From ghudson at MIT.EDU Tue Aug 2 12:06:07 2005 From: ghudson at MIT.EDU (Greg Hudson) Date: Tue, 2 Aug 2005 12:06:07 -0400 Subject: [Wocky] Jabber Kerberos patches Message-ID: <200508021606.j72G67FQ013163@egyptian-gods.mit.edu> MIT is gearing up for a pilot introduction of a Jabber service, and we were highly encouraged to see your July 7 post about Kerberos support in jabberd2 and several Jabber clients (http://mail.jabber.org/pipermail/jabberd/2005-July/002797.html). We were planning to do essentially the same thing ourselves, and this should dramatically reduce the cost of doing so. Can we get ahold of your patches for Gaim? Our current plan is to recommend Gaim on Linux and Solaris, Adium X on MacOS X, and Exodus on Windows, as those appeared to be the best-of-breed Jabber clients on those platforms. It looks like only Gaim is on your hit list, but that's still a good start. We'll also likely be interested in testing server-to-server Kerberos support, although that's less of a priority for us. Thanks for your work on this. It's always great from an MIT perspective to see application support for Kerberos authentication coming from outside of MIT. Greg Hudson Operations and Infrastructure Services Information Services and Technology From raeburn at MIT.EDU Tue Aug 2 12:28:18 2005 From: raeburn at MIT.EDU (Ken Raeburn) Date: Tue, 2 Aug 2005 18:28:18 +0200 Subject: [Wocky] Jabber Kerberos patches In-Reply-To: <200508021606.j72G67FQ013163@egyptian-gods.mit.edu> References: <200508021606.j72G67FQ013163@egyptian-gods.mit.edu> Message-ID: <036A5C47-8965-4F4C-99C4-1C802E6DD84C@mit.edu> On Aug 2, 2005, at 18:06, Greg Hudson wrote: > Can we get ahold of your patches for Gaim? Our current plan is to > recommend Gaim on Linux and Solaris, Adium X on MacOS X, and Exodus on > Windows, as those appeared to be the best-of-breed Jabber clients on > those platforms. It looks like only Gaim is on your hit list, but > that's still a good start. Adium X appears to be based on the Gaim library, so unless they've actually forked, the support should come through eventually, and we might be able to help that along... (and I had a couple of things I wanted to tweak in Adium anyways, hmm...) Ken From simon at sxw.org.uk Tue Aug 2 19:53:31 2005 From: simon at sxw.org.uk (Simon Wilkinson) Date: Wed, 03 Aug 2005 00:53:31 +0100 Subject: [Wocky] Re: Jabber Kerberos patches In-Reply-To: <200508021606.j72G67FQ013163@egyptian-gods.mit.edu> References: <200508021606.j72G67FQ013163@egyptian-gods.mit.edu> Message-ID: <42F0077B.2040904@sxw.org.uk> Greg Hudson wrote: > Can we get ahold of your patches for Gaim? Certainly. I've attached them to this email. They're also in AFS at /afs/inf.ed.ac.uk/user/s/sxw/Development/Jabber/Gaim - although our AFS is in prototype form at the moment ... The 1.3.1 patch also applies to 1.4.0 > Our current plan is to > recommend Gaim on Linux and Solaris, Adium X on MacOS X, and Exodus on > Windows, as those appeared to be the best-of-breed Jabber clients on > those platforms. It looks like only Gaim is on your hit list, but > that's still a good start. I'll be attacking Adium just as soon as my new Mac arrives, hopefully within the next week or so. My hope is that as Adium is built on libgaim, it should be fairly simple to port the Gaim patches over. I took a look at Exodus, but its written in Delphi, which I don't have immediate access to a development environment for. > We'll also likely be interested in testing server-to-server Kerberos > support, although that's less of a priority for us. Yes - Our (Informatics at Edinburgh University) interest in this is a fairly low priority item, too. > Thanks for your work on this. It's always great from an MIT > perspective to see application support for Kerberos authentication > coming from outside of MIT. That's OK. I've had a remit for a few years now to look at better integrating applications with our Kerberos infrastructure - which has led to the Kerberos support in OpenSSH, the Jabber code, and (shortly) GSSAPI support in Thunderbird. It's good to see other people interested in the code. Please keep me posted with how you're getting on. Cheers, Simon. -------------- next part -------------- An embedded and charset-unspecified text was scrubbed... Name: gaim-1.3.1-cyrus.patch Url: http://mailman.mit.edu/pipermail/wocky/attachments/20050803/3ced6650/attachment.bat From jmhunt at MIT.EDU Thu Aug 4 15:26:46 2005 From: jmhunt at MIT.EDU (Jonathan McIndoe Hunt) Date: Thu, 04 Aug 2005 15:26:46 -0400 Subject: [Wocky] GAIM - what is the dependence on GTK Message-ID: <6.2.5.1.2.20050804152531.03f386d8@mit.edu> I installed the standard GAIM for Windows and noted the GTK package also gets installed as an option. What is GTK and should we be including it in the MIT installer? Thanks, Jon From ghudson at MIT.EDU Thu Aug 4 15:28:31 2005 From: ghudson at MIT.EDU (Greg Hudson) Date: Thu, 04 Aug 2005 15:28:31 -0400 Subject: [Wocky] GAIM - what is the dependence on GTK In-Reply-To: <6.2.5.1.2.20050804152531.03f386d8@mit.edu> References: <6.2.5.1.2.20050804152531.03f386d8@mit.edu> Message-ID: <1123183711.22275.6.camel@equal-rites.mit.edu> On Thu, 2005-08-04 at 15:26 -0400, Jonathan McIndoe Hunt wrote: > I installed the standard GAIM for Windows and noted the GTK package > also gets installed as an option. What is GTK and should we be > including it in the MIT installer? GTK is a graphics toolkit. I can't imagine that gaim could work without it. From jdreed at MIT.EDU Thu Aug 4 15:33:48 2005 From: jdreed at MIT.EDU (Jonathan D Reed) Date: Thu, 4 Aug 2005 15:33:48 -0400 (EDT) Subject: [Wocky] GAIM - what is the dependence on GTK In-Reply-To: <1123183711.22275.6.camel@equal-rites.mit.edu> References: <6.2.5.1.2.20050804152531.03f386d8@mit.edu> <1123183711.22275.6.camel@equal-rites.mit.edu> Message-ID: I'm not 100% clear on the context, but in addition to what Greg said, if we're writing our own installer, it should do something sane with GTK. It's possible people might already have GTK installed for other Windows ports of GTK apps, such as Ethereal, so it should only force the user to install it if it's not already there, it shouldn't overwrite anything. -Jon On Thu, 4 Aug 2005, Greg Hudson wrote: > On Thu, 2005-08-04 at 15:26 -0400, Jonathan McIndoe Hunt wrote: >> I installed the standard GAIM for Windows and noted the GTK package >> also gets installed as an option. What is GTK and should we be >> including it in the MIT installer? > > GTK is a graphics toolkit. I can't imagine that gaim could work without > it. > > _______________________________________________ > Wocky mailing list > Wocky at mit.edu > http://mailman.mit.edu/mailman/listinfo/wocky > From jmhunt at MIT.EDU Thu Aug 4 15:50:47 2005 From: jmhunt at MIT.EDU (Jonathan McIndoe Hunt) Date: Thu, 04 Aug 2005 15:50:47 -0400 Subject: [Wocky] GAIM - what is the dependence on GTK In-Reply-To: References: <6.2.5.1.2.20050804152531.03f386d8@mit.edu> <1123183711.22275.6.camel@equal-rites.mit.edu> Message-ID: <6.2.5.1.2.20050804154948.06dde820@mit.edu> And the simple installer for