GNU Taler | Hacker News Reader

GNU Taler

taler.net

479 pointsadrianN147 comments

I was at the presentation at SHA 2017.

The demo was nice, even though i did not understand any of the crypto. Although, to be fair, i didn't even try :-)

One important take away for me was that this project aims to replace paypal, not bitcoin. Its target is to establish an open source payment infrastructure that is better than paying with credit cards, paypal, etc...

This sounds awesome if it works and takes off

Yes, I would like to have some working open source micropayment system.

However, as far as i can see there is no bank offering a backend yet. Hopefully that will change soon.

It would have been great if the taler system had been usable at SHA 2017. We had to use weird, cheap and easily falsifiable plastic coins at the food court. Using taler instead would have been a really great tech demo...

I keep hearing about Taler from a friend of mine, but there hasn't been much adoption among banks or credit unions, limiting its usefulness.

The banking sector in the US is a mosaic of banks and credit unions running a wide variety of debit networks, with ACH vaguely tying them all together. I wish it was easier to move money between banks here, seems like the quickest way is to write a check!

ACH is quicker but much less secure, because the initiator has to have the recipient's account number and bank routing number, which also suffice to initiate wire withdrawals that do not require confirmation to complete and can't easily (if at all) be clawed back.

The average person doesn't have access to an ACH console though, and most banks & credit unions offer local transfers only, with "Bill Pay" that is just them mailing checks out for you. ACH is essentially for businesses only currently.

A few years back I was able to configure payments for my Citizens' Bank credit card from a non-Citizens' checking account without any verification. Needless to say I was livid and contacted their customer service, who were basically like "yup that's how it works".

(I closed my account with them soon thereafter. Citizens' is awful for other reasons.)

Most (if not all) of the major banks and some credit unions have joined Zele (formerly clearXchange) [1] to handle person-to-person transfers. Only since the rebrand has it been obvious that each banks person-to-person service is part of a multi-bank network, but I've been using it to pay my landlady since I moved into my house last May (money is transferred instantly, so there's no worry about timing a check with bill pay's slowness - and she never gets a piece of paper with my account information on it either).

It's taken forever, but I'm hopeful with the big banks getting onboard Zele for person-to-person payments we may see something similar to the push-based systems other countries have for making everyday transactions online someday.

[1] https://www.zellepay.com/

There is also Popmoney, which seems to have a few dozen more Credit Unions (though no BECU :c), they also have the big banks.

How does Zelle compare to Venmo on things like fraud protection and customer service? (I imagine simply not being owned by PayPal is a real good start.)

Unsure on how Venmo works, but I can explain Zelle / clearXchange some. (I've done this a lot on HN; it's nice to see someone else taking up the mantle on this thread!)

With Zelle, everything is done through your participating bank. So customer service is managed through your bank, and fraud is probably the same way.

You register with Zelle through your bank, which means your bank associates your email and / or phone number with your account within the Zelle system. After being associated, other people with a participating bank can send you money using that email or phone number. The money is instantly deducted from the sender's account, and is typically in the receiver's account within minutes or hours. Since you can only initiate payments through Zelle and not withdraws, fraud is likely managed the same as wires or bill pay.

Venmo has basically no protections, its much sketchier than PayPal. If the person who sent you money wants to reverse the transaction, there is nothing stopping them if its a person to person transaction.

A "proper" setup integrates Zelle directly with your bank, aside from clicking the "Send money with Zelle" button being the only hint a third-party is involved the transaction is initiated and settled directly within your banks online account management system. When done this way your bank of choice takes responsibility for fraud and customer service, with the Zelle network silently working behind the scenes to match email/mobile numbers up with destination accounts.

If you don't have a participating bank, things get a little more annoying since you can still use the service on its own by registering an account and tying your checking account to it - then the classic 3rd-party intermediary problem comes in. On that note though, Zelle is owned by big banks, Bank of America, Capital One, JPMorgan Chase, Wells Fargo, etc., so even as sleazy as the banks are I have a lot more faith in them not screwing it up (they don't like fraud, and they want you to keep your money with them - even as greedy as they are).

In all honesty, it's probably the best setup we have for Person-to-Person transfers in the US - as more banks and credit unions adopt it you'll hopefully just open the app for your bank of choice on your phone and initiate a payment, no third party trying to take a cut, you call the bank you (hopefully) already trust for support, and as a result you remove one more account tied to your banking information that can be compromised to steal your money.

Checks also include the payer's account number and bank routing number.

Uau, checks are still in use?!

Yep, checks are very common, its how most people pay rent, larger bills, etc in the USA. Credit Unions and Banks here offer "Bill Pay", which is just them mailing out a check on your behalf.

