Is this thing on?

[Gary Kramlich]

Hey everyone!

This is Gary Kramlich current lead developer of Pidgin.  I'm reaching out to try and figure out what we can do to get Adium back on it's feet.  I have many ideas and would love to share them with you all.

Thanks,

--

Gary Kramlich <gr...@reaperworld.com>

[Gary Kramlich]

Hey everyone!

This is Gary Kramlich current lead developer of Pidgin.  I'm reaching out to try and figure out what we can do to get Adium back on it's feet.  I have many ideas and would love to share them with you all.  So who wants to help?

[Luke Hiesterman]

Maybe start with your ideas?

Luke

[Colin Barrett]

Hi Gary,

I've been trying to get our trac instance back online, or at least recover the dat from it for export. I dropped the ball a bit here, however. In my defense, life has been getting in the way, not a lack of desire.

If anyone thinks they can be particularly helpful with that, please get in touch with me here or on IRC.

-Colin

[Moses Lei]

I'm curious as to what services people are using Adium with these days? I'm not sure if this is a popular opinion, but perhaps we should just gracefully end Adium development, if our user/developer base has declined to that point. Like it or not, the days of interoperable protocols seem to be over, and there are almost no services out there that have an API for third party clients anymore.

My initial reason to contribute to Adium was because of IRC. I'm a cloud architect / IT consultant and over the last few years, almost all my use cases for Adium disappeared.

Personal use cases:

- Facebook XMPP - deprecated by provider

- AIM - no friends use anymore

- Google Talk XMPP - replaced with Hangouts mobile app

- Yahoo - no friends use anymore

- Twitter - replaced with mobile app

Business/work use cases:

- IRC - Replaced with Slack

- SIPE/Lync - Replaced with Slack in one case, HipChat in another, firewalled off in a third (so now we have to use the Windows desktop app via RDS)

So, I have not needed Adium for anything for 4-5 years now (it's not even installed on my last two machines) and thus have no impetus to work on it anymore. I'm curious to know if other people have had the same experience.

Current IM services that I use:

- SMS

- iMessage

- Slack

- WhatsApp

- Skype

- LINE

- WeChat

As far as I know, there is no support in almost any of these for third-party clients, so the raison d'etre for an app like Adium is not fulfillable-- even if we manage to get it onto mobile where it might be more useful.

Thoughts?

Moses

[Luke Hiesterman]

To me the interesting features of a modern chat system aren’t easily achievable by an Adium-like project. I want end-to-end encryption and I want cloud transcript syncing between clients as fundamental basics before I would even consider using a service. Then I need a network effect of the service so people I know are using it.

I struggle to picture a world where a project like Adium could deliver that.

Luke

[Ferhat Ziba]

What about implementing own transparent encrypted service? And deliver an opensource version ? 

[Moses Lei]

On Fri, Jan 18, 2019 at 10:41 AM Ferhat Ziba <fero...@gmail.com> wrote:

What about implementing own transparent encrypted service? And deliver an opensource version ? 

But if one wanted to do that, there's no reason to do that under a project like Adium. It wouldn't be using libpurple or even necessarily Cocoa, so there's little to no code reuse potential from the Adium codebase. And Signal I think fulfills your goals, right? (Encrypted service, open source; not sure what you mean by "transparent" in this context.)

Moses

[Matthew]

My Adium use is limited to IRC and Skype for Business. Adium still
appears to be the best option for either of these.

Outside of Adium, my chats happen in SMS/MMS/RCS, Hangouts, Slack,
Discord, Signal, and Keybase.

--

Matthew

[Carlos Morales]

Maybe the best way forward is to focus on a protocol that:

1) still has an active and relatively large user base - even if is just a niche
2) still is fully supported by the underling Libpurple platform,


Thoughts?

[Colin Barrett]

Lots of good ideas. It will be easier to start to get a critical mass around them once we've recovered our brain (trac).

Here's my take: Multi-protocol chat is always going to be a compromise in terms of experience, but I think there's definitely room for our type of client. New protocols and networks keep proliferating, after all. So we take it slow. Just getting an update out in step with Libpurple, removing (adding) accounts on any abandoned networks, and porting some new prpls to Adium, would be an upgrade for the people out there still using Adium, which is non-zero. "Adium is open for business in 2019" is a humble, but necessary first step.

Thoughts?

-Colin

[Carlos Morales]

"Adium is open for business in 2019" is a humble, but necessary first step.

Definitely!

[Matthew]

> Just getting an update out in step with Libpurple, removing (adding) accounts on any abandoned networks, and porting some new prpls to Adium, would be an upgrade for the people out there still using Adium, which is non-zero. "Adium is open for business in 2019" is a humble, but necessary first step.

