Oracle "sanitizes" Hudson's creator from wiki | Hacker News Reader

Oracle "sanitizes" Hudson's creator from wiki

jenkins.361315.n4.nabble.com

171 pointsrecampbell56 comments

Sure, I can fix it. I wanted to catch up with you on IM as we'd like to regain access to the hudson-labs infrastructure as well. There are a few other VMs we're not sure exactly where they are.

Do I read this as, "We deliberately locked you out and removed your attribution, and we hint we can fix things, and we want you to do certain things for us, but no we aren't actually saying we'll fix anything and no we didn't tell you about it in advance?"

Really, I would just walk away from anyone trying to deal with me on this basis. Come out and say what you did and what you're going to do. If you need my help with something, ask, and don't try to hint things are tied together when they aren't.

I've been following a bit the Hudson debacle and I'm not surprised things are tense there, especially after the new fork.

But I do not think it's as simple as it seems and blaming everything on Oracle's "native evilness" is a bit naive. Quite the contrary, I'm starting to get the feeling that the 'community' might have been played a bit and all the Oracle bashing used as a trampoline for the fork.

Take this message for example: why did Kohsuke feel the need to post this publicly? He left Sun/Oracle, just had closed negotiations between Oracle and his new employer Cloudbees regarding Hudson (done while working at Sun) and recently got "the community" to vote on a fork. He certainly has enough detail contacts to resolve this privately if needed, so why would he choose to post this publicly?

But yeah, the reply in itself couldn't have been more insinuating.

> Really, I would just walk away from anyone trying to deal with me on this basis.

Agreed, with one exception -- to expose the other side to the public. I think Oracle is showing its face (again) and I think it is in the community's interest to _let_ it keep showing its face.

No. If he had started something about Oracle, what would have happened to Oracle? Some redditors and HNers would seize the opportunity to complain about how horrible Oracle is and then everyone would forget about it a few days later. On the other hand, what would have happened to Kawaguchi? Some potential employer could see him complaining publicly about one of his former employers and decide that it's a bad idea to hire him.

In all, there's little good that would come out of it and a ton of risk for Kawaguchi.

That is why I emphasized the word _let_. He doesn't have to editorialize their actions and responses. Those speak for themselves. Just share them on his blog -- he is not complaining, he is just communicating to the open source community the state of the Jenkins project. Let others comment and let Oracle damage its own image.

> what would have happened to Oracle?

Well eventually the same thing that happened to SCO ;-)

They are losing credibility in the PR domain even more. This is not directed towards the general public or even Oracle's corporate customers, but towards the hackers, sysadmins, and other tech workers. The opinions of those people matter however, and eventually they will be consulted by those that make purchasing decisions. A general mistrust and dislike of Oracle will take its toll.

> Well eventually the same thing that happened to SCO ;-)

Oracle has some very good products. SCO had a Unix nobody wanted anymore (at least, nobody with two or more neurons).

For a while they had a Unix some wanted and nobody thought at the time that there will be a free better Unix variant one day. Then GNU/Linux came along...

> For a while they had a Unix some wanted

True. I wanted it.

Yeah, but Oracle's GNU/Linux came a long time ago in the form of MySQL and Postgres.

> Some potential employer could see him complaining publicly about one of his former employers and decide that it's a bad idea to hire him.

I find this very unlikely.

I read it more as being a mistake on Oracle's part and van Zyl expecting some kind of quid pro quo to fix it. I don't think van Zyl actually caused this or anything. It seems to me to be more indicative of van Zyl being an opportunist trying to get something out of fixing a mistake. After all, if this were some kind of diabolical plot, you'd imagine that van Zyl would ask for more than the location of some servers.

UPDATE: According to tweets in another comment, I'm wrong and van Zyl did cause this. I'm still inclined to attribute this to incompetence rather than malice though.

Some recent tweets from Jazon van Zyl make it sound like he's holding Kosuke's request hostage:

This wasn't anyone at Oracle. It was me after a preliminary infrastructure audit, it's not malicious or permanent. http://twitter.com/#!/jvanzyl/status/34413629902168064

This is my request to @kohsukekawa (http://bit.ly/fXsvd7) to air everything in a conversation that everyone can see http://twitter.com/#!/jvanzyl/status/34413629902168064

So in other words... let's have an IM chat for questions I have about the core architecture of a project you built, and then I'll restore your account?

