| 1 |
<pre> Network Working Group N. Borenstein, Bellcore |
| 2 |
Request for Comments: 1344 June 1992 |
| 3 |
|
| 4 |
<span class="h1">Implications of MIME for Internet Mail Gateways</span> |
| 5 |
|
| 6 |
|
| 7 |
Status of This Memo |
| 8 |
|
| 9 |
This is an informational memo for the Internet community, |
| 10 |
and requests discussion and suggestions for improvements. |
| 11 |
This memo does not specify an Internet standard. |
| 12 |
Distribution of this memo is unlimited. |
| 13 |
|
| 14 |
Abstract |
| 15 |
|
| 16 |
The recent development of MIME (Multipurpose Internet Mail |
| 17 |
Extensions) offers a wide range of new opportunities for |
| 18 |
electronic mail system systems. Most of these opportunites |
| 19 |
are relevant only to user agents, the programs that interact |
| 20 |
with human users when they send and receive mail. However, |
| 21 |
some opportunities are also opened up for mail transport |
| 22 |
systems. While MIME was carefully designed so that it does |
| 23 |
not require any changes to Internet electronic message |
| 24 |
transport facilities, there are several ways in which |
| 25 |
message transport systems may want to take advantage of |
| 26 |
MIME. These opportunities are the subject of this memo. |
| 27 |
|
| 28 |
Background -- The MIME Format |
| 29 |
|
| 30 |
Recently, a new standardized format has been defined for |
| 31 |
enhanced electronic mail messages on the Internet. This |
| 32 |
format, known as MIME, permits messages to include, in a |
| 33 |
standardized manner, non-ASCII text, images, audio, and a |
| 34 |
variety of other kinds of interesting data. |
| 35 |
|
| 36 |
The MIME effort was explicitly focused on requiring |
| 37 |
absolutely no changes at the message transport level. |
| 38 |
Because of this fact, MIME-format mail runs transparently on |
| 39 |
all known Internet or Internet-style mail systems. This |
| 40 |
means that those concerned solely with the maintenance and |
| 41 |
development of message transport services can safely ignore |
| 42 |
MIME completely, if they so choose. |
| 43 |
|
| 44 |
However, the fact that MIME can be ignored, for the purpose |
| 45 |
of message transport, does not necessarily mean that it |
| 46 |
should be ignored. In particular, MIME offers several |
| 47 |
features that should be of interest to those responsible for |
| 48 |
message transport services. By exploiting these features, |
| 49 |
transport systems can provide certain additional kinds of |
| 50 |
service that are currently unavailable, and can alleviate a |
| 51 |
few existing problems. |
| 52 |
|
| 53 |
The remainder of this document is an attempt to briefly |
| 54 |
point out and summarize some important ways in which MIME |
| 55 |
|
| 56 |
|
| 57 |
|
| 58 |
<span class="grey"> Borenstein [Page 1]</span></pre> |
| 59 |
<hr class='noprint'/><!--NewPage--><pre class='newpage'><span id="page-2" ></span> |
| 60 |
<span class="grey"> <a href="./rfc1344">RFC 1344</a> MIME and Mail Gateways June 1992</span> |
| 61 |
|
| 62 |
|
| 63 |
may be of use for message transport systems. This document |
| 64 |
makes no attempt to present a complete technical description |
| 65 |
of MIME, however. For that, the reader is refered to the |
| 66 |
MIME document itself [<a href="./rfc1341" title=""MIME (Multipurpose Internet Mail Extensions): Mechanisms for Specifying and Describing the Format of Internet Message Bodies"">RFC-1341</a>]. |
| 67 |
|
| 68 |
Mail Transport and Gateway Services: A Key Distinction |
| 69 |
|
| 70 |
Before implementing any of the mechanisms discussed in this |
| 71 |
memo, one should be familiar with the distinction between |
| 72 |
mail transport service and mail gateway service. Basically, |
