0:00
isn't it they're saying to this this this guy like can i just buy a bottle of coke and use your toilet because
0:06
otherwise it's going to be [ __ ] real world cup badly exactly like
0:12
oh [ __ ] did we just
0:20
[Music]
0:28
hello welcome to game of nodes uh the weekly podcast from independent validator teams on the cosmos and if you
0:34
live by the sword you die by the sword so i should have guessed that that was going to happen when i was asked to
0:40
describe whether i was hung over on sunday morning when i had to do a long cycle across the french and swiss alps
0:46
so um this week we've got mercifully little drama per
0:52
second so unless something happens during this podcast we've got a few things we want to talk
0:57
about um a couple of test net bits um i think that is about it um we've got
1:04
some guests coming up on future shows but we won't uh we won't spoil uh who that's going to be
1:10
um and we of course got my friend and colleague callum
1:15
on tonight's show um joining us from the bed knobs and broomsticks bread
1:21
broomsticks bread bed knobs and broomsticks bed that's a bit of a tongue twister uh
1:27
yeah genuinely is is that your bed because it is a ridiculous bed yeah yeah it is my bed
1:34
and why are you coming to us from like marry antoinette or something from
1:39
the shea long as the peasants burned down the the town nearby what's what's going on here
1:44
well kraken just listed juno so i'm a very rich man now so you know
1:51
it has gone up i'm
2:08
oh yeah actually in today's exchange rate it's probably pretty damn close i'm not gonna lie
2:15
everything's on years across the board 15 16 17 yeah yeah yeah
2:20
there's it's a good it's a good news day today uh for the price of uh of juno and
2:26
uh that that actually that news came out i think earlier i did a stream um
2:32
with our our dear eccentric friend don crotonium um and i think the news came out while
2:38
we're on the stream and we literally didn't notice all the things you came back saw my message and you're
2:44
like wait this makes my life so much easier oh yeah i mean like there's been so much
2:49
talk about fear off-ramps and stuff on junior and that's just like [ __ ] done it's just like it's just cracking and for all
2:56
well i don't know actually i i don't know if i'm the only validator users certainly we their business stuff is very easy to use
3:02
in the uk um so yeah they literally we gave
3:07
yeah right okay cool so yeah we're all the same they they were just we tried a couple of other providers and they were like
3:14
really slow and unresponsive and crack and we just like provided our company details registration all that kind of
3:19
[ __ ] and they were just like bam yeah i think you will recall when i was trying to make my binance account and i
3:26
was pulling my hair out for two weeks trying to bloody uh was it strategically pulled out in order
3:32
to make the the kind of mohawk go yeah that's the side sloping mode
3:40
is it technically a mohican no no no no it's a mohawk isn't it that's a mohawk the mohicans there
3:48
i think he's looking at that wow mohican would be like yeah it's like a people right yeah the
3:54
mohicans the big one isn't it this is this is a mohawk yeah um yeah so once on a postcard is that a
4:01
mohawk uh literally a comb over [Laughter]
4:08
i mean yeah i guess kind of i mean in the sense it's literally combed over
4:13
i don't know i've never seen your head from above null so uh it's not on google earth either so um
4:20
actually that okay i'm not gonna say that that was weird i was about to say did you know a lot of roundabouts
4:25
when viewed from above or actually elliptical shape not not round that black can't bleed my mind
4:30
um and you stopped yourself and then you continued yeah then you then you just said it anyway so anyway
4:38
it's obvious when you think about why it's because a lot of them were built some time ago are on a slope and so the
4:43
people that built it didn't didn't build it straight like so probably
4:51
i've seen some look like a peanut some some like a triangle yeah like you
4:56
know they all come in different shapes and sizes there's a there's a famous one in a market town in cheshire where it's too
5:05
uh that kind of close enough that they're they screw up each other's axis of
5:11
gravity but they're not so close that they're just one compound double roundabout
5:17
so all that happens is all these cars arrive in the center of this actually quite small town village
5:22
and then they're just like oh [ __ ] i'd have no idea how to drive through this double star-shaped semi-roundabout
5:30
situation and everybody just kind of grinds to a halt and there's a massive traffic jam there 24 7.
5:35
it's amazing that's reminds me of the uh a the is it in milton keynes where there's the the
5:42
roundabout with eight mini roundabouts around it this is the image it's the only interesting thing about
5:48
milton keynes by the way so there we go okay there's a lot of good
5:54
uh architectural facts about milton keynes my favorite one is the one about the um
5:59
you know the central train station no i've never stepped foot in milton's so it hasn't
6:05
it has an easter egg right which is that the street leading up to the train station
6:11
is on a lay line leading to a stone circle
6:16
which is a reference for the summer solstice so the the sun rises on the morning of the
6:23
summer solstice in milton keynes and illuminates the front of the the glass
6:28
front of the train station it comes up through the glass window so a little pagan temple just like
6:33
casually slotted in by the architects of milton keynes one of the weirdest that's my polite way of saying not a
6:40
particularly nice place in the uk uh we'd get we'd get tom on here properly
6:46
we need to bully him on so we've obviously met face to face and obviously my feelings on reading are pretty
6:51
they're a matter of record but um he's a lovely man from reading so you know it just goes to show you you can
6:59
the place might suck but the people are lovely that's right there you go
7:04
that's my thought on that i had to google a mohawk and mohican
7:10
so there that difference that exists there you go that's the nightmare roundabout that's the one
7:16
i was talking about so you have like about five different mini roundabouts to get onto the main one
7:22
yeah because it's like a coming together of like six different roads yeah over engineer problem guys it's it's a
7:29
this is an amazing disaster of utopian architectural practices of the mid-20th century isn't it
7:34
look at that and also oh my god how many words did you just say that i don't understand
7:41
well there was a whole period wasn't there where they were trying to build model towns like ideal towns and then they didn't get married is definitely
7:47
one of them yes it's just one massive housing estate it's one of the towns of the world
7:53
that's a huge overestimation of human intelligence right there um well the the
7:58
well in garden city is quite green but a bit weird and also
8:03
they only built they only built like what a quarter of it and then they kind of gave up i think something like that
8:09
um most uk development projects can be explained by that though it's like oh we just ran out of money and then we
8:15
stopped there is a little bit of that i have to say being being on the continent like
8:20
last week in a country france that has higher taxation and i use it or lose it
8:26
attitude as far as i gather to um public infrastructure though you'd just be like in the middle
8:31
of nowhere and there'd be a bunch of lads resurfacing a road that had some potholes because they're just like wow we got the money we got to use the money
8:38
we just like this is in the middle of nowhere like i went up one road and it was like
8:45
it was to a hamlet there were two houses in the hamlet and then the road kept going and you could literally avoid the
8:51
whole thing so it was a whole motorway that meant it wasn't necessary i didn't see a single car in like 45 minutes
8:58
and yet the road was pristine brand new it'd been resurfaced obviously within the last 12 months i was just like
9:06
in almost any other country this would have just been a dirt track by now yeah yeah which would be gravel like
9:12
that was once a paved road i feel like we've broken null with this roundabout i think this is like the the
9:19
how you get through it yeah yeah if you want to go from yeah mm-hmm that's wrong though look that great one
9:25
in the lower left is or lower rates going the opposite way oh wait no it's not no it's not so you how you how i've
9:31
been told you treat it is you go one roundabout at a time and you treat the mini roundabout as your entry onto the
9:38
next so if you haven't done roundabouts before you just break it down although i imagine it's still hellish and i
9:43
wouldn't want to drive across it yeah so you got to keep keep left on the little ones and then
9:50
yeah because it swaps directions the middle one is going a different like
9:56
direction to the um yeah so that of course you don't just slam into someone yeah
10:02
that's navigateable i already imagine you landed on that on your driving test you'll just be like well
10:08
well well i'll go again and hope they don't give me the route with it yeah you'll just be like well this is
10:14
this is obviously impossible and i failed my driving test so anyway
10:20
people stuck in the middle just how the [ __ ] do i get it forever going around looking at it
10:26
in the middle yeah get a tow truck i can't figure it out um i i saw a picture on twitter the
10:33
other day of um somebody shilling uh schilling rhino
10:39
um usurper and they were they had loads of they had a bunch of pictures of of some quite
10:45
cool looking racks and stuff and i was like a wreck are those actually your racks or are they stock photos of rats
10:53
no you can do a search those are our racks those are your racks yep oh wow we're legit got your racks out for the
10:59
lads yeah oh my gosh sorry i was going to hang around
11:05
all the time yeah also had a wireframe rhino which was pretty cool yeah that's awesome
11:10
the [ __ ] out of you is was that um
11:19
yeah we did we started uh february
11:24
january sometime earlier this year yeah where'd you get that cisco stuff off the back of the truck real cheap
11:31
sort of that's good hardware though so we're actually that that picture has like uh
11:37
it's actually old there's only eight servers in there we have ten and actually we're just getting ready to split this up into two dc's
11:44
so we're going to move it to two separate data centers we're going to try to cut the bill down actually the bill will be overall higher
11:49
but it'll be split between two dc's and then we're gonna put two horcrux nodes one in each dc they're about 30 minutes from each other
11:55
and totally i mean if the whole region gets wiped out then they're gone but assuming that doesn't happen at least
12:00
they're at different power grids and they're on you know different uh internet lines and things like that so
12:06
um we're starting to realize that one giant rack of stuff is probably not the best idea
12:11
so it's cool uh yeah not if it's all in the one place no it doesn't really help and when you
12:17
have that much going down at one time i mean we're getting around that with horcrux but when you still have that much stuff in one spot it's not great um
12:24
plus they like private dc's like this they they still have maintenance windows so like for instance we know we have a
12:30
30 minute outage coming up at like 3 a.m on august 1st and getting around that is a real a real pain for nodes that you
12:37
know aren't signing two of three type of things so it just becomes kind of a pain so having that be on two
12:43
separate dc's allows us at least to you know make sure that we still have our own nodes running somewhere type of
12:49
thing so what you're gonna like split your gear yeah one why not
12:54
have your nodes in two separate dc's but have your signing raft like on something that's more available like
13:01
um digitalocean or aws or something i wouldn't say that those are more available
13:08
honestly but we've had like big aws outages we've had linode outages i've had ovh fall off
13:14
like we've had everybody's have hurt hetsner outages things like that even for small periods but they're a little bit i don't know
13:20
they're they're probably but you you could buy more nine if you have like a three
13:26
a two or three uh raft you could just distribute it between like leno digital
13:32
ocean and aws that's what i have right now i have one at one of our one at cogent one at aws and one at ovh
13:39
all east all like within a couple hundred miles geographically and then um
13:45
and then those nodes are also in the east and so i like to bring one of those more into house and so if i'm going to
13:51
move i don't have to move i don't have to know the signing notes really i mean so you want them like um so when you're
13:56
putting them with your well i mean it makes sense to have one with your colos anyway because yeah you're close to your
14:02
notes but also if those notes go out then you've still got so you're going to have one
14:07
