There are a lot of reasons why this isn't ideal both in terms of blockchain abuse and politics. For example you have all the political pressure going on with "ohhh fullblockalypse" when the blocks are getting filled with crap and some people want to perform a power grab based on a perceived "problem".

Spam fills blocks => a "savior" comes along and says the solution is to increase block size to fill blocks with more spam under the pretext of scaling txs => you have both a power grab and more inefficiency to deal with.


I was just responding to the $$$/tx part, not trying to do a broader security analysis of incentives and disincentives, mining costs etc.

I'm in favor of the rational use of the blockchain, not necessarily pro-small blocks or big blocks. You can say I'm against block abuse. I wouldn't mind bigger blocks accompanied with a mandatory fee increase to prevent abuse and keep blockchain activity for transactions, instead of spamming or third-party storage.


The 3tx/s is what we have right now but it won't be forever, obviously. With the available software, hardware and networks we are at the low-end of tx capabilities. That will change in the future as all of these evolve. So there'll come a time* when more and more low-value txs will be processed, either directly or indirectly (sidechains), through BTC.

* Not that they aren't currently being processed. I mean that in a post-crowding-the-dust-out scenario.


By the time the subsidy goes down significantly (let's say the halving down to 900 or 450 BTC - which are 2-3 halvings ahead), 2 things will have happened

1) Higher tx capabilities - perhaps 10-20-50x or more, whether directly or with sidechains
2) Much higher BTC price to compensate for subsidy losses. As inflation lowers, BTC becomes stronger price-wise, thus mining 3600 BTCs at 400$ would give the miners less than say mining 900 BTCs at 10.000$.



Key word that Satoshi used: "currently". He realized the problems of the system in terms of scaling right now but expected this will change in the future. We can try to bring the future closer by working on software solutions and waiting for hardware solutions so that it can scale better and more txs fit without problems. However it will be crucial to not allow spam to fill the blocks and for this to happen you must have economic disincentives in place that crowds out the spam: much higher fees than we have right now.

Even at 5c mandatory fee for BTC, a 50c micropayment would be possible with BTC fees being 10% of the whole payment cost. I doubt MasterCard charges less for that. Personally, where I live, I get a warning when trying to pay certain services (post office comes to mind) with something like "an additional 0.35 euro will be charged for credit/debit card transactions". It's like "pay with cash or else +0.35 euro". So if I wanted to buy a 0.72euro stamp for mailing a small envelope inside my country, I would get a +0.35euro "cap" on top of it.

The problem with very low BTC fees is that you can't get cheap txs without getting cheap blockchain abuse. If you get first block inclusion for 0.05 USD and a spammer can get +10 to 20 blocks inclusion for his spam at 0.01 USD, the blockchain can be spammed and bloated for peanuts. So this must be fixed.

If I'm reading this correctly, you have to pay a subscription too for having a merchant/business account.

It's problematic for stuff of 1$, unless it goes much different for large merchants who have special deals.

The fees are pretty high:



So if I want to buy an mp3 from an artist, and the artist charges me 0.99$ for it, paypal will take ~0.40$ of it. Paypal becomes the artist's 60-40% partner. So this option is clearly not viable. If you go through bitcoin, the artist can keep like 98-95% of the money.

All I'm saying is that bitcoin can be viewed as different things, depending the time you live in.

If you live in 1995, it could be "impossible"
If you live in 2015, it could be "problematic to scale"
If you live in 2035, it could be "fantastic"

...and the basic algorithm could be practically the same, the only thing that will have changed is the way the hardware and networks have gone upwards in capabilities by 1000x to 10000x.

So the problems are not necessarily "inherent" to bitcoin, but rather a result of a software-hardware-network ...equation.

Obviously, by improving the software right now we can do more with scaling. For example I was reading about ASICs doing the validation. Well, why ASICs and not GPUs for start? This should be relatively easy to port, with opencl and stuff - so it would allow cpu+gpu to be exploited for max performance. The processing part could take a boost there for sure because GPUs are much much faster.

