28 ms·

Redis will remain BSD licensed

448 points by antirez 8y ago

stingraycharles 8y ago
I’m happy to hear that Redis will remain BSD, but it’s a bit weird to play the “fake news” card. Can you really blame people for interpreting this the wrong way? Especially when the original announcement was specifically targeting cloud providers, who, afaik, are just running vanilla redis. If there is anything that we can learn from this, is that a little bit of forethought when putting these announcements out there is required, and blaming this on the readers is not fair.
- lbriner 8y ago
  The parts that I noticed were missing from the original announcement include more specifics about the parts of the system that would change licence and also whether the intention was to charge people to use these new parts. It was unclear whether the intention was to stop these cloud providers or to make them pay in some way for what they were doing.
  - antirez 8y ago
    Yep, now finally this is updated to specify exactly what changes license. Incidentally: nothing I ever worked to directly. It would be way cooler if also things like RediSearch or RediGraph could still be totally open source, but it's apparently very hard, Common Clause at least will allow it to be free-as-beer and with the source code available.
    - yarosv 8y ago
      > Common Clause at least will allow it to be free-as-beer and with the source code available. The license is very vague. According to this > a product or service whose value derives, entirely or substantially, from the functionality of the Software. What if I have big online shop with huge catalog and my software value "derives from the functionality of RedisSearch"? Do my only option not to host it at Redis Labs?
      - derefr 8y ago
        > my software value "derives from the functionality of RedisSearch" Your software might just literally be 99% RedisSearch, but you're not making money from your software (e.g. by selling it); you're making money from operating your software, combined with content (e.g. by using it to facilitate e-commerce transactions.) You'd be in trouble if you wanted to white-label "your technology" and sell it as a platform for other people to run their own online shops on, because it's not really "your technology" to sell. But if your play isn't specifically that, then you're fine. Also, > Do my only option not to host it at Redis Labs? this wouldn't be an option. The whole point of this license is that the software "RedisSearch" is owned by Redis Labs. Hosting it with Redis Labs themselves, or not, is irrelevant. You have to license the software (if you're deploying it yourself)—or rely on your DBaaS provider to license it for you (if they're deploying it.) Basically exactly like a software patent, come to think of it.
- meowface 8y ago
  I don't think he's blaming people, but it does seem odd if he had no input at all into the Redis Labs announcement.
- antirez 8y ago
  Honestly my colleagues did a poor job with such an announcement there, and I did not get a chance to read it (which is actually a good thing, it means that the OSS division, I'm part of, is independent and can work to the Github repositories without caring about what the commercial side does). But it was kinda of a manifest to announce that certain things made inside Redis Labs are now Common Clause licensed. In doubt, why not asking? Also people that follow me and Redis for years getting trapped into this idea is very strange, I would never do that and without explaining, prior notice, and so forth. So yep the Common Clause page at Redis Labs sucked but in doubt ask instead of creating news out of no actual information.
  - l2dy 8y ago
    The abbreviation "CC" will be confusing if Commons Clause got popular… We already have Creative Commons licenses.
    - antirez 8y ago
      Agreed, fixed.
      - millette 8y ago
        Utterly confused. Does this have anything to do with Creative Commons licences? That's what the articles calls it in its first and last sentences.
    - jpobst 8y ago
      Particularly because the linked post calls it "Creative Common" twice and "Common Clause" elsewhere.
    - tango24 8y ago
      When I see CC, I also tend to think it means Common Criteria : https://en.m.wikipedia.org/wiki/Common_Criteria https://en.m.wikipedia.org/wiki/Common_Criteria
  - dilap 8y ago
- adrianmonk 8y ago
  I agree with your complaint but I wouldn't use the terminology "fake news". To me, that's more of a specific term that refers to media bias as a propaganda tool. Whereas this is more of a rumor mill going wild as rumor mills naturally tend to do. But yes, while hoards of people behave in a bit of a crazy manner, it is a known thing, and if you're responsible for public communication, you should understand those quirks and their predictable consequences. And you should realize that an ounce of prevention is worth a pound of cure by just making everything as clear as possible from the beginning.
  - marshray 8y ago
    To be precise, the term 'fake news' is an accusation against the media, used as a propaganda tactic, and most often false.
acjohnson55 8y ago
> However in the fake news era my attempts to provide the correct information failed Eww. This actually removes credibility from the announcement, given the way the term "fake news" continues to be used to distract from objective truth.
- royjacobs 8y ago
  I think antirez is using it in the context of people misinterpreting some facts, followed by this misinterpretation escalating fast as a lot of people seem to think its the truth.
  - misterbowfinger 8y ago
    Here's an example of a high-profile one: https://twitter.com/antirez/status/1032271799529218048 https://twitter.com/antirez/status/1032271799529218048
- antirez 8y ago
  What I mean is that even after the project leader tells you it was just some confusion, the Github project still is marked as BSD, and nothing is written in the original announcement clearly stating that Redis would switch license, it is a bit strange that the information continues to spread regardless of what I continued to say.
  - scott_s 8y ago
    I understand both your meaning and frustration. But I think acjohnson55's point is that the phrase "fake news" has become a loaded term. I avoid it because it may imply things I don't want to.
  - endorphone 8y ago
    "Redis is an example of this paradigm. Today, most cloud providers offer Redis as a managed service over their infrastructure and enjoy huge income from software that was not developed by them. Redis’ permissive BSD open source license allows them to do so legally, but this must be changed" Crying fake news (which has been co-opted to mean "real news that I don't like or that you shouldn't pay attention to") completely undermined your statement. Yes, the original announcement wasn't just unclear, it is very clearly drawing a line that "this must be changed".
misterbowfinger 8y ago
