Hacker News new | comments | show | ask | jobs | submit login
Oracle in conflict with another open source project? (hudson-labs.org)
129 points by abayer on Nov 30, 2010 | hide | past | web | favorite | 34 comments

For a little bit of context, I've done some plug-in commits in my time and met the important guys in this project.

Unlike some categorizations in the comments here, I wouldn't say that the devs are anti-Oracle, they're pro-Hudson. I think they've tried to work with Oracle as much as they can, and certainly shown more patience than I think I would have. Once it became clear Oracle didn't necessarily have Hudson's best intentions at heart, things started to go downhill.

Oracle has not been especially communicative, and it's not especially respectful to literally pull the plug on a project's infrastructure without decent prior warning. For all the failings of java.net, when it was run by Sun, they did at least make sure people knew ahead of time of planned outages (the unplanned we don't discuss).

From my POV, it looks like Oracle has two heads: there's the nasty, legal head that is happy to sue everyone with Sun's patent arsenal, and there's the plain incompetent head that doesn't have a clue how to interact with open-source development. We're seeing the latter here, and I guess the community is just fortunate that Hudson isn't patent-encumbered (as far as I know). I don't think there's any particular malice here, just a general level of incompetence and hubris.

I agree. This sounds like it's just a classic case of Not Invented Here syndrome more than anything else. I've been in a similar situation before: a big company is insistent that everyone use their own internal infrastructure because they're skeptical of the reliability of others' networks. Of course, they ignore the fact that their network is terribly unreliable in making this decision.

As someone who has been only a light user of java.net (attempting to checkout code, navigate source repositories online, etc.) I cannot imagine how anyone could have much faith that a new architecture for this service would be any better.

Very odd that the Oracle VP states that in order to remain a part of the "Java community", Hudson should remain hosted on java.net. There is no way to quantify this but it seems like a very, very tiny minority of OSS Java projects use this service for hosting.

Wonder how Apache feels about not being apart of the Java community.

Want: Github, Git, Google groups. Don't want: Java.net, kenai, Oracle messing around. Have no problems with: Fork, name change.

Silent "larger community" user has spoken.

I think by "silent larger community" with "dozens" of people, who he's really talking about is people pay Oracle a LOT of money. Oracle wants these assets to stay under as much Oracle influence as possible so customers don't go getting any bright ideas about taking their money or attention elsewhere.

I'm kind of rooting for a fork, just as a lab of top-down versus developer-driven progress, with the Oracle version staying on SVN.

Cool names might be 'NortheastPassage' or 'Passage', for the water-navigable route explorer Henry Hudson sought. Or something based on features near the Hudson river in New York, like 'Adirondack' or 'Albany'. Or something in contrast or crossing of the river, like 'EastRiver' or 'GWBridge' or 'HollandTunnel' or 'Verrazano'.

If we are talking rivers here then "Rubicon" might be an apt project fork name.

The Mohawk is a large river that feeds into the Hudson just out side of Albany. I think we should name it that :)

I vote for calling it "Jenkins": http://twitter.com/atmos/status/16880282279

> OH: "it might run weird on our butler ci. what is it? jenkins?"

Is Oracle just actively trying to destroy their acquired assets now?

Quoting many people, no need to attribute to malice what can be adequately explained by incompetence.

Fork it and rename it.

Hudson is too important to me, and lots of other developers, to be subject to the whims of Oracle executives. If ever there was a case where forking is the most obvious answer, this is it.

As for people who have built a business out of Hudson: "Hudson" is only a name and if Oracle wants the name, let them have it.

If you rename it I'll do my part in helping with the marketing of the new name by spreading the word among my peers and I suspect most other Hudson users would do the same.

How did Oracle got to own this project? I imagine that thru the Sun acquisition. But, was this a project started/sponsored by Sun. I couldn't find any information in the hudson site or in the wikipedia page.

Kohsuke Kawaguchi (the founder/lead dev) used to work for Sun, and developed Hudson for them. Oracle got Hudson via the acquisition. Kohsuke stayed at Oracle for a while (longer than some other Sun devs, certainly), then left to form Infradna instead, providing Hudson consulting.

[1] http://infradna.com/

>Permission is hereby granted, free of charge, to any person obtaining a copy of this software and associated documentation files (the "Software"), to deal in the Software without restriction, including without limitation the rights to use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies of the Software,

>package hudson;

You might own the trademark Oracle, but you just licensed the use of it to everyone.

"copies of the Software"

Doesn't say anything about the name. Trademark law and copyright law are related but very distinct. If Oracle owns and enforces the trademark "Hudson" in the context of continuous integration software, and they say "Stop using the name Hudson", you must comply regardless of any copyright licensing in effect.

Compare Firefox which has a free software license but prohibits use of the Firefox name if any modifications are made to the code. This is why it's called Iceweasel on Debian -- as I understand it, even a security patch is technically enough to violate the contract.

On the contrary: the software is called "hudson", and the license explicitly grants the right to deal in the software without restriction. Those rights include the rights to distribute and publish, but "including" doesnt mean "only". In any case, how would one publish this software without using the name "hudson"? Does this license imply that prior to publishing the software we have to replace every instane of the letters h u d s o n with something else? That would not be "without restriction".

