Hacker News Reader

You are conflating Google's code ten years ago with Google's current codebase.

I think releasing 2002's Google open source would be far less revolutionary than you think. 2002 was, like, really long ago.

C++? C++ was still largely pre-Standard in practice. GCC 3.x (which was the first release line with good C++98 support) was still in its infancy, and most universities still had GCC 2.95 installed until 2004 or 2005. Java? Java 1.4 wasn't released until February 2002. Do you remember Java from those days? Casts. Two, three, four casts per line. Generics wouldn't be introduced until late 2004, at which point Java would become usable. Python 2.2 was all the rage in 2002. I don't know about you, but I cry every time I have to write a Python script which works with 2.4 because of a busted old server. Decorators weren't even a glint in Guido's eye back then. Source control meant CVS. CVS.

How about the Web? Well, IE6 was the new thing back in 2002, so there's that. Phoenix -- later renamed Firebird, later renamed Firefox -- was released late in 2002. Opera's leading feature was that it would reload your old tabs if you restarted it, which was great because it crashed every 15 minutes. Ajax was used by, like, two sites. Webmail meant Yahoo!. MySpace didn't even exist in 2002, let alone Facebook. People still used ICQ, although AIM was still more popular. KaZaA was still spelled with random capital letters in 2002, and people still used it. "Warcraft" meant the hot new game, Warcraft III.

Ten years is a really long time in internet time. I don't think Google or Fogcreek would be hurt in the least by releasing their 10-year-old codebases, because their 10-year-old codebases are damn near useless by modern standards.

I'm totally agree with you. I think that even 5-year is enough.

What harm for Id Software was in open-sourcing Doom, Quake 1-3 ?

Several years ago Win2K code was leaked - what harm for Microsoft was in it? I suppose zero.

> After ten years all code should be made open-source.

Long digression. Which may not even be totally true, but close enough to make the point.

This isn't actually that far from the original intention of patents. It used to be that things were kept as trade secrets and hence lost to the ages, which is why no one knows how to make Damascus steel anymore. On the other hand, if someone leaked your trade secret, you could sue them for breach of contract or something, but once the secret itself was out, it was just free speech and free enterprise for people who didn't sign an NDA with you to go ahead and use it. So trade secrets were both ineffective for the secret-holder (unless they kept them very, very well) and bad for society as a whole (since none of us could use any of the secrets that didn't leak out.)

So the patent was invented. Now you had a choice: either your trade secret could remain secret or it could become a patent. If you patented, for instance, your revolutionary manufacturing process for steel, no one in the entire country could legally use that process themselves--at least not until the patent term was up. On the other hand, the process itself was published, and once your patent expired, everyone gained your knowledge.

Copyright enters the picture with software in a funny way. If you documented the process of, say, manufacturing steel, the document itself would be copyrighted. Anyone could read that document and then implement the process with no problem. The same is true for, say, the plans to some machine or appliance--the plans themselves would be copyrighted, but that wouldn't prevent you from assembling an actual device. Obviously with software this is different.

I actually think that with software, both traditional patents and the copyright process don't really work. Rather, I think software should be subject to the same kind of social contract patents successfully imposed on pre-software inventions: no copyright protection on source or executable code (copyright on documentation or string or graphic resources are acceptable), source and executable code can be "patented", which entails a limited-term (5-10 year) monopoly on the use of that source code, but after 5-10 years the code is effectively public domain, and published in perpetuity by the government.

Just to explain my downvote: you're arguing that, when locked into thermonuclear war, one side should just unilaterally disarm. You obviously know this is absurd. Unless all parties agree to a compromise (or they are forced to by higher authorities), the first mover will always be a loser.

Oh, and the 10-years copyright is extreme, of course, but current rules are also extreme and need to go. That is Joel's point: in order to get a sane compromise, we must ask for an insane target.

It's funny, I never down-vote people even when I don't agree with them. But that's just me - I am anti-censorship.

Though, I do thank you for taking the time to make an argument - I get down-votes for almost every unpopular viewpoint I take here on HN; 90% of the time the down votes are anonymous.

In order to get a sane compromise one first has to appear credible. The pro-copyright folks have no legitimacy around here because they demonstrably don't understand how the tech world works; clearly that disconnect extends to both sides.

Just out of curiosity, do you work in the tech industry? I notice that 90% of your posts have to do with copyright, usually from a perspective that is very corporation-sympathetic.

Anyhow, to address your point: I would argue that the computer software field is light years ahead of any other enterprise in human history, in terms of free, open, and shared contributions toward the betterment of mankind. That's not to say that we're without flaws or total fuckups, but a lot of us have been practicing what we've been preaching for decades. Would you disagree? Given that the music and movie industries still operate like it's 1930, what's to say they would follow suit if we "went first"? Because I would argue that we already did, 30 years ago.

