Re: [Imap-protocol] gmail fails "tag uid fetch 123 (BODY.PEEK[2.1.1])", most others OK

Brandon Long <[email protected]> Tue, 5 Mar 2019 10:55:33 -0800
Newsgroups gmane.mail.imap.general
Message-ID <CABa8R6t79-k0uwNL=j1+5FZxh8sgcmh5XUSmTaMkKnkHztMdhg@mail.gmail.com>
--===============8726061299259631914==
Content-Type: multipart/alternative; boundary="000000000000c3fc6a05835d6b63"

--000000000000c3fc6a05835d6b63
Content-Type: text/plain; charset="UTF-8"

Also interesting if it can fetch the full message, since that again might
require a lot of individual fetches to reassemble, which might time out if
we don't handle that correctly.

Brandon

On Tue, Mar 5, 2019 at 9:14 AM Brandon Long <[email protected]> wrote:

> Ugh, that is probably evil to our server.  I don't recall if we succeeded
> in optimizing the specific attachment requests, hopefully, otherwise we'd
> have to reassemble the whole message, fetching all 100 plus parts and then
> getting you what you want... With a little caching it would have to do this
> on every fetch.  That's still the fallback when the individual item fetch
> fails for some reason (message our system didn't handle right or we have
> bad offsets at least)
>
> How big is the message that caching it on your end is bad?
>
> Anyways, I've forwarded to the new team.  Obviously it should work even if
> it's a lot of work on our side.  Is it a consistent one that doesn't work,
> or any of them?
>
> Brandon
>
>
> On Mon, Mar 4, 2019, 6:14 PM Gene Smith <[email protected]> wrote:
>
>> I have been trying to fix Thunderbird so it properly accesses a large
>> mailing list digest message when fetching parts only on demand (no
>> inline display of attachments or local storage to disk of the entire
>> mesage).
>>
>> The message can be found here:
>> https://bugzilla.mozilla.org/attachment.cgi?id=10149
>>
>> I have it working OK for several servers (Dovecot, MDaemon, Cyrus) but
>> am unable to get to work with gmail.
>>
>> The bodystructure numbering of the message parts is like this:
>>
>> 1 Top level message.
>> 2 Container for the 107 rfc822 message attachments
>> 2.1 The 1st message
>> 2.1.1 Body of 1st message
>> 2.2.The 2nd message
>> 2.2.1 Body of 2nd message
>> :
>> :
>> 2.66  The 66th message having attachments (example)
>> 2.66.1 Body of 66th message
>> 2.66.2 Attachment of 66th message (text)
>> 2.66.3 2nd attachment
>> 2.66.4 3rd (last) attachment of 66th message
>> :
>> :
>> 2.107 The 107th (last) message
>> 2.107.1 Body of 107th message
>>
>> Thunderbird accesses the body of an attachment like this (uid 123):
>>
>> tag uid fetch 123 (BODY.PEEK[2.33.1])
>>
>> Several servers respond OK and return the email body of the 33th
>> attachment. But gmail responds like this:
>>
>> 48 NO Some messages could not be FETCHed (Failure)
>>
>> -gene
>>
>>
>>
>>
>>
>>
>> _______________________________________________
>> Imap-protocol mailing list
>> [email protected]
>> http://mailman13.u.washington.edu/mailman/listinfo/imap-protocol
>>
>

--000000000000c3fc6a05835d6b63
Content-Type: text/html; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable

<div dir=3D"ltr">Also interesting if it can fetch the full message, since t=
hat again might require a lot of individual fetches to reassemble, which mi=
ght time out if we don&#39;t handle that correctly.<div><br></div><div>Bran=
don</div></div><br><div class=3D"gmail_quote"><div dir=3D"ltr" class=3D"gma=
il_attr">On Tue, Mar 5, 2019 at 9:14 AM Brandon Long &lt;<a href=3D"mailto:=
[email protected]">[email protected]</a>&gt; wrote:<br></div><blockquote clas=
s=3D"gmail_quote" style=3D"margin:0px 0px 0px 0.8ex;border-left:1px solid r=
gb(204,204,204);padding-left:1ex"><div dir=3D"auto"><div>Ugh, that is proba=
bly evil to our server.=C2=A0 I don&#39;t recall if we succeeded in optimiz=
ing the specific attachment requests, hopefully, otherwise we&#39;d have to=
 reassemble the whole message, fetching all 100 plus parts and then getting=
 you what you want... With a little caching it would have to do this on eve=
ry fetch.=C2=A0 That&#39;s still the fallback when the individual item fetc=
h fails for some reason (message our system didn&#39;t handle right or we h=
ave bad offsets at least)<div dir=3D"auto"><br></div><div dir=3D"auto">How =
big is the message that caching it on your end is bad?<br><div dir=3D"auto"=
><br></div><div dir=3D"auto">Anyways, I&#39;ve forwarded to the new team.=
=C2=A0 Obviously it should work even if it&#39;s a lot of work on our side.=
=C2=A0 Is it a consistent one that doesn&#39;t work, or any of them?</div><=
div dir=3D"auto"><br></div><div dir=3D"auto">Brandon</div></div><br><br><di=
v class=3D"gmail_quote"><div dir=3D"ltr" class=3D"gmail_attr">On Mon, Mar 4=
, 2019, 6:14 PM Gene Smith &lt;<a href=3D"mailto:[email protected]" target=
=3D"_blank">[email protected]</a>&gt; wrote:<br></div><blockquote class=3D"=
gmail_quote" style=3D"margin:0px 0px 0px 0.8ex;border-left:1px solid rgb(20=
4,204,204);padding-left:1ex">I have been trying to fix Thunderbird so it pr=
operly accesses a large <br>
mailing list digest message when fetching parts only on demand (no <br>
inline display of attachments or local storage to disk of the entire <br>
mesage).<br>
<br>
The message can be found here: <br>
<a href=3D"https://bugzilla.mozilla.org/attachment.cgi?id=3D10149" rel=3D"n=
oreferrer noreferrer" target=3D"_blank">https://bugzilla.mozilla.org/attach=
ment.cgi?id=3D10149</a><br>
<br>
I have it working OK for several servers (Dovecot, MDaemon, Cyrus) but <br>
am unable to get to work with gmail.<br>
<br>
The bodystructure numbering of the message parts is like this:<br>
<br>
1 Top level message.<br>
2 Container for the 107 rfc822 message attachments<br>
2.1 The 1st message<br>
2.1.1 Body of 1st message<br>
2.2.The 2nd message<br>
2.2.1 Body of 2nd message<br>
:<br>
:<br>
2.66=C2=A0 The 66th message having attachments (example)<br>
2.66.1 Body of 66th message<br>
2.66.2 Attachment of 66th message (text)<br>
2.66.3 2nd attachment<br>
2.66.4 3rd (last) attachment of 66th message<br>
:<br>
:<br>
2.107 The 107th (last) message<br>
2.107.1 Body of 107th message<br>
<br>
Thunderbird accesses the body of an attachment like this (uid 123):<br>
<br>
tag uid fetch 123 (BODY.PEEK[2.33.1])<br>
<br>
Several servers respond OK and return the email body of the 33th <br>
attachment. But gmail responds like this:<br>
<br>
48 NO Some messages could not be FETCHed (Failure)<br>
<br>
-gene<br>
<br>
<br>
<br>
<br>
<br>
<br>
_______________________________________________<br>
Imap-protocol mailing list<br>
<a href=3D"mailto:[email protected]" rel=3D"noreferrer" target=
=3D"_blank">[email protected]</a><br>
<a href=3D"http://mailman13.u.washington.edu/mailman/listinfo/imap-protocol=
" rel=3D"noreferrer noreferrer" target=3D"_blank">http://mailman13.u.washin=
gton.edu/mailman/listinfo/imap-protocol</a><br>
</blockquote></div></div></div>
</blockquote></div>

--000000000000c3fc6a05835d6b63--

--===============8726061299259631914==
Content-Type: text/plain; charset="us-ascii"
MIME-Version: 1.0
Content-Transfer-Encoding: 7bit
Content-Disposition: inline

_______________________________________________
Imap-protocol mailing list
[email protected]
http://mailman13.u.washington.edu/mailman/listinfo/imap-protocol
--===============8726061299259631914==--