Do Buybacks Make Sense? | Lucas Bruder

I feel like buybacks are good investments when the market doesn't care about them. Like right now, buybacks are probably pretty good because token prices are depressed a lot. You're buying back tokens for pretty cheap. But no one really cares about buybacks right now. Like, they're trying to find whatever the hot narrative is. And many months ago when things were super hot and you had some DEXes on Solana and other protocols that were buying back the token, when they would announce that, it would have a big impact on the price.

People seem to really appreciate that, but the token price was super high at that time. So from the investment standpoint, it's not super great. This episode is brought to you by Sablier, the leading onchain solution for token distribution. You'll hear more about them later on in today's episode. Nothing said on Lightseed is a recommendation to buy or sell any investments or products. This podcast is for informational purposes only and the views expressed by anyone on the show are solely their opinions, not financial advice or necessarily the views of Blockworks. Our hosts, guests, and the Blockworks team may hold positions in the companies, funds, or projects discussed.

Hey guys, welcome back to another episode of Lightseed, our weekly roundup episode. Today, a little bit of a mixup. We've got Lucas from Jito joining us for what's normally the in-house Blockworks roundup for Lightseed, alongside my typical co-host here, Carlos. How are we doing today, guys?

What's up? Pretty good. Happy New Year.

Yeah, doing great. Excited for this episode. Happy New Year.

Welcome everyone to 2026 if you weren't paying attention. Happy to have you on today, Lucas. We wanted to chat about one, a new dashboard and some new data and research you guys have been digging into over there at Jito. So we'll talk about that, and then I think on the back half of this episode dig into there's a bit of a buyback debate that's been going around the community again. Feels like we go through this one every 6 to 12 months and we're back at it again. So we'll chat on that later. But maybe to kick things off, I want to talk about IBRL. Pretty common topic for the Solana space. I guess you just dropped a new dashboard and some metrics around validators and really like a block explorer essentially that digs into some things I guess. Can you give background there and then we'll dive into it?

Yeah. So yeah, we started looking into a lot of block packing metrics in like middle of last year on Solana. So how are the blocks being built? How are the transactions being sequenced? Are there any like funny games that are going on or things that are maybe not necessarily Solana aligned? And you know, as you mentioned earlier, IBRL is a common saying in Solana. It means increase bandwidth, reduce latency. This particular, at least this version of the dashboard is mainly focused on reducing latency.

So yeah what we started when we started to look at the data we wanted to look at how the blocks were being packed and none like current explorers really go into that. So, you know, if you go to Solscan or you go to Orb or the Exalona Explorer, you can just pull up a block and you just see all the transactions and they're kind of just like it's like here's just like a blob of transactions that were in the block. And basically on any every network except for Solana, that's how the network works. So on Ethereum and a lot of these L2s, they emit the entire block at the same time. And so like it literally is a blob of transactions. On Solana it's built differently. It's like the transactions are streamed out as the block producer is executing them.

So there's no wait 400 milliseconds and emit the whole thing as the block producer or the leader or the person producing the next block. You can basically emit transactions anywhere in that 400 milliseconds. And what we started to discover is that there's some validators that are start like kind of backweing the transaction packing. So it might be easiest to like walk through this what this kind of looks like if you want to pull up the dashboard here. But yeah, like let's just pull up like Helios's validator. They have a pretty high score here. So you can scroll down. You can see this recent block here. This just from some epoch here.

So we're actually looking at this is slot 391391943 and this is that single slot or that block that was produced here and you see the 1 to 64 on the bottom. Those are basically like time slices in proof of history on Solana. Essentially, each window is like 6 milliseconds on Solana. So you can see how transactions were packed within each six millisecond time window in this particular slot here. So the gray is votes transactions are votes on Solana. Green is non votes here. So yeah, you can like scroll if you go to previous, you can see some of the previous slots that they made in the same epoch. You can kind of see the green spread throughout the slot. This is kind of what we want to see.

And yeah, if we go back to the validator page and just choose one with a like kind of a lower score here and scroll down to this tick distribution. So this one looks way different, right? So you can see the there's a six millisecond time time slices here. There's not that much green in the beginning. There's kind of some like blips of green here and then all the green and all the transactions are at the end. And so, when we started looking at the data, we're like, well, I think we like knew some of this was going on, people certainly that are like super in the weeds have been chatting about it for a bit, but when we like started talking to the average person on Solana, even some validators, they didn't really know that this was going on.