| 73 |
mail transport software is responsible for moving a message |
| 74 |
within a homogeneous electronic mail service network. Mail |
| 75 |
gateways, on the other hand, exchange mail between two |
| 76 |
significantly different mail environments, including via |
| 77 |
non-electronic services, such as postal mail. |
| 78 |
|
| 79 |
In general, it is widely considered unacceptable for mail |
| 80 |
transport services to alter the contents of messages. In |
| 81 |
the case of mail gateways, however, such alteration is often |
| 82 |
inevitable. Thus, strictly speaking, many of the mechanisms |
| 83 |
described here apply only to gateways, and should not be |
| 84 |
used in simple mail transport systems. However, it is |
| 85 |
possible that some very special situations -- e.g., an SMTP |
| 86 |
relay that transports mail across extremely expensive |
| 87 |
intercontinental network links -- might need to modify |
| 88 |
messages, in order to provide appropriate service for those |
| 89 |
situations, and hence must redefine its role to be that of a |
| 90 |
gateway. |
| 91 |
|
| 92 |
In this memo, it is assumed that transformations which alter |
| 93 |
a message's contents will be performed only by gateways, but |
| 94 |
it is recognized that some existing mail transport agents |
| 95 |
may choose to reclassify themselves as gateways in order to |
| 96 |
perform the functions described here. |
| 97 |
|
| 98 |
Rejected Messages |
| 99 |
|
| 100 |
An unfortunately frequent duty of message transport services |
| 101 |
is the rejection of mail to the sender. This may happen |
| 102 |
because the mail was undeliverable, or because it did not |
| 103 |
conform to the requirements of a gateway (e.g., it was too |
| 104 |
large). |
| 105 |
|
| 106 |
There has never been a standard format for rejected messages |
| 107 |
in the past. This has been an annoyance, but not a major |
| 108 |
problem for text messages. For non-text messages, however, |
| 109 |
the lack of a standard rejection format is more crucial, |
| 110 |
because rejected messages typically appear to be text, and |
| 111 |
the user who finds himself viewing images or audio as if |
| 112 |
they were text is rarely happy with the result. |
| 113 |
|
| 114 |
MIME makes it very easy to encapsulate messages in such a |
| 115 |
way that their semantics are completely preserved. The |
| 116 |
simplest way to do this is to make each rejection notice a |
| 117 |
|
| 118 |
|
| 119 |
|
| 120 |
<span class="grey"> Borenstein [Page 2]</span></pre> |
| 121 |
<hr class='noprint'/><!--NewPage--><pre class='newpage'><span id="page-3" ></span> |
| 122 |
<span class="grey"> <a href="./rfc1344">RFC 1344</a> MIME and Mail Gateways June 1992</span> |
| 123 |
|
| 124 |
|
| 125 |
MIME "multipart/mixed" message. That multipart message |
| 126 |
would contain two parts, a text part explaining the reason |
| 127 |
for the rejection, and an encapsulated message part that |
| 128 |
contained the rejected message itself. |
| 129 |
|
| 130 |
It should be stressed that the transport software does not |
| 131 |
need to understand the structure of the rejected message at |
| 132 |
all. It merely needs to encapsulate it properly. The |
| 133 |
following, for example, shows how any MIME message may be |
| 134 |
encapsulated in a rejection message in such a way that all |
| 135 |
information will be immediately visible in the correct form |
| 136 |
if the recipient reads it with a MIME-conformant mail |
| 137 |
reader: |
| 138 |
|
| 139 |
From: Mailer-Daemon <[email protected]> |
| 140 |
Subject: Rejected Message |
| 141 |
Content-type: multipart/mixed; boundary=unique-boundary |
| 142 |
|
| 143 |
--unique-boundary |
| 144 |
