Latest / Oxide and Friends / Fork in the road for Terraform?

Transcript

Fork in the road for Terraform? · Oxide and Friends

• 0:00
  Speaker 1: I'm not in the litter box. Although I I am kind of delighting at at someone, which is to say most people who don't have context for that question.
• 0:12
  Speaker 2: No, no, no, no, no. But you don't provide context.
• 0:14
  Speaker 1: We don't provide any context.
• 0:18
  Speaker 2: I I I just wanna for for the record, in terms of not providing context, I I I listened to last week's podcast because I wasn't able to attend to listen last night. And I did really enjoy your shout out to the the light cone militia. That was very subtle and appreciated.
• 0:39
  Speaker 1: Listen, you gotta have the callbacks to previous episodes.
• 0:43
  Speaker 2: Absolutely.
• 0:44
  Speaker 1: So, hey, speaking of previous episodes, just, a really sad note that that we, just at the top here, for those that, did not see, Chris Nova, who was a guest on the podcast, last year and a very beloved technologist, died in a climbing accident last week. And really gutted to see that news yesterday. I think that that, everyone was really in in shock. Chris is someone who had a real and it is, especially, Chris would have loved this episode. Right?
• 1:20
  I mean, this is Absolutely. This is right up her alley in a lot of ways, in terms of infrastructure as code and DevOps and, a real light in the industry, who you know, I Adam, I I I relistened to that episode yesterday. I just wanted to rehear her in her own voice, and really was reminded again what a great presence she was, and really, tragic that she was taken from us so soon. So it was a very sad note over the weekend, and and her heart just obviously go out to to Chris's, family and and loved ones who I know are are devastated by the loss. So, Chris, you're gonna you you will you live on in our hearts for sure, but but we we miss you today.
• 2:09
  Absolutely. So and then, Adam, you you mentioned that you I, you you were listening to last week's episode Yes. Where, first of all, did you did you had you heard Fred Brooks' voice?
• 2:25
  Speaker 2: No. I hadn't. I really it was it was delightful not only to hear his voice, but to hear your impersonation of him, which was very well crafted.
• 2:37
  Speaker 1: My my my foghorn leghorn impersonation of of, Fred Brooks, who I was just like I just had not heard his voice before, before the podcast. I do wanna give an update on that, that the, the the silver bullet was was announced at 10:59 AM on on Tuesday. And, I feel I I I'm feeling pretty good about my prediction. Namely, there's gonna be technical depth, and it is not going to be the biggest paradigm shift in software development in the last 30 years. So, interesting stuff there for sure, but I think I think we I I feel safe with the you know, it's a little gutsy.
• 3:14
  I reported I recorded the podcast on on Monday for a Tuesday night.
• 3:18
  Speaker 2: Really calling your shot. Exactly. When you could have just waited 12 hours or something. But yeah.
• 3:23
  Speaker 3: I know.
• 3:23
  Speaker 1: And did you have any context for anything that was going on for you? Or did you had you seen any of that?
• 3:28
  Speaker 2: I I had just barely enough context because I, I had seen the proclamations of 100 Execute. So I did have enough context for the sub podcast. So, but I I have not caught up on the big announcement. So glad that there is some technical depth, but also not surprised to hear your report that it's not a silver bullet.
• 3:53
  Speaker 1: Not a silver bullet. So Fred Brooks can, be assured from beyond the grave there that he that he is he's still does not need to to, we don't need to update that essay quite yet. And then another thing you may have missed is, which gets us kinda on point today, is the the Hashi announcement. When was that? That was That was, well,
• 4:15
  Speaker 2: I I remember it well because the day that, COVID infected our house. So it was, August 10th. So 11 days ago, I guess. So just just saw that, kinda popping through.
• 4:28
  Speaker 1: And had you, but you you were, obviously had other other things to focus on.
• 4:37
  Speaker 2: Yeah. But did did was lucid enough to kinda take take a look through that, and and I was really disappointed. I mean, I I don't know. I don't know about you, but I really think of hold Hashi in a lot of respect. And so was and I have used and contributed to Hashi projects.
• 4:54
  So I was really disappointed to see them going the BSL route, which which is terrible.
• 5:00
  Speaker 1: Are we gonna distinguish the BSL route by even, like, calling it the BSL? I just feel like it's this is a, I mean, this is not open source.
• 5:09
  Speaker 2: Is is is It's not open source. And I I mean, I'm not gonna say to their credit because not to their credit, but they don't refer to BSL as open source. They have in their FAQ a bunch of euphemisms that use the word source, and availability and things like that, but but don't kind of paint it as a open source ish choice. They merely, like, imply that there is some relationship.
• 5:35
  Speaker 1: The fact is bad. Fact is really bad. I I mean, just this idea that that we well, so we we should describe what happened, the and then by way of of introducing our guests here, we'll get Josh and Malcolm here, that, actually maybe actually, we'll turn to them and have Josh, do you wanna, describe what kind of what they announced? And, I'd actually be curious if did you if you had any kind of heads up on this before they announced
• 6:08
  Speaker 3: it? Sure. So, yeah, first, it's really great to be here. This is a, yeah, a really cool podcast. So thanks so much for having me.
• 6:17
  So as far as what they announced, it really the, the key summary was that all of their major open source projects used to be MPL licensed, and it's been that way for anywhere from 5 to 9 years depending on the project. And, about 2 weeks ago, they made the fateful decision to change from that open source MPL license to the so called BSL, or business source license. And that's not a a really well known license. We so I'm the cofounder and chief product officer at GruntWerk, and we do a lot of work with Terraformal. We were not given any kind of heads up, and so it was just one of those things that was shared, like, in our company Slack.
• 7:09
  The day it came out. And So
• 7:11
  Speaker 1: you had you had no heads up on this?
• 7:14
  Speaker 3: Yeah. Yeah. No heads up. It it I don't know
• 7:17
  Speaker 1: why I'm acting surprised because so much else about this has been mishandled. But that just seems man, that seems because, I mean, this is something you've built a business around in terms of Terraform is at the kind of the foundation of what you're doing, it sounds like.
• 7:34
  Speaker 3: Yeah. Yeah. So Rotemark helps companies with, what we call DevOps foundations, and we've always thought of ourselves as agnostic to the technology. We just wanted to choose the best technology, but we always thought Terraform is the best tool for the job. And so we wound up building up hundreds of thousands of lines of code of, like, commercial grade modules, all written in Terraform along with a bunch of tools like CACB and that use Terraform and things like that.
• 8:02
  So, yeah, it it's very much a layer in our stack and, it was it was kind of shocking to see that a layer that we had just assumed was open source and would always be open source was was not. One of the jokes in our Slack, internally was, you know, we said in life there are 3 certainties, death, taxes, and TerraForm will always be open source. And, so that's the hard one. Wow. It turned
• 8:33
  Speaker 1: out to be wrong. I I I honestly thought you're gonna say death taxes and an update for our community are the the I mean, you must have just been feeling the color drain out of you reading that. I mean, that must have been just shocking then. I mean, to certainly, I mean, like you, I've always viewed, and I in fact, I love their choice of the MPLV too. We all are also big MPL fans, and I just did not see this coming at all.
• 9:07
  That must have been really gutting to see. And so I can I ask you the because I think Adam and I are trying to figure this out? Adam was like, look, I know. Like, I was sick. So I think I missed the day that they announced the rationalization for this, I think.
• 9:22
  Did I
• 9:23
  Speaker 2: Well, just we've seen this kind of thing before, right, with with, Confluent. We've seen it with Mongo, where it was kinda clear that they felt like they were getting a raw deal from cloud providers. And they, I don't know, misguidedly, but felt like they could provide some protections using this kind of legal mechanism. And it just wasn't clear at all to me who Hashi was worried about.
• 9:50
  Speaker 3: K. So can
• 9:51
  Speaker 1: you answer that question, please? Because I have the same question. Like, who is this alright. Like, look. We talked about who our we yes.
• 9:57
  We had an episode that was a sub a subtweet. Who are we talking about here? Who's the target of this?
• 10:04
  Speaker 3: You yeah. So what's what's funny is those of us who thought we might be targeted by it, kinda had no doubt. And so it is interesting to hear the other perspective. So there's a number of alternatives to some of the commercial TerraForm products out there. Just to name the the most popular ones, there's, Terra team, who's, the cofounder of him is Malcolm, who's on the line with us here.
• 10:35
  M0, Scalar, Spacelift, and then there are a number of others. And, all of them essentially compete with Terraform Cloud. And so I I think the the narrative internally at HashiCorp was something to be effective. Hey. We, HashiCorp, are funding all the development for Terraform, but look at all these providers out there competing with us earning much money.
