Fog Creek Software
Discussion Board




"SQL Server Developers Face Huge Royalties?"

I saw this article on Slashdot yesterday:

http://slashdot.org/article.pl?sid=03/02/21/1323237&mode=nested&tid=155

Apparently some proprietary, patented algorithms are embedded in SQL Server.


I chased down the link to the company that claims the patent, and the language was utterly indepherable to me. Message traffic on Slashdot (all legal experts, natch :-) ) seemed to indicate that only users that use data warehousing features are liable.


I think this is a real can of worms because it's saying that the purchaser of a shrink wrapped software product faces potentially unlimited liability due to the vendor's unauthorized use of technologies from other parties.


Somehow, this seems very wrong and severely lopsided, since the buyer is relying on a vendor's reputation as an assurance that the product is bona fide and most users don't have the means to reverse engineer or do patent or copyright searches on products they propose to buy in order to determine their own potential liability.


Does anyone have an opinion or facts on this?

Bored Bystander
Saturday, February 22, 2003

I seriously doubt it.  A large company like MSFT has many tools at its disposal to resolve this.  It is common practice for a supplier of potentially patented technology to indemnify its customers (including developers) from problems exactly like this.

Expect MSFT to step up to the plate in timely fashion.  They will make the problem go away - period.

Nat Ersoz
Saturday, February 22, 2003

Look at the press release put out by Timeline - Microsoft could have bought the rights so it's customers wouldn't have to pay royalties, but they turned down the offer.

Microsoft seems to be trying to throw around its weight in this matter - including tying it uop in the court system.

RocketJeff
Saturday, February 22, 2003

Nat, it seems that MSFT _had_ the chance to indemnify it's customers against liability and flatly turned it down. In fact, the court proceedings seem to indicate that MSFT was aware of the liability it was exposing it's customers to and yet assured them that it didn't exist. If MSFT is going to do right by customers, it's most likey going to come in the form of customer's suing for compensation for damages related to false representation once Timeline takes them to task.

I doubt very much it's something that MSFT is going to do on their own.

--
Alex Russell
alex@netWindows.org
http://www.netWindows.org

Alex Russell
Saturday, February 22, 2003

Most software licenses explicitly do not limit liability or hold harmless for copyright or patent infringement.  That's why the large financial services company that I work for always makes vendors add this to a contract.  Otherwise, some little company could sue us for infringing based on use of Microsoft's technology just because we have deep pockets. 

Mike S.
Saturday, February 22, 2003

If MS doesn't step up to the plate.  Their credibility will plummet.

Although as of late they haven't been hitting them out of the park, so they could strike out.

Cedric
Saturday, February 22, 2003

Exactly.  Something/somebody ran off ni the weeds on this one, misjudged the severity...  They will reel it in, or lose more busniess to the open source world.  And this is a fundamental they cannot ignore.  They will get it together (not that I'm pulling for them or anything, I'm not.  But I know behavior, they won't screw this up - too much).

Nat Ersoz
Sunday, February 23, 2003

Except (at least according to excerpts on /.) this only is an issue for a certain small class of SQL server developers, those who develop some particular form of tool (I forget what). Not regular database development.

Of course this is also a typical large corp strategy: take someone's 'IP' and threaten them if they don't license it to you for free. That's what patent cross licensing is all about.

mb
Sunday, February 23, 2003

I'm still foggy on the whole "users are liable" thing - does anyone know if that has *ever* happened before? It makes no sense to me whatsoever - am I now responsible for researching patents on every tool in my workshop and every piece of software I buy? Do I have to pay an attorney to do a patent search on "creating animations with predefined shapes using a timeline" before buying Flash?

I think it's an exceptionally bad holding from the judge, and should be overturned on appeal. Microsoft should be liable for the infringement of incorporating the tech as well as the subsequent infringement by those who bought the product - indemnification on a shrink-wrapped package should be a nonwaiverable warranty.

