directory-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Roy T. Fielding <field...@gbiv.com>
Subject Re: [VOTE] Directory project releases
Date Thu, 30 Dec 2004 23:10:01 GMT
On Dec 27, 2004, at 8:16 PM, Alex Karasulu wrote:
> Before discussing each specific item below I'd like to point out a few 
> things.  The ASF had through the incubator given us a link to an 
> online service for looking up trademarks.  We were told this service 
> should be used to determine if there were any infringements with 
> project names.  I personally used the service for two of the names 
> cited in the items below.

We normally start with that and also do a Google search.

> According to the ASF the existance of any trademark alone was not an 
> issue.  For there to be an infringement the trademark would have to 
> fall within the same field and domain where the names also match 
> exactly.  In the case of the Eve Directory Server for example, the 
> presence of a trademark for "Eve Online", an online game would not be 
> an infringement.  I understand your concern but this is not the same 
> name. BTW note that "Eve Online" did not exist in 2002 when we first 
> started using the name Eve Directory Server for this project.

Apparently, the same company has a trademark on "EVE" in Europe and 
they have
already used it to force a hardware company in the UK to rename their 
product.
Here is their Icelandic registration
     http://www.els.stjr.is/focal/webguard.nsf/key2/br103-2001.html

> From the language and terms on the incubator site I was completely 
> under the impression that this kind of name use would not present any 
> legal issues.

Goal #1 is "do the right thing".  Goal #2 is "protect our assets".

>>   1) Has Matt Walsh approved transfer of the SEDA name?  Is that 
>> library
>>      a direct descendant of <http://sourceforge.net/projects/seda/>, 
>> in
>>      which case the version numbering should be 4.0, or does it 
>> merely  use
>>      his architecture (in which case it is wrong to use the name)?
>>
> No there is no relation between our seda package and the SEDA project 
> on Sourceforge.  We just named this thing, seda because we had no name 
> for it at the time.  SEDA is what it is as far as the architecture 
> used.  It's like naming a cat, cat.  I think the SEDA project on 
> Sourceforge actually calls the respective code with the same fuction 
> Sandstorm.  SEDA is an acronym used for Staged Event Driven 
> Architecture as you probably already know.  It was part of Matt 
> Welsh's doctoral thesis and hence the name and the technology is not 
> something that can hold a trademark.  At least this is my 
> understanding - perhaps this can be clarified by legal.

Sorry, software products are not named "product" -- they are given 
names to
distinguish them from other software.  We do not use other people's 
existing
names without their permission because it would confuse everyone, 
including me.
[We learned that the hard way with httpd, even though we had 
permission.]

Matt's toolkit may be code-named Sandstorm, but it is distributed as 
seda on
Sourceforge.  You should name it something like "sedakit" (which 
probably means
something in another language) or "apseda" (which may be Latin, for all 
I know).

> Furthermore the concept of Threads for example was invented by some OS 
> manufacturer but there are threads implementations all over like 
> pthreads.  When does it become an infringement?  If we call a SEDA 
> implementation seda?  See what I mean? Its kind of nebulous here.  I'd 
> like some clarification myself. Regardless I'm totally fine with 
> changing the name here.

I doubt that Matt trademarked the name -- we should simply "do the 
right thing"
and avoid confusion.

>>   2) Eve is a trademark of CCP Ltd. for use in on-line software.  We  
>> cannot
>>      legally distribute software under that name.
>
>
> At a closer glance I noticed the trademark is "Eve Online" and not 
> really Eve.  "Eve Directory Server" is not the same name and it does 
> not infringe on this trademark.  I do understand your concern though 
> but where do we draw the line.  Also for the record we had the name 
> Eve before they did.  Sounds silly but true.  We've been under 
> incubation for a long time now 14 months and when I checked in the 
> begining they where not even on the map.

Not according to their registration in EU -- it is 2001.

>>   4) Snickers is a trademark of MARS, Inc.  It is actively defended 
>> and
>>      used on-line, and thus unlikely to be usable by the ASF for more 
>>  than
>>      a few months before they send us a Cease & Desist.
>
> Snickers is an ASN.1 software library which is a Snacc4J replacement.  
> It's obviously not a candy bar.  Do they really have the right to 
> enforce a Cease and Desist?  This is definately trademarked by MARS as 
> you say no doubt about that.  We need to rename this as well but just 
> for clarification how do companies like this get away with it?

They don't need the "right" -- they merely need to avoid a countersuit 
for
obstruction.  Because it is a well-established mark and that mark is 
used on the
Internet, they can reasonably constrain use of it on websites.  Since 
the
name is being used because the real Snickers is a snack, we are 
actually using
the trademarked name (not just any name) and we would lose.

In general, a good rule of thumb is to walk gingerly around companies 
who
employ teams of lawyers for the purpose of protecting their marks.  It 
simply
doesn't matter why we *might* win such a suit, since the suit alone 
would be
enough to bury our little foundation in debt.

>>   5) Kerberos is a trademark of MIT.  We cannot legally distribute  
>> software
>>      under that name.
>
> No trademark is registered with Kerberos as far as I can see.  Also 
> looks like Enrique performed some similar searches and found:
>
> "The USPTO Trademark Electronic Search System shows efforts by MIT to 
> trademark Kerberos in the early 90's were abandoned.  Published 
> articles up until the late 90's may claim a trademark by MIT, but this 
> is not the case. "

Okay, sorry that I didn't see the abandoned bit.  Once the STATUS file 
is updated
(be sure to update the table entries for the IP stuff), I have no 
objection to
releasing the non-trademarked name products.

Personally, I think it would be better if you simply released a single 
alpha
product (like ldapd) and used experience from that to improve the 
individual
libraries, but I assume you have reasons for releasing independent 
libraries
right now.  Just be sure that each released product has three +1s 
backing it up.

....Roy


Mime
View raw message