fetchmail-users — User questions and discussion

Re: [Fetchmail-users] Local delivery - smtp or sendmail

[Hans C. <fo...@gm...>]

On Wed, 19 Nov 2025, Matthias Andree via Fetchmail-users wrote:

> Am 11.11.25 um 21:24 schrieb Hans Carlson via Fetchmail-users:
>>
>>  Sooo... if I don't actually want any restrictions on fetchmail, then is
>>  there any reason NOT to use sendmail for delivery instead of SMTP?
>
> The concerns are
>
> - you're breaking up the tight coupling between the SMTP server (Postfix) and 
> the client (fetchmail), so...
>
> - that you won't notice from fetchmail's end if the Postfix service goes down 
> (which it doesn't ever do for me, Postfix is one of the nicer things in 
> software) because the sendmail command should always be able to enqueue the 
> message, even after "postfix stop", but they won't get delivered

Thanks for the clarification.  In that case, I'll setup postfix to listen 
on port 25 for fetchmail and port 587 for alpine.  That way I should be 
able to configure different restrictions based on the port.

Re: [Fetchmail-users] Local delivery - smtp or sendmail

[Matthias A. <mat...@gm...>]

Am 11.11.25 um 21:24 schrieb Hans Carlson via Fetchmail-users:
> On Tue, 11 Nov 2025, Matthias Andree via Fetchmail-users wrote:
>
>> Am 09.11.25 um 01:50 schrieb Hans Carlson via Fetchmail-users:
>>>
>>>  I'm trying to configure smtpd_sender_restrictions in postfix, 
>>> mainly so
>>>  I'll get an immediate failure if I've added a new email address that
>>>  hasn't been configured in postfix.
>>
>> Those are only available through SMTP, not through most 
>> /usr/{lib,sbin}/sendmail wrappers (certainly not Postfix's).
>
> Right, the restrictions I was planning to configure would only apply 
> to the smtp connections from alpine.  I don't think I want any 
> restrictions on the connections from fetchmail.  fetchmail should 
> process all the mail it gets and deliver it to the local user.
>
> For the smtp connections from alpine on the other hand I want to add a 
> simply table with a list of the email addresses that are allowed to 
> send email.  If in the future I add a new email address, then I want 
> the alpine SMTP connection to my local postfix SMTP server to give me 
> an immediate rejection so I know I need to go configure authentication 
> for the new email address. Without that, the postfix SMTP client 
> connection to the isp relay will eventually fail with an auth error, 
> but I won't notice it for some time because that's all done in the 
> background.
>
> Sooo... if I don't actually want any restrictions on fetchmail, then 
> is there any reason NOT to use sendmail for delivery instead of SMTP?


The concerns are

- you're breaking up the tight coupling between the SMTP server 
(Postfix) and the client (fetchmail), so...

- that you won't notice from fetchmail's end if the Postfix service goes 
down (which it doesn't ever do for me, Postfix is one of the nicer 
things in software) because the sendmail command should always be able 
to enqueue the message, even after "postfix stop", but they won't get 
delivered

[snip]


HTH

Matthias

Re: [Fetchmail-users] poll server failover with multiple DNS entries

[Matthias A. <mat...@gm...>]

Am 19.11.25 um 14:32 schrieb BASSAGET Cédric:
> Hello,
>
> I'm trying to poll a pop3 server resolving on different IP addresses :
> # dig pop.dom.tld +short
> 1.2.3.4
> 1.2.3.5
>
> when running fetchmail, it tries to reach one of the two servers, and at
> the end of the (-t) timeout, fetchmail exits without trying to poll the
> other server.

Salut Cédric,

Uh. That would be a logic bug. I won't have time immediately, so I'd 
better file an issue in Gitlab so I don't forget.
-> https://gitlab.com/fetchmail/fetchmail/-/issues/85

What's the fetchmail version you're using this with?

> So if one of my pop3 servers is unreachable and this is the one choosed by
> fetchmail, it does not work.
>
> Is there a way to tell fetchmail to poll others IP resolved by dns name
> "pop.dom.tld" if the first one fails ?

You can use the "via" keyword in your rcfile to override fetchmail's 
idea of where to connect, so you could write "poll hermes.example.org 
via 10.2.3.5..." or similar.

Hope that helps,
Matthias

[Fetchmail-users] poll server failover with multiple DNS entries

[BASSAGET C. <ced...@gm...>]

Hello,

I'm trying to poll a pop3 server resolving on different IP addresses :
# dig pop.dom.tld +short
1.2.3.4
1.2.3.5

when running fetchmail, it tries to reach one of the two servers, and at
the end of the (-t) timeout, fetchmail exits without trying to poll the
other server.

So if one of my pop3 servers is unreachable and this is the one choosed by
fetchmail, it does not work.

Is there a way to tell fetchmail to poll others IP resolved by dns name
"pop.dom.tld" if the first one fails ?

Regards
Cédric

[Fetchmail-users] The 6.6.1 release of fetchmail is available (translation updates for Spanish, Serbian)

[Matthias A. <mat...@gm...>]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA512

The 6.6.1 release of fetchmail is now available at the usual locations,
including <https://downloads.sourceforge.net/project/fetchmail/branch_6.6/>.