Philo

Philip Janus
Sunday, February 23, 2003

"I'm still foggy on the whole "users are liable" thing - does anyone know if that has *ever* happened before? It makes no sense to me whatsoever - am I now responsible for researching patents on every tool in my workshop and every piece of software I buy? Do I have to pay an attorney to do a patent search on "creating animations with predefined shapes using a timeline" before buying Flash?"

And now you realise that this is a requirement, you should understand why software patents are a bad idea. It should also be obvious why many companies like them - this will make a lot of money for Timeline.

(While there is an allegation that Microsoft mislead its customers as to their obligations, the end result is that you need to pay a patent lawyer to investigate all patents related to every aspect of the software you use or create. While Microsoft made claims that all patent issues had been dealt with, it appears they either lied or were mistaken.)

The thing that amuses me the most is that, apparantly, people who listened to Microsoft's claim that SQL Server developers didn't owe money to Timeline may suffer larger penalties than people who actually did consult a lawyer, even if they came to the same conclusion. Acting in "good faith" means "paying a lawyer". Without the lawyers receipt, you're screwed.

Btw, the GIF patent issue a few years back was similar: the GIF format (or at least part of it) was patented, but not widely known to be patented. As a result, large numbers of people had an unexpected bill, because they had failed to do an exhaustive patent search. You should, obviously, do a patent search on every algorithm you use or create, because there's no telling who already own it.

andrew m
Sunday, February 23, 2003

"You should, obviously, do a patent search on every algorithm you use or create, because there's no telling who already own it."

I wonder if anyone holds a patent for quicksort ...

"... the State of Washington finds the defendant Alyosha` in violation of patent #75162736, "Iterative Manipulation of Array Members Using An Incrementing Loop Variable" ..."

*sigh*.  Weren't patents meant to ENCOURAGE innovation?

Alyosha`
Sunday, February 23, 2003

"Btw, the GIF patent issue a few years back was similar: the GIF format (or at least part of it) was patented, but not widely known to be patented. As a result, large numbers of people had an unexpected bill, because they had failed to do an exhaustive patent search. You should, obviously, do a patent search on every algorithm you use or create, because there's no telling who already own it. "

But you see - those who used the LZW algorithm had a reason to research it (BTW, I opposed Unisys' "wait and see" approach - the court should've disallowed the case because they just sat on their rights so long). Nevertheless, the software developer, being so much closer to the problem, can be expected to be more aware of the legal issues.

And I don't believe at any time that the *purchaser* of a GIF program/toolkit/etc was liable for the programmer's infringement.

But when you buy a shrink-wrapped box off the shelf, by all means you should be indemnified from any potential infringements stemming from the contents of the box.

I'd like to stop that judge on the street and ask if he's done a copyright search on the patterns on his clothing, or should he be liable for infringement everytime someone takes a photo of him with that pattern in it?

Philo

Philo
Sunday, February 23, 2003

<quote>
And I don't believe at any time that the *purchaser* of a GIF program/toolkit/etc was liable for the programmer's infringement.
</quote>

You're wrong. If your web site contains 1 single gif created with software that has no license from Unisys you have to pay them $5000.

http://www.unisys.com/about__unisys/lzw/lzw__license__english.htm

let's use another name this time
Sunday, February 23, 2003

One thing I will gurantee. 

There will be meetings a happening this week in Redmond!

Cedric
Sunday, February 23, 2003

"You're wrong. If your web site contains 1 single gif created with software that has no license from Unisys you have to pay them $5000."

well:
[Many Web site operators use commercially available software which creates GIF images offline which are then posted on their Web sites. Since most of this commercially available software is under license from Unisys for their use of the LZW patent, users of this software are probably covered as well for this use of GIF images on their Web sites. Questions about the exact terms of the license for a particular software product should be addressed to the provider of the software.]

I actually read that as a license available for webmasters who are running software that generates dynamic GIFs. This isn't Unisys going after anyone who's got a GIF on their site, this is Unisys offering a simplified pricing model in apparent reply to a lot of inquiries and complaints.

