* Re: [bitcoin-dev] BIP Proposal: Compact Client Side Filtering for Light Clients
2017-06-01 19:01 [bitcoin-dev] BIP Proposal: Compact Client Side Filtering for Light Clients Olaoluwa Osuntokun
@ 2017-06-01 21:00 ` Eric Lombrozo
2017-06-01 21:33 ` Matt Corallo
` (5 subsequent siblings)
6 siblings, 0 replies; 39+ messages in thread
From: Eric Lombrozo @ 2017-06-01 21:00 UTC (permalink / raw)
To: Olaoluwa Osuntokun; +Cc: Arnoud Kouwenhoven - Pukaki Corp via bitcoin-dev
[-- Attachment #1: Type: text/plain, Size: 5526 bytes --]
Thanks for sending this proposal! I look forward to having a great
discussion around this.
- Eric
On Thursday, June 1, 2017, Olaoluwa Osuntokun via bitcoin-dev <
bitcoin-dev@lists•linuxfoundation.org> wrote:
> Hi y'all,
>
> Alex Akselrod and I would like to propose a new light client BIP for
> consideration:
> * https://github.com/Roasbeef/bips/blob/master/gcs_light_
> client.mediawiki
>
> This BIP proposal describes a concrete specification (along with a
> reference implementations[1][2][3]) for the much discussed client-side
> filtering reversal of BIP-37. The precise details are described in the
> BIP, but as a summary: we've implemented a new light-client mode that uses
> client-side filtering based off of Golomb-Rice coded sets. Full-nodes
> maintain an additional index of the chain, and serve this compact filter
> (the index) to light clients which request them. Light clients then fetch
> these filters, query the locally and _maybe_ fetch the block if a relevant
> item matches. The cool part is that blocks can be fetched from _any_
> source, once the light client deems it necessary. Our primary motivation
> for this work was enabling a light client mode for lnd[4] in order to
> support a more light-weight back end paving the way for the usage of
> Lightning on mobile phones and other devices. We've integrated neutrino
> as a back end for lnd, and will be making the updated code public very
> soon.
>
> One specific area we'd like feedback on is the parameter selection. Unlike
> BIP-37 which allows clients to dynamically tune their false positive rate,
> our proposal uses a _fixed_ false-positive. Within the document, it's
> currently specified as P = 1/2^20. We've done a bit of analysis and
> optimization attempting to optimize the following sum:
> filter_download_bandwidth + expected_block_false_positive_bandwidth. Alex
> has made a JS calculator that allows y'all to explore the affect of
> tweaking the false positive rate in addition to the following variables:
> the number of items the wallet is scanning for, the size of the blocks,
> number of blocks fetched, and the size of the filters themselves. The
> calculator calculates the expected bandwidth utilization using the CDF of
> the Geometric Distribution. The calculator can be found here:
> https://aakselrod.github.io/gcs_calc.html. Alex also has an empirical
> script he's been running on actual data, and the results seem to match up
> rather nicely.
>
> We we're excited to see that Karl Johan Alm (kallewoof) has done some
> (rather extensive!) analysis of his own, focusing on a distinct encoding
> type [5]. I haven't had the time yet to dig into his report yet, but I
> think I've read enough to extract the key difference in our encodings: his
> filters use a binomial encoding _directly_ on the filter contents, will we
> instead create a Golomb-Coded set with the contents being _hashes_ (we use
> siphash) of the filter items.
>
> Using a fixed fp=20, I have some stats detailing the total index size, as
> well as averages for both mainnet and testnet. For mainnet, using the
> filter contents as currently described in the BIP (basic + extended), the
> total size of the index comes out to 6.9GB. The break down is as follows:
>
> * total size: 6976047156
> * total avg: 14997.220622758816
> * total median: 3801
> * total max: 79155
> * regular size: 3117183743
> * regular avg: 6701.372750217131
> * regular median: 1734
> * regular max: 67533
> * extended size: 3858863413
> * extended avg: 8295.847872541684
> * extended median: 2041
> * extended max: 52508
>
> In order to consider the average+median filter sizes in a world worth
> larger blocks, I also ran the index for testnet:
>
> * total size: 2753238530
> * total avg: 5918.95736054141
> * total median: 60202
> * total max: 74983
> * regular size: 1165148878
> * regular avg: 2504.856172982827
> * regular median: 24812
> * regular max: 64554
> * extended size: 1588089652
> * extended avg: 3414.1011875585823
> * extended median: 35260
> * extended max: 41731
>
> Finally, here are the testnet stats which take into account the increase
> in the maximum filter size due to segwit's block-size increase. The max
> filter sizes are a bit larger due to some of the habitual blocks I
> created last year when testing segwit (transactions with 30k inputs, 30k
> outputs, etc).