Windows is already starting to add complexity. :-) I'm sure Atticus can handle this extra complexity. Thanks, Jon At 03:33 PM 8/4/2005, Jonathan D Reed wrote: >I'm not 100% clear on the context, but in addition to what Greg >said, if we're writing our own installer, it should do something >sane with GTK. It's possible people might already have GTK installed >for other Windows ports of GTK apps, such as Ethereal, so it should >only force the user to install it if it's not already there, it >shouldn't overwrite anything. > >-Jon > > >On Thu, 4 Aug 2005, Greg Hudson wrote: > >>On Thu, 2005-08-04 at 15:26 -0400, Jonathan McIndoe Hunt wrote: >>>I installed the standard GAIM for Windows and noted the GTK package >>>also gets installed as an option. What is GTK and should we be >>>including it in the MIT installer? >> >>GTK is a graphics toolkit. I can't imagine that gaim could work without >>it. >> >>_______________________________________________ >>Wocky mailing list >>Wocky at mit.edu >>http://mailman.mit.edu/mailman/listinfo/wocky From simon at sxw.org.uk Sat Aug 6 13:25:08 2005 From: simon at sxw.org.uk (Simon Wilkinson) Date: Sat, 06 Aug 2005 18:25:08 +0100 Subject: [Wocky] Re: Jabber Kerberos patches In-Reply-To: <200508021606.j72G67FQ013163@egyptian-gods.mit.edu> References: <200508021606.j72G67FQ013163@egyptian-gods.mit.edu> Message-ID: <42F4F274.706@sxw.org.uk> Hi again, I should have said in my earlier email that, once applying the patch to Gaim, you need to build with --enable-cyrus-sasl Cheers, Simon From ghudson at MIT.EDU Tue Aug 9 15:41:59 2005 From: ghudson at MIT.EDU (Greg Hudson) Date: Tue, 9 Aug 2005 15:41:59 -0400 Subject: [Wocky] gaim in the Athena release Message-ID: <200508091941.j79Jfxwe026259@equal-rites.mit.edu> I have finished putting gaim into the Athena 9.4 release (for alpha and beta machines currently; it will follow for early machines in a few days). I settled on version 1.4.0, which is the most recent version at the moment. Since users running gaim will be nudged in the direction of a Jabber service which has only pilot support, I think we need a short web page which describes the current level of support and who to contact (i.e. us) with feedback. We can point people to this page via the Athena release notes, the Athena consultants, and the SWRT software release page when we have an installer ready. Also, I notice that http://web.mit.edu/ist/start/communicating/index.html refers to Zephyr, and should probably refer to this new web page as well. I envision this page as becoming analagous to http://web.mit.edu/ist/topics/email/index.html, so perhaps it should live in the analagous place: ist/topics/jabber/index.html. Does that sound reasonable to people? If I draft some initial text for the page, can Heather edit and format it and put it up in the appropriate place? From aurora at MIT.EDU Tue Aug 9 16:17:54 2005 From: aurora at MIT.EDU (Heather Anne Harrison) Date: Tue, 9 Aug 2005 16:17:54 -0400 Subject: [Wocky] gaim in the Athena release In-Reply-To: <200508091941.j79Jfxwe026259@equal-rites.mit.edu> References: <200508091941.j79Jfxwe026259@equal-rites.mit.edu> Message-ID: <20050809161754.jcf6cmrxps4k8wgs@webmail.mit.edu> I think I own all the pages you've referenced. updates will be no problem. However, I think a service style page might be better than a topic page. Topic pages are extremely limited in the amount of text we can put on them. They're also supposed to be for broad topcis that cover a range of products/services. Here's an example of a service page: http://web.mit.edu/ist/topics/linux/rhn.html As for where it would live, I suspect something like /mit/ist/services/jabber/ might make sense. Heather Anne aurora at mit.edu Quoting Greg Hudson : > I have finished putting gaim into the Athena 9.4 release (for alpha > and beta machines currently; it will follow for early machines in a > few days). I settled on version 1.4.0, which is the most recent > version at the moment. > > Since users running gaim will be nudged in the direction of a Jabber > service which has only pilot support, I think we need a short web page > which describes the current level of support and who to contact > (i.e. us) with feedback. We can point people to this page via the > Athena release notes, the Athena consultants, and the SWRT software > release page when we have an installer ready. Also, I notice that > http://web.mit.edu/ist/start/communicating/index.html refers to > Zephyr, and should probably refer to this new web page as well. > > I envision this page as becoming analagous to > http://web.mit.edu/ist/topics/email/index.html, so perhaps it should > live in the analagous place: ist/topics/jabber/index.html. > > Does that sound reasonable to people? If I draft some initial text > for the page, can Heather edit and format it and put it up in the > appropriate place? > _______________________________________________ > Wocky mailing list > Wocky at mit.edu > http://mailman.mit.edu/mailman/listinfo/wocky > From jmhunt at MIT.EDU Tue Aug 9 16:48:26 2005 From: jmhunt at MIT.EDU (Jonathan McIndoe Hunt) Date: Tue, 09 Aug 2005 16:48:26 -0400 Subject: [Wocky] configuration settings for jabber clients Message-ID: <6.2.5.1.2.20050809164747.04044868@mit.edu> Do we have the settings, like name and port number to configure GAIM and test iChat? Thanks, Jon From ghudson at MIT.EDU Tue Aug 9 17:41:07 2005 From: ghudson at MIT.EDU (Greg Hudson) Date: Tue, 09 Aug 2005 17:41:07 -0400 Subject: [Wocky] configuration settings for jabber clients In-Reply-To: <6.2.5.1.2.20050809164747.04044868@mit.edu> References: <6.2.5.1.2.20050809164747.04044868@mit.edu> Message-ID: <1123623667.3458.243.camel@egyptian-gods.mit.edu> On Tue, 2005-08-09 at 16:48 -0400, Jonathan McIndoe Hunt wrote: > Do we have the settings, like name and port number to configure GAIM > and test iChat? (Looks like they spell it Gaim, not GAIM.) The server name is jabber.mit.edu. The port number is 5222, which is the standard Jabber port. The domain part of the Jabber IDs is mit.edu. The way this looks in gaim is that the "Server" is mit.edu and the "Connect Server" (under "Show More Options") is jabber.mit.edu. A client with SRV lookup support can just use the mit.edu SRV record to find the server. Gaim doesn't have this support; I don't know if iChat and Adium X do. From pbh at MIT.EDU Tue Aug 9 18:11:57 2005 From: pbh at MIT.EDU (Paul B. Hill) Date: Tue, 9 Aug 2005 18:11:57 -0400 Subject: [Wocky] configuration settings for jabber clients In-Reply-To: <1123623667.3458.243.camel@egyptian-gods.mit.edu> Message-ID: <200508092212.j79MCBIb013259@outgoing.mit.edu> Thanks Greg. On Windows I am able to use Exodus, using this configuration information. BTW, Mark mentioned that if someone has to use "force old SSL" then port 5223 should be used by that client. I am having a problem when using Gaim on Windows. The authentication hangs. On Exodus, I saw the dialog box that complained that the certificate of the server didn't match the name of the host. I told it to always use that certificate anyways. I am not