On 09/18/2015 03:42 AM, Max Rydahl Andersen wrote:
On 18 Sep 2015, at 9:31, Mickael Istria wrote:
> On 09/18/2015 08:56 AM, Max Rydahl Andersen wrote:
>> The jetty stuff afaik is in best case not *dependent* by SR.1, but
>> affected by it - but I wouldn't bet on it "just working".
> I would bet that the Jetty part failed because of using Jetty 9.2.10
> and 9.2.13 and facing the ClassNotFoundException on SpinLock detected
> in 4.3.0.CR1
SR0 included jetty 9.2.9, but we used 9.2.10 so my guess is that using
9.2.13 will continue to work on SR0 too as long as we have the version
range limited.
Yes, this is what we guessed when decided to restrict our jetty
dependency to 9.2.13.
>> We might have other being dependent on SR1 though, would be good to
>> hear
>> what those are.
> Several fixes depend on SR1. However, I don't know whether the version
> range are strictly set to "install" the necessary SR1 part while
> installing JBoss Tools.
Which fixes are that ? those are important to have identified.
Agree, it's important to know. I'm not aware of any besides jetty for
livereload/browsersim/cordovasim.
>> btw. the last years we built against SR0 but tested against latest
>> SR.
>> Did we stop that ?
>>
> Seems like yes, we stopped that. I can't find any reason except that
> we've forgotten to keep it in place, as the jetty issues somehow
> forced us to be more careful with SR1 than with SR0.
so I suggest we bring that back if we can't find anything that
*actually* requires SR1.
But this time we can't build against a pure SR0
because of jetty.
So we need to build against SR0 + jetty 9.2.13 and we will see if it fails.
Thanks.