Pages:
Author

Topic: Stop fuckin' around, fork the son-of-a-bitch already. - page 2. (Read 9367 times)

legendary
Activity: 2674
Merit: 3000
Terminated.
once you have researched what they all actually are.
-snip-
You yourself have no idea what they all actually are.

Interesting assertion. I'd like to read your extensive code review of each such "other implementation".
There isn't an extensive code review of such an implementation by myself (probably not by anyone, as I haven't seen one). That's not the point though. The point was that 'these people' are throwing around assertions regarding the safety of random block sizes just because 'it's fine one some test-net'.

You do, of course, realize that if one lonely cpu starts mining upon this chain, you will have been proven utterly wrong?
I was talking about the current miners of Bitcoin.

Oh wait - self-contradiction in one neat paragraph - quite an achievement!
It's not a contradiction; read the above.

Well, no. Segwit does absolutely nothing to make larger blocks safe.
Yes, it does actually, since there's no such thing as "The SegWit Omnibus Changeset".
legendary
Activity: 3080
Merit: 1688
lose: unfind ... loose: untight
no way is any miner ever going to support this.

You do, of course, realize that if one lonely cpu starts mining upon this chain, you will have been proven utterly wrong?

Quote
Not to mention the amount of botnets that will hop onto this bandwagon.

Oh wait - self-contradiction in one neat paragraph - quite an achievement!

Quote
Quote
an implementation of BitPay's adaptive block size algorithm, adjusting in the range 2MB-4MB
Unsafe without Segwit,

Well, no. Segwit does absolutely nothing to make larger blocks safe. Although other changes bundled into The SegWit Omnibus Changeset do. Which of course are available for other implementations to adopt, should they so desire.
legendary
Activity: 3080
Merit: 1688
lose: unfind ... loose: untight
The problem is that these "other implementations" only have half-baked, horribly coded improvements.

Interesting assertion. I'd like to read your extensive code review of each such "other implementation".
legendary
Activity: 2676
Merit: 2203
BitcoinPenny.com
Best. Thread title. Ever.  Grin

Regards,
Chris
legendary
Activity: 2674
Merit: 3000
Terminated.
i think its time lauda goes spend a few hours to ask his friends the difference between xthin and headers first
Again, I have no "friends" to ask. Anyhow, both of those proposals are such garbage that looking into them is a waste of time.

ill give hm a hint.
If anything, history has taught us that the information provided by you is false in most cases. Please let me know once miners measure 95% node adoption. Roll Eyes
legendary
Activity: 4424
Merit: 4794
im laughing here.

i think its time lauda goes spend a few hours to ask his friends the difference between xthin and headers first
Edit.. also compact blocks..
ill give him a hint. they all dont send the whole block in one go. they send the headers first, in both cases.

i think its time lauda goes spend a few hours to ask his friends the difference core 1mb base 4mbweight  and 2mb base 4mb weight.
ill give him a hint, they both have linear validation, they both have all the other fluffy features. but it actually increases the txdata in the baseblock to give REAL proper capacity.

once you have researched what they all actually are. then come back, without a core fanboy hat on. and instead with a coding logic hat on. basing your reply on actual real world usage, actual features involved and actual reality scenarios.
if i see you one more time basing your opinion not on code facts, but pure opinion of something just because of WHO. then you are no better than a racist who cannot see beyond the who

in short i want to see lauda approach the debate with technicals and not the simple analogy of BU=warewolves and Classic=vampire, kill kill kill
legendary
Activity: 2674
Merit: 3000
Terminated.
We never wanted a split, we just want bitcoin to be available to everyone.
If you really think idea behind these controversial fork attempts is that, then I'm sorry to tell you, but you've been deceived. You may want that, which is perfectly fine, but that's not the intent of the people who started with the controversial forks.