> I think that Redis Labs Common Clause page did not provide a clear and complete information, but software companies often do communication errors In their defense, many programmers can be so spiteful and cynical that basically nothing could've been done to change this "communication error." It's helpful that Salvatore clarified, but I suspect that there are many who won't care or won't be convinced. The headline is all they needed to trip their "GTFO" line. Personally, I think the move was fine. There are much worse companies out there.
_hyn3 8y ago
Closely related: "The Commons Clause will destroy Open Source" https://news.ycombinator.com/item?id=17818407 https://news.ycombinator.com/item?id=17818407
- Sir_Cmpwn 8y ago
  As the author of that article, I want to clarify that I was aware that Redis core was going to continue to be available as BSD and incorporated that information into my article. antirez, if you have any comments or corrections I'd be happy to include them in the article.
- JdeBP 8y ago
  And https://news.ycombinator.com/item?id=17814386 https://news.ycombinator.com/item?id=17814386 .
js2 8y ago
antirez: I understand your frustration and don’t mean to take away from the message, but please consider removing “fake news” from the wording. It unnecessarily politicizes the post and really leaves a bad taste. “Fake news” is the term of art that politicians are using to criticize true stories that they don’t like. In this case, you’re trying to correct an actual falsehood. The spreading of falsehoods and their being difficult to correct is as old as time and nothing new to this era. Thank you for listening, and for Redis, as always.
- antirez 8y ago
  Thanks, I removed fake news because I've the feeling that in US it is used by right-wing people, here in Italy it's the current government that is the master of fake news and is used by our left-wings to criticize what they do. In the doubt I changed the wording. I did not expect the term to be political.
  - js2 8y ago
    Thank you.
  - DanielGee 8y ago
    > Thanks, I removed fake news because I've the feeling that in US it is used by right-wing people It's a term used by both the left and the right. By politicians and the media. > In the doubt I changed the wording. I did not expect the term to be political. You shouldn't have. Who cares if it is political. You are the author and you should express yourself the way you want.
    - digitaLandscape 8y ago
      He cares, because he wasn't trying to be political. You could try being literate.
      - dang 8y ago
        Personal attacks will get you banned here. Please don't post like this to HN. https://news.ycombinator.com/newsguidelines.html https://news.ycombinator.com/newsguidelines.html
relaunched 8y ago
The underlying clarification makes sense, sort of. Is it correct to say that, "These specific modules were developed solely by Redis Labs and are free of any open source contributions." I understand that the previously licensed versions of the modules will retain there old license and people can choose to contribute or not, on a go forward basis (based on the common clause). However, if FOSS contributions were made (in mass), it still seems a little unctuous.
radus 8y ago
I think that this clears up questions about redis itself, but doesn't address the concern than future development will be focused on new modules rather than the core - akin to Google Play Services vs. Android.
- chii 8y ago
  > doesn't address the concern than future development will be focused on new modules rather than the core which is a very valid thing to do, since dev costs money.
  - radus 8y ago
    I'm not arguing that, and am personally not outraged by it, but in that context I do understand what seems to be a sense of betrayal on the part of OSS stalwarts, even as redis remains BSD licensed.
  - marcinzm 8y ago
    They have a right to do it but then they're making proprietary software and people who like open source software are free to not like/support them as a result.
- antirez 8y ago
  Personally I'll keep working on the Redis core, plus the Disque module, that will be developed under the AGPL license (not Common Clause). So you can expect me to be focused on the core as usually.
- kawsper 8y ago
  It makes me wonder if we would have ever received an open-source redis-cluster if it was developed today, and it makes me afraid that it won't stay open like today. InfluxDB moved their cluster implementation to their enterprise product behind an undisclosed amount of dollars if you want to self-host.
stefantalpalaru 8y ago
Too little, too late. We need a Redis fork and we need it now, because antirez is no longer reliable. Moving Redis towards an Nginx-like model where new features are exclusively commercial is nothing less than a bait-and-switch - one that affects those small-time developers who advocated for Redis in the first place, making it into what it is today. That Redis has been tainted by a greedy bastard's attempt at becoming a millionaire through licensing tricks. Long live a new Redis fork - by the community, for the community. Just like MySQL was replaced with MariaDB in most setups, Redis (and everything else touched by Salvatore Sanfilippo) needs to be replaced.
allengeorge 8y ago
It was pretty clear to me on reading the RedisLabs page that Redis itself would be BSD licensed. And, FWIW, I totally sympathize with the motivations behind their decision. I suspect we'll see variations of this in a variety of projects (if not module licensing, then enterprise-only features, reduced work in the open-source version, etc.)
- _hyn3 8y ago
  > It was pretty clear to me on reading the RedisLabs page that Redis itself would be BSD licensed. The page has been updated since the original announcement.
rauhl 8y ago
> An example of such module is RediSearch: it was AGPL and is now going to be Apache + Common Clause. Sounds to me like the right thing to do here for Free Software proponents is to download the latest AGPL RediSearch, fork it and continue development.
- codingdave 8y ago
  HN needs a feature to say, "Hey! We haven't talked in forever, wanna catch up?" when we notice comments from people we worked with many years ago but haven't talked to in forever.
- lma21 8y ago
  Who exactly would continue the development in this case? Weren't Redis Labs the biggest contributor here?
scj 8y ago
If a project is licensed "BSD + antithetical_term" or "Apache + antithetical_term", then the names of BSD and Apache should not be invoked. They are free to make a RedisLab license though.
- antirez 8y ago
  Redis core is just "BSD", nothing attached.
Yiftach 8y ago