Cash is popular here too, many of the grocers I work with see 40% to 60% of their revenue in cash, though that market segment has discouraged check use (with the exception of Costco).

In most European countries checks are not trusted anymore by common business, given the usual fraud schemes, and in some banks you need to jump through a few hoops to get them (only crossed or pre-validated).

Digital payments have replaced the majority of checks.

It depends on the trust relationship -- stores will not accept checks from customers most of the time anymore, however as the parent said, rent and utilities are still often paid via check as there is an established relationship between the payer and payee.

National retailers mostly accept checks; they just treat them as debit card transactions though, scanning the numbers and printing the check for the customer, even handing it back to the customer after it is accepted.

I guess it isn't even limited to national retailers, the independent grocer here does it too, I guess because it is worthwhile for them to pay for it.

> I guess it isn't even limited to national retailers, the independent grocer here does it too, I guess because it is worthwhile for them to pay for it.

I mean, it really just saves the headache of taking the checks to the bank to be deposited. These machines scan the account information on the check, initiate an ACH debit and print the details on the check for the customer's reference.

We do something similar when patients remit payment to us by mailing a check, we have a third-party service that provides us with check scanners that do a similar job - but instead they scan the payment amount off the check and then initiate the ACH debit, they keep an image of the check and we destroy the original copies. Makes life a lot easier, especially since we have dozens of bank accounts these checks get posted to based on what legal entity the money needs to be allocated to.

Ah, gotta love Telecheck! Their blacklist is nigh impossible to get off of though for those that are on it, but it is optional to use the blacklist.

Usually one has to have some sort of 'check writing account' setup for this with the store or present a valid state-issued identification at the very least; When the store doesn't trust the purchaser they will often manually verify the purchase by calling the bank issuing the check, etc.

Eh? Bodegas won't, but Chexsystems is a thing, and a ton of places use it. I know people who do the bulk of their shopping with checks.

I've heard about the popularity of SWIFT. ACH is nowhere near as accessible, and it has fees depending on your bank or 3rd party ACH provider.

EMV was also much more rapidly and properly deployed as compared to the US, where some javacard applets just give a static code instead of a rolling code. We have literally hundreds of banks and credit unions, plus a few thousand payment processing resellers all fighting to hold back progress in the payments industry here.

If I'm not wrong SWIFT is already on the downward. Most services that had my SWIFT code requested me to update it to my IBAN code instead because SWIFT was being replaced.

Plus, inside the Eurozone+Friends, you can use SEPA payments across borders, and it is very cheap or even free.

From what I hear from my stateside friends, e.g. american utility companies often charge extra to pay by credit or debit card.

From a UK POV, this is completely alien to me as a concept.

> Yep, checks are very common, its how most people pay rent, larger bills, etc in the USA. Credit Unions and Banks here offer "Bill Pay", which is just them mailing out a check on your behalf.

Bill Pay services will try to convert the transaction to an ACH push if the service you are paying is set up for it, saves the bank money and lets them keep it in your account (earning them money) for longer. They will always fail back to cutting a check though.

Incredibly so, in USA. Even salaries are often paid with checks. In Europe (and elsewhere?) we've used online transfers for everything since 90s and USA is a really weird in this regard. That was almost the first thing that Internet was used for.

But also to point out usage is in massive decline - most people pay bills online, and purchase either with debit/credit or cash in person.

But yes, we never really got on the 'pay via bank' train like was done in much of europe in the postwar period, this has impact of how things transitioned to digital payments..

Conversely, we have much more established credit issuing infrastructure, it is not uncommon for even teenagers to have a credit card (not just debit)..

That sounds backward to me. Here (UK) my teenagers had their own debit cards -- not credit.

Perhaps Canada, as well. I paid my rent with a check deposited into the landlord's mailbox back in 2010.

Sadly some vendors/stores don't even accept Debit/Credit here in the USA.

Personally I find it rare - usually it's because of infrastructure (e.g., the farmer's stand doesn't have a credit terminal to process payments), or tax evasion (e.g., a restaurant gives you a discount for paying cash and it doesn't report that income).

I suppose it depends what you're buying. I've also had places who use Square tack on another 3% to cover their Credit Card fees each order as well :/

In NYC it is pretty common for a mom & pop bodega to either not accept debit and credit, or have a ridiculous limit on it. The pizza place I used to eat lunch almost everyday was $25 minimum for cards. The bodega next to my office was $15 minimum. The bodega under my apartment had $5 minimum, but they usually waved it for me if I was over $3 and didn't have cash.

Also, almost every Chinese restaurant or Ramen place had Cash Only signs.

