Opening v0.1 draft of a new CHIP, proposed in collaboration w/ Alex (MiniSatoshi).
BCH yearly upgrades always on Saturday!
Pros: Mon-Fri work week doesn’t obstruct upgrade participation/celebration.
Cons: Virtually none? Of course this can be disputed and updated as according to feedback and other perspectives.
Discuss. Feedback & ideas welcome.
The biggest first initial objection is that stakeholders such as node operators, mining pools and exchanges would likely prefer the upgrade to be during the work week and NOT on a weekend. If this is a strong preference or quite important, why was it not considered or decided on as part of the “May 15th schedule” (which can and does sometimes fall on a Saturday/Sunday), and to my knowledge it has never been raised as an issue requiring adjustment until this point?
Not endorsing or rejecting this, just a point of order.
A proposal in Bitcoin Cash (aka a CHIP) is by default never accepted until the relevant stakeholders have all been contacted and they have opted in.
From the text it looks like you didn’t actually contact any stakeholders, so I’m just writing to make that part of the process very clear.
Changes to Bitcoin Cash are by default not accepted. Ever. Even such small ones.
You can make them get accepted by convincing the stateholders that may be against this. And you should get some (semi) official statement from them on their support of this change.
I imagine the major exchanges, the block-explorers and basically everyone that runs a full node to be negatively impacted by this. Those are stakeholders I think.
I will be working with stakeholders and making updates in future versions.
The reason to put this document out was that I was advised it’s better to publish early and iterate in public - giving additional time for comment & community thinking - than to work up to a much more polished draft before releasing.
Edit: CHIP updated with a prominent note making this decision & reasoning more apparent.
No, but then again not everyone is interested in such social events like celebrating a technical event, or making celebration the priority to which a technical point should adapt to. The “pros” would apply only to a limited set of people actually interested in a social event around the time of upgrade. At the same time, weekend would obstruct people’s time with their family or whatever other thing they like to do on a weekend. I for one would prefer it be defined instead as: the 1st Wednesday after May 15th.
Also, the need to set a time kinda excludes 2/3rds of the world from participating in the celebration anyway, even if it is a weekend. With current time (12:00 UTC) USA is barely awake, and Asia is going to sleep or having beers with their IRL friends or whatever they like to do at 21:00.
Cons are:
Everyone who has to be online for professional reasons would have to spend time being available on a weekend instead of doing whatever other thing they prefer to be doing. No change means this con has to be suffered 2/7 of the time. Change means 7/7.
Opinion: sends the wrong signal, that we would sacrifice our “plumbers” private life time just so some other people can have a party.
The '23 upgrade was probably the biggest and most impactful we will ever have had, and it was on a Monday, and people who were interested into celebrations, streams, etc. still had a blast!
Remember that there were issues, too, not with node stuff but with an app - Emerald DAO app not working. Pat was able to quickly fix it and allow people to mint. But what if he was just enjoying his family time on a Saturday, or just out drinking with his friends, and we’d be pinging him to intervene? I know I’d be cursing this CHIP.
A lot of the feedback so far, including this comment, seems to have a strong impression that preserving the preferences of engineers should be a high priority. While it’s true that engineers are the ones implementing the changes, they are also a tiny minority of the potential people involved in the upgrade. There has also been no complaints about weekend upgrades in the past, it was just accepted as the way things worked. This is actually consistent with industry behaviour, as many software teams do rollouts during minimal user times (1-4am in their most active user timezones), on weekends or during holidays to minimise disruption. The upgrade is only once per year, and is known years in advance. I think engineers can work around 1 BCH upgrade weekend per year.
In contrast, the vast majority of people are not engineers. General members of the public (or online crypto sphere) might be quite interested to tune into a live stream on their weekend when they were already having a relaxed day at home, far more likely then they are to make time for it during their work week. Furthermore, all of the engaged non-engineer parts of the ecosystem - such as merchant adoption teams, promotional outlets and so on, benefit in the same way. Should we prefer to make things convenient for a few hundred engineers, to disadvantage potentially millions of BCH community members and/or potential converts whose interest might be piqued by joining an upgrade livestream on a weekend when they’re more at liberty to immediately jump in and be involved?
I will address this more in the next update, but it is quite crucial to understand that I largely have to pitch the CHIP to engineers, but the benefit is actually felt by the rest of the ecosystem (particularly the marketing and product launch parts). This might not be ideal from an engineering perspective, but ultimately we all want BCH to be used and BCH’s engineering is not where its current bottlenecks lie in the holistic quest for user adoption. The marketing and promotion parts need all the help they can get, and this would make a big difference.
The CHIP addresses this, and it’s no different to the way things currently are today - about which no one complains.