Hi, this is Yiftach, CTO and Co-founder of Redis Labs. First, let me assure you that Redis remains and always will remain, open source, BSD license. For avoiding any doubt - commons clause (as defined in commonsclause.com) is applied only to add-ons (modules), on top of Redis (e.g. RediSearch, Redis Graph, ReJSON, Redis-ML, Rebloom) that were developed by Redis Labs . We initially released these modules under AGPL license but found two major drawbacks (a) AGPL does not prevent cloud providers (such as AWS) from building managed services from these modules, and (b) we got requests from developers, working at large enterprises to move from AGPL to a more permissive license, because the use of AGPL is against their company’s policy. In addition, a few people here (and other threads) have asked why we didn’t create a new proprietary license, like those offered by Elastic or MaraiaDB? Well, Commons Clause was created by a coalition of several OSS infrastructure companies, some of which use a different OSS licenses. In order to maintain a standard framework we decided to piggyback the restriction (on creating managed services by cloud providers) on any existing OSS license.
- dragonwriter 8y ago
  > we got requests from developers, working at large enterprises to move from AGPL to a more permissive license ...so you did exactly the opposite, though with naming that makes it sound like you honored that request until you actually read the license terms.
  - detaro 8y ago
    it's more permissive for some enterprise use cases if you're worried about AGPL consequences. You now can use it in internal systems. (Assuming you feel comfortable interpreting the vague bits of the "Commons Clause")
    - IshKebab 8y ago
      Yeah the commons clause does not seem like a very well written license. They even mix up i.e. and e.g. Not confidence-inspiring.
- zokier 8y ago
leshow 8y ago
IMO people in the original thread had a pretty good understanding that Redis core was going to have the same license, and some of the modules RL developed would be proprietary (sorry, "source-available"). I think many people are just plain unhappy with that situation, no misunderstanding. I read comparisons drawn between android, taking more and more out of the open source core into proprietary modules, among others.
dwheeler 8y ago
But it does appear to be true that some Redis modules used to be open source software (OSS) and are now proprietary. As I understand it, some modules used to be licensed under the AGPL (an OSS license) and they are now under this new "Commons Clause" (which is a proprietary license). So while the core of Redis is OSS, if my understanding is correct, it's absolutely true that parts of Redis have switched to being proprietary software. That's different than organizations who license the software as "AGPL or proprietary" or "GPL or proprietary" - such licenses are also unambiguously OSS. Choosing "Apache WITH proprietary-rider" is fundamentally different; Redis has every right to do that, but it's fair to note that this isn't the same thing.
- XiZhao 8y ago
  Also: there is no longer a source-disclosure requirement built into the license.
vmlinuz 8y ago
This could have been made clear in a few simple sentences: We have created a new software license which we are calling the 'Redis Labs License' (RLL), by adding the Commons Clause terms as extra restrictions on top of the Apache 2.0 License. We recognise this will no longer be a true Open Source license, but we consider it a fair trade-off between our business needs and our responsibility and commitment to the Redis community. From now and in the future, work on the following Redis Labs modules will be published under the RLL: Module A, Module B and Module C. There will be no change to the licensing of Redis core, and we remain committed to supporting ongoing development under the existing BSD license.
- mrfredward 8y ago
  Redis went out of their way to obfuscate what they were doing with deceptive license names. Now they are struggling to control the damage caused in all the confusion. There's a lesson here...
  - derefr 8y ago
    > Redis Redis Labs isn't Redis, any more than, say, Heroku is Ruby because Matz works there.
arminiusreturns 8y ago
BSD: Freedom for the developer GPL: Freedom for the user You've made your choice.
zimbatm 8y ago
Even if Redis were to change license, the ones who would be to blame are the big cloud providers that re-package OSS products and rent them without giving anything back. They are milking the work out of these projects, taking only the opportunity to make money out of it. I remember talking to Elastic and they had a similar issue where AWS would take their product, package it and not give anything back (money or patches). Obviously it's permitted by the license, it doesn't mean that it's right.
- prepend 8y ago
  Why would this tweak the developers. Isn’t this precisely what the license they chose was made for to include and allow? In this case Amazon functions like a hosting company. So would you be surprised if a hardware vendor automated the install and patching of Elastic? They’re basically just installing OSS for end users.
ChristianBundy 8y ago
Small nitpick: the "most permissive license ever" is the FPL/0BSD: https://opensource.org/licenses/FPL-1.0.0 https://opensource.org/licenses/FPL-1.0.0
keepper 8y ago
What I find interesting is all of the comments saying "Well, I was going to use RedisLabs.. but because of this, i'll use AWS/elasticache" Which is just such a weird statement. So instead of supporting the OSS creators and main contributors, who want a way to monetize their "value add" modules, while still giving the core software freely; you rather support a huge corporation that gives almost zero back to OSS ( AWS has a history of extend and embrace, make it a service/black box, with open source, that would make Microsoft of the 90s/00s proud!)
- prepend 8y ago
  “...gives almost zero back to OSS...” This is not true as Amazon not only releases it’s own projects but contributes really broadly [0]. I can’t speak for others, but I understand the decision to choose ElastiCache over CC software as I understand Amazon’s model a lot better than projects that change their license. So if I’m going to hinge on something non-OSS, I prefer the logical points where a utility that sells cloud compute so doesn’t care about software than for software companies that change their license, really substantially. My org allows AGPL as we can easily comply. But we can’t comply with CC. This shift is a big deal and kind of hoses me. You mention Microsoft in the 90s. They shared source code all the time, but you couldn’t use it and there was zero community. So if you want that, CC is way more likely to get us there as community contributions will dry up. I’d rather depend on companies that are explicit (ie, “this is open source, this is not”). [0] https://aws.amazon.com/opensource/ https://aws.amazon.com/opensource/
  - derefr 8y ago
    > But we can’t comply with CC. Why not? The point of the CC is literally "pay us to use it commercially." You can't just... pay them to use it? Even when that's exactly what you'd be doing by using an AWS service?
    - ubernostrum 8y ago
