Ethereum Cat Herders PEEPanEIP, 14 Sep 2020

Previous Meeting Next Meeting

⏯

youtube image

►

From YouTube: Peep an EIP #6: Onboarding EIP editors with Greg Colvin

Description

Peep an EIP schedule at GitHub: https://github.com/ethereum-cat-herders/PM/projects/2

Writing an EIP with Matt Garnett - https://www.youtube.com/watch?v=Vnu4m4bBqCs

Follow at twitter: Greg Colvin (@greg_colvin), Pooja Ranjan (@poojaranjan19), William (@William94029369)

Contact Ethereum Cat Herders at
Discord: https://discord.gg/sgdnxZe
Twitter: https://twitter.com/EthCatHerders
Medium: https://medium.com/ethereum-cat-herders
GitHub: https://github.com/ethereum-cat-herders
Website: https://www.ethereumcatherders.com/

A

Welcome to another episode of peep and ape and eip dedicated video series. Today's guest is greg. Colvin greg is a founder member of the fellowship of ethereum magician and has been an eip editor for over four years. I guess it's an honor to have you with us greg. You have been a great supporter of series and have participated with other speakers multiple times, but I'm not sure if we know enough about you, like your background, your first engagement with ethereum community etc. So uh let our audience know more about you.

B

um Yeah I've been in the industry for longer than I can remember, um yeah first program, the computer in 75, and uh I got pulled into ethereum by uh meng some years ago. 2015. I think, I'm not exactly sure when I was invited to be an eip editor, but uh it was relatively early on and uh at a time when we worked out uh a fair amount of the current process, and so that's where I'm coming from on that.

C

Was that actually one of the original rounds of the eip editors, like? Were you in that first cohort.

B

um No, there was an earlier cohort. um I'd have to actually check eip1 for the full list.

C

I think you appeared there, I don't, I does the ip1 actually have a list of the originals like as originals.

B

It has a list of everybody who is uh or pretty much has been an editor. I don't. I.

A

B

It's pretty much anybody with right permissions to the eip repository yeah. Do you.

C

B

C

What were you going to say? I didn't mean to.

B

The fundamental distinction is whether you have right permissions to the repository.

C

Yeah I I guess that probably could actually be the technical definition of an eip editor now yep yep yeah. uh Do you know how the original group of eip editors was chosen? Was it just sort of a natural social decision.

B

I think a few people decided um that there needed to be eips.

B

And I I wasn't part of that first, two or three people, so I don't know.

C

Going forward, um like you know, for you, for instance, what was the process of being onboarded as an eip editor? Was it some as simple as someone inviting you in and getting the permissions.

B

Somebody invited me in hudson, I think, or nick johnson or both.

C

um Has it become more difficult to become an eip editor since then? I feel, like it's been a, I seem to remember in a different conversation with you. We maybe highlighted that it's been a long time since the ip editor has been onboarded.

B

There's been few enough of us and we've gotten too busy to even on board people.

B

So we we brought on the car recently and I had trouble getting the other eip editors together online um and I think at some point I just said macau you're an editor, um then yeah, then the others were like you should.

A

Have asked and.

B

I'm like well, I tried to ask and they're like well, it's okay, but generally it's by consensus of the other editors.

C

um Has it gotten more difficult with ethereum getting larger and more complex and higher stakes.

B

No well, the work has gotten more difficult. Yes, there's there's more eips to deal with.

C

And I I'd actually had meant in the context of onboarding of onboarding new editors, so you're saying there's not much of the, I guess, fear or concern about letting people in who might be disruptive in some way. All.

B

Right, I don't know how to answer that question. um Okay, if people want to be an editor um and if there's a lot, you can do to help with outright permissions. In the repository just look around and there's a lot, you can do the editors work on just helping people work on proposals and get them ready for final call, and you don't need right permission to uh go to a pr and start commenting on the pr and helping somebody get it together.

B

You don't need right permissions to go to magicians and uh participate in discussions there. um Things like that and someone like macaw, starts doing that alex who's an editor. Now he just showed up and started going down the list of of uh issues that were just sitting there and pull requests that were just sitting there and providing us lists of ones that were ready to merge, and then I would merge them and at some point it's like.

B

uh Why isn't an alex an editor, so he can just merge them and we agreed that he should be and he was.