The source archive is available at:
<https://downloads.sourceforge.net/project/fetchmail/branch_6.6/fetchmail-6.6.1.tar.xz/download>
<https://gitlab.com/-/project/188557/uploads/5f4b7af4e98d69e1648b42e8a965811e/fetchmail-6.6.1.tar.xz>

The detached GnuPG signature is available at:
<https://downloads.sourceforge.net/project/fetchmail/branch_6.6/fetchmail-6.6.1.tar.xz.asc/download>
<https://gitlab.com/-/project/188557/uploads/9508a4087b635359fded6157dfd96e91/fetchmail-6.6.1.tar.xz.asc>

The SHA256 hashes for the tarballs are:
SHA2-256(fetchmail-6.6.1.tar.xz)= 38d01fe404e67514df394a6ed1a815bbb61aa90c0fa4402252593aced0e38a1d


Here are the release notes:
- --------------------------------------------------------------------------------
fetchmail-6.6.1 (released 2025-11-12, 32381 LoC):

## TRANSLATIONS were updated by these fine people (randomized order):
* sr:    Мирослав Николић [Serbian]
* es:    Cristian Othón Martínez Vera [Spanish]

- -------------------------------------------------------------------------------
-----BEGIN PGP SIGNATURE-----
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=Z4Xo
-----END PGP SIGNATURE-----

Re: [Fetchmail-users] Local delivery - smtp or sendmail

[Hans C. <fo...@gm...>]

On Tue, 11 Nov 2025, Matthias Andree via Fetchmail-users wrote:

> Am 09.11.25 um 01:50 schrieb Hans Carlson via Fetchmail-users:
>>
>>  I'm trying to configure smtpd_sender_restrictions in postfix, mainly so
>>  I'll get an immediate failure if I've added a new email address that
>>  hasn't been configured in postfix.
>
> Those are only available through SMTP, not through most 
> /usr/{lib,sbin}/sendmail wrappers (certainly not Postfix's).

Right, the restrictions I was planning to configure would only apply to 
the smtp connections from alpine.  I don't think I want any restrictions 
on the connections from fetchmail.  fetchmail should process all the mail 
it gets and deliver it to the local user.

For the smtp connections from alpine on the other hand I want to add a 
simply table with a list of the email addresses that are allowed to send 
email.  If in the future I add a new email address, then I want the alpine 
SMTP connection to my local postfix SMTP server to give me an immediate 
rejection so I know I need to go configure authentication for the new 
email address.  Without that, the postfix SMTP client connection to the 
isp relay will eventually fail with an auth error, but I won't notice it 
for some time because that's all done in the background.

Sooo... if I don't actually want any restrictions on fetchmail, then is 
there any reason NOT to use sendmail for delivery instead of SMTP?

>>  The problem is, if I add smtp_sender_restrictions in the postfix config
>>  (main.cf), then those restrictions apply to all connections; both from
>>  alpine and fetchmail.  I'm fairly certain there's a way to distinguish
>>  this by adding something to master.cf (still figuring that part out), but
>>  the key is, there needs to be a way to distinguish between the two.  I
>>  think if fetchmail uses sendmail instead of smtp, I can use that to setup
>>  restrictions based on smtp connections (alpine/outbound) and restrictions
>>  based on sendmail connections (fetchmail/inbound).
>
> You can add another smtpd listener (right hand side of master.cf) in Postfix 
> on a different port (left-hand side of master.cf, you can also give numbers 
> of ports instead of service names) and configure that with its own option 
> set. If you indent 2nd, 3rd, ... lines Postfix reads them as continuation of 
> the previous line in master.cf, and it should have relevant examples.

Yes, that is the other option I was looking at.  But using sendmail 
instead of a separate smtpd listener seemed like the simpler option as 
long as I don't need/want any local processing of inbound email by smtpd. 
Maybe in the future I'll think of something, but for right now I don't 
think smtpd would be adding anything to the process... basically, I just 
want fetchmail to get the mail and get it to the users INBOX.

And if I do want to process the incoming mail in some way in the future, I 
was planning to investigate some combination of postfix, dovecot and 
sieve.  At this point I don't really know anything about that, other than 
it seems to be possible.

Re: [Fetchmail-users] Local delivery - smtp or sendmail

[Matthias A. <mat...@gm...>]

Am 09.11.25 um 01:50 schrieb Hans Carlson via Fetchmail-users:
> I'm in the process of upgrading my ancient local-only little home mail 
> server and was wondering how much it matters if I use smtp or sendmail 
> for local delivery.
>
> My system is Fedora 43:
>   fetchmail: 6.5.6
>   postfix:   3.10.3
>   alpine:    2.26
>
> Keep reading for details of my setup and why I'm asking.
>
> My setup is only used for 2 users, with a few email addresses for each 
> user.  The MTA is postfix, configured to listen only on loopback port 
> 25 with a self signed cert.  There is no access to this server from 
> the outside.  The only things that access the MTA are on the same 
> host: alpine for relaying outbound mail and fetchmail for inbound mail.
>
> fetchmail is started/stopped daily for each user via cron and each 
> user has the same basic fetchmailrc:
>
>   set daemon 900
>   set logfile log/fetchmail.log
>   defaults
>   timeout 120
>   fetchall
>   nokeep
>   fetchlimit 50
>   ssl
>
>   poll ...
>   poll ...
>   poll ...
>
> I'm trying to configure smtpd_sender_restrictions in postfix, mainly 
> so I'll get an immediate failure if I've added a new email address 
> that hasn't been configured in postfix.

Those are only available through SMTP, not through most 
/usr/{lib,sbin}/sendmail wrappers (certainly not Postfix's).

> The problem is, if I add smtp_sender_restrictions in the postfix 
> config (main.cf), then those restrictions apply to all connections; 
> both from alpine and fetchmail.  I'm fairly certain there's a way to 
> distinguish this by adding something to master.cf (still figuring that 
> part out), but the key is, there needs to be a way to distinguish 
> between the two.  I think if fetchmail uses sendmail instead of smtp, 
> I can use that to setup restrictions based on smtp connections 
> (alpine/outbound) and restrictions based on sendmail connections 
> (fetchmail/inbound).