      You could offer to pay them today, yes. And get a license. But is your license that handles your use case today still going to be valid and useful a year from now? Five years from now? What if redislabs decides there are some more loopholes in need of closing? I'm not a fan of the AGPL. I think it's problematic as an example of copyright maximalism, and I think it's counterproductive in that it basically took a bunch of ideas that used to be FUD and said "yeah, what if we actually did that". But I can say this for it: you know what you're getting and you know that for all the burdens it puts on you, it (like the GNU GPL) was designed to make sure certain rugs can never be yanked out from under you.
skissane 8y ago
If you want to ship enterprise modules under a source-is-publicly-available-but-not-open-source license, a sort of limited-commercial-use-source-available-freeware license, I think that is definitely better in some ways than keeping them fully closed source, so should be applauded. But, I think this "Commons Clause" condition was a poor choice of a way of achieving that end. Having some condition which is added to an existing open source license, and takes away some of the rights that license grants, is horribly confusing. If you aren't granting the full set of rights associated with open source license X, then open source license X should not even be mentioned. If you had just picked some existing license which had the set of permissions you wanted, or even created a new one, I don't think anywhere near as many people would be complaining. "Commons Clause" is a horrible idea and should be retired.
esterly 8y ago
Best to check the source, license unchanged: https://github.com/antirez/redis/blob/unstable/COPYING https://github.com/antirez/redis/blob/unstable/COPYING Thank you Salvatore for Redis, it has benefited so many projects.
- papaf 8y ago
  This should be the top rated comment. I wish people would put the pitchforks away, and acknowledge the challenges of funding a successful Open Source project.
hannob 8y ago
Seems a severe case of "we tried really hard to obfuscate our words with marketing bullshit speak and then it backfired and we got a shitstorm instead, because noone understoond what we were really trying to say".
quotemstr 8y ago
I've seen saying for years and years that we're going to see a resurgence of proprietary software. Free software was one of those ideas that spread so successfully that people temporarily forgot that life could be otherwise, which means they stopped caring about the idea, which in turn led to a complacency that, at first, shifted software licensing to a BSD-style permissive model, and now, is starting to allow 1990s cancer like this commonscause.com license restriction to come back. There's nothing new in the software ecosystem driving this. Well, this commonsclause.com crap is such a move toward proprietary software. We're here. > The Commons Clause is a first step at starting an important discussion
- maxerickson 8y ago
  The baseline is pretty high to call anything a resurgence. I guess by revenue dollar most software is already proprietary, it's just billed as a service or whatnot. Considering all the code inside of companies, it's probably the case that most lines of code are also proprietary.
  - slavik81 8y ago
    Revenue is probably not a good measure, as an open source solution tends to reduce market size as measured by dollars spent. From Bob Young, founder of Red Hat: > The reality is most of our users don’t pay us squat [...] Our business model [showed] that if we had millions of users, we only needed one-tenth of our users to actually find something of value to pay Red Hat for. > If Microsoft’s earning $40 billion worth of operating system sales, and if we [reduce that to] a $4 billion-a-year market, because nine out of ten of the people that use our operating system aren’t going to pay us, we’re quite happy with a share of a $4 billion-a-year market. https://growthhackers.com/growth-studies/red-hat-how-they-developed-a-big-idea-that-shook-up-a-huge-market https://growthhackers.com/growth-studies/red-hat-how-they-de...
    - maxerickson 8y ago
      I think that makes it important to consider revenue. Say each company spends ¼ of revenues paying developers. Microsoft would be paying ~10x the developers to write software and probably be writing more software.
carapace 8y ago
(What I want to know is how does that #$%^ing page hijack the font even in Reader Mode!? WTF! I want serif body fonts goddamnit.)
- mrob 8y ago
  The body text is contained in a <pre> tag with newlines for paragraphs instead of the correct <p> tags. You can force Firefox to respect your font choice by adding a font:inherit to the <pre>'s css with the developer tools. This will break pages with correctly used <pre> tags.
  - carapace 8y ago
    Thank you! You're too kind.
welder 8y ago
This Redis-compatible alternative is also BSD licensed: https://github.com/ideawu/ssdb https://github.com/ideawu/ssdb And it doesn't require the dataset fit in RAM.
actuator 8y ago
This still doesn't sound good. Even I don't like the fact that AWS is extremely bad at giving back to the community but it is not like doing things like this will make people use RedisLabs. Infact, I will be wary of using these modules for the fear of getting locked with a vendor. I have used both Elasticache and RedisLabs and can say both have their advantages and annoyances. Elasticache is way easier to set up and configure. RedisLabs puts a lot of artificial restrictions which you have to work around. Also, their pricing model is kind of screwed up. The only major problems I have seen with Elasticache are: Locked APIs, no way to scale instance once created with live traffic, A-Z failover sucks and no persistence.
ummonk 8y ago
>We at Redis Labs are sorry for the confusion generated by the Creative Common page, and my colleagues are working to fix the page with better wording. So I'm not the only one mixing up "Creative Commons" with "Commons Clause".
Animats 8y ago
Isn't this what Google pulled with Android? First it was open source. Then there were closed source Play Store Services. Then Google products required the Play Store services. Then third-party Android distributions stopped working. Now you do what Google tells you to do and pay them what they want to be paid to use Android.
DickScarington 8y ago
RedisLabs needs to clearly mark which of the following versions of these modules are freely redistributable as part of a larger project or service. neural-redis BSD RediSearch AGPL rediSQL AGPL-3.0 ReJSON AGPL Redis AGPL redis-cell MIT Redis-ML AGPL redis-timerseries AGPL cthulhu BSD rebloom AGPL redis-cuckoofilter MIT redis-roaring MIT redis-tdigest MIT Session MIT countminsketch AGPL topk AGPL ReDe MIT commentDis MIT
smarnach 8y ago
Just as a curiosity, Redis Labs is offering a hosted Memcached service, so they are a cloud provider profiting from open-source software they did not significantly contribute to. Just saying. (More seriously, I think the only mistake they made was communicating this license change badly, and with too much whining about how unfair it is that people do what the license explicitly allows them to do.)
- antirez 8y ago
  It's a proxy that targets Redis actually.