>
> * total size: 585087597
> * total avg: 520.8839608674402
> * total median: 20
> * total max: 164598
> * regular size: 299325029
> * regular avg: 266.4790836307566
> * regular median: 13
> * regular max: 164583
> * extended size: 285762568
> * extended avg: 254.4048772366836
> * extended median: 7
> * extended max: 127631
>
> For those that are interested in the raw data, I've uploaded a CSV file
> of raw data for each block (mainnet + testnet), which can be found here:
> * mainnet: (14MB): https://www.dropbox.com/s/
> 4yk2u8dj06njbuv/mainnet-gcs-stats.csv?dl=0
> * testnet: (25MB): https://www.dropbox.com/s/
> w7dmmcbocnmjfbo/gcs-stats-testnet.csv?dl=0
>
>
> We look forward to getting feedback from all of y'all!
>
> -- Laolu
>
>
> [1]: https://github.com/lightninglabs/neutrino
> [2]: https://github.com/Roasbeef/btcd/tree/segwit-cbf
> [3]: https://github.com/Roasbeef/btcutil/tree/gcs/gcs
> [4]: https://github.com/lightningnetwork/lnd/
>
> -- Laolu
>
>
[-- Attachment #2: Type: text/html, Size: 7707 bytes --]
^ permalink raw reply [flat|nested] 39+ messages in thread* Re: [bitcoin-dev] BIP Proposal: Compact Client Side Filtering for Light Clients
2017-06-01 19:01 [bitcoin-dev] BIP Proposal: Compact Client Side Filtering for Light Clients Olaoluwa Osuntokun
2017-06-01 21:00 ` Eric Lombrozo
@ 2017-06-01 21:33 ` Matt Corallo
2017-06-01 22:10 ` Olaoluwa Osuntokun
2017-06-02 4:49 ` Olaoluwa Osuntokun
` (4 subsequent siblings)
6 siblings, 1 reply; 39+ messages in thread
From: Matt Corallo @ 2017-06-01 21:33 UTC (permalink / raw)
To: Olaoluwa Osuntokun, Olaoluwa Osuntokun via bitcoin-dev,
Arnoud Kouwenhoven - Pukaki Corp via bitcoin-dev
Quick comment before I finish reading it completely, looks like you have no way to match the input prevouts being spent, which is rather nice from a "watch for this output being spent" pov.
On June 1, 2017 3:01:14 PM EDT, Olaoluwa Osuntokun via bitcoin-dev <bitcoin-dev@lists•linuxfoundation.org> wrote:
>Hi y'all,
>
>Alex Akselrod and I would like to propose a new light client BIP for
>consideration:
>*
>https://github.com/Roasbeef/bips/blob/master/gcs_light_client.mediawiki
>
>This BIP proposal describes a concrete specification (along with a
>reference implementations[1][2][3]) for the much discussed client-side
>filtering reversal of BIP-37. The precise details are described in the
>BIP, but as a summary: we've implemented a new light-client mode that
>uses
>client-side filtering based off of Golomb-Rice coded sets. Full-nodes
>maintain an additional index of the chain, and serve this compact
>filter
>(the index) to light clients which request them. Light clients then
>fetch
>these filters, query the locally and _maybe_ fetch the block if a
>relevant
>item matches. The cool part is that blocks can be fetched from _any_
>source, once the light client deems it necessary. Our primary
>motivation
>for this work was enabling a light client mode for lnd[4] in order to
>support a more light-weight back end paving the way for the usage of
>Lightning on mobile phones and other devices. We've integrated neutrino
>as a back end for lnd, and will be making the updated code public very
>soon.
>
>One specific area we'd like feedback on is the parameter selection.
>Unlike
>BIP-37 which allows clients to dynamically tune their false positive
>rate,
>our proposal uses a _fixed_ false-positive. Within the document, it's
>currently specified as P = 1/2^20. We've done a bit of analysis and
>optimization attempting to optimize the following sum:
>filter_download_bandwidth + expected_block_false_positive_bandwidth.
>Alex
>has made a JS calculator that allows y'all to explore the affect of
>tweaking the false positive rate in addition to the following
>variables:
>the number of items the wallet is scanning for, the size of the blocks,
>number of blocks fetched, and the size of the filters themselves. The
>calculator calculates the expected bandwidth utilization using the CDF
>of
>the Geometric Distribution. The calculator can be found here:
>https://aakselrod.github.io/gcs_calc.html. Alex also has an empirical
>script he's been running on actual data, and the results seem to match
>up
>rather nicely.
