Hi Rob, -----Original Message-----
From: Robert Davies <[email protected]> Reply-To: <[email protected]> Date: Saturday, February 1, 2014 1:02 AM To: <[email protected]> Subject: Re: [DISCUSS] [VOTE] Move the ActiveMQ web-console to a sub-project. >Chris, > >On 1 Feb 2014, at 07:26, Chris Mattmann <[email protected]> wrote: > > >> My definition starts with a PMC that knowingly shipped a big >>front-facing >> part of >> its product that had links all over it to technical documents, youtube >> videos, company >> specific information and yet didn't have anything remotely resembling >> Apache as a first >> class citizen. > >> >> So yes, it's more than shipping a web console from another OS Apache >> License project. >> You (part of the Apache ActiveMQ PMC) help to ship an Apache endorsed >> release that >> didn't respect Apache IMO and got the attention of trademarks and the >> Apache board. > >It wasn¹t blatant, is wasn¹t deliberate - ActiveMQ has a lot of >committers and a large PMC. I'm hoping that's the case (that it wasn't blatant), but from what I've seen so far the jury is really out on that. Even statements from you that continue today suggest to me that folks were all very well aware that hawt.io was being incorporated into the sources. hawt.io out of the box does not respect Apache brands. I don't know how to put it any clearer. However, let's not pick on hawt.io. Guess what -- many external web components wouldn't respect Apache brands if we downright shipped them as part of our products. That's why we have to do things like customize them, skin or configure them, etc. Making the excuse that there's no branding policy for situations like this is a straw man and not correct (I'll get to that below). As for its size ActiveMQ's PMC is ~42 people -- that is quite a few -- similar to projects I've worked in including OODT, Hadoop, etc. >The reality is that there are only half a dozen committers who have been >consistently active on the project and have written a very large >proportion of the code. If you look at the committers who are active, >there¹s little cross over between them and committers on hawtio. A good >indication of real activity on the PMC is to look at who¹s been voting >for new committers or new releases - its consistently 6 or 7 people - >James Strachan isn¹t on that list. Thanks for the pointers. Having inactivity on the PMC is one thing (merit at the ASF doesn't expire). The PMC's responsibility is to make sure that the products it delivers and ships as "Apache releases" respect Apache brands. That's why we graduate projects out of the Incubator. That's why we elect ASF "members" who are supposed to know what's up and to ensure that the projects can operate autonomously and in a decentralized fashion so that the board doesn't have to come in and act like we're centrally managing things. When board members do step in, it's usually not with surgical precision -- it's with a bazooka. In this case -- saying most of the PMC is inactive and those that are active don't participate in hawt.io, so that's why it's OK doesn't really make me confident that this PMC knows how to manage an ASF product or set of them which isn't a good thing. I took an action from the ASF board meeting to ensure that the ActiveMQ PMC corrects this situation in which it's shipped a product that doesn't respect Apache brands. The product I'm discussing is ActiveMQ -- it's not hawt.io -- whether hawt.io respects Apache brands is a different thing depending on the Apache software it may or may not use -- I don't really care about that (Shane does though as VP, Trademarks). What I care about is that the ActiveMQ release included a sub component (hawt.io) that needed skinning and tailoring before being released. Without this tailoring we confuse our users who get Apache ActiveMQ and scratch their head why there are links to RedHat and YouTube on that web console and why there is no mention prominently of Apache. > That¹s not a good thing for the project - but its natural for a mature >project to fall into this pattern. There was a genuine intention to >improve the ActiveMQ project by including a new console (along side the >old one) - that was clearly a mistake, but it wasn¹t a blatant attempt >to circumvent the Apache brand. You continue to say that -- great. Were you the one that issued the commit to include hawt.io? If you weren't I wouldn't speculate in general -- if you were the person to issue the commit -- then I'll take your continuing statement that it wasn't intentional to heart. Whatever you or I believe is irrelevant though in the face of the very real requirement that hawt.io as shipped in Active ActiveMQ does not respect Apache brands. > > >There¹s actually no branding policy which covers these situations - I¹d >be happy to help create one. Yes there is. It's called: http://www.apache.org/foundation/marks/pmcs.html#naming This applies to the way we name our projects, to its websites and more generally to the products we ship. The way to look at it is -- first class citizens in our projects (components, libraries, middle wares, web UIs, etc.) need to respect our brands as much as the websites that we steward and maintain about them and as much as they are ultimately what the public associates in its head with our products. I'm not a lawyer. I'm not a trademarks person. But even I can make that association. This is one of the litmus tests of being a PMC member. Cheers, Chris