Content-type: text/plain; charset=us-ascii |
| 145 |
|
| 146 |
A mail message you sent was rejected. The details of |
| 147 |
the rejected message are as follows: |
| 148 |
|
| 149 |
From: Nathainel Borenstein <[email protected]> |
| 150 |
Message-ID: <[email protected]> |
| 151 |
To: [email protected] |
| 152 |
Subject: I know my rights! |
| 153 |
Rejection-reason: No mail from libertarians is |
| 154 |
accepted. |
| 155 |
|
| 156 |
The original message follows below. |
| 157 |
--unique-boundary |
| 158 |
Content-type: message/rfc822 |
| 159 |
|
| 160 |
The ENTIRE REJECTED MESSAGE, starting with the headers, |
| 161 |
goes here. |
| 162 |
|
| 163 |
--unique-boundary-- |
| 164 |
In the above example, the ONLY thing that is not |
| 165 |
'boilerplate" is the choice of boundary string. The phrase |
| 166 |
"unique-boundary" should be replaced by a string that does |
| 167 |
not appear (prefixed by two hyphens) in any of the body |
| 168 |
parts. |
| 169 |
|
| 170 |
Encapsulating a message in this manner is very easily done, |
| 171 |
and will constitute a significant service that message |
| 172 |
transport services can perform for MIME users. |
| 173 |
|
| 174 |
IMPORTANT NOTE: The format given above is simply one of |
| 175 |
many possible ways to format a rejection message using MIME. |
| 176 |
Independent IETF efforts are needed in order to standardize |
| 177 |
the format of rejections and acknowledgements. |
| 178 |
|
| 179 |
|
| 180 |
|
| 181 |
|
| 182 |
<span class="grey"> Borenstein [Page 3]</span></pre> |
| 183 |
<hr class='noprint'/><!--NewPage--><pre class='newpage'><span id="page-4" ></span> |
| 184 |
<span class="grey"> <a href="./rfc1344">RFC 1344</a> MIME and Mail Gateways June 1992</span> |
| 185 |
|
| 186 |
|
| 187 |
Fragmenting and Reassembling Large Messages |
| 188 |
|
| 189 |
One problem that occurs with increasing frequency in |
| 190 |
Internet mail is the rejection of messages because of size |
| 191 |
limitations. This problem can be expected to grow |
| 192 |
substantially more severe with the acceptance of MIME, as |
| 193 |
MIME invites the use of very large objects such as images |
| 194 |
and audio clips. Fortunately, MIME also provides mechanisms |
| 195 |
that can help alleviate the problem. |
| 196 |
|
| 197 |
One particularly relevant MIME type is "message/partial", |
| 198 |
which can be used for the automatic fragmentation and |
| 199 |
reassembly of large mail messages. The message/partial type |
| 200 |
can be handled entirely at the user agent level, but message |
| 201 |
transport services can also make use of this type to provide |
| 202 |
more intelligent behavior at gateways. |
| 203 |
|
| 204 |
In particular, when gatewaying mail to or from a system or |
| 205 |
network known to enforce size limitations that are more or |
| 206 |
less stringent than are enforced locally, message transport |
| 207 |
services might choose either to break a large message into |
| 208 |
fragments, or (perhaps less likely) to reassemble fragments |
| 209 |
into a larger message. The combination of these two |
| 210 |
behaviors can make the overall Internet mail environment |
| 211 |
appear more complete and seamless than it actually is. |
| 212 |
|
| 213 |
Details on the message/partial format may be found in the |
| 214 |
MIME document. What follows is an example of how a simple |
| 215 |
short message might be broken into two message/partial |
| 216 |
messages. In practice, of course, the message/partial |
| 217 |
facility would only be likely to be used for much longer |
| 218 |
messages. |
| 219 |
|
| 220 |
The following initial message: |
| 221 |
|
| 222 |
From: Nathaniel Borenstein <[email protected]> |
| 223 |
To: Ned Freed: <[email protected]> |