And so, we're like, oh, we should probably like build a dashboard and educate people on this. And so, that's essentially what we build. And, you know, you can poke through a lot of blocks on here and we tried to come up with like some rough scoring algorithm to score the blocks and how they're built.

Yeah, I guess maybe just at the high level if you can talk through you talked a little bit about the the transaction packing and we kind of you know we looked at the chart there for a couple validators and some of those slots. How does this kind of break out in terms of this IBRL score, the slot time score and the transaction packing?

Yeah, so this is kind of the first like iteration of this. You know, we wanted to get it out. If you go to the methodology tab at the top, it kind of breaks down how the score is achieved. And so, yeah, there's a 35% slot time score, 25% vote packing, 40% nonvote packing score. And so slot time score, it's basically like how long your slot is. So if you are I can't remember what the exact number is. If you're like over 380 milliseconds for a slot, then your score kind of exponentially decays the longer it goes.

Why 380 instead of 400?

Well, that the default slot time in Agave is actually the main validator client slot is 360 milliseconds. Everyone says 400 milliseconds, but it's actually 360. So you can see like if there's agave validators that are running more than like 380, some are running like 400 or 440 millisecond slot times, you know, they're modifying it to basically stretch their slots out. And what that does is like if you're more time as a block producer, you can you have more time to kind of ingest and get all these transactions and produce higher fees. So that's 35% of the score. The other 25% is a vote packing score. So basically this incentivizes this is showing like how votes are packed within a particular block.

So we saw instances of we can find a few examples of these later, but we found instances of like people packing votes late for some reason. And basically that just like delays the confirmation time. Like if you go to a website in Solana and you make a swap, you're waiting for that transaction to confirm. If people are packing their boats late, then that kind of delays the transaction the update on the UI for everyone. The other 40% is non-vote packing score. This is just like user transaction swaps and it's kind of like it's a little complicated, but basically it's like trying to incentivize packing throughout the slot.

So just trying to incentivize like the packing of transactions throughout all the 64 ticks instead of just at the beginning or just at the end. And so this is kind of like a we try to be like as detailed and share as much information to the score that we can. The scoring will be open source at some point but and it's this is just a vzero of it. I already see there's like some instances where we kind of need to like tweak some of the the parameters for it, but yeah, it kind of just gives you like a rough idea of like what is the score and how's this validator doing?

The this is like first of all this is incredible. The visuals are really great. I do have a question like as we saw like for instance in the Helios visual like that they it's quite uniform the way they like submit the nombo transactions whereas with the galaxy example they waited until the end to submit all the transactions like the obvious implication of this is that latency is like reduced in the Helios case and not in the Galaxy case but what is like the incentive for Galaxy and others to like wait until the end of the block to to submit all the transactions.

So yeah it essentially there's a incentive for validators to you know earlier I was talking about validators can process transactions throughout the slot and propagate them and that's what we see in Helios they're propagating it throughout the slot but you know if you essentially if the challenge is that Helios might get a bunch of super high paying transactions at the end of the slot and they're like, "Oh, I can't rewind time and like fit these in because I've already filled up my compute units for this block." So, there's essentially incentive today where it's like, well, I might as well wait till the end and see everything that I can and then peck my block.

And you know, that's not super great for Solana. We want to be the decentralized NASDAQ. We want high frequency trading to happen here. And if that's kind of going on, it's you know, not super great for that vision.

Gotcha. So it's essentially like sitting around and waiting to see at the end, you know, and then filtering out like what are what are going to be the highest paying transactions and then stuffing all those in and basically dropping the rest and to the next.

Yeah. Yeah. There's actually some let me see if I can find an example of this. There's some other like kind of interesting side effects from this too where if you if you're packing, how do I say this? Helix is packing throughout the slot. So every validator is like able to replay that live and like be essentially caught up with the the the chain and like be on the very tip. But what we see in some of these validators that are packing towards the end of the slot is it actually causes every other validator on the network to briefly run behind for a bit because they are just catching up to all this work.

So yeah, if you like pull up one of the the validators that has like a low pack score, if you look at their votes, some of the votes are actually like kind of shifted towards the middle of the slot because all the other validators, they're trying to like play catchup and like process that giant chunk of work. And so, yeah, there's been some like interesting stuff that we've kind of like it it makes sense when I explain it, but it's not something that would come to mind when you like first see this.