A rewards credit card costs the business your shopping at 2%, or all the way up to 3.5% for certain cards. Then you get network fees (from Visa, Mastercard, etc) and another few cents and tenths of a percent tacked on by the businesses Payment Processor, and poof, there goes 3% of their income!

Its no wonder so many businesses don't take credit cards or debit.

Half of my clients still pay me via Check, even though QuickBooks invoices offer a Pay Now button (with options for Credit Card or ACH).

SEPA, guaranteed wire transfers over night and in November this year a payment under 15000€ will happen in 10s (ten seconds!). Not even Bitcoin can do that anymore with transactions taking more and more time (or fees).

If the US gov would put some weight behind it, the US could easily catch up.

>transactions taking more and more time

Bitcoin blocks will always be ~10 minutes, if you pay the fee to get into the next block, it'll be 10 minutes until your payment is confirmed, before that, all the nodes will announce it anyway so 0 conf is much faster.

Although bitcoin aims to generate a block every 10 minutes, in practice it usually takes longer than that. As I'm writing this, the last block[0] took 30 minutes to mine (sometimes it spikes to hours though[1]). Most importantly, a transaction needs[2] to be confirmed 4-6 times[3] meaning it should be 4-6 blocks in the past, so you're looking at a 1-2 hour transaction time.

On top of that, if you want to pay less than a certain transaction fee[4] (right now it's $1.62[5]) you probably won't get into the next block, see https://bitcoinfees.github.io. Recall that 1 block = 20 minutes and add that on top of the confirmation time.

[0] https://blockchain.info/block/0000000000000000005f0e639f66af... look at the Timestamps of this block and the Previous Block

[1] https://blockchain.info/charts/avg-confirmation-time?timespa...

[2] https://youtu.be/bBC-nXj3Ng4?t=18m41s

[3] https://bitcoin.org/bitcoin.pdf page 8

[4] https://en.bitcoin.it/wiki/Transaction_fees

[5] https://bitcoinfees.21.co/ (a satoshi is 0.00000001 BTC, 10,000 satoshi is 40 cents right now)

https://www.gwern.net/Bitcoin-is-Worse-is-Better

It's impossible to know how much to pay to get into the next block. If there's 10 payments waiting to get into the next block and there's only space for 9, there's no way they are all going into the block, regardless of their fees.

Users don't want to play a complicated expensive, impossible to win game in order to get their payments processed in a known timeframe.

I think your 10s transaction is about equivalent to an unconfirmed bitcoin transaction or credit card authorization. How often is the actual bank-to-bank transfer?

The fact that there is an arbitrary money amount shows it's more of a fraud cutoff than a real limit.

It's not the same. SEPA doesn't allow double-spends, unlike unconfirmed bitcoin transactions.

It depends. The banks nets the transactions a number of times during the day, normally using a central bank.

In the mean time the credit on your account is debited a 'float' account that accumulates any debt to other banks or third party payment system. Each fast payment system also typically have a reserved amount deposited as a guarantee in the event that a bank runs out of money.

It's a real cost to reserve money so there is a restriction on the sums.

Can the central banks speed up the settlement? Yes probably, but it's a lot of work to do.

SEPA guaranteed overnight iff funds are actually sent by the bank that receives the transfer request. I saw several banks (French, German, Luxemburgish) taking the order online and taking up to four business days (Sent Friday, arrived Thursday).

My SEPA transfer to a UK account always takes 2-3 days according to the service provider (can't vouch for that, maybe it's the provider who's slow but they are a small IT company so ...).

It varies by country if it's zero-day credit between banks (before a cutoff time) but i don't think it's legal within the EUR zone to have more than +1 days to credit.

(I think Finland is +0. Germany and France probably not, for various cultural differences ️)

One business day is required in the Payment Service Directive. You still occasionally see two days from German banks, only few of them actually can manage a zero day transfer. I operate a Dutch bank account solely for the purpose of making fast transfers to German accounts which works really well.

Once I tested that I was able to do a zero day transfer from a German bank to the Netherlands and then to another German bank. The same without the Dutch intermediary was a day slower.

I can get similar results with my current (UK) residential banking for anything up to 25kGBP.

Once upon a time, a friend of mine needed to repay me for a loan of about $1000. Since we both banked at HSBC, we figured walking into the bank together might be a simple way to do that. Nope. I had opened my hsbc account online, he had opened his account in person. Therefore we may as well have been two separate banks, apparently. After debating the sanity of this system with a bank manager (whose fault it clearly was not), I noticed a big sign on their ATM. It said "Now accepting cash deposits!". I asked the guy if there was a limit on the size of those deposits. There was. I can't remember how much, but it was in the neighborhood of five grand I think.

