Just for fun, I worked out a way to measure the size of two received
messages -- one I had sent with html coding, the other without. What I
discovered was interesting. The text body was the same for both,
measured at 3486 characters, including spaces, and including header
information.
The plain text version came in at 5,878 characters. Header alone was
2,208. The remaining 183 characters were things such as high-ascii
quotes and apostrophes translated to text; e.g. "we're" became "we=92re"
and line breaks the mailer inserted for 2-char lines became =20 followed
by the NewLine character.
The version send with html formatting was way worse: 6,660 characters
total -- nearly double the size of the message I'd actually sent.
For one thing, every line -- every line of body text -- was preceded
by "<p class="MsoNormal">" (outside quotes are mine).
I'd have to ask why it had to redefine the text style on each line.
When I learned html, lo those many years into www history, one could
define a font or italics or ... one time and it stayed in effect until
it was turned off.
So there, for the benefit of those who didn't know, or who, like me,
knew the concept but not the extent, is information to which we can add
$1.85 and buy a cup of coffee at the locally owned and operated coffee shop.
Jeffrey Kaplan wrote:
> On Sat, 02 Jun 2007 07:53:54 -0400 J Messeder said:
>
>
>> I was amazed to learn how much space is taken simply by header/routing
>> information. I learned about that when I tried to send a 10Mb picture
>> and was refused by my ISP (Comcast) because I'd exceeded the 13M limit.
>> I wouldn't want to be quoted on the exact numbers, but the difference
>> was huge, and my message had not included any body text or formatting.)
>>
>
> As far as that goes, that 10M was the binary file on your computer,
> right? The encoding for email transmission increases the attachment
> size anywhere between 20 and 40 percent. So that 10M binary file could
> have been over 13M easily, all by itself.
>
> The message headers would only add a few kBytes at most.
>
> Email is a +text only+ medium. To send a binary file through it, the
> file has to be translated into text, which takes up more space. You
> don't see this happening because your email client takes care of the
> details for you.
>
>
>> But yes, the messages I was sending me last night were html --
>> although they actual body contained no formatting other than what
>> initial coding would have identified the font, etc. There were no
>> pictures, italics or bold to deal with in the text. I even stripped off
>> my formatted signature.
>>
>
> Properly constructed HTML messages bodies are a bit over twice the size
> of a plain text message, because it should have a plain text section
> and an HTML section. Which one is shown to the recipient depends on
> the capabilities of their email client. So if a plain text message,
> including headers, would be 2k, then writing the same message in an
> HTML writing client should be about 3.1k. HTML doesn't really bulk up
> the message more than that unless you're doing really fancy or stupid
> things like constantly changing the text layout as you go, including
> embedded elements such as graphics, java/javascript, or CSS and as far
> as I'm concerned, anyone who does THAT (the scripting in email) should
> have their fingers removed at the shoulder. Or maybe the neck, I
> haven't quite decided yet.
>
>
--
/The erasure of a human subspecies is largely painless -- to us -- if we
know little enough about it. A dead Chinaman is of little import to us
whose awareness of things Chinese is bounded by an occasional dish of
chow mein. We grieve only for what we know./
A Sand County Almanac, /Aldo Leopold, July 1949/