C

So I, um if maybe if I could rephrase that in different wording, you're saying that the original eip editors had a number of responsibilities and perhaps the definition was the right with the right permissions on the repo. But there are a number of different jobs that are associated with that and as a way into so to speak, the eip curation community. There are a number of tasks that need to be performed that the eip editors themselves are over taxed on and that's sort of like an entry into the community. So to speak.

B

Yeah, I mean we need to do a better job of of making a community of that um right. Now we're hanging on the jitter channel there. I don't know if we're gonna continue to hang out there or uh move to a discord that a lot of people are on.

C

Yeah, the ethereum r d discord does seem to be pulling in, like the majority of the technical.

B

C

I don't know about eip editor, specifically.

B

The the jitter jitter is nicely integrated with um jithub and we've been on it for a long time, but it's a discussion that we haven't had yet yeah. We do need more of a community just because there are things that people can help with that. Don't don't require an editor to do.

C

You mentioned a couple and I'm aware of a couple of initiatives for the cat herders, but in the interests of conversation, maybe maybe we could try and disambiguate a few of those um you mentioned with alex. um By the way, do you mean alex brigazassi? Yes? Is that the right alexa yeah that he would look at outstanding issues and then make the like? You know, make the changes, make the pr and wait for you to merge so.

B

Well, rather, he'd, look at the pr's um see that they were ready to merge and then let us know go ahead and merge these they're up to snuff.

C

uh Okay, so I'm glad uh so you're saying that that might that is one one entry point would be even streamlining the process highlighting what's ready for merge, what's not ready for merge.

B

C

A

I think, as a part of the eip ip group, we are like the categories and the hardware coordinator and many other people have joined together to you, know, streamline this process of eip and they are also looking forward to document uh the process of onboarding the ap editors.

A

So I think your input here could be quite valuable in the sense of like who could be the right person. As you mentioned, the people who shows up and start putting things together in terms of making highlighting issues which can be or cannot be merged. But uh has there been any instances in the past or would be like there can be some kind of consensus issue like one editor thinks that it is good to be merged, and that does not agree. How do we handle this kind of situation?.

B

It it mostly um excuse me a sec.

B

Don't want to cough into that live mic. um It hasn't come up a whole lot. There was one unfortunate time when uh uh chison did not want to merge the eip for lost funds because he believed it would put him crossways with japanese law and they're. Actually jailing programmers in japan that they believe are hacking in some way and others of us um felt uh obligated to merge it because it met our standards and as an american with first amendment rights.

B

I did not feel afraid of the law and he actually resigned as an editor over the issue and he continued to volunteer on things that didn't require actually merging, but uh that that was unfortunate, um but that wasn't a disagreement between ourselves. It was uh you know, a personal issue for.

A

I understand that part. uh I have also seen a few eips which are outstanding for a while, and uh I saw a few comments by one a few of editors saying that okay, fine, I find it okay, but I'm not agreeing to this part. If any anybody else would want to merge, go ahead and do that, but that sits for a very long time. Is there any kind of internal communication among the editors like uh how do how to proceed with any particular proposal.

B

Internal communication's been bad, I think, we've all just gotten really busy um those who are working working for pay on the project just get busy at that work and those of us who've, been volunteering only have so much time to volunteer.

B

So we need to get better on on that um and also on responding directly to authors um for any um well for all of us. As editors we either have jitter jitter tags or emails listed so to some extent. As an author, um you've got to take some initiative to contact the editors and just say: hey yeah.

B

I think this is ready to merge. You know I disagree with this one editor, but he's okay with somebody else merging it. So will someone else merge it, and so that's the answer on a number of things, as authors just need to take a little initiative to to push their eip shooter process.

A

Yeah, that's a good way. I mean, like general, understanding that they might be having some kind of internal communication to you know kind of these conflict resolving things, but as an.

B

Eip1 has that list um and we can make it more clear.

B

That um yeah, if you wind up with that situation, it's it's not likely. Actually that the objecting editor is is going to have time to contact the other editors and say there's an issue here: does someone want to deal with it? It's it's pretty likely to just sit there until the author, you know, makes a move to find another reviewer.

A

So in your experience, does any kind of uh scheduling will help like bringing all the editors on one page and see which proposal has been looked by? Some of the editors.