| 224 |
Subject: a test message |
| 225 |
Content-type: image/gif |
| 226 |
Content-Transfer-Encoding: base64 |
| 227 |
|
| 228 |
R0lGODdhQAGMAbMAAAAAAP/u7swzIu6ZiLsiEd1EM+5VRGaI3WYAAO67qkRV |
| 229 |
uwARd6q7/ywAAAAAQAGMAUME/hDISau9OOvNu/9gKI6kRJwoUa5s675wLM90l |
| 230 |
XW5YKxqPyKRygxv2dr4czwlMCZrQLFTYHBJ2hlyQYFiaz+i0WWBou7fOq1x8vXWfU |
| 231 |
qU1fJ2qEhYaHGjhZQmJ2QT1xBW1ak1xUdV0/VjtsbpUEDaEJCQOIpqeoNV+LXo5W |
| 232 |
fVN3dZKceAQPvgyhwQ2lqcXGxx5wja59eJIGUNCszF90sYp50CoqFZ4DoqMMo6M |
| 233 |
|
| 234 |
can be transformed, invertibly, into the following two |
| 235 |
message/partial messages: |
| 236 |
|
| 237 |
|
| 238 |
From: Nathaniel Borenstein <[email protected]> |
| 239 |
|
| 240 |
|
| 241 |
|
| 242 |
|
| 243 |
|
| 244 |
<span class="grey"> Borenstein [Page 4]</span></pre> |
| 245 |
<hr class='noprint'/><!--NewPage--><pre class='newpage'><span id="page-5" ></span> |
| 246 |
<span class="grey"> <a href="./rfc1344">RFC 1344</a> MIME and Mail Gateways June 1992</span> |
| 247 |
|
| 248 |
|
| 249 |
To: Ned Freed <[email protected]> |
| 250 |
Subject: a test message |
| 251 |
Content-type: message/partial; id="[email protected]"; |
| 252 |
number=1; total=2 |
| 253 |
|
| 254 |
Content-type: image/gif |
| 255 |
Content-Transfer-Encoding: base64 |
| 256 |
|
| 257 |
R0lGODdhQAGMAbMAAAAAAP/u7swzIu6ZiLsiEd1EM+5VRGaI3WYAAO67qkRV |
| 258 |
|
| 259 |
and |
| 260 |
|
| 261 |
From: Nathaniel Borenstein <[email protected]> |
| 262 |
To: Ned Freed <[email protected]> |
| 263 |
Subject: a test message |
| 264 |
Content-type: message/partial; id="[email protected]"; |
| 265 |
number=2; total=2 |
| 266 |
|
| 267 |
uwARd6q7/ywAAAAAQAGMAUME/hDISau9OOvNu/9gKI6kRJwoUa5s675wLM90l |
| 268 |
XW5YKxqPyKRygxv2dr4czwlMCZrQLFTYHBJ2hlyQYFiaz+i0WWBou7fOq1x8vXWfU |
| 269 |
qU1fJ2qEhYaHGjhZQmJ2QT1xBW1ak1xUdV0/VjtsbpUEDaEJCQOIpqeoNV+LXo5W |
| 270 |
fVN3dZKceAQPvgyhwQ2lqcXGxx5wja59eJIGUNCszF90sYp50CoqFZ4DoqMMo6M |
| 271 |
|
| 272 |
Fragmenting such messages rather than rejecting them might |
| 273 |
be a reasonable option for some gateway services, at least |
| 274 |
for a certain range of message sizes. Of course, it is |
| 275 |
often difficult for a gateway to know what size limitations |
| 276 |
will be encountered "downstream", but intelligent guesses |
| 277 |
are often possible. Moreover, an IETF working group on SMTP |
| 278 |
extensions has proposed augmenting SMTP with a "SIZE" verb |
| 279 |
that would facilitate this process, thereby possibly |
| 280 |
requiring fragmentation on the fly during message |
| 281 |
transmission. |
| 282 |
|
| 283 |
Note also that fragmentation or reassembly might reasonably |
| 284 |
be performed, in differing circumstances, by either the |
| 285 |
sending or receiving gateway systems, depending on which |
| 286 |
system knew more about the capabilities of the other. |
| 287 |
|
| 288 |
Using or Removing External-Body Pointers |
| 289 |
|
| 290 |
Another MIME type oriented to extremely large messages is |
| 291 |
the "message/external-body" type. In this type of message, |
| 292 |
all or part of the body data is not included in the actual |
| 293 |
message itself. Instead, the Content-Type header field |
| 294 |
includes information that tells how the body data can be |
| 295 |
retrieved -- either via a file system, via anonymous ftp, or |
| 296 |
via other mechanisms. |
| 297 |
|
| 298 |
The message/external-body type provides a new option for |
| 299 |
mail transport services that wishes to optimize the way |
| 300 |
bandwidth resources are used in a given environment. For |
| 301 |
example, the basic use of message/external-body is to reduce |
| 302 |
bandwidth in email traffic. However, when email crosses a |
| 303 |
|
| 304 |