So yeah you can see here if you go to previous so you know this is slot 750. You can see a bunch of these transactions are weighted towards the back that all the green stuff is kind of weighted towards the back here. And then you go to next hit next. So then you can see this is the next slot and a lot of these votes are kind of like shifted more towards the kind of towards like the middle of the slot and so that's kind of showing the side effect. This isn't as severe because they kind of like spread it out through the last few ticks. But for these ones like go to like if you go back to validators and go on like the second validator page and then go to like Coinbase 02.

Yeah. So yeah, this one's running the harmonic validator client. And so you can see here that there's a lot of votes that are kind of like shifted 100 milliseconds in and all the the non- voting transactions are emitted in a single tick here. So yeah, there's just some there's a lot of interesting data going on in chain. I feel like none of the current explorers like went into this level of detail. It's like here's all the transactions and there's no like really tight timing data on this.

Gotcha. These are these are interesting because I think yeah to your point I would say that this is data that otherwise would have had no clue about prior to a dashboard like this being kind of readily available. I mean I guess you called out some trends like do you think now that there's that the data is kind of readily out there and the community can can jump all over it. Is there an opportunity here for sort of like a kind of period of of optimization to sort of work through?

Yeah, for sure. Yeah. I think it's kind of like the slot timing stuff that happened maybe like beginning to mid of 2025 where there was a bunch of validators that started lagging their slots to like 420, 440, 500. Some were even going way above that and like ultimately it was a protocol issue that some validators were just kind of taking advantage of. But when we saw dashboards that were released that showed all the slot timing data, validators extending their slot times, a lot of validators took that back and like tuned it and you know they kind of got like socially shamed into that.

So you know people there's data here. We kind of took like a first pass at scoring. People can kind of come to their own conclusions on it, but yeah, I think that the goal is like, hey, here's more data and people can look at it and try to figure out if this is like something we should be optimizing for or not.

Do you think I saw Carlos shot over a tweet for me from Benedict Brady kind of commenting on the dashboard saying that a lot of you know maybe fixes or sort of changes to how you know block packing and TX packing are handled could be brought in protocol via future upgrades. Do you think that's sort of like the optimal solution here?

Yeah, I think the what he's talking about specifically is MCP. So multiple concurrent proposers that will bring block production into protocol. I think that's inevitable. The timeline for it is unclear depending on who you talk to. Some engineers at Holland say 6 months, some say like 12 to 18 months. So I think the the timeline for that is unclear and I think it's inevitable. It's going to happen. But I guess it's just like what do we do until then? And we think that like hopefully we can avoid these games taking place at basically these these validators are like kind of chasing yield at the detriment to the execution on the network.

And so, you know, we hope that this will kind of slow it down. And we we're pretty bullish on BAM as like the near-term solution to this, trying to essentially make everything more transparent, more deterministic, not really optimized for these like yield chasing games and for focus more on the execution. But yeah, I think that it's Anatoli mentioned MCP in his white paper and that's like Anatoli's baby and Max's baby as well. So I think it's inevitable that that happens, but the the timeline for that is kind of unclear right now.

Yeah, I guess while that is implemented, like I don't know if it will happen like in the next 12 months or or whatever. Like we really don't know. But do you expect like social shaming let's just call it like uh to make this validators change their behavior or do you expect like stake to meaningfully change based on the scores like do you think people will unstake from X validator and deposit into a validator with a higher score or or you just like don't know?

Potentially. I mean, it's like the when people didn't really know the slot time lagging was happening until people built dashboards and that kind of at least mitigated some of it. And then there's the all the work that Ghost Logs have done with sandwiching and there's been a lot of stake that's been removed from that. Ghost logs is a website called it's sandwich.me. So, it's showing validators that are sandwiching transactions like users swapping tokens. The amount that that's happening has gone down a ton since they released their dashboard because like validators are getting smarter and like getting shamed. The users are getting smarter as well.

And so yeah, the goal is to like highlight this information that this is happening and you know people can come to their own conclusions on it because I think that a lot of we can't really expect staker all stakers to be as sophisticated as those that are like super deep in the weeds on the networking things. Like we talked to the ETF issuers and people that are running super large odd air companies and when we explain this to certain people there that a lot of this just kind of like goes over their head. You know it's okay they're not they don't have to be in the weeds on everything here but I think building data or like visualizing data like this is helpful to kind of help them make better decisions and help stakers make better decisions.