As for storage, that's where we'll need a breakthrough I suspect. Compression will probably be one that gives serious gains (a simple gzip on a ~130mb block.dat takes me down to something like 100mb), but since CPUs are slow with the compression/decompression of large data sets, maybe it too should be given as a task to GPUs in multiple "blocks" of compressed data to exploit GPU parallelism. If a computer can handle compressed data with virtually no lag, then perhaps there can be a positive spillover effect to the network itself by the transmission of compressed data packages which are compressed/decompressed in realtime by the GPUs for near-zero lag. I think harnessing GPU power is definitely something worth exploring for future scaling, in more than one ways.

Along with satoshi saying bitcoin is not suitable for micropayments perhaps

Ναι το λεω ως προς το verification.

Again, there is no "paradox" here.

The technology (of bitcoin) can accommodate the accumulated number of every single tx on the planet if you give hardware and network technology the necessary time to grow. Whether that time is 10 or 100 years, nobody has a clue. Perhaps some variables and constants in the software will need "enlargement" to be able to fit everything, if they are currently "downsized" for economy, but it can happen with near 100% certainty.

What is also 100% certain is that this can never be more efficient than a hierarchical system.

What VISA could do in 2015 due to being more efficient, Bitcoin will do in +X years due to the technology compensating for the p2p / decentralized induced inefficiency. Will VISA be able to transact MORE than Bitcoin at that point? YES. Will it matter? No. Why? Because human needs for transactions are finite. You don't jump from 3 purchases a day to 1000 purchases a day. So at some point VISA will be able to process, say, a trillion txs per day but there will be no trillion txs to process. So, for VISA, the technological progression is meaningless because it doesn't allow them to do more. Perhaps they could run their data centers cheaper and that's it. For bitcoin it's another issue altogether because technology unlocks the potential which is burdened by the inefficiency.

There is nothing contradictory. You can exceed all transactions globally (the accumulated credit card, bank transfers, paypal payments, cash transactions etc) but you can still be more inefficient than banks or visa.

Maths = the ever increasing technological capabilities.

Perhaps the following example is better: Say Satoshi came along and deployed BTC in 1995, the era of Pentium 100 MHz, 16MB ram, 2 GB hard drives and network speeds that were very slow over the globe. My web/ftp speeds back then were like 1.2kb/sec on my 14.4k modem on non-peak hours.

How many tx/s could that bitcoin network do, with the means of 1995?

Fast forward 2015, how many?

Fast forward 2035, how many?

Fast forward 2055, how many?

So, it's not a matter of if, only a matter of when. Time is what makes the impossible => possible. Even with an inefficient method. Technology compensates for the inefficiency.

I disagree on the "it won't happen" part, because, well...maths... It's not a matter of if. Only a matter of when. It could be 30 years or 60 years, but at some point even 1GB or 1TB blocks will be peanuts for the standards that exist in future time-space coordinates.

I agree on the "we will never be able to process more than a centralized...", because, well.... p2p inefficiency.

If you make an assumption of ever-increasing technological means, it will *definitely* happen. The only question is "how long before it does?".

Home connections in the 80s/early 90s, were like a few hundred bps to 1200-2400 bps. That's bits. Now there are fiber 1gbps connections in some places.

CPU processing abilities that are now in a simple smartphone, used to be "supercomputer" stuff that filled up entire rooms: http://pages.experts-exchange.com/processing-power-compared/

Obviously, you can just sit and wait for the day when tech will allow you to do more, or you can optimize your modus operandi to make the most of what you have right now.

Ο αρχικος σχεδιασμος του bitcoin θεωρει οτι node = mining node. Το οτι υπαρχουν ...non-mining nodes για να κανουν καλο στο δικτυο ειναι "ανωμαλια" που προεκυψε απ'τη μεταβαση του mining σε πιο εξειδικευμενο hardware.

In theory, Bitcoin will scale even to visa level txs and beyond based solely on technological advances. The probability of that happening is actually 100% unless we somehow fail as a civilization, have a 3rd world war, a meteor hits earth etc etc. Otherwise, in 20-30-40 years, a 10gb block will be the equivalent of a 1mb one. It's like PCs went from kilobytes of ram, to megabytes, to gigabytes and home connections went from a few bytes per second to megabytes per second.