>
>We we're excited to see that Karl Johan Alm (kallewoof) has done some
>(rather extensive!) analysis of his own, focusing on a distinct
>encoding
>type [5]. I haven't had the time yet to dig into his report yet, but I
>think I've read enough to extract the key difference in our encodings:
>his
>filters use a binomial encoding _directly_ on the filter contents, will
>we
>instead create a Golomb-Coded set with the contents being _hashes_ (we
>use
>siphash) of the filter items.
>
>Using a fixed fp=20, I have some stats detailing the total index size,
>as
>well as averages for both mainnet and testnet. For mainnet, using the
>filter contents as currently described in the BIP (basic + extended),
>the
>total size of the index comes out to 6.9GB. The break down is as
>follows:
>
> * total size: 6976047156
> * total avg: 14997.220622758816
> * total median: 3801
> * total max: 79155
> * regular size: 3117183743
> * regular avg: 6701.372750217131
> * regular median: 1734
> * regular max: 67533
> * extended size: 3858863413
> * extended avg: 8295.847872541684
> * extended median: 2041
> * extended max: 52508
>
>In order to consider the average+median filter sizes in a world worth
>larger blocks, I also ran the index for testnet:
>
> * total size: 2753238530
> * total avg: 5918.95736054141
> * total median: 60202
> * total max: 74983
> * regular size: 1165148878
> * regular avg: 2504.856172982827
> * regular median: 24812
> * regular max: 64554
> * extended size: 1588089652
> * extended avg: 3414.1011875585823
> * extended median: 35260
> * extended max: 41731
>
>Finally, here are the testnet stats which take into account the
>increase
>in the maximum filter size due to segwit's block-size increase. The max
>filter sizes are a bit larger due to some of the habitual blocks I
>created last year when testing segwit (transactions with 30k inputs,
>30k
>outputs, etc).
>
> * total size: 585087597
> * total avg: 520.8839608674402
> * total median: 20
> * total max: 164598
> * regular size: 299325029
> * regular avg: 266.4790836307566
> * regular median: 13
> * regular max: 164583
> * extended size: 285762568
> * extended avg: 254.4048772366836
> * extended median: 7
> * extended max: 127631
>
>For those that are interested in the raw data, I've uploaded a CSV file
>of raw data for each block (mainnet + testnet), which can be found
>here:
> * mainnet: (14MB):
>https://www.dropbox.com/s/4yk2u8dj06njbuv/mainnet-gcs-stats.csv?dl=0
> * testnet: (25MB):
>https://www.dropbox.com/s/w7dmmcbocnmjfbo/gcs-stats-testnet.csv?dl=0
>
>
>We look forward to getting feedback from all of y'all!
>
>-- Laolu
>
>
>[1]: https://github.com/lightninglabs/neutrino
>[2]: https://github.com/Roasbeef/btcd/tree/segwit-cbf
>[3]: https://github.com/Roasbeef/btcutil/tree/gcs/gcs
>[4]: https://github.com/lightningnetwork/lnd/
>
>-- Laolu
^ permalink raw reply [flat|nested] 39+ messages in thread* Re: [bitcoin-dev] BIP Proposal: Compact Client Side Filtering for Light Clients
2017-06-01 21:33 ` Matt Corallo
@ 2017-06-01 22:10 ` Olaoluwa Osuntokun
2017-06-02 2:15 ` Chris
0 siblings, 1 reply; 39+ messages in thread
From: Olaoluwa Osuntokun @ 2017-06-01 22:10 UTC (permalink / raw)
To: Matt Corallo, Arnoud Kouwenhoven - Pukaki Corp via bitcoin-dev
[-- Attachment #1: Type: text/plain, Size: 7552 bytes --]
Eric wrote:
> Thanks for sending this proposal! I look forward to having a great
> discussion around this.
Thanks Eric! We really appreciated the early feedback you gave on the
initial design.
One aspect which isn't in this BIP draft is direct support for unconfirmed
transactions. I consider such a feature an important UX feature for mobile
phones, and something which I've personally seen as an important
UX-experience when on-boarding new users to Bitcoin. This was brought up
in the original "bfd" mailing list chain [1]. Possible solutions are: a
new beefier INV message which contains enough information to be able to
identify relevant outputs created in a transaction, or a "streaming" p2p
extension that allows light clients to receive notifications of mempool
inclusion based on only (pkScript, amount) pairs.
Matt wrote:
> looks like you have no way to match the input prevouts being spent, which
> is rather nice from a "watch for this output being spent" pov.
Perhaps we didn't make this clear enough, but it _is_ indeed possible to
watch an output for spentness. Or maybe you mean matching on the
_script_ being spent?
From the BIP draft:
> for each transaction, normal filters contain:
> * The outpoints of each input within a transaction.
> ...
Within the integration for lnd, we specifically use this feature to be
able to watch for when channels have been closed within the network graph,
or channels _directly_ under our control have been spent (either
unilateral channel closure, or a revocation beach).