|
| 305 |
|
| 306 |
<span class="grey"> Borenstein [Page 5]</span></pre> |
| 307 |
<hr class='noprint'/><!--NewPage--><pre class='newpage'><span id="page-6" ></span> |
| 308 |
<span class="grey"> <a href="./rfc1344">RFC 1344</a> MIME and Mail Gateways June 1992</span> |
| 309 |
|
| 310 |
|
| 311 |
slow and expensive boundary -- e.g., a satellite link across |
| 312 |
the Pacific -- it might make sense to retrieve the data |
| 313 |
itself and transform the external-body reference into the |
| 314 |
actual data. Alternately, it might make sense to copy the |
| 315 |
data itself to a new location, closer to the message |
| 316 |
recipients, and change the location pointed to in the |
| 317 |
message. Because the external-body specification can |
| 318 |
include an expiration date, message transport services can |
| 319 |
trade off storage and bandwidth capabilities to try to |
| 320 |
optimize the overall use of resources for very large |
| 321 |
messages. |
| 322 |
|
| 323 |
Such behaviors by a gateway require careful analysis of |
| 324 |
cost/benefit tradeoffs and would be a dramatic departure |
| 325 |
from typical mail transport services. However, the |
| 326 |
potential benefits are quite significant, so that such the |
| 327 |
appropriate use of these service options should be explored. |
| 328 |
|
| 329 |
For example, the following message includes PostScript data |
| 330 |
by external reference: |
| 331 |
|
| 332 |
From: Nathaniel Borenstein <[email protected]> |
| 333 |
To: Ned Freed <[email protected]> |
| 334 |
Subject: The latest MIME draft |
| 335 |
Content-Type: message/external-body; |
| 336 |
name="BodyFormats.ps"; |
| 337 |
site="thumper.bellcore.com"; |
| 338 |
access-type=ANON-FTP; |
| 339 |
directory="pub"; |
| 340 |
mode="image"; |
| 341 |
expiration="Fri, 14 Jun 1991 19:13:14 -0400 (EDT)" |
| 342 |
|
| 343 |
Content-type: application/postscript |
| 344 |
|
| 345 |
A gateway to Australia might choose to copy the file to an |
| 346 |
Australian FTP archive, changing the relevant parameters on |
| 347 |
the Content-type header field. Alternately, it might choose |
| 348 |
simply to transform the message into one in which all the |
| 349 |
data were included: |
| 350 |
|
| 351 |
From: Nathaniel Borenstein <[email protected]> |
| 352 |
To: Ned Freed <[email protected]> |
| 353 |
Subject: The latest MIME draft |
| 354 |
Content-type: application/postscript |
| 355 |
|
| 356 |
%!PS-Adobe-1.0 |
| 357 |
%%Creator: greenbush:nsb (Nathaniel Borenstein,MRE 2A- |
| 358 |
274,4270,9938586,21462) |
| 359 |
etc... |
| 360 |
|
| 361 |
This is an example which suggests both the benefits and the |
| 362 |
dangers. There is considerable benefit to having a copy of |
| 363 |
the data immediately available, but there also may be |
| 364 |
considerable expense involved in transporting it to all of |
| 365 |
|
| 366 |
|
| 367 |
|
| 368 |
<span class="grey"> Borenstein [Page 6]</span></pre> |
| 369 |
<hr class='noprint'/><!--NewPage--><pre class='newpage'><span id="page-7" ></span> |
| 370 |
<span class="grey"> <a href="./rfc1344">RFC 1344</a> MIME and Mail Gateways June 1992</span> |
| 371 |
|
| 372 |
|
| 373 |
a the members of a list, if only a few will use the data |
| 374 |
anytime soon. |
| 375 |
|
| 376 |
Alternatively, instead of replacing an external-body message |
| 377 |
with its real contents, it might make sense to transform it |
| 378 |
into a "multipart/alternative" message containing both the |
| 379 |
external body reference and the expanded version. This |
| 380 |
means that only the external body part can be forwarded if |
| 381 |
desired, and the recipient doesn't lose the information as |
| 382 |
to where the data was fetched from, if they want to fetch an |