one at each color and then one in the wild is that the that's the idea yeah right now we have one one in the color
14:13
and two in the wild and we'll move it from two to one and if that doesn't really matter all three can be in the wild
14:18
it's just like then i know at least like if the and i can put a priority on a ra on a raft node now so i can have that
14:23
close to the node so if it fails over it fails over far and it might miss blocks because of latency but at least if it's close then
14:30
it's at least next to maybe one of the primary nodes type of idea you'll definitely get a better hit rate if they're like in the same colo even
14:38
right now across three different um three different providers regionally it's like 20
14:45
milliseconds a piece like in between each of those which is totally fast enough like that's fine like once you
14:51
get like 100 150 like then you start it gets a little bit crazy but but anyway
14:57
yeah yeah that that's our that's our stuff the only issue with having your own hardware is that you know it ages
15:02
so some of that stuff is real fast nvme's age and all the i just we just
15:09
bought like you know a bunch of ssds for a couple big raids and things like that which are fine but
15:14
then the cpus age and all those all these big enterprise class machines what they build for width they don't build
15:20
for single core so now that all the block times are coming down on all these chains like the
15:26
kajiras of the world and everybody wants a one second block time all this hardware that we have is struggling so that's the real challenge
15:32
is like i can't i can't add any more to it like i it doesn't matter how wide how many cores i give it doesn't make a difference it comes down to some sort of
15:39
single core structure and these single cores are okay but they're not great so when you get into that situation then
15:44
then i'm looking at building more and you know that type of thing so
15:51
when you when you cost in uh that kind of wear and tear like pres what is the if you don't mind me asking
15:57
what is the price differential on running that kind of infra versus i mean
16:03
obviously the obvious thing to ask is bare metal but also you know uh kind of
16:10
i guess in my head like maybe something like digital ocean is quite often lower cost than something like aws but with
16:15
fewer guarantees maybe i think that for um
16:20
i mean overall it all depends on what you pick the hardware up at right so so this this stuff that we have is
16:27
between 2016 to 2019 gen so it's a few years old
16:33
um but they're all like dual cpu most all those machines have between 512 and a terabyte of ram each
16:42
um and they all have between 36 and 72 threads so that's two 36 core machines
16:50
are two sorry two 16 core 32 thread cpus in one box so they're
16:55
very wide um and with a terabyte of ram you can and in fast disks you can run a ton of
17:00
nodes like you can run a lot of um things that even have plenty of threads on it but so but again it
17:06
goes back to single core so there's a lot of that hardware available um and we do a lot of upgrades and that type stuff
17:12
and everything else and like try to keep it somewhat um i mean it's all we've had very little downtime within the stuff
17:19
that we've had um all our issues have been more network structure because we run some pretty complicated security
17:25
vlans and vrfs and things like that in the cisco side um so that's the only thing that we've had
17:31
you know issues with but away from that it's it's pretty cheap to be able to run so i think it's it's
17:37
the only issue with that is that it's also there's ram numbers those ram numbers are no [ __ ] joke like they're no joke to get that kind of
17:44
performance yeah like yeah to get a [ __ ] terabyte of ram yeah
17:50
yeah yeah you can't get that stuff right but again no none of our another stuff that we run uses that right so it's only
17:56
width which means how many of those chains can i run how many separate how many separate um
18:02
data stores or or drives do i have blah blah blah and that mix is that that stuff is a mix of vsphere that we we use
18:08
and also bare metal so depending on what we're using um that it's mixed because the vsphere stuff we like to use for
18:14
specific things that we could easily move across um and i want to have something that's running and move that node somewhere else and things like that and again
18:21
that's there's a lot of nvme back stuff in there so which is plenty fast but again it just depends on what the
18:26
right use of that stuff is so the only issue with that is that you're like sorry oh you're gonna cut you off but
18:32
the only issue with that like i said is like i also have a bunch of ax 101s at hetzner right um and those are month by
18:39
month one set 59.50 is no longer the fastest thing i say goodbye i launched another one within 10 minutes and that's
18:45
their issue right but now i got this asset that's i gotta you know figure out what i'm doing with and i'm paying for
18:52
and those types of things so it's just it's a balance of of of ownership
18:57
yeah we we definitely considered building building our own but like the types of servers we wanted to
19:03
get new we're going to be like 20 grand a piece and when you add them up like to say two lots of
19:10
three in like two different data centers you're like you know 120 grand and then
19:16
the like monthly bills in australia for bandwidth are [ __ ] insane like i
19:23
think well it came down to that we can buy the hardware that we want we can rent the
19:29
hardware that we want for less than what it costs to host it in australia so that's ultimately the way we ended up
19:35
having to i'm go like a fraction it must be like if you you know if you're amazing i don't know
19:41
if you're using hexa or ovh or who but it must be like a a tiny fraction of the cost of
19:47
purchase that's so cheap well um most of our nodes are on um
19:53
obh right um so there's ovh australia which i'm sure is more expensive than [ __ ] ovh anywhere else but
20:00
um so just to host uh like a one you and we would have to
20:06
have it to you but uh to host the one you in um like the cheapest one i could find
20:13
in australia was going to cost like i think it was close to 200 bucks
20:19
just for hosting that in an ip address right um but that's only with like a 300
20:24
megabit connection shared and so um that wasn't really going to work
20:30
because because for 200 bucks i can get similarly spec hardware
20:35
um plus all of the hosting fees in australia for one machine so
20:42
um and that's on in some instances like a 10 gig connection um i have not had
20:47
any bandwidth issues whatsoever the latencies are better because they're over in europe um right so
20:54
yeah it's been good for me if you look at the ovh uh
20:59
pricing for the australian nodes they're more than everywhere else so there's like specific nodes that say apac which
21:05
is asia pacific um next to them they're the only ones in all of the servers that have like
21:11
different pricing for australian ones um well in you know in the ovh lot there's
21:17
no like you know ovh france costs more than apac then you know obh germany it's
21:23
just all of the ovh stuff is the same except for the australian ones which has got a package to it because it's so [ __ ]
21:28
expensive over here although hetzner is more expensive in finland than it is in germany no it's the opposite of that isn't it
21:35
isn't it expensive in germany more expensive yeah i thought no i thought helsinki was
21:40
cheaper yeah i think yeah yeah yeah but this is like why everybody like there's
21:47
the joke about if hetzinger went down everybody be [ __ ] right all these things stop the music stops seriously
21:53
there's no joke come back it's no joke like but it's such a i don't know i mean we've seen some videos like their
21:59
process everything else but they deliver at such a low cost it's it's pretty incredible
22:05
yeah but you do i mean this is the thing though that makes me kind of when if and when we move fully to
22:11
horcrux then i think we'll consider running more stuff on hetzner but
22:18
yeah you know i've i've had like my my bare metal day has ended them
22:23
on hard drive failures and moving to ec2 back in 2012 2013 and i'm not look back and i
22:30
know and i have touchwood never had um
22:35
a major ec2 incident in all the years i've been using it
22:42
that has been amazon's fault had um had a vpc disconnect caused by them screwing
22:48
up um the entire european zone all vpc connections for about 10 minutes and
22:55
that really [ __ ] things up for a little bit but when you've got 13 hours to play
23:00
with you know in cosmetics it's such a different calculation you know right i do think
23:06
that i mean obviously we're all trying to work towards disposable notes so i think our goal is to
23:12
i mean we want to have our own part of our just how we deliver is we want to have our own machinery and our own dc's i
23:18
think that's something we like to have ownership of and and i mean we have we have stuff in head center as well but i don't want to be i don't if if eighty
23:25
percent of the chain stop because cessna goes down i don't want to be at that number i want to be on the other side of that number um and so i think we're all
23:31
moving kind of towards remote signing that can be very portable and secure and the other nodes are disposable and
23:38
whether that's at some of these providers or whether that's on akash or whether that is you know anywhere
23:44
else i think that's kind of what we're that's what we're all working towards right so this idea of like sentries and
23:49
validators and all these like all that's gone like that's all gone so i was trying to get to a point where
23:55
the core is very secure and the nodes are a disposable and that might be three nodes it might be ten it might be
24:01
whatever it is i don't really care um but i you know so i think i think
24:06
we're all kind of closely working towards that same idea yeah i think at the moment my big concern is basically how to get the
24:13
uh we're very slowly moving the direction of horcrux now but how we get that
24:20
smoothly set up in terms of designers and where and in a cost effective way because i think we have the problem that
24:26
especially for the bear market stuff the tmkms stuff has basically resulted us because we were running fourth century
24:32
architecture before we basically thirded the amount of boxes that we've got essentially because we
24:37
were running one uh stacked remote signer with some additional
24:43
dr around it um so we've got about one third of the boxes in prod
24:50
that we did three months ago like when prop 16 happened let's say and
24:56
the concern about horcruxes would be going back to a raft of signers plus
25:02
two to three plug per network probably because the way we run and
25:08
that i think that that feels like a step backwards
25:14
unless we were to do something quite radical like take all of the sentries off onto bare
25:20
metal um and just treat the bare metal so this is what i'm what i'm currently thinking now
25:26
is to go back and basically write a bunch of terraform to
25:33
uh handle scripting up big heads in the boxes like big [ __ ] heads in the
25:38
boxes and then be running sort of three to six to nine
25:43
uh maybe hetzner and a.n other bare metal provider where all the terraform is going to do
25:50
is bring up a set of boxes connect essentially open the correct ports the signer
25:56
and then say okay um i'm going to install all my upgrades and docker and then i'm
26:01
[ __ ] done right and then it's over to you lads to then provision that with the correct um
26:06
[Music] the the correct containers and uh with doing some of the akash stuff that we
26:13
did when we were evaluating running on akash um we obviously used uh tom's omnibus
26:20
quite a bit and i kind of think that that might be a good middle ground is basically pulling in
26:26
containers from omnibus that sync very quickly yeah because they're you know synced from the snapshot the day before
26:33
um and kind of put those together as yeah behind a reverse proxy and then and then
26:40
basically have that as you know running on these absolutely stacked uh bms or servers
26:45
but that may be as a way of keeping of actually potentially even increasing our performance while keeping costs down and
26:52
then just have some absolutely chunky nodes running horcrux because
26:57
that was the confusing thing to me because we had you know shorts he was talking about the performance and stuff in a different
27:02
thread and it's like horcrux is not that big a
27:07
a program it's not that heavy weight so i'm not sure where the performance hit was coming like what the op what the
27:13
problem was that it feels like it should run on a t3 micro or something on aws you know yeah my notes haven't followed
27:19
that sorry no well isn't it the lead has to talk to all of the so i think schultzy was
27:24
running like multiple networks on the same rafts but the lead i think if you get a bunch of networks on the same lead
27:31
node then they have to talk to all the other nodes um and coordinate and i think you end up
27:37
with like a situation where it's quite a lot of traffic to process
27:42