-- Laolu
[1]:
https://lists.linuxfoundation.org/pipermail/bitcoin-dev/2017-January/013397.html
On Thu, Jun 1, 2017 at 2:33 PM Matt Corallo <lf-lists@mattcorallo•com>
wrote:
> Quick comment before I finish reading it completely, looks like you have
> no way to match the input prevouts being spent, which is rather nice from a
> "watch for this output being spent" pov.
>
> On June 1, 2017 3:01:14 PM EDT, Olaoluwa Osuntokun via bitcoin-dev <
> bitcoin-dev@lists•linuxfoundation.org> wrote:
> >Hi y'all,
> >
> >Alex Akselrod and I would like to propose a new light client BIP for
> >consideration:
> >*
> >https://github.com/Roasbeef/bips/blob/master/gcs_light_client.mediawiki
> >
> >This BIP proposal describes a concrete specification (along with a
> >reference implementations[1][2][3]) for the much discussed client-side
> >filtering reversal of BIP-37. The precise details are described in the
> >BIP, but as a summary: we've implemented a new light-client mode that
> >uses
> >client-side filtering based off of Golomb-Rice coded sets. Full-nodes
> >maintain an additional index of the chain, and serve this compact
> >filter
> >(the index) to light clients which request them. Light clients then
> >fetch
> >these filters, query the locally and _maybe_ fetch the block if a
> >relevant
> >item matches. The cool part is that blocks can be fetched from _any_
> >source, once the light client deems it necessary. Our primary
> >motivation
> >for this work was enabling a light client mode for lnd[4] in order to
> >support a more light-weight back end paving the way for the usage of
> >Lightning on mobile phones and other devices. We've integrated neutrino
> >as a back end for lnd, and will be making the updated code public very
> >soon.
> >
> >One specific area we'd like feedback on is the parameter selection.
> >Unlike
> >BIP-37 which allows clients to dynamically tune their false positive
> >rate,
> >our proposal uses a _fixed_ false-positive. Within the document, it's
> >currently specified as P = 1/2^20. We've done a bit of analysis and
> >optimization attempting to optimize the following sum:
> >filter_download_bandwidth + expected_block_false_positive_bandwidth.
> >Alex
> >has made a JS calculator that allows y'all to explore the affect of
> >tweaking the false positive rate in addition to the following
> >variables:
> >the number of items the wallet is scanning for, the size of the blocks,
> >number of blocks fetched, and the size of the filters themselves. The
> >calculator calculates the expected bandwidth utilization using the CDF
> >of
> >the Geometric Distribution. The calculator can be found here:
> >https://aakselrod.github.io/gcs_calc.html. Alex also has an empirical
> >script he's been running on actual data, and the results seem to match
> >up
> >rather nicely.
> >
> >We we're excited to see that Karl Johan Alm (kallewoof) has done some
> >(rather extensive!) analysis of his own, focusing on a distinct
> >encoding
> >type [5]. I haven't had the time yet to dig into his report yet, but I
> >think I've read enough to extract the key difference in our encodings:
> >his
> >filters use a binomial encoding _directly_ on the filter contents, will
> >we
> >instead create a Golomb-Coded set with the contents being _hashes_ (we
> >use
> >siphash) of the filter items.
> >
> >Using a fixed fp=20, I have some stats detailing the total index size,
> >as
> >well as averages for both mainnet and testnet. For mainnet, using the
> >filter contents as currently described in the BIP (basic + extended),
> >the
> >total size of the index comes out to 6.9GB. The break down is as
> >follows:
> >
> > * total size: 6976047156
> > * total avg: 14997.220622758816
> > * total median: 3801
> > * total max: 79155
> > * regular size: 3117183743
> > * regular avg: 6701.372750217131
> > * regular median: 1734
> > * regular max: 67533
> > * extended size: 3858863413 <(385)%20886-3413>
> > * extended avg: 8295.847872541684
> > * extended median: 2041
> > * extended max: 52508
> >
> >In order to consider the average+median filter sizes in a world worth
> >larger blocks, I also ran the index for testnet:
> >
> > * total size: 2753238530
> > * total avg: 5918.95736054141
> > * total median: 60202
> > * total max: 74983
> > * regular size: 1165148878
> > * regular avg: 2504.856172982827
> > * regular median: 24812
> > * regular max: 64554
> > * extended size: 1588089652
> > * extended avg: 3414.1011875585823
> > * extended median: 35260
> > * extended max: 41731
> >
> >Finally, here are the testnet stats which take into account the
> >increase
> >in the maximum filter size due to segwit's block-size increase. The max
> >filter sizes are a bit larger due to some of the habitual blocks I
> >created last year when testing segwit (transactions with 30k inputs,
> >30k
> >outputs, etc).
