Microsoft quietly buys Netscape browser technology (and patents) | Hacker News Reader

Microsoft quietly buys Netscape browser technology (and patents)

slashgear.com

70 pointsgioele30 comments

I don't see a scenario where Microsoft will actually use these patents against Firefox or Chrome. They're doing everything they can to curry favor with web developers, and they'll destroy it in one fell swoop if they even think about it.

That being said, I'm curious why they felt these patents were worth $1B. There must be something really special in there.

(Disclaimer: Mozilla employee with no knowledge outside of what's been posted on the Internet about it.)

Back in the sands of time when Netscape begat Mozilla Foundation and released their source, they did so under the MPL, a license specifically written to grant patent rights where required to operate the software. I'm not sure how to interpret that, but I suspect Firefox would fare better in any war than Chrome or Safari.

Safari will do fine. Microsoft and Apple have a broad patent cross-licensing agreement. Chrome, on the other hand...

Motorola Mobility wants 4bn for patents relating to XBox and Win7 - 1bn for a better bargaining position could be worth it.

Is not like Google uses SSL, javascript, cookies or anything like that :)

Apparently Netscape patented them all once upon a time.

There was also mention of patents for messaging. Microsoft outright bought Skype, wich was bound to have patents in that area as well. Adding these patents to their portfolio must be helpful knowing some of their competition (Google, Apple).

> They're doing everything they can to curry favor with web developers

Only because they need it. If they can use those patents to achieve the same effect being nice would, I have no doubt they'll use force. As far as I know them, being nice isn't in their DNA.

In point of fact, they couldn't if they wanted to.

You see, once you have owned a patent, no one can use that particular patent against you. Ever.

So Firefox is safe, at least from those patents.

I'd love to believe that, but they paid a billion dollars for them, it's hard to imagine a public company shelling out that cash unless they expected to make back that money somehow...

Google can sue Bing to the ground, it's a cold war anyway. They will never go directly against each other. There are just proxy wars.

Maybe they did it for the same reason Google bought Motorola, to prevent a havoc AOL.

I can't imagine MS losing any goodwill they have gained in the past few years by attacking Firefox. Then we're talking about anti-trust all over again.

Chrome is another story, Motorola wants to ban Windows, XBox and other sales all over EU.

http://news.cnet.com/8301-13506_3-57407991-17/microsoft-ends... "on a typical $1,000 laptop, Microsoft must pay Motorola $22.50 in royalties for 50 patents related to H.264. In order to get full H.264 functionality, Microsoft must also license 2,300 other patents from 29 different companies. The company pays 2 cents for each $1,000 laptop to use those patents."

They even refused the bond MS was willing to post while the matter in on appeal. Looks like MS bought specially designed nukes to use against a particular and well-armed adversary.

With all the patent lawsuits flying around among the big tech companies, stockpiling patents seems like a prudent strategy from a corporate risk POV. I have no idea which patents are included in the list, nor do I know what MSFT would do with them. Just generally speaking.

SSL, cookies + Javascript are/were Netscape patents.

Better article about it: http://www.zdnet.com/blog/networking/microsoft-buys-netscape...

SSL http://news.cnet.com/Netscape-patents-crypto-protocol/2100-1...

cookies http://worldwide.espacenet.com/publicationDetails/biblio?CC=...

Javascript http://www.computerworld.com.au/article/255293/-z_programmin...

Sweet. They only expire in 3 years or so. Brilliant move on AOL's part.

I suppose it's only a matter of time until they do something obnoxious with these patents. Who will be their first target: Firefox or Chrome?

As long as Firefox's code heritage can be traced back to code released by Netscape under GPL (or Apache), Firefox will be fine. Chrome, OTOH, may have some problems.

Also, it's unlikely Chrome will be Microsoft's target - it's more likely Android will. Chrome is much less of a threat to Microsoft's future and Android includes a WebKit-based browser.

As long as Firefox's code heritage can be traced back to code released by Netscape under GPL (or Apache), Firefox will be fine.

Regardless the license, aren't they still vulnerable to claims of patent violation?

