Proposal to revise forest graph and integration practices for JDK 9

[Joe Darcy]

On 11/27/2013 12:52 AM, David Holmes wrote:
> On 27/11/2013 3:13 AM, Joe Darcy wrote:
>> On 11/25/2013 7:19 AM, Chris Hegarty wrote:
>>> I am really happy to see this issue being discussed. I'm in favor of
>>> fewer, simpler structured, forests, and this proposal seems to give 
>>> that.
>>>
>>> There is one potential issue I see. Having done several bulk
>>> integrations into jdk8/tl over the past year, I found it nearly
>>> impossible to get a stable/quite jdk8/tl. Ignoring the stability issue
>>> for now ( being discussed in another thread ), if hotspot is to
>>> integrate into -dev it will be nearly impossible to for the integrator
>>> to actually build and test, the latest source, before pushing. As the
>>> underlying repos in -dev are bound to be moving at a fast pace.
>>>
>>> It is worth noting that currently this is not an issue as master is
>>> quite, apart from the scheduled/well known integration slots.
>>>
>>> Apart from the bulk integrations I did into jdk8/tl, I'm not sure that
>>> anyone else downstream is doing anything similar. If so, then their
>>> experiences here would be useful.
>>>
>>> This said, I'm still in favor of the current proposal, just maybe
>>> needs more specifics around integrations.
>>
>> Just a quick comment for now, for this reasons and others, I think it
>> would be helpful if we moved the JDK to a more continuous integration
>> model. The sort of challenges we have in JDK integration are exactly the
>> sort of situations CI can help.
>
> I often hear CI being touted as some kind of silver bullet but I am 
> yet to see how it actually helps. I would be interested to get some 
> details both on what "CI" means exactly and what problems it resolves 
> (and what problems it introduces).
>


We've found the "Continuous Delivery" book by Jez Humble and David 
Farley to be useful.

Some relevant quotes from the first chapter or two of that text:

>
> If It Hurts, Do It More Frequently, and Bring the Pain Forward
> "This is the most general principle on our list, and could perhaps 
> best be described as a heuristic. But it is perhaps the most useful 
> heuristic we know of in the context of delivering software, and it 
> informs everything we say."
>
> Integration is often a very painful process.
> "If this is true of your project, integrate every time somebody checks 
> in, and do it from the start of the project.
> If testing is a painful process that occurs just before release, don’t 
> do it at the end. Instead, do it continually from the beginning of the 
> project."

HTH,

-Joe