• 11:03
  And some of them are venture backed and have raised multiple rounds of money. And I I'm I'm guessing that, that that was not seen as a positive sign, inside of HashiCorp. So I and if you look at some of the FAQ language as well, around how they define embedded, it it's sort of oddly precise in my opinion to target, really all these, like, Terraform for CICD type vendors.
• 11:34
  Speaker 1: So so the target is all of them?
• 11:38
  Speaker 3: Yeah. I mean, so I
• 11:40
  Speaker 1: I what I'm a little weird
• 11:42
  Speaker 3: about this line of questioning is so a bunch of us, in in this named group, we all got together and, as I think you're working up to, you know, we we ultimately formed a a consortium. But, one of the the kind of philosophical tenets of that consortium is we are not anti HashiCorp. In fact, we're we're all very grateful for what HashiCorp created. I mean, Terraform was an amazing technology and by most measures, they're an amazing company. So it's really important to all of us not to be anti HashiCorp just because they made a decision that happens not to meet our needs or that we disagree with as members of the TerraForm community.
• 12:28
  But but, yeah, ultimately, there are some implications of that and that's it's worth talking through what those are because they're
• 12:36
  Speaker 1: Yeah. I mean, and I think I first of all, I should say that I I love the the the spirit of the OpenTF effort in terms of of really, trying to guide Hashi back to to kind of the right decision here, and also totally appreciate just the the the spirit by which I look. We we are grateful for this. This is kind of the there's this narrative that, we're we're not seek seeking to kind of take advantage of the development of others. It is, that said, I just think it is, it's very strange to me that Hashi would kinda go after this was effectively a startup ecosystem, which I mean, generally, startups should not go after other startups.
• 13:20
  This is kind of one of those. When you see a startup spending time on other startups, it's like, look. That's a wrong you should be going after incumbents. And I kinda thought there might be some incumbent on so do you do you think TerraForm and again, and not to not to kinda push you out of your comfort zone here, but you was Terraform the motivation? Because I kind of assume that Vault would have been the motivation.
• 13:41
  Speaker 3: You know, great question. I'm not as familiar with the Vault ecosystem. And then it it felt like for those of us in the Terraform ecosystem, there were these obviously thriving companies that, you know, maybe HushCorp felt were counter to its goals for for some reason, but I I'm just not familiar with, the other ecosystem. But even so, the the the classic example they always give in the FAQ is always like you can't do vault as a service. And I didn't know if that was like a a straw man example or if there really are, like, Vault as a service offerings that that companies have been offering that, without a partnership with HashCorp that they've been trying to crack down.
• 14:26
  So as far as I have seen, it is focused on Terraform, but I also feel like I'm not personally aware nearly as much of the other systems that they plan, compared to the Terraform one.
• 14:38
  Speaker 4: Yeah. I'm I'm not I don't know of a a manageable outside of HashiCorp, but I knew I do know that other secrets management companies were concerned, particularly those that that wrote their own, secrets management software, but they use Terraform to manage their infrastructure. Like, now they're a competitor to Ashycorp. Can they use can they use Terraform? And, like, this was an early concern, but this is another one of those things that's been amended in the FAQs.
• 15:06
  So if you're not managing that tool or making a managed offering of that tool, then you're not considered in, violation of the the the the BUSL license.
• 15:18
  Speaker 1: So I I missed the amendment to the FAQ. So what was the what was their amendment?
• 15:21
  Speaker 4: Yeah. They keep they've they've added a number of amendments, but there was one the other day that was added that said that if you if you use a Hashi tool let me look up what number it is real quick. Sorry. I should have these all memorized by now. I've been at this page about 7,000 times.
• 15:37
  Yeah. Chrome's recommending it to me constantly.
• 15:43
  Speaker 1: Thanks, Chrome. I hadn't seen this. Yeah. Thank you.
• 15:46
  Speaker 4: Give me one second here. I can't remember.
• 15:47
  Speaker 1: And are they hopefully, that language is gonna be in the license itself. Samantha, just again, I know my my perspective on this is the maybe slightly sharper than than than the rest of yours. But the, I I mean, I really think that there's a lot, whether it's deliberate or not, There's a huge amount of ambiguity about I mean, these are products that are kind of like span the stack, I mean, deliberately, really. And so there's a bunch of questions about who competes with what. The language is pretty vague.
• 16:19
  And, Adam, I mean, we had this issue with with Cockroach. We are effectively using BSL code from Cockroach in oxide. And I'm just trying to remember if that if the more specific language came before or after our use of cockroach. But cockroach's language is really very, very specific about, because we had the question of like, hey. We are using this as the foundation of the oxide control plane.
• 16:45
  Are we gonna be violating the the BSL, or the I believe, Josh, you called it the so called BSL, which is what I I insist that we all call it, the so called BSL. But they were, to Cockroach's credit, they were very explicit about what it what constitutes database as a service. And in particular, if a user cannot create a schema or modify a schema, it does not they consider it to be noncompetitive. It's like, okay. Great.
• 17:12
  That's very that's very specific and technical. That's not a entry in a fact. That's, like, in the license. It's very concrete. We don't really have that here, but correct me if I'm if that's incorrect.
• 17:24
  Speaker 3: You know what? I'd love to speak to that actually because I've done a lot of thinking about their their license. So so just to be clear, the way that they structure their license, they use this, you know, business source license, which, by the way, sometimes people say BSL because that's the acronym for business source license, And sometimes people like Corey, just a moment ago say Busil because that's the official SPDX, short handle for the business source license. So just in case anyone's a a nerd on that kind of stuff like that.
• 17:53
  Speaker 4: I I also say Busil because it sounds gross, and I want people to realize that.
• 17:57
  Speaker 1: I like it. I I like I I just got out of abusive relationship.
• 18:06
  Speaker 3: That's a
• 18:06
  Speaker 4: good one. I think I think I think Metabucil every morning.
• 18:12
  Speaker 3: I've never appreciated all these puns. Well, yeah. So the I like, I'm just gonna adopt them that, I guess. But usual, it as a as a standard license, it just says absolutely no production usage is allowed, but you can do anything you want for nonproduction usage. And then the way it's structured is they have this thing called an additional use grant.
• 18:31
  And if you actually look at the license text, it's like this header that they have there. It's almost like metadata to the text. And the one additional use grant that they give you is use is they say you can use it in production as long as, a, you're not a competitor, and, b, if you are a competitor, you don't embed Terraform, or host Terraform. And so to go back to your original point, Brian, one of the major issues with that is that it's vague. What does it mean to be a competitor?
• 19:01
  What does it mean to embed? And what does it mean to host? And to HunchCorps credit, they have addressed that, but kind of. Like, the the first issue I have is that they address it in the FAQ. And, you know, an FAQ is not a license.
• 19:15
  And so now you're you're left to wonder as an end user, like, okay. So I read the license and it's unclear. So I go to the FAQ. The FAQ adds additional clarity, but is that binding and cannot change? You know, so so it's it's it's vague.
• 19:32
  And not only that, but even the definitions that they give in the FAQ, it's like more words, but it it still has ambiguity. And and so They thought it made
• 19:43
  Speaker 2: it clear that it can change. Right? They just changed it. They just changed the license. So they've made it clear the license for the software is what they feel like.
• 19:50
  Speaker 4: They the FAQs, to be clear, when you go to the FAQs, each of the FAQs has a last. Each one has a last updated date, and the embedded one was updated today. So there's a there's a more clear definition that was added today of what embedded means. By the way, the one we were talking about earlier was 0.26. So question 26 is not great.
• 20:11
  So this
• 20:11
  Speaker 3: is If I may,
• 20:12
  Speaker 2: part you know, part of the reason this is so disappointing is that so much of, you know, Terraform, for example, Terraform success, to me at least, has been on the back of its openness. And people may I'm sure the the, you know, folks, who are embedded Terraform into their products, who are represented here, made that choice in part because of its openness. You know, when I started using Terraform years years ago, I wouldn't have used it if it was some proprietary software or even if it was some Busil software. It was because of its openness. So it's so disappointing to have the ladder pulled up.
• 20:48
  Speaker 1: You know, Adam, you're really right. And I I because I feel and this is part of the reason why open source became so important for infrastructure software is that, you know, it you got investment protection. And one of the things that, you know, folks don't necessarily remember from the bad old days of, proprietary software. But if you've got proprietary software now that that that remains, you will see bad behavior from rent syncing companies where you will build a product based on a proprietary product. And now when you go to re up your license or or suddenly, you've lost all of your leverage.