yeah and my stuff like i have um on the set that i have i have i have um
27:48
five separate wraps going and so that that lead does move around and i'm i only have a couple of those where i'm
27:53
trying to push it to uh to a primary but um i haven't i didn't see the same amount of usage that that
28:00
that schultzy did so i'm not sure what that was or maybe that's it's fighting with something or something else but
28:05
we've been configured to maybe they could maybe i don't know yeah so i
28:11
are we gonna talk about that comment are we gonna let that one go
28:22
the issue is that this is a audio podcast so you know you have to read it out to everybody who's listening and not watching what was the comment that came
28:29
up from the uh so it was sort of just saying that's a bit of a man who enjoys handcuffs and
28:34
i'm just going to say that i can't remember what it said in horror the the if you show something it's a lot less
28:41
than letting the imagination play out so i'm just going to say no comment and let people imaginations run wild because
28:47
that is the true terror here so no comment so no king shaving man
28:52
there's 20 22. yeah yeah yeah that's right yeah exactly that's why it's no comment yeah
28:59
i'm just saying who knows who knows what these have seen
29:04
who knows what these pipes have seen
29:09
there's something about the structure of that sentence that did not sit well with me
29:16
though thank you there we go also no where are you you
29:21
look like you're in the break room of like uh
29:29
he's either in a holding cell or he's in like the isolation room of like a high school in like england yeah it's got it
29:35
has the vibe of like an adult education center yeah
29:42
saying it looked like a mental ward or some [ __ ] no i i said it looked like the conjugal visit room of a women's prison
29:48
[Laughter]
29:53
uh frey you know the um series in the uk where it's educating they go in different schools in all the different
29:59
um different parts of the uk on channel four basically there's just documentaries that go into schools in
30:05
the uk and just film for like a year this reminds me of like those schools where they set up clearly because you
30:10
can get quite a lot of content in the isolation booths at school but bad students just go sit down me like nullis
30:16
sat in one of those boobs in like the isolation room at school and just there's gonna be a teacher that uh comes
30:21
through what are you talking about yeah yeah so what the old schools didn't have it so you just you would just go and i
30:28
think sometimes sometimes it's not obvious that there is that big an age gap between us but like you know i may
30:34
as well have gone to school with mary [ __ ] poppins apparently because callum i just have to say mate what you
30:39
just described sounded like a description of a person who should be
30:45
going to jail is that people are going into schools and filming children in isolation rooms
30:52
yeah yeah yeah no that's just the uk schooling system i don't know yeah so to be fair the uk schooling system is
30:58
pretty [ __ ] at the moment they basically sold it all off to private companies called academies so yeah yeah
31:05
you could sell off pretty much most of the school to a private company i i think literally it's something not
31:12
statistic like 98 of the of secondary schools in the uk are now owned essentially by private companies and
31:18
they get to part-set the curriculum and some of those private companies are just like essentially aristocrats who have
31:24
put together a company and they they literally just teach like whatever the [ __ ] they feel like and they're just
31:29
like oh yeah and all of these lords and ladies were just like totally great you know totally great 100 awesome
31:37
let's not mention the slave owning i mean i'm okay there's a little bit of exaggeration yeah it's not they have a rough curriculum
31:44
but they can [ __ ] around with it yeah they don't influence it yeah i've seen some i've got some friends who are teachers and i've seen like what they
31:50
teach and they the teachers don't the teachers are literally like as in they show you it and they're like
31:56
look at this compared to what we were taught and you're like holy [ __ ] this country is [ __ ] like [ __ ]
32:04
like 12 years of tory rule and it literally is like too fair it was labour that did that
32:09
actually so you know just [ __ ] near liberals man just like sell off the schools just sell off the schools for
32:15
profit why the [ __ ] not what what's the worst thing that can happen yeah okay like quite a bad thing actually lads
32:20
because you end up teaching [ __ ] [Music] maintains
32:26
like it was bad enough when i was at school where you just all the only thing you learned you learned about prohibition that was all of your
32:32
american history um you might have done the french revolution because they are just the
32:37
channel and then you did the nazis you did nazis over and over and over and over and over and over again but with weirdly little
32:44
moral judgment just lots of facts and you're kind of like i feel like we should be mainly talking
32:49
about how in every video game i play they're the bad guys and for a reason because they're [ __ ] bad
32:55
rather than just being like and what happened in 1937 in november and you're like
33:01
this is like yeah yeah yeah like the lead up of like post world war one germany to like yeah exactly and all of
33:07
these weird facts where you're just like and which politician said that we need to squeeze germany into like a lemon
33:13
until the pips squeak and you're just like these are not useful facts for understanding how genocide happened
33:19
which is really the top line thing we should be trying to learn from this history but don't but don't google what
33:24
churchill did in india by the way don't google it don't google it oh yeah and china and japan
33:31
may as well not have had any history because [ __ ] it we don't want to teach it but yeah we're not going to teach it african
33:37
history now they just existed when we when when when when we
33:42
went over there you know you might literally yeah you might get one one module on the scramble for africa which
33:48
is absolutely the zulu or the boa war and the only yeah was that was when the nhs stopped
33:54
providing optician optical and dentistry stuff for free that was the level of african history in my gcse history yeah
34:02
i remember there was it just started with the ground nut thing because they ground nuts and palm oil right because
34:08
they needed um oil for for machinery uh to you know to help the the working class they were
34:14
pressing back home who were building their engines of war um they were like oh we need some we need more oil palm oil suddenly we have
34:21
a use for africa let's just occupy the whole thing uh the other abuse they had for africa of course besides you know
34:26
enslaving the population and generally terrorizing them for like 200 years anyway britain british history not great
34:33
there's some relationship between britain and australia right you guys want to talk about that here
34:38
australia has a really interesting history uh which which i think you know we'll be able to i'm glad you brought up africa
34:45
and uh and india but you failed to bring up australia i don't know why you would do that we literally heard nothing about
34:51
australia either we just go like this it's kind of i mentioned like oh we
34:56
should we may have shipped off some comments yeah those guys there were some real bad apples because those guys they were
35:02
quite racist and then they were very nice the aboriginals and and you know real bad apples that's why we shipped
35:07
them off because we weren't racist we were busy civilizing a bunch of places and that was very unracist you know we
35:13
were building bridges which is it's like the opposite of racism really that's like
35:19
how could how could that possibly be racist and you're just like
35:24
awesome awesome history is it time for lunch yeah yeah yeah it's fun
35:32
yeah and then at some point you go to university or whatever and they're just like oh yeah all that history you learned uh it's actually quite bad and
35:37
you're like i did have that feeling that there was another side of the story yeah yeah yeah so yeah it's pretty
35:43
inspiring i finished uh school as in pre-university education four years ago
35:49
so yeah there we go that's when i finished with the uk education system
35:56
um yeah well you know sure we're convicts but he's [ __ ] up
36:03
here we go sure cut us out at the end at the end you're like okay so you've done you you we sent you on a on a boat half
36:10
a year died coming over here we put you in a box for like 20 minutes
36:15
and uh okay now have some land in a country full of awesome resources and sell it to
36:20
everyone else the real punishment is having to pay
36:27
yeah the real punishment is having to pay more for ovh boxes
36:34
long play it's a really long play to be able to have it's like it's like the domino's meme it's the domino's
36:40
thing where it's a sending convicts to australia null struggling to to to rent boxes
36:47
that's the one a real punishment for the convicts would have been to make them [ __ ] stay in england and drink warm beer for the next
36:54
200 years well a lot of there is there's an argument
37:01
that goes a lot of the reason that p so one of the one of the many things that resulted in so many people getting transported was
37:07
a massive crackdown on the rules about common land and especially poaching like poaching
37:13
uh became a very very common reason to get deported i gather which um was largely about essentially trying to deny
37:20
well the uh there is an argument there is a historical argument that goes something like um
37:26
early stage capitalists and also the ruling class who were trying to profit off of the new industry were trying to
37:32
not only break the back of uh labor that had had quite a good time after black death in terms of like
37:38
trying to renegotiate the relationship between landowners and uh sort of uh serfs especially the end of serfdom was
37:44
was the end of the black death right because there were so few people they said well we're not gonna accept the system anymore you've got to pay us
37:50
and there's a kind of an argument that goes okay well people also used to forage for a lot of their living so they didn't have to be reliant on a wage to
37:56
live fully and like by breaking people fully into relying on a wage to live that's like
38:02
the ultimate subjugation into modern capitalism um and the thing that was stopping that was like being able to
38:09
hunt basically um on common land or grow stuff on common land
38:14
foreign wood and stuff so hence poaching law is becoming more extreme and then
38:19
sending people to australia so yeah but that's also why in english folklore there's so many um popular folk
38:26
songs where the heroes are poachers um because there's a kind of historical there's like a folk memory of how
38:32
important that was to everyday life and survival and how sad it was when it went
38:37
i guess or maybe we just romanticized all the things that finished in the past i don't know
38:42
um what does all this have to do with juno and kraken
38:48
well i was gonna say should we should we move on to talk about the kraken listing shall we
38:53
we're gonna come back to that and that's how juno ended up on kraken before we get sidetracked that still
38:59
wasn't really before before um so you surfer i just want to know uh
39:07
so where do you have your side node now say say no because it you got quite good uptime at the moment i'm saying my say
39:13
node is at hetzner because our our current stuff that's there is not uh the stuff in the colo is not fast enough for
39:19
it so same thing with my evos node and my say nodes so
39:25
are you signing locally on those or are you running no i'm signing locally in both those so that that will both i'm i'm
39:32
building new machines for the for the new colo that we're standing up and so those will eventually move back but for
39:38
right now those are remote because i i don't i don't own hardware fast enough to do that yeah so what is says like config isn't
39:44
it a bunch of validators and quite fast blocks is like is say the one that's trying to push like 500 validators that
39:51
i saw i'm not too caught up on it no intentionally keeping it pretty tight
39:56
right they are yeah i think they're they're trying to pull um well originally they were trying to pull 500
40:01
like half second blocks that was kind of like where they started they're i'm not sure what it was set to initially or
40:07
what they're trying to be able to do with that um right now i think i don't know you're probably looking at it but i think it's like maybe a second or point
40:13
eight seconds something like that 0.9 yeah so they're they're trying to
40:18
get that one 1.3 now i think oh no you guys
40:25
[ __ ] it's all over the place yeah i was just going to say isn't that because they're trying to do on-chain order books or something right is that yeah
40:31
yeah yeah yeah so they want they want fast culture on those types of orders and things like that so i think i think that's the goal and and honestly from we
