I had an idea and the beginnings of a pet project in which a repo could 
"subscribe" to updates of another repo and kick off Travis builds of that 
project. One concern I had was overloading Travis: if the source project has 
frequent commits, you'd have a frequent stampeding herd (you could limit to a 
certain # per day/hour/etc). The other issue is if you have a repo with a large 
number of subscriptions, you'd flood Travis's queue and prevent other projects 
from getting timely builds. 

> On Aug 3, 2017, at 1:31 PM, Bryan Joseph <[email protected]> wrote:
> 
> That is an interesting way to handle it if possible. My coworker, Luke had 
> the idea of using a docker container per package. So each container would 
> have the latest Elixir build. Either on push to master or a daily build. It 
> would download the latest version of a package from hex, run tests or 
> whatever and then report back to something with the results. Repeat for each 
> package in hex (or a subset). And that way work could be distributed.
> 
>> On Thu, Aug 3, 2017 at 12:08 PM, Josh Austin <[email protected]> wrote:
>> That's interesting, I wonder if you could have an integration where a 
>> webhook from elixir-lang/elixir could trigger Travis test runs on individual 
>> package repos but direct them to use a new build of Elixir and post the 
>> results to a central location (S3 bucket or something).
>> 
>> I'm not an expert on Travis but that sort of configuration might go a long 
>> way toward making the regression check workload manageable. Then we'd just 
>> need to aggregate the results and diff them.
>> 
>> My biggest concern was around deploying the temporary compute resources to 
>> run all these tests, but if we could get the existing Travis configs to run 
>> that could be great.
>> 
>> 
>>> On Thursday, August 3, 2017 at 10:24:49 AM UTC-4, OvermindDL1 wrote:
>>> I'd also be willing to run such a build node as well, though perhaps travis 
>>> could do the heavy-lifting?
>>> 
>>> 
>>>> On Thursday, August 3, 2017 at 7:25:50 AM UTC-6, Bryan Joseph wrote:
>>>> I'm also interested in seeing something like what Rust does exist for 
>>>> Elixir and willing to help make it happen
>>>> 
>>>>> On Wed, Aug 2, 2017 at 5:46 PM, Allen Madsen <[email protected]> wrote:
>>>>> I personally love the idea of releases running against commonly used 
>>>>> packages. I'm not sure how rust does it, but it would be cool if the work 
>>>>> could be distributed. For example, you do x, y, and z in your project and 
>>>>> submit a URL to the elixir team and they have something that hits those 
>>>>> URLs and expects them to ping back success or failure.
>>>>> 
>>>>> Allen Madsen
>>>>> http://www.allenmadsen.com
>>>>> 
>>>>>> On Wed, Aug 2, 2017 at 1:23 PM, Josh Austin <[email protected]> wrote:
>>>>>> Any thoughts about building a regression testing tool for new Elixir 
>>>>>> releases? I'm thinking about something like cargobomb which is used for 
>>>>>> Rust release regression testing.
>>>>>> 
>>>>>> Other cargobomb info:
>>>>>> - blog post: https://brson.github.io/2017/07/10/how-rust-is-tested#s-ds
>>>>>> - example report: 
>>>>>> http://cargobomb-reports.s3-website-us-west-1.amazonaws.com/nightly-2017-07-07/index.html
>>>>>> 
>>>>>> I'm thinking having excellent tools like mix and hex.pm could enable 
>>>>>> something similar for Elixir. I'm interested in knowing your thoughts 
>>>>>> about this.
>>>>>> 
>>>>>> Best,
>>>>>> Josh Austin
>>>>>> 
>>>>>> -- 
>>>>>> You received this message because you are subscribed to the Google 
>>>>>> Groups "elixir-lang-core" group.
>>>>>> To unsubscribe from this group and stop receiving emails from it, send 
>>>>>> an email to [email protected].
>>>>>> To view this discussion on the web visit 
>>>>>> https://groups.google.com/d/msgid/elixir-lang-core/7387f1f0-7124-495d-81b2-a94fcb7efbde%40googlegroups.com.
>>>>>> For more options, visit https://groups.google.com/d/optout.
>>>>> 
>>>>> -- 
>>>>> You received this message because you are subscribed to the Google Groups 
>>>>> "elixir-lang-core" group.
>>>>> To unsubscribe from this group and stop receiving emails from it, send an 
>>>>> email to [email protected].
>>>>> To view this discussion on the web visit 
>>>>> https://groups.google.com/d/msgid/elixir-lang-core/CAK-y3CsPJ6EH4kEOub4rW93zH1WHDrp-fjNWusdw65THdJ0CAQ%40mail.gmail.com.
>>>>> 
>>>>> For more options, visit https://groups.google.com/d/optout.
>>>> 
>> 
>> -- 
>> You received this message because you are subscribed to the Google Groups 
>> "elixir-lang-core" group.
>> To unsubscribe from this group and stop receiving emails from it, send an 
>> email to [email protected].
>> To view this discussion on the web visit 
>> https://groups.google.com/d/msgid/elixir-lang-core/d674917a-cdb0-47f1-a7e4-243042c23e26%40googlegroups.com.
>> 
>> For more options, visit https://groups.google.com/d/optout.
> 
> -- 
> You received this message because you are subscribed to the Google Groups 
> "elixir-lang-core" group.
> To unsubscribe from this group and stop receiving emails from it, send an 
> email to [email protected].
> To view this discussion on the web visit 
> https://groups.google.com/d/msgid/elixir-lang-core/CAKRLFpbWbeJig4EcO3iNex5GdeBSwMBy%3D9qjHDLeXpeWkCYEsw%40mail.gmail.com.
> For more options, visit https://groups.google.com/d/optout.

-- 
You received this message because you are subscribed to the Google Groups 
"elixir-lang-core" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to [email protected].
To view this discussion on the web visit 
https://groups.google.com/d/msgid/elixir-lang-core/2B711905-4DFF-4722-8AF2-C1B1A8C61919%40danmcclain.net.
For more options, visit https://groups.google.com/d/optout.

Reply via email to