Put the knife down and take a green herb, dude. (c) Ruffin Bailey 2001-2021

$fileNames = new-object collections.arraylist $license = @" // This Source Code Form is subject to the terms of the Mozilla Public // License, v. 2.0. If a copy of the MPL was not distributed with this // file, You can obtain one at https://mozilla.org/MPL/2.0/. "@ Get-ChildItem C:\temp\*.js | %{ $name = $_.Name #Write-Host $name $content = new-object collections.arraylist Get-Content $_.fullname | %{ $ignore = $content.add($_) } $newPath = ("C:\temp\out\" + $name) write-host $newPath $content = $($content -join [Environment]::NewLine) ($license + $content) | set-content $newPath } Write-Host $($fileNames -join [Environment]::NewLine)

Recently, Real Software posted a blog post that said that Real Basic works with MySQL, but that you'd have to open source any Real Basic app that used it.

Real Studio is able to connect to the Community (free) edition of MySQL, but this edition usually requires you to open source software that connects to it (due to its GPL license).

I didn't think that was right. Obviously, you can use MySQL without open sourcing code the same way you can write an app with GNU/Linux and release it closed source. If MySQL was that "viral", almost every two bit app on some cheap php hosting environment is now GPL'd.

Turns out it's more complicated, apparently. It might be that the Real Software plugin that talks to MySQL is GPL'd. I don't know how that'd happen. Can't Real just get a MySQL commercial license and start coding?

Regardless, the post got me thinking about MySQL licensing. And that, though I'm over two years late to the party, seems to be a lot of fun.

Groklaw - Monty Program AB's Suggestion to EU Commission to Get Rid of the GPL on MySQL - Updated:

So that is why they care. They have big plans for a business around MySQL, and they want to make some money from it. MariaDB is their fork of MySQL. Of course, there's nothing wrong with making money. Notice the role of the Open Database Alliance in all this, in case anyone tells you there is no connection. There is.

I'm not sure I understand how the sort of dual-licensing many open source but commercially funded apps work. There's no way that MySQL-Cash isn't somehow benefitting from impressive changes and fixes to MySQL-Altruism. (My guess is that they pretend to have a "clean room", have someone walk in, view the OSS MySQL code, then walk out, across the hall, and do the same thing from memory to the closed version.)

But companies do this, pretending that the open source version is completely and irrevocably downstream of the closed source version, and one of the MySQL co-founders apparently wanted to have his cake and eat it twice. He wanted to relicense MySQL under the Apache license so, essentially, he could do whatever the heck he wanted to do with the closed MySQL codebase at his new company without releasing that code rather than have to open source his fork of MySQL. That is, unless he bought a license to develop MySQL from the closed source, his only avenue to fork was to fork the GPL version. Oh noes! Which, of course, suggests that there's a fair bit of code in MySQL-Cash that we're not seeing in MySQL-Altruism.

Apparently that threw the world on its ear. IBM said that Oracle could buy Sun (and, with it, MySQL), but some little schmoe (above) disagreed. The sale was held up, and millions of bucks allegedly lost during the deliberation. Fine.

The worst part? Some wicked pixelers started saying stuff like this:

And in the ultimate irony, Richard Stallman himself joined the fray against... Richard Stallman?

/sigh

No, no he didn't. But that didn't stop the world from deciding that the GPL prevented you from forking code. WTF? RLY? Come on. Get a new job writing about whatever it is that you're really interested in, because it sure ain't the GPL.

Here's a quote from a horribly written piece on CNet called "Stallman: GPL doesn't guarantee software freedom".

Even Richard Stallman, co-author of the GPL and founder of the free-software movement, and not someone that spends much time worrying about monetization of open-source software, gets this.

As noted in a letter co-drafted with Open Rights Group and Knowledge Ecology International, Stallman notes that Oracle's proposed acquisition of MySQL could hurt its development because the GPL reduces incentives to commercialize the code.

Come on, that wasn't enough for you to wonder what was going on? You really thought RMS was arguing for Apache over GPL? To what ends? (Sorry -- I don't usually bash like this, but the FUD here is insane, and so easily seen through.)

A commenter on this story has a much better answer/handle on this situation. It's so good, I'm posting it all, in case the story disappears. "mbenedict" is the author.

There are many issues here getting mixed up.

First, in fairness to Stallman, when he talks about a "lack of a more flexible license" for MySQL he's really talking about GPL v3 -- or more specifically mixing v2 and v3 code together in a possible future fork. Matt's selective quoting above seem to misrepresent Stallman's position by removing its context.

Second, even under GPL v2, there is *nothing* preventing a "MySQL 2" to "arise, take the code, hire all of the developers, and development of the open-source database would not miss a beat." A new entity under GPL *can* fork MySQL code and do all that. What the new entity *cannot* do is re-license the forked code under a different license -- including GPL v3, or a separate Commercial License under a dual-licensing scheme.

So we finally get into the crux of the issue which Matt curiously omits: dual-licensing of GPL v2 code. We all knew it was "evil". Now we're all acting surprised that it could be "evil". Dual-licensing allows companies to give out crack to babies, and then charge them $$$ once they're hooked... completely against the spirit of GPL.

Now, I'm not a fan of GPL. I've been a long proponent for BSD-style licenses (which Apache derives from). Lots of companies "talk" about relicensing from GPL to Apache... but actions speak louder than words.

Okay, okay, okay. Let's stop mbenedict at one spot -- BSD? Insane. "Please steal my code! Make hats of cash! Give me nothing! Just know that you can't sue me." Or, as I've said before, those licenses "essentially enable legalized plagiarism". That's crazy. Well meaning, perhaps (more likely too business oriented and not written by a guy who actually contributes to open source code), but crazy.

There's a perfect license for releasing code into the world for commercial companies to use: The LGPL. It's fair. It encourages passive collaboration (OSS' biggest boon) from commercial enterprises (a rare but powerful thing). And, get this, it fairly requires that if they improve the functions you gave them, they must give back to you those -- and only those -- improvements. Brilliant! That's what The GNG Manifesto is all about.

I think you get the point. RMS is anti-GPLv2 insofar as it's not the GPLv3. He likes GPLv3, but thinks it should have more flexibility for users to change their project's license as changes to the GPL are made. He does not like Apache licenses more than GNU licenses. MySQL should, for RMS, be licensed under GPLv3, though v2 and v3 allow you to fork.

What a media fail.

