[Standards] MAM & Offline messages

[Sergey Dobrov]

Hello folks,

I am wondering what should happen if server has no support of offline 
messages storage but does have support of MAM and someone sends offline 
message, should server send back "service-unavailable" or not?


Thanks.

Re: [Standards] Last connection time of friend

[Peter Waher]

Hello Tobias

Thanks a lot. This is exactly what I need.

Best regards,
Peter Waher

From: Tobias Markmann [mailto:tmarkm...@googlemail.com]
Sent: den 3 mars 2014 14:30
To: XMPP Standards
Subject: Re: [Standards] Last connection time of friend


Hi Peter,

On 2 Mar 2014 20:10, "Peter Waher" 
mailto:peter.wa...@clayster.com>> wrote:
>
> Hello
>
>
>
> Is there a means to figure out the last time a friend connected to its XMPP 
> server? Can you ask the server hosting the user account for last connection 
> time? (You might not have been online to receive the presence message when 
> the user went offline.)
>

The following XEP might also be of interest: 
http://xmpp.org/extensions/xep-0318.html

Re: [Standards] Last connection time of friend

[Peter Waher]

Hello Ben

Thanks for your input. You're correct that you can always send the probe. What 
worries me is how many servers will actually route it. It only says servers 
SHOULD route the probe in RFC 6121.

How would you do this use case better, than by sending a probe?

Best regards,
Peter Waher


From: Ben Langfeld [mailto:b...@langfeld.me]
Sent: den 3 mars 2014 14:20
To: XMPP Standards
Subject: Re: [Standards] Last connection time of friend

The point here is that this is a normative recommendation against doing 
presence probes, because there's often a better way to achieve the result. It 
is not denying you the option of doing so.

On 3 March 2014 13:47, Peter Waher 
mailto:peter.wa...@clayster.com>> wrote:
Hello Dave

Thanks for the quick response. This looks like exactly what I need.

I'm a little worried about this statement in RFC 6121 though (and the 
explanation a little circular):

   Presence probes SHOULD NOT be sent by a client, because in general a
   client will not need to send them since the task of gathering
   presence from a user's contacts is managed by the user's server.
   However, if a user's client generates an outbound presence probe then
   the user's server SHOULD route the probe (if the contact is at
   another server) or process the probe (if the contact is at the same
   server) and MUST NOT use its receipt of the presence probe from a
   connected client as the sole cause for returning a stanza or stream
   error to the client.

Sending such probes seems to me to be a great way of finding connection 
problems in Internet of Things networks. Have to do some experiments to find 
out how it works out. Perhaps the explanation for why this should not be done 
by clients can be revised in the RFC?

Best regards,
Peter Waher

From: Dave Cridland [mailto:d...@cridland.net]
Sent: den 2 mars 2014 16:21
To: XMPP Standards

Subject: Re: [Standards] Last connection time of friend

Some servers also respond to a probe with the last offline presence (with a 
timestamp).

On 2 March 2014 19:17, Philipp Hancke 
mailto:fi...@goodadvice.pages.de>> wrote:
Am 02.03.2014 20:08, schrieb Peter Waher:

Is there a means to figure out the last time a friend connected to its XMPP 
server? Can you ask the server hosting the user account for last connection 
time? (You might not have been online to receive the presence message when the 
user went offline.)

http://xmpp.org/extensions/xep-0012.html#offline -- not sure how many servers 
implement it though.

Re: [Standards] Last connection time of friend

[Peter Waher]

Hello Steven

Thanks for your input. The point here is to find a mechanism or a collection of 
tests to help finding out the reason why a collection of devices are not 
online. Is it because they cannot connect to the XMPP server? Or because one of 
many other conditions. One way to estimate if they cannot connect is to look at 
the last time they connected (or disconnected). It’s very different if it was 
only an hour ago, or a month ago.

Best regards,
Peter Waher


From: Steven Lloyd Watkin [mailto:ll...@evilprofessor.co.uk]
Sent: den 3 mars 2014 13:54
To: XMPP Standards
Subject: Re: [Standards] Last connection time of friend

If you know the full jid of the device, then you could always ping it? 
http://xmpp.org/extensions/xep-0199.html

_

Steven Lloyd Watkin
Software Engineer
PHP ::: Java ::: Ruby ::: Node.js ::: XMPP
ll...@evilprofessor.co.uk (email+jid) ::: 
http://www.evilprofessor.co.uk
Facebook / Twitter / Flickr: lloydwatkin

Organiser of WORLD RECORD breaking charity event:
Scuba Santas ::: http://www.scuba-santas.co.uk
Supporting the RNLI & DDRC - 15th December 2013 - NDAC, Chepstow

On 3 March 2014 16:47, Peter Waher 
mailto:peter.wa...@clayster.com>> wrote:
Hello Dave

Thanks for the quick response. This looks like exactly what I need.