40:38
talked about this i think a few weeks ago but but even from the team like they they're okay with block misses and like
40:44
they like their ideas like even if you're missing a third of blocks that means we should be pushing faster like we want to be able to see where this can
40:50
go um which is very different from how you know we've traditionally approached you know a
40:56
six-second block and if i miss one like you know you're you're you're asking questions in
41:01
the mirror right the tungsten rod gets dispatched from space if you validate the penembra zone
41:07
you know about the rods from god people who like i i love the penumbra zone
41:13
and also rods from gods is amazing yeah people always think i'm crazy when i say rose from god and i have to like
41:18
reference the the whole that that was an orbital weapon system thing yeah no that was written about
41:25
yeah yeah there was they weren't they were really trying to yeah i know that yeah okay i thought good [ __ ] you
41:30
don't want if you don't want this for the new top gun film put your fingers in your ears now right
41:37
there are some right okay it's time for a little moment here the aside from the bit where the navy
41:44
for some reason is testing a mark 10 spy plane which obviously wouldn't happen because duh it would be like the air
41:50
force and there's why would the navy have a spy plane that takes off on the land that's
41:56
and why would they be supplying a test pilot it would be lockheed or whoever the primary contractors would be using their top they would be using their test
42:02
pilot it would be not maverick and it would also be a guy who's reliable not a guy whose call sign is maverick it would
42:09
be somebody who is very calm like a test pilot maybe who's not going to do something dumb and blow up a [ __ ]
42:15
plane and whatever anyway that's not the important bit the important bit is the entire plot has at least one major
42:21
glaring hole one of the major glaring holes is that they fire tomahawk missiles at an airfield
42:26
which shows they know where this airfield is and they're like oh the reason we can't just use these is because we can't just like fly a plane
42:32
over and bomb this thing it's because there's all these anti-aircraft missiles and it's like well you know where the airfield is and you just blew that up
42:38
with tomahawk cruise missiles why don't you blow up the sam sites but more than all of that they
42:44
have a mark 10 spy plane and they don't have rods from god and you're just like literally this entire
42:50
mission is predicated on the idea that you cannot just drop a tungsten rod into this like
42:56
hard to find place and blow it up and and like clearly you would just use
43:01
rods from god i was literally watching the whole top gun thing being like princess frost for god right there i
43:08
don't like joy how do we possibly blow this thing up and i'm like brother oh god
43:15
you know where my suspension of disbelief ends the rods from the god that is that's the movie
43:20
after that that's it no no fun zero out ten i put my beer down i got up and left
43:26
i was like i don't know how the movie ended yeah all right so again back to our
43:35
it's like hurting [ __ ] cats man back to the thing we were talking about um so
43:41
i have a big burning question with these fast blocks right so what happens to all the history because
43:48
this is going to add up super quick and interesting fun fact um notion notional
43:53
helped uh psy with their uh code i think to make those blocks happen super fast but what's gonna
44:00
happen with the history here like no one's gonna keep history from a bloody
44:06
one second block time after six months it's just going to be astronomical and if you imagine this going for years
44:13
you there's not nvm is big enough on the planet to like hold it so
44:19
what do we do do we just sell taxes like why don't we just why aren't we
44:25
programming these blockchains to just cost the history
44:30
well you you would also have to assume that there's gonna be a lot more uh zero transaction blocks right so you can drop
44:36
those out of there and so i would assume that if it's if it's once if it's a six second block or if i have six one second
44:43
blocks on average if it's you know one or two transactions per every six seconds that
44:48
most of those blocks are gonna be empty but i think the point i mean it's so different than what we talked about before right well it is different but it
44:54
might be along the same lines what we talked about before which is at some point you're not gonna be able to do it they're gonna like at some point you
45:01
would have to cut a new genesis right and you have like what todd has with
45:06
junoclassic.blockpain.com which is some sort of archive note and every four or five years there is these different history
45:13
type of nodes that have that before some sort of regen and then it runs up again i guess i
45:19
don't know so there's there's two things they need to like start working on
45:25
and one is that you need to be able to aggregate archives um
45:30
so if you like just re-cut it on the same um network name or even if the
45:36
network name doesn't matter you need to be able to aggregate archives as long as the accounts match up um because you're going to end up with
45:42
this situation people are going to be trying and you know people's fys finish on different dates too so you can't even
45:48
have um you know just cut it every year at the end of the fy and start a new
45:53
network because you're going to still have crossover between
45:58
you know australia's finishes on june and america finishes whenever and i don't know when
46:05
um england finishes but then april so the point is you're gonna have all
46:11
this history right but it's gonna be huge and no one is maintaining history anymore anyway people are state sinking
46:18
people are using pruned um uh snapshots so
46:24
there would only for each network there would only be a handful of people keeping history so i'm pretty sure how many how many
46:31
cosmos networks could you do you think do you think for certain at this point
46:37
all of the live cosmos networks actually have complete history i don't i don't
46:43
think so i i i would think so i would think the foundations would run like i run two for stargaze that we have two
46:49
full notes that we keep history on but like target is a pretty easy network as we haven't gone through that structure
46:54
um you would think that those groups would want that history right
46:59
so even the archive nodes in the history nodes i've tried to query blocks
47:05
back in the history and i don't get responses out of them
47:10
like um when i when you try to query the vowel set specifically from history like
47:16
uh i know a block a long time ago yeah you can never get data out of them it's that only keeps like the last 5 000
47:22
blocks or something um i think no isn't that also how that's set up because they're like a query time and
47:28
there's some other things like like how long that query could run on that note because there's some rpc settings in there i thought you'd like jack all that
47:34
weight this is a good talk question but i thought you'd jack a lot of those networks those things up because some of those queries take
47:40
some um decent amount of uh kind of transaction hopping to figure
47:45
out what exactly the state is to be able to return it back to you so that that query time might get timed out or something similar to that
47:51
yeah well i guarantee like none of the validators are going to be keeping right history for these fast blocks because
47:58
like your performance degrades as you keep more history oh yeah for sure so
48:04
yeah i don't know what the [ __ ] they're gonna do about that um but if they're you know i think if
48:09
that's the way they're gonna go and they don't care so much about history they should just toss it
48:16
um so maybe have like a trailing you know you you have to set a flag to
48:22
keep history basically so if you're like rpcs you can set a flag to keep it if you're a validator it just
48:28
you only need to know the hash of the last block and the transactions from this block and then you move on anyway so or maybe
48:36
like you know the last hundred blocks and then just have an open door out the back and off it goes
48:41
um so people don't have to do all this stuff with um pruning and and um state syncing all the time to try and keep up
48:48
performance well yeah i mean especially if people start moving to things like akash anyway where they you have to use omnibus like
48:55
to do that in a sense well like any kind of reasonable way you want to use omnibus right and so you're going to get
49:01
history from the last day plus the snapshot yeah and that that is increasingly just
49:06
the way what you see everybody doing like if you look to everybody's snapshots half of them we poked you
49:13
because obviously polka is like the hardest working snapshot provider in the cosmos
49:18
um but yeah yeah i do i don't know um it's an
49:23
interesting question like i i mean actually i guess it's that thing of like um
49:31
there you go so todd just said if it's only got the valve set for the last five thousand blocks it's
49:36
probably not a history node even if it has all the blocks history nodes have state for every block okay so just on
49:43
that fun fact um todd i i was querying off your history notes so
49:51
um it's uh i don't know it must be maybe the particular query i was using
49:58
but i was i was trying to query the validator set from history and it
50:03
no no juno history note i could find would give it to me um i didn't try other networks but
50:10
uh anyway sorry continue uh the fray well i guess i guess it's just like
50:15
that's this thing of like there are certain cases like taxes where you very much do need the whole history
50:21
um but like other than that they're it's not really the end of the i mean
50:26
even in a hypothetical world where no history no archive nodes existed
50:32
it's actually maybe not the end of the world for a lot of use cases are the the integrity at state now is
50:39
the most important thing right by a massive massive and especially in proof of state proof
50:45
of stake networks where it's just the validator set right now signing the block that matters and then
50:52
the finality on that proof of work you know obviously different very different but proof of
50:58
stake i think like the history is fundamentally less important is my feeling
51:04
so it's it's less important as far as determining whether the the block and the transactions are valid right that's
51:11
that the current state is valid right from that point of view but you know obviously some people will
51:18
need history and if they don't have it the governments will kick up the stink about this particularly yeah sure i guess i'm
51:25
also talking like philosophically i think we have this attachment to the idea that proof of stake should have that completely obviously we
51:31
need we need it for for tax reasons but let's ignore that for a second but that's one side and go like you know we're we're also attached to this i
51:37
think for conceptual reasons because of familiarity with what came before and actually like proof of stake is kind of
51:43
its own thing um maybe it's like it's like very different to like a stream where you know if you
51:50
have like you know a large streaming data store where is every single event or or event
51:57
sourcing whereas every single delta that's ever existed in the in the data store um and you can reconstruct the state i
52:04
mean actually event event sourcing is a really interesting model to think about like where
52:10
everything that's stored is just a delta from previous state to next state so you can always reconstruct any point any
52:16
point in the history and that's excuse me that's maybe closer to the the utxo model
52:23
right um [Music] in fact isn't that literally what the utxo model is
52:30
just a lot of deltas completely lost what you're talking about
52:35
uh i was talking about event sourcing no absolutely database strategy for streaming databases
52:41
right yeah it's a thing i don't know about that again what does this have to do with juno kraken yeah yeah i'm struggling to make the
52:48
link here junior on kraken is going to make us all rich um that's not financial advice uh
52:54
uh but uh yeah it's it's do you think now that adam is now irrelevant
53:04
yes i would say i would say there we go alex said it for me but i would have got a
53:10
step step a step further and uh yeah i mean the reason i say dogs too whoops
53:17
the reason that we were on track again we're always getting doxxed man it was oh god always getting dogs by
53:24
collaborators and friends anyway it's like do you know that drill tweet where it's like uh i think the original tweet
53:29
is actually like turning the dial turning the dial on and then in inverted commas racism
53:35
and looking around at the crowd for approval then turning it more
53:40
i feel like it's like turning the dial on hot take and then looking around at the crowd for approval before doing more
53:48
but but calm didn't really raise the ante he just kind of let me publicly shoot myself in the head
