consensus hardfork does not mean splitting the network.
intentional split is not the default end result of a hardfork

hardforks just leave the minority unsynced and stuck.. where by the sensible thing is to join the majority by upgrading with the majority and continue with the majority.. rather than be stuck in limbo..


however, opposers completely adamant to not join the majority.. can add additional code intentionally.
additional code is then needed for the unsynced minority to ignore consensus, ignore the majority, ignore the orphans and start their own network.
meaning they intentionally start their own chain and reboot their own syncing of new blocks from a different source than the majority.

that tweet was a question, not a statement.

and it is actually gmaxwell that wants to split the network

here i will spell it out for you.
here is blockstream gmaxwell trying to get anyone not blockstream to fork away..  and they replied they wont be that dumb to intentionally split the network... and before you reply BullSh*t.. its from the horses mouth himself.


even funnier.. gmaxwell also says he willing to intentionally split the chain just to implement the soft fork

as for the exponential vs linear.. its neither.

its not linear (straight diagonal line)

its not exponential (horizontal that curves vertically)

it IS an S-curve..


the issue is that we are still in the early days of bitcoin so we are still at the bottom section of the S-curve, so it appears like its exponential

only problem is that growth has been halted to not see the bigger picture

referring to your frustration of setting up a full node.

the reason your frustrated and many people are. is that while its setting up its not really 'usable' straight away, you cant see current balance and cant really spend anything until its synced..

right, thats the main frustration.

this can be solved so easily.

if the devs just got their implementation to not sync first then check unspents(utxo) second.. but instead grabbed some litewallet code that grabs unspent's data from other nodes first. and then done syncing second. the syncing then becomes just a background/unnoticeable thing. while the node is actually functional straight away.

bam!. easy to code, lets users just get on with using bitcoin straight away. problem solved

EG emulate electrum or other litewallets as soon as you open the node. then syncing is not a critical, thumb twiddling wait. its just a background function users dont realise is happening, because they are no longer forced to wait until synced before properly using it

how?

was it him saying he is happy to let blockstream do their LN(paypal2.0) if they let bitcoins mainnet dynamically grow.
which then got lost in translating and fed to the wolves to vomit out as the opposite?

update of my previous post which had graphical description of growth..

here is blockstreams Sheep (in red) impression of what they want to tell the world will happen if blockstream doesnt start its commercial service
"1gb blocks by midnight"

vs

my opinion (in blue) of onchain natural dynamic scaling over the next 33 years

1tb a month lol

ok.. uploading. 8gb/month to send live relay of tx's and new blocks. per connection

for instance.
100 connections to fully synced node requires 800gb
10 connections to fully synced node requires 80gb

as for helping other nodes sync. that is the killer data wise.

100 connections to 0 synced node(where your its only seed) requires 9.88tb  (yea doom)
10 connections to 0 synced node(where your its only seed) requires 988gb  (yea doom)
1 connections to 0 synced node(where your its only seed) requires 98.8gb  (manageable)

the solution.
dont have 100connections..
reduce the connections= reduce the data needed to upload.
especially dont connect to nodes that are not near/already synced.. thats the main trick.

to emphasise this point..
if 75 nodes had 75 connections 5625nodes would get the data in 1 relay.
so anything over 80 is overkill/not required.

i can safely assume atleast 75 people have good internet. so not all 5600 nodes need to upload to that many. as the recipient probably already got the data via someone elses node. so if bandwidth is an issue.. bring your connections down. dont act like a 'supernode' if you are only an average node.

this is also why blockstream are doing their 'fibre' (supernodes) so that they get the data out in one relay so that other nodes dont have to do all the work.. which means you can happily play '6 degrees of separation' with 6connections(48gb) just to relay tx data and blocks so the nod can get second opinions from other sources.

to add to this point, (to help explain it)

ethereum was NOT a consensus hard fork... it was a intentional split

some call Ethereums split a controversial split hard fork
some call Ethereums split an altcoin maker
some call Ethereums split a bilateral fork..