• 21:23
  And that investment protection was really, really important. I mean, Adam, part of the reason that you had that disposition around Terraform is like, hey. I'm gonna build some infrastructure. It's gonna stick around for a while. And I it's kind of my responsibility to the future to make sure that that we're building on something where they don't have to worry about this.
• 21:40
  And that and I I I really think you know, I gave this talk a decade ago on corporate open source antipatterns. And I I said actually that I was I would come back in in 2022 to talk about, you know, the new mistakes that we had made. And I I really feel that first among them is relicensing. I I I think that there is a social contract here that is, being set aside. And I think especially for those folks that contributed.
• 22:14
  And, Adam, you mentioned you'd you'd contributed at some point to Terraform. I'm sure Josh, you, Malcolm, and Corey, and you've got plenty of folks who have. Those folks assigned their copyright to Hashicorp. And to me, that assign because Hashicorp demanded it. And to me, that assignment of copyright bounded that relationship by trust.
• 22:36
  Like, I am now trusting you to, to exercise my intent, which is to keep this open source under the MPLB 2.
• 22:45
  Speaker 5: I also think,
• 22:47
  Speaker 1: Go ahead.
• 22:48
  Speaker 5: One reason there's so much talk about Terraform versus bulk or anything like that is Terraform is different in that it benefits for more users. You get more people writing providers, more people writing tools on top of it. Whereas Vault and console, those sort of benefit from enough people to keep the the code being developed. But Terraform is really it's a it's a language. It's a compiler to turn HCL into infrastructure, and I think that's different.
• 23:15
  And those are more community projects rather than just products.
• 23:22
  Speaker 1: Absolutely. And I I do think and that's why I am I mean, I think this was a a bad idea, poorly implemented. But one of the the kinda complexions of poor implementation here is the fact that this was done across this entire very disparate suite. So it's very hard to kinda suss out the intent. And I had kind of assumed, although now, you know, I'm really questioning that assumption that Terraform was kinda collateral damage on the way to going to monetize alt, to or or or to prevent AWS, for example, or a cloud provider, from having a vault based service.
• 23:58
  But, it kinda sounds like that's not the case, honestly, and which is really it's it it it however you cut it, it it violates the trust that is so important in open source. I mean, Adam, you'd kind of said this point earlier about, like, you know, when they're everything they kinda say in the FAQ is like, well, I'm I'm having some trust issues with you now. So, and, Josh, for you and Malcolm and Corey and and the the kind of the broader Terraform community, I mean, you almost feel like if at this point Ashley were to walk it back, I don't you feel that you still are would have some trust issues there?
• 24:36
  Speaker 4: I I definitely would. I probably have like the spiciest takes out of the group. So take, take some, take some, some Harper.
• 24:44
  Speaker 5: You're definitely the spiciest. Sorry.
• 24:49
  Speaker 4: But, like, the trust I mean, I I wrote a blog post on Friday, kind of about this. I spent a lot of time watching, some videos of, Mitchell and Arman. I don't I don't know them personally, just to be clear. I I've watched a bunch of videos. I was trying to get, like
• 25:03
  Speaker 3: I mean, I'm
• 25:03
  Speaker 4: a founder of a very early stage company. Right? And so I was trying to, like, get an idea of, like, what, like, what happened. Right? And, like, trying to make a through line as to, like, where it went from this team of people that really cared about open source to where we landed.
• 25:17
  And I don't know if you guys talked about this before I joined. I'm old, so it took me forever to get into Discord. But, like, there's a fallout that's happening right now. Like, CNCF is looking at pulling, HashiCorp tools from multiple, dev workflows for containerd, Kubernetes kind, Flux 2, Argo project. Like, there's a bunch of people that now have busy work, but they're gonna try to pull these tools out and replace them with CNCF equivalents so there's no issues.
• 25:43
  Right? And so it's like like, this has ended up in a very weird place for a company that's so committed to open source. But I think that what's really telling about where HashiCorp is as a company today is, like, they were committed to open source. Sure. But we had the same panic 2 years ago when they announced they weren't taking PRs anymore because they were understaffed.
• 26:03
  Does everybody remember that? Wow. Yeah. Oh, yeah. They stopped taking contributions.
• 26:09
  They didn't ask for help. They didn't do a call for maintainers. They stopped accepting community is what they did. Right? Like, that's really what they did.
• 26:18
  They stopped accepting community. And now in the light of of this happening, you see competitors and people in the space come together and offer resources. We're offering them the community they say that they want. But that that's, like, that's the thing that's really weird for me about, like, their commitment is, like, I feel like they don't have a commitment to open source. I don't think they really care about the community as much as they try to make it out to be in in this boost will change.
• 26:41
  I think there's something else probably driving it, but, like, the community is here. Like, we want to help. We don't want a fork. We're contributing resources. We're offering things, and HashiCorp isn't accepting them.
• 26:52
  And that's what we really, I think, wanna push for is, like, getting this to not be a fork and really getting them to come back to where they used to be. And and will we be able to trust them from there? Like, I mean, I I think if they truly handed it over to a foundation, yeah, but simply a license reversal, reverting a commit isn't gonna do it. I I I think we got it to be a land in a foundation's hands.
• 27:17
  Speaker 1: Well and so and and, Corey, I I can't judge by the latency in terms of you getting into Discord in terms of your exact age. It's like tree rings terms of, like, aging. So determining the exact vintage. I don't know if you remember Hudson. And I feel like Hudson has been kind of culturally lost.
• 27:34
  So, Hudson was a ubiquitous platform for CI. And, one of the the the the many horrific things that Oracle did, the kind of the the mechanized Oracle did after the acquisition of Sun was they decided that, actually, there was that Hudson what they were going to enforce the trademark of Hudson, and it was effectively gonna be proprietary. And the community turned on a dime and turned that into Jenkins. And Hudson was Jenkins was actually was Hudson for many years. And, and it was remarkable to me how quickly just for all the same reasons you're outlining because this there's a there's a real broad grassroots momentum around Jenkins.
• 28:21
  And, I mean, Adam, you remember that, the whole and and it happened very, very quickly.
• 28:27
  Speaker 3: We we
• 28:27
  Speaker 2: had just started we had just started using Hudson, and and, and it just the whole it was like schooling fish. Everyone turned at the same instant. It was remarkable.
• 28:38
  Speaker 1: And and Jenkins, like, survived and thrived. So I I do think that we I mean, fortunately, Oracle provides a a case study for the survival of of open source projects after the the most mendacious corporate behavior. Thank you, Oracle, I guess. And just in terms of of the, you know, the strength of the community being able to kind of propel, an alternative forward or because I think it's, you know, another by the way, another and this is a great segue into the OpenTF work that that that you all have have kind of brought to come together to form. The one potential and maybe even likely outcome here is that open TF does have to become a fork, is a fork for some number of months.
• 29:25
  And Hashi realizes, like, actually, we wanna participate. So how can we and then, you know, the terms would be very different. It's like, well, this is gonna be in a foundation and so on. This is not dissimilar. This happened at, a joint where I was the the joint, the we were the the stewards of Node.
• 29:42
  Js, and the community was really fracturing over a bunch of issues. And there was a fork that started there. I as I told them, it's like, you actually didn't need to fork it to get our attention. But we, you know, I think it for me, it was eye opening in terms of of, how difficult it was going to be for any single company to keep that to keep what was a pretty fractured community together, and that a foundation was gonna be the right route for, for Node. The so I I I think that the foundation route ultimately I mean, I'd love to get your your all take on this, but it feels like the foundation route has to be the route or this software.
• 30:21
  And I and, you know, one of the things that I would revisit from my my talk 10 years ago is I was I was kind of jaundiced about foundations, and that was before a lot of this, the the frankly abusive trust. And I now I kinda feel like, actually, foundations are when you when you've got a broad based community, you really it really needs to be in a foundation. Do you wanna maybe in terms of, like, a segue to Open TF, so, how how quickly after this did you I mean, this it sounds like it took everybody, by surprise. Everyone was kind of on the back foot. How I I how did you start reaching out to one another?
• 30:56
  How did you find one another? How did the the OpenTF effort start?
• 31:01
  Speaker 3: Yeah. I I can comment on that. In my opinion, that was actually one of the most magical things that happened from this. So we've had a number of companies that were either competitors or, like, they kind of used our stuff and we sort of competed. And and these are people I've maybe spoken to once or never met, but, you know, you see online when they post something.
• 31:23
  All of a sudden within a day of that, announcement taking place. So literally by that Friday, we had a Slack community among all these so called competitors where we were all sort of united, by a common cause. The word that kept kept coming to my mind was, like, fellowship. We're like a fellowship of, of companies who all, were, looking to a a need for an open source Terraform that we could continue to innovate and and build our companies on. And then from there, the idea that we would need to fork, but maybe present Hashi with the opportunity to first revoke.