Agreed! But I'm guessing that should happen after our current
infrastructure is replaced with something that isn't currently broken
and will be more maintainable. So unless the people ready to do the
work are the ones asking whether it's worth the effort, I'd like to
see our immediate discussion focused on the short term how than the
long(er) term where and why.

(Not directed at anyone in particular...)

--

Matthew

[Colin Barrett]

100% agreed there. We need our brain back!

[erythronium23]

My coworkers and I all use Adium for OTR chat over Google Talk and generic XMPP. Who can't picture a world where Adium provides end-to-end encryption? I use it for that purpose every day.

Google will never deliver a compact, efficient, no-frills, usable, end-to-end encrypted chat solution on the desktop. Neither will WhatsApp, Signal, Slack, HipChat or any other app coming from a culture of full-screen iPad fondling. Every few years I try Pidgin and Psi+ and find them lacking in usability and decent visual compactness compared to Adium as well.

Adium with yMous is irreplaceable for me. At the very least, I think it can prosper as the premiere desktop client for Google Talk and XMPP, protocols which remain open and widely-used.

[Moses Lei]

Colin/Matthew: do we have enough development resources to continue on with the project, in your opinion?

If we don't have people to maintain infrastructure, perhaps we should look at some more creative solutions? For example, how about exporting our trac data and importing it into Pidgin's trac as a component? That saves us from having to manage trac, and gives us direct linking to the pidgin/libpurple community's bugs and issues.

On a similar note, could Adium rebrand and become Pidgin for MacOS? @Gary, I'm curious as to what you think of that.

@erythronium23: I'm not saying Adium isn't useful for people, clearly it is. But we seem to have gotten to the point where the user base is dwindling, and the developer base clearly hasn't gone up in a while since we haven't been putting out releases. Additionally, we don't support voice/video, which is a basic feature of IM systems these days. So, where does that leave the project?

Moses

[Colin Barrett]

I think unfortunately the project we're facing now right now is going to be how to recover trac from online archives, will be an annoying one. I'm not sure what the best way to proceed with something like that is, having not done it myself. In my imagination, a solution would look something like wget downloads HTML, plugs into a python script to scrape the HTML, and it gets written into some neutral file format; then you clean the data, fixup problems; then the data would be imported into wherever we're going next.

-Colin

[erythronium23]

I work at a fairly large company (>2000 engineers on Macs) and very time I introduce a Google Talk user to Adium, they adopt it. I think part of the adoption/growth problem is simply visibility. I think that a clear value proposition, and presence in the app store, would go a long way. I could be wrong, but I think it's not expensive to test that hypothesis.

On Fri, Jan 18, 2019 at 6:06 PM Moses Lei <bu...@moseslei.net> wrote:

[Moses Lei]

Colin, I didn't understand what the issue was; I thought that we had some problems with trac's configuration or hosting, not that we had completely lost the database!

Yes, you are right, restoring from online archives of just rendered web pages *would* be a big challenge.

Moses

[Colin Barrett]

Unfortunately, I believe we cannot be in the App Store due to licensing issues. Correct me if I'm wrong.

[Moses Lei]

On Fri, Jan 18, 2019 at 3:16 PM erythronium23 <erythr...@gmail.com> wrote:

I work at a fairly large company (>2000 engineers on Macs) and very time I introduce a Google Talk user to Adium, they adopt it. I think part of the adoption/growth problem is simply visibility. I think that a clear value proposition, and presence in the app store, would go a long way. I could be wrong, but I think it's not expensive to test that hypothesis.

We've already had the discussion of whether we can be in the App Store years ago. We can't because it violates the GPL and people are not willing to grant exemptions for Adium or libpurple in order to fulfill the App Store's distribution requirements. 

[Matthew]

I'm more interested in the documentation than the tickets. Docs help
current and future users for the product we already provide, no
development necessary. Next would be improvements to enable
incremental development. But even with zero future development, we can
do better for our users.

And while it would certainly be nice to have the tickets, it's a lot
of work, and isn't necessary to do things like "...getting an update

out in step with Libpurple, removing (adding) accounts on any

abandoned networks, and porting some new prpls to Adium".

Is there any reason we shouldn't "simply" move development, tickets,
and wiki to BitBucket?

--

Matthew

[Gary Kramlich]

On Fri, Jan 18, 2019 at 11:06 AM Colin Barrett <co...@springsandstruts.com> wrote:

Hi Gary,

Heya Colin,

 

I've been trying to get our trac instance back online, or at least recover the dat from it for export. I dropped the ball a bit here, however. In my defense, life has been getting in the way, not a lack of desire.