The facts still state that you are wrong, and 4MB blocksizes are perfectly safe. 20MB with xthin is also perfectly safe. But i'm happy with 4MB for now, or even 2MB until we need more.
Saying "xx MB block size is safe" is wrong. Saying "xx MB block size is safe if we limit the TX size to xx or less" may be true. These two statements are inherently different, ergo I'm not wrong.

No one has the money to set up a several months long non-stop spam attack.
This got to be a bad joke, right? A fair amount of people have enough money to spam up the network for a very long time.

This is all legitimate users and you know it.
Please post the testing methodology that extracts 'real user transactions' from the pool of all transactions, i.e. excludes 'spam transactions'. I'm sure everybody would like to know how this revolutionary method works.

Quadratric validation time doesn't matter if you just limit the transaction size to 1MB or lower.
The statement contradicts itself. It matters until you add even more limits to Bitcoin.
legendary
Activity: 1106
Merit: 1005
Exactly, it's just a spin-off of the actual bitcoin blockchain Smiley.
I don't care about the name bitcoin classic, unlimited or I don't know...
So introducing confusion in addition to the media portraying Bitcoin as a joke due to that is what you want?

the important thing is that they increase the blocksize and add other few interesting things.
Fun fact: Neither one of those teams have developed anything worth incorporating into Bitcoin Core. Don't get me started on idiotic ideas such as "header-first mining". Roll Eyes

We never wanted a split, we just want bitcoin to be available to everyone.

It's you who wants to artificially limit bitcoin. Go ahead, limit bitcoin, but we will leave you and your Cripplecoin behind and fork bitcoin to get bitcoin back as it was supposed to be. A currency for the people.

No it's not. We can safely increase the limit to at least 4MB right now, and there won't be a problem.
Yes it is and no we can not.

It has been tested on testnet already. Stop spreading misinformation. There's nothing dangerous about bigger blocks.
The BU testing methodology is garbage and should not be used as 'evidence' for anything.

The reason of the blocksize limit was not decentralization, not safety, but anti-spam.
Both safety and anti-spam.

The blocks aren't full of spam
You can't prove this and you know it.

Also it's ridiculous to say 2MB blocksize is dangerous while saying we should get 2MB later. f it's not dangerous later it's not dangerous now either.
No, it is not. You clearly have no idea what you're talking about. Quadratic validation time without Segwit. I highly doubt you even understand the big O notation.

Of course everything you disagree with is garbage in your eyes.

The facts still state that you are wrong, and 4MB blocksizes are perfectly safe. 20MB with xthin is also perfectly safe. But i'm happy with 4MB for now, or even 2MB until we need more.

I do know (and so does everyone else) that the blocks grew over time and for the past several months most blocks have been 900+ kB. mostly legitimate transactions.

No one has the money to set up a several months long non-stop spam attack. This is all legitimate users and you know it.

If you claim to not know this is legimate users, you're just plain stupid.     

Quadratric validation time doesn't matter if you just limit the transaction size to 1MB or lower.
legendary
Activity: 2674
Merit: 3000
Terminated.
No it's not. We can safely increase the limit to at least 4MB right now, and there won't be a problem.
Yes it is and no we can not.

It has been tested on testnet already. Stop spreading misinformation. There's nothing dangerous about bigger blocks.
The BU testing methodology is garbage and should not be used as 'evidence' for anything.

The reason of the blocksize limit was not decentralization, not safety, but anti-spam.
Both safety and anti-spam.

The blocks aren't full of spam
You can't prove this and you know it.

Also it's ridiculous to say 2MB blocksize is dangerous while saying we should get 2MB later. f it's not dangerous later it's not dangerous now either.
No, it is not. You clearly have no idea what you're talking about. Quadratic validation time without Segwit. I highly doubt you even understand the big O notation.
legendary
Activity: 1106
Merit: 1005
2 MB is inherently dangerous

No it's not.

We can safely increase the limit to at least 4MB right now, and there won't be a problem.

It has been tested on testnet already. Stop spreading misinformation. There's nothing dangerous about bigger blocks.