Labels: gpl, hats of money, licensing, mysql, open, OSS

The LGPL and Java:

If you distribute a Java application that imports LGPL libraries, it's easy to comply with the LGPL. Your application's license needs to allow users to modify the library, and reverse engineer your code to debug these modifications. This doesn't mean you need to provide source code or any details about the internals of your application. Of course, some changes the users may make to the library may break the interface, rendering the library unable to work with your application. You don't need to worry about that -- people who modify the library are responsible for making it work.

That's why the LGPL is no good; this is an exploitable hole to allow people to modify a library and start reverse engineering your code, it would seem. I'm not real sure why that's bad, but I do know it's not something I'd particularly want butting up next to my proprietary code.

Perhaps this is why a number of pay-to-play companies run a secondary, "Free as in Mozilla Public License" codebase. Having grown out of the need to recreate Netscape, not to become part of an idealistic Hurd, the MPL watches companies' backs well. Too well.

Here's the metaphor I heard that got me interested in the LGPL (IANAL):

You're making a calculator. You've got every button down except the square root button. You decide to look at external libraries to save time.

GPL Square Root -- You can embed the function and sell your software, but now every bit of your calculator, plus, minus, 1s, 5s, the whole thing, must be released with full source code. Now anyone can take your code and sell your application. Anyone. Not necessarily a bad thing, unless you like to make money. Everyone can use your first-rate natural log function, but do remember that they, too, can't sell this [legally] without showing the code to anything they've added. Great for promoting passive collaboration.

LGPL Square Root -- Originally, before the Java issue, I thought the LGPL went like this: You can add this square root function, release and sell your app. You only have to include source to the square root function; everything else can stay proprietary. Let's say, however, that you find a bug in the square root function, say it thinks the square root of 49 is 8, but gets everything else right -- or that it only goes to 5 decimal places. If you fix the bug or extend its accuracy to 10 decimal places, you have to release the source to those improvements with your app. Personally, that's great so far. I like it. If you only have the function b/c of someone else's contribution, you owe it to him/her/them to release your improvements for all to use. It enables passive collaboration, but still allows you to protect your investment. Nobody "deserves" your fancy natural log function if you don't want to give it to them. It was entirely your work. Keep it. The "rights" to the code are protected from both sides, and fairly. You'd also owe the 'LGPL' a big thank you, of course, but you get the point.

BSD Square Root -- Take it. Sell your app. You might have to acknowledge the project whose work you took, but that's it. It's yours. Take it from 5 to 10 decimal places of accuracy, heck, take it to 2000, and you don't have to share a keystroke. Great for greedy people who don't want to be bothered, but I'd hope they'd feel a little bad about taking work to which they didn't -- nor will -- contribute.

MPL Square Root -- You can take this square root function, release, and sell your app. In fact, you could (afaict) take your own square root implementation, take a few files from the MPL version, and put those together to be released. Your own, wholly original code could stay closed, and only changes you made to MPL files would need to be released. Initially, I thought the MPL was like BSD. As is obvious here, it's not. It's sort of like the LGPL in that your code can be released in support of and as part of a closed source product, but it's "file-level" in that any file of your library could potentially be taken and used in a new context. But it's as close to the LGPL as you can get if you want to allow someone to compile your code alongside proprietary code, and it doesn't seem exceptionally evil. It's a good deal less protective than the LGPL, but just good enough that I still like it. (MPL section added 20131011.)

Okay, comparison over. But now, following the relatively dated quote about Java issues, above, (iff the Java issue is legit) the LGPL forces anyone using the square root function to allow anyone to decompile their entire app that uses the square root function to "bug check" against their code.

That's simply not a place I want people going. Short of the square root function, it's my code. Stay out. If you want my code to work, use the old library that thinks 49's square root is 8. If you can upgrade the lib to work with 49, great, but why should I have to give you any rights to my code downstream of the square root function to make that happen? You didn't make the investment in app, and I want to protect it. That should be my right. You shouldn't get to peek at my natural log function if I've used LGPL code elsewhere. If your change to the square root -- and who knows how inane that change might be; perhaps it also now doesn't do 36 correctly -- breaks compatibility with my app, that's my problem. I'll fork the square root lib if I have to.

Oh well, guess I'll eventually have to start my GNG (GNG's Not GNU) movement with an LGPL done, if not right, then fair. Might be as easy as slashing Section 6 in the LGPL; I'll have to check sometime. Now, on to UO (I'm on sabbatical from WoW)!

Labels: java, lgpl, licensing

title: Put the knife down and take a green herb, dude.	descrip: One feller's views on the state of everyday computer science & its application (and now, OTHER STUFF) who isn't rich enough to shell out for www.myfreakinfirst-andlast-name.com Using 89% of the same design the blog had in 2001.
FOR ENTERTAINMENT PURPOSES ONLY!!! Back-up your data and, when you bike, always wear white. As an Amazon Associate, I earn from qualifying purchases. Affiliate links in green.
x MarkUpDown is the best Markdown editor for professionals on Windows 10. It includes two-pane live preview, in-app uploads to imgur for image hosting, and MultiMarkdown table support. Features you won't find anywhere else include... MarkUpDown Multiline Table & Bootstrap Grid support. Beautiful Easy Actions that keep the Markdown flowing. HTML paste to paste HTML source into your documents. You've wasted more than $15 of your time looking for a great Markdown editor. Stop looking. MarkUpDown is the app you're looking for. Learn more or head over to the 'Store now!

Thursday, April 15, 2021
Insert a license header into each file of your code with Powershell Okay, this is just a start -- I haven't made it recursive yet -- but just in case I get this bee in my bonnet again, I want to remember what I'd done to this point. Got a library that needs licensing? You'll want to change the save location from `C:\temp\out\` (later I'll make that a param I guess). You might also change what's in `$license` from the MPL (though it is my favorite), but, past that, throw into a Powershell file and you're on your way. $fileNames = new-object collections.arraylist $license = @" // This Source Code Form is subject to the terms of the Mozilla Public // License, v. 2.0. If a copy of the MPL was not distributed with this // file, You can obtain one at https://mozilla.org/MPL/2.0/. "@ Get-ChildItem C:\temp\*.js \| %{ $name = $_.Name #Write-Host $name $content = new-object collections.arraylist Get-Content $_.fullname \| %{ $ignore = $content.add($_) } $newPath = ("C:\temp\out\" + $name) write-host $newPath $content = $($content -join [Environment]::NewLine) ($license + $content) \| set-content $newPath } Write-Host $($fileNames -join [Environment]::NewLine) Labels: licensing, noteToSelf, powershell posted by ruffin at 4/15/2021 08:30:00 AM