> >
> > * total size: 585087597
> > * total avg: 520.8839608674402
> > * total median: 20
> > * total max: 164598
> > * regular size: 299325029
> > * regular avg: 266.4790836307566
> > * regular median: 13
> > * regular max: 164583
> > * extended size: 285762568
> > * extended avg: 254.4048772366836
> > * extended median: 7
> > * extended max: 127631
> >
> >For those that are interested in the raw data, I've uploaded a CSV file
> >of raw data for each block (mainnet + testnet), which can be found
> >here:
> > * mainnet: (14MB):
> >https://www.dropbox.com/s/4yk2u8dj06njbuv/mainnet-gcs-stats.csv?dl=0
> > * testnet: (25MB):
> >https://www.dropbox.com/s/w7dmmcbocnmjfbo/gcs-stats-testnet.csv?dl=0
> >
> >
> >We look forward to getting feedback from all of y'all!
> >
> >-- Laolu
> >
> >
> >[1]: https://github.com/lightninglabs/neutrino
> >[2]: https://github.com/Roasbeef/btcd/tree/segwit-cbf
> >[3]: https://github.com/Roasbeef/btcutil/tree/gcs/gcs
> >[4]: https://github.com/lightningnetwork/lnd/
> >
> >-- Laolu
>
[-- Attachment #2: Type: text/html, Size: 10365 bytes --]
^ permalink raw reply [flat|nested] 39+ messages in thread
* Re: [bitcoin-dev] BIP Proposal: Compact Client Side Filtering for Light Clients
2017-06-01 22:10 ` Olaoluwa Osuntokun
@ 2017-06-02 2:15 ` Chris
2017-06-02 2:28 ` Gregory Maxwell
0 siblings, 1 reply; 39+ messages in thread
From: Chris @ 2017-06-02 2:15 UTC (permalink / raw)
To: bitcoin-dev
On 06/01/2017 06:10 PM, Olaoluwa Osuntokun via bitcoin-dev wrote:
> One aspect which isn't in this BIP draft is direct support for unconfirmed
> transactions. I consider such a feature an important UX feature for mobile
> phones, and something which I've personally seen as an important
> UX-experience when on-boarding new users to Bitcoin.
Totally agree. My first thought is maybe you could keep bip37 filtering
as optional for unconfirmed transactions. Since you're only interested
in incoming transactions in this case you can create one big filter with
all your wallet's addresses and reuse that filter. The bip37 privacy
issues mainly come up when trying to get the filter to match both
incoming and outgoing transactions, which is not needed in this case.
Otoh, if you download the block from the same peer that you gave a bip37
filter then they could probably test the txs in the block against both
filters. :/
^ permalink raw reply [flat|nested] 39+ messages in thread
* Re: [bitcoin-dev] BIP Proposal: Compact Client Side Filtering for Light Clients
2017-06-02 2:15 ` Chris
@ 2017-06-02 2:28 ` Gregory Maxwell
2017-06-02 3:35 ` Alex Akselrod
2017-06-02 16:07 ` Chris Pacia
0 siblings, 2 replies; 39+ messages in thread
From: Gregory Maxwell @ 2017-06-02 2:28 UTC (permalink / raw)
To: Chris; +Cc: Bitcoin Dev
On Fri, Jun 2, 2017 at 2:15 AM, Chris via bitcoin-dev
<bitcoin-dev@lists•linuxfoundation.org> wrote:
> On 06/01/2017 06:10 PM, Olaoluwa Osuntokun via bitcoin-dev wrote:
>
>> One aspect which isn't in this BIP draft is direct support for unconfirmed
>> transactions. I consider such a feature an important UX feature for mobile
>> phones, and something which I've personally seen as an important
>> UX-experience when on-boarding new users to Bitcoin.
>
> Totally agree. My first thought is maybe you could keep bip37 filtering
> as optional for unconfirmed transactions. Since you're only interested
Really bad for privacy. Data for transactions at the tip is only
14kb/s-- potentially less if segwit is in use and you're not getting
witnesses. Is that really that burdensome?
FWIW, leaving a mobile browser just running while pointed at some
websites seems to use more traffic than that just loading advertising.
:)
^ permalink raw reply [flat|nested] 39+ messages in thread
* Re: [bitcoin-dev] BIP Proposal: Compact Client Side Filtering for Light Clients
2017-06-02 2:28 ` Gregory Maxwell
@ 2017-06-02 3:35 ` Alex Akselrod
2017-06-02 16:07 ` Chris Pacia
1 sibling, 0 replies; 39+ messages in thread
From: Alex Akselrod @ 2017-06-02 3:35 UTC (permalink / raw)
To: Bitcoin Dev
[-- Attachment #1: Type: text/plain, Size: 2032 bytes --]
I agree with Greg and Laolu; BIP-37 filtering for transactions is no better
than for blocks and completely destroys privacy.