• 32:08
  Like, all of these discussions were happening. We had, like, our own little de facto governance mechanism, like, little slack poles. It we're and we're all commenting on how much fun it was to, like, be collaborate with all these competitors. So and to this day, like, I mean, to this day, it was like a week ago, to this day, like, it I don't know. It it remains one of my most positive, like, business memories because it's just so fun to go from this kind of, like, distant competitive mindset to collaborating with a group of people all united in a common cause.
• 32:41
  And what I was surprised to learn was how many more resources and willingness we all had, to contribute to Terraform open source than what was actually currently being contributed to Terraforma long source. It it sorry. I I know I'm at risk of rambling, but just to have this one point. I've had a bunch of customer conversations in the last week, and a common refrain that I've heard from customers is, yeah, but how could your consortium possibly have the resources to maintain the giant that is Terraform? And I I'm always, like, chuckling with that because, you know, I get it.
• 33:19
  Like, the nice branding, the the the publicly traded company in Edge, the regular blog posts, the regular releases. It looks like there's dozens or hundreds of people working on this tool. And when we did the internal estimation, we could we estimated that it was a small fraction, at HashiCorp that that told them people actually working on Terraform open source, as the number of people we would be able to marshal as a consortium. So that's what I think given us all a lot of pause as we think about and what the path forward might look like.
• 33:54
  Speaker 2: You know, because you say that, you know, I think back to now, like, 13 years ago, Brian, when we were kicking off at the Lumos project, there there was never any thought of, like, presenting a demand letter to Oracle because we knew what the answer was gonna be there. But yeah. But like it felt sort of in some ways daunting. And this is when Brian, you were at Joanna and I was at Delphix and Garrett was, I think at Vicenta. And, there was something very inspiring that what you're describing Josh of that moment of kind of becoming this fellowship or collaborating across company borders to get this thing done to, like, erect this monolith was it was kind of inspiring.
• 34:37
  Speaker 1: It definitely was. And it was it was clarifying, obviously. But it's like, yeah, we're on our own. Like, we don't have like, all of these and it honestly felt like a weight had been lifted where we had been and, you know, with with or Oracle was on such a a speed run of open source malfeasance, and it's all happening very, very quickly. But the you know, we were very concerned about how do we accommodate Oracle with now so many Sun folks leaving, and we're all part of this big exodus.
• 35:12
  And, one of the things that we there were lots of things that we did not like about the way that it was being prosecuted internally when it was open. And, you know, people are like, well, you mean you can't have your own operating system? It's like, no. Actually, we can. It's like, actually, sorry.
• 35:27
  It's like, well, who writes the device drivers? It's like, yeah. We write the device like, you would not write that as it turns out like we actually know how
• 35:33
  Speaker 2: to create. Yeah. It turns out like that. I mean, for split up some more wind in your sails, turns out 13 years later, like that experiment is going fine for us. Like we have not died on some mountain of complexity that only a team of 10,000 engineers could take on.
• 35:48
  So if if it's any consolation to your to your customers is they ask reasonable questions saying, how are you gonna be named this thing? It's, like, actually quite well.
• 35:57
  Speaker 1: And I would say that there was so much friction that we had for so long about who what's important, who does what. And Sun, back when it was pre oracle, Sun had this outsized influence that was kinda uncontrollable. Like, ultimately, it wasn't a real community because Sun dominated it too much. And it only became a community when it was kinda forced to become 1, really, a true community. And so Josh totally get what you're the the the the vibes of, like, really kinda coming together and the the the clarify.
• 36:32
  And I so one thing I I would love you all to speak with too because it's something that I loved in the so you put out the Open TF manifesto. The, it was I'd it obviously came together it was to come together pretty quickly, it sounds like. The, certainly, by the time, and and one of one of your number had actually pinged me. I was really honored, honestly, to be pinged, being like, hey. Would you mind signing this thing?
• 37:02
  And I'm, I mean, for us, we've got a Terraform provider, so Terraform definitely plays a role for us. But we're nowhere near, as, kind of at the epicenter as you are. And honestly was honored to to that that, you know, we'd love to be able to to help out any way we can because I really loved all the energy. And in particular, what I loved and, Josh, maybe you can walk me through kind of when this started happening. As people are signing it, they're also talking about resources that they can actually dedicate to it.
• 37:32
  And it's, like, significant. People are talking about, you know, dedicating 2 3 folks to this effort. That kind of evolved? How did that start?
• 37:42
  Speaker 3: Yeah. You know, initially, there was this realization that we all had that a busil, an busil I'm gonna have to check on that to I'm from him on. Anyway, a a Beusel licensed Terraform was just not in the world any of us wanted to live in. And and it wasn't just about commercially, like, is this something that we could make commercially viable? It it's actually really important to share this.
• 38:11
  A lot of people have said, well, hey. I'm not a competitor to HashCorp that this doesn't matter to me. Why should I care? From the perspective of someone who is considered, supposed competitor to Hashicorp, we realized that, okay. So in order to actually participate in this Busil licensed ecosystem, the the license is so vague, almost by design, that we can't really operate in compliance with it unless we get HashiCorp's explicit permission.
• 38:39
  So now in order to operate our business, presumably on an annual basis, we have to approach them on bended knee and ask for permission to continue operating as a business and pray that they don't raise licensing prices on us too much.
• 38:54
  Speaker 1: And absolutely by the way, just to just to to kind of underscore that point, it is so vague that we are not embedding, any, any Hashi products in the oxide rack. But if we were, we oxide would feel that it it is far too vague because, in particular, we we don't know who's gonna acquire HashiCorp. And take it from adamantly, the acquirer can really and, you know, it could easily be an Oracle acquisition. It could be a Broadcom acquisition. It could be a Dell acquisition.
• 39:30
  And those are all companies that that could easily say that they compete with Omniside. And I mean, so it's like, Adam, I don't know what you're taking back. I'm like, this is like too big for us.
• 39:41
  Speaker 2: Let alone be something. I can't be not not only that, but imagine a diligent customer of ours would say, hey. Do you have anything fishy inside that box? Like, well, we do have this busil thing. It's like, well, we don't know what that is.
• 39:54
  We can't accept this product. They don't bring this into our environment. Like, for all I know that, you know, HashiCorp is gonna show up and demand their extortion money from from our customers. So it creates too much way too much risk.
• 40:06
  Speaker 1: Way too much risk. I'm actually glad you brought up that point because that's something that's kinda it gets kinda lost on folks is that having gone through an acquisition, from someone who's doing a lot of due diligence with Samsung on Joyant, and they are I mean, this is I can't remember what 20 7. So, like, we need to do a license audit on all of the node modules we're using. I'm like, well, this is gonna be interesting. And I'm like, and it was actually, it's a huge credit to the node ecosystem that we've it I mean, it was, like, it was tens of thousands of modules.
• 40:40
  It was actually great because we overwhelmed the, the firm that does open source due diligence. Is like, we cannot do due diligence
• 40:50
  Speaker 3: on this.
• 40:50
  Speaker 1: We've never seen as much open source. Like, we
• 40:52
  Speaker 2: I love the idea of being, like, well, we're out of compliance because we have to send this guy a cup of coffee. I don't know. It's in the license. But otherwise, we can't lift that or whatever.
• 41:02
  Speaker 1: You know, like, it's gonna it will take us, 6 years $35,000,000 to get the pure deal, all this open source software. And so they they they go, like, alright. We're just gonna do a license on it. We're just, like, we're gonna stop. And they we had a license on it.
• 41:16
  And amazingly, we had I'm like, oh my god. We're gonna find an a GPL stowaway that's gonna kill this thing. But there was there was nothing. It was all MIT and BSD, and there was not even any GPL or LGPL. It was really clean, And some Apache was basically all clean.
• 41:36
  But, man, you hit something like that in an audit from a big potential customer or an acquirer where they are thinking, like, look. Maybe you can make this decision, but this does not fit my risk appetite as someone with much deeper pockets. And you you it's like that this is part of the reason that that all of you can't adopt this. This is like, no. This is crazily risky.
• 41:59
  Speaker 4: So so funny little side story. We we actually just finished our first priced round a few days before this announcement, and we did a full license audit that probably would have tanked our round if this would have happened. We haven't even we haven't even announced the round because we got so caught up in this, this initiative.
• 42:18
  Speaker 1: Oh, man. Yeah. Well yeah. I mean, god. Thank god.
• 42:22
  I I think I you got it done. Congratulations. I got it done on that, obviously, on the on the raise. And, good. But, yeah, glad you glad you sold the house before the flood.