Cool, I've been working with Matthew a bit, he was under the impression the trac backup was rather old?  Doesn't really matter, but I can get something stood up pretty quickly.  Also, while I haven't mentioned it, I did notice your life events (sorry don't know what else to call them) which I why I reached out to others rather than poking you (again) :)

 

If anyone thinks they can be particularly helpful with that, please get in touch with me here or on IRC.

See above ;)

 

-Colin

[Gary Kramlich]

On Fri, Jan 18, 2019 at 11:02 AM Luke Hiesterman <grav...@gmail.com> wrote:

Maybe start with your ideas?

1. move the bitbucket.org/adium account from a single use account to a team account, provide access to everyone that needs it

2. move primary repo from hg.adium.im to bitbucket.org/adium.  We did the same with pidgin about 2 years ago now and it's been working well.

3. Either fix up buildbot or integrate adium into https://bamboo.pidgin.im.  I have a mac mini running high sierra i think, i'd have to check to be sure, and get that setup to do CI which has been planned to get added to the pidgin bamboo already.

4. if we're going to stick with the existing cpanel instance, we can leave it where it is, or we can migrate it to a hosting provider where a pidgin dev works.

5. Get trac back up and running, either on the existing cpanel instance or pull it into the kubernetes cluster I've been migrating pidgin services to.

I had more, but that should be a good start?

 

Luke

[Gary Kramlich]

On Fri, Jan 18, 2019 at 11:27 AM Moses Lei <bu...@moseslei.net> wrote:

I'm curious as to what services people are using Adium with these days? I'm not sure if this is a popular opinion, but perhaps we should just gracefully end Adium development, if our user/developer base has declined to that point. Like it or not, the days of interoperable protocols seem to be over, and there are almost no services out there that have an API for third party clients anymore.

If you all just want to end development, that's your decision.  I'd hate to see it happen, but again it's your call.  As for third party client apis... there never really was any.. I mean there was TOC for AIM way back when... but we (pidgin) have been mimicing the official clients for 20 years no.. 

My initial reason to contribute to Adium was because of IRC. I'm a cloud architect / IT consultant and over the last few years, almost all my use cases for Adium disappeared.

Personal use cases:

- Facebook XMPP - deprecated by provider

- AIM - no friends use anymore

- Google Talk XMPP - replaced with Hangouts mobile app

- Yahoo - no friends use anymore

- Twitter - replaced with mobile app

Facebook: https://github.com/dequis/purple-facebook/wiki

AIM: Shutdown by AOL on December 15th 2017

Google Talk/Hangouts: being discontinued by Google in the near future, but https://bitbucket.org/EionRobb/purple-hangouts/ in the mean time

Yahoo: Protocol rewrite in 2016, later abandoned by the provider for new product "Squirrel"

Twitter: https://github.com/mikeage/prpltwtr

There are many many more at https://d.pidgin.im/wiki/ThirdPartyPlugins#AdditionalProtocols

 

Business/work use cases:

- IRC - Replaced with Slack

- SIPE/Lync - Replaced with Slack in one case, HipChat in another, firewalled off in a third (so now we have to use the Windows desktop app via RDS)

Slack: https://github.com/dylex/slack-libpurple

SIPE/Lync: http://sipe.sourceforge.net/

 

So, I have not needed Adium for anything for 4-5 years now (it's not even installed on my last two machines) and thus have no impetus to work on it anymore. I'm curious to know if other people have had the same experience.

Current IM services that I use:

- SMS

- iMessage

- Slack

- WhatsApp

- Skype

- LINE

- WeChat

Whatsapp bans 3rd party client users on sight, so that's a no go.  SMS could use push bullet, but there are prpls on that page for WeChat, Line, Skype, and Slack.

 

As far as I know, there is no support in almost any of these for third-party clients, so the raison d'etre for an app like Adium is not fulfillable-- even if we manage to get it onto mobile where it might be more useful.

See all my previous statements ;)  The mobile thing is tricky as the iOS app store is not gpl compatible and libpurple is gpl'd and granting an exception to the license for it is unrealistic.

 

Thoughts?

Moses

[Gary Kramlich]

On Fri, Jan 18, 2019 at 11:41 AM Luke Hiesterman <grav...@gmail.com> wrote:

To me the interesting features of a modern chat system aren’t easily achievable by an Adium-like project. I want end-to-end encryption and I want cloud transcript syncing between clients as fundamental basics before I would even consider using a service. Then I need a network effect of the service so people I know are using it.

We are working on tackling many of these issues head on in libpurple3.  That said, porting to it is non trivial.