A constant stream of transactions is OK, but even cheaper for light clients
would be Laolu's proposal of streaming more tx data than existing inv
messages but less than existing tx messages.
We could make a bit field of things to include in every inv-with-metadata
message, such as:
- witness data
- scriptSig data pushes
- scriptPubKey
- hash of scriptPubKey (unnecessary if full scriptPubKey is sent)
- scriptPubKey data pushes
- etc.
This way a full node might be able to tell what application (or type of
application) a light client is running, but not the client's addresses or
outputs, except maybe when the client originates transactions.
On Thu, Jun 1, 2017 at 10:28 PM, Gregory Maxwell via bitcoin-dev <
bitcoin-dev@lists•linuxfoundation.org> wrote:
> On Fri, Jun 2, 2017 at 2:15 AM, Chris via bitcoin-dev
> <bitcoin-dev@lists•linuxfoundation.org> wrote:
> > On 06/01/2017 06:10 PM, Olaoluwa Osuntokun via bitcoin-dev wrote:
> >
> >> One aspect which isn't in this BIP draft is direct support for
> unconfirmed
> >> transactions. I consider such a feature an important UX feature for
> mobile
> >> phones, and something which I've personally seen as an important
> >> UX-experience when on-boarding new users to Bitcoin.
> >
> > Totally agree. My first thought is maybe you could keep bip37 filtering
> > as optional for unconfirmed transactions. Since you're only interested
>
> Really bad for privacy. Data for transactions at the tip is only
> 14kb/s-- potentially less if segwit is in use and you're not getting
> witnesses. Is that really that burdensome?
>
> FWIW, leaving a mobile browser just running while pointed at some
> websites seems to use more traffic than that just loading advertising.
> :)
> _______________________________________________
> bitcoin-dev mailing list
> bitcoin-dev@lists•linuxfoundation.org
> https://lists.linuxfoundation.org/mailman/listinfo/bitcoin-dev
>
[-- Attachment #2: Type: text/html, Size: 2873 bytes --]
^ permalink raw reply [flat|nested] 39+ messages in thread
* Re: [bitcoin-dev] BIP Proposal: Compact Client Side Filtering for Light Clients
2017-06-02 2:28 ` Gregory Maxwell
2017-06-02 3:35 ` Alex Akselrod
@ 2017-06-02 16:07 ` Chris Pacia
1 sibling, 0 replies; 39+ messages in thread
From: Chris Pacia @ 2017-06-02 16:07 UTC (permalink / raw)
To: Gregory Maxwell; +Cc: Bitcoin Protocol Discussion
[-- Attachment #1: Type: text/plain, Size: 598 bytes --]
On Jun 1, 2017 10:28 PM, "Gregory Maxwell" <greg@xiph•org> wrote:
Really bad for privacy.
As I mentioned the issue is the potential collision with the block filter,
but bloom filters by themselves aren't inherently bad for privacy. They
just reduce the anonymity set. The reason bip37 doesn't work for privacy is
how the filters have to be used/abused to make it work. That's not what I
mentioned above.
Data for transactions at the tip is only
14kb/s-- potentially less if segwit is in use and you're not getting
witnesses. Is that really that burdensome?
Your not running that on mobile.
[-- Attachment #2: Type: text/html, Size: 1250 bytes --]
^ permalink raw reply [flat|nested] 39+ messages in thread
* Re: [bitcoin-dev] BIP Proposal: Compact Client Side Filtering for Light Clients
2017-06-01 19:01 [bitcoin-dev] BIP Proposal: Compact Client Side Filtering for Light Clients Olaoluwa Osuntokun
2017-06-01 21:00 ` Eric Lombrozo
2017-06-01 21:33 ` Matt Corallo
@ 2017-06-02 4:49 ` Olaoluwa Osuntokun
2017-06-09 3:59 ` Olaoluwa Osuntokun
2017-06-02 6:00 ` Karl Johan Alm
` (3 subsequent siblings)
6 siblings, 1 reply; 39+ messages in thread
From: Olaoluwa Osuntokun @ 2017-06-02 4:49 UTC (permalink / raw)
To: Arnoud Kouwenhoven - Pukaki Corp via bitcoin-dev
[-- Attachment #1: Type: text/plain, Size: 6061 bytes --]
> In order to consider the average+median filter sizes in a world worth
larger
> blocks, I also ran the index for testnet:
>
> * total size: 2753238530
> * total avg: 5918.95736054141
> * total median: 60202
> * total max: 74983
> * regular size: 1165148878
> * regular avg: 2504.856172982827
> * regular median: 24812
> * regular max: 64554
> * extended size: 1588089652
> * extended avg: 3414.1011875585823
> * extended median: 35260
> * extended max: 41731
>
Oops, realized I made a mistake. These are the stats for Feb 2016 until
about a
month ago (since height 400k iirc).