448 points by antirez 8y ago

I’m happy to hear that Redis will remain BSD, but it’s a bit weird to play the “fake news” card. Can you really blame people for interpreting this the wrong way? Especially when the original announcement was specifically targeting cloud providers, who, afaik, are just running vanilla redis. If there is anything that we can learn from this, is that a little bit of forethought when putting these announcements out there is required, and blaming this on the readers is not fair.

lbriner 8y ago
The parts that I noticed were missing from the original announcement include more specifics about the parts of the system that would change licence and also whether the intention was to charge people to use these new parts. It was unclear whether the intention was to stop these cloud providers or to make them pay in some way for what they were doing.
- antirez 8y ago
  Yep, now finally this is updated to specify exactly what changes license. Incidentally: nothing I ever worked to directly. It would be way cooler if also things like RediSearch or RediGraph could still be totally open source, but it's apparently very hard, Common Clause at least will allow it to be free-as-beer and with the source code available.
  - yarosv 8y ago
    > Common Clause at least will allow it to be free-as-beer and with the source code available. The license is very vague. According to this > a product or service whose value derives, entirely or substantially, from the functionality of the Software. What if I have big online shop with huge catalog and my software value "derives from the functionality of RedisSearch"? Do my only option not to host it at Redis Labs?
    - derefr 8y ago
      > my software value "derives from the functionality of RedisSearch" Your software might just literally be 99% RedisSearch, but you're not making money from your software (e.g. by selling it); you're making money from operating your software, combined with content (e.g. by using it to facilitate e-commerce transactions.) You'd be in trouble if you wanted to white-label "your technology" and sell it as a platform for other people to run their own online shops on, because it's not really "your technology" to sell. But if your play isn't specifically that, then you're fine. Also, > Do my only option not to host it at Redis Labs? this wouldn't be an option. The whole point of this license is that the software "RedisSearch" is owned by Redis Labs. Hosting it with Redis Labs themselves, or not, is irrelevant. You have to license the software (if you're deploying it yourself)—or rely on your DBaaS provider to license it for you (if they're deploying it.) Basically exactly like a software patent, come to think of it.
meowface 8y ago
I don't think he's blaming people, but it does seem odd if he had no input at all into the Redis Labs announcement.
antirez 8y ago
Honestly my colleagues did a poor job with such an announcement there, and I did not get a chance to read it (which is actually a good thing, it means that the OSS division, I'm part of, is independent and can work to the Github repositories without caring about what the commercial side does). But it was kinda of a manifest to announce that certain things made inside Redis Labs are now Common Clause licensed. In doubt, why not asking? Also people that follow me and Redis for years getting trapped into this idea is very strange, I would never do that and without explaining, prior notice, and so forth. So yep the Common Clause page at Redis Labs sucked but in doubt ask instead of creating news out of no actual information.
- l2dy 8y ago
  The abbreviation "CC" will be confusing if Commons Clause got popular… We already have Creative Commons licenses.
  - antirez 8y ago
    Agreed, fixed.
    - millette 8y ago
      Utterly confused. Does this have anything to do with Creative Commons licences? That's what the articles calls it in its first and last sentences.
  - jpobst 8y ago
    Particularly because the linked post calls it "Creative Common" twice and "Common Clause" elsewhere.
  - tango24 8y ago
    When I see CC, I also tend to think it means Common Criteria : https://en.m.wikipedia.org/wiki/Common_Criteria https://en.m.wikipedia.org/wiki/Common_Criteria
- dilap 8y ago
adrianmonk 8y ago
I agree with your complaint but I wouldn't use the terminology "fake news". To me, that's more of a specific term that refers to media bias as a propaganda tool. Whereas this is more of a rumor mill going wild as rumor mills naturally tend to do. But yes, while hoards of people behave in a bit of a crazy manner, it is a known thing, and if you're responsible for public communication, you should understand those quirks and their predictable consequences. And you should realize that an ounce of prevention is worth a pound of cure by just making everything as clear as possible from the beginning.
- marshray 8y ago
  To be precise, the term 'fake news' is an accusation against the media, used as a propaganda tactic, and most often false.

acjohnson55 8y ago

> However in the fake news era my attempts to provide the correct information failed Eww. This actually removes credibility from the announcement, given the way the term "fake news" continues to be used to distract from objective truth.

royjacobs 8y ago
I think antirez is using it in the context of people misinterpreting some facts, followed by this misinterpretation escalating fast as a lot of people seem to think its the truth.
- misterbowfinger 8y ago
  Here's an example of a high-profile one: https://twitter.com/antirez/status/1032271799529218048 https://twitter.com/antirez/status/1032271799529218048
antirez 8y ago
What I mean is that even after the project leader tells you it was just some confusion, the Github project still is marked as BSD, and nothing is written in the original announcement clearly stating that Redis would switch license, it is a bit strange that the information continues to spread regardless of what I continued to say.
- scott_s 8y ago
  I understand both your meaning and frustration. But I think acjohnson55's point is that the phrase "fake news" has become a loaded term. I avoid it because it may imply things I don't want to.
- endorphone 8y ago
  "Redis is an example of this paradigm. Today, most cloud providers offer Redis as a managed service over their infrastructure and enjoy huge income from software that was not developed by them. Redis’ permissive BSD open source license allows them to do so legally, but this must be changed" Crying fake news (which has been co-opted to mean "real news that I don't like or that you shouldn't pay attention to") completely undermined your statement. Yes, the original announcement wasn't just unclear, it is very clearly drawing a line that "this must be changed".

misterbowfinger 8y ago

> I think that Redis Labs Common Clause page did not provide a clear and complete information, but software companies often do communication errors In their defense, many programmers can be so spiteful and cynical that basically nothing could've been done to change this "communication error." It's helpful that Salvatore clarified, but I suspect that there are many who won't care or won't be convinced. The headline is all they needed to trip their "GTFO" line. Personally, I think the move was fine. There are much worse companies out there.