There are however some issues on the hardware market. Issues like the fact that CPUs are not advancing the way they should, given that AMD is lagging and Intel doesn't have to push too much (a 9 year old quad core is currently cheaper and quite competitive with low end chips of today by AMD and Intel. This would be impossible between, say, 1991 of a 486 @ 50ΜHz and 2000 with a low end Celeron or K6 running 300-500 MHz)

Optical drives are also stagnant. If someone saw the CD->DVD progression of the 90s, where DVDs had capacity near what a hard disk had at the time (in '95 a CD was ~700mb and a disk was ~500-800mb, in 98-99 a disk was ~4-10GB and the DVD was 4-8GB) you'd expect we would have >1TB optical discs right now but they are buried in some lab, despite having read about such advances multiple times. We didn't use to have ...HDDs for "storage", we had floppies, then CDs, then DVDs... now we have HDDs - which in a way is artificial demand for HDDs and still HDDs don't offer cost effective ways of storage.

I bought a WDC 1TB 7200rpm (blue / ealx sata2) around 5 years ago for ~62 euro and its price is not much different currently (59 euro). In the 90s, if I paid the same money I paid on my wdc 80mb in 92, I would get a Fireball of 6.4GB in 97.

If manufacturers were running full speed we'd be seeing a whole different market. Now with the content stored online, you download a movie, see it, erase it. 10-15 years ago, people hoarded videos etc and stacked them in DVDs, etc. Then they run out of space and started buying HDDs and external HDDs. Then they stopped hoarding, just downloaded and erased. So no motive for HDD manufacturers to go dramatically lower on $/GB, plus they have a cartel nowadays that works as inefficiently as the cpu one.

TLDR: While it's a mathematical certainty that over a long-enough timeline bitcoin will scale as technology evolves, in the short-to-mid term, we have to face some market anomalies that create an artificial slowdown in significant (and affordable) increases of cpu power and hdd storage space.

Για legit / non-spamming συναλλαγες, υπεραρκετο. Πιστευω θα αρκει για κανα χρονο ακομα - ισως και παραπανω.

Το spec του πρωτοκολλου ειναι δεδομενο. Στην αρχη διανεμεται η νομισματικη βαση και οσο αυτη η διανομη μειωνεται το fee πρεπει να παιζει μεγαλυτερο ρολο, αλλιως το δικτυο θα γινει χωνι. Ε η νομισματικη βαση εχει ηδη διανεμηθει σε ποσοστο ~72% και απομενει ~28%.

Καποιοι λενε οτι σημερα το κοστος ανα συναλλαγη (αμα διαιρεσεις το block reward δια 2000 συναλλαγες πχ που γινονται σ'αυτο) ειναι πχ 25 BTC + 0.3 τα fees  ΔΙΑ 2000 txs = 25.3 / 2000 => 0.01265 BTC κοστος ανα συναλλαγη x 370$ = 4.68$ παιρνει ο miner για τη καθε συναλλαγη. Επεκτεινουν το σκεπτικο στο μελλον με τη λογικη οτι εφοσον το block reward πεσει απ'το 25 στο 12, τοτε ο miner θα παιρνει πχ 2.3$ αρα τα αλλα 2.3$ θα πρεπει να μπουν απο fees. Υπαρχει ομως και η παραμετρος της τιμης. Οτι αν πχ το bitcoin διπλασιαστει σε τιμη, τοτε δεν αλλαζει κατι για τον miner. Ειτε βγαλει 25 btc με 400$, ειτε 12.5 btc με 800$, σε $$$ (με τα οποια πληρωνει ρευμα, εξοπλισμο κτλ) τα ιδια θα παρει.


