Never-ending Travis build failures

[James Le Cuirot]

I'm still sat in #adhearsion and several times a day, we get reports
from Travis that the same changes for blather and ruby_ami are failing
over and over again. I don't know why Travis keeps retrying these but
it is a community resource so please be good citizens and make it stop.

Regards,
James

[Ben Langfeld]

I have been working on the assumption that someone still has a stake in Adhearsion and might fix it. If that's no longer the case then I'll switch off as much as I can.

Ben

Enviado do meu iPhone

> --
> You received this message because you are subscribed to the Google Groups "Adhearsion" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to adhearsion+...@googlegroups.com.
> For more options, visit https://groups.google.com/d/optout.

[Luca Pradovera]

Hello,

there are a lot of things that need to be brought back under control over the various Adhearsion projects.

We have been swamped with work or I would have gone through the releases and fixes myself earlier.

I will try to get a roadmap to a 3.0 release and the various gem fixes drafted next week.

Best regards,

Luca

> To unsubscribe from this group and stop receiving emails from it, send an email to adhearsion+unsubscribe@googlegroups.com.

> For more options, visit https://groups.google.com/d/optout.

--
You received this message because you are subscribed to the Google Groups "Adhearsion" group.

To unsubscribe from this group and stop receiving emails from it, send an email to adhearsion+unsubscribe@googlegroups.com.

[James Le Cuirot]

On Mon, 24 Jul 2017 17:56:44 +0100
Ben Langfeld <b...@langfeld.co.uk> wrote:

> > Em 24 de jul de 2017, às 15:30, James Le Cuirot
> > <ch...@aura-online.co.uk> escreveu:
> >
> > I'm still sat in #adhearsion and several times a day, we get reports
> > from Travis that the same changes for blather and ruby_ami are
> > failing over and over again. I don't know why Travis keeps retrying
> > these but it is a community resource so please be good citizens and
> > make it stop.

> I have been working on the assumption that someone still has a stake
> in Adhearsion and might fix it. If that's no longer the case then
> I'll switch off as much as I can.

It was a while before I realised that it was the same changes being
retried. I now see that these are nightly tests. I guess the idea there
is to check that they still work with newer gems? That is useful but
perhaps less so when development has stagnated.

As for stakeholders, it helps if you tell the community what is
(or isn't) going on. We were left to figure out for ourselves that
Adhearsion had effectively been put on life support. I contacted Ben
Klang about it a year ago as we were evaluating our options going
forward. He explained that there had been a change in circumstances but
that maintenance would continue, albeit at a slower pace. To be honest,
what I have seen since hasn't filled me with confidence. I know that
you poured your heart and soul into this project so I trust you would
do more now if you had the time but please don't keep the community in
the dark.

We have now developed our new system on Adhearsion 3. It's live, it
works, and we're happy with it. I hardly had to make any changes and
Ben Klang merged the ones I did make so thanks for that. I did raise a
couple of issues that I wanted your opinion on to ensure I fixed them
in the right way. If I'm not going to receive that then perhaps you
should grant commit access to me and whoever else needs it so that we
can just go ahead and fix these things. I'm not planning to make any
large changes but I can at least deal with any issues as I find them.

Regards,
James

[Ben Langfeld]

You now have commit access along with a few other members of the community who have had for a while after requesting it. Luca Pradovera has release access and is effectively managing the project at this point since he uses it every day.

I have not had any need to use Adhearsion for around 2 years. After much usage and maintenance and knowing all the darkest secrets I am left frustrated at the architecture of the framework down to the detail of language choice. This is not relevant to most casual Adhearsion users, but is one of several reasons that I do not make an active effort at maintenance.

Among others are exhaustion, other demands on my time and frustration with limited community engagement over the 7 years I have been involved in the project (as of this month!).

I will do my best to review any outstanding requests for my comments. Please ping me on those to remind me if I don't respond within a few days.

Ben

Enviado do meu iPhone

[Luca Pradovera]

Everybody,

welcome to the Adhearsion project as committers!

As I said, there is plenty to do and I would appreciate the help. Part of the reason why development has stagnated is that, with the Bens not using Adhearsion any more, I have been working on it alone for a while.

Development stoppage has also resulted in plenty of mutually incompatible PRs, CI failures, and other issues that are honestly too much to manage for only one developer.

It has been my goal for a while to clean up the situation, release Adhearsion 3 (which I use daily) and all gems in an 3-compatible version, and create a 2.x.x support branch for the gems.

In the interest of that, I ask everyone to provide any and all feedback regarding what is TODO on the mailing list.

Let's set a deadline to July 31st, as from what I see issues and requests should be well known by now.

The goal is to release Adhearsion 3 by mid-August.

Does that work for you all?

Best regards,

Luca

> To unsubscribe from this group and stop receiving emails from it, send an email to adhearsion+unsubscribe@googlegroups.com.

> For more options, visit https://groups.google.com/d/optout.

--
You received this message because you are subscribed to the Google Groups "Adhearsion" group.

To unsubscribe from this group and stop receiving emails from it, send an email to adhearsion+unsubscribe@googlegroups.com.

[Andrius Kairiukstis]

Thank you,

I think most important things would be following:

– up to date gems

– asterisk 14 support

– ringback and network indications in dial command (currently only fake ringback or some audiofile can be played back to the user)

To unsubscribe from this group and stop receiving emails from it, send an email to adhearsion+...@googlegroups.com.