_hyn3 8y ago

Closely related: "The Commons Clause will destroy Open Source" https://news.ycombinator.com/item?id=17818407 https://news.ycombinator.com/item?id=17818407

Sir_Cmpwn 8y ago
As the author of that article, I want to clarify that I was aware that Redis core was going to continue to be available as BSD and incorporated that information into my article. antirez, if you have any comments or corrections I'd be happy to include them in the article.
JdeBP 8y ago
And https://news.ycombinator.com/item?id=17814386 https://news.ycombinator.com/item?id=17814386 .

js2 8y ago

antirez: I understand your frustration and don’t mean to take away from the message, but please consider removing “fake news” from the wording. It unnecessarily politicizes the post and really leaves a bad taste. “Fake news” is the term of art that politicians are using to criticize true stories that they don’t like. In this case, you’re trying to correct an actual falsehood. The spreading of falsehoods and their being difficult to correct is as old as time and nothing new to this era. Thank you for listening, and for Redis, as always.

antirez 8y ago
Thanks, I removed fake news because I've the feeling that in US it is used by right-wing people, here in Italy it's the current government that is the master of fake news and is used by our left-wings to criticize what they do. In the doubt I changed the wording. I did not expect the term to be political.
- js2 8y ago
  Thank you.
- DanielGee 8y ago
  > Thanks, I removed fake news because I've the feeling that in US it is used by right-wing people It's a term used by both the left and the right. By politicians and the media. > In the doubt I changed the wording. I did not expect the term to be political. You shouldn't have. Who cares if it is political. You are the author and you should express yourself the way you want.
  - digitaLandscape 8y ago
    He cares, because he wasn't trying to be political. You could try being literate.
    - dang 8y ago
      Personal attacks will get you banned here. Please don't post like this to HN. https://news.ycombinator.com/newsguidelines.html https://news.ycombinator.com/newsguidelines.html

relaunched 8y ago

The underlying clarification makes sense, sort of. Is it correct to say that, "These specific modules were developed solely by Redis Labs and are free of any open source contributions." I understand that the previously licensed versions of the modules will retain there old license and people can choose to contribute or not, on a go forward basis (based on the common clause). However, if FOSS contributions were made (in mass), it still seems a little unctuous.

radus 8y ago

I think that this clears up questions about redis itself, but doesn't address the concern than future development will be focused on new modules rather than the core - akin to Google Play Services vs. Android.

chii 8y ago
> doesn't address the concern than future development will be focused on new modules rather than the core which is a very valid thing to do, since dev costs money.
- radus 8y ago
  I'm not arguing that, and am personally not outraged by it, but in that context I do understand what seems to be a sense of betrayal on the part of OSS stalwarts, even as redis remains BSD licensed.
- marcinzm 8y ago
  They have a right to do it but then they're making proprietary software and people who like open source software are free to not like/support them as a result.
antirez 8y ago
Personally I'll keep working on the Redis core, plus the Disque module, that will be developed under the AGPL license (not Common Clause). So you can expect me to be focused on the core as usually.
kawsper 8y ago
It makes me wonder if we would have ever received an open-source redis-cluster if it was developed today, and it makes me afraid that it won't stay open like today. InfluxDB moved their cluster implementation to their enterprise product behind an undisclosed amount of dollars if you want to self-host.

stefantalpalaru 8y ago

Too little, too late. We need a Redis fork and we need it now, because antirez is no longer reliable. Moving Redis towards an Nginx-like model where new features are exclusively commercial is nothing less than a bait-and-switch - one that affects those small-time developers who advocated for Redis in the first place, making it into what it is today. That Redis has been tainted by a greedy bastard's attempt at becoming a millionaire through licensing tricks. Long live a new Redis fork - by the community, for the community. Just like MySQL was replaced with MariaDB in most setups, Redis (and everything else touched by Salvatore Sanfilippo) needs to be replaced.

allengeorge 8y ago

It was pretty clear to me on reading the RedisLabs page that Redis itself would be BSD licensed. And, FWIW, I totally sympathize with the motivations behind their decision. I suspect we'll see variations of this in a variety of projects (if not module licensing, then enterprise-only features, reduced work in the open-source version, etc.)

_hyn3 8y ago
> It was pretty clear to me on reading the RedisLabs page that Redis itself would be BSD licensed. The page has been updated since the original announcement.

rauhl 8y ago

> An example of such module is RediSearch: it was AGPL and is now going to be Apache + Common Clause. Sounds to me like the right thing to do here for Free Software proponents is to download the latest AGPL RediSearch, fork it and continue development.

codingdave 8y ago
HN needs a feature to say, "Hey! We haven't talked in forever, wanna catch up?" when we notice comments from people we worked with many years ago but haven't talked to in forever.
lma21 8y ago
Who exactly would continue the development in this case? Weren't Redis Labs the biggest contributor here?

scj 8y ago

If a project is licensed "BSD + antithetical_term" or "Apache + antithetical_term", then the names of BSD and Apache should not be invoked. They are free to make a RedisLab license though.

antirez 8y ago
Redis core is just "BSD", nothing attached.