B

We we need better communication that the editors need to talk to each other better, which we used to do better when we had more time- and we have to keep an eye, keep an eye on the uh on the jitter channel right now, uh which I've been trying to do uh to every day or every few days at most uh take a look at what's there and and try and move it forward. um Man mccall was real busy there, which is why we invited him to just be an editor yeah.

B

So he didn't have to keep asking me to do things.

C

I mean the the process. Is one editor to make a pr? It's not like. I don't know if it does have. Even oh yeah github does support needing multiple signatures for merge, but um right now the process just means find any author who's able to merge any editor who's able to merge. No.

B

um Only any editor has fright permission, so any any editor can merge it. um Any author of a pr in the system can um can say that certain people have to give it a review yeah before it can be merged. So if there are people who's whose opinion you you want to have uh before you merge it, you can ask for that within the github. uh A pr system.

C

Is that a is that a common practice with the eyepiece to tag a specific editor to have them review? It.

B

um I'm not sure um uh it would be more up, it could be. Anyone doesn't have to be an editor. It could be any expert that you want to to review a pr before it goes somewhere somewhere on github there's a little box where you say reviewers, and if you click somebody off as a reviewer, they have to approve before jit uh before jana will merge interesting.

C

Okay, um so we did sort of discuss the ability to sort of enter. I guess what we called the the community of vip curation by helping to streamline the process of what needs to be merged and whatnot. um What other? What other tasks do you think community newcomers, people who want to like sort of join the band of eip curation?

C

uh What other tasks do you think there are for for them to do.

B

um Just looking for interesting eips to review, I mean authors try to reach out. They don't necessarily know um who, in the community, is going to be a good person for feedback and help. um So if you're interested just you know, keep an eye on the issues keep an eye, especially on open pr's yeah, and uh you know, look where the discussion is, and you know either participate on the magicians to move things forward.

B

If pr's are waiting for review uh review, the prs um uh send a message to an editor. If it looks ready or you need some help- yeah it's just not that well defined like a lot of ethereum stuff. It's just show up find something that needs to be done and do it.

C

I I guess one of the things I'm sort of fishing for is not necessarily a strict definition, but ideas I mean there were, I think, a few in what you just said. uh Maybe I'll try.

A

C

A little bit, one is creating technical conversation on the fellowship like on the magician's discourse.

A

C

So there's generating conversation around eips as a way of getting a review on them.

B

Every eip's editor has a link to where the discussion is happening.

C

um There's looking for people who have the kind of technical expertise can look for eips in those fields of expertise and review them themselves and then try to draw attention to their readiness.

B

Anyone who speaks anyone who speaks english can help with the editing.

C

Oh, that's a good point and thought of that just getting being able to help people who are coming from english as a non-first language background that.

B

Can happen yeah.

C

Statistically, it should even happen a majority of the time. I suppose.

B

Even if one person speaks english, somebody else could be better.

C

B

Yeah, I never know where to put commas somebody has to fix it.

C

Yeah yeah yeah actually get me a bunch awesome uh interesting. um This is actually a really relatively tangential point. um Eips are written in english. Is there actually a specific english they're written and can people write color with you.

B

um I think the spell checker accepts both english, um english, english and american english, um so interesting one or the other, um but not not other dialects.

C

um I was aware of one other initiative which I figured I'd mention. um I know that there have been cat hurters, who have tried to go back to like sort of the the um the the ancient history of vips and go through.

A

Vips that have been.

C

Open since, like you know the paleolithic blockchain age, which was like three years ago uh and contact authors just to see if they are still interested in seeing these sort of stalled and zombie eips see what can be closed and what can be moved forward. Also,.

B

C

So I'd assume that might be another good point for coming in uh yeah.

B

Mikko's been talking about a bot to just look for old ones and close them. um But if you contact the author, it's always nice.

C

um What are the reasons that you think the ips do sometimes get stuck like that.

B

um Sometimes the author loses interest and goes away. Sometimes they've actually been rejected by the core devs, but they're still sitting there.

B

Sometimes it just takes time I've I have one eip, that's been in various stages since uh 2016 and it's just not quite not quite time so.

C

What would that mean? It's not quite time.