You can add another smtpd listener (right hand side of master.cf) in 
Postfix on a different port (left-hand side of master.cf, you can also 
give numbers of ports instead of service names) and configure that with 
its own option set. If you indent 2nd, 3rd, ... lines Postfix reads them 
as continuation of the previous line in master.cf, and it should have 
relevant examples.

HTH

Matthias

[Fetchmail-users] Local delivery - smtp or sendmail

[Hans C. <fo...@gm...>]

I'm in the process of upgrading my ancient local-only little home mail 
server and was wondering how much it matters if I use smtp or sendmail for 
local delivery.

My system is Fedora 43:
   fetchmail: 6.5.6
   postfix:   3.10.3
   alpine:    2.26

Keep reading for details of my setup and why I'm asking.

My setup is only used for 2 users, with a few email addresses for each 
user.  The MTA is postfix, configured to listen only on loopback port 25 
with a self signed cert.  There is no access to this server from the 
outside.  The only things that access the MTA are on the same host: alpine 
for relaying outbound mail and fetchmail for inbound mail.

fetchmail is started/stopped daily for each user via cron and each user 
has the same basic fetchmailrc:

   set daemon 900
   set logfile log/fetchmail.log
   defaults
   timeout 120
   fetchall
   nokeep
   fetchlimit 50
   ssl

   poll ...
   poll ...
   poll ...

I'm trying to configure smtpd_sender_restrictions in postfix, mainly so 
I'll get an immediate failure if I've added a new email address that 
hasn't been configured in postfix.

The problem is, if I add smtp_sender_restrictions in the postfix config 
(main.cf), then those restrictions apply to all connections; both from 
alpine and fetchmail.  I'm fairly certain there's a way to distinguish 
this by adding something to master.cf (still figuring that part out), but 
the key is, there needs to be a way to distinguish between the two.  I 
think if fetchmail uses sendmail instead of smtp, I can use that to setup 
restrictions based on smtp connections (alpine/outbound) and restrictions 
based on sendmail connections (fetchmail/inbound).

Soooo, back to the original question... on my little local-only server are 
there any pros/cons I should be aware of if I add a fetchmail mda config 
line to use sendmail (or whatever the postfix equiv is) instead of the 
default smtp?

An alternative, might be to distinguish the two by the fact that alpine 
connects to smtp via ipv4 and fetchmail connects via ipv6.  Maybe I could 
use that - although I was considering completely disabling ipv6 since I 
don't plan to ever use it... at least not locally.  Then again, maybe this 
IS a use for ipv6.

Thanks for any advice.

Re: [Fetchmail-users] IDLE Socket error After IP reconnect

[Matthias A. <mat...@gm...>]

Am 01.11.25 um 23:42 schrieb Mailu via Fetchmail-users:
> Oh sorry, trying to do my best on the go with iOS mail :/
> Hope this is better even though not quoting this time anyways...
>   
> I just wanted to ask if the reason can really be the server employing short timeouts.
> Timeouts seem rather reasonable: I noticed that *exactly* every 200mins I receive the socket error (except for the additional errors in the time slots around the IP reconnect).

Hum. 200 min without receiving mail? Can you get me debug logs? See the 
fetchmail FAQ, <https://www.fetchmail.info/fetchmail-FAQ.html#G3>

> I would still say that this is expected as eventually every server will disconnect on their side. Or do you mean that fetchmail should actively reconnect every 30mins by default? If so, this seems not to work for me...

It's certainly so that fetchmail has had support for a long time to deal 
with servers disconnecting anytime - that's just a nearly normal 
situation in networks.

Re: [Fetchmail-users] IDLE Socket error After IP reconnect

[Mailu <lol...@we...>]

Oh sorry, trying to do my best on the go with iOS mail :/
Hope this is better even though not quoting this time anyways...
 
I just wanted to ask if the reason can really be the server employing short timeouts.
Timeouts seem rather reasonable: I noticed that *exactly* every 200mins I receive the socket error (except for the additional errors in the time slots around the IP reconnect).
I would still say that this is expected as eventually every server will disconnect on their side. Or do you mean that fetchmail should actively reconnect every 30mins by default? If so, this seems not to work for me...

> Am 01.11.2025 um 13:14 schrieb Matthias Andree via Fetchmail-users <fet...@li...>:
> 
> Your quoting is extremely broken. Please use a *WORKING* mailer to quote properly and mark quotes.
> 
> Being "professional" (aka paid for what they do) does not mean "free of surprises" or "violations of standards".
> 
> 
>> Am 01.11.25 um 13:04 schrieb Mailu via Fetchmail-users:
>> This would be a server employing short timeouts... you can work around that with idletimeout 280 or similar (in rcfile or command line argument) on supported versions of fetchmail. Check after what time the server gives you a TIMEOUT and then set idletimeout several seconds short of that.
>> 
>> Oh really? Shouldn‘t I receive this Message more often then and Not only 1-2x a times a day?
>> 
>>> Yes, assuming the server is compliant and does _not_ try to be extra smart and delete the message after the first attempt to read it.
>> I Hope it is not. I am Talking about Web.de. A large German mail provider. Basically the Same as GMX. Should be professional, but you never know in the Details…
>> 
>>> The elegant ways are 1. don't use IDLE if the server employs nonconforming timeouts, 2. get a static IP.
>>> 
>>> The practical way is "ignore and move on".
>> Thanks for confirming!
>> 
>> 
>> _______________________________________________
>> Fetchmail-users mailing list
>> Fet...@li...
>> https://lists.sourceforge.net/lists/listinfo/fetchmail-users
> 
> 
> 
> 
> _______________________________________________
> Fetchmail-users mailing list
> Fet...@li...
> https://lists.sourceforge.net/lists/listinfo/fetchmail-users

Re: [Fetchmail-users] IDLE Socket error After IP reconnect

[Matthias A. <mat...@gm...>]

Your quoting is extremely broken. Please use a *WORKING* mailer to quote 
properly and mark quotes.