• 42:31
  That's, you
• 42:32
  Speaker 3: can I'm
• 42:34
  Speaker 4: sensitive on that one. I got I got hit by this tropical storm, and it took my roof.
• 42:38
  Speaker 1: Oh, man. Oh my gosh. I'm so sorry.
• 42:41
  Speaker 4: No. It's cool. It's a it's a it's a small patch. It's it'll be it'll be fun.
• 42:45
  Speaker 3: That's quite a quite
• 42:46
  Speaker 5: a few weeks ago.
• 42:49
  Speaker 1: All the
• 42:49
  Speaker 5: hacker news and Reddit comments on this, I think, so there's lot of comments saying I'm just a regular Terraform user. Why should I care? And I think to what to what you said, Brian, like, there's just this endless list of small places where this could trip you up, and they're hard to predict. And it's it does impact you a lot as regular user, and it's hard to know what will happen in the future, like maybe another license change. The the FAQ is apparently binding, but it gets changed every day.
• 43:16
  So who knows what's gonna go in there?
• 43:20
  Speaker 1: Yeah. Totally. And I also think that for those folks who are asking that question, like, well, I don't give you like that. It doesn't really affect me. It's like, it definitely does affect you because the the the folks that are gonna that are gonna contribute to the community and its vitality, it affects them.
• 43:35
  And that vitality very much affects you. I mean, you are that is to say someone who feels that they're more removed from this. It's like unless you're, like, a really active contributor, it is gonna affect you. And now, the good news is, like, you're probably not gonna, like, when the world moves to open TF, you're like, alright, fine. We're doing this now.
• 43:54
  It's just not gonna feel as disruptive because that's where the community is gonna go.
• 43:59
  Speaker 3: Yeah. It You Oh, I'm sorry. It it didn't mean to interrupt.
• 44:02
  Speaker 1: Go ahead, Josh.
• 44:03
  Speaker 3: I was gonna say, there was a lot of debate about what the vision for the port should be. And the 2 primary paradigms that we were choosing from, were, number 1, were essentially just a legal clone of what the kind of Hashi TF fork is doing. Well, I guess they're not a fork. Maybe I'll just call it Hashi TF. And and the model there would have been, like, we we just look at what they release.
• 44:28
  We make sure not to ever look at the source code, and then we just implement it ourselves. But then we realized that, you know, we can actually innovate. Like, one of the things that, HushCorp, kinda locked down was the only real interface to the Terraform libraries is the CLI interface itself. But they don't expose the libraries in a way that would make them easily usable by tools that we're embedding them because, you know, that didn't really make sense for their commercial interest presumably. So with open TF, we realized, like, wow.
• 45:00
  If we expose, for example, Terraform apply functionality as a library call, then people can do all kinds of cool things with it. And, it really expands the notion of what Terraform could be and how you could use it and becomes a lot more interesting and exciting.
• 45:19
  Speaker 1: I think absolutely. And I would counsel Adam, I know you counsel the same way that, like, you should not be you should be looking to your own use cases, your own customers for your own community, for what you do, not to not to Hashi. Yeah. And, I mean, we know we felt this way. We got I mean, DTrace too, but especially ZFS Adam, got a real burst of innovation when in in particular, the because all of the the the, the the featuring, the the feature identification item.
• 45:46
  Speaker 2: Exact exactly right. It was really interesting, actually. I was thinking about that example because, ZFS prior to, you know, in open Solaris, even prior to a Lumos envisioned a straight line of features. It was just version 1, 2, 3, 4, 5, and each version carried new features with it. And in this kind of diaspora, then we had different versions of ZFS and then open ZFS that had different features enabled.
• 46:12
  And before we had seen that, you know, even from inside, working on open slayers, we had seen that as sort of a bug. And then in a community, it was really a feature to enable lots of people to go their own ways and do their own experiments and enable their own products and customers without necessarily requiring everyone to move in lockstep on every decision.
• 46:35
  Speaker 1: I mean, it was a huge one. I feel like we had a bunch of those. And it was also helpful, honestly, and I imagine that there will be people like this from Hashi. I mean, I I can say that that I mean, I don't think I'm divulging any great confidence when I say there are lots of people in Hashi who are pretty, disillusioned with their employer. And I've heard from a lot of folks, gotten a lot of DMs from folks who are, and by the way, the advice I if you are at Hashi, my advice to you would be, this is gonna be a this is really gonna be very divisive.
• 47:08
  And, you I mean, Adam, you and I lost relationships permanently, over this issue, over the the the the acute disagreement of what the future of the software should be. And, you know, there's, there are gonna be like, these these disagreements are gonna be pretty sharp, unfortunately, try to find a, it's always important that you that we whatever entity you're working with, be it a community or company, you wanna find one that reflects your values. And I I think that you've done a lot of people at Hashi who are really wondering, you know, what happened to this place? Is this still the place that reflects my values?
• 47:55
  Speaker 3: Well, you know
• 47:56
  Speaker 1: But I think that I think that, Josh, sorry, though, a a long way of saying, I I I I I would I think it's almost a walk that you will have some number of people who love Terraform, and that's what brought them to Hashi, who leave Hashi and and contribute to an open TF, with their visions too. And it's like, that's it. Those are people you wanna have in the community, bringing a lot of expertise.
• 48:21
  Speaker 3: Yeah. You know, one of the interesting things that we were thinking about on the OpenTF side was about the philosophical decision, that HashiCorp made. And interestingly, the company, I represent Frontwork, we had a a sort of parallel philosophical decision. So, essentially, you know, the main concern that Hashi had is from their perspective, they were saying, hey. We're pouring all these resources into Terraform and other people are profiting from it.
• 48:46
  That's not fair. And and so the the philosophy comes in, into how you choose to respond to that. Personally, the philosophy that I endorse is, if you're the creator of an open source project or you're you're kind of sitting at the epicenter of it, that's a unique competitive advantage, and it and it's incumbent on you, as the the creator of the open source tool or or the setting of the epicenter, to figure out how to leverage that to make a better product, and and that's how you make it work. I think Hashi took the other tech, which was basically to say, okay. We don't like what's happening here.
• 49:28
  We we don't necessarily have confidence that we can compete, with the the products that we've got. And so we're going to just disable competitors from competing in this ecosystem. What's interesting add on to that is, groundwork has an open source tool called Tera grant, which is a thin wrapper around Terraform that is MIT licensed with, by the way, no plans to change that. But what's interesting is there are, companies out there that, leverage grunt work, or leverage Tera grant, in the open t f consortium. And, they make a bunch of money from it, and we don't see any money from that.
• 50:09
  And, but we spend money to create TeraBrun. And we think that's okay because we also get some benefits from creating TeraBrun, and we think we have some unique advantages by being the maintainers. And if we ever want to do more there, then, we can try to leverage those benefits and exercise our business muscle and and offer a valuable product to our customers. So Yeah. I Yeah.
• 50:33
  I'm sorry. Go ahead.
• 50:35
  Speaker 1: No. Go ahead, Josh. Sorry. Maybe I did. No.
• 50:37
  Definitely. I was gonna say that I actually, you know, we went through all of this when the open sourcing of the operating system was being contemplated at Sun. And, I mean, this is really the the beauty of software. Because the way I would flip it around is so you've got someone out there who's using Tera Grunt and is selling services on or what have you. There, it didn't cost you anything to engage them.
• 51:04
  Right? That that that's a that those entities have kind of identified themselves. Like, hey. Like, we actually, we share values. We share interests.
• 51:13
  We wanna we we're we're kind of partnering with you. And, you know, to I think it it is, it it it is, reductive to say, like, well, they're they're taking our work and profiting on it. It's like, well, you know, actually, the the the there are they are incorporating your work into something else that they're selling. But that is also it didn't you didn't have to have, you know, someone a business development relationship kinda build that. And on the one hand, it's not transactional one way, but it's not transactional the other way either.
• 51:42
  You didn't have to pay for ads. You didn't have to like, you this kind of this other entity is kinda self identified. And, boy, there's a lot of opportunity now to go, figure out what we can go build together. And if there's nothing to build together, it's like, well, it didn't cost you anything anyway. And I think that the the the big fallacy that I always saw inside the corporate walls is people like the choice that I actually have is that, that the this entity, pays me for Terraform, or this entity, does not pay me for TerraForm.
• 52:13
  And it's like, well, that's actually not the the choice you've got. The choice you've got is that because that entity probably has already, for Adam, all the reasons that you adopt the Terraform. It's probably already decided they're gonna adopt open source infrastructure. So it's really your open source infrastructure or someone else's open source infrastructure. And which of those do you want?