seeing a similar dialog box on Windows. I wonder if it is waiting for me to confirm the certificate, but doesn't have a way for me to confirm it? Was there anything that you had to do with the configuration on Athena to deal with the certificate of the server? Paul -----Original Message----- From: wocky-bounces at MIT.EDU [mailto:wocky-bounces at MIT.EDU] On Behalf Of Greg Hudson Sent: Tuesday, August 09, 2005 5:41 PM To: Jonathan McIndoe Hunt Cc: wocky at mit.edu Subject: Re: [Wocky] configuration settings for jabber clients On Tue, 2005-08-09 at 16:48 -0400, Jonathan McIndoe Hunt wrote: > Do we have the settings, like name and port number to configure GAIM > and test iChat? (Looks like they spell it Gaim, not GAIM.) The server name is jabber.mit.edu. The port number is 5222, which is the standard Jabber port. The domain part of the Jabber IDs is mit.edu. The way this looks in gaim is that the "Server" is mit.edu and the "Connect Server" (under "Show More Options") is jabber.mit.edu. A client with SRV lookup support can just use the mit.edu SRV record to find the server. Gaim doesn't have this support; I don't know if iChat and Adium X do. _______________________________________________ Wocky mailing list Wocky at mit.edu http://mailman.mit.edu/mailman/listinfo/wocky From pbh at MIT.EDU Tue Aug 9 18:34:56 2005 From: pbh at MIT.EDU (Paul B. Hill) Date: Tue, 9 Aug 2005 18:34:56 -0400 Subject: [Wocky] configuration settings for jabber clients In-Reply-To: <200508092212.j79MCBIb013259@outgoing.mit.edu> Message-ID: <200508092235.j79MZDH8017221@outgoing.mit.edu> I figured out what was causing the hang during authentication when using Gaim on Windows. When setting up the account profile I was using the "register" button as a way of saving and authenticating. When I used the save button, and then chose "login" the authentication completed. Paul -----Original Message----- From: wocky-bounces at MIT.EDU [mailto:wocky-bounces at MIT.EDU] On Behalf Of Paul B. Hill Sent: Tuesday, August 09, 2005 6:12 PM To: 'Greg Hudson' Cc: wocky at mit.edu Subject: RE: [Wocky] configuration settings for jabber clients Thanks Greg. On Windows I am able to use Exodus, using this configuration information. BTW, Mark mentioned that if someone has to use "force old SSL" then port 5223 should be used by that client. I am having a problem when using Gaim on Windows. The authentication hangs. On Exodus, I saw the dialog box that complained that the certificate of the server didn't match the name of the host. I told it to always use that certificate anyways. I am not seeing a similar dialog box on Windows. I wonder if it is waiting for me to confirm the certificate, but doesn't have a way for me to confirm it? Was there anything that you had to do with the configuration on Athena to deal with the certificate of the server? Paul -----Original Message----- From: wocky-bounces at MIT.EDU [mailto:wocky-bounces at MIT.EDU] On Behalf Of Greg Hudson Sent: Tuesday, August 09, 2005 5:41 PM To: Jonathan McIndoe Hunt Cc: wocky at mit.edu Subject: Re: [Wocky] configuration settings for jabber clients On Tue, 2005-08-09 at 16:48 -0400, Jonathan McIndoe Hunt wrote: > Do we have the settings, like name and port number to configure GAIM > and test iChat? (Looks like they spell it Gaim, not GAIM.) The server name is jabber.mit.edu. The port number is 5222, which is the standard Jabber port. The domain part of the Jabber IDs is mit.edu. The way this looks in gaim is that the "Server" is mit.edu and the "Connect Server" (under "Show More Options") is jabber.mit.edu. A client with SRV lookup support can just use the mit.edu SRV record to find the server. Gaim doesn't have this support; I don't know if iChat and Adium X do. _______________________________________________ Wocky mailing list Wocky at mit.edu http://mailman.mit.edu/mailman/listinfo/wocky _______________________________________________ Wocky mailing list Wocky at mit.edu http://mailman.mit.edu/mailman/listinfo/wocky From pbh at MIT.EDU Tue Aug 9 18:50:18 2005 From: pbh at MIT.EDU (Paul B. Hill) Date: Tue, 9 Aug 2005 18:50:18 -0400 Subject: [Wocky] gaim in the Athena release In-Reply-To: <20050809161754.jcf6cmrxps4k8wgs@webmail.mit.edu> Message-ID: <200508092250.j79MofLc019743@outgoing.mit.edu> Hi Heather, I agree that a service style page makes sense. BTW, I note that two existing IS&T pages already make a passing reference to the Gaim client and jabber: http://web.mit.edu/answers/zephyr/zephyr_non_mit.html http://web.mit.edu/acs/www/communication.html These should be up dated as well. Paul -----Original Message----- From: wocky-bounces at MIT.EDU [mailto:wocky-bounces at MIT.EDU] On Behalf Of Heather Anne Harrison Sent: Tuesday, August 09, 2005 4:18 PM To: Greg Hudson Cc: wocky at mit.edu Subject: Re: [Wocky] gaim in the Athena release I think I own all the pages you've referenced. updates will be no problem. However, I think a service style page might be better than a topic page. Topic pages are extremely limited in the amount of text we can put on them. They're also supposed to be for broad topcis that cover a range of products/services. Here's an example of a service page: http://web.mit.edu/ist/topics/linux/rhn.html As for where it would live, I suspect something like /mit/ist/services/jabber/ might make sense. Heather Anne aurora at mit.edu Quoting Greg Hudson : > I have finished putting gaim into the Athena 9.4 release (for alpha > and beta machines currently; it will follow for early machines in a > few days). I settled on version 1.4.0, which is the most recent > version at the moment. > > Since users running gaim will be nudged in the direction of a Jabber > service which has only pilot support, I think we need a short web page > which describes the current level of support and who to contact > (i.e. us) with feedback. We can point people to this page via the > Athena release notes, the Athena consultants, and the SWRT software > release page when we have an installer ready. Also, I notice that > http://web.mit.edu/ist/start/communicating/index.html refers to > Zephyr, and should probably refer to this new web page as well. > > I envision this page as becoming analagous to > http://web.mit.edu/ist/topics/email/index.html, so perhaps it should > live in the analagous place: ist/topics/jabber/index.html. > > Does that sound reasonable to people? If I draft some initial text > for the page, can Heather edit and format it and put it up in the > appropriate place? > _______________________________________________ > Wocky mailing list > Wocky at mit.edu > http://mailman.mit.edu/mailman/listinfo/wocky > _______________________________________________ Wocky mailing list Wocky at mit.edu http://mailman.mit.edu/mailman/listinfo/wocky From jdreed at MIT.EDU Wed Aug 10 00:13:03 2005 From: jdreed at MIT.EDU (Jonathan Reed) Date: Wed, 10 Aug 2005 00:13:03 -0400 Subject: [Wocky] configuration settings for jabber clients In-Reply-To: <1123623667.3458.243.camel@egyptian-gods.mit.edu> References: <6.2.5.1.2.20050809164747.04044868@mit.edu> <1123623667.3458.243.camel@egyptian-gods.mit.edu> Message-ID: Just for kicks, I tried this in iChat (3.0.1 v 392 that ships with Tiger). iChat does not appear to have SRV record support - looking at the wire, I don't even see it attempting an SRV lookup. Also, interestingly enough, in iChat, if I choose SSL, it sets the port to 5223, however it then insists that I'm connecting to Jabber in an insecure way and that my password is exposed on the network. tcpdump and ethereal disagree with this statement, and admittedly I find it confusing that iChat would say that when "SSL" is checked in the options (and indeed it is using SSL, I can see the certificate being exchanged over the wire). I thought perhaps it was because I did not have the MIT CA in my Keychain, but I added that and it still gives that error. -Jon At 5:41 PM -0400 8/9/05, Greg Hudson wrote: >On Tue, 2005-08-09 at 16:48 -0400, Jonathan McIndoe Hunt wrote: >> Do we have the settings, like name and port number to configure GAIM >> and test iChat? > >(Looks like they spell it Gaim, not GAIM.) > >The server name is jabber.mit.edu. The port number is 5222, which is >the standard Jabber port. > >The domain part of the Jabber IDs is mit.edu. The way this looks in >gaim is that the "Server" is mit.edu and the "Connect Server" (under >"Show More Options") is jabber.mit.edu. A client with SRV lookup >support can just use the mit.edu SRV record to find the server. Gaim >doesn't have this support; I don't know if iChat and Adium X do. > >_______________________________________________ >Wocky mailing list >Wocky at mit.edu >http://mailman.mit.edu/mailman/listinfo/wocky -- ------------------- Jonathan Reed jdreed at mit.edu ------------------- From hallisey at MIT.EDU Wed Aug 10 14:54:26 2005 From: hallisey at MIT.EDU (Joanne M. Hallisey) Date: Wed, 10 Aug 2005 14:54:26 -0400 Subject: [Wocky] Next team meting Message-ID: Hello, Understanding that there are other priorities, vacations, deadlines, can you tell me which of the following dates at 2:00 PM work best for our next update meeting: Tuesday, Aug. 16 Tuesday, Aug. 23 Tuesday, Aug. 30 I will be talking to individuals in the meantime about individual tasks. Some questions to consider: What do we want people to test? Who do we want to test in IS&T (all fo SDIT)? If Gaim is included in the Athena release, do we need to coordinate some of the documentation with that release? Likewise, do wee need to provide information for the Athena consultants (olc, others) for the purposes of this pilot? Please add any other questions we need to be addressing before release. Thanks, Joanne -- Joanne Hallisey Sr. Project Manager MIT - Information Services and Technology 617-253-1894 From ghudson at MIT.EDU Wed Aug 10 15:43:51 2005 From: ghudson at MIT.EDU (Greg Hudson) Date: Wed, 10 Aug 2005 15:43:51 -0400 Subject: [Wocky] configuration settings for jabber clients In-Reply-To: References: <6.2.5.1.2.20050809164747.04044868@mit.edu> <1123623667.3458.243.camel@egyptian-gods.mit.edu> Message-ID: <1123703032.30038.40.camel@egyptian-gods.mit.edu> On Wed, 2005-08-10 at 00:13 -0400, Jonathan Reed wrote: > Just for kicks, I tried this in iChat (3.0.1 v 392 that ships with > Tiger). iChat does not appear to have SRV record support - looking > at the wire, I don't even see it attempting an SRV lookup. Also, > interestingly enough, in iChat, if I choose SSL, it sets the port to > 5223, however it then insists that I'm connecting to Jabber in an > insecure way and that my password is exposed on the network. It might be the server spitting out this error. Gaim has a wonderful debug option which lets you figure these things out; I assume iChat probably does not. From jdreed at MIT.EDU Thu Aug 11 09:59:08 2005 From: jdreed at MIT.EDU (Jonathan Reed) Date: Thu, 11 Aug 2005 09:59:08 -0400 Subject: [Wocky] configuration settings for jabber clients In-Reply-To: <1123703032.30038.40.camel@egyptian-gods.mit.edu> References: <6.2.5.1.2.20050809164747.04044868@mit.edu> <1123623667.3458.243.camel@egyptian-gods.mit.edu> <1123703032.30038.40.camel@egyptian-gods.mit.edu> Message-ID: At 3:43 PM -0400 8/10/05, Greg Hudson wrote: > >It might be the server spitting out this error. Gaim has a wonderful >debug option which lets you figure these things out; I assume iChat >probably does not. That sounds right, actually. At least, I cannot find that message in any of the localizable resources in the application bundle, unless Apple chose not to localize that one error message. Do we know why the server does that? Along those lines, what server are we using, so that I may UTSL? Or is there not a readable copy of the code we're using (including any local patches)? -Jon -- ------------------- Jonathan Reed jdreed at mit.edu ------------------- From ghudson at MIT.EDU Thu Aug 11 10:38:07 2005 From: ghudson at MIT.EDU (Greg Hudson) Date: Thu, 11 Aug 2005 10:38:07 -0400 Subject: [Wocky] configuration settings for jabber clients In-Reply-To: References: <6.2.5.1.2.20050809164747.04044868@mit.edu> <1123623667.3458.243.camel@egyptian-gods.mit.edu> <1123703032.30038.40.camel@egyptian-gods.mit.edu> Message-ID: <1123771087.3447.4.camel@egyptian-gods.mit.edu> On Thu, 2005-08-11 at 09:59 -0400, Jonathan Reed wrote: > Do we know why the server does that? Along those lines, what server > are we using, so that I may UTSL? Or is there not a readable copy of > the code we're using (including any local patches)? Well, it would suggest a bug or a misconfiguration. I'm still owed a pointer to the server source and local patches so I can better look into this sort of thing (and set up a repository, and start integrating SASL support). Ideally, though, iChat would be properly negotiating SSL over port 5222, not using old-style SSL over port 5223. It doesn't sound like it has support for that, though. From jdreed at MIT.EDU Thu Aug 11 11:55:46 2005 From: jdreed at MIT.EDU (Jonathan Reed) Date: Thu, 11 Aug 2005 11:55:46 -0400 Subject: [Wocky] Buddy Authorization Message-ID: I understand that one of the big selling points of jabber is "individual privacy" and that they want it to be like meeting in real life, so both people are aware of the interaction, but I find it rather frustrating that each time I add someone to my buddy list (or I get added to theirs) I have to deal with the authorization, or 'handshake' as the jabber docs call it. Is there a setting on the server for this? Or is the idea that I should only use my 'buddy list' for people whom I want to authorize (and monitor their status), and if I wish to chat with someone not in my buddy list, I should simply open a new chat directly with that person and bypass the list entirely? This seems like it might be difficult to migrate from the AIM mindset (or even the xzewd mindset, if you will), where your buddy list is more like an address book so you can quickly start an IM with someone. In the Jabber world, it seems like it's more of a trust relationship. I understand the concerns obviously (online stalking by knowning how long someone is idle, etc), but in the Zephyr world, zlocate and finger provide