[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [ide-dev] IDE working group [WAS: Improving Eclipse JDT - Ecosystem]
- From: Doug Schaefer <dschaefer@xxxxxxx>
- Date: Tue, 15 Oct 2013 16:59:59 +0000
- Accept-language: en-US, en-CA
- Delivered-to: firstname.lastname@example.org
- Thread-index: AQHOycf7taKv0hfFFkajqS5zi5J7kw==
- Thread-topic: [ide-dev] IDE working group [WAS: Improving Eclipse JDT - Ecosystem]
- User-agent: Microsoft-MacOutlook/188.8.131.52913
In general, I guess I like this proposal, especially if it brings new
funding into the community. I also wonder how this interacts with
companies already funding contributions directly through the projects.
What role, if any, do they have in the working group?
As a senior member of the community, I'd like to have a say in what the WG
works on. But that's maybe a separate issue revolving around architectural
control over how the Eclipse IDE evolves and how do we ensure the WG and
the existing contributors work together towards common goals and how do we
decide what those goals are. We've always left that to the masses which
while great on paper, isn't always a path to success. A little benevolent
dictatorship, or at least consensus building, goes a long way :).
On 2013-10-15 8:11 AM, "Gunnar Wagenknecht" <gunnar@xxxxxxxxxxxxxxx> wrote:
>Am 15.10.2013 um 13:44 schrieb Aleksandar Kurtakov <akurtako@xxxxxxxxxx>:
>I have one comment about the proposal though:
>> "However, those members do not receive road map ranking points and must
>>agree on allocating the development resources to work items as decided
>>by the IDEWG."
>> This is gonna be a clean showstopper for joining such a working group
>>as we usually contribute only development time and having someone else
>>deciding what to do is something that doesn't make any sense from
>>company POV and many wouldn't even dare to propose such a thing to their
>>management ( I for one wouldn't). Maybe I misundertand it entirely?
>Maybe this needs rewording. My rational is the following: I don't want
>the IDEWG to get into competition with projects for development
>resources. If any member has its own agenda, with own work items he wants
>to contribute resources to, then I'd say he can go straight to the
>projects and let his developers contribute patches/gain committership. He
>doesn't need the WG for that. That's the project development model we
>The primary goal of the IDEWG (IMHO) should be funding the development of
>the Eclipse IDE. Ideally, that can be done by generating enough funding
>so that the IDEWG can pay a full time resources to work on IDE projects
>(eg. JDT, Platform, Egit, etc.). Simply put, those resources should not
>be driven by the interests of their employer but by the interests of the
>> Overall, I find it weird to prefer $$ contributions instead of code
>Of course, this model makes a fundamental assumptions: There are
>companies out there that want to support the development of the Eclipse
>IDE but can't/don't want to hire developers. I want to collect that money
>in the IDEWG and hire developers for that. Most likely, those developers
>can be found in existing member companies. The IDEWG must ensure that the
>collected money is well spent in the interest of the donator.
>> Another concern I have is that such approach would make it even harder
>>for code contributions getting in if it's not on the IDEWG list as
>>committers/reviewers would spend their time on this list and that's the
>Maybe here is the misunderstanding the IDEWG is not supposed to replace
>any planning/contribution practices in place by projects today. I think
>it's a separate channel to facilitate contributions by establishing an
>attractive sponsoring model around them. Major benefits of the IDEWG are
>that (a) there is actually a long-term roadmap and a vision for the IDE
>spanning across multiple projects, (b) interests aren't dominated by a
>single project and/or company and (c) it acts completely open and
>transparent as an Eclipse IWG.
>Maybe that needs to be clearer pointed out?
>ide-dev mailing list