-- Laolu
On Thu, Jun 1, 2017 at 12:01 PM Olaoluwa Osuntokun <laolu32@gmail•com>
wrote:
> Hi y'all,
>
> Alex Akselrod and I would like to propose a new light client BIP for
> consideration:
> *
> https://github.com/Roasbeef/bips/blob/master/gcs_light_client.mediawiki
>
> This BIP proposal describes a concrete specification (along with a
> reference implementations[1][2][3]) for the much discussed client-side
> filtering reversal of BIP-37. The precise details are described in the
> BIP, but as a summary: we've implemented a new light-client mode that uses
> client-side filtering based off of Golomb-Rice coded sets. Full-nodes
> maintain an additional index of the chain, and serve this compact filter
> (the index) to light clients which request them. Light clients then fetch
> these filters, query the locally and _maybe_ fetch the block if a relevant
> item matches. The cool part is that blocks can be fetched from _any_
> source, once the light client deems it necessary. Our primary motivation
> for this work was enabling a light client mode for lnd[4] in order to
> support a more light-weight back end paving the way for the usage of
> Lightning on mobile phones and other devices. We've integrated neutrino
> as a back end for lnd, and will be making the updated code public very
> soon.
>
> One specific area we'd like feedback on is the parameter selection. Unlike
> BIP-37 which allows clients to dynamically tune their false positive rate,
> our proposal uses a _fixed_ false-positive. Within the document, it's
> currently specified as P = 1/2^20. We've done a bit of analysis and
> optimization attempting to optimize the following sum:
> filter_download_bandwidth + expected_block_false_positive_bandwidth. Alex
> has made a JS calculator that allows y'all to explore the affect of
> tweaking the false positive rate in addition to the following variables:
> the number of items the wallet is scanning for, the size of the blocks,
> number of blocks fetched, and the size of the filters themselves. The
> calculator calculates the expected bandwidth utilization using the CDF of
> the Geometric Distribution. The calculator can be found here:
> https://aakselrod.github.io/gcs_calc.html. Alex also has an empirical
> script he's been running on actual data, and the results seem to match up
> rather nicely.
>
> We we're excited to see that Karl Johan Alm (kallewoof) has done some
> (rather extensive!) analysis of his own, focusing on a distinct encoding
> type [5]. I haven't had the time yet to dig into his report yet, but I
> think I've read enough to extract the key difference in our encodings: his
> filters use a binomial encoding _directly_ on the filter contents, will we
> instead create a Golomb-Coded set with the contents being _hashes_ (we use
> siphash) of the filter items.
>
> Using a fixed fp=20, I have some stats detailing the total index size, as
> well as averages for both mainnet and testnet. For mainnet, using the
> filter contents as currently described in the BIP (basic + extended), the
> total size of the index comes out to 6.9GB. The break down is as follows:
>
> * total size: 6976047156
> * total avg: 14997.220622758816
> * total median: 3801
> * total max: 79155
> * regular size: 3117183743
> * regular avg: 6701.372750217131
> * regular median: 1734
> * regular max: 67533
> * extended size: 3858863413 <(385)%20886-3413>
> * extended avg: 8295.847872541684
> * extended median: 2041
> * extended max: 52508
>
> In order to consider the average+median filter sizes in a world worth
> larger blocks, I also ran the index for testnet:
>
> * total size: 2753238530
> * total avg: 5918.95736054141
> * total median: 60202
> * total max: 74983
> * regular size: 1165148878
> * regular avg: 2504.856172982827
> * regular median: 24812
> * regular max: 64554
> * extended size: 1588089652
> * extended avg: 3414.1011875585823
> * extended median: 35260
> * extended max: 41731
>
> Finally, here are the testnet stats which take into account the increase
> in the maximum filter size due to segwit's block-size increase. The max
> filter sizes are a bit larger due to some of the habitual blocks I
> created last year when testing segwit (transactions with 30k inputs, 30k
> outputs, etc).
>
> * total size: 585087597
> * total avg: 520.8839608674402
> * total median: 20
> * total max: 164598
> * regular size: 299325029
> * regular avg: 266.4790836307566
> * regular median: 13
> * regular max: 164583
> * extended size: 285762568
> * extended avg: 254.4048772366836
> * extended median: 7
> * extended max: 127631
>
> For those that are interested in the raw data, I've uploaded a CSV file
> of raw data for each block (mainnet + testnet), which can be found here:
> * mainnet: (14MB):
> https://www.dropbox.com/s/4yk2u8dj06njbuv/mainnet-gcs-stats.csv?dl=0
> * testnet: (25MB):
> https://www.dropbox.com/s/w7dmmcbocnmjfbo/gcs-stats-testnet.csv?dl=0
>
>
> We look forward to getting feedback from all of y'all!