Thursday, October 17, 2013
How can Microsoft be eating HP's lunch if HP can't cook on its own? I need to go back and do a true link aggregation with prior posts on the subject, but in the past month or two, I've seen too many "Mac" bloggers -- and there's a community that I enjoy that seems to include John Gruber, Marco Arment, and, recently, Ben Thompson -- say that Microsoft has been eating hardware manufacturers' lunches for years. They're not wrong (warning: youtube vid with cursing), they've just gone waaaay too far down the Apple rabbit hole. Sure, Apple won for several years, but what's the true cost of using Microsoft? And what's the value? How much has HP, Dell, and friends had to know about creating and updating operating systems for the last twenty-plus years? How to burn an image to a hard drive. That's right. Essentially nothing. How much has HP, Dell, and friends had to spend to create an operating system? Yes, billions to license one (and, okay, very little to use the other when they've bothered), but essentially much closer to nothing to develop Windows. There's a ton of obvious user maintenance, but R and D? Relatively speaking, zero. How big of a risk was it for Apple/NeXT to continue to develop their own OS? That's an interesting question that deserves more investigation. Let's not talk so much about OS 9. That OS played itself out. It essentially failed. Apple took FreeBSD and started essentially over, building in NeXT/Cocoa and OS 9/Carbon layers on top. How difficult is it to pull off creating an OS from scratch, or, more accurately, on top of a stable NIX kernal? Enter exhibit WebOS. Ask Palm, Blackberry, Nokia, and HP how creating OSes contributes to the bottom line. How can Microsoft be eating HP's lunch if HP can't cook on its on?* To quickly turn the metaphor, if you're not willing to take on the risk of creating an OS (or decided it's a dumb risk not worth taking), I guess you're stuck with cooks that are 70% pure. There's a company that's able to create cooks that are 99.1% pure (as I catch up on my DVR'd Breaking Bad Marathon), and they're proving difficult to kill. Actually, Breaking Bad metaphors work better for Microsoft. They're like the cartel's chemist complaining about Jesse's, um, artistic approach to cooking meth (warning: more cursing from youtube). What matters is that Jesse gets to 96% purity and the cartel's team is, hrm, doing much worse. A few years ago, Ballmer was the chemist putting down the iPhone. Microsoft knows how to synthesize phenylacetic acid, but they can't cook. The point is that hardware manufacturers were unable to commoditize their complements -- nor was Microsoft able to make more than mobile equivalents to those manufacturers' beige boxes to put their OS on -- and Linux and OSS have been unable to create a replacement-level OS alternative. Unless you count Android. For today, let's leave Google out of it. Google's eating everyone's advertising lunch. Microsoft has earned the money they're collecting on Windows. It's not extortion when you're the only risk-reducing game in town. Labels: apple, hardware, hats of money, licensing, linux, microsoft posted by ruffin at 10/17/2013 09:20:00 AM

Saturday, October 05, 2013
Arguing that minimized JavaScript is in "Executable form" MPL 2.0 FAQ: Q16: Is "minified" JavaScript Source Code? No. Minified JavaScript, while not an "executable" in the software engineering sense of the word, is difficult for humans to read, edit, and modify. As such, it is not "the preferred form for modification" and so it is not Source Code as defined by the license. Therefore, minified JavaScript is the Executable form, and the responsibilities set out in the license for distribution of the Executable form should be met when you distribute minified MPL-licensed JavaScript. Compare to OOXML. Like interpreted code, you could (and, I believe) should argue that purposely obfuscated human-readable code and standards are operating as executable/proprietary code. Labels: javascript, licensing, mozilla, ooxml, open posted by ruffin at 10/05/2013 06:05:00 PM

Tuesday, October 01, 2013
The "ASP Loophole" TV is SOCIAL, again: The Honest Public License: What motivated this new license now? We have a general availability version of Funambol coming out in September. I already know there are commercial companies that are live with our code and do not return anything. More than two years ago, we did something similar, switching Sync4j from BSD to GPL. There were companies taking our code and running away with it, without returning anything. One even managed to get public with software based on our code, and our community never saw a line of their modifications. Now is no different. I've seen this argument for using GPLv2 libraries with edits on commercial web servers before, and from some large companies and government organizations. Most interesting to me in that post was seeing someone with good, FSF values outside of FSF. But I'm also reminded of my "GNG Manifeso" post years ago (over 11. What traction!). You can't really get mad at someone for "getting public" if you release under BSD. From me: This is also why I dislike the X11, BSD, and MIT licenses. These licenses don't do enough to protect the contributions of the people that made the code -- they essentially enable legalized plagiarism. It's certainly one's right to make code that's this unregulated, but these licenses are nearly overly altruistic motivations. And, as we see here, the mistaken use of the BSD didn't match the intent either. But does that require GPL? I don't think so. I'd like to think you can strike a balance between making a company release everything and making a company release any improvements to the logic you helped create in your OSS lib. Labels: F/free, licensing posted by ruffin at 10/01/2013 10:59:00 AM

Sunday, September 16, 2012
How Free software isn't enough to guarantee f/Freedom FOSS Patents: If Google can cancel Acer's license, why should Apple have to grant one to Google?: Acer is a member of the Google-dominated "Open Handset Alliance" and was about to release a smartphone running an Android fork named Aliyun, which was created by Alibaba, a Chinese Google competitor, but Google essentially says: "you're with us or you're against us". You can be a member of the OHA and an official licensee of Android, or you want to distribute forks (derivative programs), in which case we'll throw you out of the OHA and cancel your official Android license. Google got its way. Android's non-openness is old news, and nothing I could say would counterbalance the deafening silence and open double standards of various "free software" and "open source" luminaries and entities. It's strange, but endlessly fascinating, how corporations configure themselves around Free and open source licensing. Some places use GPL'd software on their server, but argue that's not distribution, so they're not under any obligation to release the source (I've forgotten how the argument goes, but I've worked for people who claim it). Others, like that bizzarre kickup around MySQL, have strange forks of code into Free and non-Free codebases. Did MySQL really never incorporate anything submitted by the community? Or there's the dual licensed stuff, which really isn't a huge deal unless you're doing it retroactively, like Mozilla, where, iirc, some stuff had to be rewritten before the new licensing was done. And here, it's not the code so much as the access to the coders. Support is worth as much as the code in some cases, especially when the code is especially complex, like I'm assuming Android is. Probably more important here is early access to Google's R&D, which Acer stands to lose if its most favored nation status is revoked. It's hard not to exploit Free software. Apple didn't always play nice with WebKit and Darwin (and still doesn't give back everything, I'm sure), and, in Darwin's case, wasn't under any compulsion to do so. With webkit, we're lucky to have an Apple that follows the LGPL as closely as they do, I'm afraid. Android, in my opinion, pretty obviously stole some of Java. GPL'd software has been discovered in a number of closed source products. Google's maneuvering of cultural affiliation, loosely de/coupled with and from open source software, is the obvious move. I wonder if RMS would (or, perhaps better asked, should) even be against what Google is doing. Should you get to see Free software as it's being developed, before it's released? Is Android the perfect example of how open source can power commerce? Labels: android, apple, free, Google, gpl, lgpl, licensing, linux, open posted by ruffin at 9/16/2012 02:11:00 PM