that information pseudo-anonymously, perhaps our Jabber server should be configured the same way? -Jon -- ------------------- Jonathan Reed jdreed at mit.edu ------------------- From pbh at MIT.EDU Thu Aug 11 13:59:24 2005 From: pbh at MIT.EDU (Paul B. Hill) Date: Thu, 11 Aug 2005 13:59:24 -0400 Subject: [Wocky] Buddy Authorization In-Reply-To: Message-ID: <200508111759.j7BHxVj6012885@outgoing.mit.edu> Hi, Jabber's "buddy" feature is closely tied to presence, hence has some privacy concerns and requires a level of authorization. However, Jabber also has the concept of a "roster" which provides some of the functionality that people may be used to from AIM's buddies and some of the Zephyr features for discovering people. Below is some information that I found online. It is primarily aimed at developers, but has some useful background information for people on this list. 5. Managing Your Roster Roster management is one of the most confusing areas of the protocol, or at least of its current documentation. Here's an overview. How the roster works. The roster is the list of Jabber entities with whom the user has a presence relationship. This includes people to whose status the user is subscribed ("buddies"), people who are subscribed to the user's status ("watchers"), and people to whom the user wants to subscribe but who haven't yet approved the request. The roster is thus the model of the GUI "buddy list". The Jabber server stores the user's roster and notifies all logged-in clients whenever it changes: when the user adds or removes a buddy, when another user adds or removes the user as a buddy, and when a potential buddy approves or rejects a subscription request. This is known as a roster push. The change notifications are sent as elements using the jabber:iq:roster namespace. The client can also manually request a complete copy of the roster: it's necessary to do this just after logging in to detect changes that may have been made since the last logout at this computer. [JPO 1.6.12] Subscribing & unsubscribing buddies. Confusingly enough, adding and removing buddies is not done via roster manipulation but via elements. Here the type is subscribe or unsubscribe and the request is sent to the buddy's address. Confirmation or rejection is received as a reply element (with the same ID) with a type of subscribed or unsubscribed. Conversely, you'll also deal with other entities wanting to subscribe to your presence. In this case it's exactly the opposite: you receive the subscribe or unsubscribe and (after consulting with the user) reply with subscribed or unsubscribed. [JPO 1.4.1.3 -- 1.4.1.7] In all these cases, since your roster is changed, the server will also send another roster push to tell you about the roster element to change. Manually updating the roster. It's possible for you to update the server-side roster by sending elements. You would do this not to add or remove buddies (see the previous section for how to do that) but to add or change metadata associated with a buddy, such as their nickname or group[s]. [JPO 1.6.12] Actually, it is possible to add people to the roster this way, but it won't subscribe you to their presence. They'll just be there passively, as in an e-mail address book. You may or may not want to support this in your UI; it doesn't seem especially useful to me, but it might be handy if there are people the user wants to send messages to but doesn't need to know their online status. The documentation says it's also possible to remove people directly from the roster by setting the "subscription" attribute to "remove". [JPG p.65] This is the only way to remove an item added without a subscription (as in the previous paragraph); but I'm not sure what happens if you remove a subscribed buddy this way. Presumably it should unsubscribe you from their presence notifications just as if you had sent a element. Implementation tip. Since the server will notify you of all changes in the roster, you can drive the management of the GUI buddy list entirely from these pushes. For example, when the user wants to add a buddy, you send out a element but don't update the buddy list yet; you'll immediately receive a roster push from the server telling you about the change and should update the buddy list when you receive it. By using roster pushes to manage the buddy list, you ensure that the buddy list stays up to date whether it's you or another simultaneously logged-in client making the changes. More roster info. You may want to display more buddy information in your UI than simply the Jabber ID, presence state and status message. As described above, a full name/nickname can be associated with a buddy in the roster. A whole lot more information about a buddy can be obtained by retrieving their vCard via an query, if they've stored one on the server [JPO 1.6.26]. vCards are the Internet equivalent of business or Rolodex cards and can contain any imaginable sort of personal and contact information such as photos, birthdays, phone numbers... Still more generally, the Jabber server as of version 1.4 allows users to store arbitrary XML data tagged with any namespace, accessible to anyone who queries their account using the same namespace. [JPO 1.6.10] I'm not aware of any concrete usage of this, but it has great potential. The preceding information comes from . From nemo at MIT.EDU Thu Aug 11 14:42:20 2005 From: nemo at MIT.EDU (Joel N. Weber II) Date: Thu, 11 Aug 2005 14:42:20 -0400 Subject: [Wocky] Buddy Authorization In-Reply-To: (message from Jonathan Reed on Thu, 11 Aug 2005 11:55:46 -0400) References: Message-ID: I think a more general question is whether the goal is for an Athena Jabber server to provide the same semantics as zephyr currently does, or whether jabber is supposed to provide support for some other culture. I get the impression that there are many zephyr users who would be unhappy if jabber were to replace zephyr without providing a similar user experience. But I'm not clear on what problem the Athena jabber server is supposed to solve. (On the other hand, if it's not supposed to be a zephyr replacement, what does it offer that AIM doesn't?) How does jabber's current buddy authorization interact with jabber conferences? Occasionally I participate in conversations on zephyr classes where it ends up being useful to move to personals, and on rare occasions this involves someone who I wouldn't happen to have already authorized as a buddy, and so the semantics of letting anyone send are useful. Every now and then we see a webzephyr user carrying out an extensive conversation on a zephyr class trying to get someone to subscribe to webzephyr so that they can send a personal message, which is somewhat annoying for everyone else who's watching the class. I think it's also the case that jabber was designed with the idea that anyone anywhere on the Internet would normally be able to connect, without any strong mechanism for authenticating the users. In such a context, buddy authorization probably makes a _lot_ of sense. If you have a system that