53:54
was it ever relevant oh yeah well i mean it was the it was
54:00
the default uh on off ramp right because it's easily transferable
54:06
and now what's it ever relevant oh my god i still think it is i still think it
54:12
is purely because this atom has of course got the maturity and the deep pools of liquidity on sexes
54:18
so that is the only thing to be done it has time
54:24
yeah yeah exactly it has it's been on all these exchanges for years now so
54:30
of course juno is gonna have to wait a good few years and juno is only on one exchange and i don't know what kraken's
54:35
availability is i just saw that it's not on it's not live in japan that was in kraken's tweet unlucky ccn
54:43
oh he's saying all the [ __ ] today man
54:50
you know use your real name on the internet you want to be careful uh but yeah no um yeah obviously actually i
54:57
think that it is a fair point that the liquidity on on asthma is obviously huge and also their inflation model is very
55:04
different and that has i think a pretty big implication for its uses not on off ramp judo is
55:11
it does have a problem in the the low supply i think does have an impact on us on his usefulness for an on off ramp as
55:17
a general purpose thing as opposed to just for
55:23
junior notes drama grammars
55:28
isn't it get just derivatives though or am i getting it confused with like by bit or something there's so many [ __ ]
55:34
exchanges i don't know i was going to say rama says that juno is on bit get i would say
55:40
big is irrelevant because no one knows what the [ __ ] it is i i'm hearing about it just now yeah
55:46
i've never heard of it bit get out that's all i'm saying thanks rama
55:52
i i think it's it could be on other like djinn exchanges too but um
55:57
you know yeah well isn't it unlike dj exchanges did you decentralized exchanges yeah that's what i meant
56:04
that's what he meant just ggl exchanges okay good sensational
56:11
um so yeah anyway so now that uh well put it this way
56:17
now that juno is on kraken uh will you be
56:23
utilizing adam for anything in the next month well well so this is the thing so
56:30
there's been a lot of talk recently about atom 2.0 and no one knows what the [ __ ] is going on with that so i guess
56:36
maybe that will come back and bite us all in the ass about it but i would argue that we're actually up to more
56:41
like adam 5.0 or adam's yeah no no no what is it is it cosmos hub dash 4 we're
56:47
currently on so yeah but also the use case changes monthly on
56:53
uh on adam so far oh that's that's what that's what the twitter accounts will tell me
56:59
oh my god did you see jaby's talking spicy i know the neutron one the neutron one right
57:05
yeah right and friend oh the show um charles jabster iii
57:11
known by his handle jabby um in fact everybody knows his name's joe so i don't know why i called him
57:17
charles but whatever he kind of seems like he could be a charles i always thought that about joe
57:22
anyway um he i think he basically like quote tweeted
57:28
you'll have to correct me on the the correct zuma language for this uh callum despite uh twitter originally being a
57:34
millennial thing i gather it is now zoomer thing so we do about a quote tweet says that a sub tweet no it's a
57:40
ratio it's a ratio no i don't even ratio no so do you know
57:46
no sub tweeting is say if i put a tweet and saying say you'd annoyed me
57:52
yeah i i reference it without actually mentioning it directly i don't i don't i just say like oh
57:58
a founder has just front-runned me for 50 percent of my allocation where everyone knows it's you but i haven't
58:04
just added you but yeah i don't even hit a ratio but he quote tweeted setting from i can't remember the name of the
58:10
guy it was some terror account formally unfortunate isn't it but um
58:16
but yeah um then uh but yeah so he quoted her and he wrote a
58:21
good thread on neutron to be fair and i'm nothing against neutron he just put what makes neutron unique and then he
58:27
sort of put a thing on juno which was completely everyone took it a bit too hot everyone
58:33
thought his bashing juno no he wasn't he was just sort of saying like oh this is why this exists and this is
58:39
why um this is this is why yeah juno like you know they can still both exist
58:45
so and then jaby retweeted it and just put so and one of the reasons he listed why
58:51
neutrons better is just that atom is readily available on these exchanges and um
58:57
yeah so they then jaby retweeted and said um how's that looking now and then jaby
59:04
then also replied quite quickly after saying i still think it's a good point and i imagine that's due to the
59:09
availability of atom on exchanges yeah but it's still also funny to see
59:15
when jabba gets spicy because no no it was it was good i i like that tweet that is the sort of
59:22
tweet i would write so speaking of tweets that you might
59:28
want to write um hey uh actually you know what you know it
59:33
was it was a second so i actually also want to cover one more thing that is actually technical and is within the purview if you like of
59:41
the actual purpose of the show which is that we were talking a few weeks ago about we've talked actually several times about a cache and running on
59:47
running on a cache and running on akash which obviously both completely different services um but also the the limitations of the
59:54
sdl in terms of validator ip white listing which is a problem for all of us that
59:59
run remote signers and i've been talking to the akash team about this
1:00:06
they don't plan on fixing that within the short term on their roadmap but sdl
1:00:12
changes for that kind of stuff are potentially in the medium term but this is again where we kind of get
1:00:17
back to our good friend tom um that they're going to basically roll out um changes in omnibus that uh cover
1:00:26
simple setups for reverse proxying so that you can configure your own um white lists using a reverse proxy layer
1:00:34
um while they because they i guess they just don't think enough people
1:00:39
have the sdl problem outside of validators to me it seems like actually there's
1:00:45
probably a lot of use cases where you do want uh ingress white listing
1:00:50
but they are i think reasonable to say well you can just obviously run a reverse proxy in a container so
1:00:57
um but they're going to make it easy for you to do that so that's quite cool in
1:01:02
terms of those of us that are interested in running validators and infrastructure on a cache in that way
1:01:07
and akash for that matter um i really need to decide to consistently pronounce
1:01:13
it correctly at some point um so that's that's something i found out really in the last couple of days which i'm
1:01:18
excited about in terms of actually giving it a proper try um and maybe that also be the thing that pushes us to
1:01:24
more seriously start evaluating a horcrux setup because we kind of got the boxes lying around at the moment but
1:01:30
we're been doing fine with tmkms so i haven't had like a a reason to break off from all the
1:01:36
various things we're doing and be like lads this is the priority this week so did you in the end determine that
1:01:44
using a cash is uh economically viable
1:01:50
uh yes but it's not loads cheaper than aws in
1:01:56
its current form um it is a bit cheap loads cheaper than aws
1:02:03
that's not really the bar though is it yeah if that's the bar yeah well you're talking like the the
1:02:09
uh to get a validator running with a reasonable hard disk um from provided that was definitely available
1:02:15
um you can see in the video i did but the bids were about seventy dollars per
1:02:20
month that was that was what i got to um i mean
1:02:25
that's for what that's for how many cores and ram and everything else
1:02:30
um enough to run akash i think it was like four four cause
1:02:36
32 right i don't know it wasn't pretty it's not going to hack it for juno that right right um it was enough to run
1:02:43
akash but then akash is the bit that's the blockchain is actually quite simple compared to
1:02:49
sorry well quite low footprint let's say compared to a cosmosome chain so
1:02:54
that is that if you were going to run juno then i at that point i don't know
1:03:00
um but this is kind of back to you i think we talked about this before this is back to the difficulty of if your product is a marketplace
1:03:07
uh fixing the supply side of that is like a multi-year road map just to bring on enough
1:03:13
providers right yeah there's a lot of uh there's a lot
1:03:18
of spare um supply at the moment in a cache
1:03:24
but the problem with um akash is is that they
1:03:29
i haven't been able to find decent reliability anywhere through a cache so i've tried a
1:03:35
bunch of different servers just for websites and they keep going down it's been quite annoying we ended up
1:03:41
having to move the juno website back to versel because oh right yeah i think i saw you
1:03:47
mentioning this in one of the chats like so is the juno website now back on versailles
1:03:52
yeah because oh wow i didn't realize that well we lost the provider three times in
1:03:58
a month so i don't know if it was like in a lot of cases we could still
1:04:03
access it through the provider's reverse proxy but um and it wasn't a cloudflare
1:04:09
problem um it was something sometimes they would just go dark and you couldn't access them through the web
1:04:15
at all other times like the the reverse proxy would stuff up on them
1:04:21
so i don't know um but we had enough issues with it that we decided to move it off
1:04:26
there because we can't just keep having the main website go down so um
1:04:32
but i did find out after i moved it to versailles that it is quite
1:04:39
quite data-heavy you mean just the page load and everything else well you get there's a lot of graphics and movement those
1:04:45
videos those videos are pretty big
1:04:51
actually so it's been back on um so i only put it uh i moved it to a new team yesterday on
1:04:58
versailles is this still this is still going through cloudflare
1:05:03
oh yeah still that's it yeah so cloudflare's caching some stuff here like the edge too for selling
1:05:08
we run our stuff and we're selling a couple different sites i've been always happy with that um the sale is good
1:05:17
for 20 bucks a month yeah unless you blow through your data
1:05:23
then not so much but ah man it's the the web interface is so confusing on uh
1:05:29
first cell i can never find out how to get into the bloody the correct um
1:05:37
settings like there's they got the settings across the top and then if you click the settings it's like the team settings
1:05:43
oh no now i'm in the but now i can't find the analytics like they go the individual site then go to
1:05:48
analytics yeah it's a little bit goofy like yeah but so it doesn't happen sorry you have to go back to the team
1:05:56
to get the usage right okay right oh yeah so we put it up yesterday
1:06:02
and in the last 24 hours it's used 7.5 gig of bandwidth
1:06:09
which yeah it's pretty good wow that's quite a bit yeah yeah so like is that
1:06:16
cat is that cached i mean it's at the edge but yeah that's proxy through uh so for
1:06:23
comparison i mean obviously my my site's got a lot less uh requests but like my websites used 15
1:06:30
meg i think it's these are those videos back there what are those things back in these backgrounds
1:06:36
yeah yeah so um you know for example the
1:06:43
you know the main video there has used 1.8 gig yeah that's a bit much
1:06:49
and it's like it's 46 megabytes i think really that should be maybe
1:06:58
anytime it takes 46 minutes to load a website that might be something you want to think about well i mean
1:07:05
wait what does that wait wait wait wait you broke you broke the fray
1:07:11
wait what is what
1:07:18
like hang up what did
1:07:24
are we serving i'm sorry like wait where are we
1:07:29
completely welcome somebody said the judo website is serving 46
1:07:36
megabytes what's a webinar you're a mobile browser hang on
1:07:42
i don't know let me let me let me recheck that i see both a mp4 and a webm on that thing let me see if i can pull
1:07:48
it by them size how are you saying this we're just off just not the source
1:07:53
all right i'm now going to load it on my phone to be fair actually i've just loaded the website here the fan is now
1:08:00
on in my laptop the macbook is taking off this is this is there's not there's
1:08:05
something not quite right about this because this is this is a powerful this laptop is basically a web server
1:08:13
let me let me write let me throw the lighthouse on it see what's up there might be just some just some work to
1:08:18
everyone's laptops just starts taking off i guess there's an intel laptop there i can't even find the videos oh
1:08:25
[ __ ] right uh just oh yeah it's gotta be the videos jesus
1:08:32
what does this have to do with juno and kraken so kraken would have gone to juno's
1:08:40
website to do their due diligence and that's why it's been 46 gigs because or
1:08:46
whatever it was because they've been doing a lot of their due due diligence and going back to the juno network io site
1:08:53
it looks like it it doesn't play the videos on mobile we're okay we haven't like killed everybody's data that really