I’m a little worried about this statement in RFC 6121 though (and the 
explanation a little circular):

   Presence probes SHOULD NOT be sent by a client, because in general a
   client will not need to send them since the task of gathering
   presence from a user's contacts is managed by the user's server.
   However, if a user's client generates an outbound presence probe then
   the user's server SHOULD route the probe (if the contact is at
   another server) or process the probe (if the contact is at the same
   server) and MUST NOT use its receipt of the presence probe from a
   connected client as the sole cause for returning a stanza or stream
   error to the client.

Sending such probes seems to me to be a great way of finding connection 
problems in Internet of Things networks. Have to do some experiments to find 
out how it works out. Perhaps the explanation for why this should not be done 
by clients can be revised in the RFC?

Best regards,
Peter Waher

From: Dave Cridland [mailto:d...@cridland.net]
Sent: den 2 mars 2014 16:21
To: XMPP Standards
Subject: Re: [Standards] Last connection time of friend

Some servers also respond to a probe with the last offline presence (with a 
timestamp).

On 2 March 2014 19:17, Philipp Hancke 
mailto:fi...@goodadvice.pages.de>> wrote:
Am 02.03.2014 20:08, schrieb Peter Waher:

Is there a means to figure out the last time a friend connected to its XMPP 
server? Can you ask the server hosting the user account for last connection 
time? (You might not have been online to receive the presence message when the 
user went offline.)

http://xmpp.org/extensions/xep-0012.html#offline -- not sure how many servers 
implement it though.

Re: [Standards] Last connection time of friend

[Tobias Markmann]

Hi Peter,

On 2 Mar 2014 20:10, "Peter Waher"  wrote:
>
> Hello
>
>
>
> Is there a means to figure out the last time a friend connected to its
XMPP server? Can you ask the server hosting the user account for last
connection time? (You might not have been online to receive the presence
message when the user went offline.)
>

The following XEP might also be of interest:
http://xmpp.org/extensions/xep-0318.html

Re: [Standards] Last connection time of friend

[Ben Langfeld]

The point here is that this is a normative recommendation against doing
presence probes, because there's often a better way to achieve the result.
It is not denying you the option of doing so.


On 3 March 2014 13:47, Peter Waher  wrote:

>  Hello Dave
>
>
>
> Thanks for the quick response. This looks like exactly what I need.
>
>
>
> I'm a little worried about this statement in RFC 6121 though (and the
> explanation a little circular):
>
>
>
>*Presence probes SHOULD NOT be sent by a client*, because in general a
>
>client will not need to send them since the task of gathering
>
>presence from a user's contacts is managed by the user's server.
>
>However, if a user's client generates an outbound presence probe then
>
>the user's server SHOULD route the probe (if the contact is at
>
>another server) or process the probe (if the contact is at the same
>
>server) and MUST NOT use its receipt of the presence probe from a
>
>connected client as the sole cause for returning a stanza or stream
>
>error to the client.
>
>
>
> Sending such probes seems to me to be a great way of finding connection
> problems in Internet of Things networks. Have to do some experiments to
> find out how it works out. Perhaps the explanation for why this should not
> be done by clients can be revised in the RFC?
>
>
>
> Best regards,
>
> Peter Waher
>
>
>
> *From:* Dave Cridland [mailto:d...@cridland.net]
> *Sent:* den 2 mars 2014 16:21
> *To:* XMPP Standards
>
> *Subject:* Re: [Standards] Last connection time of friend
>
>
>
> Some servers also respond to a probe with the last offline presence (with
> a timestamp).
>
>
>
> On 2 March 2014 19:17, Philipp Hancke  wrote:
>
> Am 02.03.2014 20:08, schrieb Peter Waher:
>
>
>
> Is there a means to figure out the last time a friend connected to its
> XMPP server? Can you ask the server hosting the user account for last
> connection time? (You might not have been online to receive the presence
> message when the user went offline.)
>
>
>
> http://xmpp.org/extensions/xep-0012.html#offline -- not sure how many
> servers implement it though.
>
>
>

Re: [Standards] Last connection time of friend

[Steven Lloyd Watkin]

If you know the full jid of the device, then you could always ping it?
http://xmpp.org/extensions/xep-0199.html

_

Steven Lloyd Watkin
Software Engineer
PHP ::: Java ::: Ruby ::: Node.js ::: XMPP
ll...@evilprofessor.co.uk (email+jid) ::: http://www.evilprofessor.co.uk
Facebook / Twitter / Flickr: lloydwatkin