| 383 |
up-to-date version in the future. Such information could be |
| 384 |
represented, in MIME, in the following form: |
| 385 |
|
| 386 |
From: Nathaniel Borenstein <[email protected]> |
| 387 |
To: Ned Freed <[email protected]> |
| 388 |
Subject: The latest MIME draft |
| 389 |
Content-type: multipart/alternative; boundary=foo |
| 390 |
|
| 391 |
--foo |
| 392 |
Content-Type: message/external-body; |
| 393 |
name="BodyFormats.ps"; |
| 394 |
site="thumper.bellcore.com"; |
| 395 |
access-type=ANON-FTP; |
| 396 |
directory="pub"; |
| 397 |
mode="image"; |
| 398 |
expiration="Fri, 14 Jun 1991 19:13:14 -0400 (EDT)" |
| 399 |
|
| 400 |
Content-type: application/postscript |
| 401 |
--foo |
| 402 |
Content-type: application/postscript |
| 403 |
|
| 404 |
%!PS-Adobe-1.0 |
| 405 |
%%Creator: greenbush:nsb (Nathaniel Borenstein,MRE 2A- |
| 406 |
274,4270,9938586,21462) |
| 407 |
etc... |
| 408 |
--foo-- |
| 409 |
|
| 410 |
Similarly for the case where a message is copied to a local |
| 411 |
FTP site, one could offer two external body parts as the |
| 412 |
alternatives, allowing the user agent to choose which FTP |
| 413 |
site is preferred. |
| 414 |
|
| 415 |
Image and other Format Conversions |
| 416 |
|
| 417 |
MIME currently defines two image formats, image/gif and |
| 418 |
image/jpeg. The former is much more convenient for many |
| 419 |
users, and can be displayed more quickly on many systems. |
| 420 |
The latter is a much more compact representation, and |
| 421 |
therfore places less stress on mail transport facilities. |
| 422 |
|
| 423 |
Message transport services can optimize both transport |
| 424 |
bandwidth and user convenience by intelligent translation |
| 425 |
between these formats (and other formats that might be added |
| 426 |
later). When a message of type image/gif is submitted for |
| 427 |
|
| 428 |
|
| 429 |
|
| 430 |
<span class="grey"> Borenstein [Page 7]</span></pre> |
| 431 |
<hr class='noprint'/><!--NewPage--><pre class='newpage'><span id="page-8" ></span> |
| 432 |
<span class="grey"> <a href="./rfc1344">RFC 1344</a> MIME and Mail Gateways June 1992</span> |
| 433 |
|
| 434 |
|
| 435 |
long-haul delivery, it might reasonably be translated to |
| 436 |
image/jpeg. Conversely, when image/jpeg data is received |
| 437 |
for final delivery on a system with adequate storage |
| 438 |
resources, it might be translated to image/gif for the |
| 439 |
convenience of the recipient. Software to perform these |
| 440 |
translations is widely available. It should be noted, |
| 441 |
however, that performance of such conversions presumes |
| 442 |
support for the new format by the recipient. |
| 443 |
|
| 444 |
Although MIME currently only defines one audio format, more |
| 445 |
are likely to be defined and registered with IANA in the |
| 446 |
future. In that case, similar format conversion facilities |
| 447 |
might be appropriate for audio. |
| 448 |
|
| 449 |
If format conversion is done, it is STRONGLY RECOMMENDED |
| 450 |
that some kind of trace information (probably in the form of |
| 451 |
a Received header field) should be added to a message to |
| 452 |
document the conversion that has been performed. |
| 453 |
|
| 454 |
Some people have expressed concerns, or even the opinion |
| 455 |
that conversions should never be done. To accomodate the |
| 456 |
desires of those who dislike the idea of automatic format |
| 457 |
conversion. For this reason, it is suggested that such |
| 458 |
transformations be generally restricted to gateways rather |
| 459 |
than general message transport services, and that services |
| 460 |
which perform such conversions should be sensitive to a |
| 461 |
