09:49 < rlpowell> I think I've figured out how to describe my dislike of the GPL in a more constructive way.
09:49 < rlpowell> (IIRC bancus, at least, was against me on this)
09:50 < rlpowell> It seems to me that there are two types of software: software that is appropriate for re-use in may general
                  situations, and software that isn't.
09:50 < rlpowell> It seems obviously valuable to me that as much of the former be as available as possible, but not so for the
                  latter.
09:51 < rlpowell> The former: math libraries, database code, memcached, that sort of thing.
09:51 < rlpowell> Although that again is divided into two categories: stuff you mostly want to access as an application (web
                  servers, memcached, mysql) and stuff you actually want to crack open and/or directly incorporate in your own code
                  (math libraries, GUI libraries, that sort of thing).
09:53 < rlpowell> The latter: a recipe tracking program; Company X's timesheet program; basically all computer games.
09:53 < rlpowell> I'm making a (very fuzzy) distinction between three (I thought it was 2; oh well) categories:
09:54 < rlpowell> Code that people are likely to want to re-use directly (by incorporating it into their own code); applications
                  that are of broad enough usefulness that having them be freely available is Really Good; everything else.
09:54 < rlpowell> Each category has fuzzy boundaries to the next category; Apache being open-source as well as just free is A Good
                  Thing because then people can write add-ons easily.
09:55 < rlpowell> My problem with the GPL is that it's designed to increase the size of the first two pools, but it does so at the
                  expense of the third pool.
09:55 < rlpowell> As an example:
09:56 < rlpowell> I'm working on a piece of code to simplify internal auditing at my company.
09:56 < rlpowell> At some point I discover that my script would be *far* easier if "tail" had Feature X.
09:56 < rlpowell> It is Good And Right that I be forced to give Feature X back to the community.
09:57 < rlpowell> It is Bad And Wrong that the community can assert ownership of my auditing software as a result, for two reasons:
09:57 < rlpowell> 1.  This gives me a strong motivation to not use GPL software, which means that this giant Category 1 pool is
                  useless to me, which *defeats the entire purpose*.
09:58 < rlpowell> 2.  No-one *actually wants* my auditing software.
09:58 < rlpowell> cizra: This is very true, but let's say that I make the modification by grabbing tail's code and inserting it
                  into my program.
09:59 < rlpowell> I should *still*, ethically, refactor the modification so it can be part of regular tail, if it's generally
                  useful.
09:59 < rlpowell> (regardless of tail's license, btw; I assert that adding to the pool of human-reusable source code is a general
                  ethical imperative)
10:01 < rlpowell> And then we get to the sticky point: I assert that being able to sell software that other people are unlikely to
                  want to re-use *as code* is Right And Good.
10:02 < rlpowell> That there is an ethical imperitive to add to the general code-reuse pool, but that *not all software belongs in
                  that pool*
10:02 < rlpowell> Not just because I made it and I want to make money from it, but because no-one *really* wants to re-use that
                  code.
10:03 < rlpowell> Yes, I'm aware that this POV cannot be made into a legal document.  :)
10:03 < rlpowell> But the GPL seems to me to be very much the wrong solution.
10:04 < rlpowell> It's always struck me as "Well, if you won't play my game, I'll take my ball and go home!".  As far as I know,
                  RMS is not *actually* a 6 year old, but the GPL sure sounds like it.
10:05 < rlpowell> End rant.  :)
10:05 < bancus> For me, releasing software under the GPL is a compensation thing.
10:05 < bancus> Yes, you can use my code, but I want the right to use your code in return.
10:05 < cizra> I would prefer BSD for my software.
10:06 < cizra> (not that I've written anything useful..)
10:06 < bancus> These days, that's basically what I do.
10:06 < bancus> But mostly what I right these days are quick hacks that I don't care so much about.
10:06 < rlpowell> bancus: I understand that; my point is that in many case, you won't ever *want* to use my code, so why should I
                  be forced to choose between giving it to you and not using your code?
10:07 < bancus> Because having the ability to use your code, whether or not I do so, is the condition I ask for letting you use
                *my* code.
10:07 < rlpowell> Yeah, that seems really petulant to me (no offense intended).
10:07 < bancus> Would you complain if I charged you $10?
10:09 < rlpowell> If you charged me $10 but I didn't have to release my code to the world?  I'd prefer that in theory; the problem
                  is that $10 x many pieces of code adds up pretty quickly.