but not a consensus hard fork.. because thats a different thing altogether

(note: there's many forms of hard forks that have differing end results/goals/intentions. dont think harkfork=split. its an umbrella term for different possible results)

but ethereum nodes intentionally avoided consensus by actually ignoring each others chain data (google: --oppose-dao-fork) thus causing an altcoin, by disconnecting opposing nodes.

a consensus fork is ONE chain. where the minority dont create new blocks automatically. they are left orphaning what they dont like and never syncing.

in the case of a 95% agreement. of say 5600 nodes... 5300 agree 300 dont... what then happens is pools then do a consensus vote of their own. which gives time for the 300 to upgrade...
when pools get to their 95%. a grace period happens.. whereby both the 300nodes and 1pools not voting still have time to get onboard before being left unsynced and 'stuck'.

ok.. you might be about to rebuttle that losing 1 pools and UNDER 300full nodes is bad,, but..

blockstreams soft fork can lose 1 pool too.. and, would currently have over 2000 nodes no longer at 'full validation status' and demoted to half validation status.

i get what your thinking. and this has already been solved.. its the empty block concept, people still dont get.

firstly. pool A solves a block hands it out.. lets say pool B receives it.
immediately.
pool B tells the asics. to start working on an empty block. (no tx apart from blockreward claim(coinbase tx))
because poolB does not know what tx's to add into blockB because its yet to validate blockA

what happens is a pool can take upto 2 minutes to validate a block. (as long as its not got a super bloater 1mb tx)

and so before validating a blockA to know what unconfirmed tx's remain in mempool to start adding to pool B's empty block, if blockB 'luckily' gets a solution then Block B looks 'empty'.

obviously if BlockB is not solved after 2 minutes and if all is good with BlockA... poolB starts adding tx's from unconfirmed mempool to blockB and resending a hash to the ASICS to hash away at.

this is why most 'empty' blocks are found within a couple minutes of the previous timestamp. because they are only mpty because they were luckily solved so quick

..

now thats dealt with.
instead of letting 1tx have 20,000sigops, letting it have say 500sigops max. means that we will never ever ever have a risk of it taking 10minutes to solve a block by someone having a 1mb tx.

secondly pools can right now. in their own settings. without needing to get a consensus or an vote or permission. set their pool node to only allow tx's with < xxx sigops. and then those trying to send such clunky tx's will soon realise their tx's are not getting accepted. and will react accordingly (being more leaner with their tx's)

thirdly making leaner tx's also helps more peoples tx get accepted.
EG 2500tx's of 400bytes. instead of 1tx of 1mb

leaving the sigops soo high, even when (lets say segwit activated)  quadratics is not an issue. may save processing time.. but still allows a bloater to send a tx of 1mb. so reducing the sigop count of tx's should still be done anyway.

so knowing it should be done anyway. we might aswell just do it. and solve the problem sooner. rather than waiting for segwit to not do a full job

why is it that blockstream sheep use all of blockstream failures to sound like someone else is doing it.

Gmaxwell loves monero.. he is the lead dev.. he even has a monero address in his profile!! https://bitcointalk.org/index.php?action=profile;u=11425
gmaxwell actually wanted the other implementations to fork off, the other implementations laughed and aggressively said no
gmaxwell has helped add tx maturity(CLTV: bank 3-7day funds not available)
gmaxwell has helped add tx revocation(CSV: bank paypal chargebacks)
gmaxwell has pushed the fee war (advocated not to increase blocksize.. added 'average' fee estimator that actually increases fee even with no demand)
gmaxwell /rusty russell wants commercial permissioned services (LN Hubs)
gmaxwell/adamback organised many all expense weekends for miners to bribe miners
gmaxwell/adamback are members of the bankers hyper ledger (bankers blockchain)
gmaxwell/adamback are paid by bankers

as best as unbiased as i can be..
here are some options.