[James Le Cuirot]

On Tue, 25 Jul 2017 13:22:22 +0100
Ben Langfeld <b...@langfeld.co.uk> wrote:

> You now have commit access along with a few other members of the
> community who have had for a while after requesting it. Luca
> Pradovera has release access and is effectively managing the project
> at this point since he uses it every day.

Thank you very much.

> I have not had any need to use Adhearsion for around 2 years. After
> much usage and maintenance and knowing all the darkest secrets I am
> left frustrated at the architecture of the framework down to the
> detail of language choice. This is not relevant to most casual
> Adhearsion users, but is one of several reasons that I do not make an
> active effort at maintenance.
>
> Among others are exhaustion, other demands on my time and frustration
> with limited community engagement over the 7 years I have been
> involved in the project (as of this month!).

I'm sorry to hear that. Many have moved on from Ruby and I don't intend
to be stuck with it forever but I don't feel it's time to move on yet.
Regarding community engagement, I think you guys did a great job at the
time but the project was always going to be somewhat niche.

> I will do my best to review any outstanding requests for my comments.
> Please ping me on those to remind me if I don't respond within a few
> days.

To save you time, here they are. Admittedly none of these were directed
specifically at you but you used to chime in on almost every discussion
so I didn't feel the need to ping. I also understand if they are tricky
to answer now that you're out of the loop.

https://github.com/adhearsion/ruby_speech/issues/48
https://github.com/adhearsion/adhearsion/issues/618
https://groups.google.com/forum/#!topic/adhearsion/YZdxZFveS00
https://groups.google.com/forum/#!topic/adhearsion/DPiFuvbjCW0

Thanks,
James

[Giorgio Bianchi]

Hi,

Do all the features present till Adhearsion 2.6.4 will be ported in Adhearsion 3? E.g. : https://github.com/adhearsion/adhearsion/pull/614

Is there a porting guide to convert applications based on Adhearsion 2.6.4 to Adhearsion 3?

Regards

Giorgio

> To unsubscribe from this group and stop receiving emails from it, send an email to adhearsion+...@googlegroups.com.

> For more options, visit https://groups.google.com/d/optout.

--
You received this message because you are subscribed to the Google Groups "Adhearsion" group.

To unsubscribe from this group and stop receiving emails from it, send an email to adhearsion+...@googlegroups.com.

[James Le Cuirot]

On Tue, 25 Jul 2017 09:04:32 -0400
Andrius Kairiukstis <k...@andrius.mobi> wrote:

> I think most important things would be following:
>
> – up to date gems

Agreed, particularly Nokogiri as that is vulnerable.

https://github.com/adhearsion/adhearsion/issues/621

> – asterisk 14 support

I don't know what the current state of Asterisk support is but if 14
doesn't work well already then this seems like a big ask for a small
bunch who may not even use Asterisk. I don't (any more) and I suspect
Luca doesn't either. I know that working around Asterisk's quirks has
always been tricky and FreeSWITCH has always worked better.

> – ringback and network indications in dial command (currently only fake
> ringback or some audiofile can be played back to the user)

I'm not sure about network indications dialling is pretty flexible
already. The ringback option can be a proc that returns any object that
responds to #stop!. This can be used to do more than play simple audio
files. I abused it to implement a local buffer that prevents a remote
agent destination from being overloaded.

[Luca Pradovera]

Hello,

Asterisk 14 support consists of a few small fixes, a couple of which I have ready to go.

The main problem there is that I am not really sure we got all of it, as we only have one big project using Asterisk 14 and the fixes I did were aimed at that. There would be a need to help testing.

Ringback indications would be surprisingly complicated to implement due to how we use call origination.

I would suggest that, if you are on Asterisk, to just use `execute` and the Dial() application. I did that many times.

Best regards,

Luca

[James Le Cuirot]

On Tue, 25 Jul 2017 08:39:24 -0400
Luca Pradovera <luca.pr...@gmail.com> wrote:

> In the interest of that, I ask everyone to provide any and all feedback
> regarding what is TODO on the mailing list.
> Let's set a deadline to July 31st, as from what I see issues and requests
> should be well known by now.
> The goal is to release Adhearsion 3 by mid-August.
>
> Does that work for you all?

Thanks, this is the kind of proactive attitude that we need! :) Sorry
for the slight delay. Although I do open source for work and pleasure,
Adhearsion falls quite firmly into the work bracket and I was off on
Friday. I also won't be back until the 8th when I'll no doubt have a
ton of other stuff dumped on me but I'm very keen to see this happen so
I'll do whatever I can.

I don't think there are actually any major showstoppers. As I said, we
are using v3 in production already. Here are some noteworthy issues that
should be relatively easy to fix:

Definitely get the gems up to date, as Nokogiri is vulnerable.
https://github.com/adhearsion/adhearsion/issues/621

Exception events are triggered twice per exception and most people will
notice these annoying double log entries.
https://github.com/adhearsion/adhearsion/issues/618

I've now had Ben confirm that Call#controllers probably doesn't do what
is documented. This should be resolved one way or another.
https://groups.google.com/forum/#!topic/adhearsion/YZdxZFveS00

This concerns ruby_speech rather than Adhearsion so it's not strictly
needed for a v3 release but it is breaking the Rayo spec here.
https://github.com/adhearsion/ruby_speech/issues/48