Monday, July 23, 2012
Why BSD is evil in nine words or less. Recently, there was an OSNews article called "Sparrow's acquisition highlights the dangers of closed source". I think there are issues enough with the logic in that statement alone, so I got baited into reading. Sparrow, I guess I should add, is a pretty neat (though low featured, afaict) stand-alone email application for OS X. People like it. Google bought it. Now there will be no new Sparrows and, more importantly, no Sparrow for iPad, which they'd been working on recently. Buying Sparrow was very likely, I believe, a talent acquisition move for Google. They're buying the programmers as much or more than the code. Regardless, apparently the OSNews author believes that if Sparrow had been open source, it'd be reasonable to expect the next version of Sparrow to come out on time and under budget in spite of the developers leaving. And I guess the implication is that Sparrow would have grown just as quickly and been just as personally lucrative for the developers if it were an open source project as it did as a closed source one (wait... what?). Such an argument is pretty sad. Open Source only protects you from your favorite app's going kaput if you have programmers ready to contribute, and usually to contribute for free. I've posted a couple of times that I'm reading Dreaming in Code, the book about the Chandler Project. Some folks seem to enjoy the app. It's open source. Chandler's also essentially dead. Thunderbird is open source. It's in stasis mode too. Open source does not protect you from a project's death. It does allow a group of programmers to pick up where the original team stopped. And if you've ever inherited code, you know what a bear coming to a brand new codebase can be. Let's just say it's often a theoretical possibility, but practical nightmare. Every program is a cyborg. Completely remove the original human element -- the original developers -- and the cyborg likely dies. Google's hiring Open-Sparrow's developers kills Open-Sparrow just as surely as this kills Closed-Sparrow. The difference? At least Closed-Sparrow's programmers made enough cash to keep them interested to this point in the game. Show me a project that went -- entirely unfunded -- from commercial to successful open source project. Closed source protects small programming shops. There's no way around it. There are a few exceptions that prove the rule where donations fund projects, but I don't know that any were initially closed-source commercial projects. More importantly, it's hard to sell software as a business model using free software. Services, fine, you can sell yourself as someone who provides services for OSS, but there's not a huge market for email-app-related services. Not many calls from people waiting to spend $120/hour for you to configure Sparrow. Lots more willing to pay for the app or to click your, in my case at least, surprisingly well-targeted ads. Regardless, replying to a comment to the OSNews post, I believe I convinced myself of why BSD is evil in nine words or less. OSNews > Thread > "RE: Whining because they are Apple fanboys?" by Macrat: RE: Whining because they are Apple fanboys? by Macrat on Sun 22nd Jul 2012 02:33 UTC in reply to "Whining because they are Apple fanboys?" Member since: 2006-03-27 Your suggestion - that Apple users consider open source - is like asking a group of religious fanatics to convert to atheism. All Mac users are using open source as open source apps come with OS X. http://www.apple.com/opensource/ Reply Parent Score: 3 RE[2]: Whining because they are Apple fanboys? by rufwork on Mon 23rd Jul 2012 13:52 in reply to "RE: Whining because they are Apple fanboys?" Member since: 2012-07-23 The important thing to remember is that the OS is using BSD-licensed stuff under the hood. That means it was open source, but isn't any more. The BSD is far too permissive, and doesn't protect software. I'm surprised it's considered an Open-as-in-Free license at all sometimes. "Please plagiarize my code and call it your own! I'm begging you!" Reply Parent Score: 1 And that's really the rub for me. BSD allows a situation where *it was* open source, but it isn't any more.** Thank you, me, for putting that succinctly. Based on the length of your typical post, including the preamble to this one, I'm surprised you were able to come through like that for us. Labels: apple, bsd, copyright, gng, licensing, open, os x, OSS posted by ruffin at 7/23/2012 10:38:00 AM

