Conversation
Notices
-
!RedHat releases #OpenShift Origin under #Apache License 2.0 http://ur1.ca/951s9
about a year ago from web- Evan Prodromou, Tyng-Ruey Chuang and Christopher Allan Webber and 3 others like this.
- Marcio B. Jr., Christopher Allan Webber and Răzvan Sandu repeated this.
-
Red Hat doing its bit for the decline of the GPL I see. #obvioustroll
-
Red Hat Cloud Computing team on #OpenShift Origin: http://ur1.ca/951uy
-
Richard Fontana
GNU Affero General Public Licence
, The GNU General Public License
, Not a Troll
, Simon Phipps
@webmink !GPL wasn't even in the running. !AGPL was considered, though. I consider your dent and this reply !notatroll
-
@fontana hmm a permissive license with code hosted on GitHub (don't you have an rpm for Gitolite?): are you soliciting Brogrammers?
-
@tmarble I assume that your dent is !notatroll. Red Hat rejects #brogrammers movement.
Evan Prodromou likes this. -
@fontana impact is push towards proprietary web services: a less DoubleSpeak title might be #OpenForBizShift or #SecretSauceInTheCloudShift
-
Evan Prodromou likes this.
-
@fontana no forking can undo Red Hat's !disturbing decision to not stand for (as in defend) users' freedoms this time
Jure Repinc (JLP) likes this. -
@lxoliva WDYM? The Apache License is a !freesoftware license, specifically recommended by !fsf for contexts such as that here
-
@fontana it is well known that any permissive license, such as Apache, does nothing to *defend* users' freedoms. defend != respect
-
@fontana what case this context fits in that makes the fsf recommendation a permissive license, rather than a strong copyleft one?
-
@fontana and why does Red Hat care about what the FSF has to say when it comes to offering users weaker defenses than possible?
-
@lxoliva possible weasel word warning
-
@lxolivia OpenShift involves deployment scripts and etc right? LAGPL might make more sense, but AGPL might require releasing system scripts?
-
ie, APLv2 makes sense to me here, from my vague understanding.
-
Los Angeles General Public License #LAGPL
-
Christopher Allan Webber
Free Software Foundation
, Free Software
, Evan Prodromou
, Christopher Allan Webber
Btw, I thought "APLv2" meant Apache License v2 but that's probably wrong. I blame @evan or my misreading of http://ur1.ca/954rj
-
YES. I agree.
-
@evan a) AGPL for SN a mistake b) no desire for SN to be basis of standard fs fsw impl c) ? #uncertaintyretrollnotatrolltrulyobtainsthistag
Christopher Allan Webber and Richard Fontana like this. -
I agree that Apache 2.0 is a good license for software trying to set up new networks and ecosystems.
-
I don't think AGPL was a mistake for StatusNet.
Arne Babenhauserheide and Christopher Murtagh like this. -
-
@evan license of competitor projects (if any) also relevant.
-
I'm having computation problems!
-
@fontana thanks. from that, I infer there were other contenders to such a standard, but OpenShift sounded unique to me. what did I miss?
-
@fontana trust on whom or what? I don't know what issue it is to tell whether I understand it ;-) never been a Cygnus employee BTW
-
@lxoliva the desire is for this code to become the standard. Only other FLOSS PaaS project I know of is #CloudFoundry; uses ALv2 as well
-
Doesn't CloudFoundry depend on VMWare, though?
-
@evan No, CloudFoundry is its own thing. They do distribute a "Micro Edition" of it as a vmware image, though.
-
@fontana if not, the solution is for the rightsholder to not be a company...
-
@cwebber maybe ALGPL would be a more suitable acronym, although it might be mistaken for some Arabic license ;-)
-
@fontana I was a contractor with “Cygnus, a Red Hat company” for a couple of years, but employment came later, after Red Hat arrived in .br
-
@fontana I don't get that argument. ALv2 doesn't take away any doubt that the software will remain free in the future; quite opposite
-
@fontana so Red Hat missed the opportunity to set up a level playing field like e.g. Linux's, furthering proprietary forks instead :-(
-
@lxoliva ah, I did not know that
-
@fontana how can Red Hat, after years of Linux experience, not see that “AL is better to bring .com contributors” is FUD by abusers?
-
@fontana aoliva at cygnus.* shortly became at redhat.*, yeah. it all started in feb 2000, few months after the merge
-
Cool. cloudfoundry.com is like, "Step 1: Install VMWorkHappySpaces"
-
Also, HEY.
-
@lxoliva I think it may be true right now. If Red Hat had chosen AGPL, little chance of there being non-RHT contributors.
-
@fontana if OpenShift gets any external contributions under AL, how could AGPL have averted that?