Being "professional" (aka paid for what they do) does not mean "free of 
surprises" or "violations of standards".


Am 01.11.25 um 13:04 schrieb Mailu via Fetchmail-users:
> This would be a server employing short timeouts... you can work around that with idletimeout 280 or similar (in rcfile or command line argument) on supported versions of fetchmail. Check after what time the server gives you a TIMEOUT and then set idletimeout several seconds short of that.
>
> Oh really? Shouldn‘t I receive this Message more often then and Not only 1-2x a times a day?
>
>> Yes, assuming the server is compliant and does _not_ try to be extra smart and delete the message after the first attempt to read it.
> I Hope it is not. I am Talking about Web.de. A large German mail provider. Basically the Same as GMX. Should be professional, but you never know in the Details…
>
>> The elegant ways are 1. don't use IDLE if the server employs nonconforming timeouts, 2. get a static IP.
>>
>> The practical way is "ignore and move on".
> Thanks for confirming!
>
>
> _______________________________________________
> Fetchmail-users mailing list
> Fet...@li...
> https://lists.sourceforge.net/lists/listinfo/fetchmail-users

Re: [Fetchmail-users] IDLE Socket error After IP reconnect

[Mailu <lol...@we...>]

This would be a server employing short timeouts... you can work around that with idletimeout 280 or similar (in rcfile or command line argument) on supported versions of fetchmail. Check after what time the server gives you a TIMEOUT and then set idletimeout several seconds short of that.

Oh really? Shouldn‘t I receive this Message more often then and Not only 1-2x a times a day?

> Yes, assuming the server is compliant and does _not_ try to be extra smart and delete the message after the first attempt to read it.

I Hope it is not. I am Talking about Web.de. A large German mail provider. Basically the Same as GMX. Should be professional, but you never know in the Details…

> The elegant ways are 1. don't use IDLE if the server employs nonconforming timeouts, 2. get a static IP.
> 
> The practical way is "ignore and move on".

Thanks for confirming!

Re: [Fetchmail-users] IDLE Socket error After IP reconnect

[Matthias A. <mat...@gm...>]

Am 01.11.25 um 11:44 schrieb LolliPOPsystem--- via Fetchmail-users:
> Hi,
>
> New User here.
>
> Setup: Running a Python script starting one instance for each of multiple mail accounts in IDLE mode leveraging individual $FETCHMAILHOME environment variables like this:
> https://www.linuxquestions.org/questions/linux-software-2/correct-fetchmail-with-idle-and-multiple-servers-config-642089/
>
> Fetchmail runs behind a router with Dynamic IP.
>
> Occationally (and particularly After IP resets), I receive the following error for each instance and the script restarts the fetchmail instances:
>
> Nov 01 08:48:54 fetchmail: Received BYE response from IMAP server:  TIMEOUT

This would be a server employing short timeouts... you can work around 
that with idletimeout 280 or similar (in rcfile or command line 
argument) on supported versions of fetchmail. Check after what time the 
server gives you a TIMEOUT and then set idletimeout several seconds 
short of that.

> Nov 01 08:48:54 fetchmail: re-poll failed
> Nov 01 08:48:54 fetchmail: socket error while fetching from…
>
> I assume this is due to the IP change and the connection cannot be re-established. I think this will Not harm fetchmail’s functionality and I can safely restart it without fearing any mail will be lost or not fetched.

Yes, assuming the server is compliant and does _not_ try to be extra 
smart and delete the message after the first attempt to read it.

Worst case if the IP changes in the middle of a transfer you might end 
up with two copies of a message, one complete and one incomplete.

> Can you confirm this understanding and maybe explain better what is Happening and what an elegant way is to mitigate?

The elegant ways are 1. don't use IDLE if the server employs 
nonconforming timeouts, 2. get a static IP.

The practical way is "ignore and move on".

[Fetchmail-users] IDLE Socket error After IP reconnect

[<Lol...@we...>]

Hi,

New User here.

Setup: Running a Python script starting one instance for each of multiple mail accounts in IDLE mode leveraging individual $FETCHMAILHOME environment variables like this:
https://www.linuxquestions.org/questions/linux-software-2/correct-fetchmail-with-idle-and-multiple-servers-config-642089/

Fetchmail runs behind a router with Dynamic IP.

Occationally (and particularly After IP resets), I receive the following error for each instance and the script restarts the fetchmail instances:

Nov 01 08:48:54 fetchmail: Received BYE response from IMAP server:  TIMEOUT
Nov 01 08:48:54 fetchmail: re-poll failed
Nov 01 08:48:54 fetchmail: socket error while fetching from…

I assume this is due to the IP change and the connection cannot be re-established. I think this will Not harm fetchmail’s functionality and I can safely restart it without fearing any mail will be lost or not fetched.

Can you confirm this understanding and maybe explain better what is Happening and what an elegant way is to mitigate?

Thanks!