Friday, July 06, 2012
Missing MySQL's move to Oracle -- Copyleft and Hats of Money Recently, Real Software posted a blog post that said that Real Basic works with MySQL, but that you'd have to open source any Real Basic app that used it. Real Studio is able to connect to the Community (free) edition of MySQL, but this edition usually requires you to open source software that connects to it (due to its GPL license). I didn't think that was right. Obviously, you can use MySQL without open sourcing code the same way you can write an app with GNU/Linux and release it closed source. If MySQL was that "viral", almost every two bit app on some cheap php hosting environment is now GPL'd. Turns out it's more complicated, apparently. It might be that the Real Software plugin that talks to MySQL is GPL'd. I don't know how that'd happen. Can't Real just get a MySQL commercial license and start coding? Regardless, the post got me thinking about MySQL licensing. And that, though I'm over two years late to the party, seems to be a lot of fun. Groklaw - Monty Program AB's Suggestion to EU Commission to Get Rid of the GPL on MySQL - Updated: So that is why they care. They have big plans for a business around MySQL, and they want to make some money from it. MariaDB is their fork of MySQL. Of course, there's nothing wrong with making money. Notice the role of the Open Database Alliance in all this, in case anyone tells you there is no connection. There is. I'm not sure I understand how the sort of dual-licensing many open source but commercially funded apps work. There's no way that MySQL-Cash isn't somehow benefitting from impressive changes and fixes to MySQL-Altruism. (My guess is that they pretend to have a "clean room", have someone walk in, view the OSS MySQL code, then walk out, across the hall, and do the same thing from memory to the closed version.) But companies do this, pretending that the open source version is completely and irrevocably downstream of the closed source version, and one of the MySQL co-founders apparently wanted to have his cake and eat it twice. He wanted to relicense MySQL under the Apache license so, essentially, he could do whatever the heck he wanted to do with the closed MySQL codebase at his new company without releasing that code rather than have to open source his fork of MySQL. That is, unless he bought a license to develop MySQL from the closed source, his only avenue to fork was to fork the GPL version. Oh noes! Which, of course, suggests that there's a fair bit of code in MySQL-Cash that we're not seeing in MySQL-Altruism. Apparently that threw the world on its ear. IBM said that Oracle could buy Sun (and, with it, MySQL), but some little schmoe (above) disagreed. The sale was held up, and millions of bucks allegedly lost during the deliberation. Fine. The worst part? Some wicked pixelers started saying stuff like this: And in the ultimate irony, Richard Stallman himself joined the fray against... Richard Stallman? /sigh No, no he didn't. But that didn't stop the world from deciding that the GPL prevented you from forking code. WTF? RLY? Come on. Get a new job writing about whatever it is that you're really interested in, because it sure ain't the GPL. Here's a quote from a horribly written piece on CNet called "Stallman: GPL doesn't guarantee software freedom". Even Richard Stallman, co-author of the GPL and founder of the free-software movement, and not someone that spends much time worrying about monetization of open-source software, gets this. As noted in a letter co-drafted with Open Rights Group and Knowledge Ecology International, Stallman notes that Oracle's proposed acquisition of MySQL could hurt its development because the GPL reduces incentives to commercialize the code. Come on, that wasn't enough for you to wonder what was going on? You really thought RMS was arguing for Apache over GPL? To what ends? (Sorry -- I don't usually bash like this, but the FUD here is insane, and so easily seen through.) A commenter on this story has a much better answer/handle on this situation. It's so good, I'm posting it all, in case the story disappears. "mbenedict" is the author. There are many issues here getting mixed up. First, in fairness to Stallman, when he talks about a "lack of a more flexible license" for MySQL he's really talking about GPL v3 -- or more specifically mixing v2 and v3 code together in a possible future fork. Matt's selective quoting above seem to misrepresent Stallman's position by removing its context. Second, even under GPL v2, there is nothing preventing a "MySQL 2" to "arise, take the code, hire all of the developers, and development of the open-source database would not miss a beat." A new entity under GPL can fork MySQL code and do all that. What the new entity cannot do is re-license the forked code under a different license -- including GPL v3, or a separate Commercial License under a dual-licensing scheme. So we finally get into the crux of the issue which Matt curiously omits: dual-licensing of GPL v2 code. We all knew it was "evil". Now we're all acting surprised that it could be "evil". Dual-licensing allows companies to give out crack to babies, and then charge them $$$ once they're hooked... completely against the spirit of GPL. Now, I'm not a fan of GPL. I've been a long proponent for BSD-style licenses (which Apache derives from). Lots of companies "talk" about relicensing from GPL to Apache... but actions speak louder than words. Okay, okay, okay. Let's stop mbenedict at one spot -- BSD? Insane. "Please steal my code! Make hats of cash! Give me nothing! Just know that you can't sue me." Or, as I've said before, those licenses "essentially enable legalized plagiarism". That's crazy. Well meaning, perhaps (more likely too business oriented and not written by a guy who actually contributes to open source code), but crazy. There's a perfect license for releasing code into the world for commercial companies to use: The LGPL. It's fair. It encourages passive collaboration (OSS' biggest boon) from commercial enterprises (a rare but powerful thing). And, get this, it fairly requires that if they improve the functions you gave them, they must give back to you those -- and only those -- improvements. Brilliant! That's what The GNG Manifesto is all about. I think you get the point. RMS is anti-GPLv2 insofar as it's not the GPLv3. He likes GPLv3, but thinks it should have more flexibility for users to change their project's license as changes to the GPL are made. He does not like Apache licenses more than GNU licenses. MySQL should, for RMS, be licensed under GPLv3, though v2 and v3 allow you to fork. What a media fail. Labels: gpl, hats of money, licensing, mysql, open, OSS posted by ruffin at 7/06/2012 09:20:00 AM

Saturday, December 24, 2011
Google Music Manager's License Links -- Not Good Enough (but it does do FLAC!) I'm playing around with Google's Music Manager a little, uploading a few hundred tracks just for fun and downloading a few I grabbed from a recent promotion. I'm not sure how I feel about cloud music, but it seems like a decent idea. Surprised Apple & Google don't care about the bandwidth. How much can a dollar a track really buy you? But the link to the OSS used in the Music Manager isn't quite enough. Many OSS apps that include or are themselves Free software display the GPL or LGPL in their entirety when you install, making you "accept" the GPL before using the software. I always thought that was the wrong terminology -- you're not really "accepting" it so much as the developers who made your software did. In retrospect, however, this in-your-face license makes a lot of sense. Many applications only have the required OSS licenses in their About boxes, and that just barely seems to meet the letter of the law for some OSS license requirements, imo, even though I'm guilty of doing the same thing. I think you're required to give the license to your user and a link to code, and hiding just the license in an obscure menu item isn't the way to do that. Google does even worse in the Music Manager. In the About, there's a link to a page (admittedly a page copied locally by the install, though it is an html page that requires a browser to view) that has more links to licensing realted to the software they've used in Music Manager. In some cases, like libmpg123, the link is directly to the LGPL. In others, like id3lib, the link is to that software's home page, not the license. That seems bogus. I will credit that page with an appropriate link to the tarball (http://dl.google.com/dl/androidjumper/src/current/music-manager-source.tar.bz2), which is nicely done. But the tab for About with a link to YA page of links, on which you may have to hunt to find the license, is closer to Kevin Bacon than the letter of the LGPL. EDIT: But it's hard to stay mad when Google Music handles FLAC with its Cloud player. Awesome. Labels: copyright, Google, gpl, lgpl, licensing, OSS posted by ruffin at 12/24/2011 06:47:00 PM