• 52:30
  Because that's actually the choice you're making. And if you at least if they adopt your open source infrastructure, there's an opportunity to go do something later together. If they adopt someone else's, it's like, yeah. You're just out of the conversation. And I feel like that's the the fallacy that that Hashi is is falling prey to, in part because I do feel that, like, for a generation that simply came up post open source, They don't understand.
• 52:56
  I mean, I I god. I wanna be the the like, a reread. I I don't know if you'd read it out today, but but Bill Gates' an open letter to hobbyists, 1976, looks like it was peeled right from Hashi's back or vice versa. Like, I don't think that they realized the degree to which you sound like a sniveling Bill Gates in 1976 talking about people pirating their basic. And, you know, I I I I think that there's this is probably the reason why I think the the the reverence for history is so important in the industry because it's like you gotta like, you're you're not the the first entity that has kinda struggled with this, but the the conclusion you're drawing is not the right conclusion to draw.
• 53:39
  And now that we I don't know. What point are we gonna talk about the video?
• 53:45
  Speaker 2: Yeah. We gotta talk about that video.
• 53:46
  Speaker 1: It was alright. Question for for Josh and and Nachman and Corey. Did you all watch the video?
• 53:54
  Speaker 2: It's a video of she made. I guess yeah. I guess being interviewed by a developer advocate there.
• 54:02
  Speaker 5: I couldn't make it all the way through.
• 54:04
  Speaker 4: Yeah. I hit it about a quarter of the way through, and I just I lulled myself into closing, the tab.
• 54:10
  Speaker 1: Great. Josh?
• 54:12
  Speaker 3: Yeah. It it was you know, the problem was it was like a full conversation. But,
• 54:19
  Speaker 1: obviously, like,
• 54:20
  Speaker 3: teed up questions.
• 54:22
  Speaker 1: We are here to report something shocking and delightful and open for debate,
• 54:28
  Speaker 2: I feel. That's right. So so, Brian, you watched it. You sent it over to me. And I was watching it, and and I had the same reaction as the other folks.
• 54:36
  I actually paused it, got up and took a walk, and then got my notebook because I I I wanted to write down some of the particularly odious things. But then I noticed something, which is that the developer advocate, and I'm not gonna name her, had this wild background. And and I and I DM'd it over to Brian. And, Brian, what was your reaction when I sent you
• 55:00
  Speaker 1: this picture? Hand on heart. Hand on heart, my reaction was Adam has had DALL E generate an image of, of formerly open source company announcing license change. And he has he has photoshopped it in this abstract image.
• 55:21
  Speaker 2: I was
• 55:21
  Speaker 1: just like, damn, he did that quickly. I was just like
• 55:24
  Speaker 4: and then a part of
• 55:26
  Speaker 1: my brain was like, no. No. Wait a minute. We recognize that from the video we watched moments ago. And indeed, it's so Adam has dropped this into the chat.
• 55:36
  Like, you've you've got this kind of, like, okay. So someone is saying, like, this is a psyop. Okay. So, Adam, why don't you talk about your hypothesis of this?
• 55:46
  Speaker 3: I think that
• 55:47
  Speaker 2: this dev advocate is is staging a quiet rebellion here. So the the picture to describe it is a, like, sort of blobby, baseless, mysterious, sort of human ish looking thing.
• 56:03
  Speaker 1: Looks like the offspring of Grimace and Jabba
• 56:06
  Speaker 2: Like the Loch Ness monster or something. Yes. Holding a goldfish and squeezing out gold coins.
• 56:15
  Speaker 1: The goldfish is vomiting gold. Literally vomiting gold.
• 56:23
  Speaker 3: It so Wait.
• 56:24
  Speaker 1: I mean I just love I love that there's this person who's sick.
• 56:27
  Speaker 3: Tell me.
• 56:27
  Speaker 1: I love that there are people right now that were, like, washing their dishes or walking the dog or driving, who just like, what the hell? Who just pulled over or stopped doing the dishes or stopped walking the dog and are now on their phone being like, I gotta see this. Like, this, it surely cannot be what they're describing.
• 56:43
  Speaker 2: It's like And so this is over the right shoulder of the dev advocate who's who's, ostensibly interviewing Armin, the CEO and cofounder. And I don't know, like, either she sits there and this has been over her right shoulder for years years, like the last scene in The Usual Suspects or something.
• 57:03
  Speaker 1: Or this is dead as a prisoner. It's like Cool. It's or this is some David David film.
• 57:09
  Speaker 2: Quiet, brilliant act of rebellion. She's like, okay. I'll interview you, I guess. Let me just change my Zoom background real quick before we do that. Because if you as you're saying, Brian, it's one step removed to having this blobby entity slicing open a goose and plucking out the golden eggs.
• 57:30
  That would be just only slightly more on the nose.
• 57:34
  Speaker 1: I feel it legit would have been subtler if they had some Mackenziean era miser wearing a top hat, holding an orphan by its ankle, shaking him loose of any bread in his pockets. I feel that would be subtler than this.
• 57:49
  Speaker 2: Oh, man.
• 57:50
  Speaker 1: It's just, shocking. So, yeah, I I'm sorry to distract us in this. Adam and I just knew that, like, we gotta get this out. We gotta we gotta so you and you all had you all seen this, first of all?
• 58:02
  Speaker 5: I did not notice this upon looking.
• 58:05
  Speaker 1: Right. You're too too big. Hey. Understandably. Like, I I I mean, we could Adam and I could barely get through the video, and we are nowhere near in the crosshairs of this thing, as as you all are.
• 58:17
  But, yeah. It's it's definitely, it's definitely something else, and it's it's, I mean, a little too on the nose. That's right.
• 58:26
  Speaker 2: And and and to folks at Hashi, listen to this. Resistance is alive and thriving, apparently.
• 58:35
  Speaker 1: The sign we will give the sign of grimace to know that we that we are one with the resistance inside of HashiCorp.
• 58:44
  Speaker 5: Everyone has, like, a goldfish tank on their desk.
• 58:47
  Speaker 1: Ain't that totally? Exactly.
• 58:49
  Speaker 4: I mean, I I get the coin, and I get the I get the gold coins, but I really do not know what the purple thing is. Like, what is the purple thing?
• 58:56
  Speaker 1: Oh, we know what the purple thing is. The purple thing is an announcement to our community.
• 59:00
  Speaker 2: That's right. The capital
• 59:01
  Speaker 1: That's the fact. That is the fact. You can see 0.26 on there.
• 59:07
  Speaker 3: Oh, man.
• 59:08
  Speaker 1: So when you all announced this, I mean, with the, I mean, I I certainly and I it sounds like there's a lot of kind of positive energy around the formation of this. But I you definitely saw that it feels like you saw that positive energy continue as you announce it. It seems like a lot of lot of grassroots enthusiasm for the for the effort here. Is that a fair summary?
• 59:28
  Speaker 3: Yeah. 100%. Yeah. Definitely. I mean, like I said earlier, the the the vendors themselves were really excited about working with each other.
• 59:38
  And then there are all these people that reach out to us to express support. There are some high profile folks who reached out and privately expressed their support as well. And, we had members of the Terraform core contributor team from times past express their support. Yeah. I mean, it the it was kinda neat to see, that manifesto get more and more names, companies, projects, and individuals endorsing it.
• 1:00:08
  And, you know, in in the beginning, I think we published it with, like, what do you got what do you guys think? Like, 7 names? Seven companies, maybe?
• 1:00:18
  Speaker 4: Yeah. That was just handful. It was about 7. And the first the first sign, like, signature came in and I was like, oh, somebody else cares.
• 1:00:26
  Speaker 3: Yes. Exactly.
• 1:00:27
  Speaker 2: It was
• 1:00:27
  Speaker 3: it didn't
• 1:00:28
  Speaker 1: just like
• 1:00:28
  Speaker 4: the wall of PR started coming in, and it was like, yes.
• 1:00:32
  Speaker 5: Yeah. And and, like, a a tip, if this happens to anyone else, do not do it like we did where people pledge by making an HTML PR.
• 1:00:42
  Speaker 3: Oh. Because the I'm talking
• 1:00:44
  Speaker 1: with the other elephant
• 1:00:45
  Speaker 3: in the room. Like, I listen I wanna tell you
• 1:00:47
  Speaker 2: the way
• 1:00:47
  Speaker 1: to manifesto, but, like, literally, every PR is a merge conflict.
• 1:00:51
  Speaker 3: It was brutal. Yeah. You know, that was my personal contribution, to this, project is I took, an hour of my life, and I remember to handle those merge conflicts.
• 1:01:03
  Speaker 4: And somebody's got some silly indentation, so I've had to turn off my, my save on format or format on save.