B

um The core devs aren't ready to make that bigger change to the evm. The the community is going to need to want it banned enough.

C

I definitely have seen that there are certain eips that sort of um seasonally pop back up again, I suppose it's a similar phenomenon. Community sentiment can change over time. There could be more drastic need if new things are introduced into the ecosystem.

B

Yeah I mean like prague, pow's been proposed in may of 2018 and its status bounces back and forth, um but it wouldn't yet be approved. It wouldn't yet be appropriate to just remove it, because it's old.

C

I mean if we revealed tomorrow that there's a large asic cartel that's controlling a significant amount of the hash power. I mean, I believe, everybody's view of it would likely change.

B

There is but that's another story.

C

I I use the word revealed. um I.

B

To a large enough.

C

Population in the ecosystem.

B

Yeah, we don't get into it it's hard to prove um yeah yeah.

C

A

uh Talking about the efforts like uh I understand doing, editing job can be like quite stressful in terms of looking into every details, the technical details that should be correct, the formatting should be correct. So what do you think? How much effort is needed? Maybe an estimate in terms of per week or per month.

B

I I can't tell you, there's, there's editors that haven't done anything for for years, uh but they're still there and we haven't, we haven't bothered to revoke their permissions and we'd be happy if they became active again, but they're sort of honorary yeah. So really it's, however, much time you had to give.

B

So you could show up every every few months and there'll be an eip. You were particularly interested or skilled in and you sort of worked on that and then you went away for a while. It's not a big deal, it's nobody's paying you to do it, so nobody can make you do anything. Oh.

A

That was like my next point. I was trying to come up over there, so this job has been most of the time voluntarily, like done by people. Do you.

B

A

We can like streamline the process, I mean not me. I mean like the group that that are working towards it. uh We can also go for like fundraising and if we can define certain timelines certain jobs in my mind, it makes good sense to be paid for the editors as well. What are your thoughts on it? Like your experience and.

B

I don't have a strong feeling there um in most work with technical standards. Not many people are paid that directly. You know they'll often be working for another company in the industry. Yeah. The company might support them to use part of their time to help with the standard, usually not all of their time, because they gained and maintained their expertise by actually working in the industry so to a large extent, they're working on their own time, but it's because they have a company supporting them.

B

They can afford to do that, um but it's more rare that somebody's just flat out paid to do a job, and this and this kind of thing right, we'd, be happy to have such people, but um we try and automate it as much as we can.

C

I suppose that itself would raise the questions of exactly who's footing the bill and what the incentive structure becomes. Then yeah.

B

Yeah, um there's ethics there as an editor. You know that whatever your source of money, um you you do what's best for ethereum, not for your employer and where there's a conflict, you you have to be careful to act ethically. There.

A

So let us know about uh the challenges that are faced by editors, because this is an open source like, as you mentioned, many people shows up. Many people do not, but things things has to move on, like eips has to be approved and merged. So what are the challenges in the current process?.

B

Oh, the biggest challenge for me in ethereum has always been, but uh so.

C

Wait a second, you usually don't use version control.

B

No, I've used lots of version control over the years. I just find jet to be the least intuitive I've ever used, but people seem to vary on that um if you've, if you've seen the webpage that generates um phony jit man pages.

C

I have been sometimes.

B

Yeah it's there somewhere, it generates phony ones and they're almost they're. Almost right. You know it's just this switch takes the upstream of whatever unless you have a merged remote, in which case you know it's just like this is this: is one command? Why doesn't it do just one thing, but the whole notion of this is a wonderful decentralized system, except it all got re-centralized on github. So why am I going through?

B

Why am I going through all these hoops to say that it's decentralized uh and why is it so hard to just take out one file and change it and put it back, and why is it so hard when I changed the file and changed my mind to restore an earlier version?

B

It's like it's not actually version control. It's I don't know what.

A

B

But we don't have to get too far into that. One of the biggest challenges for me is just operating the machinery, um so people who are good at that um that helps.

B

And you know the other challenge is you'll just get eips. You can't understand because they're outside your technical competence, so you're either looking for another editor to help looking for other people to help um things like that um from.

C

What you were saying.

