zlacker

[parent] [thread] 29 comments
1. lostms+(OP)[view] [source] 2020-05-27 23:10:09
That is exactly why I am not open sourcing my .NET bindings for TensorFlow.
replies(2): >>Wowfun+a8 >>solark+dc
2. Wowfun+a8[view] [source] 2020-05-28 00:10:14
>>lostms+(OP)
I don't know the details of your project, but what if you open sourced under the GPL?

I feels as though the world has soured on the GPL in recent years, but whenever I see this type of sentiment—that open sourcing work is just a gift from small developers to big tech companies, or something thereabouts—I think, wouldn't the GPL solve that problem?

Sure, big companies can still use GPL'd code, but they're forced to give back as much as they take, which is exactly the outcome you want.

(This doesn't necessarily apply to the situation with AppGet, however.)

replies(3): >>bluGil+F8 >>lostms+zb >>na85+Ec
◧◩
3. bluGil+F8[view] [source] [discussion] 2020-05-28 00:15:10
>>Wowfun+a8
Big companies are more likely to give back to non gpl code in my experience. The gpl scares them enough that when they use it they isolate it carefully so they don't have to make changes. Non gpl code (depending on license terms of course) get embedded where it is easier to need changes and then a desire to keep up with some other feature means they contribute back.

Of course either way they are a business and will be careful. Their core business will not be contributed.

replies(2): >>SeeThe+ZB >>rustic+lL
◧◩
4. lostms+zb[view] [source] [discussion] 2020-05-28 00:42:20
>>Wowfun+a8
That would not have helped MongoDB. I considered their new license, but it seemed tailored to their specifics. So maybe more research is required.

Perhaps a variant of CC-BY NC would be better, but it is not tailored to code.

replies(2): >>colejo+2f >>fabian+rn
5. solark+dc[view] [source] 2020-05-28 00:48:44
>>lostms+(OP)
How much will that really change the likelihood of MS reimplementing it?

They didn't take any of the AppGet code, just the ideas.

replies(2): >>colejo+9f >>lostms+Gh
◧◩
6. na85+Ec[view] [source] [discussion] 2020-05-28 00:52:38
>>Wowfun+a8
By Eben Moglen's own admission[0], the GPL has never been successfully enforced in a court of law, even against criminal infringers.

It seems toothless to me. A relic from a byegone era when companies cared about ethical behavior.

[0] https://www.gnu.org/philosophy/enforcing-gpl.html

replies(3): >>colejo+Ie >>ASalaz+cg >>ryukaf+At
◧◩◪
7. colejo+Ie[view] [source] [discussion] 2020-05-28 01:09:08
>>na85+Ec
The GPL has been used successfully in courts of law. It’s not common as most violators just open source when called out, or settle when sued, but it has worked before. A famous example is D-Link using the Linux kernel in their routers. After a lawsuit in Germany, D-Link had to release their changes.[0]

I don’t understand this idea that a license that the author willingly released his code under is unenforceable, but EULAs are?

[0]: https://web.archive.org/web/20141007073104/http://gpl-violat...

replies(1): >>tracke+4j
◧◩◪
8. colejo+2f[view] [source] [discussion] 2020-05-28 01:12:41
>>lostms+zb
What’s wrong with AGPL?
replies(1): >>lostms+Qf
◧◩
9. colejo+9f[view] [source] [discussion] 2020-05-28 01:13:54
>>solark+dc
Depending on who you ask, ideas are still protected by intellectual property laws. Patents are a good example.
◧◩◪◨
10. lostms+Qf[view] [source] [discussion] 2020-05-28 01:19:53
>>colejo+2f
AFAIK AGPL permits linking without disclosing linked code like LGPL does. Meaning they can still distribute it as their own, train models and serve them with a simpler framework. Just can't build services powered directly by it.
replies(1): >>leonid+Gk
◧◩◪
11. ASalaz+cg[view] [source] [discussion] 2020-05-28 01:22:23
>>na85+Ec
Ahem https://www.gnu.org/philosophy/enforcing-gpl.html
replies(1): >>na85+Dn
◧◩
12. lostms+Gh[view] [source] [discussion] 2020-05-28 01:32:59
>>solark+dc
They might. But it is not as simple as WinGet. It would take a team or two a couple of years to reimplement. ML.NET is 2 years old now, still has very limited deep learning capabilities, and is barely visible in trend graphs.

