4 ms·
IANAL but this seems like a pretty clear cut case of copyright infringement, I.e. a situation where an ordinary person might be confused into thinking that Powe
by asah 10y ago
IANAL but this seems like a pretty clear cut case of copyright infringement, I.e. a situation where an ordinary person might be confused into thinking that PowerShell curl/wget provides the real thing, but does not.
Although I suppose that PowerShell could fall under "parody."
EDIT: this was sarcasm.
- kuschku 10y agoYou mean trademark infringement.
- jdbernard 10y agoPowerShell isn't copying the curl or wget code. This would be more like trademark infringement but I don't know if curl or wget are registered trademarks. Still not a cool move.
- dublinben 10y agoYou don't need to register a trademark to have trademark protections. If you use a 'mark' in 'trade' and can demonstrate consumer confusion, then you have a case.
- wyldfire 10y agoThat's absolutely not the standard for copyright infringement (under US code). If curl or wget were trademarked, however, it could definitely be "confusingly similar" which would probably make a valid trademark infringement claim.
- dublinben 10y agocurl and wget automatically earn trademark protection simply by using a 'mark' in 'trade.' They do not need to be registered to earn this protection.
- thomasahle 10y agoAlternatively you could see curl not as a program, but part of the shell API. Microsoft has then reimplemented the API. Sure, from the Oracle vs Google case, even this may be infringement, but only in a more anal way.
- erlehmann_ 10y agoThe point is that they have not reimplemented it. If the Microsoft versions of curl and wget were compatible with the original utilities, there would not be any problem.
- saurik 10y agoWhich actually also came up in the Oracle case: Google claimed to be building a compatible implementation of the Java API, but they couldn't come up with a single example of an app that worked on the JVM which could also run on Dalvik, or the other way around: they were pulling the exact same "embrace extend extinguish" stuff with the Java API that Microsoft had with Visual J++. In that case Sun sued Microsoft and won, leading people to cheer from the sidelines.
- opendomain 10y agoWhat you are describing is trademark, not copyright. There is no trademark on 'curl' or 'wget', so there is not legal status to force anyone to remove features based on these names.
- rgbrenner 10y agotrademarks (at least in the US) are acquired automatically when they are used (just like copyrights). you can't use someone else's name just because they didn't file a trademark app.
- ihuman 10y agoAre curl and wget copyrighted? I thought they were free software under a copyleft-style license.
- Sylos 10y agoHe mixed up copyright and trademark, as far as I can tell. And trademarking is then no problem with copylefted software. For example, Linux is trademarked by Linus Torvalds and Firefox is trademarked by Mozilla. And at least in the case of Firefox, it's actually pretty important to have, as otherwise someone could bundle malware with Firefox and offer that for download as "Firefox", with nothing that Mozilla could do about it.
- wyldfire 10y agoYes, their source code is protected by copyright. What does this mean? It means that the people have granted the author of that code a limited-term monopoly on making copies of that source code. In turn, as you said, the author has made it available under a restrictive license requiring redistribution of source. What did Microsoft do? Well, apparently they created aliases that mimicked some subset of the curl/wget behavior for PowerShell. This might allow simple *nix shell scripts to work with little or no porting to PowerShell. The names of the executables are not protected by copyright. Although -- who knows, if APIs can be considered protected by copyright maybe the names of the binaries themselves are a shell-style API. It seems like a case you'd be unlikely to win. :/
- Kihashi 10y agoThat doesn't mean that they aren't copyrighted. It just means that you can freely use them under the terms of their copyright license. If the code for them were not copyrighted (usually referred to as Public Domain), then you could do whatever you wanted with them-- things like making derivative works and using a commercial license.
- Lennie 10y agoLet's be very clear here. This is not true: No copyright notice == Public Domain. Everything You make gets Your copyright by default. This means the default is: nobody can use what You make. You can put other copyright terms on your works by adding a copyright notice. You can put something in the Public Domain by adding a notice to your work. Things can also become Public Domain automatically after a certain amount of time. But that time has been extended several times in the US to something unreasonable: like 75 years after the death of the author. You can thank Disney/Mickey Mouse for that.
- dalke 10y agoThere's a problem with your argument (replacing 'copyright' for 'trademark' as others have pointed out). What level of name protection applies to command-line tools? For example, can the GNU project replace existing Unix tools with their own implementations which have the same executable name?
- wyldfire 10y agoyes, absolutely they can. In fact, they almost certainly DID! GNU's legacy was in creating tools that were compatible reaplcements for Solaris/Unix/etc ones.
- umanwizard 10y agoJust because someone did something without getting sued doesn't necessarily mean it was legal.
- paulmd 10y agoI think you're thinking of trademark infringement. Copyright infringement would be if Microsoft had used GPL'd code (for example) in violation of its license. I don't think there's any suggestion that's the case here. Your contention that there is a "real thing" would have tremendously negative impacts for the Linux/BSD community. Your typical Linux/BSD contains many utilities that share names and functionality with closed-source System-V utilities but are not in fact the "real thing" (in their case, for licensing purposes). Is everyone supposed to go back and purge those just because they aren't perfectly equivalent in options and functionality to the original System-V, lest someone like the SCO group go after them legally? I don't think there's any doubt that PowerShell provides a bad clone of wget and curl, as there are many complaints about it. And as a matter of user-friendliness there should be a way to easily access the real executable if desired (I can't comment either way on if there is, I don't know PowerShell). However, I see Microsoft as just as within their rights to provide a "wget" alias as a Linux distribution is to provide a "ps" utility to their users.
- saurik 10y agoI feel there is a big difference between making a replacement for a tool that comes with a generic name in a suite and replacing an entire tool. If you are going to make a reimplementation of Microsoft's spreadsheet software that is open source, you don't get to call it Excel. If you make a reimplementation of cURL, you also shouldn't get to reuse the name of their project.
- paulmd 10y agoSo again, what is the difference between making a new spreadsheet app and calling it "Excel", and reimplementing a suite of smaller applications and reusing all their same names? Say, how coreutils steals the name of a large number of SysV utilities? It certainly feels like there's some kind of a difference between the two, but I can't put my finger on exactly what puts one into either category. IMO it may simply be a matter of willingness to sue/defend your trademark. AT&T didn't defend their trademark and now it's genericized. To throw extra fuel onto the fire, let's add Oracle v Google into this discussion. To me, the coreutils are almost like the API of a programming language - you string together application calls just like function calls, to produce a useful application. Oracle v Google established that the API itself is not copyrightable. Should an attempt to defend based on trademark of the function names have succeeded? If not, what differentiates the "ls" utility from Excel?