-
@fontana How do you mean? #interested
-
@fontana AL just enables some of the improvements to never see the light of day. rather than encouraging, it discourages contributions
-
@fontana my feeling is that Red Hat chose preemptive defeat, rather than leadership on the path that has served it so well
-
“if you copyleft, we won't contribute” often means “we won't contribute anyway, but with copyleft we can't abuse so it's no use”
Bradley M. Kuhn and Jure Repinc (JLP) like this. -
@lnxwalt such FUD was spread about GPLed software in pre-cloud days too. but just look at the Android kernel, and Java ;-)
-
@lxoliva, Extremely insightful. I think you're right. Sad thing is, it's one of those hypotheses that can't be tested w/out alt. universes.
-
Never thought @fontana'd be someone who'd attack copyleft just to corporate shill for !RedHat: RH could've prevented pwnage in this case!
Jure Repinc (JLP) likes this. -
@lxoliva Well, when I say industry standard, I mean that GPL/AGPL OpenShift would not likely become the basis for others’ implementations.
-
@bkuhn well, given the existence of AL contender, it could have been tested now, but Red Hat chose the pyrrhic preemptive defeat instead
-
@bkuhn by preemptive defeat I mean giving up core values to get popularity for something that won't be aligned with the core values
-
@lnxwalt is your assumption that this someone big doesn't want the software to be Free? if so, why go along with their wishes?
-
@lnxwalt ... because they can't make them proprietary? how would proprietary derivatives benefit us or Red Hat, anyway?
Rui Seabra likes this. -
@lnxwalt this line of argument comes across as “keep your enemies close, and push your real friends away”
-
@lnxwalt it makes no sense to dual license when one license is strictly more permissive than the other
-
@fontana sounds like Red Hat is uncertain about the OSS proposition that OSS is more beneficial to all involved parties
-
@fontana if it was sure, it would know others would join in regardless of conditions to keep derivatives OSS
-
@evan it is obviously trivial, for there's license compatibility. but that won't get users copyleft defenses from sidestream abusers
-
@lxoliva It isn’t meant to carry that message. Dual licensing is a perfectly legitimate way to increase contributors to a project …
-
@lxoliva ... while still protecting against proprietary abuses.
-
@lxoliva Tell that to our friends at Mozilla.
-
@lnxwalt when licenses are compatible, it achieves the opposite: stronger copyleft loses teeth but keeps abusers' contributions away
-
@lnxwalt again? :-)
-
or at least that should be the effect of dual licensing on rational players who understand the legal meaning of this practice ;-D
-
@evan you are correct sir
-
@bkuhn choice of #AGPL -> de facto RHT pwnage, as if © assignment were required. BTW, unlike #CloudFoundry, no #CLA for #OpenShift Origin.
-
@fontana because they misunderstand the license, or because they don't want to respect users' freedoms?
Jure Repinc (JLP) likes this. -
Nope. The only way to do that is to put value into the copyleft fork.
-
@lxoliva Don't know. Right now it's still all theoretical.
-
@fontana that sounded like an admission of preemptive defeat :-(
-
@lxoliva not really. Apache License is actually stealth copyleft here.
-
@fontana Wow, I had no idea I or #MediaGoblin were on an #AGPL rescue mission! (Not actually sure how to take that.)
-
@fontana No, I didn’t mean proprietary.
-
@cwebber what you need to do is set up the #Affero Software Foundation. The rest will take care of itself.
Christopher Allan Webber likes this. -
@lnxwalt it is interesting to note that Apache License doesn't seem to allow mere relicensing of unmodified sources under GPL/AGPL
-
@fontana Actually, I’ve got to admit that #RedHat‘s cloud platform is the only one I’ve seen that didn’t make me feel queasy.
Evan Prodromou likes this. -
@fontana I think a trivial change like s/OpenShift/GNUShift/ or what have you would be enough of a change to be a derivative work.
-
It's really nice, actually! I've started working on getting StatusNet running there.
-
@evan I also got crazy idea to run #statusnet on top of #JBossAS on #OpenShift. It *must* be possible.
-
Yes, on the big OpenShift service. Couldn't run identi.ca there yet (no queue server module), but otherwise should work.
-
@fontana I think we could probably make the cron module work, though.
-
I don't even know what that is.
-
But I'm definitely going to try some of the NodeJS stuff I've been doing there.
-
One thing that I think PaaS providers need to do is allow a "fallback" to IaaS. If you don't have a module for RabbitMQ (say), ...
-
...let me spin up my own virtual server in the same data centre and run it on that.
-
@fontana btw, personally I think AGPL here would have been a strong business move.
-
@evan that sounds like it might be like what #OpenShift calls "DIY" - see e.g. http://ur1.ca/95f16
-
If running OpenShift is to be a direct revenue stream, AGPL better. If goal is to get others to adopt, then sell them RHEL, AL better.
Marcio B. Jr. likes this. -
Along those lines, I am really interested in both a) MediaGoblin working on OpenShift and b) seeing if OpenShift can deploy to Debian ;)
Richard Fontana likes this. -
Somewhat separate curiosities; I'd be happy with MediaGoblin deployments via OpenShift to RHEL only if it was easy
-
@evan that can only protect downstream, not sidestream. unless it's so compelling that sidestream dries out, of course
-
@fontana huh? stealth copyleft? now you totally lost me. can you elaborate?
