Could somebody post the logs of yesterday's meeting on getmonero?

I heard back about the impact of the proposed upgrade schedule on Ledger integration

and am told that the integration of CLSAG into their application can't be done by September 17, but may be done by October 17

and that no one else on the Ledger team can work on the Monero application at the moment

I don't suppose their app is open source?

It is

so if necessary, someone else could write the update

It is, and I'm writing back to ask if the relevant branch containing CLSAG work is posted anywhere

sarang: Honestly, I find that answer a bit frustrating

so end users can build it and install it themselves?

I need clarification, since my understanding was the necessary changes are already done, and just need to be rebased/merged at the appropriate time

The company has boatloads of funding, yet no one can supposedly work on an important upgrade for one of the more popular coins

hyc: Yes, but needs to be compiled, which is a PITA

Anyway, I'll ask for further clarification

I assume that there would not be consensus to delay the proposed timeline?

I don't see a problem with October. the last hardfork was November1 wasn't it?

Accomodating plenty of time for third-party wallets to upgrade is one thing, having to postpone it for a few months because one provider cannot work on it is another

with PoW changeover on Nov 30

I mean only because it's now unknown if Ledger would be ready

hyc: Release was Nov 1 afaik and actual fork date Nov 30

yeah that's what I meant

so another release on Nov 1 seems fine to me

But last year we kind of had to postpone stuff because we weren't ready

As far as I can see, we (+ the ecosystem) can be ready in two months, give or take

Personally, if Ledger can't be ready with a huge heads up/runway, we push ahead with clear messaging on why ledger support is lacking at HF

Very clearly mentioning it in all messaging about the release, as well as providing estimates on the Ledger-support update afterwards.

And just make it abundantly clear that we gave them a huge heads up but they weren't able to meet it, not that Monero itself failed to integrate it etc.

I don't see that as an option either to be honest

all that's going to do is irritate everyone. what's the benefit of an approach like that?

We already set teh date keeping in consideration that they needed more time. I'm strongly against moving the date up because of ledger. I agree with dEBRUYNE

There will be tons of support requests every day as well as frustrated users

I'll ask what the _earliest_ possible schedule could be in order to guarantee support

Then do we just delay until every wallet has full support on all network upgrades?

without promising that there will be support for such a schedule

Seems like a slippery slope IMO

we only have trezor and ledger to worry about at the moment

is trezor ready to go already?

hyc: MyMonero too

They have to make changes for CLSAG afaik

sethsimmons: I was more thinking of assisting Ledger with some stuff so we can move the timeline up

Probably Cake and Monerujo as well

don't they just use monero's libs?

<dEBRUYNE "sethsimmons: I was more thinking"> If this is possible, absolutely

Cake and Monerujo use the GUI lib basically

Ok, good to know

trezor said to let them know the height and they will be good

I don't mean that we shouldn't try to do everything we can to get Ledger ready in time, but rather if that fails, do we delay the network upgrade until whenever they choose to/are ready to release?

Trezor will have a small PR to the Monero codebase when we merge the CLSAG code and decide the height, I am told

But they made it sound like they are otherwise good to go

and apparently they've already been doing testing with their own test suite

At the end if they don't care of moving Monero up on their priorities, people will just go to the competing company that will do it, as it should be. We can give ledger as much support as they need, but adapting to the needs of a company which makes money on Monero is absurd.

sethsimmons: I just mean, here we are in July, and they're telling us October.

should be their priority not ours

if they miss their promised October date, *then* we rake them over the coals

but for now, they've given us advance warning. what's the problem?

Thats what I'm saying

I am asking the following things of the Ledger team now (the head of the coin integration team is now looped in on the emails)...

Not to move UP the date, but to keep the date we set yesterday, try to help them how we can, and be clear about the lack of support if they aren't ready

1. Will any additional changes to the Monero codebase be needed (aside from SarangNoether/monero 91e77ce) for Ledger support?

2. Is the new Monero app for Ledger (that includes WIP CLSAG support) available? If nobody else at Ledger can/will be able to work on it, perhaps someone else will if it's made available.

What is the earliest possible schedule (code freeze, binary release, upgrade) that could guarantee Ledger support? There seems to be a fair amount of developer frustration that Ledger support could be unavailable given the currently-proposed timeline, especially given all the earlier work that had been done on CLSAG Ledger integration.

