For those complaining that Google should have donated more, that's just ridiculous. Google already does donate more; they support the Eclipse Foundation financially.
There are performance issues being reported in the latest Eclipse, and it emerged that automated performance tests that used to be run are not currently being run because there aren't machines available. Google steps up - within 48 hours - with more than enough cash to solve the problem.
The problem needed a small amount of resources now, rather than a huge donation in three months time. A bigger donation might not even have solved this problem - obviously there was some resource allocation failure in the first place.
As an Eclipse user, let me say: thank you, Google.
"Within 48 hours of Mike’s post, a representative of Google’s Open Source Programs Office let the developers know they would be taking up his challenge of helping getting the performance tests running again"
That is awesome. Many companies could not get 20k approved for an internal project in that amount of time.
Eclipse is the primary and recommended dev environment for Android, and it's quite a sluggish beast. This is no gift horse. It's a spoonful of charity where several metric tonnes are called for.
Did you read the article? Google donates quite a bit to Eclipse in money and man power. This one time donation was specific to a need for new machines _today_.
I am a fan of how much Google has put in Eclipse. I think Google employees are on the Eclipse board, they pay Eclipse committers, and might be one of the bigger contributors to the foundation.
However, I do think that Google (and most companies) should step it up even more. Consider how many employees Google has using Eclipse, and how many products Google has building on Eclipse. They are doing a lot with Eclipse and should contribute a lot.
Yes, Eclipse is open source, and companies do not need to contribute to it. But the benefits of open source is a lot more than the free price, it is the freedom to improve bugs that affect your development, and most importantly the freedom to make sure that some single company is not leading the product/project in a strange direction.
I do think that every company that uses Eclipse as an IDE for its employees should have a moral obligation to donate $500 (the typical price of an IDE) to the Eclipse Foundation. And if you are building a product on top of Eclipse, then perhaps 10% of the $500 per user of your product.
Yes, I know I am asking for a lot. Yes, donating to the foundation does include the full time salaries of committers. Yes, there are more companies than Google that need to be doing this. And, yes, I do fear that the tragedy of the commons is to be expected by default for any successful open source project.
As was mentioned in this thread 2 hours ago, Google has 9 committers to Eclipse. http://dash.eclipse.org/dash/commits/web-app/commit-count-lo... They've also paid members of the foundation for over 6 years. And there is never a moral obligation to pay for something that is explicitly licensed for free.
Perhaps Eclipse and the open source community as a whole needs to consider asking for more than just making their code/products free.
I know of way too many people giving so much to open source projects, building 'successful' communities, but having a hard time getting enough investment into the community to guarantee the future growth of such ecosystems.
The problem there is, if Google donated $7.5 million (15,000 engineers * $500) to the Eclipse Foundation, that would certainly give the impression that Google had an unfair amount of influence on the direction of project, compared to a small company only paying a couple of hundred $ (or an open source developer, not paying anything at all).
I think that goes directly against what you say is the primary benefit of open source: that no one company has an undue amount of influence on the product.
I think it's much better that they contribute a few engineering resources directly to the product, rather than just handing over money. That way, they're guaranteed to be engaging directly in the open source "process".
In the case of Eclipse, I think you're missing that IBM is already the dominant contributor, and that indeed the project originated from them.
So I doubt anyone is going to accuse Google of trying to purchase influence.
But, it's an open source project, people and companies are free to contribute as much as they like or not at all. I don't use Eclipse regularly at the moment, but I have in the past and so I'm grateful to everyone who has supported it.
I'm especially grateful for this particular donation, because one of the reasons I don't like to use Eclipse is its sluggishness (although anecdotally I do think it has gotten faster and more stable in recent years, and it doesn't seem any worse than Visual Studio in this regard).
I'm sometimes surprised at how much most FOSS projects -- including prominent ones with large user bases -- struggle for money, test hardware, hosting and even manpower. Look at The GIMP for instance. Just a small band of developers doing their best to release a free image editor that people have incredibly high expectations from. How much money, hardware and manpower does the team working on Photoshop have?
Open source projects deserve our money more than some random platformer game on Kickstarter. Truth is, though, that most FOSS projects do not actively solicit donations, or seem to not know how to do so. People are more inclined to donate when you tell them that their donations will be used to pay 3 developers who will implement features A, B, C and D and make a new release in X months. Nobody will click that small donate button tucked away in the corner of your project website. Nobody will donate when you tell them "financial help is needed for many different reasons", quoting from the GIMP website.
This move from Google is a perfect example of how donations should ideally work. The developers mentioned a concrete problem the project was facing, Google stepped up and donated enough money to solve that particular problem. Nothing more, nothing less.
Running Juno on my work laptop (a recent ThinkPad i5 system) is close to unusably slow while 3.8 runs great (for a Java UI-heavy app).
Hopefully this helps sort things out (though donating time of Google engineers might be a better practical solution). In any case, I think I'll stick with Sublime Text 2 for Android development when doing anything other than on-device debugging.
The community edition of IntelliJ makes a great Android development environment and is free. I can't imagine doing Android dev in a straight text editor.
I love IntelliJ, but I keep coming back to Eclipse because of all the plugins available for it. I would love to see IntelliJ somehow build the 3rd-party community of plugin developers.
They should donate a lot more, and actually get them to make a version that is highly optimized for building Android apps. It should be like "Eclipse for Android", rather than just "Eclipse for Java", with an Android development add-on, as it is now.
This is something Google could do pretty easily themselves right now - I'm a little surprised that they haven't. Installing the plugins isn't that hard, but a nice little package that contained everything you needed for Android development (and none of the stuff you didn't) would probably make the onboarding process for people not accustomed to Eclipse a bit better.
$20K seems reasonable, since it's directed at acquiring hardware for testing Eclipse's performance. You can buy a lot of simple workstations for $20k, especially given that they don't need SSDs or GPUs.
There are performance issues being reported in the latest Eclipse, and it emerged that automated performance tests that used to be run are not currently being run because there aren't machines available. Google steps up - within 48 hours - with more than enough cash to solve the problem.
The problem needed a small amount of resources now, rather than a huge donation in three months time. A bigger donation might not even have solved this problem - obviously there was some resource allocation failure in the first place.
As an Eclipse user, let me say: thank you, Google.