My friend took a withdrawal in cash. He handed me the money. I walked over to the machine and deposited it.

Makes sense if you look at it sideways.

There are also checks for that kind of thing.

Cheques don't typically clear immediately (to crudely attempt to prevent overdraft/double spending attempts) and can have a cost associated with their purchase.

Cheques are uncommon in Europe now-a-days.

They take five days to clear, however.

About ten years ago I moved from New York State to Washington State. I wanted a bank account that would follow me and Chase had not yet bought WaMu so I went to BoA in NY, opened an account, moved my money into it.

I arrived in Seattle and at the BoA branch there they said they didn't recognize BoA of New York. I couldn't deposit into the WA branch. I could write myself a check and it would be processed with the reduced speed of a check from an out of state bank.

Eh? How long ago was this? He couldn't write a cheque? He couldn't do a simple BACS/CHAPS/FPS transfer?

I'd have thought it simple without even walking into the bank...

It's not that long ago that doing a transfer even between different subsidiaries of the same UK bank still took 3 days.

Usually it'd be faster if it was within the same bank, but even then there was often no guarantee it'd be same day.

If you wanted a fast transfer, withdrawing and depositing the cash again was the fastest option for many combinations of accounts until just a few years ago.

Long ago? TODAY i can't pay my credit card balance, which is simply another account in my own bank, in less than 2 business days!

I log on their site, i see both accounts under my login, i just transfer money from one to another. Two. Business. Days.

It is probably the maximum amount of time they can get away with. Because there is an obvious benefit for the bank to delay. They can use your money as their security longer. They can charge you late fees. etc.

Well I didn't realise we were talking about speed, OP spoke of ease.

Cheques are not common in Europe now-a-days.

Bringing agpl into a banks systems would cause lots of fears. Adoption there is unlikely.

I read "vaguely tying them together" as "usually it works, sometimes not". This latter is my experience.

Electronic transfers enjoy a certain priority also, direct deposit, wire transfer, cash or check deposit at atm or branch are all different for different banks. In other words, it is a vague description when things clear. Some banks are better to their customers some are not so much. I live in the US.

An improvement is most certainly welcome and appreciated.

> One important take away for me was that this project aims to replace paypal, not bitcoin. Its target is to establish an open source payment infrastructure that is better than paying with credit cards, paypal, etc...

Well actually if you read the initial paper from Satoshi Nakamoto, bitcoin was invented as a replacement for paypal, not money :

> Commerce on the Internet has come to rely almost exclusively on financial institutions serving as trusted third parties to process electronic payment [1]

But bitcoin became really popular among the libertarian movement which hated the Fed and other central banks because they saw them as threats to freedom, and now people mostly think about bitcoin as a replacement for money.

[1] : https://bitcoin.org/bitcoin.pdf

Neat! I'm of the opinion that journalism, open source development, etc. would be much closer to self-sustaining if only users had the option of dropping a quarter in the hat — performing a one-click micropayment — when visiting a webpage. I'm not likely to subscribe to the NYT, but I'd gladly pay 25 cents after reading a great article if it didn't require entering CC info and bouncing between three different webpages. Unfortunately, this would require immense coordination between web standards, browsers, OS vendors, and banks to implement. Taler sounds like it's targeting a similar use case; I hope it sees wide adoption so that at least the idea could be tested at scale!

(Sadly, with most traffic coming from mobile these days, I'm not sure how well the wallet concept would play with iOS's closed ecosystem. Is it against the App Store guidelines to release a wallet-type extension? Sigh, I miss the days when hacky tech like Napster and BitTorrent would set the world on fire without anyone having to gatekeep...)

> but I'd gladly pay 25 cents after reading a great article if it didn't require entering CC info and bouncing between three different webpages... Unfortunately, this would require immense coordination between web standards, browsers, OS vendors, and banks to implement

If you're okay with customers choosing to make piece-work payments after consuming the material, the problem becomes much easier because micro-payments don't have to be instantaneous or binding.

All you need is a not-terribly-insecure ledger that the consumer has full read/write access to, and approves of every month (perhaps implicitly for bills under $X). This removes a lot of the security/trust/coordination issues -- an 90% solution is good enough if you just double-check with the consumer before doing anything actually important. Similar to the proposal here: https://techcrunch.com/2017/09/17/the-learned-helplessness-o... Micro-payments made by browser software (sim. SSNs) become spam-reducing mechanisms rather than trusted tokens.

Of course, this doesn't work if you want true micro-payments; i.e., where the consumer pays for access to the material, instead of making a piece-work contribution after having already consumed the material.