In a global community, there is no ideal timezone for everyone, but of the options, this does a pretty good compromise of giving most people the chance to be part of the upgrade (and immediate releases afterwards) if they’re willing to adjust their sleep schedule minimally to be involved.
By “some other people”, you’re talking about the entire rest of the world who we are trying to attract as interested, curious & engaged Bitcoin Cash users. It’s the opposite, a service has to run for maximum benefit & engagement of end users, not engineers. The reason sales, marketing, customer support & product departments exists in companies is all to facilitate end user happiness and ensure their needs are a top priority. At the end of the day, no users = failed project. Bitcoin Cash, as a decentralised community providing a product/service in a similar manner to a company is no different. We can make the most compelling money, or we can make a kind of good effort but not if it means working 1 weekend per year. I hope the BCH community can set the bar higher, and that those involved in the technical workings are able to be convinced of the bigger picture - particularly in light of BCH’s historically poor track record at attracting and retaining quality promoters, marketers & less hardcore end users. It’s critical we make things easier for those parts of the ecosystem if we are to expand beyond the very technical and very engaged diehards.
We did the best we could, but there was significant hassle involved in it. Don’t mistake “a lot of committed people making akward arrangement meant it went ok” for “it could have been a lot easier and a lot better than it was with just some slight adjustments to the date planning”.
This cuts to the same point mentioned above. Engineers frequently do rollouts at non peak times, like Saturday morning, so there is spare time to fix things before the Monday morning rush of regular commerce. As BCH becomes more and more integrated into the general economy, this extra weekend buffer to fix any last minute production exposed niggles is actually ideal, rather than a problem.
I understand the marketing opportunity here, but I think we need to be cautious about doing something like this.
I’ve had clients in the past who insisted on go-live’s/upgrades on a Saturday which meant I had to be around to supervise. I detested them for it and the employees at exchanges, etc, might feel similarly.
Rightfully or wrongfully, I don’t think BCH has enough weight to do this and the (in this case, deliberate as opposed to incidental) support burden might brew up bad sentiment in the they deliberately had to choose to do the upgrades on a f'n weekend sense.
People who are passionate about BCH would probably be forgiving but I don’t think that’s an accurate representation of who this might impact in practice.
It’s not preserving the preference if engineers would prefer a workday. Now they have to tolerate non-workday 2/7 of the time.
It’s also not just engineers, and it’s also not just people from our ecosystem - since BCH is supported by many multi-crypto services.
Also, it could be a larger affected set than you think: imagine a BCH merchant network and their POS system having a hiccup, and them then having to wait on tech support who’d have limited availability on a weekend and users would be delayed in their use of service, so end users would be affected too - and it would reflect badly on all.
I think podcasters and anyone interested in podcasts can work around 1 BCH upgrade on a non-weekend per year. See what I did here?
That’s your conjecture, here’s mine: general populace is not as affected by the choice of day as you seem to think. If I didn’t have technical interest, upgrade would be a non-event for me: I’d just go do something with my friends & family who are free on the weekend rather than tune into a podcast, and would read the news of the upgrade on the next day, or watch a recording, or just glance over Telegram to see if went as smooth as expected. If I was some staff - the upgrade would negatively affect my friends & family, not just me. “Oh babe pls understand, I have to be online now, I can’t go do the thing with you.”
How would we even determine if “rest of the ecosystem” actually cares about tuning into a live stream? There are plenty alternatives to market an upgrade - without needing it to be a live stream or a count-down or whatever. This countdown social event is a drip in the ocean of potential marketing activities, why single out this particular one which demands another set of people to make sacrifices?
If this argument is so compelling and key, why has no-one made a CHIP saying “Network upgrades only on weekdays”? Why have no support systems or engineers complained about this in the past? Yes there’s a difference between 2/7 and 7/7, but there’s also a difference with 0/7. 2/7 is clearly not optimal, it just accidentally happens to be the status quo. We shouldn’t accept the status quo just because it’s the status quo, if we’re going to win we need to make changes to the best possible solution and fix the problem once upfront, using the exact same logic as something like a blocksize auto adjusting upgrade.
Yes, I do - and we have. But my argument is not for the podcasters, it’s for the general public. They cannot work around an upgrade, in fact they are not aware of it until the second it’s already happening, instead they flow to what fits their time schedule and what crops up on their Twitter feed at the exact moment they can click on it and be interested.
This is equally your conjecture. You don’t happen to care about the upgrade besides technical involvement, but in a world where lots of exciting new announcements & hype is going on in the BCH community coinciding with upgrade day (as we have already seen in previous upgrades) it will be more interesting to the general person on that day.