[Fetchmail-users] The 6.6.0 release of fetchmail is available (TLS support for SMTP, incl. STARTTLS)

[Matthias A. <mat...@gm...>]

The 6.6.0 release of fetchmail is now available at the usual locations,
including <https://downloads.sourceforge.net/project/fetchmail/branch_6.6/>
and <https://gitlab.com/fetchmail/fetchmail/-/releases#6.6.0>


Distributors: please install the revived Italian translation!
=============


The source archive is available at:
<https://downloads.sourceforge.net/project/fetchmail/branch_6.6/fetchmail-6.6.0.tar.xz/download>
<https://gitlab.com/-/project/188557/uploads/390ec2599b4f1f47bb18f37b7e5b601b/fetchmail-6.6.0.tar.xz>

The detached GnuPG signature is attached and available at:
<https://downloads.sourceforge.net/project/fetchmail/branch_6.6/fetchmail-6.6.0.tar.xz.asc/download>
<https://gitlab.com/-/project/188557/uploads/5ef2d3ea82be1133e3630ad0d7e789dc/fetchmail-6.6.0.tar.xz.asc>

The SHA256 hashes for the tarballs are:
SHA2-256(fetchmail-6.6.0.tar.xz)= 7b9c19e6683e827d556751aa5db5d44b961e87be8b3087535b4909ba1b59321c


NOTE that fetchmail-6.5.X will go out of support at the end of 2025.

The 6.6.0 release does not intentionally change configuration files,
command line arguments, or behavior, and should therefore be a drop-in
compatible replacement for 6.5.X releases.


Here are the release notes:
--------------------------------------------------------------------------------
fetchmail-6.6.0 (released 2025-10-27, 32381 LoC):

## FEATURE:
* SMTP TLS and STARTTLS support. By default, this works opportunistically,
  attempting to set up a TLS connection to the smtphost if it understands EHLO
  and offers STARTTLS, but will not enforce peer certificate validity for
  compatibility, esp. because "localhost" (the default SMTP host) usually
  isn't listed in the X.509 certificates.
    Behavior can be tweaked by adding /notls (cleartext connection), /tls
  (TLS-wrapped connection, negotiating TLS before conversing otherwise),
  or /starttls (requiring EHLO to offer STARTTLS, requesting the latter and
  requiring the server certificate to validate) to the SMTP host's name.
    Also, you can add /tlsproto=... where ... accepts the same parameters
  as the --sslproto option, which see.
  Ports, if not specified, default to 25 for opportunistic and /notls modes,
  465 for /tls and 587 for /starttls, but can be overridden either by giving,
  say /25 or /smtp for /starttls.

## TRANSLATIONS were updated by these fine people (randomized order):
* it:    Luca Vercelli [Italian]
* pl:    Jakub Bogusz [Polish]
* fr:    Frédéric Marchal [French]
* sv:    Göran Uddeborg [Swedish]
* sq:    Besnik Bleta [Albanian]
* ro:    Remus-Gabriel Chelu [Romanian]
* ja:    Takeshi Hamasaki [Japanese]
* de:    Matthias Andree [German]
* cs:    Petr Pisar [Czech]
* eo:    Keith Bowes [Esperanto]

-------------------------------------------------------------------------------

Re: [Fetchmail-users] Migrating from 6.3.24 to 6.4.38

[Dennis P. <da...@be...>]

On 10/26/2025 2:50 PM, Andrew C Aitchison wrote:
> On Sun, 26 Oct 2025, Dennis Putnam via Fetchmail-users wrote:
>
>> I have fetchmail (6.3.24) running on CentOS 7 and am moving it to 
>> Ubuntu fetchmail (6.4.38). I thought the move would be pretty much 
>> seamless but was wrong. Apparently there are differences in the 
>> fetchmailrc file that I am not understanding. I am using it for 
>> gmail. Here is a sanitized portion of my fetchmailrc:
>>
>> *set logfile "/var/log/maillog"
>> set postmaster "mailman"
>> set bouncemail
>> set no spambounce
>> set properties ""
>> defaults:
>>         timeout 240
>> poll pop.gmail.com
>>         proto POP3
>>         user "uuu...@gm..."
>>         is uuuuuuuuu
>>         password 'kkkkkkkkkkkkkkkr'
>>         preconnect "echo `date '+%b %d %H:%M:%S'` uuuuuuuuu"
>>         ssl
>>         nokeep
>>         fetchall
>
>> Trying to connect to 127.0.0.1/25...connected.
>>
>> *At this point it seems to hang and does not proceed to the next 
>> mailbox.
>> Can someone help me fix my fetchmailrc? TIA.
>
> I think (but am not sure) that the problem may be with your local mail
> server (MTA - probably postfix or exim) rather than fetchmail.
> I would look in the logs for that.
>
Hi Andrew,

You were exactly right. I never would have made that connection myself. 
Thanks.

Re: [Fetchmail-users] Migrating from 6.3.24 to 6.4.38

[Andrew C A. <fet...@ai...>]

On Sun, 26 Oct 2025, Dennis Putnam via Fetchmail-users wrote:

> I have fetchmail (6.3.24) running on CentOS 7 and am moving it to Ubuntu 
> fetchmail (6.4.38). I thought the move would be pretty much seamless but was 
> wrong. Apparently there are differences in the fetchmailrc file that I am not 
> understanding. I am using it for gmail. Here is a sanitized portion of my 
> fetchmailrc:
>
> *set logfile "/var/log/maillog"
> set postmaster "mailman"
> set bouncemail
> set no spambounce
> set properties ""
> defaults:
>         timeout 240
> poll pop.gmail.com
>         proto POP3
>         user "uuu...@gm..."
>         is uuuuuuuuu
>         password 'kkkkkkkkkkkkkkkr'
>         preconnect "echo `date '+%b %d %H:%M:%S'` uuuuuuuuu"
>         ssl
>         nokeep
>         fetchall

> Trying to connect to 127.0.0.1/25...connected.
>
> *At this point it seems to hang and does not proceed to the next mailbox.
> Can someone help me fix my fetchmailrc? TIA.

I think (but am not sure) that the problem may be with your local mail
server (MTA - probably postfix or exim) rather than fetchmail.
I would look in the logs for that.

-- 
Andrew C. Aitchison                      Kendal, UK
                    an...@ai...