And american signature-less swipe transactions and now the contactless sub-30-quid payment thing in .uk seem to make it pretty clear that below a certain amount, consumers really don't give a shit about explicitly approving things.

TBH I don't think conscious micro-payments would work well. Something like spotify where you pay a flat fee for 'all you can eat' and proceeds are distributed via what you consume would be more practical. Medium is going towards that, but they still require you to press a like button to pay someone.

Most people will choose free with ads, but a significant amount will chose premium if it is compelling enough and that can represent a lot of money.

I don't know — I think many people have an allergic reaction to subscriptions of any kind. (I know I do.) Also, no one service could possibly encompass every blog, project, or publication that one might want to "tip", and people aren't going to subscribe to more than a handful of services at once. (Especially when things like Medium posts are already provided for free.)

Ultimately, I think people are very charitable when it comes to trivial transactions, but not when they're inconvenienced, and so I think it would be best for everyone to make that usability gap as tiny as possible. Even better if you could actually get something for your quarter: stats showing your tip% compared to peers or friends? Different site styles for high tippers? Personal stats tracking your interests over time? Monetized, unskeezy, free-as-in-freedom "likes": lots of possibilities here.

> I think many people have an allergic reaction to subscriptions of any kind.

Many people? I don't think it's a lot. Look at the subscriber numbers for Netflix. Or consider how many are willing to sign a contract for their cell phone. People will pay if the value is high.

I think the truth is most people don't care very much about the news. Time is scarce and people have a lot of choices about how to spend their time.

Yes, because Netflix and Spotify are amazing, amazing deals. The news? You can read that anywhere. Why would anyone subscribe to a service that offers them what they already have?

And yet, people are more than happy to tip a busker or leave a few bucks for the waiter or bartender. (Even outside the US, where tipping isn't expected.)

There's more than one thing at play here though.

Netflix and Spotify are giving people something they want without a lot of friction. I can get everything on those services for free other places but the experience is worse.

People are tipping buskers because they are sharing a physical location with that person. Put the video of the street performance on YouTube with a tip button and the dollars per view number is going to drop drastically.

Bartenders and waiters are tipped well because they are providing personal service.

The news fails because to be useful these days you need to consume news from a variety of sources. And you explicitly don't want to only get news from one media empire because then you're just getting whatever bias the CEO wants distributed.

I haven't heard a subscription news model which doesn't feel like its also trying to sunk-cost me into being less critical of the reporting.

I subscribe to Netflix, cable TV, MLBtv, Amazon Prime Video, and YouTube Red. The only newsy thing I subscribe to is New Yorker magazine.

Subscribing to multiple services isn't that unusual. There just has to be some value to it.

That's would be an awesome alternative to PayPal. I wish them luck.

Edit: btw, we really need a public and free payment system, why are we giving to paypal and credit cards this incredible privilege? Paying with digital cash should be as inexpensive as paying with real cash.

Because the credit/debit card systems are horribly insecure and we need to pay them so they can pay people who were scammed

Are you serious? Do you think "we should", really? Have you ever considered the problems and probably impossibility in achieving what you want?

TFA sort of proves that it is theoretically possible. But I believe that mints in the Taler model can take a flat fee to pay for server costs and so on.

On the other hand, it's not like handling physical cash is free, really.

Think about time spent counting, ensuring that there are no falsified notes, bringing excess notes to the bank and making sure there is change, etc.

I think it is quite feasible to create an electronic payment system that has comparable overhead.

Consider that there are existing electronic payment systems, e.g. electronic cash in europe, that already work and are quite secure (being smartcard-based)

I didn't say that. I've just said that it is not possible to have an "open" web payments standard. Someone has to maintain it and link it to the real world money.

The only way besides having centralized for-profit organizations maintaining the payments infrastructure is to use cryptocurrencies.

You have to pay for development and upkeep of the service. So it will never be as inexpensive.

Cash also has production and upkeep costs.

Taler is free in the sense of libre software, but it is not a free payment system. In their demo they added a small transaction fee. However, because it is not a crypto currency they can greatly reduce the transaction fees.

> Private When you pay with Taler, your identity does not have to be revealed. Just like payments in cash, nobody else can track how you spent your electronic money. However, you obtain a legally valid proof of payment.

> The exact details of the underlying contract that was signed between customer and merchant. However, this information would typically not include the identity of the customer.

IRS: The term “money laundering” refers to the activities and financial transactions that are undertaken specifically to hide the true source of the money.

https://www.irs.gov/compliance/criminal-investigation/overvi...

Am I missing something?