Η πιεση απ'το στρατοπεδο του gavin προβλεπω οτι θα αποδωσει... με την εννοια οτι εχουν πουλησει τοση προπαγανδα οτι το bitcoin πεθαινει / θα πεθανει κτλ κτλ αν δεν ανεβουν τα μπλοκ, που στο τελος θα κερδισει την αυξηση των 2mb με consensus ολων οι οποιοι αμα δε το δεχτουν θα φανουν "παραλογοι". Ετσι το spamming με χαμηλο κοστος θα συνεχισει ανενοχλητο (οπως και οι πολυ χαμηλου κοστους συναλλαγες χαμηλης προτεραιοτητας). Αυτη ειναι η δικη μου προβλεψη.

https://bitcoinfees.21.co/

The fastest and cheapest transaction fee is currently 40 satoshis/byte, shown in green at the top.
For the median transaction size of 340 bytes, this results in a fee of 13,600 satoshis.

13600 satoshi = 0.05$.

Με 5 cents του δολαριου, ή 4.4 cents ευρω, αυτη τη στιγμη μπαινεις στο πρωτο μπλοκ με υψηλη προτεραιοτητα. Και αυτο υπο συνθηκες backlog συναλλαγων / πηξιματος.

Με 1-2-3 cents θα μπεις, αλλα αργοτερα.

Το θεμα ειναι οτι fees στυλ 0-1-2-3-5 cents ειναι γελοια και ευνοουν το abuse του συστηματος - πχ φτιαχνουν ορισμενοι καποια σκριπτ και κουνανε τα ιδια λεφτα περα-δωθε 500 φορες για να γεμιζουν τα μπλοκς... για τη πλακα τους.

Ειτε θα εχεις υψηλο fee ως αντικινητρο abuse,
ειτε θα εχεις χαμηλα fees και θα τρως το abuse στη μαπα,
ειτε θα εχεις ενα block limit οπου το λιμιτ θα αναγκαζει αυτον που θελει να παιξει σοβαρα να βαλει μεγαλυτερο fee για να μπει η συναλλαγη του και ο αλλος θα μενει συνεχεια εκτος / unprocessed.

Αν ανεβασεις το block size limit και υπαρχουν ακομα χαμηλα fees, τοτε απλα ευνοεις το abuse του blockchain. Και αυτο εχει μακροπροθεσμες συνεπειες.

Αυξηση των fees χρειαζεται για να κοπει το spam. Η πραγματικη χρηση δεν ειναι ουτε μισο megabyte per block.

Ελεγχο γνησιοτητας των συναλλαγων, ναι. Αν δουν πχ οτι ο προηγουμενος εκανε mine μια συναλλαγη που ο αποστολεας εστειλε 10 φορες τα ιδια bitcoin σε καποιον αλλο, τοτε προφανως πολλαπλασιασε τα bitcoins και αρα αυτη ειναι ακυρη και απορριπτεται. Απ'τον επομενο ομως που θα κανει mine το block. Αυτος που παραποιησε το block δεν εχει ιδιαιτερο οικονομικο κινητρο γιατι θα χασει το block reward οταν οι επομενοι ανακαλυψουν (βαση του πρωτοκολλου) οτι εχει κανει παραποιηση και ετσι θα του βγαλουν το μπλοκ ως μη-γενομενο (ορφανο), οπως και την ακυρη συναλλαγη.


Οσο το ορισουν. Σε default ρυθμισεις απλα θα βαλουν οσο χωραει μεχρι 750kb, απ'το υψηλοτερο προς το χαμηλοτερο fee. Το default block setting ειναι 750, οχι 1mb.


Δεν ειναι βεβαιο το τι θα γινει. Εξαρταται την ποιοτητα των συναλλαγων. Οταν αυξανεται η προσφορα "χωρου" και υπαρχει αφθονος χωρος, τοτε το κινητρο να πληρωσει καποιος αυξημενο fee μικραινει. Απ'την αλλη αν καποιος επεξεργαζεται πχ τις διπλασιες συναλλαγες, τοτε και με τα μισα fees θα βγαζει τα ιδια. Αλλα υπαρχει και αλλη μια παραμετρος που ειναι το ποσο γρηγορα "εκπεμπει" το πιστοποιημενο block των 2mb. Αν πχ αυτο καθυστερει στη διαδοση του στο δικτυο κατα 1-2 λεπτα, τοτε αυτος ειναι χρονος ο οποιος μπορει να στερησει απο τον miner την πρωτια οτι αυτος ελυσε το block και αρα να χασει τα 25btc του block reward. Σ'αυτη τη περιπτωση, αν ο miner εχει αργο propagation λογω αργης συνδεσης πχ, τον συμφερει πολυ περισσοτερο να βγαζει blocks με 0 συναλλαγες και να μην τον ενδιαφερει καθολου το fee, τα οποια ουτως ή αλλως κυμαινονται σε αστεια ποσα σε σχεση με τυχον απωλεια του 25 btc block reward.