 

I struggle to picture a world where a project like Adium could deliver that.

Is this not what libpurple is used for in Adium...?

 

Luke

[Gary Kramlich]

On Fri, Jan 18, 2019 at 3:57 PM Colin Barrett <co...@springsandstruts.com> wrote:

Lots of good ideas. It will be easier to start to get a critical mass around them once we've recovered our brain (trac).

Good first step of course!  But you need a way to collaborate and patches on mailing lists/issue trackers is not people want to do anymore.  They know the pull request work flow and bitbucket provides that workflow.

 

Here's my take: Multi-protocol chat is always going to be a compromise in terms of experience, but I think there's definitely room for our type of client. New protocols and networks keep proliferating, after all. So we take it slow. Just getting an update out in step with Libpurple, removing (adding) accounts on any abandoned networks, and porting some new prpls to Adium, would be an upgrade for the people out there still using Adium, which is non-zero. "Adium is open for business in 2019" is a humble, but necessary first step.

I agree with this completely, but open for business in 2019, means being able to do business the 2019 way.  That is, in a way that people are familiar with ;)

 

Thoughts?

-Colin

[Gary Kramlich]

On Fri, Jan 18, 2019 at 4:53 PM erythronium23 <erythr...@gmail.com> wrote:

My coworkers and I all use Adium for OTR chat over Google Talk and generic XMPP. Who can't picture a world where Adium provides end-to-end encryption? I use it for that purpose every day.

Google will never deliver a compact, efficient, no-frills, usable, end-to-end encrypted chat solution on the desktop. Neither will WhatsApp, Signal, Slack, HipChat or any other app coming from a culture of full-screen iPad fondling. Every few years I try Pidgin and Psi+ and find them lacking in usability and decent visual compactness compared to Adium as well.

As I Pidgin dev I'd love to hear your complaints (seriously, can't make it better if I don't know what people don't like :) ) off thread of course ;)

 

Adium with yMous is irreplaceable for me. At the very least, I think it can prosper as the premiere desktop client for Google Talk and XMPP, protocols which remain open and widely-used.

[Matthew]

Perhaps this is ironic considering questions about whether Adium is
still relevant, but when I enabled 2FA on a Google account today, I
noticed this bit of text.

--

Matthew

[Gary Kramlich]

On Fri, Jan 18, 2019 at 5:06 PM Moses Lei <bu...@moseslei.net> wrote:

Colin/Matthew: do we have enough development resources to continue on with the project, in your opinion?

If we don't have people to maintain infrastructure, perhaps we should look at some more creative solutions? For example, how about exporting our trac data and importing it into Pidgin's trac as a component? That saves us from having to manage trac, and gives us direct linking to the pidgin/libpurple community's bugs and issues.

Pidgin's trac will be dying at somepoint... When exactly, I don't know as other options kind of suck...  But trac is a pig on resources and slow to boot.  It needs to be replaced.

 

On a similar note, could Adium rebrand and become Pidgin for MacOS? @Gary, I'm curious as to what you think of that.

I mean.. that's possible sure.. but Pidgin3 uses Gtk3 and Gtk3 has a native Quartz backend, so Pidgin3 will literally just work natively (read no XQuartz) on OSX.

 

[Gary Kramlich]

On Fri, Jan 18, 2019 at 5:15 PM Colin Barrett <co...@springsandstruts.com> wrote:

I think unfortunately the project we're facing now right now is going to be how to recover trac from online archives, will be an annoying one. I'm not sure what the best way to proceed with something like that is, having not done it myself. In my imagination, a solution would look something like wget downloads HTML, plugs into a python script to scrape the HTML, and it gets written into some neutral file format; then you clean the data, fixup problems; then the data would be imported into wherever we're going next.

That sounds absolutely wretched.. Do you still have access to the disk that it was running on?

 

-Colin

[Gary Kramlich]

On Fri, Jan 18, 2019 at 5:19 PM Colin Barrett <co...@springsandstruts.com> wrote:

Unfortunately, I believe we cannot be in the App Store due to licensing issues. Correct me if I'm wrong.

If the OSX app store has the same terms as the iOS app store, then yes, it's hostile to the GPL and it's not feasible to grant an exception for it.

 

[Gary Kramlich]

There's slightly more to it..  There are about 600ish people I think currently in the COPYRIGHT file in the Pidgin repo.  Figuring out who worked on libpurple code is on the surface trivial, untill you remember a lot of credit was given via commit message.  Even then, if we do have email addresses, some of them go back 20 years and all it will take is 1 that covers enough code to make it not worth it.