10:10 < rlpowell> Also note that if it was I have to let *you the author of the code I'm using* use my code, that'd be fine.
10:10 < rlpowell> Because that doesn't make it impossible for me to make money from it.
10:10 < rlpowell> It's the "release to the whole world" that's the problem.
10:11 < Tene> If nobody would ever care about modifying the code you're writing to sell, why would anyone care to buy it in the
              first place?
10:11 < Tene> Or are you claiming that you can get it completely right, with no bugs, and perfectly appropriate for every purpose
              for every potential user?
10:11 < Tene> Or am I misunderstanding the situation that's relevant here?
10:12 < rlpowell> Wanting a modification to the code and having the inclination/ability/desire to sit down and modify it are *not*
                  the same thing.
10:12 < rlpowell> If I found a bug in winamp, I would *not* be interested in fixing it myself (random example).
10:14 < bancus> I'm drawing comparisons between types of compensation.
10:15 < rlpowell> Tene: And even if you want to modify the code to fix a bug, that's very different, to me, from the code being
                  something that should be in the general code-reuse pool.
10:16 < bancus> I think, from a corporate standpoint, the ability to hire someone to add a feature for me is not to be
                underestimated.
10:16 < rlpowell> Huh?
10:16 < bancus> Also the ability to keep using an old version when a company wants to force a major upgrade.
10:16 < bancus> rlpowell: You're talking about *personally* modifying code.
10:16 < bancus> I'm talking about paying someone to do that for you.
10:16 < bancus> Sure, the company might be willing to do it for you, if you buy enough of their stuff.
10:17 < rlpowell> I understand what you're saying, but I'm failing to see how it fits in with the discussion.
10:17 < bancus>  10:12 < rlpowell> Wanting a modification to the code and having the  inclination/ability/desire to sit down and
          modify it are  *not* the same thing.
10:17 < rlpowell> Ah.
10:19 < bancus> The bigger trouble I've found, rather than having the inclination and ability, is tracking your modifications over
                new versions.
10:19 < rlpowell> My issue is that if someone writes, say, a nice graph theory library and it's GPL, and I need that functionality
                  for my commercial product, my choices are: 1.  Use the code, release my code, make little or no money.  2.  Write
                  again myself, diminishing productivity.  In either case, the economy has been harmed, which would be OK *if*
                  there was equal value to me releasing my code, and I assert that there usually isn't.
10:19 < bancus> Enough modifications, and tracking your changes becomes a fulltime job in itself.
10:19 < rlpowell> True dat.
10:20 < bancus> I once changed gaim so that actions appeared as "* person does shit" Instead of "***person does shit"
10:20 < bancus> A little change, for a personal preference.
10:20 < bancus> But it was a pain in the ass to keep rerolling my changes into new debs.
10:20 < bancus> Gentoo might make that a bit easier.
10:20 < bancus> But at the expense of making everything a bit harder.
10:21 < bancus> (and/or more time consuming)
10:21 < florolf> http://youtube.com/watch?v=T62E-_pQt3c <- sweet.
10:22 < rlpowell> *nod*
10:22 < rlpowell> bancus, cizra, tene: may I publish this conversation?
10:23 < rlpowell> (don't know if I'm going to, but I want to at least keep a copy for my notes)
10:24 <@timonator> frightening!
10:24 < bancus> Go for it.
10:25 < cizra> rlpowell: Have fun
10:26 < bancus> Like I said, for me it's a matter of preference. Having GPL'd code returned to the wild is the compensation I ask.
                People are, of course, free to negotiate other licenses, assuming that it's all my code. If it's not, then it's not
                my problem either.
10:26 < bancus> Using the GPL also allows multiple people to contribute to a project and assure that they all get access to the
                compensation.
10:27 < rlpowell> I understand that.  I assert that by requiring that compensation, you are harming the economy without
                  *necessarily* giving any reas benefit to anyone, including yourself.
10:27 < bancus> Hm.
10:28 < rlpowell> s/reas/real/, obviously.
10:28 < bancus> How does this differ from writing code that I simply never release?
10:28 < rlpowell> Not at all.
10:28 < rlpowell> WHich is *exactly* my point.
10:29 < bancus> Except that, with the GPL, while I may not get any benefit, I may get benefit after all.
10:29 < rlpowell> You might, but is that benefit worth the harm caused?
10:29 < rlpowell> Note that I'm assuming here that the point of the GPL is to increase the size of the general code-reuse pool.