The reason of the blocksize limit was not decentralization, not safety, but anti-spam.

The blocks aren't full of spam, and the ant-spam measure is doing more harm than good right now, it's time to remove it while we work on a solution to keep the blocksize small without sacrificing security or throughput.      

Also it's ridiculous to say 2MB blocksize is dangerous while saying we should get 2MB later. f it's not dangerous later it's not dangerous now either.

Quote
though the internet is proven to be fast for hundred of MILLIONS of people as shown by all the countries doing livestreaming to prove its not a dream. the community accept some places can get over 100mb/s

500 Mb/s (Down and Up) is available to the public as well. (I have that internet as a regular user).
Pretty sure that's available for reasonable prices in most countries nowadays.

I know it's not the 'average internet', but it's available, and it will become standard eventually. As with all technology, the 'rich' are only a couple of years ahead of the mainstream.
legendary
Activity: 2674
Merit: 3000
Terminated.
Quote
optional change of POW (modified scrypt, CPU mineable and hopefully ASIC-resistant)
This obviously wouldn't be Bitcoin anymore, and no way is any miner ever going to support this. Not to mention the amount of botnets that will hop onto this bandwagon.

Quote
an implementation of BitPay's adaptive block size algorithm, adjusting in the range 2MB-4MB
Unsafe without Segwit, although I like the idea of adaptive block size algorithm with an lower and upper bound (as long as the upper one isn't beyond safe limits).

I have no idea why they want this to be associated with Bitcoin when it clearly isn't Bitcoin. I'll take a wild guess: Manipulation.
legendary
Activity: 1778
Merit: 1043
#Free market
the important thing is that they increase the blocksize and add other few interesting things.
Fun fact: Neither one of those teams have developed anything worth incorporating into Bitcoin Core. Don't get me started on idiotic ideas such as "header-first mining". Roll Eyes



I don't know but check this: https://ww.reddit.com/r/btcfork/comments/53tfnb/hfp0_an_early_prototype_of_a_bitcoin_hard_fork/
legendary
Activity: 2674
Merit: 3000
Terminated.
Exactly, it's just a spin-off of the actual bitcoin blockchain Smiley.
I don't care about the name bitcoin classic, unlimited or I don't know...
So introducing confusion in addition to the media portraying Bitcoin as a joke due to that is what you want?

the important thing is that they increase the blocksize and add other few interesting things.
Fun fact: Neither one of those teams have developed anything worth incorporating into Bitcoin Core. Don't get me started on idiotic ideas such as "header-first mining". Roll Eyes
legendary
Activity: 1778
Merit: 1043
#Free market



Interesting, freedom of choice Smiley no one will force you to run a software... never.
For every bitcoin you have, you will now have one bitcoin, and one bitcoin classic!  May the best coin win!

Exactly, it's just a spin-off of the actual bitcoin blockchain Smiley.
I don't care about the name bitcoin classic, unlimited or I don't know... the important thing is that they increase the blocksize and add other few interesting things.
legendary
Activity: 1596
Merit: 1026



Interesting, freedom of choice Smiley no one will force you to run a software... never.
For every bitcoin you have, you will now have one bitcoin, and one bitcoin classic!  May the best coin win!
legendary
Activity: 2674
Merit: 3000
Terminated.
says the guy who hasnt even read a line of code.
nor even knows a line of core code without spoonfeeding
I don't need to read nor know a single line of code in BU. There's something called third party code review (which unfortunately isn't as common as it should be). Oh wait, you wouldn't know that since you don't resort to knowledge nor rational arguments, but rather character assassination.
legendary
Activity: 4424
Merit: 4794
rather than all implementations coming to a joint agreement making all implementations all on the same level playing field coming to a joint consensus, which the community thought we reached before last christmas.
The problem is that these "other implementations" only have half-baked, horribly coded improvements.
says the guy who hasnt even read a line of code.
nor even knows a line of core code without spoonfeeding
legendary
Activity: 4424
Merit: 4794
I just mean that Africa is not a country, goober.  People are tightly wound these days.   Tongue
Cheesy ok i get your point about generalizing, well played,
hero member
Activity: 493
Merit: 518
The most interesting thing about the above-quoted list is that nobody wants to play the obvious zinger in response. Have a little fun. Please.