[I see by the downvotes this is an unpopular perspective, but I can't see how Taler's claims "Taler is not suitable for illegal activities" and "Taler prevents tax evasion and money laundering" are justifiable. It seems ideally designed for money laundering as defined by the IRS - "hiding the source of the money". Please explain it to me (ELI5)]

My guess? Tax evasion is a different problem from money laundering, and Taler seems to care about the former, not the latter. Whether this turns into a legal obstacle for adoption remains to be seen.

"Taler prevents tax evasion and money laundering." https://taler.net/en/merchants.html

I just don't see how they can claim that and claim to hide the payer identity.

How is this different from cash? Money laundering is about intent, not means of payment.

It is very similar to cash in that it is an anonymous means of payment. However, Taler tries to legitimize the transfer and explicitly claims "Taler is not suitable for illegal activities." https://taler.net/en/index.html

But by hiding the source of funds, it seems eminently suitable for illegal activities. Putting a (partially) legitimate business as front to process illegal profits is a textbook money laundering structure.

> Putting a (partially) legitimate business as front to process illegal profits is a textbook money laundering structure.

Yes, but how does Taler equate to doing this? Again, how is it any different than cash? Money Laundering is about intent, not instrument.

The primary uses of Taler will likely be legitimate - I never equated it to money laundering, just expressed my doubt for the assertion that it cannot be used for money laundering.

I can't tell you how Taler differs from cash with respect to money laundering. That's why I asked the question.

Can you tell me why Taler cannot be used for money laundering?

I think you're reading the "Taler is not suitable for illegal activities." wrong. They're not saying you can't (as like you said, of course you could), they're saying you shouldn't

I guess what they're saying is, "we don't want you to doing anything illegal on our service, but we can't really stop you easily if you were to do that".

"Taler prevents tax evasion and money laundering." https://taler.net/en/merchants.html

Perhaps it is private between transacting parties but still trackable per account (in line with know your customer laws).

> Can you tell me why Taler cannot be used for money laundering?

Can you tell me why cash can't be? My point is that money laundering is a crime of intention, not instrument.

I'm sorry, I misunderstood and thought that your point had something to do with my ancestral comment.

Money laundering is a crime of intent ("transactions... undertaken specifically to hide the true source of the money.") Cash is one mechanism that can be used for money laundering.

(Are we good? Violent agreement so far, right?)

The Taler web site claims "Taler prevents tax evasion and money laundering." https://taler.net/en/merchants.html

How?

First of all, cash is not "ideally designed for money laundering".

Second of all, Taler improves upon cash for combatting money laundering. Money laundering by its very nature creates a dishonest paper trail to hide the real source. When laundering cash, the paper trail and the cash are separate, and because the serial numbers of bills are rarely recorded, there is no proof that the paperwork presented to regulators explaining the cash income actually resulted in said cash income.

For example, imagine a cash-rich business being audited by the IRS. The business shows daily bank deposits of $10,000 per day, with approximately 100 daily receipts for 100 daily transactions worth about $100 each. The money launderers were actually trying to hide large single transfers of $300,000 monthly that were being buffered into daily deposits, but because the dishonest paperwork shows 3,000 separate transactions, that's what's "real" until proven otherwise.

With Taler, the cash transfer and the contract are cryptographically bound. Which means that money launderers can't just draw up some fake paperwork after one large transaction, they need to make 3,000 real transactions actually happen, from real Taler wallets, which themselves must appear reputable since all Taler wallets are independently auditable (what good is having a money laundering business to hide your dishonest Taler wallets if your dishonest Taler wallets are in danger of being randomly audited themselves at some other point in time?) This drastically increases the cost of establishing an apparently honest paper trail, which in turn makes money laundering using Taler less appealing.

I seem to remember that the Cannabis industry in the US was having difficulty establishing banking relations with reputable firms. They could be a good start for this project to see usage given the open source philosophy that might resonate with some dispensaries and the friction with traditional payment processors.

Users of those commerces wouldn't be too hostile to the alternative payment option I'd imagine, and maybe with the combination of INRIA & GNU pedigree, backend relations would be less risky for the "real finance" interface.

In EU it is really trivial and very cheap, fast and safe to transfer payments between individual bank accounts and business bank accounts. It would be really hard to further simplify it.

For me, the allure of bitcoin and any other newer payment transfers systems is to be able to get rid of the "In $somelocation" clause. I want internet money to be as universal as email.

Please take a look at the demo they have set up.

They show an example micro payment system as browser plugin, where you pay for reading articles.

I am excited at the prospect of paying for articles without spying adware and coin-mining javascript.

Back to your point: Even in the EU paypal has still users, because it is still easier to use than SEPA - transactions.