discourages annoymous messages from being sent through relatively strong authentication of users, needing to authorize each user you individually talk to probably becomes much less of a problem. From hallisey at MIT.EDU Thu Aug 11 15:15:54 2005 From: hallisey at MIT.EDU (Joanne M. Hallisey) Date: Thu, 11 Aug 2005 15:15:54 -0400 Subject: [Wocky] Client Choices Message-ID: From what we said before, we are testing the following for the pilot: Adium X for the Mac Gaim for Windows and Linux/Unix Do we want people with Tiger to test iChat? Joanne -- Joanne Hallisey Sr. Project Manager MIT - Information Services and Technology 617-253-1894 From ghudson at MIT.EDU Thu Aug 11 15:26:28 2005 From: ghudson at MIT.EDU (Greg Hudson) Date: Thu, 11 Aug 2005 15:26:28 -0400 Subject: [Wocky] Client Choices In-Reply-To: References: Message-ID: <1123788388.3447.29.camel@egyptian-gods.mit.edu> On Thu, 2005-08-11 at 15:15 -0400, Joanne M. Hallisey wrote: > From what we said before, we are testing the following for the pilot: > Adium X for the Mac > Gaim for Windows and Linux/Unix Correct. > Do we want people with Tiger to test iChat? Yes. Although the initial results don't sound too promising; it sounds like it doesn't support the preferred kind of SSL connection, and the kind it does support doesn't work right with our server (although we may be able to fix that). From ghudson at MIT.EDU Fri Aug 12 02:20:31 2005 From: ghudson at MIT.EDU (Greg Hudson) Date: Fri, 12 Aug 2005 02:20:31 -0400 Subject: [Wocky] [Fwd: [Gaim-devel] Gaim 1.5.0 has been released] Message-ID: <1123827632.3447.32.camel@egyptian-gods.mit.edu> According to this announcement, Gaim 1.5.0 contains a plethora of security fixes, so I plan to upgrade Athena to this version, and I suggest SWRT use it as well. -------------- next part -------------- An embedded message was scrubbed... From: "Mark Doliner" Subject: [Gaim-devel] Gaim 1.5.0 has been released Date: Thu, 11 Aug 2005 23:06:24 -0500 Size: 3372 Url: http://mailman.mit.edu/pipermail/wocky/attachments/20050812/88a28570/attachment.eml From pbh at MIT.EDU Fri Aug 12 14:41:13 2005 From: pbh at MIT.EDU (Paul B. Hill) Date: Fri, 12 Aug 2005 14:41:13 -0400 Subject: [Wocky] Buddy Authorization In-Reply-To: Message-ID: <200508121841.j7CIfLPJ022085@outgoing.mit.edu> Hi, Thank you for your comments and questions. I believe we are approaching jabber with an open mind, and for several different purposes. First there are several communities of users at MIT that have never broadly adopted Zephyr. Many people do use AIM, but as more work is done over IM, relying on AIM, Yahoo, or MSN becomes problematic. By providing an institutional IM service we hope that people will appreciate the consistent namespace and find it of value. We are also somewhat uncomfortable with some people using these 3rd party commercial services to carry out institutional business. The jabber protocol appears to offer support of most of the Zephyr features that MIT users feel are important. We realize that today, few, if any, of the clients support many of these features. This may change overtime, either through the efforts of people at MIT or elsewhere. I feel there are reasons to explore alternatives Zephyr. Many people at MIT need to collaborate with others around the world. Although Zephyr is used at some other sites, it has never had the wide spread deployment that some other IM systems enjoy. Furthermore NATs are a reality, and Zephyr does not work well when using address translation. As GSS support is added to jabber client and servers it remains to be seen how developers will deal with the authorization issues. Although we expect to use GSS within the campus community, which might argue against the current authorization model, we also want the jabber service to provide an IM foundation for extra-institutional IM. Not all of those other sites will necessarily being using authentication methods that we find acceptable. The UIs will have to evolve over time to address these different needs. At this time we are not talking about replacing Zephyr with jabber. We are trying to create an environment where jabber can compete with Zephyr and determine what needs to be changed to best meet our customer's needs. Paul -----Original Message----- From: wocky-bounces at MIT.EDU [mailto:wocky-bounces at MIT.EDU] On Behalf Of Joel N. Weber II Sent: Thursday, August 11, 2005 2:42 PM To: Jonathan Reed Cc: wocky at mit.edu Subject: Re: [Wocky] Buddy Authorization I think a more general question is whether the goal is for an Athena Jabber server to provide the same semantics as zephyr currently does, or whether jabber is supposed to provide support for some other culture. I get the impression that there are many zephyr users who would be unhappy if jabber were to replace zephyr without providing a similar user experience. But I'm not clear on what problem the Athena jabber server is supposed to solve. (On the other hand, if it's not supposed to be a zephyr replacement, what does it offer that AIM doesn't?) How does jabber's current buddy authorization interact with jabber conferences? Occasionally I participate in conversations on zephyr classes where it ends up being useful to move to personals, and on rare occasions this involves someone who I wouldn't happen to have already authorized as a buddy, and so the semantics of letting anyone send are useful. Every now and then we see a webzephyr user carrying out an extensive conversation on a zephyr class trying to get someone to subscribe to webzephyr so that they can send a personal message, which is somewhat annoying for everyone else who's watching the class. I think it's also the case that jabber was designed with the idea that anyone anywhere on the Internet would normally be able to connect, without any strong mechanism for authenticating the users. In such a context, buddy authorization probably makes a _lot_ of sense. If you have a system that discourages annoymous messages from being sent through relatively strong authentication of users, needing to authorize each user you individually talk to probably becomes much less of a problem. _______________________________________________ Wocky mailing list Wocky at mit.edu http://mailman.mit.edu/mailman/listinfo/wocky From hallisey at MIT.EDU Fri Aug 12 15:02:46 2005 From: hallisey at MIT.EDU (Joanne M. Hallisey) Date: Fri, 12 Aug 2005 15:02:46 -0400 Subject: [Wocky] Adium Config note Message-ID: In setting up AdiumX on a Mac running 10.3.9, the username needs to be juser at mit.edu or it will default to @jabber .org. Joanne -- Joanne Hallisey Sr. Project Manager MIT - Information Services and Technology 617-253-1894 From hallisey at MIT.EDU Tue Aug 16 11:07:02 2005 From: hallisey at MIT.EDU (Joanne M. Hallisey) Date: Tue, 16 Aug 2005 11:07:02 -0400 Subject: [Wocky] Next team meting Message-ID: Hello, Ok, Greg is on vacation until Aug. 30, then I hope to be away on Sept. 6. So, if possible, can we meet on Thursday afternoon, Sept. 1 at 3:00 in W92 ( I will reserve a room when the date is set. If you are not sure if you