---

1. if segwit truly was 'backward compatible'. core could make some segwit transactions now and relay them directly to BTCC or Slush..(the two pools in support) get the Tx's added to the block and see if any orphans occur when that block is relayed to the network.
if no orphan occurs, then its backward compatible and that raises the confidence level.

core cough blockstream cough have $90m so they can happily pay BTCC $15k (price exceeding more than a blockreward) as a safeguard to btcc if the block orphans.
cheap 'test' in reality.

however it does not solve the second bit of confidence/risk that the majority of pools are not happy jumping forward at.. (lack of node full validation count)

---

2. doing it as a full community consensus. getting the community to have a consensus, rather than just the pools.
this is a consensus (hard)fork.. not an intentional split. .. consensus. there is a difference.

and if pools see a high majority of nodes able to fully validate these new blocks, then pools will gain confidence to vote for it and make them.

however it then means the community gain power to veto aswell as show devotion for (yep core dont want decentralised free choice).

---

3. so knowing they would need 95% community acceptance they would need to give the community something the community really want.
real true onchain transaction scaling thats not just a one time boost and also not having to let the devs decide what nodes should or shouldnt do.

so by also including dynamic blocks as a core release. nodes regain the independant diverse decentralised power. and it grows only when nodes show they can cope with.

then its just a 2 birds 1 stone election event where everyone gets what they want. and all different (10 different implementations) can be diversely accepting and consenting to the same things on an even playing field, rather than a one brand dominance/control event

DrSeuss

if you have the answers to the topic it may be worth you locking the topic, otherwise you will get people with colourful signatures making random comments to spam the topic forever.

the chinese actually pay for electric.
what the chinese get is free ASICS.

for $2100 they get from retail customers, asic manufacturers can make upto 5 ASICS..
thus give 1 asic to a poor schuck that pays the retail $2100 for one rig.
keep 4 for themselves..

well the smart option they choose
for $2100 they make 3 rigs .. give one to the retail customer that handed them $2100,,
keep $840.. and run the 2 spare rigs themselves
and spend the $840 on electric and wages.

that $840 works out as electric for the 2 rigs for 11 months netting them a further 3btc in pure profit for running the 2 miners for those 11 months

---

as for the powerplants electric.
the world is trying to develop area's like mongolia.
so they set up a power plant. ready to cope with future housing, retail and industry..
but while there are no houses, shops and industry. the electric that leaves the power plant is not being used (pure waste)
but the power plant is still producing a minimum electric output as thats just how plants work.

so while the power plant is producing and wasting XXXmegawats. and the energy is wasted going down cables that dont plug into anything..
they might aswell let anyone that wants to use it,  use it. the power plant atleast gets paid for it then and its not a complete waste while they develop new towns and cities

bitcoin is just code.
but its how PEOPLE treat it and use it that changes things.

bitcoin does not have a voice to tell someone to not pull the trigger.
bitcoin does not have a hand to take a gun from someones grasp

its PEOPLE that can stop wars. but right now PEOPLE like blockstream are trying to start another economic civil war, by commercialising bitcoin.

bitcoin was suppose to be the financial system with no barriers of entry. but blockstreams motives have caused things like fee wars. and now atleast a dozen developing countries that would benefit most from bitcoin cant/wont use bitcoin due to fee's being more than an hours wage.

and dont blame it on data issues.. blockstream are happy with 4mb of data per block. but dont want it to be 4x transaction capacity to equal 4mb of data. they are pretending to give a one time gesture of 2x capacity. but as fast as people adopt the new keys when it activates. they will take away this boost by adding in extra bytes of data which bloat the data but decrease the tx count back to what we get now.

EG a 360byte 2in 2out tx,  will become 1.370kb tx for the same 2in 2out. once you add the bloat of confidential payments, csv cltv and other features blockstream want.

so while they bloat up the data onchain without offering much in the way of real transaction scaling they will push people into LN commercial service hubs to rule bitcoin users lives.
only ever increasing the onchain block capacity as a gesture to allow the commercial LN settlements to settle without a bottleneck..
yes its already been thought about talked about and conceived as their ideal plan for the future. keep bitcoin held back until the commercial services needs the utility of onchain scaling..

bitcoins ethos is being eroded by PEOPLE who want control and to be at the "core" of bitcoin they even said so by naming their brand as such.

so while ethical people sit on their hands and trust bitcoin (voiceless handless code) will save the planet.. corporations with banker investment are corrupting the people that have their hands and voices of bitcoin

even as a non profit, they can be corrupt.