10:30 < rlpowell> Which probably isn't the case; making it impossible to make money selling software was probably part of the goal.
10:30 < bancus> I don't see that it is harm, necessarily.
10:30 < rlpowell> But I consider that goal *horrifyingly* unethical.
10:30 < bancus> The economy is no worse off than if my code had never existed at all.
10:31 < rlpowell> Right, but I consider contributing to the general code-reuse pool *where appropriate* an ethical imperative.
10:31 < rlpowell> So "hoarding" generally useful code is still harm.
10:31 < bancus> And some people consider contributing to the poor an ethical imperative.
10:31 < bancus> So "hoarding" your money is harm?
10:32 < rlpowell> Yeah, but I don't agree with that.  :)
10:32 < rlpowell> (and I can point to good reasons, but that's a whole different discussion)
10:33 < bancus> And I don't agree that giving people my code is an ethical imperative.
10:33 < bancus> It's a nice thing to do, certainly.
10:33 < rlpowell> I can go with that.
10:33 < bancus> It's my code, I wrote it, and I should be able to benefit from it in the manner that I choose.
10:33 < rlpowell> Yes, I don't disagree with that in any way.
10:34 < rlpowell> I'm just asserting that the GPL is a petulant and harmful form of benefit to request.
10:34 < bancus> Hm.
10:34 < rlpowell> Given my overal goal of "as much good code getting written as possible, making as much money as possible".
10:34 < rlpowell> Huh, I didn't realize that was my overall goal.  Interesting.
10:34 < bancus> And requiring you to pay an assload of money isn't petulant and harmful?
10:35 < rlpowell> Depends on the code.
10:35 < rlpowell> If the code is general utility code that would be useful to hundreds of programs around the world, then yes,
                  that's harmfus.
10:35 < rlpowell> l
10:35 < Tene> Wait, how does you releasing commercial software which includes my graph theory library contribute to the code-reuse
              pool?
10:36 < rlpowell> It doesn't, but unless the commercial software is, itself, something that would save other people time building
                  their own software (copycatting does *not* count here), so what?
10:36 < bancus> Theoretically anything could save other people time.
10:36 < Tene> Then how does you not being able to re-use my graph-theory library prevent contributions to the code-reuse pool?
10:36 < rlpowell> Theoretically.  It's a maximization problem, in my mind.
10:37 < Tene> Yes, feel free to republish my contributions here.
10:38 < Eimi> Why doesn't copycatting count?  If your commercial software is part of the code reuse pool, then I can use it.  Else
              I have to build my own.  The fact that "my own" would be functionally a copy of yours doesn't seem like it would
              enter into things.
10:39 < rlpowell> I want to maximize the economic output of the software industry.  Some code's maximum value to that output is in
                  being available for re-use.  Some code's maximum value to that output is actually generating money (i.e. being
                  sold).
10:39 < rlpowell> Eimi: Does my last comment answer that?
10:40 < rlpowell> I assert, with some limited evidence, that maximizing the economic output of the software industry == lots of
                  good software being available at, at worst, reasonable prices.
10:41 < rlpowell> Having actually done the math on multivariable maximization problems is probably helpful here.  :)
10:41 < bancus> So, you're saying that frameworks and libraries and such should be freely available, to enable more end-user apps
                to be made quickly and sold for fat profits?
10:41 < Eimi> Probably.  My understanding of economics has always been pretty poor.  I'm not sure what is meant by "the economic
              output of the software industry".  I think the designers of the GPL had as a goal maximizing the size of the reusable
              code pool.
10:41 < rlpowell> Bingo.
10:41 < bancus> So, framework developers should work for free, then?
10:41  * rlpowell thinks about that for a minute.
10:41 < bancus> A good framework is arguably more difficult to make than end-user software, in a lot of cases.
10:42 < rlpowell> What do you mean by "framework" here?
10:42 < rlpowell> Eimi: Yes, I'm aware that the GPL's goals do not match mine; I find the GPL's apparently goals offensive, and I'm
                  trying to explain why.
10:42 < bancus> System-level APIs, or, really anything that offers a generic way to accomplish task X.
10:42 < Tene> I believe that the common counterargument is that "lots of software that I can't work on or fix bugs on and that I'm
              dependant on the original author for fixes of" is actually a bad situation to be in.
10:43 < rlpowell> But is it worse than hamstringing the software economy?  I don't know.  I don't think anyone does.
10:43 < rlpowell> It *feels* less worse to me, but that has no value, of course.