are on the team, send me an email. If you are doing any tasks for the project, please consider yourself on the team and if you just want to attend the meeting for information, please feel free to come. Some questions to consider: What do we want people to test? Who do we want to test in IS&T (all fo SDIT)? If Gaim is included in the Athena release, do we need to coordinate some of the documentation with that release? Likewise, do wee need to provide information for the Athena consultants (olc, others) for the purposes of this pilot? Please add any other questions we need to be addressing before release. Thanks, Joanne -- Joanne Hallisey Sr. Project Manager MIT - Information Services and Technology 617-253-1894 _______________________________________________ Wocky mailing list Wocky at mit.edu http://mailman.mit.edu/mailman/listinfo/wocky -- Joanne Hallisey Sr. Project Manager MIT - Information Services and Technology 617-253-1894 From hallisey at MIT.EDU Thu Aug 25 08:31:36 2005 From: hallisey at MIT.EDU (Joanne M. Hallisey) Date: Thu, 25 Aug 2005 08:31:36 -0400 Subject: [Wocky] FYI: NYTimes Article on Google's chat Message-ID: Hello, Just passing along an article in today's NY Times Technology section. TECHNOLOGY / CIRCUITS | August 25, 2005 David Pogue: Google Gets Better. What's Up With That? EVER heard the old joke about the two psychiatrists who pass in a hallway? One says, "Hello there." The other thinks, "I wonder what he meant by that?" An improved Google Desktop Search, including Gmail, news, photos and Web clips. Both free Google tools are for Windows. In high-tech circles, that's pretty much what people are saying about Google these days. If you hadn't noticed, Google is no longer just an Internet search tool; it's now a full-blown software company. It develops elegant, efficient software programs - and then gives them away. In today's culture of cynicism, such generosity and software excellence seems highly suspicious; surely it's all a smokescreen for a darker, larger plot to suck us all in. What, exactly, is Google up to? The mystery only intensified this week, as Google announced two more free software tools for Windows: a new version of Google Desktop Search and a free instant-messaging program called Google Talk. The original version of Desktop Search, which Google unleashed last fall, brought the speed and effortlessness of Google's Internet search to your own PC. You'd type a few letters, and in a fraction of a second, you'd be looking at a complete list of files that included your search term - even if that term appeared inside the body of a document. It could even search e-mail, chat-session transcripts and the contents of Web pages you'd seen. Google Desktop 1.0 certainly blew away Windows' own built-in search tool, which operates with all the speed of an anesthetized slug. But it was limited in three ways. First, you had to operate it from within your Web browser, limiting its convenience. Second, because it could call up Web pages, e-mail messages and chat transcripts, Google Desktop alarmed people who, ahem, had something to hide from bosses or spouses. And finally, it could see inside only a limited number of document types. For example, it couldn't search PDF files, Web sites visited with any browser except Internet Explorer, or e-mail messages except those in Outlook or Outlook Express. VERSION 2 , now available at google.com in what's technically in a public beta test version, attacks all of these drawbacks with a vengeance. In Version 2, you can begin a search with a keystroke or by clicking in the search box that's always on the screen. A pop-up menu of search results appears as you begin to type and narrows itself with each additional keystroke. When you see the item you want, you can open it by clicking or by walking up the list with the arrow keys and pressing Enter. In other words, you can now find and open a certain program, document or control panel entirely from the keyboard, with blazing speed and simplicity. This is old news to Mac fans, of course; the Spotlight feature in Mac OS X 10.4 works the same way. But for Windows XP and 2000 veterans, getting such an omniscient, speedy search feature free is truly liberating. ( Microsoft plans something similar for the next version of Windows, due at the end of 2006.) Google has also beefed up your privacy options. You can omit search categories like secure Web sites (banking sites, for example), password-protected Microsoft Office files, and so on, and you can even flag individual files so that they'll never appear in the search results again. Finally, the program now recognizes many more document types: e-mail from Gmail, Outlook, Outlook Express, Netscape Mail, Thunderbird and Mozilla Mail; chat transcripts from AOL or MSN Messenger; Web pages you've visited using Internet Explorer, Firefox, Netscape or Mozilla; PDF files; and your Outlook calendar and address book. (And speaking of Outlook, Google Desktop now installs its own search bar right into Outlook, meaning that you can search your e-mail collection in the blink of a cursor.) The company expects to add more kinds of files to this list, thanks to a public plug-in protocol it has published online. Yet believe it or not, the little search box is the last thing you'll notice when you install Google Desktop. The first thing you'll see is the Sidebar, a column of rectangular panels hugging the right edge of your screen. Each is a window onto a different kind of real-time information from the Internet. Some are ho-hum, like your latest incoming Gmail and Outlook e-mail, news, stock and weather tickers. Others are refreshingly quirky: the Photos panel shows a continuous, two-inch-tall slideshow of pictures from your own collection, and the surprisingly useful Scratch Pad is a blank box where you can type casual notes throughout your workday (they're saved automatically). Each panel expands horizontally, drawer-like, to reveal more details when clicked. The Sidebar is about as clean-looking as anyone could make it, but it's still a lot of clutter in a very small space, especially if you add new panels as they become available. On the other hand, you can tidy things up quite a bit: drag your Sidebar panels into a different order, hide the ones you don't use, or collapse them into one-line summaries. Skip to next paragraph Enlarge This Image A buddy list, and an exchange, on Google Talk. Multimedia Audio Slideshow Inside Google Desktop Search Readers Forum: David Pogue's Columns Once again, Google isn't the first company to dream up a modular, Internet-connected suite of miniprograms; the Sidebar is a lot like Mac OS X's Dashboard or the shareware programs Desktop X and Konfabulator. But never mind that; you can't keep a good idea down, and this is a good one indeed. Google's second revelation this week, Google Talk, lets you communicate with your buddies either by typing or, if your PC has a microphone and speaker, by speaking. As long as you and your conversation partner are at Windows computers, you can converse with