This is a good question. I will try and figure out some way to objectively demonstrate general interest in weekend events. Perhaps there is some survey data or precedents that can prove my point. Why single out this particular one? Well, because it’s a pain point I have directly experienced, which can be improved, which is easier to fix once now than suffer with forever & which I believe will make a massive difference.
It’s not mutually exclusive with other marketing efforts, which as you know I largely dedicate my life to. Those other efforts aren’t stopping, this is just a problem I need to get fixed now so that it’s not a recurring pain in the ass for every upgrade livestream for the next 20 years.
Because it’s at worst an annoyance which the affected people have to tolerate 2/7 of the time, not worth the effort of trying to change it. People had better things to do? Like, cost of trying to make the change (“meta cost”) to 0/7 is bigger than the benefit of having 2/7 reduced to 0/7. The “meta cost” for defending against 7/7 is smaller, though, and the benefit of successful defense is bigger (5 vs 2).
Why has nobody complained all this time? I’d say because it was tolerable and they’d see fixed datetime as reasonable - due to consistency of it, and nothing prompted them to even give it consideration to some alternative. Now you brought everyone’s attention to it, you made a prompt, and are proposing to make it worse for that set of stakeholders, well - I am complaining - complaining right now Does that count?
I’m not even the affected party, I’m not responsible for any service impacted by node software, I can sleep through the next upgrade with no consequence. But since you opened the topic, maybe we’ll be able to make a stronger argument about setting it on a workday, hah.
General public tunes into many live-streams at various times and the general public gives these events varying priorities. Why would they be at a significant loss for not being able to participate in one particular yearly live-stream?
With exception of '23, upgrades were only technical Schelling points, only recently did we create celebratory Schelling points for the same date, and I played a part there, too, because I like to mark the upgrades, too: for '22 I had a bounty contract and for '23 I had Emerald DAO.
Thing is - I didn’t expect anyone else to adapt to me by adjusting the schedule to make celebration more convenient, I just did it without caring which day it will be, and if people wanted to participate - they could, if not, no biggie. Why is a live-stream deserving of a different attitude?
Summary of my arguments boils down to this:
Benefits
Unknowable and a drip in the ocean of grand scheme of things. Upgrade live-stream on a weekend won’t make a big difference VS upgrade live-stream on a weekday, especially with the context of all podcasts during a year. If you do 2/week that’s less than 1% of podcasts affected.
Of course you and some others prefer the weekend, but how do we quantify and weigh it against preferences of others? +(number_pro) * (5/7) ?
Costs
More easily provable, we have the rough numbers of services which would be affected. How do we quantify? -(number_con) * (5/7).
Which number is greater? Sure, you can speculate about benefits to general populace tuning into the live-streams, but then we can also speculate about potentially affected user-base by slower tech support response on a weekend.
If the benefits were convincing, then maybe the BCH tech staff would be convinced, but what about non-BCH tech staff? There’s a risk of causing grievance and them making bad PR for BCH, that would nullify the benefits. Fixed date is proven tolerable by the past upgrades, but here you’re proposing to intentionally cause grievance for marketing purposes. I just can’t see the net positive here, sorry.
I think it’s fair to say that GP has negative interest in this CHIP, but I will speak for only myself from here. The reason is not at all because marketing isn’t important. Marketing is extremely important. But because IMO the cost/benefit of this proposal is weighted toward certain and high cost with unknown benefit which I don’t expect to be high.
For example, I’m imagining the fallout from shopping this to all the people we need to regularly interact with - exchanges, miners, pools, services, infrastructure… and the reaction I expect - “Why are you taking my time with this?”. To which I wouldn’t have a good response. Maybe if it were a Tuesday or Wednesday, it would go “ok”. Could be better. Could be worse. I don’t know. Right now it’s “not a problem, don’t fix it” territory.
However, I want to say that my overall opinion is in contrast to @bitcoincashautist who, as I understand it, is suggesting a modification to a weekday. Barring some big and high certainty benefit that we haven’t seen yet, in addition to the potential stakeholder fallout, the significant meta cost of modification to the CHIP process itself is not desirable to me for any day of the week or other similar modification I have heard at this time.
Having said all that, this situation is super interesting and worthwhile. First proposal to change specifics of the process.
I have said it elsewhere but will repeat here - nobody knows the right thing for sure. We are doing our best to pick our path into the future, each with our own history, biases, vision and fallible skill at predicting the future.
My thoughts and concerns on this topic are well-expressed by Tom, bitcoincashautist, and emergent_reasons.
My opinion can be summarized as “if it isn’t broken, don’t fix it.”
I think the cost of discussing and implementing this change is far greater than the benefit gained from the change, which I also anticipate to be marginal.