[Fetchmail-users] Migrating from 6.3.24 to 6.4.38

[Dennis P. <da...@be...>]

I have fetchmail (6.3.24) running on CentOS 7 and am moving it to Ubuntu 
fetchmail (6.4.38). I thought the move would be pretty much seamless but 
was wrong. Apparently there are differences in the fetchmailrc file that 
I am not understanding. I am using it for gmail. Here is a sanitized 
portion of my fetchmailrc:

*set logfile "/var/log/maillog"
set postmaster "mailman"
set bouncemail
set no spambounce
set properties ""
defaults:
         timeout 240
poll pop.gmail.com
         proto POP3
         user "uuu...@gm..."
         is uuuuuuuuu
         password 'kkkkkkkkkkkkkkkr'
         preconnect "echo `date '+%b %d %H:%M:%S'` uuuuuuuuu"
         ssl
         nokeep
         fetchall

*Here is the output when I run it:

*Old UID list from pop.gmail.com:
  <empty>

Scratch list of UIDs:
  <empty>

fetchmail: 6.4.38 querying pop.gmail.com (protocol POP3) at Sun 26 Oct 
2025 01:00:02 PM EDT: poll started
Oct 26 13:00:02 uuuuuuuu
Trying to connect to 2607:f8b0:4002:c05::6c/995...connected.
fetchmail: SSL verify callback depth 2: verify_ok == 1, err = 0, ok
fetchmail: Certificate chain, from root to peer, starting at depth 2:
fetchmail: Issuer Organization: Google Trust Services LLC
fetchmail: Issuer CommonName: GTS Root R4
fetchmail: Subject CommonName: GTS Root R4
fetchmail: SSL verify callback depth 1: verify_ok == 1, err = 0, ok
fetchmail: Certificate at depth 1:
fetchmail: Issuer Organization: Google Trust Services LLC
fetchmail: Issuer CommonName: GTS Root R4
fetchmail: Subject CommonName: WE2
fetchmail: SSL verify callback depth 0: verify_ok == 1, err = 0, ok
fetchmail: Server certificate:
fetchmail: Issuer Organization: Google Trust Services
fetchmail: Issuer CommonName: WE2
fetchmail: Subject CommonName: pop.gmail.com
fetchmail: Subject Alternative Name: pop.gmail.com
fetchmail: pop.gmail.com key fingerprint: 
A3:FD:1E:8B:34:9B:B3:89:BF:E9:67:5B:42:DE:4E:8A
fetchmail: SSL/TLS: using protocol TLSv1.3, cipher 
TLS_AES_256_GCM_SHA384, 256/256 secret/processed bits
fetchmail: POP3< +OK Gpop ready for requests from 
2600:1700:5cac:4600:431:e582:3ec6:f485 956f58d0204a3-63e8b2c032
amb144186743d50
fetchmail: POP3> CAPA
fetchmail: POP3< +OK Capability list follows
fetchmail: POP3< USER
fetchmail: POP3< RESP-CODES
fetchmail: POP3< EXPIRE 0
fetchmail: POP3< LOGIN-DELAY 300
fetchmail: POP3< TOP
fetchmail: POP3< UIDL
fetchmail: POP3< X-GOOGLE-RICO
fetchmail: POP3< SASL PLAIN XOAUTH2 OAUTHBEARER
fetchmail: POP3< .
fetchmail: POP3> USER **uuuuuuuu**@gmail.com
fetchmail: POP3< +OK send PASS
fetchmail: POP3> PASS *
fetchmail: POP3< +OK Welcome.
fetchmail: selecting or re-polling default folder
fetchmail: POP3> STAT
fetchmail: POP3< +OK 3 22429
3 messages for **uuuuuuuu**@gmail.com at pop.gmail.com (22429 octets).
fetchmail: POP3> LIST 1
fetchmail: POP3< +OK 1 7418
fetchmail: POP3> RETR 1
fetchmail: POP3< +OK message follows
reading message **uuuuuuuu**@gma...@po...:1 of 3 (7418 
octets) About to rewrite Return-Path: <xx...@dd...>...
...rewritten version is Return-Path: <**xx...@dd...**>.
About to rewrite Cc: **xx...@dd...**...
...rewritten version is Cc: **xx...@dd...**.
About to rewrite To: **uuuuuuuu**@gmail.com...
...rewritten version is To: **uuuuuuuu**@gmail.com.
About to rewrite From: **xx...@dd...**...
...rewritten version is From: **xx...@dd...**.net.

Trying to connect to 127.0.0.1/25...connected.

*At this point it seems to hang and does not proceed to the next 
mailbox. Can someone help me fix my fetchmailrc? TIA.

Re: [Fetchmail-users] Upgrade to fetchmail 6.5.6

[<t_...@ti...>]

On Tue, 21 Oct 2025 19:48:39 +0200
Matthias Andree via Fetchmail-users
<fet...@li...> wrote:


> Angelo,
> 
> you can either put --syslog on your fetchmail command lines in the 
> fetchmail.service unit files, or you can add
> 
>     set syslog
> 
> early (before skip and poll lines) to your various .fetchmailrc files.
> In that case, fetchmail will send directly to syslog without
> timestamps, and on systemd-based systems, syslog also usually ends up
> where journalctl looks.
> 
> HTH
> 
> Best regards,
> Matthias

OK Matthias,

I'll use --syslog on the command lines

Thanks you all for taking the time to answer my question.

