I was wrong (part 2)

https://liorpachter.wordpress.com/2017/08/03/i-was-wrong-part-2/

by adenadel • 9 years ago

95 38 9 years ago

liorpachter.wordpress.com

I've been about half a block down this road. Thank God, the university office determined it wasn't worth the universities time to invite the lawyers -- and my project is happily open source.

mbreese• 9 years ago

One of my least favorite experiences in grad school was dealing with our school's tech transfer office. They fought hard to avoid using the GPL for licensing anything. Just getting them to agree that a simple (to be published) tool should have an open source license was difficult. (And took many meetings)

Fast forward to two different Bay area universities and it was a completely different experience. Open source licenses were agreed upon over a quick email.

hprotagonist4• 9 years ago

I had a fun time with the library copyright office. 6 months before my thesis defense I emailed them with "hi what's policy and procedure for creative commons licenses on theses". Nobody had ever asked before...

edit: my thesis is cc-by licensed.

dsjoerg• 9 years ago

Thanks for sharing your painful journey!

adenadel• 9 years ago

I can't take credit, I just follow Lior's blog :)

glangdale2• 9 years ago

Good to see this change, but I wish people wouldn't go down this road to begin with.

Speaking as someone who sits in industry now, I have found it almost impossible to successfully conclude negotiations between not-very-interested academics who don't really understand how licensing works (and the staff who supposedly help them) and a big, paranoid company.

When you put some rando commercial license or complicated terms on your (likely tax-payer funded) data or software, what it essentially requires from big companies is a extremely expensive legal review to see what we would be signing up for. This review might be considerably more expensive than the amount being asked for the data/software. What's more, if a small company signs up for this, and ever wants to be acquired by a big company, the review may not happen (small companies and startups are usually more relaxed about this stuff) but will have to take place during acquisition.

Bonus points for when the academics or their representatives wander away mid-discussion.

As a final bonus, I have seen our commercial competitors in this area just essentially fund university researchers to do thinly veiled advertising by accessing the same dataset for an "academic" paper. Meanwhile I sit on my hands respecting "the latter and spirit" of the law.

glangdale2• 9 years ago

oops "the letter and spirit" of the law.

anjc5• 9 years ago

What are these not-very-interested academics expecting? For you to agree to whatever commercial license or...?

I don't get why you'd expect academic institutions to offer you permissive licenses though, when in all likelihood you have your own complicated partnership/IP/licensing processes for industry partners. Why is it ok for them but not for you?

And what difference does it make that their research is tax-payer funded? That doesn't make it free. That means that the institute has an obligation to make returns and, academic research being somewhat open, lock down IP rights with certainty.

folli• 9 years ago

As someone also on the industry side of things, what would be your recommended solution to the problem? On the one hand we don't want the legal mess that licensing entails, on the other hand I'm absolutely in favor of reimbursing the academics in some way. But how?

Some new, standardized, industry accepted commercial license (who will invest the required legal effort to come up with such a license that's accepted by both sides?)? A concept based on voluntary donations (good luck getting that by the finance department)?

anjc5• 9 years ago

Follow up to my other comment...I just noticed from your bio that you work at Intel, and now have to lol at you complaining about academics trying to license commercial products. I presume you don't mind when tax-funds are directed towards academic R&D when it's conducted for Intel (and plenty of it is in Europe) :)

xfer• 9 years ago

Regarding "tax-payer funded data/software" From the blog:

I believed then, and still do now, that it’s right that institutions that support software development should benefit from its commercial use (UC Berkeley receives 2/3 of the royalties for commercially licensed products), that students are entitled to renumeration for software engineering work that does not directly support of their own research goals, and that funds are needed to support specialized personnel who can maintain/improve code and service user requests.

Also i believe, most students in academia are under-paid.

ballenf• 9 years ago

It's off-topic but a very similar argument exists for why the Facebook license + patent type arrangements(1) can cause problems for large companies or small startups that may want to be acquired later.

And I too wish that the Facebook-of-the-future don't go down this road. Number two on devs wishlist should be transparency about exactly which patents cover the code at hand. IMHO.

I draw the analogy due to the similarities of "open source" and government funded findings. Another area where these issues arise is charitable contributions that come with strings attached (like naming a building or agreeing to continue a certain program).

People like getting full recognition for the act of generosity without the cost of their strings being acknowledged or subtracted from the total.