Yeah, exactly. Like I would assume like valvators with a heavy institutional stake to not really like be affected by this. But more so with future network upgrades. You mentioned like BAM could potentially help with this. Like I'm curious if that's because of application controlled execution or any other like variables that BAM introduces.

Yeah, that and then just the like the verifiability of the sequencing. So knowing what sequencing is happening and when it's happening and just being like super transparent about that and not really trying to we don't really want to play those games like these kind of like yield chasing games at least on this like in this dimension. I think there's other ways to play these kind of games to grow the pie versus like trying to you know play spot timing games and late packing games.

Thanks to Sablier for sponsoring today's episode. Sablier is the leading onchain solution for token distribution trusted by teams like Maple, Immutable, and Axi Infinity. The number one distribution protocol on EVM is now live on Solana. Automate your airdrops and token investing streamed by the second, transparently, and trustlessly. Secure, audited, and ready for Solana Builders. Get started at sablier.com. As always, investments in blockchain technology involve risk. Terms and conditions apply. Do your own research.

One question that comes to mind from my end is like we've we've talked about like MCP, we've talked about BAM, like do you think that reducing slot times could potentially help with this like holding the slot time?

Yeah, I think it helps mitigate it to a point. So yeah, if we have if we cut the slot times in half, then you will likely still have people that try to play these games, but you know, when when you when we were showing like the single tick packing, like this super tall vertical green line, right now that's happening every like 350 or 400 milliseconds. So if you half slot times, then you would probably half that, too, which seems like a step in the right direction.

So yeah, I think that shortening slot times and making the slot times more deterministic, which should come through Alpenlow later this year, will be helpful for these Alpen glow will help with the slot times. The late packing stuff. It's like you I think you just need to show data and then explore some potentially in protocol options on like a super long time horizon. But that stuff takes a while to ship for sure.

On cutting slot times, like what what do you think is the limitation? Like I'm assuming at some point you kind of run up into a wall for the you know the hardware or for the network to kind of operate at kind of small slot times.

Yeah, there's probably some there's probably some like data propagation changes that you could do to help shorten slot times. So yeah, there's I think there's one change where like you basically propagate shreds like the the block the mini blocks faster. Maybe you could potentially use something like double zero to like multicast them out so the validator isn't constrained by networking. And then past that there's some other stuff that might come like async execution which I think is I think Alpenlow is a requirement for that but async execution you should be able to get much shorter slot times.

So yeah I think there's some engineering stuff that can be done to fix this.

Yeah. And what are the like second order effects of solving this? Like let's say we reduce lot times we implement MCP gotobam is widely adopted in the network like what would you see change from the users perspective or from a market makers perspective or like yeah from the end user?

Yeah I think that users will get better pricing and market makers will have more certainty over how their transactions are ordered when we first found this data we talked to a lot of market makers on Solana. So like you know all the big ones you've heard of. We talked to prop amms and they had no idea that a lot of this stuff was going on and we're like oh this explains actually a lot of what we're seeing on the network with like some specific validators. So you know when they when when those things are implemented then it will just give market makers better certainty in the comfort to quote tighter and ultimately that will result in better pricing for users.

Yeah. I I wonder if you think or how you think transaction landing services come into play at in this because for instance like I'm thinking of humidify which is closely benefits from Nomi right or there are other transaction landing services on the network like do you think this is this will remain like relevant going forward?

I mean I think this I guess we're kind of like the first transaction landing service, but I do think on a long time horizon, it's kind of a commodity service, I think. So, I think we'll just see those kind of there need to be like new business lines for them. Cuz in the past it was really hard to land transactions and the the current state of the network's like not super great with all these transaction landing services cuz there's like six or there's probably like 10 transaction landing services and so people will send the same transaction to all 10 of them.

So if you actually like look at the validators and the the number of transactions they're getting, there's a ton of duplicates, but also there's something called durable nonses that people are using where they basically only pay the fee once. So what they'll do is they'll be like, "Oh, Danny has a transaction landing service. I'm going to send my swap to him and tip his address tip his like business account because he's going to try to land in it." and I'll send this the same transaction or a different transaction with the same durable nons to Carlos in his landing service and tip that and both of those make it to the validator but only one can win and so there's a ton of like duplicates of the exact same transaction but also there's duplicates of different transactions with the same nonses that validators have to filter through and yeah it's just a really it's causing some interesting kind of side effects on salon on the networking and the scheduler and everything because you have all these transactions that make to the validator that are just never going to land because like people are sending the same thing to like 10 different parties.