10:43 < bancus> rlpowell: To tack onto tene's comment, would it be alright for someone to build a house, and then decree that you
          may not repair it or make any modifications?
10:43 < bancus> Or make a car, for instance.
10:43 < rlpowell> It's not comparable.
10:43 < rlpowell> Making a modification to software == being able to make copies for zero marginal cost.
10:43 < Tene> Is a software economy that primarily produces proprietary software not already hamstrung?
10:44 < bancus> There is a law that says that car companies cannot require you to only use first-party service.
10:44 < rlpowell> This is not true of houses or cars.  At all.
10:44 < rlpowell> Tene: Yes, it is, which is why it's a *multi-variable* maximization problem.  :)
10:44 < Tene> Exactly the point that most people don't explicitly mention!
10:45 < rlpowell> Woohoo, I've said something useful.  :)
10:45  * bancus is still waiting for an answer about frameworks.
10:45 < rlpowell> I still don't know what you mean by framework.
10:46 < bancus> Say, http://audiere.sf.net/
10:46 < bancus> This is a library that enables loading and playback of music and sounds.
10:46 < Tene> bancus: a framework is just a software product targeted towards software developers.  It falls under "end-user
              software" not "libraries"
10:46 < bancus> Presumably saving time for authors of software.
10:46 < bancus> Tene: I'm conflating framework and library here.
10:46 < Tene> At least, that's the primary relevant distinction I can think of.
10:46 < rlpowell> http://www.overcomingbias.com/2007/03/policy_debates_.html -- the reason why arguments like this tend to simplify
                  instead of admitting that it's a multi-variable problem.
10:46 < bancus> Or possibly using framework to mean "a system of libraries".
10:47 < bancus> :o
10:47  * bancus is paying attention to the iPhone SDK talk.
10:47 < bancus> They're talking about inviting some developers from various companies to try out the SDK for a few weeks.
10:47 < kpreid> THe most useful definition I know is that a framework is a library which, if you use it, you tend to use all over.
10:47  * Tene snorted coke upon reading "Dr. Snakeoil's Sulfuric Acid Drink"
10:48  * Tene curses at rlpowell.
10:48 < rlpowell> bancus: My *suspicion* is that maximized output by my criteria (hereafter "maximization") occurs when people
                  noticing the need for something like audiere say "*FUCK*.  That's going to be a PitA.", then go write it, then
                  release it under LGPL (as, in fact, audiere is), then get on with the real code.
10:48 < kpreid> i.e. an extension to the platform rather than a module to hook on the side
10:48  * rlpowell laughs at Tene.
10:48 < bancus> "Travis Boatman from EA is up, so he had two weeks to play with the new SDK. They took an existing game called
                Spore."
10:48 < rlpowell> 0.o
10:48 < Tene> I need to become cuter so that my misery elicits sympathy instead of entertainment from rlpowell.
10:48 < rlpowell> bancus: URL, please.  Now.
10:48 < bancus> "You control the spore by moving the iphone around"
10:48 < bancus> http://live.phonemag.com/
10:48 < kpreid> http://arstechnica.com/news.ars/post/20080306-live-coverage-of-the-iphone-software-roadmap-announcement.html (not
                same)
10:48 < bancus> It's just comments made by livebloggers.
10:49 < bancus> http://live.gizmodo.com/
10:49 < bancus> has pictures
10:49 < bancus> This SDK looks fucking sweet, BTW.
10:49 < rlpowell> How much?
10:50 < bancus> For what?
10:50 < bancus> The SDK?
10:50 < bancus> They haven't said, but they don't charge for any others.
10:50 < bancus> Although Xcode is Mac-only.
10:51 < bancus> And maybe this is Leopard-only.
10:51 < bancus> Which might convince me to upgrade finally.
10:51 < bancus> Nothing else has.
10:51 < rlpowell> OK.  Usually SDK's aren't free, and I didn't know apple's stance.
10:51 < bancus> "But you can get Safari 3" "Um, already god it. They released it for Tiger, too."
10:51 < bancus> got it
10:52 < bancus> I went through the list of new features on Apple's site, and most of the ones I was interested in, I already have,
                usually through 3rd party software.
10:52 < bancus> Except for the unified theme, which I like, but not enough to drop $129 on.
10:53  * Broca reads backscroll
10:53 < Broca> Whew, you sure talked a lot.