There are several licenses written that say "you may not use our name to promote your version of the software" or something like that. MIT is not one of them.

Oracle is welcome to fork it and give it a tougher license. Dont expect it will see many commits after that tho.

Is that revocable though? Usually grants that are irrevocable will say so...

"Without restriction". Not that that matters. If a license doesnt have a term, then its irrevocable. Oracle can always change the license for future releases.

You must bear in mind that this is open source. Oracle does not own this. It doesn't appear to be aware of this.

hmm the community should apply some fun and humor..

name the fork SSBounty

"Oracle in conflict with another open source project?"

I know it is chic to dump on Oracle these days. Can we please get a neutral title or the same title as the original post? I disagree with the use of the word "another" and the question mark. A better title would be "Conflict between Oracle and Hudson project".

EDIT: It looks like Andrew Bayer, a core developer on the Hudson project, submitted the story. It also looks like any comments calling for reasonable, rational debate are being down-voted. That's all I'm asking: reasonable, rationale debate.

It looks like that Oracle doesn't like the way that the project is run and this isn't an attack on open source. This is a disagreement about how the project is run. But we can't even have that debate because the topic has been slanted.

1. The title is not untrue.

2. It's not known whether this will lead to a full blown "conflict" (although, who are we kidding, it probably will)

I don't know anything about Hudson. But when I see stories like this (big bad corporate imposes their will on open source), I always think the drama queens (on both sides) are busy having "meetings" instead of doing actual work.

Want to fork? Then fork! You don't need anyone's permission. In this day of DVCS it's about two commands to fork back should relations improve. Using github is different only in degree, not in kind, from using a different IDE on your local machine. You don't have to have a holy war about it. If your development is better as a result you will win, if not you won't. It might be fun to throw a little tantrum, but it won't make the code any better.

If I'm a developer, especially a developer on a project for which I'm not getting paid, I'll do the development using whatever the hell tools I want, thanks. And if someone tells me I can't, frankly, it's not really up to them.

> I don't know anything about Hudson. But when I see stories like this (big bad corporate imposes their will on open source), I always think the drama queens (on both sides) are busy having "meetings" instead of doing actual work.

>If I'm a developer, especially a developer on a project for which I'm not getting paid, I'll do the development using whatever the hell tools I want, thanks. And if someone tells me I can't, frankly, it's not really up to them.

This is exactly the position of the Hudson developers, so why call them drama queens? They can't "just fork it". It's a large open source project with 100+ developers. They need to plan, get consensus, etc. Those "meetings" you seem to disdain aren't just for people to bitch.

I hope they do change the name and fork it. Oracle needs to learn a lesson here and they are absolutely powerless (for once) other than having the TM on Hudson.

>Those "meetings" you seem to disdain aren't just for people to bitch.

As Cory Doctorow put it in For The Win, [1] meetings, and everything that go with them are the price of being superhuman. (Where "superhuman" is defined as "not just sitting in a tree eating berries".)

[1] http://craphound.com/ftw/Cory_Doctorow_-_For_the_Win.htm

In the email from the Oracle VP, he pretty much invites the developer community to fork the project under a different name:

Because it is open source, we can't stop anybody from forking it. We do however own the trademark to the name so you cannot use the name outside of the core community. We acquired that as part of Sun.

The issue though is of course not that cut and dry. The community would want to think over and debate any such move as it would have a long-term impact on the overall working relationship with Oracle, Sonatype, other Hudson contributors/sponsors, etc...

Invitations are usually welcoming. "We can't stop you" doesn't really strike me as such.

I dunno, maybe I'm reading into this too much, but it almost sounds like a "hint" so as not to upset others at Oracle. Kinda like "I don't have that thing you're looking for, but if I did, it would be in my unlocked office in the third desk drawer on the left. I have to go to lunch now."

I'm not sure I can see that, since the sentence following that quote is »We hope that everyone working on hudson today will do as they claim to want, and work with us to make hudson stronger.« This reads quite passive-aggressive to me, indicating that either you go the Oracle way, or you're clearly not really interested in bringing the project forward.

And even if the above were a subtle hint intended to be below the Oracle radar, it would be a suggestion or opinion of a single person, not an invitation.

I also read a bit of sarcastic "good luck making a Hudson without the name, we'll still have it". Presumably there would not be any Oracle changes that break plug-ins, so a lot of users would have to manually migrate to a fork. I'm guessing Oracle feels that puts them in a good position.

> Those "meetings" you seem to disdain aren't just for people to bitch.

Yes they are. If you read through the posts in the actual mailing list, consensus has been built--the consensus is "move to github".

See several developers' comments here: http://groups.google.com/group/hudson-users/msg/5c6ec5888594...



This entire thread:



The only reason it hasn't happened already is because Oracle wants to bitch. I'm calling the developers drama queens too because instead of forking after consensus was built, which was the logical thing to do, they are still talking to Oracle and/or blogging about what Oracle will or will not let them do.

A move to github doesn't constitute a fork, though, does it?

My guess is that many of the active, non-Oracle developers have a significant financial interest in the project's continued success and open-ness. Either they themselves make money via Hudson (consulting, add-ons, etc.) or their employers have many projects depending on Hudson.

Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | Legal | Apply to YC | Contact