Yeah. I I wonder like I I I'm like going on a tangent here probably, but like the goal of SIMD96 was originally to like cause these transaction landing services to drop. And I've seen from the data that priority fees have been a bit more more favored like recent in recent months. Do do you expect that trend to continue and and like me tips to decline over time?

Yeah. Yeah. I think the the one thing that I'm not the biggest fan of with CID96 is that none of it goes to the stakers and protocol. SIMD23 was approved and hopefully that gets rolled out this year. Ultimately, like the stakers are giving their stake to the validators to give them the power to produce blocks and the stakers should see some of that revenue. Doesn't have to be all of it, but it should be some of it. And there's no in protocol way to do that. And yeah, we just haven't been spending a bunch of time trying to optimize tips right now.

So yeah, I think that the the amount of priority fees will just the percentage of them compared to tips will just keep going up.

Maybe to kind of close out the discussion on the IBRL dashboard. I guess Lucas kind of just want to give you the floor like going forward here now that you have this Vzero out and I think this episode will be out a few days after it's already been publicly announced as of the airing. But I guess what will you be looking to kind of refine there and you know maybe dig a bit deeper into the data to find solutions for?

Yeah. So, yeah, I think we will probably be like tweaking some of the scoring parameters a little bit. So, you know, keep an eye out on the website for those and we'll probably announce those as well. So, yeah, trying to tweak the scoring for some of the packing stuff. And, there's a ton of other interesting data like grouping by client ID. There's definitely some like specific ballot air clients that are not packing blocks in a super favorable fashion. So being able to group by that. And then yeah, there's a ton of other data that will look into like account locks and cost like how validator is actually packing different types of transactions throughout slots.

And there's some interesting like clustering that we've done to kind of like cluster different validators together or clusters them based on how they pack their blocks so you can kind of get an understanding of who's doing what. And yeah, I guess whatever the community tells us to add here. We've already gotten some good feedback and suggestions on things to add to. So, we're going to keep investing in that and basically building what the the community wants for this.

I do I I do wonder since you listed off some things there, like I know D0ero also has, you know, sort of their interesting kind of latency dashboard and I wonder if maybe the latency information combined with kind of the slot data might also reveal some some trends that you could see.

Yeah, we were we were talking to them earlier about some stuff here. So, we'll I think that'll shake out to be pretty interesting. But yeah, there's been it's been like a pretty helpful tool for us internally on BAM and just seeing like what the BAM valid errors are doing because we've kind of discovered some like performance bottlenecks in it. So, it's been helpful from that standpoint. And then I mentioned in the blog post, but there's just like some other interesting stuff that we've seen where it's not necessarily validators doing something malicious. It's like there's actual like for instance in Asia basically every data center provider in Tokyo in like the Asia region there all their vote scores went really bad like a month ago.

And I think it's because there's a lot of there's a lot of submarine cables that are running fiber between countries that are down right now. And so the there actually have like a kind of a not super great vote score because there's actually like network like worldwide networking issues. So we've been talking to some of the data center providers on like hey what happened here and are trying to help them figure out what happened and fix it. So, yeah, it's kind of just a there's a lot of data to kind of gro on the Salon blockchain and, you know, we'll kind of see where the data takes us.

Cool. Love it. We will, uh, stay tuned and stay updated on, uh, what's going on in the data. I think, uh, we couldn't have Dan Smith today, but, I know he's going to want to geek out on a future roundup as well, so we'll probably dive into uh, some more examples of that data as well.

Thanks to Sablier for sponsoring today's episode. Sablier is the leading onchain solution for token distribution. Trusted by teams like Maple, Immutable, and AXI Infinity. The number one distribution protocol on EVM is now live on Solana. Automate your airdrops and token vesting streamed by the second, transparently, and trustlessly. Secure, audited, and ready for Solana builders. Get started at sablier.com. As always, investments in blockchain technology involve risk. Terms and conditions apply. Do your own research.