Yiftach 8y ago

Hi, this is Yiftach, CTO and Co-founder of Redis Labs. First, let me assure you that Redis remains and always will remain, open source, BSD license. For avoiding any doubt - commons clause (as defined in commonsclause.com) is applied only to add-ons (modules), on top of Redis (e.g. RediSearch, Redis Graph, ReJSON, Redis-ML, Rebloom) that were developed by Redis Labs . We initially released these modules under AGPL license but found two major drawbacks (a) AGPL does not prevent cloud providers (such as AWS) from building managed services from these modules, and (b) we got requests from developers, working at large enterprises to move from AGPL to a more permissive license, because the use of AGPL is against their company’s policy. In addition, a few people here (and other threads) have asked why we didn’t create a new proprietary license, like those offered by Elastic or MaraiaDB? Well, Commons Clause was created by a coalition of several OSS infrastructure companies, some of which use a different OSS licenses. In order to maintain a standard framework we decided to piggyback the restriction (on creating managed services by cloud providers) on any existing OSS license.

dragonwriter 8y ago
> we got requests from developers, working at large enterprises to move from AGPL to a more permissive license ...so you did exactly the opposite, though with naming that makes it sound like you honored that request until you actually read the license terms.
- detaro 8y ago
  it's more permissive for some enterprise use cases if you're worried about AGPL consequences. You now can use it in internal systems. (Assuming you feel comfortable interpreting the vague bits of the "Commons Clause")
  - IshKebab 8y ago
    Yeah the commons clause does not seem like a very well written license. They even mix up i.e. and e.g. Not confidence-inspiring.
zokier 8y ago

leshow 8y ago

IMO people in the original thread had a pretty good understanding that Redis core was going to have the same license, and some of the modules RL developed would be proprietary (sorry, "source-available"). I think many people are just plain unhappy with that situation, no misunderstanding. I read comparisons drawn between android, taking more and more out of the open source core into proprietary modules, among others.

dwheeler 8y ago

But it does appear to be true that some Redis modules used to be open source software (OSS) and are now proprietary. As I understand it, some modules used to be licensed under the AGPL (an OSS license) and they are now under this new "Commons Clause" (which is a proprietary license). So while the core of Redis is OSS, if my understanding is correct, it's absolutely true that parts of Redis have switched to being proprietary software. That's different than organizations who license the software as "AGPL or proprietary" or "GPL or proprietary" - such licenses are also unambiguously OSS. Choosing "Apache WITH proprietary-rider" is fundamentally different; Redis has every right to do that, but it's fair to note that this isn't the same thing.

XiZhao 8y ago
Also: there is no longer a source-disclosure requirement built into the license.

vmlinuz 8y ago

This could have been made clear in a few simple sentences: We have created a new software license which we are calling the 'Redis Labs License' (RLL), by adding the Commons Clause terms as extra restrictions on top of the Apache 2.0 License. We recognise this will no longer be a true Open Source license, but we consider it a fair trade-off between our business needs and our responsibility and commitment to the Redis community. From now and in the future, work on the following Redis Labs modules will be published under the RLL: Module A, Module B and Module C. There will be no change to the licensing of Redis core, and we remain committed to supporting ongoing development under the existing BSD license.

mrfredward 8y ago
Redis went out of their way to obfuscate what they were doing with deceptive license names. Now they are struggling to control the damage caused in all the confusion. There's a lesson here...
- derefr 8y ago
  > Redis Redis Labs isn't Redis, any more than, say, Heroku is Ruby because Matz works there.

arminiusreturns 8y ago

BSD: Freedom for the developer GPL: Freedom for the user You've made your choice.

zimbatm 8y ago

Even if Redis were to change license, the ones who would be to blame are the big cloud providers that re-package OSS products and rent them without giving anything back. They are milking the work out of these projects, taking only the opportunity to make money out of it. I remember talking to Elastic and they had a similar issue where AWS would take their product, package it and not give anything back (money or patches). Obviously it's permitted by the license, it doesn't mean that it's right.

prepend 8y ago
Why would this tweak the developers. Isn’t this precisely what the license they chose was made for to include and allow? In this case Amazon functions like a hosting company. So would you be surprised if a hardware vendor automated the install and patching of Elastic? They’re basically just installing OSS for end users.

ChristianBundy 8y ago

Small nitpick: the "most permissive license ever" is the FPL/0BSD: https://opensource.org/licenses/FPL-1.0.0 https://opensource.org/licenses/FPL-1.0.0

keepper 8y ago

What I find interesting is all of the comments saying "Well, I was going to use RedisLabs.. but because of this, i'll use AWS/elasticache" Which is just such a weird statement. So instead of supporting the OSS creators and main contributors, who want a way to monetize their "value add" modules, while still giving the core software freely; you rather support a huge corporation that gives almost zero back to OSS ( AWS has a history of extend and embrace, make it a service/black box, with open source, that would make Microsoft of the 90s/00s proud!)

prepend 8y ago
“...gives almost zero back to OSS...” This is not true as Amazon not only releases it’s own projects but contributes really broadly [0]. I can’t speak for others, but I understand the decision to choose ElastiCache over CC software as I understand Amazon’s model a lot better than projects that change their license. So if I’m going to hinge on something non-OSS, I prefer the logical points where a utility that sells cloud compute so doesn’t care about software than for software companies that change their license, really substantially. My org allows AGPL as we can easily comply. But we can’t comply with CC. This shift is a big deal and kind of hoses me. You mention Microsoft in the 90s. They shared source code all the time, but you couldn’t use it and there was zero community. So if you want that, CC is way more likely to get us there as community contributions will dry up. I’d rather depend on companies that are explicit (ie, “this is open source, this is not”). [0] https://aws.amazon.com/opensource/ https://aws.amazon.com/opensource/
- derefr 8y ago
  > But we can’t comply with CC. Why not? The point of the CC is literally "pay us to use it commercially." You can't just... pay them to use it? Even when that's exactly what you'd be doing by using an AWS service?
  - ubernostrum 8y ago
    You could offer to pay them today, yes. And get a license. But is your license that handles your use case today still going to be valid and useful a year from now? Five years from now? What if redislabs decides there are some more loopholes in need of closing? I'm not a fan of the AGPL. I think it's problematic as an example of copyright maximalism, and I think it's counterproductive in that it basically took a bunch of ideas that used to be FUD and said "yeah, what if we actually did that". But I can say this for it: you know what you're getting and you know that for all the burdens it puts on you, it (like the GNU GPL) was designed to make sure certain rugs can never be yanked out from under you.