1:08:59
freaked me out for a second there funny so get this right so the ecosystem mp4
1:09:05
is 20 meg the footer mp4 is 12 meg
1:09:10
the hero mp4 is 10 meg the hero 2 webm is 4 meg so when i when
1:09:18
i load it in lighthouse i get it funny enough it doesn't do that bad but it's not awful it's a 82 performance
1:09:24
which isn't bad but the because i think it's just it's prioritizing structure it's prioritizing
1:09:29
paint right but the the ecosystem webm is 12.9 megs and the
1:09:35
footer is 5.7 in webm which is pretty significant so the overall page
1:09:40
load is 22.8 megs that could be probably reduced a little
1:09:46
bit i would think find some ways to be able to do that you could probably sacrifice some detail
1:09:53
in those but but i mean i mean but but i think next i'm sure this is next.js behind the scenes but
1:09:59
next and also vercell's doing a good job of actually loading items in order and you know not not making it a
1:10:05
an issue from a from a viewer perspective it's just a lot of data behind the scenes yeah so i'm beginning to wonder if like
1:10:12
a cache uh sort of you know just turfed the websites
1:10:17
because of how much data they were using or something maybe well there's no sla there and you have there's no contractor right so any provider can just kill it
1:10:24
at any time which is really on you as a as someone as a content provider that
1:10:30
has to monitor the fact if my provider is still serving this data is that right
1:10:37
um there's no sla right so like whoever serving that like it's too much data they're just gonna shut that thing off
1:10:43
right they're gonna kill that and you have you like you just you don't really know about that until the site's down
1:10:48
yeah people message you and go sights down again you're like [ __ ] permissions cuts both ways i guess yeah so it's a
1:10:55
permission to run yeah what's lighthouse is that something i
1:11:00
don't know about that is a google delivered kind of like uh page loading and paint and
1:11:09
um metrics structure by google that determines and gives some score around
1:11:15
load time and best practices etc how does a king node score in there
1:11:21
i can tell you in a second because it's only a single page right so i'm guessing it's gonna be pretty fast so
1:11:27
while they do that i think that that concludes the conclusion there is that
1:11:34
when when omnibus lands its changes can do more experimentation with akash
1:11:39
i'm personally looking forward to doing that but obviously there's some stuff i think with which is going to need to
1:11:44
happen i mean obviously this is the thing like i i think with overclock i said this
1:11:50
actually i was talking to uh don kryptonium earlier i have absolutely
1:11:56
no problem believing that overclock have the technical chops to pull off
1:12:01
everything they need to to make all of the tech work for akash i think the hard thing for them is going
1:12:06
to be actually finding that supply and working out how things like like you said was saying
1:12:12
no sla cool can't run any critical infrastructure on it
1:12:17
you know how do you work out that stuff in a marketplace where
1:12:22
you aren't going to be able to get an sla but you can enforce economic incentives
1:12:28
what is the mechanism design there to make it actually viable for you as an infrastructure provider who needs
1:12:35
high availability to have faith in that system whilst it's still being cheaper than aws
1:12:40
which is honestly not that expensive in the grand scheme of things
1:12:46
yeah i i mean my point is it's early it'll grow those like that
1:12:52
structure will occur over time um i also don't know kind of early on if
1:12:57
it's a requirement that it needs to be cheaper like there's a there's there's some trade-offs there in terms of
1:13:03
decentralization and not paying for one provider like it's kind of like early days of electric cars like it was always
1:13:09
like a measurement of but is it cheaper than gas well maybe it's not like maybe that is that the yardstick that we're
1:13:14
using here like at least initially in overtime et cetera et cetera so um i'm not so concerned if it's exactly
1:13:21
cheaper it was more expensive than i there's a lot of talk about how it's how cheap it is and i think your
1:13:27
your data showed that it's not you know costs are costs and so it wasn't it
1:13:32
wasn't um a hundred percent cheaper or something similar to that like it was cheaper than aws but it still
1:13:38
still um was pretty close and so i think i don't know if if price is the only benefit to this
1:13:44
right so i think you might be paying for some other areas as well yeah it's trade-offs on there um
1:13:51
so we've got we've got callum on here um
1:13:58
so
1:14:09
so to be fair callum does currently run 100 of his nodes on akash
1:14:14
yeah yeah a hundred percent of my nodes are on a cache and they have and a hundred percent of up time on all the chains i run nodes on
1:14:23
so uh so we wanted to to to pick your brains uh calum on how you've got
1:14:29
across every chain that callum validates he's got 100 up to up time and
1:14:34
he's never missed a block so i have never in my lifetime missed a block so exactly so we were we we had we
1:14:42
we had we saw his tweet on twitter and we were like we've got to get him on to explain how he's done that so calum do
1:14:47
you want to explain your your magical strategy don't run notes
1:14:54
that's good yeah yeah but i've never i've never missed a block locally because it's all
1:15:01
you know very in a contained environment so i've run a june i know before
1:15:06
one of one yeah my my the perfect network one validator
1:15:14
did you just like a bunch of juno to yourself just to like make you number one on the list
1:15:20
of one of one
1:15:36
you get a jaunty little thing where it it provisions you the thing and then it spins up the node in docker and then you
1:15:42
get the logs where it's like this is a validator and then one block two blocks yeah it's
1:15:47
literally that easy guys i don't know what you guys are complaining about i run it
1:15:54
which was entirely too long when you can just like connect your thing to the juno test net i'm confused
1:16:01
yeah there you go yeah it works both ways you see banter works for both ways guys [Laughter]
1:16:09
yeah i mean but you also weren't awake to troll when we were actually doing the the podcast so um
1:16:17
no dice that would've been funnier in the live chat by the way what's that oh in your uh man it's 3 a.m
1:16:24
i'm not getting up at 3 a.m after that what what is that 3am for you now yeah yeah yeah shouldn't you be doing chain
1:16:30
upgrades yeah i simply didn't have any spare time at 3am
1:16:37
oh did say have another update yeah okay that excuse works all the time doesn't it yeah yeah exactly
1:16:44
um no seriously what time is that though uh utc uh
1:16:49
we do it like free it's even like three or four utc or two like it changes i think we very variated varied a little
1:16:56
bit it's like three two four utc one of those times it's pretty early for me and it's
1:17:03
and it's like it's like every other every other week every three weeks pretty pretty inconsistent is what i'm
1:17:09
hearing you've got to work on your holidays right you guys already yeah yeah that's the only way i've missed blocks you could say
1:17:17
yeah just all those holidays it's not like keynotes which is every
1:17:22
week at 2100 utc on wednesdays rain hail or shine even if we have
1:17:28
absolutely [ __ ] nothing to talk about i think i think you are actually the
1:17:33
only person that has maybe made almost actually i don't think you have made every episode of g of gaming notes have
1:17:39
you i don't think he's been lagging he's been most of the time so you know see
1:17:44
i think i've missed one maybe really that's it yeah i've missed three it's a couple yeah
1:17:50
actually i've i've missed 20.
1:17:55
yes you've heard calallen the most inconsistent yeah i'm actually one of the founders by the way guys i've just
1:18:00
not shown up for the 20 shadows shadow banned him
1:18:06
the nothing had been every single one i'm the only one who's been every single one i'm i'm putting that out there yeah
1:18:12
i think you've i think you've been in every single one but i think you've like dropped like a few times but you've been in every single one whereas i think jaby
1:18:19
kind of jabby kind of dipped after the first whenever it was after the first two
1:18:25
yeah after the early drama jaby was like this is actually too much i want this
1:18:30
i kind of i'm kind of done with this at least to the conjugal cell to get the good internet though so government
1:18:37
yeah yeah you're in isolation that's that's that's the trade-off for good internet your mohican has made some sort
1:18:43
of appearance on at least 21 episodes yeah wait what one did you surfer miss
1:18:49
i've missed a few oh i have has
1:18:56
it's kids or something else it always comes up that's it yeah so we since there were 79 minutes in can we talk a
1:19:02
little bit about howell yeah like you're both here and like you know only got 10 minutes left like
1:19:09
should we people who have who have listened this far who can't be can't be very large number we could we could
1:19:15
strengthen that is there anything you want to gift the people who have gone through 79 minutes of roundabout slash
1:19:22
kraken slash whatever else type of conversation we've had
1:19:27
i'm going to give you the keys to the lamborghini [Music]
1:19:38
[Applause] [Laughter] um yeah i mean this isn't the first this
1:19:44
isn't the first application i guess that i've launched on juno so i kind of knew that it would be even on test net would
1:19:50
be a bit nuts like how's your how's your day big calum uh well you decided to dip for an
1:19:56
interview straight after launching it you're like yeah i'm going to tweet about it and then go for it when it's after
1:20:02
so so this is this is the sort of he dropped the tweet five minutes later he was in a call
1:20:09
so yeah but no it was actually fine like um it wasn't bad uh just i was actually kind of blown away by the
1:20:15
amount of people that are willing to play around with something like in a way that everyone everyone's
1:20:20
been quite nice about it we've had a few smaller issues some good feedback already to be fair
1:20:26
and just yeah just got chaotic really quickly like because we've probably had like quite a few you know peaks and interests
1:20:33
and dips than peaks because of just how our communication is as a project so yeah everyone seems to then everyone
1:20:40
like we're getting a bunch of people loads of questions the faq has already built uploads because we're getting
1:20:45
asked a lot the same thing but it is what it is and yeah it's quite exciting
1:20:51
it's weird to see people play around with something that you've been exposed to for like two months and just kind of just been there and you're like i'm used
1:20:58
to this and then they're like oh how do you do this this doesn't make sense you're like no it's just this but it doesn't make sense to them so yeah
1:21:05
can i actually use this is this usable uh yeah yeah maybe not it's subjective
1:21:11
isn't it but yeah it's that's kind of a big question
1:21:17
i missed the the announcement because i was probably sleeping yeah uh
1:21:22
and oh you're on don kryptonian as well with your uh man i've had a long a long day it's a
1:21:29
long day what made you look good in the uh in the yeah he did you and i he did do a nice
1:21:35
one with the uh image he used for you i i he asked for an image and i don't even
1:21:40
have it on this i'd have a laptop but i i sent him one um that i made up
1:21:47
uh it's it's basically just me but with a dance cap on which was my response to
1:21:54
somebody it was i can't remember who it was like there's some
1:22:00
it was some [ __ ] [ __ ] project or something that basically was it for the terror developer fund um
1:22:07
i love i love dan's uh hero and it just always cracked me up like i love him
1:22:13
yeah he basically i think clipped that straight out of game of modes uh i don't know i i sent him a picture
1:22:20
he was like that picture's rubbish i'm just gonna clip one out of game of those basically
1:22:27
um but yeah there was like some you know when um we
1:22:32
the test net halted sorry the mainnet halted on upgrade and we had to just uh patch it because of
1:22:38
the order of upgrades there was basically i can't remember where it was that somebody came out the woodwork from an
1:22:44
uh a project or a chain or something that was given a [ __ ] about it was like well anybody could have tested for this
1:22:50
we found it
1:22:57
i was not going to engage with it and then i think in the end i basically like made i made a um
1:23:03
yeah i made photoshop myself in a dance hat and i was like you [ __ ] happy or something like that and then uh
1:23:09
be like you're a good one you [ __ ] knew you couldn't know and and then i deleted the tweaks i was like it's actually quite childish but i have the
1:23:16
the image of being a duds hat lying around on my hard drive so i said it to don and he was just like
1:23:21
be serious [Music] which is where you know that's when you know like because don is