non profit means at the end of the financial year they cant have made any profit.
this is simply.. 'its month 11, lets go on a spending spree to get rid of the funds in the account'

oxfam have retail shops. but are a non profit because they make sure its always spend by year end.
oxfam CEO get a big fat salary that could have been used to make 1000 water wells. but he buys a house, a boat and a shiny car instead.
oxfam only put 10% of fund towards "worthy projects" once you take away all the money syphoning tricks.

same for red cross. ceo on over $1m a year. giving a low % of income to actual front end services.

thus being a non profit is not as 'ethical/moral' as any other corporation. it just simply means dont make a profit to not have to pay tax on.

even if set as a non profit.. blockstream can still have 'sponsor' contracts to halt bitcoin expansion and instead develop commercial services. as long as they dont make profit and spend every penny by year end

take linux foundation.. hyperldger foundation.. they are at foundation status which meant to be reserved for charitable causes.. but its just used as a tax free scheme.. and yea hyperledger is the bankers blockchain that blockstream is helping to make

here is the issue.

instead of making it a 2 part consensus
nodes first
pools second

blockstream decided to skip the hard consensus and went soft requiring just pools to form consensus.

the problem.. smart and risk averting pools wont flag desire for something if the nodes they send this new data to are not full validating nodes to double check the new data.

so while the node count is low. pools are not in a rush to change.. as there are risks of letting unupdated nodes just pass data on blindly.

it should have always been a real consensus vote. and core(blockstream) just thought they could slip it under the rug before christmas by bypassing real consensus by making it soft.. but pools wont risk it

dont blame pools when it was blockstream paid devs that decided to make the pools the decision makers. rather than a full community vote

so now what you should be doing is

we know you have 0.00202562 from 68a8fd9710594334cf960402fdca66dea88affa61d70baca1cff04230f7d3fe6
say you wanted to spend 0.0001 to someone else
leaves you 0.00192562
take off the tx fee 0.00015820
leaves you 0.00176742

so the example would be:
in
using 68a8fd9710594334cf960402fdca66dea88affa61d70baca1cff04230f7d3fe6  -  0.00202562 unspent

out
destination - 0.001
address you own - 0.00176742

knowing the missing 0.00015820 is the fee

not fixed. it varies depending on demand.(its a bidding war concept.. like an auction)
https://bitcoinfees.21.co/  helps find a good estimate.

the green line tells you the biggest chance of getting into the next block. at the moment it is 70sats a byte
which if you are just doing a 1in 2out is ~15820sats fee

some people get lazy and just put a fixed amount of say 80sats a byte knowing it wil most of the time outbid everyone to get in the next block.. only problem is if everyone done that. you then have to increase it.. (its a bidding war concept)

you can have multiple inputs. and multiple outputs.

easy guide calculation for approximate tx byte size

(148*in)+(34*out)=
this is an estimator as the answer can be +-10 different

eg
1 input if you only want to spend the $1.75(0.00202562)
2 output to go to one destination and the rest back to you

(148*1)+(34*2)=216
it actually results in 226byte when you actually build the tx

---

for instance if a 226byt tx costs 14cents
there is no point adding the other unspents (listed below) because
(148*2)+(34*2)= ~364byte = ~$0.23 tx fee
so why add in the 9cent unspent to just be eaten up in the extra fee by adding it.

---

you also have
0.00003 ($0.03) unspent a76f50b8ab76d2a0880be73306753423cc0cbea9a3aca20f13dfc51bdf3d4954
0.0001 ($0.09) unspent 5b9044c61a5c3d68f5bef7b67093e4109264e4aba836204669da42d5caeb12a1

---

lastly remember when you spend X to the destination you want. calculate whats left. then work out a fee you want to pay. take that from whats left and then put the remainder back to yourself

you spent the 0.005 from this
81bb5df4040f98793875591dcc9f93755b94425b9d1a6cc4c86871b246cc9185

but you still have 0.00202562 - about $1.75 in this one
68a8fd9710594334cf960402fdca66dea88affa61d70baca1cff04230f7d3fe6

just be sure when you use the unspent output 68a8 blah blah blah as a new input.. you set 2 output. one to the destination and one back to yourself.