Organiser of WORLD RECORD breaking charity event:
Scuba Santas ::: http://www.scuba-santas.co.uk
Supporting the RNLI & DDRC - 15th December 2013 - NDAC, Chepstow


On 3 March 2014 16:47, Peter Waher  wrote:

>  Hello Dave
>
>
>
> Thanks for the quick response. This looks like exactly what I need.
>
>
>
> I’m a little worried about this statement in RFC 6121 though (and the
> explanation a little circular):
>
>
>
>*Presence probes SHOULD NOT be sent by a client*, because in general a
>
>client will not need to send them since the task of gathering
>
>presence from a user's contacts is managed by the user's server.
>
>However, if a user's client generates an outbound presence probe then
>
>the user's server SHOULD route the probe (if the contact is at
>
>another server) or process the probe (if the contact is at the same
>
>server) and MUST NOT use its receipt of the presence probe from a
>
>connected client as the sole cause for returning a stanza or stream
>
>error to the client.
>
>
>
> Sending such probes seems to me to be a great way of finding connection
> problems in Internet of Things networks. Have to do some experiments to
> find out how it works out. Perhaps the explanation for why this should not
> be done by clients can be revised in the RFC?
>
>
>
> Best regards,
>
> Peter Waher
>
>
>
> *From:* Dave Cridland [mailto:d...@cridland.net]
> *Sent:* den 2 mars 2014 16:21
> *To:* XMPP Standards
> *Subject:* Re: [Standards] Last connection time of friend
>
>
>
> Some servers also respond to a probe with the last offline presence (with
> a timestamp).
>
>
>
> On 2 March 2014 19:17, Philipp Hancke  wrote:
>
> Am 02.03.2014 20:08, schrieb Peter Waher:
>
>
>
> Is there a means to figure out the last time a friend connected to its
> XMPP server? Can you ask the server hosting the user account for last
> connection time? (You might not have been online to receive the presence
> message when the user went offline.)
>
>
>
> http://xmpp.org/extensions/xep-0012.html#offline -- not sure how many
> servers implement it though.
>
>
>

Re: [Standards] Last connection time of friend

[Peter Waher]

Hello Dave

Thanks for the quick response. This looks like exactly what I need.

I'm a little worried about this statement in RFC 6121 though (and the 
explanation a little circular):

   Presence probes SHOULD NOT be sent by a client, because in general a
   client will not need to send them since the task of gathering
   presence from a user's contacts is managed by the user's server.
   However, if a user's client generates an outbound presence probe then
   the user's server SHOULD route the probe (if the contact is at
   another server) or process the probe (if the contact is at the same
   server) and MUST NOT use its receipt of the presence probe from a
   connected client as the sole cause for returning a stanza or stream
   error to the client.

Sending such probes seems to me to be a great way of finding connection 
problems in Internet of Things networks. Have to do some experiments to find 
out how it works out. Perhaps the explanation for why this should not be done 
by clients can be revised in the RFC?

Best regards,
Peter Waher

From: Dave Cridland [mailto:d...@cridland.net]
Sent: den 2 mars 2014 16:21
To: XMPP Standards
Subject: Re: [Standards] Last connection time of friend

Some servers also respond to a probe with the last offline presence (with a 
timestamp).

On 2 March 2014 19:17, Philipp Hancke 
mailto:fi...@goodadvice.pages.de>> wrote:
Am 02.03.2014 20:08, schrieb Peter Waher:

Is there a means to figure out the last time a friend connected to its XMPP 
server? Can you ask the server hosting the user account for last connection 
time? (You might not have been online to receive the presence message when the 
user went offline.)

http://xmpp.org/extensions/xep-0012.html#offline -- not sure how many servers 
implement it though.

Re: [Standards] Last connection time of friend

[Peter Waher]

Hello Philipp

Thanks for the quick reply. Should have been able to find this myself, but 
missed it.

Best regards,
Peter Waher


-Original Message-
From: Philipp Hancke [mailto:fi...@goodadvice.pages.de] 
Sent: den 2 mars 2014 16:17
To: standards@xmpp.org
Subject: Re: [Standards] Last connection time of friend

Am 02.03.2014 20:08, schrieb Peter Waher:
> Is there a means to figure out the last time a friend connected to its 
> XMPP server? Can you ask the server hosting the user account for last 
> connection time? (You might not have been online to receive the 
> presence message when the user went offline.)

http://xmpp.org/extensions/xep-0012.html#offline -- not sure how many servers 
implement it though.