• 1:01:13
  Speaker 1: I but I can just imagine, like, you the you know, before you guys went live being like, alright. Look. I mean, if we have hundreds of PRs, I mean, you know, hurt me with that problem. It's like, oh, here we are. We're being hurt with that problem now.
• 1:01:24
  It's a good it's a it's a great problem to have in terms of, having so much interest that you've got merge conflicts. I was I was definitely man, I'm having a hard time landing this thing, so I'm having a hard time rebay like, having to rebase 6 different times to actually, like, touch down here.
• 1:01:39
  Speaker 3: But
• 1:01:40
  Speaker 4: but that's Think how many think how many signatures
• 1:01:43
  Speaker 3: we'd have if it didn't require a rebase. Like, we But
• 1:01:47
  Speaker 1: and I also I have to say, I one of the things that that you were, you know, I I you kind of see as sort of thinking, like, how we can structure this or whatever, structure the list. One of the things that I was kind of, privately hoping you would do, and it looks like you did do, Mike, you got folks that are really committing resources to this. And I really love the fact that that those folks were kinda at the top. Like, we've got a there's a bunch of signatories to this, a bunch of people that agree with it. But, I I it was really heartwarming to see the folks that were really being willing to to to make commitments.
• 1:02:18
  I'm sure they were seeing the commitments that others were making. And, you know, it was great to see. And and any, I mean, I any reaction from any official reaction from Hashi, I'm gonna assume the answer is no. You know what? Feel free to not answer that question.
• 1:02:39
  Speaker 3: So just to be clear, we actually tweeted from the open TF Twitter account, that we're officially waiting until this Friday, August 25th for a response. Because I remember in the manifesto, we said, look, we we think that no fork is better for everyone, if if that's possible. Now there's some legitimate questions you raised earlier about to what extent would we all trust, such a reversion to an open source license. But, you know, there's there's some overhead to a fork as well. It creates confusion.
• 1:03:10
  Speaker 1: Well, it definitely creates confusion. And to be clear, I think that there I I definitely think that there shouldn't be a fork. I just think that that the foundation needs to be in the driver's seat. I think that OpenTF needs to be the ones. And if Hashi wants to come back, they should come back.
• 1:03:22
  They should come back to participate in the new project, and the new project is is the Open TF project.
• 1:03:28
  Speaker 3: Exactly. So we're waiting for them officially to reply by, by August 25th, this Friday. But then they just published this sort of clarification of their, FAQ earlier today. So I personally interpreted that as a sort of implicit rejection, and I guess we'll see if if they respond at all. But, yeah, if if we don't get an official response by, this 25th, this Friday, then we plan to proceed with the fork.
• 1:03:58
  In fact, we've got several engineers in the consortium already working, on a private version of that and getting it ready. And there's a ton of energy. CI jobs are running. Terraform names are removed to make sure we're compliant with all the trademark requirements. Yeah.
• 1:04:15
  I mean, it's I think if if we were to have to revert, there'd be a lot of sadness. Like, it's it's gonna be Interesting. A critical project actually.
• 1:04:24
  Speaker 1: Well, and I think that that I don't think you again, I think that the the the end state here is not a reversion. It's it's Hashi coming to the community as a community member. Yeah. And I and I I think that, you know, part of the other reason I think that this is important is because I do think that Hudson has drifted too far away in the collective memory. I mean, to the point where it doesn't even make like, hey.
• 1:04:46
  Folks that are making a list of Oracle's top 10 open source transgressions, that's definitely Hudson belongs on that list. Like, I know, like, there's a there are a lot of things kind of vying for the maybe, you know, a top twenty list. But, you know, people have kind of forgotten about it. And I I do feel that that companies that have benefited from open source have benefited from that social contract, and indeed, that the whole reason that your technology was ever adopted by anybody was because it was open source. Those companies need to not take their communities for granted.
• 1:05:18
  And, it's important that the that the community, speak with its voice and as a check because you you know, what you want is that, you know, the the kind of the the the future company that is contemplating this. It it would be helpful to have an object loss and being like, boy. Yeah. But, you know, we could we create another could we end up losing everything in the process of doing this? And, you know, this should really right now, this is kinda thought of as being 0 risk, these relicensing changes.
• 1:05:48
  And they are, they're really divisive. They're very disruptive for the community. And we we really need to this needs to be thought of more like litigation. Like, you go into litigation, like, you can you're going to war, you can lose a lot, even if you're, if you're a plaintiff. And that's what you're doing.
• 1:06:07
  When you're relicensing, you are spinning the roulette wheel with your company on the board, and it it it's in your community, and you could lose it all. And, because the open source itself, like, the community is not gonna be vanquished. The community you on, Josh, sounds like this is kind of the realization you all had. It's like, wait a minute. We're we may be competitors, but we're they are or we're kind of competing in this broader infrastructure as code ecosystem.
• 1:06:35
  But we got a lot of shared values here, and we can we can bond together, to reflect those shared values in a shared project.
• 1:06:43
  Speaker 3: Yeah. Yeah. Absolutely. One of the things we actually did, in the first couple days of Open TF was we started thinking, with the product manager's mindset and working backward from the outcome that our customers would want. Our customers being people who choose technology, to deploy their infrastructure.
• 1:07:02
  And, we were thinking, okay, such a person would want legal clarity. They'd wanna know the tool is popular. The project is active. There's a clear vision, high investment in the project. It's publicly endorsed by, names that they recognize or foundations they recognize.
• 1:07:20
  There's commercial support available. There's a broad ecosystem available. It's compatible with their favorite tools, and it has features that they want. And, yeah, ultimately, we feel like in addition to the legal clarity that OpenTF can offer, a lot better than the Busil Hash Corp, TerraForm, we actually think the project can compete on its own merits even separate from that, pretty effectively too.
• 1:07:50
  Speaker 1: Yeah. That's terrific. There's a question in the chat about, and I and and maybe too a little too early to think about the the specific foundation of this. I don't know if you already have, whether you do OpenTF as its own foundation that would kind of pull in other infrastructure software or have you have you given kinda thoughts to that or are you deferring all of that until after Friday?
• 1:08:11
  Speaker 3: Oh, no. It it's just focused on Terraform. Like like all enterprises, focus is hard, but it it yields the best results. So this is a pure Terraform effort. It is unrelated to vault or console or anything else.
• 1:08:28
  We're just focused on making, the best possible version of OpenTF that that we can make.
• 1:08:35
  Speaker 1: That's great. And would you do that as its own 501c6? Or how would you, why am I uttering Rob? The,
• 1:08:48
  Speaker 3: I I it if you
• 1:08:52
  Speaker 2: want do you have opinions about the Linux Foundation? Do you wanna share about the Linux
• 1:08:56
  Speaker 3: Foundation? Anything?
• 1:08:57
  Speaker 1: You know, I just feel like I'm about to talk to my teenagers about a bad decision they're gonna come in. Like, look, you're gonna be at a party where there's a lot of Linux Foundation going on. And, you know, I just want you to know who you are.
• 1:09:08
  Speaker 2: That's right. Just be safe.
• 1:09:09
  Speaker 1: Be safe. And if you need me to come pick you up, I will. Have you given thought to the the Canada Pacific Foundation vehicle, Josh?
• 1:09:21
  Speaker 3: Yeah. So, you know, we wanted we we recognize that we need a governance structure of some kind because there's going to be multiple different companies all contributing resources. And there has to be some mechanism for how we set the road map, how we are reviewing PRs, how we respond to GitHub issues, and, you know, and things like that. So, we want to move ideally to the, CNCF. That's something that we've been looking into, and some of our members have had some discussions there.
• 1:09:58
  And, but most really most of all, we just want a sort of governance structure that we're we're not innovating on that. Like, let's find whatever the best open source governance structure is that can accommodate lots of different companies, and let's just adopt that. So that that's the real need. The the foundation is sort of the solution to that need.
• 1:10:22
  Speaker 1: And and truthfully, I mean, being in the OF, the CNCF does I mean, that that does solve a bunch of problems for you. So, I think that that may, like, now I'm pouring the kids a drink here, Adam.
• 1:10:35
  Speaker 2: Yeah. Just drink here with us where it's safe. Exactly. For the record, not one
• 1:10:40
  Speaker 1: of those houses. The, but I I mean, I I think you're wise to kinda have stay focused on what you're trying to the problems you're trying to solve and the problems you're not trying to solve. You don't wanna reinvent, obviously. Yeah. Well, I'm hoping that those those talks are fruitful because I think you do have I mean, this is software that is at the foundation of a lot of stuff, and it's actually it's really important, and it's important that it stays open.