While they would be doing that, Azure will continue to lag behind (still does not even offer T4) on GPU offerings in comparison to AWS and GCP, because there's no compelling reason to run custom deep learning workloads on Azure. Azure is loosing $$$ millions of potential revenue to that. It would be a totally different story if .NET folks would get full TensorFlow and/or PyTorch in place of partial ports available now.

◧◩◪◨
13. tracke+4j[view] [source] [discussion] 2020-05-28 01:42:52
>>colejo+Ie
Of course with all/most US routers generally taking only signed ROMs, it's mostly useless to even have the code.

I wish they'd just create something closer to tomato than the flashy, less useful interfaces.

replies(1): >>colejo+Ko
◧◩◪◨⬒
14. leonid+Gk[view] [source] [discussion] 2020-05-28 01:56:34
>>lostms+Qf
AGPL is as strict as GPL. It's just that GPL allows you to modify the source, serve them as a web application and not have to disclose the code.

AGPL addresses that by putting code distributed as a web application in the same category as compiled code.

replies(2): >>hnick+Ou >>lostms+Sw
◧◩◪
15. fabian+rn[view] [source] [discussion] 2020-05-28 02:27:44
>>lostms+zb
The P2P Foundation has some writings on licenses they label as "copyfair", "copyfarleft", and "copyjustright"

https://wiki.p2pfoundation.net/CopyFair_License

https://wiki.p2pfoundation.net/Copyfarleft

https://www.metamute.org/editorial/articles/copyfarleft-and-...

replies(1): >>lostms+cH
◧◩◪◨
16. na85+Dn[view] [source] [discussion] 2020-05-28 02:29:31
>>ASalaz+cg
Have you actually read the essay to which you're linking?

>Meanwhile, much murmuring has been going on in recent months to the supposed effect that the absence of judicial enforcement, in US or other courts, somehow demonstrates that there is something wrong with the GPL, that its unusual policy goal is implemented in a technically indefensible way, or that the Free Software Foundation, which authors the license, is afraid of testing it in court.

(implying it has never been tested in any court)

>We do not find ourselves taking the GPL to court because no one has yet been willing to risk contesting it with us there.

(Explicitly stating it has never been tested in court)

>I have assisted free software developers other than the FSF to deal with such problems, which we have resolved—since the criminal infringer would not voluntarily desist and, in the cases I have in mind, legal technicalities prevented actual criminal prosecution of the violators—by talking to redistributors and potential customers.

And lastly, the money quote. In the cases of criminal, malicious infringement, they were not able to prosecute.

replies(1): >>triple+DA
◧◩◪◨⬒
17. colejo+Ko[view] [source] [discussion] 2020-05-28 02:43:08
>>tracke+4j
That problem is called TiVoization. It’s addressed with the 3rd version of the GPL. Sadly, Linus is adamantly against that clause.
replies(2): >>chii+Ev >>tracke+Mu3
◧◩◪
18. ryukaf+At[view] [source] [discussion] 2020-05-28 03:38:13
>>na85+Ec
Note the date on that essay: it's from 2001. It has indeed come up in court in the last 19 years. For example:

https://wiki.fsfe.org/Migrated/GPL%20Enforcement%20Cases

◧◩◪◨⬒⬓
19. hnick+Ou[view] [source] [discussion] 2020-05-28 03:52:42
>>leonid+Gk
If my web server calls out to AGPL software (such as executing a transform on a PDF using Ghostscript) as an external binary is this covered even though the AGPL code is not directly compiled, linked or modified? I'd assume so otherwise it would be a very easy loophole and code could be firewalled into separate modules but I'm not sure what exact language covers this in the license.
replies(1): >>chii+yv
◧◩◪◨⬒⬓⬔
20. chii+yv[view] [source] [discussion] 2020-05-28 04:00:47
>>hnick+Ou
It's untested in court.