-
If ALv2 has "copyleft" against AGPL it's that relicensing healthy permissively licensed community to copyleft for its sake is a "dick move"
-
And a GnuShift thats main purpose is to AGPL the codebase wouldn't win friends to AGPL, would probably even alienate friends like me.
-
I'm an AGPL supporter! But friends don't let friends destroy communities, & relicensing would make one of the anti-copyleft arguments true
Richard Fontana likes this. -
@cwebber yeah, but we know that's FUD by abusers&friends, right? how can relicensing be ok to non-Free, but not to copyleft?
drew Roberts likes this. -
@cwebber abusers go like this “if you copyleft, I won't contribute” (asposted before), then “if you add copyleft, you're a dick”
-
@cwebber but can we settle on counting this as further damage Red Hat just inflicted on software freedom by going permissive?
-
@lxoliva I believe that under some circumstances using noncopyleft license is better route to achieve social aims of copyleft.
-
redhat hasn't inflicted damage on sw freedom by going permissive. Many permissive projects have done pretty well with j…
-
@fontana I love all forms of Free Software, but I oppose giving up copyleft defenses without strong evidence it will achieve more freedom
-
@fontana yeah, thank you! even OSI, whose defn obsesses about properties of licenses, goes beyond that in some bullets
-
@fontana [software] freedom is about the [4] freedoms one retains; licenses are just one of many means to take them away
Christopher Allan Webber and Klaus like this. -
@fontana I (&FSF) agree it may be in some cases, but that doesn't make it so in all cases, let alone without plausible evidence
-
@fontana <friendly> (-: oh, give me a break! don't you dare labeling the abusive power to take away others' freedom as freedom </...
-
@tekk yes, it has taken away copyleft defenses. that's damage regardless of other benefits that may arise
-
@cwebber I don't follow this, if you want restrictions on what can be done with the source, put those restrictions in the liscense.
-
I'm not okay with relicensing to non-free, though a lot of ALv2 decision here seems to do with OpenShift being like a library
Richard Fontana likes this. -
I fail to see how removing a feature is inherently damage. if I want to race a car, is removing the seats to decrease w…
-
Proprietization would be bad, though most proprietary (and copyleft!) tie-in will probably happen on associated scripts I'd anticipate.
Richard Fontana likes this. -
@lxolivia I think if LAGPL existed, I'd probably agree; since it doesn't, ALv2 probably makes more sense to me
-
@lxolivia btw I don't think we should give up on copyleft licesnses! I <3 copyleft. I think it's strategy, though.
-
@tekk I'm speaking of a defense against freedom-depriving abuse, not any random feature. rendering one defenseless is damaging in my book
-
@cwebber LGPL is just additional permissions on top of GPL; if Red Hat wanted, it could've offered similar permissions on top of AGPL
-
@cwebber [frequent typo alert: my surname is spelled with a singular I, lest it would be olWEva :-]
-
@fontana of course! freedom means *no* obstable prevents its enjoyment; -ENOSRC and ©opyrestrictions are just two common obstacles
-
@fontana there are various other kinds of legal, technical, social, etc obstacles to freedom, even when it comes to software
-
@fontana I wish more FLOSS proponents would see that. too bad OSI didn't expand Perens' criteria like Debian did
-
Erk, sorry! :)
-
don't worry, I used to do lxolivia as well :)
-
@evan, I avoid forking a permissively licensed program under copyleft license unless there's clearly a tragedy of the commons in progress.
-
Why would you want do to that?
-
How could I miss this conversation :) Entertaining as usual even if it's flat /me missing the old threaded view
-
You can turn that on in user preferences if signed in!
-
Richard Fontana
Free Software Foundation
, Evan Prodromou
, Mike Linksvayer
, Christopher Allan Webber
I think @evan should have written #statusnet in #Python and @cwebber should have written #Mediagoblin in #PHP. That would set things aright.
Christopher Allan Webber likes this. -
@bkuhn one important thing you overlook: the human developers of #OpenShift Origin wanted to use the Apache License.
Christopher Allan Webber likes this. -
I think that makes sense.
-
I think the interesting part of this discussion is that #RedHat is treated very much like a community insider.
-
Holding RedHat to the same expectations as, say, a loose group of individual developers is a good sign of the company's engagement.
-
@evan it is promising that companies CAN be treated as a community insider. I want to found companies, at least one, and this gives me hope.
Evan Prodromou likes this. -
Plus there's whatever mysterious thing @spot was asking about
-
Richard Fontana
Red Hat
, GNU Affero General Public Licence
, Bradley M. Kuhn
, Christopher Allan Webber
@bkuhn As James #Stewart said in another film "One doesn't often get a second chance". @cwebber and #MediaGoblin is #AGPL's second chance.
All Identi.ca content and data are available under the