Ciao,
Angelo 

> 
> 
> 
> _______________________________________________
> Fetchmail-users mailing list
> Fet...@li...
> https://lists.sourceforge.net/lists/listinfo/fetchmail-users

Re: [Fetchmail-users] Upgrade to fetchmail 6.5.6

[Matthias A. <mat...@gm...>]

Am 21.10.25 um 12:45 schrieb Andrew C Aitchison:
>
> On Mon, 20 Oct 2025, t_...@ti... wrote:
>
>> I've read the FAQ and I have to say that my fetchmail works very fine.
>> I only wonder why it includes timestamps in logs since the last
>> upgrade.
>
> Probably related to this * fetchmail, with --logfile, now logs time 
> stamps
> into the file, in localtime
>   and in the format "Jun 20 23:45:01 fetchmail: ". It will be localized
> through
>   the environment variables LC_TIME (or LC_ALL) and TZ.
>   Contributed by Holger Hoffstätte.
> from /home/src/ubuntu/fetchmail/fetchmail-6.5.0/NEWS
>
> I think it is for the benefit of IMAP IDLE.
> In that mode fetchmail keeps an open connection to the server
> and logs whenever the server offers/pushes a new messages.
>
> IMAP IDLE saves fetchmail from starting a new connection (including 
> setting up encryption) every few minutes and allows it to be completely
> idle until there is a new message.
> It doesn't support more than one remote mailbox
> (unless you set up separate fetchmail processes with separate configs).
>
I've received relevant parts of Angelo's configuration directly, and 
part of the .service file is this section:

> [Service]
> Type=oneshot
> StandardOutput=journal
> StandardError=journal
> SyslogIdentifier=fetchmail
> SyslogFacility=mail
> SyslogLevel=warning

According to the .fetchmailrc file, Angelo does not use a logfile, so 
Holger's contribution does not apply;
instead, fetchmail logs to stdout/stderr, adds a timestamp as it would 
when logging to console
-- and systemd captures the output and adds its own timestamp for the 
journal.

Angelo,

you can either put --syslog on your fetchmail command lines in the 
fetchmail.service unit files, or you can add

    set syslog

early (before skip and poll lines) to your various .fetchmailrc files.
In that case, fetchmail will send directly to syslog without timestamps, 
and on systemd-based systems, syslog also usually ends up where 
journalctl looks.

HTH

Best regards,
Matthias

Re: [Fetchmail-users] Upgrade to fetchmail 6.5.6

[Andrew C A. <fet...@ai...>]

On Mon, 20 Oct 2025, t_...@ti... wrote:

> I've read the FAQ and I have to say that my fetchmail works very fine.
> I only wonder why it includes timestamps in logs since the last
> upgrade.

Probably related to this * fetchmail, with --logfile, now logs time stamps
into the file, in localtime
   and in the format "Jun 20 23:45:01 fetchmail: ". It will be localized
through
   the environment variables LC_TIME (or LC_ALL) and TZ.
   Contributed by Holger Hoffstätte.
from /home/src/ubuntu/fetchmail/fetchmail-6.5.0/NEWS

I think it is for the benefit of IMAP IDLE.
In that mode fetchmail keeps an open connection to the server
and logs whenever the server offers/pushes a new messages.

IMAP IDLE saves fetchmail from starting a new connection (including 
setting up encryption) every few minutes and allows it to be completely
idle until there is a new message.
It doesn't support more than one remote mailbox
(unless you set up separate fetchmail processes with separate configs).

-- 
Andrew C. Aitchison                      Kendal, UK
                    an...@ai...

Re: [Fetchmail-users] Upgrade to fetchmail 6.5.6

[<t_...@ti...>]

On Mon, 20 Oct 2025 20:16:34 +0200
Matthias Andree via Fetchmail-users
<fet...@li...> wrote:

> Am 18.10.25 um 22:51 schrieb t_...@ti...:
> > On Sat, 18 Oct 2025 15:01:08 -0500
> > axreios <rp...@am...> wrote:
> >
> >> On Sat, Oct 18, 2025 at 07:36:52PM +0200, t_...@ti... wrote:
> >>> Hi list,
> >>> after upgrading to 6.5.6 I've noticed that syslog messages include
> >>> the timestamp. Since I'm starting fetchmail thru systemd, the
> >>> timestamp is already included in journal. Is it possible to remove
> >>> timestamp from syslog message ?
> >>>
> >>> Here is my uname -a :
> >>>
> >>> Linux agmob3 6.16.12-100.fc41.x86_64 #1 SMP PREEMPT_DYNAMIC Sun
> >>> Oct 12 18:44:08 UTC 2025 x86_64 GNU/Linux
> >>>
> >>> Thanks in advance for any answer.
> >>> Ciao
> >>> Angelo
> >>>
> >> Are you getting the timestamp listed twice?  I am not seeing that,
> >> though I am not using systemd.  I do start fetchmail from the
> >> command line with the --syslog option.  FWIW.
> >> 	ax
> >>
> >>
> > Hi,
> > thanks for your answer.
> > here is an example of messages from my journalctl
> >
> > Oct 18 15:03:14 fetchmail[1601465]: Oct 18 15:03:14 fetchmail: 77
> > messages (77 seen) for ...... Oct 18 15:03:14 fetchmail[1601465]:
> > Oct 18 15:03:14 fetchmail: 43 messages (43 seen) for ...... Oct 18
> > 15:03:15 fetchmail[1601825]: Oct 18 15:03:15 fetchmail: No mail for
> > ...... Oct 18 15:03:15 fetchmail[1601825]: Oct 18 15:03:15
> > fetchmail: No mail for ......
> >
> > with previous fetchmail version (6.4) this was not happening.
> 
> 
> Please heed <https://www.fetchmail.info/fetchmail-FAQ.html#G3>
> 
> and please include information on how your systemd file looks like.
> 
> 

Thanks for answering.

I've read the FAQ and I have to say that my fetchmail works very fine.
I only wonder why it includes timestamps in logs since the last
upgrade.


here is my systemd service (triggered by timer)
=======================================================
[Unit]
Description=Fetch My Mails  
Documentation=man:fetchmail(1)
ConditionPathExists=/tmp/XISON

[Service]
Type=oneshot
StandardOutput=journal
StandardError=journal
SyslogIdentifier=fetchmail
SyslogFacility=mail
SyslogLevel=warning

### check for full internet connection
ExecStartPre=/Scripts/ck_full_conn

### fetch mails 
ExecStart=fetchmail -f /home/.MYFMRC
SuccessExitStatus=1

### fetch spam
ExecStart=fetchmail -f /home/.MYFMRC_SPAM
SuccessExitStatus=1

Restart=on-failure
RestartSec=90
=======================================================


On the config files I use the follwing options :
=======================================================
set invisible
set no bouncemail
defaults proto imap timeout 20 bad-header accept
=======================================================

Should I provide any other info ?

Thanks again.
Regards
Angelo


> _______________________________________________
> Fetchmail-users mailing list
> Fet...@li...
> https://lists.sourceforge.net/lists/listinfo/fetchmail-users

Re: [Fetchmail-users] Upgrade to fetchmail 6.5.6

[Matthias A. <mat...@gm...>]

Am 18.10.25 um 22:51 schrieb t_...@ti...:
> On Sat, 18 Oct 2025 15:01:08 -0500
> axreios <rp...@am...> wrote:
>
>> On Sat, Oct 18, 2025 at 07:36:52PM +0200, t_...@ti... wrote:
>>> Hi list,
>>> after upgrading to 6.5.6 I've noticed that syslog messages include
>>> the timestamp. Since I'm starting fetchmail thru systemd, the
>>> timestamp is already included in journal. Is it possible to remove
>>> timestamp from syslog message ?
>>>
>>> Here is my uname -a :
>>>
>>> Linux agmob3 6.16.12-100.fc41.x86_64 #1 SMP PREEMPT_DYNAMIC Sun Oct
>>> 12 18:44:08 UTC 2025 x86_64 GNU/Linux
>>>
>>> Thanks in advance for any answer.
>>> Ciao
>>> Angelo
>>>
>> Are you getting the timestamp listed twice?  I am not seeing that,
>> though I am not using systemd.  I do start fetchmail from the command
>> line with the --syslog option.  FWIW.
>> 	ax
>>
>>
> Hi,
> thanks for your answer.
> here is an example of messages from my journalctl
>
> Oct 18 15:03:14 fetchmail[1601465]: Oct 18 15:03:14 fetchmail: 77 messages (77 seen) for ......
> Oct 18 15:03:14 fetchmail[1601465]: Oct 18 15:03:14 fetchmail: 43 messages (43 seen) for ......
> Oct 18 15:03:15 fetchmail[1601825]: Oct 18 15:03:15 fetchmail: No mail for ......
> Oct 18 15:03:15 fetchmail[1601825]: Oct 18 15:03:15 fetchmail: No mail for ......
>
> with previous fetchmail version (6.4) this was not happening.


Please heed <https://www.fetchmail.info/fetchmail-FAQ.html#G3>

and please include information on how your systemd file looks like.

Re: [Fetchmail-users] Upgrade to fetchmail 6.5.6

[<t_...@ti...>]

On Sat, 18 Oct 2025 15:01:08 -0500
axreios <rp...@am...> wrote:

> On Sat, Oct 18, 2025 at 07:36:52PM +0200, t_...@ti... wrote:
> >Hi list,
> >after upgrading to 6.5.6 I've noticed that syslog messages include
> >the timestamp. Since I'm starting fetchmail thru systemd, the
> >timestamp is already included in journal. Is it possible to remove
> >timestamp from syslog message ?
> >
> >Here is my uname -a :
> >
> >Linux agmob3 6.16.12-100.fc41.x86_64 #1 SMP PREEMPT_DYNAMIC Sun Oct
> >12 18:44:08 UTC 2025 x86_64 GNU/Linux
> >
> >Thanks in advance for any answer.
> >Ciao
> >Angelo
> >
> 
> Are you getting the timestamp listed twice?  I am not seeing that,
> though I am not using systemd.  I do start fetchmail from the command
> line with the --syslog option.  FWIW.
> 	ax
> 
>

Hi,
thanks for your answer.
here is an example of messages from my journalctl

Oct 18 15:03:14 fetchmail[1601465]: Oct 18 15:03:14 fetchmail: 77 messages (77 seen) for ......
Oct 18 15:03:14 fetchmail[1601465]: Oct 18 15:03:14 fetchmail: 43 messages (43 seen) for ......
Oct 18 15:03:15 fetchmail[1601825]: Oct 18 15:03:15 fetchmail: No mail for ......
Oct 18 15:03:15 fetchmail[1601825]: Oct 18 15:03:15 fetchmail: No mail for ......

with previous fetchmail version (6.4) this was not happening.

Thanks again.
Ciao,

Angelo

Re: [Fetchmail-users] Upgrade to fetchmail 6.5.6

[axreios <rp...@am...>]

On Sat, Oct 18, 2025 at 07:36:52PM +0200, t_...@ti... wrote:
>Hi list,
>after upgrading to 6.5.6 I've noticed that syslog messages include the timestamp.
>Since I'm starting fetchmail thru systemd, the timestamp is already included in journal.
>Is it possible to remove timestamp from syslog message ?
>
>Here is my uname -a :
>
>Linux agmob3 6.16.12-100.fc41.x86_64 #1 SMP PREEMPT_DYNAMIC Sun Oct 12 18:44:08 UTC 2025 x86_64 GNU/Linux
>
>Thanks in advance for any answer.
>Ciao
>Angelo
>

Are you getting the timestamp listed twice?  I am not seeing that,
though I am not using systemd.  I do start fetchmail from the command
line with the --syslog option.  FWIW.
	ax