I'm a new arrival in the tech industry; used to work in the content industry. I have NO love for corporations - none at all - it's just that I have really gotten to know the people on both sides and find it frustrating that the fight ends up being about which sides corporations can win - not how we can all win. I might add, as an aside, that we've seen with Google's recent turn that no corporation can be trusted to 'not be evil'.

I would certainly agree that the people that make up the content side of the tech industry - programmers, devs, designers, etc . - really do take open and shared contribution seriously, to an impressive extent.

But that's true in Hollywood too - forget about the business structure - the real work happens across thousands of small disparate shops with groups of people working together on the things they love. This is true from development all the way to post-production. It's an open and creative environment.

I do see a clear way that copyright can be reformed to create a win-win scenario. It's a lengthy idea and if I can find the time to put it on paper, I'll happily post it here.

Fine, the copyright protection for Google's algorithm has evaporated. Where do you plan on getting a copy of it? And even if you did somehow acquire it, how do you plan on using it without violating Google's trade secret protection on it?

There are existing legal frameworks in place (trade secret, trademark, service marks, etc.) that complement copyright and would enable software (and movies!) to remain viable enterprises even with drastically shorter copyright terms (or no copyright at all.)

What's good for the goose is good for the gander. If my business and my copyright are the same thing, why should I be forced to 'open-source' my business after ten years?

Neither trade secret nor trademark nor service marks protect my creative work; only copyright. But my copyright is my business just as the source code is the foundation of Joel's business. So what Joel thinks is good for my business is surely acceptable for his own.

If anti-copyright folks are as concerned as they claim to be about adding value back into society - and as quickly as possible at that - it makes overwhelming sense that valuable services such as software should be scrutinized first, before mere entertainment. We should obviously be talking about limiting a term on trade secrets - not just copyright - people will still have trademarks and service marks to defend their reputation and authenticity.

>We should obviously be talking about limiting a term on trade secrets

That doesn't make any sense. Copyrights and trade secrets are two very different creatures. In order to benefit from a copyrighted work, the creator has to show the work, but this is not the case with trade secrets.

I honestly don't understand how one could even begin to compare the two. Copyrights and patents have to be revealed specifically in order to be granted protection, and you are certainly within your rights to keep your patented creations and copyrighted works to yourself.

The limiting of copyright and patent is supported by advocates of such limitations on the basis of benefits from the work accruing in a more fair balance between the individual and society. It is stated on this thread and many others that it is society that chooses to allow patent and copyright to exist exactly in order to bring forth the work of individuals. The good of society is placed above that of the creator of the work.

If one likes that idea, it works even better in the case of 'trade secrets'. Abolishing trade secrets and instead substituting a time limit like that of a patent would REALLY let those benefits flow to society at large. We force pharmaceutical firms to come up with new ideas every patent cycle; why should software - after all, it's 'eating the world' and is a highly-maleable product - be exempt from that process of creative renewal?

Case in point: Microsoft used it's dominance gained from Windows to directly stifle competition, to the point where the law had to get involved.

Google was a disruptive force initially, but now they have begun to tweak their algorithm to fortify the walled-garden they wish to build - while also using the Microsoft model of buying up potentially disruptive companies.

How exactly did you go from expiring copyright to open sourcing?

Expiring copyright is just that: you can copy the distributed result (binaries) as you want. It does not entail open sourcing any more than it would force the studios to release their raw footage and CGI files.

Oh another thing: very important:: a short term of copyright like a few years would be the biggest boon to Hollywood ever as they could simply sit and wait for works to drop into the public domain before turning around and producing them without paying the creators a penny. There would be tons of creators strung along via a studio option - just long enough till the work dropped into the public domain. It would harder than ever for individuals to profit from their creative work and easier than ever for Hollywood to make money off of it.

What creators, exactly, are you talking about? If you mean screenwriters, they could simply sign a simple contract ("You shall not produce a movie based on the work without authorization") before showing it to the studios. Why would they need copyright?

Expiring copyright and open-sourcing are not the same thing - but they can have the same result. That's a point I think tech-people don't think about enough. If my business is the creative work I've done, what right does an expiring copyright have to put me out of business? By placing my work in the public domain?

Open-sourcing code - though different from an expiring copyright - would create the same result in a business based on the ownership and public non-availability of that code.

Regarding the 'simple contract': contracts mean very little - leverage means everything. Granting very short-term copyrights removes all leverage from the owner of the work. His commercialization window grows very small and he is dependent on organized outside entities to 'make it happen' for him.

Patent coverage isn't even as short as a ten-year term mentioned in the OP. And patents are usually produced by it integrated firms that already have a huge commercialization apparatus running 24/7.

You might consider actually reading what he wrote instead of having an emotional reaction.

Just because a copyright has expired doesn't mean you have to actively distribute it