1:23:27
not a serious guy so when he's going like come on man take this a little bit seriously sure
1:23:33
i i jumped on the i jumped on there with don the other day and literally talked about hong kong for
1:23:39
like half an hour while i was eating a steak i was like
1:23:44
we talked about bread we've talked about british history for about half an hour this one so we can't talk like us he he
1:23:50
sometimes takes the first 50 minutes to actually get to the subject but i i don't think that's a stone we can throw
1:23:56
from our glass house we're doing it again we're literally doing it again right now yeah he's asked
1:24:02
me a few times i'm like i don't know if i have anything really to talk about and then you guys have paid you guys have you you know you've paid that goat path
1:24:08
of having no topics to have a good conversation so yeah i feel pretty sometimes you don't need to have an agenda to have
1:24:15
a good time i know that but his podcast seems like a real one
1:24:23
seems different than this one i don't know it's slightly different in some sort of way oh yeah i mean that's
1:24:30
yeah i think you said but you do have to go on it now otherwise you know i know like really no friends so don't reach up
1:24:36
let's go uh so all right so back to howell what give us a give a give it
1:24:42
for folks who don't understand what it is or not been playing with it all day today we're going to listen to this give a give a 120 second overview of what you
1:24:49
guys are trying to build or what you're building i shouldn't say
1:24:54
try to that's not that's not that's a that's a bad phrase what you are building and what you have i should say
1:25:00
that right so now that howl is live it is a microblog platform of 140 characters
1:25:07
that you can write and using the how token you can stake to liking is staking on a
1:25:13
on a post and basically the you the delegator gets 60 of the inflation the author of the
1:25:20
post you delegated to gets 20 of the inflation 10 percent of the weight of which goes to the community dao and 10
1:25:26
which goes to the dev fund which is sort of very similar to the uh commun the community tax on a cosmos sdk chain
1:25:35
and yeah so basically it's just this sort of market model which the frey explained
1:25:42
really well in don's uh podcast about how it creates this market of the
1:25:48
authors and the uh delegators of you know buying and selling how
1:25:53
to support each other and building this market very naturally but basically you can support smaller creators you like it also has a patreon
1:26:00
model built in it's just yeah it's awesome come check it out come play with it come break it
1:26:06
you might get you might get rewarded and also i guess it's worth mentioning there's a couple of things we can't
1:26:12
build until main net um because so it doesn't work on mobile yet because
1:26:17
you kept on mobile doesn't work for what we need um for test net applications so
1:26:23
that will come we have like a version 1.1 uh which is like the things that we can
1:26:29
do after mainnet and there's like three i think three major things on that
1:26:34
one is mobile support two is a bridge to twitter so
1:26:39
essentially just like posting from how to twitter to duplicate it so you you can just
1:26:44
stay in one rather than posting to the other so it's just gonna kind of re repost it to twitter
1:26:50
and then the third one and this is the one i'm probably most excited about actually is the
1:26:56
long form posts so um the kind of actual like long form uh
1:27:01
blog post because we we actually started building this kind of more back to front like the long form was what we were doing first
1:27:06
but then you needed replies and comments and stuff and uh especially because obviously callum's
1:27:12
involved in dow dow um like replying to at the moment you currently comment on dow doubt on
1:27:18
proposals and things like that and it was kind of that sort of use case like the replies to a long-form blog
1:27:24
post and the replies to like dao proposals that kind of kicked us in the direction of doing that
1:27:29
bit first like the shorter posts um and also the fact that you need ipfs to
1:27:35
do longer stuff so we almost need to launch essentially see if people use it see if
1:27:41
there's any money in the bank and then we can afford to buy some [ __ ] off heads in the boxes and
1:27:47
run ipfs nodes and reverse proxy and build out for some infrastructure around
1:27:52
um [Music] the stuff we need to to run the the long-form blogging stuff which we've
1:27:57
kind of won't take us long to it's kind of already partly implemented in various bits and pieces so
1:28:03
yeah on the chain on chain it's there but of course we don't actually run the infrastructure to
1:28:09
support it so yeah yeah we need to build the ui for it basically so but i i i'm pretty excited about the
1:28:16
blogging thing if i'm honest that's kind of cool um can you just leave firefight for
1:28:23
the viewers uh how is its own blockchain or how is built on juno so yeah it's uh it's its
1:28:30
own no it's about to say the completely wrong thing it's um it's it's it's a con
1:28:35
it's a set of contracts on juno so it's written in cosmosome it's deployed on juno that means you have to use juno for
1:28:42
the gas fees right and that's probably the do you know testnet yeah it's currently on uni free the juno
1:28:48
test net and it's currently broken uni3 or we've broken juno tools so i've
1:29:00
so i've tweeted a couple we have dosed multiple things today completely unintentionally so firstly we
1:29:07
dost the juno force it for testnet and i've been posting some good memes about that today and then secondly we
1:29:14
dost juno tools test net because of the amount of requests that people were making
1:29:21
so i did the bart simpson chalkboard meme but he's just writing i will not dos do you know juno tools
1:29:28
or i will not dos the juno faucet so yeah we broke a lot of stuff today
1:29:33
but not us can we i mean a couple uni uni-3
1:29:39
validators on this podcast right can we provide some infrastructure should you get off those types of things or is there other ways
1:29:45
to be able to do that that we can be able to provide some additional infrastructure for that structure
1:29:52
yeah so we had some i remember pop marks quite early on in our testing actually
1:29:57
recommended some pt they they they also offered to
1:30:03
run some infrastructure so i guess it's just saying we kind of need to um
1:30:08
sort out uh frey don't we we need to we haven't have we done actually any analysis on
1:30:14
how heavy we're hitting our pcs because we are hitting them very heavily we're probably hitting them pretty
1:30:20
heavily i mean i think there's a whole bunch of stuff that we've talked about a lot on here about infrastructure dowels
1:30:26
and all that other kind of stuff and i think howell is an interesting use case other
1:30:31
than other maybe junior swap i don't think there's another smart contract um on
1:30:37
other than maybe forces fortis was a [ __ ] for rpc calls as well that's going to be doing the kind of
1:30:43
numbers and the kind of hits that howl might be if people even slightly get into it
1:30:49
and i think when going all the way back to the very first
1:30:54
idea of how which was just like okay what is the most simple thing you can do like uh
1:31:01
that's going to try and that's going to basically break cosmasm was my original thought and it was basically like blog
1:31:07
platform or twitter clone are the two quickest ways to evaluate whether something a database a ui technology
1:31:14
whatever is actually usable for a consumer use case um
1:31:20
that was like in lisbon last year so it's taken a long time to have time to work on this to find other
1:31:26
people that are interested in working on it and to build this out i mean we've been working on it since what february march
1:31:33
so it's like five months of work to even get to an mvp um
1:31:39
and see if we can yeah break cosmetic basically but assuming we don't actually break the
1:31:44
underlying runtime we're going to have to like prov i mean howl itself is going to have to provide infrastructure um
1:31:52
that's the only reasonable thing to do and the hal dao can do that and that's also partly where my interest in akash
1:31:57
and all that stuff is coming from at the moment as well i'm kind of thinking longer term that you know the hal dao is going to need to
1:32:03
pay for some of that stuff or you know longer longer term it becomes its own chain and then problem solved
1:32:09
within its own marketplace it pays for itself out of its own naked token purely out of inflation or whatever and
1:32:15
actually you know we're thinking really long term the sort of stuff you might want to do with the ownership model for something
1:32:21
like this to make it actually meaningfully decentralized it probably should be its own chain and
1:32:32
i don't know that it would be a cosmos chain i know that it would have to have cosmasm
1:32:38
and it would one i think we'd want it to have ibc but
1:32:43
whether it would have the sdk or not is it
1:32:48
it depends where the development of the sdk goes yeah you want to get to one second block
1:32:53
times is that what you're saying well i mean one second it would be benefit it would be beneficial in this context wouldn't it
1:32:59
it would it would i know i i saw that like just in posting and messing around a little bit like i mean you have the the standard kind of like i'm waiting for a
1:33:06
block to post type structure right plus my i was using a ledger so my right thumb is totally worn down to a nub
1:33:12
because every single thing i'm 22 hits to be able to get to that point that it's not really the best hardware wallet
1:33:18
structure yeah yeah well i always suggest maybe as weaver has done a bunch of stuff on
1:33:24
essentially like the the bit that does the patreon mechanic hal pay um [Music]
1:33:30
is kind like the beginnings of actual wallet abstraction and account abstraction and weaver's done a lot of
1:33:37
thinking and built a proof of concept which like callum can probably tell you about
1:33:42
yeah he's yeah we've has done some great stuff i think actually the proof of concept of
1:33:48
that was prior to the pay like the subscription model so yeah yeah
1:33:53
it's really cool he's already written that but we've figured we can't throw loads and loads
1:33:59
yeah we can't use them and and it's got to build up trust right like i understand why someone with this
1:34:05
system when i explained to it i'd understand why people are a bit hesitant it's like whoa like who am i trusting
1:34:10
here so he built a very small and it's live on the test net right now a smart contract where you can
1:34:17
or is it i think it's using off z actually in its authorizations contract where you can
1:34:23
authorize uh this contract you deposit juno in and it can make the gas payments for you
1:34:30
so you don't have to click the kepler pop-up you just have to keep this thing topped up so i could keep i could top it
1:34:36
up with 10 juno and forget about it probably for a couple months before it even because of just how gas on juno
1:34:42
works you know it's like 0.001 juno so right yeah like and it just means you don't
1:34:47
have to go approve approve it just goes and you should only have to do one big top up which is exactly all like however
1:34:54
whatever top up you're feeling comfortable doing for some people that might be one for some it might be five but yeah that's so we've got a lot of
1:35:01
sort of mvps like that actually behind the scene all from weaver who's amazing and is very underrated as a uh well
1:35:08
there's another thing that's behind the scenes which is the the uh the path controller contract
1:35:14
which callum wrote which is quite important to uh like if you're a dow and you want to
1:35:20
mint arbitrary sub paths for stuff in your project or if you're a wallet and you want to provide free names
1:35:26
for your users or something like that um you need callum's contract where you basically buy a root name send it to
1:35:33
that contract and then the contracts can essentially control it for you um
1:35:38
so there's actually there's a lot of different bits and pieces that we have knocking around for
1:35:44
road map items uh but yeah i think it all depends on
1:35:49
whether people actually use it right first we have these we have these floating pieces that we want to put in the puzzle
1:35:56
but in what order depends on what people want yeah so yeah and i deal with one we've
1:36:03
already i've already seen a few suggestions where it's like oh i don't really like having to approve tx's each time
1:36:08
and it's like we've got a poc solution for that we've also got a solution for
1:36:14
onboarding another one which we've sort of we haven't written any of the code for but it will be possible we have to
1:36:19
again wait until how goes live the trading but setting where howe can be used for gas
1:36:25
so if someone just gets sent how no juno on chain even though i think that's
1:36:30
technically impossible because the account needs to be active if they don't have juno for gas we could allow hal for
1:36:36