Forgot to mention before for packaging that there's nothing wrong with a homebrew cask either...

[Gary Kramlich]

On Fri, Jan 18, 2019 at 9:39 PM Matthew <mathuae...@gmail.com> wrote:

I'm more interested in the documentation than the tickets. Docs help
current and future users for the product we already provide, no
development necessary. Next would be improvements to enable
incremental development. But even with zero future development, we can
do better for our users.

Funny you mention documentation.  Pidgin is currently in the process of converting pidgin.im and developer.pidgin.im into a static site generated by hugo.  You can see an early version of it at http://nest.pidgin.im.  All the docs are being converted from the genshi template format that trac uses to markdown which hugo then renders.  Hugo has a huge theme database and it's pretty straight forward to get going with.

 

And while it would certainly be nice to have the tickets, it's a lot
of work, and isn't necessary to do things like "...getting an update
out in step with Libpurple, removing (adding) accounts on any
abandoned networks, and porting some new prpls to Adium".

The other thing about the old tickets is... If they're really bad.. they'll reappear :)

 

Is there any reason we shouldn't "simply" move development, tickets,
and wiki to BitBucket?

Pidgin only moved the code to bitbucket primarily for pull requests.  The wiki isn't bad, but it's not great either.  See my earlier comments about moving to hugo and https://nest.pidgin.im for the way Pidgin went.  Issues are problematic too as Bitbucket's issues are super nerfed as Atlassian wants you to use Jira... That said, Atlassian does offer open source licenses for all of their products.  That's how Pidgin is using Bamboo.

 

Thanks,

--

Gary Kramlich <gr...@reaperworld.com>

[Gary Kramlich]

On Fri, Jan 18, 2019 at 10:01 PM Matthew <mathuae...@gmail.com> wrote:

Perhaps this is ironic considering questions about whether Adium is
still relevant, but when I enabled 2FA on a Google account today, I
noticed this bit of text.

Heh, traditionally you create an "app password" for these "insecure clients", but I'd love to find a way to make libpurple support this setup.  That of course depends on the protocol, but we can of course add api for it.

[Amelia Rose Khan]

What is the benefit of Adium these days?  I like the interface and having multiple services in one place but people have moved on to individual software that do 10 times more.  Has the user base of Pidgin grown over the years?

Are these protocols still growing or used by enough people to warrant them in one place?  AIM Bonjour Gadu-Gadu Google Talk Groupwise ICQ IRC SILC SIMPLE Sametime XMPP Zephyr.  Even people said in this thread they moved to Slack or some similar software and not using IRC.

I feel the issue is the number of developers who work on the project.  A lot of the developers got busy with life and that is great.

[Gary Kramlich]

On Sat, Jan 19, 2019 at 12:38 PM Amelia Rose Khan <shawn...@gmail.com> wrote:

What is the benefit of Adium these days?  I like the interface and having multiple services in one place but people have moved on to individual software that do 10 times more.  Has the user base of Pidgin grown over the years?

The user base of Pidgin has been waning over the years, but a lot of that can be accounted for by lack of releases/support because contributors have "lifed out".  That is, they started careers and/or families and they just don't have time to contribute anymore.  Three years ago (January 2016) when I took over Pidgin my focus has been on getting Pidgin3 and libpurple3 released while making the code base more approachable so that we can attract and maintain new contributors which is something we've been struggling to do since the early 2000's.

 

Are these protocols still growing or used by enough people to warrant them in one place?  AIM Bonjour Gadu-Gadu Google Talk Groupwise ICQ IRC SILC SIMPLE Sametime XMPP Zephyr.  Even people said in this thread they moved to Slack or some similar software and not using IRC.

Third party plugins have always been the life blood of Pidgin, adding new unique features that wouldn't necessarily be considered into mainline immediately.  As such, there is a large list of third party protocol plugins at https://developer.pidgin.im/wiki/ThirdPartyPlugins.  You'll notice there's a slack plugin there as well.  We're still working on cleaning up the voice and video bits in the 3.0 code base, but that is a priority now as people use it every day.

 

I feel the issue is the number of developers who work on the project.  A lot of the developers got busy with life and that is great.

Yep! See my earlier comments about "lifing out" :)

 

[Matthew]

For reference, Sparkle shows 2000 active users this week.
Historically, Sparkle's been ~50% the actual number, but I have no
clue if that's still a useful estimate. Sparkle recorded 5019 users
the week after the most recent release Apr 27, 2017).

I feel confident that would go up if Adium didn't look like a dead
project, and if it supported protocols people are using these days. As
Gary has said, libpurple supports quite a bit more than what's
currently available in Adium.

--

Matthew