Thursday, April 22, 2010
The NY Times' "Ethicist" is Anything But From The Ethicist - E-Book Dodge - NYTimes.com: [Q:]... The publisher apparently withheld [a Stephen King book's eBook version] to encourage people to buy the more expensive hardcover. So I did, all 1,074 pages, more than three and a half pounds. Then I found a pirated version online, downloaded it to my e-reader and took it on my trip. I generally disapprove of illegal downloads, but wasn’t this O.K.? C.D., BRIGHTWATERS, N.Y. [A:]An illegal download is — to use an ugly word — illegal. But in this case, it is not unethical.... Thus you’ve violated the publishing company’s legal right to control the distribution of its intellectual property, but you’ve done no harm or so little as to meet my threshold of acceptability... What crap. Is the book out of print? Is that hardcopy somehow obsolete now? Of course not. There's obviously value added with the eBook or ole C.D. wouldn't've wanted it. How does Mr. Cohen (the "Ethicist") decide when you've paid enough into the system to begin illegal civil disobedience? Can I pay for the movie in the theater and then download? Better parallel: Can I burn a Blu-Ray b/c I've purchased the DVD at full price? What if I paid a clearance price for the DVD? Have I still paid in enough for someone to forward me a bootleg Blu-Ray in high def? If I've read the book, can I sneak into a theater showing the movie that's not quite filled? Where's the harm in that? Would it really put poor, poor CD out to take along that "more than three and a half pounds" of codex on his trip? Really? Look, if you want a law changed because, in this case, you feel superior enough to remark "the anachronistic conventions of bookselling and copyright law lag the technology", then start lobbying. Now show me one fair law that's anticipated a specific technology perfectly. Sort of another anachronism, ain't it? Honestly, I think eBooks are an interesting way to leverage your ownership of IP into more profit. As long as we're not EULAing hardcopies, knock yourselves out. Furthermore, in this case we have easier solutions for CD. Wait for the g******d eBook to be released. Trade time for money. Read another book on your trip. I just finished Water is Wide* by Pat Conroy on my iPod. You'll enjoy it. If you want to read a new book now, ya gotta pay. Or why not go to your local library and reserve a copy to read while you're waiting. That's a pretty good deal, isn't it? You're not out a buck. Now you read Conroy on your trip and you get to know the latest and greatest from that sick-o King[1]. And guess what, you've already paid for the privilege. Take advantage of it. Had Cohen even so much as said, "Though the risk of being caught is low, it does exist, and in NY the penalty is [X]. I would also say that you need to delete the eBook as soon as you return from your trip, when its marginal utility is gone, and that once the eBook is released, you should stop using this rationalization immediately," I would have felt a little better. As it stands, it bothers me that a representative of what's essentially the record of the United States could show such a simpleton's approach to ethics and encourage his readers to break the law without understanding the ramifications on themselves and the corporations that provide them with their goods. I'm no corporate cheerleader, but when a "ethicist" rationalizes stealing in officially sanctioned e-print, you know society's gauge of right and wrong had long since made a turn for the gutter. [1] Actually, I'm suspicious King is one of the best authors alive. I've read a few of his books that aren't about blood and guts, and they're all exceptionally well written. Still, I tried that city in a bubble book and couldn't get past the first few chapters. SICK. It's all about how you apply yourself, I guess. Labels: blu-ray, business, control, copyright, digitization, DRM, ethics, evil, hats of money, kindle, licensing, nyt, web 2.0 shiite, zealousness posted by ruffin at 4/22/2010 12:05:00 PM

Thursday, February 11, 2010
HTML 5 Video: Apple's Next USB AppleInsider \| Hulu to make videos available on iPad without Flash - rumor: 'The TV shows on Hulu would be perfect on the iPad. There is just one hitch: the iPad doesn’t support Flash, and all of Hulu’s videos currently run inside a Flash player,' states TechCrunch 'But that could change by the time the iPad launches in March. One rumor I’ve heard from an industry insider is that Hulu is working on an iPad-friendly version of its site that should be ready by the time the iPad hits the market,' continues TechCrunch's Erick Schonfeld. If there's one thing Apple's figured out, it's how to leverage the 3-10% of folks that use Macs to push new standards over the tipping point. The iMac all but killed the PS/2-serial port mismash that came before it. USB was happening, but it wasn't until the iMac supported only USB (and at the same time merged the Apple peripheral market with WinPC) that USB became the dominant connection tech. Compare to Apple and Firewire, for which Apple collected license fees for a while. Firewire didn't go nearly so well as USB because firewire was an add-on, not a go-for-broke replacement for USB. USB 2.0 became so much more dominant that Jobs himself was eating his Firewire a few years after its introduction. Honestly, to make Firewire work, Jobs probably should have killed USB and gone 100% Bluetooth for peripherals, but Bluetooth wasn't ready for the weight then, and it's not now. Now, the iPad is killing Adobe. I'm not sure what Jobs has against Flash other than it's not something Apple controls. But he is killing it. HTML 5 video support in Firefox is a better solution, imo, than closed Flash plugins. It's, in a sense, a level less complicated. Once codec compression, speed, and resources are up, up, and down, respectively, to Flash levels, why not use HTML 5? The iPad helps make that "why not" an easier question to answer. If I'm worried about paying licensing for encoding and having end users have the proper Flash plugin installed, now I have a good excuse (aka, "business plan") to develop an alternative: a new platform that only supports the alternative that helps me hedge my bets against Flash. I can develop for HTML 5, pull an id and Quake 3 (where they released a test version of their latest game on Mac first) with the iPad to help bug test a relatively stable platform, and be ready to roll things out if Adobe doesn't give me a sweater deal. Reminds me a lot of AOL and Mozilla. At one point, AOL's licensing deal with Microsoft over using IE as the browser engine in AOL's client was coming due. AOL actually released a version of AOL based on Mozilla for the Mac, showing Microsoft that they didn't really need IE. MS no longer had AOL over a barrel, and a licensing deal was worked out. And b/c of this, in large part, we now have Firefox. AOL's bet played out perfectly for us, though perhaps not so well for them. HTML 5 video is the same move all over again. s/Microsoft/Adobe. s/AOL/[Hulu\|YouTube] Labels: aol, apple, control, F/free, hats of money, id, iPad, licensing, microsoft, mozilla posted by ruffin at 2/11/2010 10:28:00 AM

Sunday, November 22, 2009
Is there such thing as "practically free" software formats? I enjoyed thinking through Should we fight for Ogg Vorbis?, a contribution to the Linux Journal by Glyn Moody back in 2007. The most problematic statement in the piece, I think, is this one. So my doubts about this campaign have nothing to do with any weaknesses in Ogg. It's just that I wonder whether this is really something the free software world should be expending much energy on when there are other more pressing problems. Whereas DRM and software patents, for example, are manifestly and unequivocally bad for free software (and indeed for everyone), that doesn't seem to be true for the MP3 format. Is there a reason to have an open and free format when patent holders don't seem to care much about the folks who are making free software and aren't paying royalties/license fees? Rather, aren't there more pressing places where license holders are worried about enforcing patents where someone could be turning their OSS coding resources? I'm not sure how to feel about this one. I know that I'm getting to the point that I prefer PDF over any other file format for printed works. I'm so freakin' tired of dealing with the way doc, docx, rtf, etc keep fookin' slightly whenever I open them in the wrong application. I used to make do with Microsoft Word, 1998 and 2000, and as long as those apps kept working I figured I'd make do. They don't work so well any more. Now that Word 2100 or whatever it's at now can save in pdf, I'd rather just see pdf files. It's harder for me to edit a pdf than even a wacky docx at times, but there are a wealth of fairly reasonably priced apps that'll allow one to mess with pdfs. At worst, I just print them out and scrawl. Perhaps ODF is the best alternative, but PDF is the practically open format that seems to be doing best, and I don't even have to Google LAME to display it on most OSes. Does this disinclination to support ODF more directly comprise my politics? Yes, I believe it does. We need a standard that will display well outside of its contemporary platform, and display that way for the foreseeable future and beyond. That seems to be pdf to me (and yes, I realize pdf can sometimes be no more descriptive than avi; you really don't know what's in the wrapper. Again, egg + face). Still, is there "practically free" that should be good enough? I'm not sure. I don't like the mp3 reasoning any more than I did for gifs years earlier or pdfs, even after they've been declared an open standard (thanks wikipedia) in 2008. OOXML is open too, you know. Yet there's a certainly practicality to using these formats not designed to be open to humans and machines at the same time. I hate bluffing. Is GNU/hurd ready yet? Labels: control, copyright, DRM, ethics, F/free, licensing, pseudo-academic, Word posted by ruffin at 11/22/2009 01:00:00 PM