Jumping along to I guess our our back half of the episode, second conversation to dive into. One that I think we've had probably plenty of times on Blockworks podcast. But I wanted to talk about it again since it's been so hotly debated again. The buyback debate. And I think probably the broader debate here is like more of a growth versus buyback debate. But generally speaking there is this there has been this ongoing question of you know should a crypto protocol should these early startup protocols be turning on buybacks buying back their tokens with revenues.

I think a bit more of a heated discussion got kicked off recently with Helium the H&T token as well as Jupiter both talking about I think within the same week or a few days saying you know considering turning off buybacks or we're going to stop doing buybacks of the token. They either haven't had an impact on the market or in hindsight maybe it was just a waste of money. I guess I I maybe just to level set like what are our general thoughts on like this you know doing buybacks in crypto with some of these early stage protocols or buybacks versus growth like what are our thoughts there?

Yeah like I can take a first stab at this. I think the main reason like protocols do buybacks like as a general rule is to just send a credible signal to the market that they care about the token. And the reason for that is that there's no other way or there wasn't any way to do that like six months ago other than doing buybacks. So I I think bump is probably the most clear example here where when they launched the token like people were really like unsure about how value occurred between the labs entity and the and the pump token. So they started like buying back 100% they started to buy back the token with 100% of the revenues.

The problem is that they're essentially like destroying capital to some extent in the sense that that buyback is programmatic and they are buying back the token regardless of the market cap or the PS it's trading at. And also like it's unclear to me if the market is actually rewarding tokens that are doing the the buybacks. As we've seen like with Helium and Jupiter, I think part of the reason they're trying to or like proposing stopping buy is because the tokens are like down 80 plus% from all-time highs and we haven't really seen like the market reward that activity and that that money could have been used otherwise for growth or or other initiatives. But yeah, curious like to hear like actually from Luca's perspective what he thinks of all of this.

Yeah. Um yeah, a few months ago we went like super deep in the weeds here. So we added buybacks a few months ago. It's unclear if that really affected the price. I think it's hard to make that distinction. And it's kind of and then we were kind of we've also just discussed like a staking model as well. So you can imagine that's kind of like another alternative here where you provide value back to stakers by kind of taking some of the the revenue and redirecting it to people are staking the token.

Yeah, I don't know. I think I it's kind of hard cuz like on one hand you want to like I'm here for the long term. I think most builders are and you want to reinvest in growth. That's like the most important thing. You don't see startups that are like buying back shares or like high growth tech companies that are buying back shares. They're kind of like reinvesting in growth. At the same time, it's like you have a a market that's trading based on a few factors and you know, maybe people are counting buybacks in there. So, you kind of have to like balance the two. And I guess the the other point is like I feel like buybacks are good investments when the market doesn't care about them.

Like right now buybacks are probably pretty good because token prices are depressed a lot. You know you're buying back tokens for pretty cheap but no one really cares about buybacks right now. Like they're trying to find whatever the hot narrative is. And you know like many months ago when things were super hot and you had like some DEXes on Solana and other protocols that were buying back the token like when they would announce that it would have a big impact on the price. And people seem to really appreciate that, but the token price was like super high at that time. So, from the investment standpoint, it's not super great.

So those are kind of my thoughts on it. Like, I don't know if I'm like for it or against it. I think I'd probably be for it in like a market like this, but I think it also depends on the phase of your project. Like we actually just put out a JIP geo improvement proposal I think it was like two weeks ago on pausing buybacks and using the the revenue for growth. So yeah, I think we're kind of in growth mode with BAM and like would rather spend the money incentivizing the adoption of that versus continuing to buy back JTO. But that will like probably change in the future depending on what phase we're in.

Yeah, I I think you both raised great points. I think you know, Carlos, you started off with this idea and I think we we talked about this quite a bit honestly on ZeroX as well as Lightseed like this idea that for many protocols like just announcing or committing to a buyback was kind of this like good faith way to signal that you know the token wasn't just like a memecoin. I guess but you know that be kind of became it almost became like a bit of a requirement from the market of like well if they're not going to you know if the protocol isn't going to announce a buyback then why is why would I buy the token or why is the token price going to go up and so then it kind of became like you know a necessary card to even play in the game so to speak like over the past maybe year or so.

And so then we just saw kind of like in large part a big chunk of you know all revenue generating protocols kind of committed maybe in some form or other or at least talked about you know at some point hey yeah we'll probably turn on the buybacks. But I think the the nuance