Καθε μπλοκ που προστιθεται ειναι +1 εγκριση. Το 6 conf δεν ειναι κατι το οποιο εχει νοημα σε σχεση με το mining. Ειναι κατι που αφορα τους συναλασσομενους περισσοτερο και την αξιοπιστια των συναλλαγων.

Το να βγαλει ενα μεγαλο pool 3 πειραγμενα μπλοκ σερι για παραδειγμα οπου το 2ο και 3ο επιβεβαιωνουν το πειραγμενο πρωτο block ειναι κατι το οποιο ειναι πιθανο (εννοω ειναι πιθανο να βγαλει καποιος 3 μπλοκ σερι, οχι απαραιτητα να το πειραξει). Γι'αυτο λενε 6 conf... σου λεει θα βγαλει ο αλλος 6 μπλοκ σερι για να "αλλαξει" την ιστορια του blockchain με παραποιημενη? Αλλα και 10 conf να εβγαζε σερι, παλι δε λεει κατι. Αν ο 11ος βγαλει σωστο μπλοκ και δει οτι ο αλλος εχει κανει μαιμουδια εδω και 10 μπλοκς, θα τον ορφανεψει και θα του ακυρωσει και τα προηγουμενα 10 μπλοκ και ο μαϊμουδιαρης ο miner θα χασει 25x10 = 250btc σε rewards.


Δεν αλλαζει κατι και στα 2 ερωτηματα.


Λογω πιο αργου propagation time στο δικτυο. Οσο πιο μεγαλα τα block τοσο πιο πολυ αργουν να διαδωθουν και αυτο δημιουργει προβλημα ορφανων μπλοκ.


50.000000000001% των miners

Το 75% που αφορα το classic fork attempt ειναι ενα hard-coded trigger, που λεει οτι ΑΝ το 75% των blocks γινονται mine με το classic, τοτε το classic απο μονο του θα παει στα 2mb και θα σκισει το δικτυο στη μεση, δημιουργωντας 2 bitcoins... το bitcoin που εχουμε τωρα, και το "classic" που θα βγαζει 2mb blocks. Αν το 75% δεν επιτευχθει, τοτε θα παραμεινει στο 1mb.

Θεωρητικα και με 51% αυτο μπορει να γινει, το 75% ειναι και καλα για να υπαρχει αυξημενο consensus. Αλλα το 75 ειναι λιγο γιατι αυτα τα ζητηματα πρεπει να εχουν ομοφωνια. Γι'αυτο και βλεπεις οτι ο ιδιος ειχε προτεινει 90 στο XT και τωρα το ...κατεβασε στο 75%. Το 90% εχει πολυ μικροτερη πιθανοτητα να δημιουργησει ρηγμα με 2 παραλληλα νομισματα. Το 75% ειναι σχεδον σιγουρο οτι θα σκισει το bitcoin σε 2 νομισματα τα οποια θα κυκλοφορουν παραλληλα.

I haven't studied cryptsy's wallets on doge, but it could have been affected in a similar manner to dash, so it was probably way oversold. Given that anyone who couldn't take out their btc was getting the doge with a huge "penalty" and moving them out to sell them, this must have created a continuous selling pressure due to the moving and dumping for BTCs.

See all these limits, hours, charges, fees mentioned even for much better services than the US?

Still 100 light years behind what BTC offers right now, with 4-5cents in fees for first block inclusion and avg 10 minutes time, for practically no limit in terms of value transacted.

And that, with a 100mb queue/backlog in the mempool and "blocks are full" bullshit.