Friday, September 11, 2009
The presumption of representation From the iTunes 9 readme, emph mine: This software may be used to reproduce materials. It is licensed to you only for reproduction of non-copyrighted materials, materials for which you own the copyright, or materials you are authorized or legally permitted to reproduce. If you are uncertain about your right to copy any material, contact your legal advisor. Corporate society in the United States assumes that the standard citizen has, in theory, access to resources that most don't. In fact, they assume that real entities (people) have the same sorts of resources the collective, ficticious, legal entities (corporations) have and utilize on a day to day basis. When these assumptions are made, the individual is hurt. Who is my legal advisor? If we hadn't built a house, I'm betting I wouldn't even be able to say "my lawyer" with any confidence. Even then, I'm not calling to ask about CD ripping. He's nice enough in this case that he'd probably talk to me for 10-15 minutes without charging, but that's rare in my limited experience. Very rare. When corporations assume people operate as they do, with the same sorts of access to legal representation, cash and credit lines, and other resources that its employees, other than a few past CEOs, would never have on their own, they should be, in select cases, held liable for providing such little legal guidance for their tools. If ripping a CD is illegal -- ever -- Apple should tell its users in the readme that nobody short of weirdos like me read anyhow. And when iTunes' installer has a button labeled "Read license" that doesn't require clicking before the app installs, well, they should be liable for whatever users do after moving forward. I don't like the RIAA. I don't like it much at all. I despise the current length of copyright protection that essentially neuters the growth of the public domain. But when it comes to such willful negligence as you see in quotes like the above and installers like iTunes 9's, I can't help but wonder how protections and rights in the US have swung so far over into the corporations' favor. A hint: The people, by not staying vigilant, let it happen. I'm not sure how one makes protections shift back. Labels: idiot, iTunes, licensing, public domain, Randmo posted by ruffin at 9/11/2009 11:57:00 AM

Tuesday, July 07, 2009
I've broken Netflix For months now, Netflix hasn't so much as ventured a recommendation -- or, if it has, they have been few and quickly dismissed. This despite my having rated 444 movies. Seriously? You've got nuttin? If I've got Bloodrayne on my list, isn't it likely I'll watch anything? Maybe my queue's long enough and my plan sorry enough they don't feel they need to entice me into using Netflix more. Still, I'd expect recommendations and an upsell. In other news, I'm at least temporarily tired of blogging about much of anything, really really like what I've seen of Paintbrush for OS X (which does what Seashore should have done for OS X: provide a Paint replacement), and believe that Ogg Theora will eventually be used more for commercials than anything else. That's right, the new built-in video codec in Firefox will be used to ensure everyone in that browser sees what are now Flash adverts without a hitch. Until bandwidth costs < licensing, I'm not sure why anyone would walk away from what's already on the scene -- Flash, Silverlight, h.264. In fact, Apple's serving the h.264 Kool Aid as quickly as possible, putting hardware support into every consumer Mac it makes in addition to the iPhone. I'd like to see Theora do to video what png did for images on the net. Still, the only place they make sense -- file size is small enough and the desire to get the message out there more pull than push -- is advertisements, and even then only for adverts being pushed to folk with Firefox browser strings. Labels: apple, control, copyright, firefox, iphone, licensing, paint, web 2.0 shiite posted by ruffin at 7/07/2009 12:47:00 PM

Wednesday, July 09, 2008
Music distribution approaches (already was) GPL-ish I'm going to go out on a limb and say that music distribution services are essentially in the same position as Red Hat. They can't really sell any digital good that's their unique content. All they can really do is support its use and integration with hardware they also sell. I've said before that the reason I go to iTunes' or Amazon's mp3 stores is because it's essentially worth a buck not to have to search for a particular tune. In my case, "search" means look for, order, and wait for a used CD to come in. In other cases, it's probably the time to fire up Limewire or whatever the kids are using now, and find a copy of a song with a bit rate high enough to be worth listening to -- or maybe to go next door and rip their suitemate's CD. There's no real way to protect the music anymore, the same way Red Hat, etc, can't protect the changes they make to Linux. All you can do is provide superior service and hardware integration. Labels: DRM, licensing, online distribution posted by ruffin at 7/09/2008 01:01:00 PM

Wednesday, August 01, 2007
Fake Steve on FreeBSD The Secret Diary of Steve Jobs has this to say about the inclusion (and closing up) of FreeBSD in OS X: And I'm a big fan of open source software and we use a lot of it at Apple (wrapped up in our own proprietary code and sold for big money, of course -- thanks, suckas!) I continue to have very little idea why anyone would release an OS worth of code using a BSD license other than the hope to reach nirvana. The difference between what's happened with FreeBSD (sucked into OS X; good luck seeing that again) and KHTML (sucked into Safari, but LGPL licensed, now unforking with the original codebase) is the perfect compare and contrast for how Free software works in a capitalistic society. Buy used books, listen to RMS 80% of the time, and, at night, always wear white. Labels: apple, lgpl, licensing, rms posted by ruffin at 8/01/2007 11:10:00 AM