header field that indicates that the sender does not wish to |
| 462 |
have any such conversions performed. A suggested value for |
| 463 |
this header field is: |
| 464 |
|
| 465 |
Content-Conversion: prohibited |
| 466 |
|
| 467 |
User agents that wish to explicitly indicate a willingness |
| 468 |
for such conversions to be performed may use: |
| 469 |
|
| 470 |
Content-Conversion: permitted |
| 471 |
|
| 472 |
However, this will be the default assumption of many |
| 473 |
gateways, so this header field is not strictly necessary. |
| 474 |
It also should be noted that such control of conversion |
| 475 |
would only be available to the sender, rather than to any of |
| 476 |
the recipients. |
| 477 |
|
| 478 |
|
| 479 |
|
| 480 |
|
| 481 |
|
| 482 |
|
| 483 |
|
| 484 |
|
| 485 |
|
| 486 |
|
| 487 |
|
| 488 |
|
| 489 |
|
| 490 |
|
| 491 |
|
| 492 |
<span class="grey"> Borenstein [Page 8]</span></pre> |
| 493 |
<hr class='noprint'/><!--NewPage--><pre class='newpage'><span id="page-9" ></span> |
| 494 |
<span class="grey"> <a href="./rfc1344">RFC 1344</a> MIME and Mail Gateways June 1992</span> |
| 495 |
|
| 496 |
|
| 497 |
Robust Encoding of Data |
| 498 |
|
| 499 |
In addition to all the reasons given above for possible |
| 500 |
transformation of body data, it will sometimes be the case |
| 501 |
that a gateway can tell that the body data, as given, will |
| 502 |
not robustly survive the next step of transport. For |
| 503 |
example, mail crossing an ASCII-to-EBCDIC gateway will lose |
| 504 |
information if certain characters are used. In such cases, |
| 505 |
a gateway can make the data more robust simply by applying |
| 506 |
one of the MIME Content-Transfer-Encoding algorithms (base64 |
| 507 |
or quoted-printable) to the body or body part. This will |
| 508 |
generally be a loss-less transformation, but care must be |
| 509 |
taken to ensure that the resulting message is MIME- |
| 510 |
conformant if the inital message was not. (For example, a |
| 511 |
MIME-Version header field may need to be added.) |
| 512 |
|
| 513 |
User-oriented concerns |
| 514 |
|
| 515 |
If a gateway is going to perform major transformations on a |
| 516 |
mail message, such as translating image formats or mapping |
| 517 |
between included data and external-reference data, it seems |
| 518 |
inevitable that there will be situations in which users will |
| 519 |
object to these transformations. This is, in large part, an |
| 520 |
implementation issue, but it seems advisable, wherever |
| 521 |
possible, to provide a mechanism whereby users can specify, |
| 522 |
to the transport system, whether or not they want such |
| 523 |
services performed automatically on their behalf. The use of |
| 524 |
the "Content-Conversion" header field, as mentioned above, |
| 525 |
is suggested for this purpose, since it it least provides |
| 526 |
some control by the sender, if not the recipient. |
| 527 |
|
| 528 |
References |
| 529 |
|
| 530 |
[<a id="ref-RFC-1341">RFC-1341</a>] Borenstein, N., and N. Freed, "MIME |
| 531 |
(Multipurpose Internet Mail Extensions): Mechanisms for |
| 532 |
Specifying and Describing the Format of Internet Message |
| 533 |
Bodies", <a href="./rfc1341">RFC 1341</a>, Bellcore, June, 1992. |
| 534 |
|
| 535 |
Security Considerations |
| 536 |
|
| 537 |
Security issues are not discussed in this memo. |
| 538 |
|
| 539 |
Author's Address |
| 540 |
|
| 541 |
Nathaniel S. Borenstein |
| 542 |
MRE 2D-296, Bellcore |
| 543 |
445 South St. |
| 544 |
Morristown, NJ 07962-1910 |
| 545 |
|
| 546 |
Email: [email protected] |
| 547 |
Phone: +1 201 829 4270 |
| 548 |
Fax: +1 201 829 7019 |
| 549 |
|
| 550 |
|
| 551 |
|
| 552 |
|
| 553 |
|
| 554 |
Borenstein [Page 9] |
| 555 |
</pre> |