of silly letters used only by a tiny group > of people sitting atop a volcano in the north atlantic, plus things like > the german umlaut letters, etc. This 8 bit character set is the next > widest supported. You most likely are talking about either a Microsoft-specific charset (used by default in earlier Windows versions, 1252) or ISO-8859-1 (which is close to the Microsoft charset, but not identical). ISO-8859-1 is one of the charsets allowed by the Mime spec. > You can send just about anyone a german O-umlaut, for example, and it > will be correctly displayed. Beyond that, and the omega falls in this > category, you are relying on increasingly less universal support. Your system supports it. Just go to http://en.wikipedia.org/wiki/ISO_8859 and look at the charset encodings there. Maybe that gives you a clue what I'm talking about. > I think what happened is that your system specified some alternate font No. A plain-text email /never/ specifies a font. My email did specify a character set (an encoding), as I have told you multiple times. You don't have to guess what happened, you just have to read what I tell you -- and, of course, try to understand it. And maybe read up a bit on Mime, so that you actually know what you're talking about. > for the upper half of the 8 bit character set, and my system supports > only the normal 128-255 characters, There is no "normal" 8-bit character set. As far as plain-text email is concerned, there is only US-ASCII and the various ISO-8859-* encodings. (This is part of the Mime specification.) > This is why I was curious what character code you sent for the omega. Which is what I've responded various times now. How often do you need it being told? ISO-8859-7... is this so difficult? > Once again, whether anyone likes it or not, if you want to be universally > understood, stick to the 7 bit ASCII. Ah... how "universal" is this? This is as "universal" as the "world series" is "world". ASCII may be enough to spell American, but it's not enough for what most people speak. > The vast majority of the time you'll be OK with the common 8 bit > character set. Now here starts the trouble. There is no single standard 8 bit charset; it's as simple as that. Once you start assuming that there is, you're in for encoding/decoding trouble. Gerhard -- http://www.piclist.com PIC/SX FAQ & list archive View/change your membership options at http://mailman.mit.edu/mailman/listinfo/piclist