(1) The reality of Facebook is that it's unlikely to ever be an issue, but if the arrangement becomes more commonplace the open source community could become a minefield of unquantifiable risks. I realize this is a slippery-slope argument and is thus easy to overstate.

drefgert2• 9 years ago

There won't be too many comments on this post cause the internet ain't interested when someone says they are wrong. The internet wants to TELL them they are wrong, when they think they're right.

csydas• 9 years ago

I'm not sure that's so true. Everyone enjoys a good argument over what license the author should have used. )

I do get what you're saying but I think people discuss positive things all the time on even the worst sites. Admitting you're wrong isn't all that novel, but the changing of license is.

drefgert2• 9 years ago

So you're saying I'm wrong?

anjc5• 9 years ago

I don't understand why you think you were wrong, based on the article. Surely it depends upon your objectives. If your goal is for researchers to use the software then one license might be preferable, if your goal is commercialisation then another license might be.

In the same way that some researchers wont use software which doesn't have permissive licenses, some companies wont use software which does.

I don't get it. I also don't get why you expected license agreements to be signed and completed in a single day by commercial entities. You couldn't sign and complete a 5 minute coffee date with most company's IP staff within a single day, nevermind have licensing arrangements agreed upon.

ubernostrum• 9 years ago

The wrongness I read from it is a certain level of arrogance, that there can't really be anything too complex in licensing and copyright issues, and all those lawyers just must be posturing or engaging in pointless make-work. The discovery that well-understood standard license terms are preferred for a reason, and that there are very real issues lurking in non-standard license agreements, and corresponding realization that maybe all the people saying "you're wrong" had a point... is a step toward correcting that arrogance.

anjc5• 9 years ago

That makes total sense but it sounds as if his realisation tended towards open licenses being preferable in all academic scenarios, due to a now-more firmly held belief that lawyers etc are a waste of time. This obviously isn't a rule which applies generally, and I'm not sure why it even applies here when his goal was to commercialise, not to please researchers.

dm3192• 9 years ago

Here's the license for stampy, a popular read aligner from Oxford:

12. License ===========

This is a release version. Permission is granted for the normal use of the program and its output in an academic setting, including in publications. If the program is used to generate data for a publication, please cite this paper:

G. Lunter and M. Goodson. Stampy: A statistical algorithm for sensitive and fast mapping of Illumina sequence reads. Genome Res. 2011 21:936-939.

The program itself may not be modified in any way, and may not be reverse-engineered.

This license does not allow the use of this program for any commercial purpose. If you wish to use this program for commercial purposes, please contact the author.

No guarantees are given as to the program's correctness, or the accuracy or completeness of its output. The author accepts no liability for damage or otherwise following from using and interpreting the output of this program. The software is supplied "as is", without obligation by the author to provide any services or support.

http://www.well.ox.ac.uk/~gerton/README.txt

Gatsky2• 9 years ago

Some context:

The blog author just wrote another post which was essentially a hatchet job on a 'rival' software package. One of the major differences between his software kallisto and the rival Salmon was the license - Salmon has a GNU General Public License.

dm3192• 9 years ago

'I was wrong' part 1 was a really a discussion about how wrong / fraudulent some authors were who published in Nature in 2004. It looks like it turned into a bit of flame-fest[1], with the original authors chipping in.

[1] https://liorpachter.wordpress.com/2015/05/26/pachters-p-valu...

Gatsky2• 9 years ago

No I mean this blog post: https://liorpachter.wordpress.com/2017/08/02/how-not-to-perf...

sevensor• 9 years ago

I was an onlooker -- for four months -- as a university IP office struggled to license source code developed at the university, to one of its own researchers. When the drama was concluded, the IP office couldn't find the source code.

dankohn12• 9 years ago

I'd like to commend the author for realizing that a commercial license was not achieving the goals he had for his software and being willing to own up publicly about why and how he changed directions.

However, while I have used MIT/BSD-style licenses in the past, I can no longer recommend them, as they potentially open up software users to malicious submarine patent attacks. That is, a contributor can intentionally contribute code that (purports to) read on a (not-even-issued-yet) patent, and then later threaten to sue users of the software.

That's why I instead recommend Apache-2.0. My arguments are more fully described here: https://www.cncf.io/blog/2017/02/01/cncf-recommends-aslv2/

paulgb• 9 years ago

Interesting, has this actually happened?

dankohn12• 9 years ago

Yes, a patent-owning entity has tried to contribute code reading on that patent. Apologies that I can't give the details.

nnutter• 9 years ago

BSD does not have a patent license clause so if Lior, et al. have a patent on any part of the algorithm then won’t commercial use will still be ambiguous? And even if they don’t it could still be using (infringing) on a patent of a third party. Basically software patents make using any software a risk but possibly more so when you’re using something based on novel ideas.

Edit: GPL and ASL have explicit patent clauses (obviously only applies to patents the software licensor has authority to license).

I was wrong (part 2)

Related Stories