skissane 8y ago

If you want to ship enterprise modules under a source-is-publicly-available-but-not-open-source license, a sort of limited-commercial-use-source-available-freeware license, I think that is definitely better in some ways than keeping them fully closed source, so should be applauded. But, I think this "Commons Clause" condition was a poor choice of a way of achieving that end. Having some condition which is added to an existing open source license, and takes away some of the rights that license grants, is horribly confusing. If you aren't granting the full set of rights associated with open source license X, then open source license X should not even be mentioned. If you had just picked some existing license which had the set of permissions you wanted, or even created a new one, I don't think anywhere near as many people would be complaining. "Commons Clause" is a horrible idea and should be retired.

esterly 8y ago

Best to check the source, license unchanged: https://github.com/antirez/redis/blob/unstable/COPYING https://github.com/antirez/redis/blob/unstable/COPYING Thank you Salvatore for Redis, it has benefited so many projects.

papaf 8y ago
This should be the top rated comment. I wish people would put the pitchforks away, and acknowledge the challenges of funding a successful Open Source project.

hannob 8y ago

Seems a severe case of "we tried really hard to obfuscate our words with marketing bullshit speak and then it backfired and we got a shitstorm instead, because noone understoond what we were really trying to say".

quotemstr 8y ago

I've seen saying for years and years that we're going to see a resurgence of proprietary software. Free software was one of those ideas that spread so successfully that people temporarily forgot that life could be otherwise, which means they stopped caring about the idea, which in turn led to a complacency that, at first, shifted software licensing to a BSD-style permissive model, and now, is starting to allow 1990s cancer like this commonscause.com license restriction to come back. There's nothing new in the software ecosystem driving this. Well, this commonsclause.com crap is such a move toward proprietary software. We're here. > The Commons Clause is a first step at starting an important discussion

maxerickson 8y ago
The baseline is pretty high to call anything a resurgence. I guess by revenue dollar most software is already proprietary, it's just billed as a service or whatnot. Considering all the code inside of companies, it's probably the case that most lines of code are also proprietary.
- slavik81 8y ago
  Revenue is probably not a good measure, as an open source solution tends to reduce market size as measured by dollars spent. From Bob Young, founder of Red Hat: > The reality is most of our users don’t pay us squat [...] Our business model [showed] that if we had millions of users, we only needed one-tenth of our users to actually find something of value to pay Red Hat for. > If Microsoft’s earning $40 billion worth of operating system sales, and if we [reduce that to] a $4 billion-a-year market, because nine out of ten of the people that use our operating system aren’t going to pay us, we’re quite happy with a share of a $4 billion-a-year market. https://growthhackers.com/growth-studies/red-hat-how-they-developed-a-big-idea-that-shook-up-a-huge-market https://growthhackers.com/growth-studies/red-hat-how-they-de...
  - maxerickson 8y ago
    I think that makes it important to consider revenue. Say each company spends ¼ of revenues paying developers. Microsoft would be paying ~10x the developers to write software and probably be writing more software.

carapace 8y ago

(What I want to know is how does that #$%^ing page hijack the font even in Reader Mode!? WTF! I want serif body fonts goddamnit.)

mrob 8y ago
The body text is contained in a <pre> tag with newlines for paragraphs instead of the correct <p> tags. You can force Firefox to respect your font choice by adding a font:inherit to the <pre>'s css with the developer tools. This will break pages with correctly used <pre> tags.
- carapace 8y ago
  Thank you! You're too kind.

welder 8y ago

This Redis-compatible alternative is also BSD licensed: https://github.com/ideawu/ssdb https://github.com/ideawu/ssdb And it doesn't require the dataset fit in RAM.

actuator 8y ago

This still doesn't sound good. Even I don't like the fact that AWS is extremely bad at giving back to the community but it is not like doing things like this will make people use RedisLabs. Infact, I will be wary of using these modules for the fear of getting locked with a vendor. I have used both Elasticache and RedisLabs and can say both have their advantages and annoyances. Elasticache is way easier to set up and configure. RedisLabs puts a lot of artificial restrictions which you have to work around. Also, their pricing model is kind of screwed up. The only major problems I have seen with Elasticache are: Locked APIs, no way to scale instance once created with live traffic, A-Z failover sucks and no persistence.

ummonk 8y ago

>We at Redis Labs are sorry for the confusion generated by the Creative Common page, and my colleagues are working to fix the page with better wording. So I'm not the only one mixing up "Creative Commons" with "Commons Clause".

Animats 8y ago

Isn't this what Google pulled with Android? First it was open source. Then there were closed source Play Store Services. Then Google products required the Play Store services. Then third-party Android distributions stopped working. Now you do what Google tells you to do and pay them what they want to be paid to use Android.

DickScarington 8y ago

RedisLabs needs to clearly mark which of the following versions of these modules are freely redistributable as part of a larger project or service. neural-redis BSD RediSearch AGPL rediSQL AGPL-3.0 ReJSON AGPL Redis AGPL redis-cell MIT Redis-ML AGPL redis-timerseries AGPL cthulhu BSD rebloom AGPL redis-cuckoofilter MIT redis-roaring MIT redis-tdigest MIT Session MIT countminsketch AGPL topk AGPL ReDe MIT commentDis MIT

smarnach 8y ago

Just as a curiosity, Redis Labs is offering a hosted Memcached service, so they are a cloud provider profiting from open-source software they did not significantly contribute to. Just saying. (More seriously, I think the only mistake they made was communicating this license change badly, and with too much whining about how unfair it is that people do what the license explicitly allows them to do.)

antirez 8y ago
It's a proxy that targets Redis actually.