10:53 < rlpowell> Oh no; and it was just settling down.  :)
10:53 < bancus> Oh, aside from the SDK, they annouced enterprise support, including Exchange and Cisco VPN stuff.
10:53 < Broca> rlpowell: I sure would like to hear rms' response to the assertion that there are some kinds of software to which
          people don't *want* the source.
10:53 < rlpowell> I think the only useful thing I said was "This is a multi-variable maximization problem; here's my maximization
                  criteria."
10:54 < rlpowell> Broca: I wouldn't; I can imagine it pretty clearly.
10:54 < Broca> Then your model of rms is more detailed than mine. Let me hear.
10:55 < Tene> rlp: you assert "If I let you have the source, you'll give it to all your friends for free.", right?
10:55 < rlpowell> I don't assert, by the way, that there exists software that *no-one* wants the source to, but that there is
                  software for which general source code availability really isn't very helpful, where "isn't very helpful" is WRT
                  my maximization criteria.
10:56 < rlpowell> Tene: No, I assert "If I let the world have the source, *someone* will distribute it widely; for free or not, I
                  sure won't be seeing the money."
10:56 < rlpowell> Broca: I think he would categorically assert that I'm full of shit and that software is near-useless without the
                  source code.
10:56 < rlpowell> But I haven't read any RMS in years; it makes me break out in hives.
10:56 < Tene> You saw his craigslist ad, though, right?
10:57 < rlpowell> Tene: I'm fully aware that there is counter evidence to this argument, but then, would you like me to count the
                  number of OEM software spams I get each day?
10:57 < rlpowell> I think so, yes.
10:57 < Tene> rlpowell: and you assert that that number would be higher if the source was released?
10:58 < rlpowell> No.  I'm simply using that as evidence that there *are* people who will take software without paying the authors
                  if given the chance.
10:58 < Tene> So, *which* large software product isn't widely available illegally for free?
10:58 < bancus> Heh. From Gizmodo's coverage: Epocrates. This is an app for doctors. Jesusphone about to gain app that can cure
                cancer.
10:59 < Broca> Huh?
10:59 < bancus> New iPhone apps.
10:59 < rlpowell> Tene: "illegally" is the key word there.  Are you saying it's possible to release source in such a way that
                  getting a free copy would still be illegal?
10:59 < Broca> What's the Jesusphone?
10:59 < Tene> rlpowell: As I understand it, the traditional response is that the people who would do that are already doing it just
          fine with every proprietary software product they get their hands on anyway, so it would be no different.
10:59 < rlpowell> Because I'm not aware of any such system.
10:59 < bancus> A joke name for the iPhone.
10:59 < bancus> Because everyone was waiting for it like the second fucking coming.
10:59 < Broca> Heh.
10:59 < rlpowell> Heh.
11:00 < bancus> Although with this SDK, it may well be the second fucking coming.
11:00 < bancus> (In a different sense, of course.)
11:01 < bancus> Hm. Sega's demoing now.
11:01 < Broca> bancus: you should start quoting Amos 5:18 to people who are waiting for the second coming.
11:01 < cizra> Broca: What does it say?
11:01 < bancus> "So you can make me come, that doesn't make you Jesus."?
11:02 < bancus> Oh, sorry, s/Jesus/Je-he-sus/.
11:02 <@timonator> mythtv wants mysql :((
11:02 < bancus> Actually, I think that's the fourth track.
11:02 < bancus> And probably no one here got that joke.
11:03 < Broca> Am 5:18: Woe to you who long / for the day of the LORD! / Why do you long for the day of the LORD ? / That day will
               be darkness, not light.
11:05 < cizra> Broca: A comment on self-righteousness and false confidence of people, would be my guess.
11:06 < Broca> Indeed.
11:10 < bancus> I guess I could instead make a joke about Revelations 22:21 instead.
11:10 < bancus> He who testifies to these things says, “Surely I am coming quickly.”
11:10 < bancus> Amen. Even so, come, Lord Jesus!
11:10 < Broca> http://www.codeplex.com/singularity -- MICROSOFT LAUNCHES SINGULARITY
11:10 < bancus> Oh, sh--
11:11 < bancus> By the way, my first joke was a reference to the Tori Amos song Precious Things.
11:11 < Broca> Oh, okay. Never heard that one.
11:14 < bancus> Sega developer about putting Monkey Ball on the iPhone: “This is not a cellphone game. This is a full console game.
                And we underestimated the power of the device. We had to fly in a developer to upscale the art for the iPhone.”