I also think this topic has a high likelihood of becoming a “bikeshed” topic, further illustrating the high cost for marginal benefit.
I think we should just keep the 15th as it is. It’s static, consistent, and unbiased. And if it happens to fall on a weekend or a weekday, so be it. We mortals can’t control time, but we can control whether or not we impose arbitrary protocol on others.
There has been quite a lot of discussion about this, and my thanks to everyone involved. This is my first attempt at doing a CHIP, and it’s been an enlightening experience certainly.
One thing that has become an obvious problem is that the CHIP process is largely reviewed & discussed by engineers - so it’s very poorly suited to pitching an idea that is against some engineers’ preference(s) but improves BCH as a holistic product. That might just be the way the CHIP process is and will be, I don’t know, but I think it’s worth exploring further. “If it ain’t broke, don’t fix it” is one manifestation of this - it obviously already IS broken (at least for some parts of the ecosystem), that’s how Alex and I experienced this pain point and came up with this idea in the first place. Discussing the benefits or compromises between different interests is a separate problem we’ll have to address on top of even convincing people there IS a problem (and there is).
Another interesting and somewhat disappointing learning is how many questions and opinions we got (across a variety of public and private media channels) from people who didn’t even read the CHIP. We addressed several concerns very prominently that people brought up over and over again having clearly not read or considered the arguments presented. But that is also beholden on us to clarify the format and document in such a way as to make that clearer.
Alex and I have taken a ton of feedback and notes on what has been said and we will work on a version 2.
That’s because writing the CHIP kinda comes last - it writes itself by testing the ideas in public discourse first. Your “job” as a CHIP owner is not to get the change that you yourself prefer or write down just your preference and perspective, but to figure out what change is good for BCH and just be a vehicle for it, it’s a humbling process.
Discovering this is the whole point of the CHIP process. It doesn’t mean that you have to get what you want though, but sometimes it means BCH gets what it needs.
This is probably the best point made in ithis whole discussion. Using this point I’d like to suggest a slightly different approach to Jeremy and others.
What about having the party discussion and stream on the first Saturday after the actual upgrade?
It helps becaues those engineers that are hard at work monitoring stuff etc are currently unable to actually attend the streams, due to being hard at work. Most of the hard working “plumbers” made an NFT in that first block, to prove my point.
There is in actual fact no compelling reason to have a live stream. I mean, there is literally nothing to see anyway. Nothing physical happens. Nothing to observe. Except maybe a website specifically built for this.
I have full faith in people like Jeremy that are good at this kind of stuff to instead make those PR moments a couple of days later, on a Saturday. With that special made website showing something relevant.
Because not only is BCAs point accurate, it goes even further in that those plumbers are hard working and can’t actualy enjoy the party themselves, so we have to ask why. Why have it at the same time…
This is certainly a suggestion worth further consideration. It’s true that “nothing” happens, particularly in the ideal / likely case of a smooth, successful upgrade. There is still a “live” element to celebrating the first actual block that comes through and the first usage made of the new features and so on. I’m not sure that can be replicated with a later on promotional stream, but it’s a good idea to reflect on.
Just finished the BCH Day #6 stream for Satoshi’s Angels.
It was great I think in the end, but there was a ton of production problems & distractions from the fact that I (and everyone else involved) was pulling this together last minute on a Tuesday afternoon when we’d already had work & other committments. All of this could have been avoided if it was on a weekend.
This is a perfect illustration of what I am trying to point out with the CHIP. This kind of basement stream might be good (and even ideal) for the hardcore, but it isn’t going to get us to global reserve currency.
More to come on that, I am working on a revised version of the CHIP, although I don’t expect to make the 2024 May upgrade (or 23 November lock in) but perhaps the year after.
Arguably, if this stream isn’t valuable enough for its production team to make sure they have August 1 available for themselves, then this demonstrates the marginal benefit given to changing the upgrade day to a Saturday.
If this production is so important, then it should not be done on a volunteer basis. It should be funded appropriately - such that its production team can afford a single day off known a year in advance.
Funding is a separate point, and one I agree with. We do need to get better about funding BCH promotion (although I am clearly self interested in saying that).
However the argument is only partially on behalf of the media production team & mostly on behalf of the potential audience. Saturday is a good day to get maximum possible interest from the general public, which is ultimately who we need to buy in to the project.
This will be discussed & addressed in the updated CHIP.
don’t want to be rude, but for 203 views (YT 9 days after streaming), I would suggest simply doing it on Saturday (or not doing it at all). I mean, its your time. You can decide how you spend it. But it doesn’t look like its worth the investment.
Does that count?