Once again, I think the MS holding causes a huge problem - if it is enforced, it could effectively end the COTS software market tomorrow. Not the result I suspect we're looking for here.  I would think the preferred result (if supported by the legislation) is that if MS doesn't buy the appropriate patent implementation licenses, then they're issued an injunction to stop selling the product.

Philo

Philip Janus
Monday, February 24, 2003

It threatens the closed source model.  If you can't see the source code, how are supposed to know what patents it's infringing?

Ged Byrne
Monday, February 24, 2003

I think there's some confusion about what a patent infringement can and can't be.

If we take an example from the physical world which is what the patent system was originally intended to protect, it might help to make it clearer.

Suppose Inventor A patents a method for cleaning clothes using a rotary crank with water flowing around the inner surface of a cylinder the clothes being moved in a random manner with a central vaned spindle.

Manufacturer B designs and builds a washing machine which relies upon the patent of A (whether intentionally or not).  B sells the machine generally to the populace.

User C has a washing machine sold by A and washes their clothes by it.

Inventor A discovers this and successfully sues manufacturer A.  User C continues to wash their clothes regardless, they haven't breached the patent and they aren't using the patent.

Unless this software patent and any application that implements it involves the final user in also using the process described to create another object or process the final user is not in breach of that patent.

This was where the GIF patent problem ended up, almost all users that distributed GIF files did so using another application that created the file in that format.  The user was not in breach, creators of the application were.

Even if the creation of a GIF file (or any comparable software process), was perfomed by someone writing code which called a library which performed the conversion and writing of the file, the use of the library is not the breach of patent.  The original writing of the library was, unless it was under licence. 

If the original licence restricts the licensee from passing on that right to others, the breach is the licensee's not the final user's.

A conceivable result from a patent suit could be the ordering of the destruction of all means of infringement and all objects created by that infringement.  But that destruction would be the responsibility of the infringer and the final user would have no obligation to do anything.

So, if this patented process embedded inside SQL Server doesn't require the final user/developer to recreate the process patented directly there is no breach by them and no exposure.

I of course speak authoritatively with no qualification to do so, Caveat Lector.

Simon Lucy
Monday, February 24, 2003

Ummm, User C obviously bought from B, not A

Simon Lucy
Monday, February 24, 2003

I like your ABC analogy so let's continue it ...

In this case C can be one of two things.  C could be an end user simply using the product (who has no problems for the very reasons you suggest) or C could be a developer (since it is a database) using the product to create an entirely seperate product which s/he would sell.  From my understanding (I am not a lawyer, etc)  it is these people that have the problem here.  The license for DTS and some of the data analysis services in SQL server does not allow for further development in this manner. 

So if C is washing clothes then s/he is doing what the product was designed to do, however if C is engineering a clothes washing process based on A's work, which was bundled for use but not for further innovation by B, then there are problems.

Since this product is sold mostly to innovators, MSFT probably should have been more clear about where the boundaries were, but I am not sure how much real trouble they are in.

Anyway, that is my understanding of the issue.

Ran Whittle
Monday, February 24, 2003

This stuff never goes anywhere.

pb
Monday, February 24, 2003

Ran Whittle:

Essentially yes.  It doesn't actually matter what the product was designed for, its whether the user infringes on the patent directly (knowingly or unknowingly) or not.

Simon Lucy
Monday, February 24, 2003

Its also true that no other third party was enjoined in this action and that each infringement would have to be proved individually by the patent owner.

If that infringement is obvious and the infringer has deep pockets then they may think it worthwhile.  But given that they signed a licencing deal with Microsoft for a product to be distributed they could (on an individual case basis) be considered to have known that its use would have been such and such and so that they had lost their patent by allowing the use.

Anyhow, reading the case there's more than a few incompetents involved in this.  It is of course pursuable only within the US jurisdiction, software patents are not recognised in Europe, for example.

Simon Lucy
Monday, February 24, 2003

*  Recent Topics

*  Fog Creek Home