Some licenses explicitly state that contributors license relevant software patents that they control. Whether GPL 2.1 includes a patent grant is unclear, but the Mozilla Public License certainly does.

>As long as Firefox's code heritage can be traced back to code released by Netscape under GPL (or Apache), Firefox will be fine.

Does the GPL contain a patent grant and/or indemnification clause? I think only the GPL v3 does which wasn't around then.

As was discussed a hundred times, GPLv2 includes an implicit patent license (because it's impossible to grant the required rights without granting the usage of the patented technology and not granting those rights is a license violation).

An explicit grant is much better than an implicit one (and is one reason why the university where I did my grad work refused to let us use the GPL).

Besides, wasn't the Netscape code released under the NPL/MPL, partially for this reason?

> wasn't the Netscape code released under the NPL/MPL, partially for this reason?

Exactly. When I wrote the original comment, I didn't remember Netscape invented its own license.

There are two potential problems with this.

1. Given an implicit patent license between A and B, and a subsequent sale of A's patents to C, is there an implicit license between C and B?

It's certainly not obvious that there would be such a license, as the basis of the original implied license between A and B is A's acting in a way that only obeys licenses A is subject to if A licenses out A's patents. C would not be in that position, and so it is harder to find implied licensing by C.

2. If there is an implicit license between C and B, what is its scope?

It is quite possible that such a license would be limited to existing products that are practicing the patents.

If I give you an irrevocable right to use my summer home and then I sell it to someone else, I must disclose to the new buyer the fact you have the right I gave you. The rights granted by the GPLv2 are similarly irrevocable - when Oracle acquired Sun and MySQL, our ability to continue forking and redistributing the GPL-licensed code was never in question.

Quite relevant, BTW, is that Oracle continued to distribute MySQL's code under the GPL, and so Oracle also granted use of every patent that covers the code they acquired. Since the grant extends to derived versions (or it wouldn't be possible to license them similarly), it also covers forks released under GPL.

GPL is, perhaps, Stallman's most brilliant hack.

(Upvoting you to try to cancel out the idiotic down vote someone gave you)

You can't really make analogies to land, because estates in land are special in law. (I'm going to discuss things from the point of view of US law. This will be somewhat applicable in other Common Law countries, although details vary, and the devil is in the details. Someone else will have to cover Civil Law countries).

A patent license would be covered under contract, not under estates in land. Whether a patent license is binding against future patent owners depends on the nature of the license. If it is an actual license, then there is a good chance it will bind future patent owners. If, however, it is characterized as a covenant not to sue, then there is a good chance that it will NOT bind future patent owners.

There are no hard and fast rules here--the case law is murky, with cases both supporting and contradicting nearly everything I said in the previous paragraph. Non-exclusive licenses are more likely to be characterized as covenants not to sue. (And that's explicit non-exclusive licenses--we are on even shakier ground with an implied non-exclusive license).

As you note, it is relevant that Oracle continues to distribute the GPL code they bought. This means that the conditions that might imply an implicit patent license in GPLv2 apply to them. The case that is interesting, and much more uncertain, is where party A releases GPLv2 code covered by a patent they control, and later they sell the patent to party B, and B does NOT distribute any GPLv2 products that use the patent, and so the only way to find that B is providing an implicit patent license is to find that they are bound by such licenses that A granted.

In general, it is simply bad practice to rely on implied licenses. Even if a court finds that they survive a transfer, there is a serious risk that the scope of the transferred license will be limited to those who already had a license at the time of the transfer.

A good software license, free or non-free, should be explicit about patents.

You can't sell a right you no longer have. At least you shouldn't.

GPLv2 includes an implicit patent license

Has been tested in court? I think the answer is no. The logic Stallman brings up seems valid, though.

I think it wasn't, but I think the reasoning is strong enough to prevent it form ever being tested. It's easier to settle.

The Netscape Public License contains a patent grant, AIUI.

Won't be either, watch for Moto and every other device manufacturer (save Apple who has a cross licensing agreement) that ships a browser on a phone to have to pay $22.50 to MSFT.

Reply on news.ycombinator.com