C

Sorry and meant to cut you off with that, I was gone yeah. What are you saying um from what you were saying before? It sounds also like um there's been a heavy increase in the pace that eips are are are submitted with, so I'm assuming there's also a bit of uh trying to keep up with the firehouse phenomenon. Is that correct.

B

I'm not sure I I haven't counted, uh somebody else might know.

C

I found myself.

B

C

An infographic for like eip is released over time. Like you know what the what the rate is.

B

Yeah, I think somewhere jit hub, will give you that number, um uh I think, for a repo. It makes a graph of of how things are coming in, but um there is a distinction like between corey by aips and uh ercs, because the corey ips have to go through the core devs and there's going to be more more expertise there, both applied to reviewing them outside the editors, um whereas for given erc uh there might be a more narrow audience and um the approval process is easier.

B

It doesn't affect the protocol, so it's just a standard, it sits there. It says if you want to do things this way, here's the details of how to do it and if we all do it uh according to these details, then we get along, um whereas if something has to go in their protocol, all of the clients have to agree to do it exactly the same way.

B

All the clients have to implement it. So it's a much higher bar.

A

This is interesting conversation like I'm, enjoying it getting to know more about insight of how editors and things work so for uh for those who have tuned it for the first time and generally, our conversation is around the ethereum improvement proposal. We talk and try to take community questions and uh bring in an author or an experienced person to answer these questions, and on that uh I would like to take some suggestions from you greg on for the people who are actually trying to streamline this eip and eip editors and other onboarding processes.

B

I don't know how much more there is to say: we've gone through a lot on that um the the core erc distinction is important. uh Just well erc is a good example. um It's not part of the core protocol doesn't affect the clients, but it's an important standard. Everyone who agrees to use it is doing things in the same way. Lots of other programs can make use of the rc20 standard, but it's not the only way to do that kind of thing.

B

There's other ways to do it, there's other standards for that, so there's, there's sort of a community just around erc20 and similar token standards, and that's to some extent up to the author um to sort of find the the other people who are interested in doing that, because if there aren't other people yeah, um why bother.

A

I I don't want to put you on the spot, but I can't let you go without asking this question about one major change that is going and that is like under progress about separation of eip.

A

Network upgrade and eips on technical specification. What are your take on that.

B

You're definitely different, I don't I don't. I don't know what the technical issues are and separating them in the system right now, um beyond the you know, if there's a status field, saying whether it's a core eip yeah, whether we need to start calling them something different, um I don't have a strong opinion. There.

A

That's good and that's like kind of getting an easy way out, but that's okay, because that is something under process, and I I think that the time will tell us what's going to happen with that. uh We are trying.

B

I would rather call them ercs, I just don't know how much pain that might cause to existing systems, so.

A

Yeah, I think, as for my understanding, the separation is for the uh you know, the network related and the general eips. Oh yeah, I mean I'm sorry, are you proposing the the rest of this standard should be renamed as erc's instead of eips, because uh they would be.

B

A

B

It's a discussion we need to have where I don't have really strong opinions, but but it is it's confusing to people. I can see.

A

Right, that's a different angle that that certainly requires a.

B

A

There maybe next time.

B

A

Yeah, so with this uh um series uh we are trying to uh let people know about current process of ethereum how things move around proposals. If somebody would want to propose something new, we have done an episode on writing of the eip, as well with the light line. That is very helpful to understand how one can write a proposal and submit in a way that the author and the ap editors do not have to push it back back and forth like to do work right.

A

So so I find that, uh like you, know, very helpful in the same way like this episode talking about how editors work is quite insightful for people with that, we have almost reached to the end of the episode and uh before we leave, I would like to let people know that greg has been showing up for this series multiple time and I hope to should see you can't see you with more and more proposals in the next episode.

A

We are planning to discuss one of the erc's uh that is eip2535 the diamond standard track with nick. uh We have already run an episode with him on the eip173, which is actually in the last stage, but uh last call, but that is there for a while.

A

So uh uh if I can bring it to a detailed attention that that is something also to be pushed forward, I mean I'm not saying that you have to merge it, because just because I'm saying so, but if it's waiting just because of what it has to look, give it a look, this may be the end of the episode with greg, but not with the conversation, I hope to keep on seeing you in many more episodes.

A

I would like to thank you greg and thank you everyone for joining us today. This is pooja london. Signing off, see you next week.

B