Sean, I think you're coming from the predisposition that somehow I am reflecting my current employer's competitive stance with either of the companies in question. Let met tell you - nothing can be farther from the truth. The company I am working for right now doesn't compete with any of the commercial vendors in the business of the packaging and distributing Apache Hadoop stack. Far from it - the company is _completely_ vendor agnostic.
Another presumption you made is that I can be easily forced or manipulated to express or approve of something that isn't of my own opinion or deduction. Absolutely not true! You can easily figure out that exactly these expectations made me to part the ways with the very employer you're affiliated with right now. But enough about me - this isn't a thread to discuss high moral grounds of course. As we have the factual conversation now let me clarify the background: 0/ Wasn't factually wrong, but had an unfortunate choice of words. Here's why: HDP2.2 has stopped providing standard Linux service handles, hence one can not rely on traditional Linux mechanisms to server services life-cycle. Same is true and in the higher extent with Cloudera's parcels. That last qualifier hasn't been expressed clear enough, but now it is. 1/ the links to the tweet is incorrect, so I have hard time put it in the context 2/ Ah, it seems that #1 above is a part of #2 here. What the tweet is conveying is that _none_ of the commercial vendors that are consuming Bigtop aren't making an effort to have Bigtop as an open integration point for the distro works. While it is fine from ASL legal stance, it might be considered of a questionable practice and put off new contributors. But I guess commercials are commercials - they have the right to lock-out their competitors, so we really can not do anything about it. I guess that put the matter at rest as the technical ground of the tweets has been clarified. I think it'd be more clear if the discussions on 'public expressions principle' and 'facts-based' are separated. So I will share my stance on the actual topic of this in a separate email. With regards, Cos On Mon, Mar 09, 2015 at 08:25AM, Sean Mackrory wrote: > Re Nate's question, you can see a list of contributor's affiliations here: > https://bigtop.apache.org/team-list.html. They're not always up to date for > everyone, but mine is, and for most of the active contributors whose status > I know it is. > > >> Affiliations are ispofacto should be irrelevant in any Apache project > > Which is exactly why I think these tweets are unacceptable. I mentioned my > employer for no reason other than full disclosure about what bias I'm > bringing to the table - I explained that very clearly. But I still > absolutely believe that what I'm saying is best for the Apache community. > > Now whoever is writing these tweets has a clear bias towards GridGain's > recent contribution to the project (although I don't think any of those > tweets cross any kind of line) and a very negative bias towards Cloudera > and now Hortonworks. Now, I believe those tweets were written by you (Cos). > If that's the case, don't you think they're clearly influenced by who > you're affiliated with? If you publish those thoughts under your own name, > they're understood in the context of who you are. When you publish them > under ASFbigtop, you (mis)represent them as community consensus, which is > clearly not the case. Just pretending you don't have an affiliation doesn't > mean you don't have one. > > Now there exist facts that have led you to the opinions you have, and I > respect those opinions, but if you really did write these tweets, you were > retweeting your own criticism of companies who often compete with your own > company and replying to your own tweets to create a contrived conversation > under the name of the project? I really can't see how you can defend doing > that in the first place, let alone continuing to allow that. I really hope > I'm wrong about who wrote those tweets. But if I am, let's please get that > out in the open because the use of the ASFbigtop account should also not be > confidential. Regardless of accuracy or author, I still don't think the > tweets contribute to building community or advancing the project. > > >> I believe Apache CloudStack has done a great job at defining some > guidelines for its social media presence. > > I think these guidelines are pretty good. I'd +1 adopting them for our own > project. > > > On Sun, Mar 8, 2015 at 11:35 PM, Bruno MahИ <[email protected]> wrote: > > > Alright, I am falling for it :) > > > > > > 0/ https://twitter.com/ASFbigtop/status/572641141322993664 > > > > @*c0sin* <https://twitter.com/c0sin> at this point only @*ASFbigtop* < > > https://twitter.com/ASFbigtop> is a choice if you're devops & use > > @*puppetlabs* <https://twitter.com/puppetlabs> or #*Chef* < > > https://twitter.com/hashtag/Chef?src=hash> 2 deploy. @*hortonworks* < > > https://twitter.com/hortonworks> @*cloudera* <https://twitter.com/cloudera> > > aren't for you > > > > 1/ https://twitter.com/ASFbigtop/status/572641141322993664 > > > > Yet @*mattbrandwein* <https://twitter.com/mattbrandwein> both @*cloudera* > > <https://twitter.com/cloudera> @*hortonworks* <https://twitter.com/ > > hortonworks> are posing as a big friends of the project. > > #*whatsupwiththat* <https://twitter.com/hashtag/whatsupwiththat?src=hash> > > ? > > > > 2/ https://twitter.com/ASFbigtop/status/567910541009661952 > > > > What's really odd @*mattbrandwein* <https://twitter.com/mattbrandwein> is > > that @*cloudera* <https://twitter.com/cloudera> (and @*hortonworks* < > > https://twitter.com/hortonworks> too) have dropped the contributions to > > @*ASFbigtop* <https://twitter.com/ASFbigtop> long time ago. > > > > > > > > 0/ is factually wrong because: > > * It got corrected later on :) > > * This general statement is wrong. I see cases where it is better to use > > Apache Bigtop when using puppets or chef. I see cases where it is better to > > use hortonworks' or cloudera's distributions when using puppet or chef. > > > > > > 1/ is factually wrong because: > > * if they are posing as friends, then they are not our friends. However I > > do not see any fact to support such claim. But I do see facts that Cloudera > > is friendly to Apache Bigtop. See the recent blog posts about Apache Bigtop > > from Cloudera. See also the contributions and support from Cloudera. > > > > 2/ is factually wrong because: > > * http://bigtop01.cloudera.org:8080/ . Do I need to say more? > > * Other members of this community working for Cloudera can probably give > > more details if needed > > > > > > But besides how factual are these tweets, the issues I see with them are: > > * They are all factually wrong. But it may be too late in people minds. > > * They do not represent the consensus among the PMC or community > > * As part of the PMC, I am de facto associated with these tweets. And as I > > consider them unprofessional and contrary to the Apache Way, I do not wish > > to be associated with them. Unfortunately I do not know what to do besides > > publicly denouncing such tweet every time or removing myself from the > > project. > > * They are shaming contributors. Not only there is nothing wrong with > > contributing or not, but nothing positive can come out of it. I also don't > > believe shaming contributors is something a member of an Apache community > > would do. > > * They go against http://www.apache.org/foundation/policies/conduct.html > > * This creates a hostile environment for the same contributors/entities > > and could scare away potential contributors and users. They have less > > incentive to participate and an even more difficult time to justify > > participating. > > * They do not further Apache Bigtop as a project or a community > > * 0/ 1/ and 2/ are mere opinions that should belong to a private tweeter > > account and not an ASF branded one > > * I do not know the brand/marketing implications of 0/ since it could be > > constructed as the ASF making some sort of judgment of the quality of these > > products. > > > > > > Thanks, > > Bruno > > > > > > On 03/08/2015 09:38 PM, Konstantin Boudnik wrote: > > > >> What I think we'll benefit us all is a subject matter discussion - and not > >> just an exercise in demagogue. So, if one says he/she believes that fact > >> A was > >> technically incorrect - it'd be courteous to the rest of the participants > >> to > >> explain what was the fact A and why it was perceived to be incorrect. > >> Otherwise how such a statement can be taken into consideration to build a > >> consensus? > >> > >> Cos > >> > >> On Sun, Mar 08, 2015 at 09:17PM, Bruno MahИ wrote: > >> > >>> Hi Cos, > >>> > >>> Let's not distract the discussion by reading too much into that sentence. > >>> It looked to me as if Sean simply listed all the possible cases (you > >>> missed the part about opinion, which is different from affiliation). > >>> > >>> If you really insist, I can copy paste the tweets that are not > >>> factual from the private thread that you have already been made > >>> aware of. > >>> However I would rather focus now on driving consensus on some guidelines. > >>> > >>> Thanks, > >>> Bruno > >>> > >>> Note: Contributors are supposed to not be influenced by their > >>> affiliation on ASF premises. However it does not mean that we cannot > >>> mention them if this is the case. > >>> I am not saying it is the case here and I don't think it is, but I > >>> wanted to clarify this misunderstanding. > >>> > >>> > >>> On 03/08/2015 08:37 PM, Konstantin Boudnik wrote: > >>> > >>>> Sean, > >>>> > >>>> in the interest of being factual when you say > >>>> > >>>> On Sun, Mar 08, 2015 at 05:25PM, Sean Mackrory wrote: > >>>> > >>>>> represent consensus among the Bigtop PMC. In my view, some of the > >>>>> tweets on > >>>>> that account have been factually incorrect, overly biased by the > >>>>> author's > >>>>> own affiliations and opinions, and they are harmful to the community. > >>>>> There > >>>>> > >>>> Could you please be more specific what were technical incorrectness or > >>>> an > >>>> expression of affiliation? Affiliations are ispofacto should be > >>>> irrelevant in > >>>> any Apache project. Establishing the factual technical ground would, I > >>>> am > >>>> sure, will help to have a discussion in a more subject matter. > >>>> > >>>> Regards, > >>>> Cos > >>>> > >>>> > >