GCC is written specifically to prevent this sort of firewalling of the compiler to prevent it from being plugged into another piece of software and avoid GPL requirements.

replies(1): >>hnick+eB
◧◩◪◨⬒⬓
21. chii+Ev[view] [source] [discussion] 2020-05-28 04:02:02
>>colejo+Ko
> Linus is adamantly against that clause.

oh i didnt know he was adamant against the clause - i thought that he didn't want to force it upon the many existing users of linux.

replies(1): >>colejo+Fw
◧◩◪◨⬒⬓⬔
22. colejo+Fw[view] [source] [discussion] 2020-05-28 04:14:20
>>chii+Ev
https://youtube.com/watch?v=PaKIZ7gJlRU

He also has refused to release any of his projects under GPLv3 (or even “v2 or later”).

replies(1): >>kasaba+z16
◧◩◪◨⬒⬓
23. lostms+Sw[view] [source] [discussion] 2020-05-28 04:16:09
>>leonid+Gk
You seem to be correct. Still the note about training a deep learning model using the framework, then publishing it without applies.
◧◩◪◨⬒
24. triple+DA[view] [source] [discussion] 2020-05-28 05:04:11
>>na85+Dn
Your implication that "untested in court" means "likely unenforceable" is quite wrong--unless you see lots of people openly violating a license in cases where it would be economic to pursue them, the more likely explanation for "untested" is simply that the accused infringers get competent legal advice, and comply without wasting their money on a losing court case. Though as other comments note, since the writing of that essay someone (D-Link) finally did refuse to comply without a court order, at which point Harald Welte took them to court and D-Link indeed lost.

And what do you think "not able to prosecute" means? District attorneys (or non-USA equivalents) decide when to prosecute crimes, not private copyright owners. It's very rare for the criminal justice system to intervene in complicated white-collar stuff, especially when a straightforward civil remedy is available. A legal realist might say that means the GPL--and indeed most copyrights beyond those infringed by warez/torrentz sites--is effectively unenforceable criminally, and would in a useful sense be right; but it's enforceable civilly, so no one cares much.

◧◩◪◨⬒⬓⬔⧯
25. hnick+eB[view] [source] [discussion] 2020-05-28 05:13:16
>>chii+yv
Good to know, it'll be interesting if it ever does get tested.

Even then there's always the analogue/human loophole I suppose, similar to captcha mechanical turks.

◧◩◪
26. SeeThe+ZB[view] [source] [discussion] 2020-05-28 05:20:27
>>bluGil+F8
So it's not worth it. Why be less careful than companies in the hope they will be good citizens, when it's easier to be careful and not have to hope?
◧◩◪◨
27. lostms+cH[view] [source] [discussion] 2020-05-28 06:08:11
>>fabian+rn
These are pretty interesting. Thanks for the links.

I like how this general direction is discussed a lot in France, and sad about learning a language taking so much effort.

◧◩◪
28. rustic+lL[view] [source] [discussion] 2020-05-28 06:46:44
>>bluGil+F8
Dual licensing helps. AGPL or Paid licence.
◧◩◪◨⬒⬓
29. tracke+Mu3[view] [source] [discussion] 2020-05-29 00:12:16
>>colejo+Ko
Unfortunately, it's come out from the FCC, not just the hardware companies in this case, and it really sucks... Raally need to drain the swamp in the FCC.
◧◩◪◨⬒⬓⬔⧯
30. kasaba+z16[view] [source] [discussion] 2020-05-29 19:29:24
>>colejo+Fw
Which is too bad because CDDL is compatible with GPLv3 thus that would make integration of ZFS possible
[go to top]