I like that prospect too, but many media companies have proven to be scummy enough with their malvertising and tracking that they're likely to add micropayments and keep their spy/adware. A recent example is German television switching to a new digital terrestrial transmission format, whereupon the commercial (ad-ridden) channels now require a yearly subscription to watch, but have kept the ads. The enabling technology was the possibility to encrypt said channels in the new format, and they, of course, jumped on it, pocketing the extra money but keeping all their ads.

10x times better than the US? Maybe yes - but not cheap and fast enough for micropayments.

Why do micropayments need to be fast? Seems like exactly the kind of thing that could be held on to and batch processed.

I mean, I have used Giropay and SOFORT, both of which have been always confirmed instantly with relatively low fees (about 2% for my purchases, tho I'm not certain how that scales for sub-10€ purchases).

Success of payment system is extremely dependent on the network effect.

The value for users comes from the ubiquity of the payment system. Nobody wants to transfer money into the wallet for infrequent or one time buys.

They seem to have the right idea and might get the implementation just right, but if they can't get the momentum, it's all for nothing. I really hope this takes off.

That's what I don't get. Publishers have all the incentive to get behind a solution like this, but somehow I never see a "send a tip" or Flattr button anywhere (maybe I'm not looking?).

I just wish a bunch of well-known publishers would get together and agree on _one_ solution that they'd all add to their site.

It's just a matter of adding a widget on their article pages. They already have large readership. It could kick things off and maybe get over that network effect cold-start issue.

They could add a nice bounty like "make a first payment and don't see ads on our site for a month". Or similar.

A solution like this could be a good candidate because it's neutral. With Flattr or similar you're encouraging that particular middleman. However I think to inspire confidence in publishers they'd need to prove that this solution will have strong support for years to come (which is always unsure with open source projects).

EDIT: To be clear I'm mostly thinking of micropayments to news sites. The strategy is less clear when talking about electronic commerce in general.

"Once the wallet is charged, payments on websites take only one click, are never falsely rejected by fraud detection and do not pose any risk of phishing or identity theft."

Pretty bold claims. Anyone can give ELI5 how it works? Merchant needs a bank account, customer needs to load funds to the wallet... Sounds quite traditional to me. Either these people are geniuses or they don't know what they are doing.

GNU are quite famous for their graphic design skills.

I hate to admit it, but I do think their ideas and projects would gain wider adoption if they packaged them in a more visually appealing way. I do like the minimalist style, however the logo could use some work.

PS: I am a programmer, not a designer.

Many GNU hackers simply don't have those skills. I don't. Someone should offer their time/skills.

Some FOSS projects put a lot of work into UX, which is really great!

Compare two implementations of OStatus-compatible social networks: GNU-Social and Mastodon.

GNU-social: https://upload.wikimedia.org/wikipedia/commons/8/8a/GNU_soci...

Mastodon: http://wikigeek.co/wp-content/uploads/2017/01/mastodon.png

Mastodon is a lot better for the average joe, but the important piece from this is that its predecessors, GNU-social and StatusNet, helped show that this sort of application can work--and I'm sure the development of Mastodon was easier because many of its challenges were solved already by other free software projects.

The neat thing about all these social networks is that they're federated, which means a user on mastodon can send a message to someone using gnu-social or pump.io and vice versa.

That's the magic of free software!

The https://github.com/hannesmannerheim/neo-quitter theme is quite popular for GNU social, on sites like https://quitter.se

Hey, that's pretty neat!

Open source software needs to find a way to fund designers. There's not nearly as many "hobbyist UX designers" as there are programmers, unfortunately.

Ironically, this very project might help out with funding all sorts of developers.

There are a lot of hobbyist UX designers out there, judging by the amount of unsolicited redesigns and blogs about what could have been if...

The problem is that when you do something as a hobby, you usually want full control. This puts the developer/designer duo at odds. I think this is why we have many non-functional pretty designs and functional ugly applications/sites.

How does this compare to the upcoming Payment Request API?

Once the Payment Handler API [1] is completed and implemented in browsers and browser extensions can register themselves as payment handlers, GNU Taler could be used through the Payment Request API.

[1] https://www.w3.org/TR/payment-handler/

As I understand it, PR-API is a glue for current payment methods. So your browser would simply contain a hash that is connected to a payment method (i.e. a credit card). It's effectively like Stripe built in to your browser. Before you use it you need to put in a payment method.

GNU Taler, however, is a payment method. It federates with other financial institutions to allow you to make online purchases without using credit cards. In the SHA2017 demo, they keep saying "it's like cash" and that's very close to accurate. The one difference is that payment recipients are forced to keep a paper trail of what money they've received (not from whom, but just that it was received). So in that sense it's almost better than cash in that it's more transparent for tax purposes, but maintains the anonymity of cash for the buyer.

  with extremely low transaction costs

Does anyone here happen to know what the pricing structure is?