if your talking about why i switched from talking about hard drive to bandwidth, obviously hard drives are cheap so the other 'cost' is internet speed
if your talking about my reasoning to talk about UPLOAD instead of download.. there is a technical reason for this
if your talking about why i am mentioning uploading live video. as oppose to a tweet there is an obvious reason for this too

the reason i used live streaming is because its a real life scenario people can understand of large data moving..
(upload 10minutes of SD(0.5mbIT/sec) quality video=~37mbyte of data)

months ago lauda attempted to debunk WATCHING videos as that was only download bandwidth to which i replied months ago with UPLOAD stats of livestream recording at that time. and today was reminding him of the facts that the internet as a whole is not a problem for 2-4mb blocks upload and by default definitely no issues for download

though the internet is proven to be fast for hundred of MILLIONS of people as shown by all the countries doing livestreaming to prove its not a dream. the community accept some places can get over 100mb/s, but as a safe level the majority feel 2-4 is acceptable.. even core believe 4mb is acceptable now.
im not going to get into the debate of someone else advocating 20mb, as that is just poking laudas bear and not something the community as a whole could consider right now(though technically possible).

i personally have only been advocating for 2mb this year knowing rationally that in the near future technology progresses to allow for more (even if the technology is already available now. its best to stay in the safe zone)

in short 4mb is internet safe, 2mb is even safer so data speed debate of 2mb should be considered resolved as it was before late 2015 (even in cores 4mb eyes, 2mb is safe)

which is why i laugh hard when lauda was saying 2mb was bad.. yet his friends are saying 4mb is acceptable, and lauda has now backtracked to say 4mb is acceptable "because its core". but 2mb is still bad.

even if its a 2mb base 4mb weight linear validation rules.. lauda will still not be happy and will always try to debate some crap to keep core as the overlords, rather than all implementations coming to a joint agreement making all implementations all on the same level playing field coming to a joint consensus, which the community thought we reached before last christmas. and then tried to get core back inline in spring. and then again in summer..

I just mean that Africa is not a country, goober.  People are tightly wound these days.   Tongue
legendary
Activity: 2674
Merit: 3000
Terminated.
-snip-
I suppose we can keep musing about how irrational it would be for such blocks to mined... but, food for thought.
But, but, BU team promised me everyone would be playing nice and creating sigop friendly blocks?  Roll Eyes

-snip-
months ago lauda attempted to debunk WATCHING videos as that was only download bandwidth to which i replied months ago with UPLOAD stats of livestream recording at that time. and today was reminding him of the facts that the internet as a whole is not a problem for 2-4mb blocks upload and by default definitely no issues for download
Comparing live-streaming to running a node is also a 'false analogy fallacy'. One has incentives, the other one doesn't for example.

im not going to get into the debate of someone else advocating 20mb, as that is just poking laudas bear and not something the community as a whole could consider right now(though technically possible).
Sure, even 1 TB blocks are technically possibly. This doesn't make it safe.

which is why i laugh hard when lauda was saying 2mb was bad.. yet his friends are saying 4mb is acceptable, and lauda has now backtracked to say 4mb is acceptable "because its core". but 2mb is still bad.
I have no friends, ergo this statement is an outright lie.

lauda will still not be happy and will always try to debate some crap to keep core as the overlords,
As said many times, not that it matters, I have no relationship to any Core contributor whatsoever.

rather than all implementations coming to a joint agreement making all implementations all on the same level playing field coming to a joint consensus, which the community thought we reached before last christmas.
The problem is that these "other implementations" only have half-baked, horribly coded improvements.
Pages:
Jump to: