Saturday September 30, 2006 Browsers of Power?
Now here's interesting. I've just been reviewing my log files and for once I looked at the browser version list. No surprise to find that two-thirds of my visitors are using Mozilla browsers, but one entry caught my eye because I'd not seen it before. It said "NSA Power-Browser,gzip". I have spotted "gzip" as a browser before but nothing that said "NSA".
I've done the obvious and googled for obvious hits - nothing. Is someone playing a prank, or should I expect a special interview next time I pass through a US port of entry?
Posted by webmink
( Sep 30 2006, 07:57:52 AM PDT )
Permalink
Comments [3]
links for 2006-09-30
-
52,000 processor super-computer in Dell's home town is built from Sun servers using AMD Opteron chips. I gather this will be one of the most powerful systems in the world. Wow!
-
.. until now! The IBM software patent on LZW expires this weekend, so developers no longer have to worry about it and can start using LZW compression and the formats they require freely without fear of being bullied.
Posted by webmink
( Sep 30 2006, 05:19:52 AM PDT )
Permalink
Thursday September 28, 2006 Startup Camp
You may recall that I recently wrote about attending FOO Camp, the archetypal "unconference" that Tim O'Reilly runs each year to cultivate the community that powers his publishing business. An unconference is a gathering of the sort of people you'd find at a conference doing the sorts of things people at conferences do when they aren't in sessions. I wrote:
Unconferences are less about deconstructing authority and more about expressing the connectedness that comes about when "the network is the computer" express itself when people come together too.I'm not saying the more structure sessions weren't valuable ... But the world will be changed by the relationships formed at FOO this year, not by the sessions that were presented. That's what good unconferences do - rescue us from content and return us to the power of relationships.
So I'm delighted to say that Sun is supporting an interesting new unconference that's being organised by the people behind Mashup Camp, David Berlind and Doug Gold. It's Startup Camp, and registration is open now. The event will be held at the Computer History Museum in Mountain View, California, on November 2nd and 3rd 2006. I hope I'll be able to find a way to be there, not least becuase I want an open source startup to win the Best Startup Contest. And yes, Robert, you are welcome at this one!
Posted by webmink
( Sep 28 2006, 06:16:10 AM PDT )
Permalink
links for 2006-09-28
-
Linus' antipathy for the FSF is clear in this lucid e-mail, where he makes clear he has no time for "Free software".
-
"Free Speech" - for small values of "free"? What next, confiscation of books that indicate you oppose the regime?
-
"Free Speech" - for small values of "free". Is there a theme developing here?
Posted by webmink
( Sep 28 2006, 05:22:41 AM PDT )
Permalink
Wednesday September 27, 2006 The Jurisdiction Paradox
I've been getting requests from a few people to remove the "choice of venue" clause from the CDDL. I'd like to discuss why I would rather not do that if I can avoid it. This is naturally my own opinion and intended as legal advice to no-one.
I trained as an electronic engineer in England, many years ago. Through one of those strange quirks of fate, when on graduation I applied for my first job the application was received by a software engineering manager rather than a hardware guy as I had anticipated. I thus ended up learning how to boot a mainframe using data-bus switches on the front panel and gradually became expert in the grey area between what's clearly hardware and what's clearly software - operating systems and data communications. It's therefore spectacularly ironic that these days I spend much of my life having to understand and manage issues related to software licensing law.
Keeping in mind that I'm not only not a lawyer, I have never even been stunt double for an actor playing one on TV, I feel forced to comment on stuff I've heard about "choice of venue" clauses in open source software licenses. In particular there are folk who think that these clauses take away their freedoms related to software. The rationale when applied to the Mozilla license goes something like:
- The license says litigation has to be in California:
any litigation relating to this License shall be subject to the jurisdiction of the Federal Courts of the Northern District of California, with venue lying in Santa Clara County, California
- But I don't live in California
- If I get sued by $CORPORATION I will have to travel there at my expense so their lawyers can roast me
- That's not fair, they are bigger than me, why should they be allowed to bully me just because I want to use this software
- That clause is therefore an abridgement of my freedom
That sort of logic is pretty easy to follow and seems compelling. However, it is wrong - the clause is highly unlikely to affect an individual developer like this. Here's why.
- First let's discuss jurisdiction and venue. Jurisdiction is a question of whether a court has the power to hear the claim. Jurisdiction must be had over both the subject matter and the parties involved. Once the parties have determined jurisdiction, then the question of venue comes forward. Venue is merely a choice of where, geographically, the case should be heard.
- When two companies that both trade in more than one jurisdiction and have more than one jurisdiction in common engage in litigation, before they can start they have to agree which jurisdiction and then which court within that jurisdiction will hear the case. They may well engage in a pre-lawsuit lawsuit to set the venue (aka "motion practice", apparently). This is a costly and uncertain exercise that can bias a case unfairly.
- A clause like the one I cite above, described as a "choice of venue" clause but actually selecting both jurisdiction and venue, allows a licensor to pre-agree which venue will be used in the case described above. It's pretty reasonable for two businesses agreeing a contract to make any future litigation faster by this sort of pre-agreement, apparently. It's like a default value in the case evaluation of the variable is required.
- If, however, you are an individual, or a company that trades in only one of the places in which the other party also trades, the only jurisdiction that can hear the case is the one you share in common with the other party. In this circumstance, the "choice of venue" clause has no effect - no reasonable court would hear a case involving a party with no connection to the court. The idea of corporations doing the equivalent of extraordinary rendition on the strength of a choice of venue clause is a literalist fantasy.
The main beneficiary of a "choice of venue" clause in an open source license is actually the smaller US business. Despite notionally trading in several states, a smaller business will only want to engage in litigation against a larger company in their home state. Without "choice of venue" clauses, the larger corporation may well be unfairly advantaged over the smaller company. Thus, paradoxically, the desire to be rid of choice of venue clause out of a sense of loss of freedom actually represents siding with large corporations over the smaller guy when he's the copyright owner.
So that's why I am in favour of "choice of venue" clauses in open source licenses (at least the ones that use contract law). I'm still open to discussion on it in licenses I can affect, like CDDL, where I notice the useful clarifying language about needing to be a US citizen that Mozilla has in section 11 was removed rather than reworked (probably for clarity because it's unnecessary in lawyer eyes).
That may have to go back in, but before I'll agree the clause should go altogether I'll need to hear why reducing the freedoms of small businesses like this is justified for the sake of an argument that revolves around a false premise about individuals having to travel elsewhere to defend a lawsuit.
Posted by webmink
( Sep 27 2006, 07:44:49 PM PDT )
Permalink
Comments [1]
links for 2006-09-27
-
Scoble's not using the name I suggested (ScobleVision) but the opening podcast is worth watching - not least for me to see Jonathan explain what I do at Sun.
-
Two days in a row and Dalibor has one of the day's best posts, this time about GPL v3, Apache compatibility and Linux.
Posted by webmink
( Sep 27 2006, 05:23:49 AM PDT )
Permalink
Tuesday September 26, 2006 links for 2006-09-26
-
This will be worth watching - a group blog by the OpenOffice.org engineering team at Sun. From the tone of the first few entries they obviously mean business.
-
"If you've ever wondered where all your disk space has gone, Disk Inventory X will help you to answer this question." I've been using jDiskReport but this is so much faster.
-
Dalibor Topic on the GPL v3 "debate".
-
This is useful, although it would be more useful as part of the conversation (with quotes from the Kernel maintainers' document, and even links to it) rather than as a politburo statement. People have to remember we're on draft 2 of 5...
Posted by webmink
( Sep 26 2006, 05:20:24 AM PDT )
Permalink
Monday September 25, 2006 links for 2006-09-25
-
This looks great! I have always wanted a boat like this one, for messing about on the canals, rivers and navigations in the south of England. I can imagine shuttling to and from Winchester in this on a summer weekend.
-
There's a new Pratchett out - we picked up a copy locally on Friday and it's already been devoured by two family members...
-
The latest MAKE Magazine has made it to the UK. Should have picked one up at FOO I suppose.
-
Every science geek's dream girlfriend.
-
"Instead of spending nearly endless programmer time locking out users, Sony should focus work on a more positive solution ... and encourage legal uses of homebrew software." Most vendors would kill for customers this passionate.
Posted by webmink
( Sep 25 2006, 05:25:34 AM PDT )
Permalink
Comments [1]
Sunday September 24, 2006 Engaging Positively On GPL v3
If you've not been reading the comments under my daily links, you will have missed out on some excellent stuff relating to the new GNU General Public Licence proposals, GPL v3. Just to catch up on the story so far, Richard Stallman and Eben Moglen are in the middle of a proposed major revision of the GPL. To see where things have reached, visit the FSF GPL v3 web site - there's an awesome AJAX-powered review version of the license there, for example. That review version allows readers to add comments, and quickly scrolling through will show you where the main areas of controversy are found (they show up as "hot spots). They include accusations of political motives (duh), language that deals with DRM, language to defuse software patents and language to affect those who deploy modified code without distributing it. If you have the patience to wait for the Draft 1 comment page to load, you'll see these have been the controversial topics all along.