His second reply on the thread sounds even more like he's holding the account hostage:

http://jenkins.361315.n4.nabble.com/Hudson-Dev-Can-I-get-my-...

This shouldn't be surprising at this point. Everyone should know by now that Oracle doesn't give two fucks about anything other than making money.

I happy that Oracle's goal is to make money, but they seem to only care about doing so in the short term. How much goodwill has Oracle blown with influential technical folk over what is really a very minor piece of software in the grand scheme of things? How much harder will it be for Oracle to promote a new technology/standard/whatever if the first thing everyone's thinking is "How are they going to screw me?"

Oracle believes they aren't selling to technical staff, but to non-technical management, and their sales processes are built around that belief. So far, it's working for them.

and after working for ten years in one of the largest IT companies around, trust me that thing about selling to management works!At the end of the day it is the guy in suit who makes most decisions on buying software

True, but the guy in the suit is influenced by people other than salesmen as well, most of all by hype. I remember when Java came onto the scene. Within a few years everyone was busy porting everything to Java. If something like that happens again, Oracle (and others) will have to write down billions.

If I were Oracle, I would not want to provoke someone like Google to drop Java and really throw its weight behind a different mainstream applications platform. Oracle is playing a dangerous game right now. No army of salespeople will save them if and when the tide turns. These things are non linear.

I agree but I think that applies to technologies like Java that are insanely popular.For example I know my company bought a front end proxy software which kinda sucked , I asked the guy who in my opinion was the best front end proxy guy in the company to see if they took his opinion and he said no.So that decision was made way way high up than him so in this case the critical mass of tech experts were so low that the guy in suit did not even bother to consult the best tech guy he had

Agree although I wonder why Oracle then seems to care so much about Hudson -- It's quite un-enterprisey compared to competing build systems which include Sarbanes Oxley compliance features, etc.

This is what I do not get about Oracle. Are we (tech influencers, i guess) all commie hippies that can't make a damn difference at all to their long term business?

Because the answer is either yes or no, and if there is a good possibility that it is no (Oracle's behavior is long term financially detrimental) you might think shareholders would care. But I do not see $ORCL shareholders marching in the street in protest about the apparent pillaging of the company's future. Where are the pitchforks?

See rst's comment. $ORCL doesn't sell "technology". It sell's a "Business Solution".

// Which in itself is generally a good strategy for most startups to emulate.

This is effective in the short term, yes. But "businessmen" do not create the future, they sell (or buy) the present. And the people creating the future are almost entirely aligned against the Oracle ecosystem.

The next generation of business leaders for most companies is inthe same meetings with Oracle's salespeople. They will pick "solutions" for business over their IT staff's opinion. Oracle thinks like them, the IT staff doesn't.

But most people don't want technology. I don't want a website or any complicated stuff - I want to communicate with and see what my friends are doing (facebook). I don't need an amzingly clustered spider and ranking engine, I just want to find sutff (google). I don't want to talk about feeds and speeds, I just want a computer that "works" (apple). I don't want to talk with some IT Nerd about how Open Source software is superior and about technological lock-in, I just want a positive ROI and want to grow my business (Oracle).

// Oh - and HN's favourite -- I just want to play Bingo with my students.

Last I heard though, Oracle was generating mountains of spare cash. The sort of profits they can lay their hands on can buy a lot of Next Big Things to let others take the initial risk and then sell out to Oracle when they're more established.

(Which, from what I understand, isn't far from what Google has done, with the bulk of non-search success having been from acquisitions.)

Didn't we have a fairly decent conversation regarding this idea less than a day ago, "Selling pickaxes during a gold rush"? [1]

I agree, however this is only a short term strategy. Soon, the gold rush will be over and there will be an abundance of pickaxes on the second hand market. Case in point, so called "dark fiber", or the unused infrastructure expansions of the early dot-com bubble.

[1] http://news.ycombinator.com/item?id=2183207

Oracle doesn't sell technology. They sell solutions to {accounting, erp, mrp, etc} problems. Technology is just how they do it, the people buying it (and me, if I were buying oracle) don't care in the least about the technology.

Do you care about the fuel injectors in your car? What about the magnetron in your microwave?

I don't care because they don't seem like details I need to worry about — but if all the mechanics or microwave repairmen said the one I had was a terrible choice, I might care.

Except that Oracle is a platform company. As a CIO you are buying into a platform on top of which your in-house developers, or overpaid consultants, will be building a solution. ERP platforms are not Bingo Card Creator.