spectacular sound quality. Now, Google Talk 1.0 is probably the most stripped-down chat program on earth. No conference calling, video chats or direct person-to-person file transfers. (Features like these are common in rivals like Skype, iChat and the messenger programs from AOL, MSN and Yahoo.) So what, exactly, is Google trying to prove here? Its mission, in fact, is far grander. Google Talk aims to end the ridiculous era of proprietary chat networks. At the moment, AOL, MSN and Yahoo each maintain separate, incompatible networks. The big boys each want to be alone in the sandbox, and the losers are their customers. Google Talk, however, is based on an open, published standard that the company is making available to all. Already, Google Talk communicates with popular chat programs like iChat, Trillian, Adium, Psi and GAIM, but that's just the beginning. Google is making overtures to Yahoo, AOL and Microsoft about making their chat programs compatible; EarthLink has already agreed to join the federation; and Google is also inviting the makers of games, collaboration tools and even cellphones to join in what it hopes will one day be a grand, unified chat network. In the meantime, Google Talk is significant for another reason: it requires a Gmail account. (Gmail is Google's free, Web-based e-mail service, whose two most famous aspects are its vast capacity - over two gigabytes of storage for each account - and the ads that appear, in small type, off to the right side of each message you read. The ads are computer-matched to keywords in the body of the message, which disturbs some privacy advocates.) Until now, Gmail accounts were available by invitation only. Google let the service spread gradually and virally, giving each existing member a few additional invitations to extend. At one point, people were actually selling these invitations on eBay. As of yesterday, however, all that has changed. Now anyone can get a Gmail account - and can therefore use Google Talk. But to prevent spammers and other abusers from snapping up Gmail accounts by the thousands, Google has designed a clever safeguard: when you apply for a Gmail account, you must provide a cellphone number. Google sends a code to your phone, which you use to complete the registration. (Actually, you don't have to own a cellphone; you just have to know somebody with a cellphone. They can get the code for you, because each cellphone number is good for a number of registrations - just not hundreds of them.) In a single week, then, Google, the software company, addressed deficiencies in Windows, tried to create a grand unified chat and voice network, and opened its clean, capable, capacious e-mail system to all comers. All of this software is beautifully done, quick to download and fun to use - not to mention free and (apart from the Gmail service) entirely free of ads and come-ons. Wish they'd cut it out. Trying to figure out what this company's really up to is enough to drive you crazy. -- Joanne Hallisey Sr. Project Manager MIT - Information Services and Technology 617-253-1894 -------------- next part -------------- An HTML attachment was scrubbed... URL: http://mailman.mit.edu/pipermail/wocky/attachments/20050825/9f7b90bf/attachment.htm From jmhunt at MIT.EDU Thu Aug 25 11:51:47 2005 From: jmhunt at MIT.EDU (Jonathan McIndoe Hunt) Date: Thu, 25 Aug 2005 11:51:47 -0400 Subject: [Wocky] anyone else having problems signing onto Jabber server? Message-ID: <6.2.5.1.2.20050825114937.0419ce20@mit.edu> I just tried signing on three times and it failed everytime. The error says: Couldn't connect to host. I pinged jabber.mit.edu and got responses. Thanks, Jon From mark at MIT.EDU Thu Aug 25 12:22:46 2005 From: mark at MIT.EDU (Mark Silis) Date: Thu, 25 Aug 2005 12:22:46 -0400 Subject: [Wocky] anyone else having problems signing onto Jabber server? In-Reply-To: <6.2.5.1.2.20050825114937.0419ce20@mit.edu> References: <6.2.5.1.2.20050825114937.0419ce20@mit.edu> Message-ID: Hi Jon, Thanks for the report it should be available once again. -- Mark On Aug 25, 2005, at 11:51 AM, Jonathan McIndoe Hunt wrote: > I just tried signing on three times and it failed everytime. The > error says: > > Couldn't connect to host. > > I pinged jabber.mit.edu and got responses. > > Thanks, > Jon > > _______________________________________________ > Wocky mailing list > Wocky at mit.edu > http://mailman.mit.edu/mailman/listinfo/wocky > From jmhunt at MIT.EDU Thu Aug 25 12:53:41 2005 From: jmhunt at MIT.EDU (Jonathan McIndoe Hunt) Date: Thu, 25 Aug 2005 12:53:41 -0400 Subject: [Wocky] anyone else having problems signing onto Jabber server? In-Reply-To: References: <6.2.5.1.2.20050825114937.0419ce20@mit.edu> Message-ID: <6.2.5.1.2.20050825125315.04152f68@mit.edu> Yep...thanks At 12:22 PM 8/25/2005, Mark Silis wrote: >Hi Jon, > >Thanks for the report it should be available once again. > >-- Mark > >On Aug 25, 2005, at 11:51 AM, Jonathan McIndoe Hunt wrote: > >>I just tried signing on three times and it failed everytime. The >>error says: >> >>Couldn't connect to host. >> >>I pinged jabber.mit.edu and got responses. >> >>Thanks, >>Jon >> >>_______________________________________________ >>Wocky mailing list >>Wocky at mit.edu >>http://mailman.mit.edu/mailman/listinfo/wocky > From ghudson at MIT.EDU Mon Aug 29 15:20:56 2005 From: ghudson at MIT.EDU (Greg Hudson) Date: Mon, 29 Aug 2005 15:20:56 -0400 Subject: [Wocky] Google Talk and the future of IM Message-ID: <200508291920.j7TJKucG022741@egyptian-gods.mit.edu> Google has recently introduced an instant messaging service (still in beta) called Google Talk. It uses the Jabber client-to-server protocol, which is good news because it will encourage improvements in Jabber client support. Unfortunately, it could have been much better news--they could have supported the Jabber server-to-server protocol so that MIT Jabber users could talk to Google Talk users. As it stands, Google Talk will provide no incentive for users to use the MIT Jabber server; instead, it will just add another closed instant messaging network to the existing mix. Google's FAQ at http://www.google.com/talk/about.html#open says: Today, with instant communications, you can't talk to your contacts or buddies in one service while using another service. We hope to change that. We want to work with other willing service providers to enable their users to communicate directly with Google Talk users. http://www.google.com/talk/developer.html#service_5 says: 5. I am a communications service provider and want to federate with the Google Talk service. How do I proceed? Please contact us at federation at google.com. While I'm sure they're mainly hoping to get mail from directors at AOL and MSN, we might consider drafting a note to send to federation at google.com asking them to open their server to other Jabber servers. Enough polite and well-worded requests from people like us might nudge them in the right direction. For more on this, there's an excellent piece at http://www.livejournal.com/users/nugget/97081.html which I agree with fully.