gas using this other contract and they're similar to the authorizations module where it has a juno balance which
1:36:42
is funded by us but it will use the how token juno swap to try and keep the level they'll always try and come back
1:36:48
to like 10 juno so it always has 10 juno's enough for let's say a thousand transactions he'll
1:36:54
always try and keep go to that target 10 juno eventually and yeah that creates some cell pressure
1:37:00
for how token but i don't think it would be a problem really like the pressure would probably
1:37:06
be quite minimal i mean yeah i used it for probably
1:37:12
two three hours today and i mean for for some of you guys launched today like i didn't i didn't come across any blockers
1:37:17
like there was nothing there that didn't work like everything kind of worked does it was there i had a lot of functionality i wanted right like there
1:37:24
was a lot of things that that were like either confusing or like just trying to understand how you guys are thinking through it or those types of things but
1:37:29
that's that's natural right um but everything like for uh for something you guys just launched today i thought
1:37:34
it was extremely put together like everything just worked you know what i mean uh so do i need hell to be able to post
1:37:42
no so you could earn how by writing a good post
1:37:48
which is something i quite like i quite like that as a model as in a way of onboarding people you don't i mean
1:37:53
caleb's going to be rich off of this model so i've wasted too many bangers on this
1:37:59
podcast guys you have to pay me it's my turn
1:38:04
finally the ship posting pays off he's been training for this for many years how do i make money by ship posting
1:38:11
yeah i mean that's actually so is this worth saying they're going to be some challenges next week um to break
1:38:17
how and there are rewards in how which will be applied on main net to be
1:38:23
gained for for breaking it and it's going to be stuff like uh you know become a howl whale
1:38:29
a crew the most subscribers stuff like that you know basically stuff where you form a whole style take over
1:38:36
like like literally if somebody if somebody rugs the dao will be very impressed we have talked
1:38:44
about it quite a bit i think we know of a few vulnerabilities that if you worked
1:38:49
it out and got organized with some other people you could probably rug the dow um that would be funny um but we're going
1:38:56
gonna publish some of these um sort of on probably monday and then uh see whether hal is still standing on
1:39:04
tuesday uh it would be interesting to see how quickly people can work out ways of of gaming the system
1:39:10
um because uh we obviously have some moderation and stuff in place but it's quite rudimentary at the moment so
1:39:17
uh you know that's why we kind of want to incentivize people to break it because
1:39:22
um it's we we obviously are going to do stuff around you know maybe traditional
1:39:27
auditing but at the same time like bugs and code integrity are one thing um and people
1:39:34
using the system that you design the incentives wrong on is like totally another and i think
1:39:39
we're more concerned about the latter case at the moment um which is probably a false sense of confidence from the rust compiler i
1:39:45
don't know yeah yeah agreed
1:39:51
null's just zoned out to the middle distance so he's he's using hell right now he's lost
1:39:58
i'm just playing around with how there you go and you know something that was
1:40:04
really evident to me is that you have to buy a dens address to [ __ ]
1:40:13
well if you're listening now you would know that eventually my contract can fix that and provide it for free
1:40:19
oh that way freezies yeah yeah it's just uh yeah i mean it's good
1:40:25
it's a [ __ ] you know different uh
1:40:31
integration model to pump up those returns on the dens
1:40:38
it's like yeah yeah well i mean like well too fast too like at the moment it
1:40:45
is yeah but it's pretty cool spring is everybody's finished dosing judo tools they're kind of down the rpc's back up
1:40:53
uh that char limb though no it's old school twitter used to be
1:40:59
140. so yeah i like it yeah i'm okay with that i like two five five
1:41:04
better like two five five oh what like a like a bite yeah okay i can see that i can see the appeal of that but it's too
1:41:10
late now we've committed to 140. man too late
1:41:16
i don't know raise the ticket maybe we can argue about it like are you guys functionality here to like you know do
1:41:23
uh tweets yeah yes so yeah like fred's so that's something i would quite like
1:41:28
because that would be a cool way for before we get long form away a lot of people on sort of cosmos twitter even
1:41:35
crypto twitter in general you know they write a thread so that's definitely saying we want to support so yeah we do
1:41:41
we have read some replies but at the moment they're kind of not grouped together in the ui
1:41:46
so the question the question we have is like whether we should recursively like
1:41:52
always be grouping those together or whether we should because obviously if this is successful in any way we're going to have to like put out some
1:41:58
infrastructure build out some um indexing and stuff and then we can kind
1:42:05
of arbitrarily decide how to make that work and that that's that kind of secondary layer of stuff
1:42:11
do we just throw all our threads in there is the question you know and just have that served out of the secondary
1:42:17
thing i i don't know we we will kind of cross those bridges when we come to them i guess all the data exists and the
1:42:24
machinery if you like to make threads totally works already that's just the replies just thread of replies same as
1:42:29
twitter um yeah how it's displayed is different at the moment because we don't really have threads
1:42:35
we need to follow the linked list yeah yeah yeah the graph is perfect for our
1:42:42
and even like sub query anything all of those indexing solutions are literally perfect for how we can just you know
1:42:48
persist right to chain as in the only time you'd interact with the chain ideally would be when you're
1:42:54
writing a post and then it would all be done through this graphql thing yeah unless i'm really overselling subquery
1:43:01
but i think that's sort of the ideal model right yeah yeah they've been chasing us a
1:43:06
little bit about when we go live are we going to use sub query and the answer is probably yes
1:43:14
there's probably uh unless nobody uses hal in which case then in which case
1:43:20
on to the next project i would but that that piece i would kind of think would be part of mvp like i the
1:43:28
it'd be difficult to launch without that querying like i can't even like find people i was trying to find you guys but
1:43:34
i can't do that until i either find a post you guys are in or something somewhere that so i don't know how you would launch without that because i
1:43:39
don't think yeah well we actually we we the search thing we sort of all
1:43:45
thought that the other one or one of the other ones had actually just implemented it there is a ticket there's bit i think
1:43:50
it's the second it's like it's like it's like from march it's like from march which is like a month after we started
1:43:56
properly developing yeah and it just gives you an idea of like how kind of tunnel vision sometimes
1:44:03
we have been on this is that we're just like like literally i think today i was just like hang on lads i can't find the
1:44:08
search bar yeah and then callum was like there is no search bar and i was like i'm pretty sure i saw a search bar and
1:44:14
counts like there is no search bar and i was like i was like but we have the endpoint to do search it's like yeah but there is no
1:44:20
search bar like okay well i'll make a ticket for it and i made a ticket the calendar was like there's already a
1:44:25
ticket for the search bar and i was like [ __ ] sake this is probably like page as well yeah yeah yeah
1:44:33
but yeah so yeah it's just the pattern matching on blockchain is hard because
1:44:38
it's impossible sure so we couldn't we couldn't do like you know on twitter where if i if i type cal you would get
1:44:44
like me and then you know whoever all of my burner accounts that uh i used to pump up uh pump up
1:44:52
impressions on my posts uh but yeah so currently the the very basic search implementation would be if
1:44:58
it's an exact match show it which is exactly how dens works if you go on dens and you go to register a name
1:45:06
right that's how you can check if it's taken but you're doing some search on the hashtag right so there's something there
1:45:12
yeah but that's only an exact match as well i'm very
1:45:26
i mean the core of it is the structure around being able to stake on a post and that's yes
1:45:31
pretty awesome at the center of it is this this model where yeah you know it
1:45:37
you as a content consumer get rewarded but also the content you're consuming gets rewarded which is very
1:45:44
very different and i've been on a few things now where i've said it's for a lot of people who don't have the largest
1:45:50
following you know they're not big enough to get a conduct sponsorship but they're um they're big enough to
1:45:56
pull in some impressions uh they can they can get some get some rewards on how which i think is awesome
1:46:02
yeah i i think i mean that's where this fits in which is like kind of this great this great flattening of
1:46:09
bring it down to content content and value right um and being able to reward that i think that's awesome
1:46:16
and really curating your feed around that and we've also got some more stuff on sort of
1:46:21
in like sort of grouping and having custom feeds or you can even you can even separate your feeds based on elias's you could have your own eliases
1:46:28
and that could be your each feedback right so you just say that again
1:46:35
you can say that word again aliases there you go elias is yeah sorry
1:46:41
uh yeah you have the emphasis you had the emphasis on the long than
1:46:46
the wrong i just feel like there was a bit of a glitch in the matrix there and i was like am i the only one hearing
1:46:52
this what does this report
1:46:58
post do where the [ __ ] does that go oh that goes to the moderator so do you want me to ban someone
1:47:05
but i won't be a moderator on mainnet i've just made myself a moderator for testing are we moderating
1:47:11
i am i i made a multi-sig proposal to make myself a moderator just so i could see if it was working so am i banned
1:47:18
already then am i banned from how did you do that oh yeah yeah yeah i've taken your money don't worry it's mine
1:47:24
how do i stake a post i can't say the link the heart the heart
1:47:30
oh is that it you just press yeah what i what i didn't pick up which
1:47:35
obviously these guys help me out with it's not like like on twitter you just like hit a like and you have 490 likes but but it could be 490 likes from like
1:47:42
one person right so yeah like one day is one like right so you can stake
1:47:48
so i could stake all of my how on something that you have and not just 500 likes or whatever right
1:47:56
yeah and it's maybe i should say it's one howl at one arbitrary point of time
1:48:02
is the like because your likes will keep on increasing because your like will expire
1:48:07
because you are liking how which is a stake is you lock up somehow for for an
1:48:13
amount of epochs which is currently 24 hours on testnet and then it comes back to you
1:48:18
so you have to keep on coming back and i think a maintenance you guys are talking like two weeks or something like that yeah yeah so like
1:48:26
on testnet it's currently an epoch is an hour so you get rewards every hour and the epochs end after 24 so it's a day
1:48:34
or whereas on main net would do any epoch is a day 14 14 epochs which is two
1:48:39
weeks that is subject to change still we could we could easily change it that's all
1:48:44
configurable imagine this imagine if your like button
1:48:49
was actually staking juno and it was backed back ended by a liquid
1:48:55
staking contract uh i don't want a headache
1:49:02
those are all words i understand but that was 2d gen for me i just blew it so so what we're going to be introducing
1:49:10
is a how derivative called a wu
1:49:16
and it's going to be a stable coin which is tied to the value of treats okay
1:49:23
oh there's the value of carbon and it's algorithmically pegged to
1:49:28
treats through some minute and burnt mechanism to be fair maybe we should actually have
1:49:33
two tokens in hal just so that we could claim it's a stable coin and then somebody'll be like there's literally
1:49:38
nothing stable about this desire is they have a relationship there's
1:49:44
actually very complicated math it's actually very very gross clearly
1:49:49
yeah we just have y equals zero complicated these two these two stable coins
1:49:56
like a double helix yeah it's like the it's like the cobras head from
1:50:01
it's top gun again what is the [ __ ] is wrong with me today this is the new target
1:50:10
yeah whatever viper cobra whatever then they give the twizzle twizzle if you've seen the film
1:50:16
you know i'm talking about it they go towards the ground the two things yeah
1:50:22
yeah yeah