The system is open, thus payment processors are in competition to provide the lowest fees possible. This is, in my opinion, one of the important reasons why a payment system needs to be open: competition forces fees to be as low as realistically possible.

This sounds a lot like the Ripple network. Or is there a substantial technological difference?

Agreed. The architecture diagram [1] describes a flow that is identical to how a user on Ripple buys IOUs from a gateway, trades or spends them with other users, and how a receiver redeems them with the gateway for the original currency.

Ripple captured my attention for years, but its backing company has taken it in a direction of commercial adoption which is good for them, but antithetical to the applications I was excited about.

If GNU Taler is going to revive that vision, then that's really exciting!

[1] https://taler.net/en/architecture.html

I think this suffers from the same issue that I have with a lot of crypto/magic currencies.

It isn't immediately clear to me why I'd want to use it. They could advertise it as: 'Pay online, without Paypal!' and a big 'Sign Up' button, and it would do more for me than anything their current page does.

Having GNU in the name doesn't really help either, as I associate it with nobody adopting the thing whatsoever.

> They could advertise it as: 'Pay online, without Paypal!' and a big 'Sign Up' button, and it would do more for me than anything their current page does.

Taler requires banks to cooperate, so "signing up" doesn't make any sense in this context. Taler support would have to be offered by the bank; a user doesn't sign up for Taler --- they just need to use the Taler wallet extension.

> Having GNU in the name doesn't really help either, as I associate it with nobody adopting the thing whatsoever.

Now you're just insulting. (I'm also a GNU hacker.)

>> Having GNU in the name doesn't really help either, as I associate it with nobody adopting the thing whatsoever. >Now you're just insulting. (I'm also a GNU hacker.)

I see how it can be taken that way. But I can't really help my impression of the name.

Is this an official GNU project?

Yes.

https://www.gnu.org/software/

I think the official terminology is that there is only one GNU project with many GNU packages, of which Taler, gcc, and Gnome are: just a package.

That terminology doesn't seem to be used very consistently, even by project maintainers:

https://duckduckgo.com/?q="is+an+official+gnu+project"

Yeah, typically "part of the GNU project"

It varies, even internally. Officially, yes, it's "package". "GNU Project" (capital "P") refers to GNU itself. "Project foo is part of GNU" is also common. "A GNU project" isn't uncommon informally. On Savannah, they're "projects". The maintainers guide uses both---you can tell what was written early on, possibly by rms, vs. changes that happened later on.

In the future, perhaps it would be useful to put what the thing actually is in the title. Just "GNU Taler" doesn't tell me anything about what something is, other than it's a GNU project.

Usually if you editorialize the title of the link submission the mods will reset it to the title of the page.

I can understand why, but would saying "GNU Taler is blah blah" really editorializing? As long as you aren't embellishing?

Yes, the mods have been known to reset it anyway. It’s too bad, really.

I would hope "GNU Taler (one-click cash payments)" would've survived because it took a subtitle from the page rather than editorialising.

I sincerely doubt you can draw an accurate line for what's editorialising and what isn't though and I think I'd prefer the mods to continue to err on the side of "nope" simply because I think the failure modes are less annoying overall. But it's always going to be a trade-off.

I think a blanket "nope" stance is too harsh, and leaves no room for nuance. It also assumes that the mods are not intelligent people.

The nice thing about bureaucracy is that it doesn't leave any room for debate. Not everything needs to be debated. If you're on HN, you're here to waste time. The click is free.

To be clear, "err on the side of" is still a long way from "blanket", and I did say it was a trade-off. Sorry if my choice of idiom was confusing.

I guess we should be thankful it wasn't just a link to a GitHub project with a skeleton README.

"the wallet's balance must be charged in the desired currency"

I wonder if this was written by someone from Central Europe. This use of the word charge is no good in English because it's a heteronym - it can mean to add money or take it!

What would be the correct word for only adding?

Deposit to put money, withdraw to take it.

"Deposit" or in Ireland "lodge" or make a lodgement.

"Add" itself is acceptable: "the desired currency must be added to the wallet's balance"

English has many heteronyms. I think that property is horrible and should be avoided whenever possible. Why is it accepted by english-speaking people?

I think it's common to use "(re)charge" when adding credits to something like a Steam wallet. This feels like that type of transaction.

Next: GNU Privacy Guard implements the Clipper Chip algorithm.

Why did GNU accept this state surveillance system as a project?

Reply on news.ycombinator.com