• 1:11:08
  And, you know, I loved the the the kind of the tenor you struck in your manifesto about, like, we are trying to assure an open source Terraform. That is the objective of OpenTF. And, you know, useful ain't that. So
• 1:11:23
  Speaker 2: Just to dovetail that just to, especially for naive users, for folks who aren't familiar with any of this stuff to make the path unencumbered in a way where you know, new folks, could easily walk themselves into a situation they don't mean to. So I think the the work that you're doing here is so important to create or to actually take take that torch that Hashi carried for so many years and and built up with Terraform and these other projects and, lives up to that commitment or or that you're able to carry forward that commitment for new users, existing users, and for this really big and thriving community. Like, we're building Terraform providers, you know, for the oxide rack. We wanna make those open TF providers, and we're gonna guide our users there because it's gonna be the right place to use it.
• 1:12:15
  Speaker 4: Yeah. As hectic as
• 1:12:16
  Speaker 5: this,
• 1:12:17
  Speaker 3: Oops. I got it.
• 1:12:19
  Speaker 5: As hectic as hectic as this last 11 days has been, I do think in some sense, like, the the sort of silver lining is I do think, you know, OpenTF or TerraForm, whatever you wanna call it, is gonna come out a much stronger, better product. For sure. And like so I go if you go the foundation route, you don't have a single vendor in control of all the the features that go into it. Because there's, you know, a lot of people competing that would be part of this foundation, you don't, you get a lot of innovation and but everyone benefits from it. And so I I really think we we all win, despite not being able to sleep a lot from that we can do.
• 1:13:03
  But we're all gonna win in the end, I think.
• 1:13:06
  Speaker 1: Totally agree, Malcolm. And just for whatever it's worth, I absolutely think it needs to be in a foundation. I think the question is, you know, is it just mechanically, is it an existing foundation? Do you use the LF? I think that which sounds like do you use another 501c6?
• 1:13:19
  The the c three route is a much harder route, and I would definitely recommend the c six route. So, no, I think you're doing exactly the right thing by going to Foundation. The, for sort of trademark and so on, One, question that that that that did come up in the chat is like, alright. Well, so how do you recommend? Like, when someone does use your software and they are using it to as part of an offering that is commercial, like, how are you supposed to deal with that?
• 1:13:47
  And, Adam, you and I had this happen because the software that we developed is used by the one of the most profitable companies on earth, in terms of Apple. And, you know, you have to have the kind of self awareness to realize that, like, your software is not the only thing that that there's a product and a service that that company is developing of which your software plays a role. And it's kinda natural for me to say, well, no. My software is the only thing that's like, well, maybe maybe not. Probably not.
• 1:14:16
  It's probably more nuanced than that. And certainly in Apple's case, it's like, no. Apple's delivering a product to their customers. They're very they've been they're very innovation oriented. And, you know, the fact that they chose to integrate our technology, I think, we viewed as exciting and uplifting.
• 1:14:32
  I don't think we, again, because it's like the opportunity was not Apple, you know, the the a son or whomever, a licensing fee for this. Like, that was never gonna be on the table. It's like they use it or they don't, and I'd much rather they use it than not. I do feel that, like, those open source companies, I think that, you know, you and I know they got a bunch here. You really wanna focus on the the the the goods and services that that you make and serving your customers.
• 1:14:59
  Josh, I loved hearing that you've been talking to your own customers about it because, ultimately, like, that's just important for all of us. It's like the people who are gonna gonna pay you for your product or service, those are the ones who you ultimately we all ultimately are all serving. And that that's part of what it feels like is the the entity that's not being well served by Hashi's decision are are their own customers.
• 1:15:22
  Speaker 3: Yeah. Exactly. You know, every company creates value and captures a small portion of that value. And in order for the company to be successful, you have to have a well defined business model of how you're gonna capture value. And so that there's nothing wrong with that.
• 1:15:39
  Like, those business models that make companies successful allow them to create really great tools. I think the confusion comes in when companies mistake what the role of open source is in their business model. Yeah. Is it lead generation? Like, okay.
• 1:15:56
  Great. It's great lead generation. Or is it, a free tier of your product, or something else? And and so I I think that's one of the things that was so shocking about the HashiCorp element is I always saw it as as lead generation for them. Like, Terraform is the the gateway drug, and then you could try some of their commercial offerings, but there are alternatives too.
• 1:16:16
  And and they they've kind of turned it off to the point where, like, it is the product. You know, that it that it it's just the thing you purchased directly. So with Open TF, we're we're kind of stepping outside of that paradigm. We're saying, look, all these different companies need this thing to exist, so we think it can support a prosperous ecosystem. And each of the companies, well, we'll figure out our own business model.
• 1:16:41
  But the the core thing that you're using, OpenTF, you know, that's gonna be thriving independently into itself, which will hopefully inspire confidence when when users use it. Yeah.
• 1:16:52
  Speaker 1: Yeah. Amen. And, Josh, to which I would add in terms of the other kind of importance of open source, it's also an opportunity for a conversation with someone who with whom you share values. It's opportunity for potentially folks that that that wanna work for you. There's a lot of opportunity with open source, and there's a lot of energy that you get to harness.
• 1:17:11
  I know when when, Dave McJanet was announced, the new CEO of HashiCorp was out was at HashiCorp. I think it was in 2016. And, Adam was with you. I met Scott Hammond, I think. Did you meet Scott?
• 1:17:22
  Yeah.
• 1:17:22
  Speaker 3: She met Sure. Yeah. Absolutely.
• 1:17:23
  Speaker 1: Yeah. So Scott I was with Scott, who was the CEO of Joyant at the time. And, you know, there's so much energy at HashiConf. And the new CEO is there, but, like, nowhere to be found. Like, not on sync.
• 1:17:36
  And Scott is just like, oh my god. This is CEO malpractice. Like, let me be the CEO of Hashi. I'm like, hey. You already got a job, pal.
• 1:17:43
  Like like, hey. Check your check your ring finger, pal. Like, you're not the CEO.
• 1:17:48
  Speaker 4: But he's like, god, I
• 1:17:49
  Speaker 1: wanna be the CEO because I would be having all these conversations with people that are using this technology. There's so many conversations to be had. And his instinct is not, oh, I wanna go sell to all these people. It's, I wanna learn what they're doing. I wanna understand what they're doing.
• 1:18:03
  I wanna understand how the technology that we have developed is being used by you. And there's a lot that can come from those conversations. And, you know, there's a lot of opportunity there. It's just your Josh, just to your point, like, you're just gonna have to get creative, and in a really exciting and uplifting way.
• 1:18:24
  Speaker 3: Yeah. Well, this yeah.
• 1:18:27
  Speaker 1: This has been great. I this has been a, I we're again, I was super flattered. We are flattered at Oxide that that we that, you all looked to us to to contribute our name to it. We were, honored to do so, and, really excited to be a part of wherever you take Open TF. You know that you got a you got a kindred spirit and a supporter here at Oxide, and we wanna help you all out, any way you can.
• 1:18:58
  It's energizing to see see this community come together.
• 1:19:01
  Speaker 3: Oh, thank you so much. You know, honestly, a lot of us are fans of of Oxide and, you know, your, your your founders, including yourself, Brian, and, and Jesse. So, yeah, we're we're really honored when when you guys joined. So means a lot to us as well.
• 1:19:19
  Speaker 1: Well, that was awesome. And, Malcolm, I remember when we were talking about the podcast, like, don't like, I I'm a fan of the podcast. Like, I'm gonna I
• 1:19:25
  Speaker 3: mean, you didn't notice that was,
• 1:19:27
  Speaker 5: Yeah. You actually got the back catalog, and I was like, like, again? Exactly.
• 1:19:35
  Speaker 1: Well, that was I mean, honestly, that's so great, and we're just, excited that we can have this conversation today. And I'm really looking forward to to what's next for you all and and OpenTF. And, Corey, good luck with your roof. I'm sorry to hear about your roof.
• 1:19:51
  Speaker 4: Yeah. I was gonna get that straightened out here the next day or so.
• 1:19:53
  Speaker 1: No. Awesome. And and congratulations on your successful round. And, Josh and Malcolm again, great work with OpenTF, and can't wait to see where it goes.
• 1:20:03
  Speaker 5: Thank you so much for having us.
• 1:20:05
  Speaker 3: Alright. Definitely. This is great. Thank you.
• 1:20:08
  Speaker 1: Thanks, everyone. See you next time.
• 1:20:10
  Speaker 3: Let's see if I can figure out how
• 1:20:11
  Speaker 4: to log out of this
• 1:20:13
  Speaker 3: thing. I
• 1:20:15
  Speaker 5: guess, turn your phone off.