Wednesday, August 02, 2006
FreeBSD, give until it helps? I somehow managed to create a small ripple at apple.slashdot.com with a post a while back essentially giving my standard GNG rant "against" BSD licenses. I've been meaning to find time to blog about this adequately, but haven't yet. So here goes the important point to me... It appears that there are some that believe that FreeBSD accomplishes its work precisely by being so "free" that people appropriate it for their own commercial purposes -- but the initial work has so much momentum, its legacy contributions essentially create an interoperable conversation or standard. More simply, FreeBSD's openness meant Apple could choose it as the base of OS X, versus GNU/Linux. FreeBSD's stability, etc, meant that Apple did choose it. And now, no matter how much OpenDarwin whines (again, my characterization of one type of FreeBSD champion-er) about not getting the sort of collaboration from Apple they felt they deserved (apparently some packages from Apple wouldn't even compile, iirc), more Mac users now know how to use, say, the "ls" command than ever before. Apple users now know *NIX, where before they were simply masters of the Extension Manager. This is a very interesting point. It's a much less idealistic motivation than my reasons behind GNG, and much MUCH less idealistic than Stallman's for GNU. One must also admit that it's been at least partially [very] successful. I'm not sure of an example more powerful than OS X & Free BSD, but that's a very convincing one. Oh, don't worry. I'm not convinced, personally. Perhaps many FreeBSDers feel the way I've described, above, but many don't. Thus OpenDarwin's frustration. My post, entitled, "BSD's Fault," was only meant to speak directly towards OpenDarwin's predicament, which was the subject of the Slashdot story about which everyone was ostensibly speaking. If you don't want to end up where OpenDarwin is, failing because you expected a commercial company to return the favor they received when grabbing very open BSD'd code, then you don't use nor support BSD. I'm also one that believes GNU/Linux's (and also Tomcat's and MySQL's, etc) adoption by many businesses is a much more powerful set of success stories than those of BSD's. The GPL does a decent job of delineating where its influence ends. With GNU/Linux, etc, where the app stops, many have been able to make a living programming new code (Perl, php, etc) or creating very impressive documentation (ora.net, anyone?) about these more closed, Freely licensed techs. That these techs are more prevalent on the server side means they are relatively invisible compared to, say, OS X, but their power in dollars (as a quickly picked, stereotypical measuring stick) is much much greater. I believe Apple should feel their ethics demand they give more back to BSD, and help ensure that Darwin, OS X's FreeBSD, compiles for most serious hobbyists. Apple, not caring what these hobbyists add to apps in the future (and in some respects, I don't blame them), is treating and will continue to treat them as third class citizens. That's the fault of an idealistic interpretation of the way BSD works that GNU (or GNG) protects programmers from experiencing, and they still manage to get a whale of a lot of work done. This leaves me wondering if GNU will ever be a serious contender for the desktop. I don't know. Labels: F/free, gng, licensing posted by ruffin at 8/02/2006 01:09:00 PM

Sunday, July 10, 2005
The LGPL and Java The LGPL and Java: If you distribute a Java application that imports LGPL libraries, it's easy to comply with the LGPL. Your application's license needs to allow users to modify the library, and reverse engineer your code to debug these modifications. This doesn't mean you need to provide source code or any details about the internals of your application. Of course, some changes the users may make to the library may break the interface, rendering the library unable to work with your application. You don't need to worry about that -- people who modify the library are responsible for making it work. That's why the LGPL is no good; this is an exploitable hole to allow people to modify a library and start reverse engineering your code, it would seem. I'm not real sure why that's bad, but I do know it's not something I'd particularly want butting up next to my proprietary code. Perhaps this is why a number of pay-to-play companies run a secondary, "Free as in Mozilla Public License" codebase. Having grown out of the need to recreate Netscape, not to become part of an idealistic Hurd, the MPL watches companies' backs well. Too well. Here's the metaphor I heard that got me interested in the LGPL (IANAL): You're making a calculator. You've got every button down except the square root button. You decide to look at external libraries to save time. GPL Square Root -- You can embed the function and sell your software, but now every bit of your calculator, plus, minus, 1s, 5s, the whole thing, must be released with full source code. Now anyone can take your code and sell your application. Anyone. Not necessarily a bad thing, unless you like to make money. Everyone can use your first-rate natural log function, but do remember that they, too, can't sell this [legally] without showing the code to anything they've added. Great for promoting passive collaboration. LGPL Square Root -- Originally, before the Java issue, I thought the LGPL went like this: You can add this square root function, release and sell your app. You only have to include source to the square root function; everything else can stay proprietary. Let's say, however, that you find a bug in the square root function, say it thinks the square root of 49 is 8, but gets everything else right -- or that it only goes to 5 decimal places. If you fix the bug or extend its accuracy to 10 decimal places, you have to release the source to those improvements with your app. Personally, that's great so far. I like it. If you only have the function b/c of someone else's contribution, you owe it to him/her/them to release your improvements for all to use. It enables passive collaboration, but still allows you to protect your investment. Nobody "deserves" your fancy natural log function if you don't want to give it to them. It was entirely your work. Keep it. The "rights" to the code are protected from both sides, and fairly. You'd also owe the 'LGPL' a big thank you, of course, but you get the point. BSD Square Root -- Take it. Sell your app. You might have to acknowledge the project whose work you took, but that's it. It's yours. Take it from 5 to 10 decimal places of accuracy, heck, take it to 2000, and you don't have to share a keystroke. Great for greedy people who don't want to be bothered, but I'd hope they'd feel a little bad about taking work to which they didn't -- nor will -- contribute. MPL Square Root -- You can take this square root function, release, and sell your app. In fact, you could (afaict) take your own square root implementation, take a few files from the MPL version, and put those together to be released. Your own, wholly original code could stay closed, and only changes you made to MPL files would need to be released. Initially, I thought the MPL was like BSD. As is obvious here, it's not. It's sort of like the LGPL in that your code can be released in support of and as part of a closed source product, but it's "file-level" in that any file of your library could potentially be taken and used in a new context. But it's as close to the LGPL as you can get if you want to allow someone to compile your code alongside proprietary code, and it doesn't seem exceptionally evil. It's a good deal less protective than the LGPL, but just good enough that I still like it. (MPL section added 20131011.) Okay, comparison over. But now, following the relatively dated quote about Java issues, above, (iff the Java issue is legit) the LGPL forces anyone using the square root function to allow anyone to decompile their entire app that uses the square root function to "bug check" against their code. That's simply not a place I want people going. Short of the square root function, it's my code. Stay out. If you want my code to work, use the old library that thinks 49's square root is 8. If you can upgrade the lib to work with 49, great, but why should I have to give you any rights to my code downstream of the square root function to make that happen? You didn't make the investment in app, and I want to protect it. That should be my right. You shouldn't get to peek at my natural log function if I've used LGPL code elsewhere. If your change to the square root -- and who knows how inane that change might be; perhaps it also now doesn't do 36 correctly -- breaks compatibility with my app, that's my problem. I'll fork the square root lib if I have to. Oh well, guess I'll eventually have to start my GNG (GNG's Not GNU) movement with an LGPL done, if not right, then fair. Might be as easy as slashing Section 6 in the LGPL; I'll have to check sometime. Now, on to UO (I'm on sabbatical from WoW)! Labels: java, lgpl, licensing posted by ruffin at 7/10/2005 02:27:00 PM

<< Older \| Newer >>