^ that's question 3

[end questions]

Thats great

Seem reasonable? I'm trying to communicate the frustration in a constructive way, and offer possible solutions

looks good

Hopefully having the team lead in the loop will help to get more specific answers about timing

After this is all done, I'll make a much better effort to understand the hardware wallet development cycle

and avoid these kinds of scheduling problems in the future

sounds like the prblems happen on their side

If there were another developer who could work on the application, that might solve it

The problem is only on their side. I don't understand why we should bend to the needs of a company who cannot make a schedule to accomodate the needs of their customers

This may be primarily a case of bad timing, where the sole developer is unavailable

I don't want to throw anyone under the bus here

I'm operating on my clearly incomplete understanding of their broader development cycle

I feel like we are trying to solve somebody elses problems

It sounds like with a slightly different schedule, there may not be an issue

I'll wait to hear back on that

sarang: As far as I know, there are multiple people that have sufficient capabilities to work on XMR

Within the Ledger team?

Or more broadly

If their working branch is made available, maybe someone else can finish it up

Yes

At least cslashm and the person that took over

The person who took over is the person unavailable

And cslashm handed over responsibility already

My solution: 1) Announce the date and warn people that ledger might not be ready for the hard fork 2) wait for ledger's customer to complain 3) Ledger realize its paying customers are pissed if they cannot use tehir hw 4) ledger gets its shit together. Profit

Perhaps wait until I hear back from the team lead

If only a small schedule change would be needed, that'd be great

Not ideal, but great for users

ErCiccione[m]: That sounds like a mutually destructive course of action

I think it's a bad idea. We already set a date keeping in consideration their internal problems (because that's what it is). It's simply a wrong approach IMO.

It will only sour the relation with Ledger

sarang: Yes, but afaik he still works at ledger

I don't see why he cannot jump in

This is basically an 'emergency' case

Part of setting a date was to see if it would be reasonable for those teams, no?

What are you referring to?

The strategy you outlined

It will merely result in pissed off users, a pissed off Ledger team, and a soured relationship between Monero and Ledger

oh, that wasn't really serious :)

but i stand by my point that adapting too much to the needs of a company that makes money on Monero is wrong. Should be the other way round and also sets a bad precedent. Support as much as they need. Moving our schedule, no.

Well, the schedule was pretty arbitrary

Sticking with it simply on some kind of principle, if it harms users, seems unwise

At least having a candidate schedule has moved the discussion forward with Ledger

and there's some benefit to that

No. We are telling them that they do whatever they want, because we will adapt to "not harm users"

I see your argument, but I would at least like to wait and hear from the team lead

Perhaps the response will be "ok, we can make that work after all"

Perhaps not

I still think it's a wrong approach, but sure we can wait for their reply. The release date is set in the meantime.

I think sarang has a good middle ground.

We've set a date that we will do our best to stick to, and use to leverage Ledger to try and meet our generous timelines

If they respond that they can't do it until some very late date, we have another discussion on what to do then

But in the meantime we certainly owe them the chance to respond etc.

The ideal is that they realize we mean business and they have no good reason to hold up our network upgrade, and they figure out how to not have an entire coin support rely on a single person and meet our dates

If that doesn't happen we can move from there with more clarity.

I'll shut up after this, but we don't owe them anything. It's totally the other way round. Let's keep that in mind.

We owe them common decency to communicate

I don't think we owe them delaying the deployment because they have a single point of failure

OK, I reached out to cslashm again as well

Thanks sarang

For spearheading the comms around this

No problem; I want to make sure it's all handled in a civil way

sethsimmons: Sure, of course and since we care, we gave a generous timeline, we offerd support and we are even trying to fix their problems. I think that's where we should stop.

I agree with you

Just hoping it doesn't come to moving forward without support

So the status is that I've asked those questions of Francois (who now does the Monero development for Ledger) and the integration team lead, and also just looped in Cedric (who says he is quite busy but may be able to jump in with select tasks)

So let'

Is Cedric = cslashm?

Yes

Assuming so

thanks

yes

OK, he's contacting Francois as well to get more clarity on this

kinda missed the discussions in the last days but we can’t announce a hard fork without Ledger support, though that seems to be consensus anyway

Personally I see it as much more complicated. Seems to me there is much confusion and diverging opinions how to best react to this difficult situation.

monero includes ledger source code so if Ledger updates later we have to release again anyway

and I’m not interested in doing 2 releases :P

Well, there wasn't exactly a dearth of point releases. That could well go into one of this, in a worst-case scenario that is. Hardfork would be only once of course.

I see the whole story as growing pains, which will only increase with success, with bigger "ecosystem", more players, more interdependencies. It's hard to stay capable of acting.

Nobody plans to make life hard for anybody else, but how to decide if push comes to shove. Very difficult questions.

There are people who do.

Knowing how software dev works in reasonably big companies... If if fork date is moved to November then December etc. someone will always ask for more time

*Even if

Just set a deadline for everyone once, and then help whoever asks for help

3 months out, not to forget

There is already ledger CLSAG code in the Monero branch