Various media pundits worked out a while back that there was likely to be a problem with the Linux kernel and GPL v3. It was inevitable, both because Linus Torvalds had expressed pragmatic contentment with GPL v2 long ago (and had ensured all the file headers for the kernel were licensed under that version alone and not the FSF-recommend "or any later version") and because the copyrights to the various components of the kernel belong only to their individual authors and are not aggregated anywhere, thus making a license change an arduous administrative task. Contacting all the contributors to allow a relatively uncontroversial license change took the Mozilla Foundation ages, so the task for the Linux kernel will be sisyphean should it ever happen. That latter fact means the kernel authors have little to lose by being negative about v3. And negative they have been. They just published an agreed position statement on GPL v3 which, despite an emphasis throughout on all the issues I know independently that their employers worry about, they assert was made independently of their various employment. That brings us up to date.
I linked to that statement a few days back and got a reply from Luis Villa, of GNOME fame, as well as from Dalibor Topic. Luis has written two exceptionally good articles. The posting to read first, What the kernel guys got wrong, looks at the statement itself and finds most of the statements in it wanting. Read the comments too. The other post, What the kernel guys are and aren’t (and really should be) saying about GPL v3 digs under the surface and exposes the social dynamics of the complaints. Very insightful analysis, no less than I would expect from Luis who is a very sharp cookie indeed.
I've been following the GPL v3 process since it started, and despite a certain amount of initial scepticism I'm pretty impressed. Eben Moglen in particular has shown himself to be a statesman and a scholar and the GPL v3 review process is no hollow plebiscite. The FSF team has engaged, listened and responded to all the comments they have received, and the draft 2 text shows great improvement over the draft 1 text, especially concerning the stance on DRM.
I have a growing confidence that what will appear from the process after another 3 drafts could well form the basis of a unification of the Free and Open Source software communities. The compatibility mechanism created by section 7 is brilliant, and what was initially cold-war-era posturing on patents and DRM is evolving well into balanced approaches to handling both issues. There's certainly more progress needed - the patent language is still too imbalanced against large portfolio holders, for example. But the track record to date shows that the goal is reachable if everyone engages positively.
Which of course raises the question: if I can see that, if others like Luis, Dalibor and Mark can see it, why can't the kernel developers see it?
Update: One more post on the subject from Luis, What FSF Got Wrong, has an extremely interesting comment from an OSDL employee that says "The kernel developers did not directly particpate in the GPLv3 review process because it was clear from the beginning that any direct comments would not cause change." Nice excuse.
Posted by webmink
( Sep 24 2006, 04:25:18 PM PDT )
Permalink
links for 2006-09-24
-
Hem has a new, simply sublime album.
-
I'm blown away by these recordings, dating back to the late 19th century.
Posted by webmink
( Sep 24 2006, 05:20:10 AM PDT )
Permalink
Saturday September 23, 2006 links for 2006-09-23
-
This seems a rock-throwing exercise and spends a remarkable amount of time defending software patent holders and DRM implementers.
-
Just the book for you if you're considering trying OpenOffice.org
-
Hey, I like this one. All four of the free tracks are good but the first ("Blue Light Room") especially so.
Posted by webmink
( Sep 23 2006, 05:18:50 AM PDT )
Permalink
Comments [2]
Friday September 22, 2006 links for 2006-09-22
-
The Zune won't play them. Now v11 makes sure they are almost worthless in the place you still can play them. I do hope you didn't invest too much in that music collection, it just became worthless.
-
The details from the developers involved. This looks like it could be a huge step forward, both because of the ease of package and extension delivery and because of the introduction of online updates.
Posted by webmink
( Sep 22 2006, 05:23:14 AM PDT )
Permalink
Thursday September 21, 2006 links for 2006-09-21
-
JetBrains developer points out the sheer idiocy of eWeeks' article and the lameness of Mike's sniping. Slow news day or something more sinister? Whatever it is, the article is worthless.
Posted by webmink
( Sep 21 2006, 05:22:16 AM PDT )
Permalink
Wednesday September 20, 2006 links for 2006-09-20
-
Pandora is really very good, give it a try.
-
"... scores of threats are far more likely to kill an American than any terrorist -- at least, statistically speaking."
-
This is pretty cool. They are giving a free sample pack, including free shipping, to Flickr Pro users, I just ordered some.
-
So if you have been unlucky (or foolish) enough to build a library of WMA files, you are out of luck (and probably can't even burn s CD from them). Maybe they should rename their program "Plays For Now"?
-
Sun's been working with the University of Kent on this. BlueJ has been around since the early days of Java, and NetBeans true platform independence makes it a great future vehicle.
Posted by webmink
( Sep 20 2006, 05:20:13 AM PDT )
Permalink