Well, okay. but for a CEO the business comes first, developer comfort second. If it can't meet business objectives (which oracle certainly can) it doesn't matter how much developers love it.

Developer comfort is the difference between being woefully understaffed, and having a quality team supporting your business objectives.

If productivity drops, or you suddenly only have 50% of your development staff because management bought an inferior product that developers hate, I don't see how that helps the business?

Wow, I guess I'm the naive techno-utopian of HN.

Considering that the person who responded to Kohsuke is Jason van Zyl, creater of Maven, and not an Oracle employee as far as I know, I have to question the title of this story.

It appears van Zyl is with Sonatype, a company working with Oracle to continue Hudson under its original name:

http://www.sonatype.com/people/2011/02/hudsons-bright-future...

Fantastic excerpt from it

----

The Hudson lead, Winston Prakash from Oracle, is highly skilled, very thoughtful, and he cares about the community. He is also the first person to create detailed, comprehensive architectural documentation.

----

So now the main accomplishment of the current project lead is writing detailed documentation. Those that can - code, those that can't - write very detailed documentation. Yeah, I think I'll stick with Jenkins...

On the other hand good documentation is essential if you want many users. I think that O'Reilly, Coriolis, New Riders, and the other publishers have had more to do with the wider use of open source in the last 15 years than ESR and the FSF itself.

Architectural documentation and user documentation are two completely different things. User documentation is what O'Reilly does. Architectural documentation is what Oracle is claiming to have done.

>> So now the main accomplishment of the current project lead is writing detailed documentation. Those that can - code, those that can't - write very detailed documentation.

So we're now running people down for decent documentation? Seriously?

No, we are looking for a pretext to run down people who work for Oracle.

Because Oracle is behaving like bully we don't need much of a pretext.

Then go with that rather than picking on people for basic good practice.

Let's just say that if I see a resume on my desk and the _main_ accomplishment of the candidate is writing detailed documentation, that person probably won't get the job as a lead developer.

See this recent Sonatype blog post for more background: http://www.sonatype.com/people/2011/02/hudsons-bright-future...

IMO, it sounds like Sonatype (which was already bundling a version of Hudson in their premium product) sees this recent fork as an opportunity to ramp up their contribution to the "offical" Hudson project and take some control.

If this was just an honest mistake, the response should have been worded differently.

At this point, perhaps the account problem could be fixed immediately, and then a separate discussion started to address the hudson-labs issue. To do otherwise seems unnecessarily infantile.

...assuming there really is any 'issue' on either side here, that is.

...and reading through things, I guess there is. Awful.

I agree... Until there is another follow-up post then it's very ambiguous.

Ah yes, Oracle is carrying on the grand tradition of Pharaohs, ancient Rome's damnatio memoriae, and Stalin's Great Purge.

The removal was done by a Sonatype employee; company which could be potentially more vulnerable to this tactics as their products are subject to be chosen by developers. I would think twice before buying them a product after this.

Since Sonatype's offering is centered around Maven, you should think twice regardless of this issue and then look for something better (gradle or buildr, for example, or the old reliable ant).

I'm confused. Based on the second entry in the thread, it sounds like it was just a mistake. Am I missing something?

The response is bringing that into question. It sounds like Oracle is holding his account hostage.

Why would Oracle keep his account with full functionality after he left the company, badmouthed it, and started a fork of the product the company paid him to develop?

I agree removing his name was dumb, but it could be an unpredicted side effect of modifying the account.

If you are familiar with ESR's writings you will know that the primary motivator of a lot of people writing code in the gift economy is recognition.

When it comes to recognition and proper attribution therefore, there are no 'just mistakes'. It is like if you go on holiday and come back to find that the bank has misplaced all of your money, foreclosed on your mortgage and now there are a bunch of smelly windows developers living in what used to be your house.

... and when you go to complain to the bank they say "I don't see why you're so upset, it was just a mistake"...

... except that, in this case, every time you ask the bank to put it right and give you your house back, the bank is saying "sure, no problem, but can we talk about you doing us a little favour first?"

----

I won't apologise for the extremity of the above dramatisation either - stealing someone's credit is just about the worst crime in open source.

The original thread http://jenkins.361315.n4.nabble.com/Hudson-Dev-Can-I-get-my-...

Reply on news.ycombinator.com