>
> -- Laolu
>
>
> [1]: https://github.com/lightninglabs/neutrino
> [2]: https://github.com/Roasbeef/btcd/tree/segwit-cbf
> [3]: https://github.com/Roasbeef/btcutil/tree/gcs/gcs
> [4]: https://github.com/lightningnetwork/lnd/
>
> -- Laolu
>
>
[-- Attachment #2: Type: text/html, Size: 8576 bytes --]
^ permalink raw reply [flat|nested] 39+ messages in thread* Re: [bitcoin-dev] BIP Proposal: Compact Client Side Filtering for Light Clients
2017-06-02 4:49 ` Olaoluwa Osuntokun
@ 2017-06-09 3:59 ` Olaoluwa Osuntokun
2017-11-09 23:44 ` Olaoluwa Osuntokun
0 siblings, 1 reply; 39+ messages in thread
From: Olaoluwa Osuntokun @ 2017-06-09 3:59 UTC (permalink / raw)
To: Arnoud Kouwenhoven - Pukaki Corp via bitcoin-dev
[-- Attachment #1: Type: text/plain, Size: 7316 bytes --]
Hi y'all,
Thanks for all the comments so far!
I've pushed a series of updates to the text of the BIP repo linked in the
OP.
The fixes include: typos, components of the specification which were
incorrect
(N is the total number of items, NOT the number of txns in the block), and a
few sections have been clarified.
The latest version also includes a set of test vectors (as CSV files), which
for a series of fp rates (1/2 to 1/2^32) includes (for 6 testnet blocks,
one of
which generates a "null" filter):
* The block height
* The block hash
* The raw block itself
* The previous basic+extended filter header
* The basic+extended filter header for the block
* The basic+extended filter for the block
The size of the test vectors was too large to include in-line within the
document, so we put them temporarily in a distinct folder [1]. The code
used to
generate the test vectors has also been included.
-- Laolu
[1]: https://github.com/Roasbeef/bips/tree/master/gcs_light_client
On Thu, Jun 1, 2017 at 9:49 PM Olaoluwa Osuntokun <laolu32@gmail•com> wrote:
> > In order to consider the average+median filter sizes in a world worth
> larger
> > blocks, I also ran the index for testnet:
> >
> > * total size: 2753238530
> > * total avg: 5918.95736054141
> > * total median: 60202
> > * total max: 74983
> > * regular size: 1165148878
> > * regular avg: 2504.856172982827
> > * regular median: 24812
> > * regular max: 64554
> > * extended size: 1588089652
> > * extended avg: 3414.1011875585823
> > * extended median: 35260
> > * extended max: 41731
> >
>
> Oops, realized I made a mistake. These are the stats for Feb 2016 until
> about a
> month ago (since height 400k iirc).
>
> -- Laolu
>
>
> On Thu, Jun 1, 2017 at 12:01 PM Olaoluwa Osuntokun <laolu32@gmail•com>
> wrote:
>
>> Hi y'all,
>>
>> Alex Akselrod and I would like to propose a new light client BIP for
>> consideration:
>> *
>> https://github.com/Roasbeef/bips/blob/master/gcs_light_client.mediawiki
>>
>> This BIP proposal describes a concrete specification (along with a
>> reference implementations[1][2][3]) for the much discussed client-side
>> filtering reversal of BIP-37. The precise details are described in the
>> BIP, but as a summary: we've implemented a new light-client mode that uses
>> client-side filtering based off of Golomb-Rice coded sets. Full-nodes
>> maintain an additional index of the chain, and serve this compact filter
>> (the index) to light clients which request them. Light clients then fetch
>> these filters, query the locally and _maybe_ fetch the block if a relevant
>> item matches. The cool part is that blocks can be fetched from _any_
>> source, once the light client deems it necessary. Our primary motivation
>> for this work was enabling a light client mode for lnd[4] in order to
>> support a more light-weight back end paving the way for the usage of
>> Lightning on mobile phones and other devices. We've integrated neutrino
>> as a back end for lnd, and will be making the updated code public very
>> soon.
>>
>> One specific area we'd like feedback on is the parameter selection. Unlike
>> BIP-37 which allows clients to dynamically tune their false positive rate,
>> our proposal uses a _fixed_ false-positive. Within the document, it's
>> currently specified as P = 1/2^20. We've done a bit of analysis and
>> optimization attempting to optimize the following sum:
>> filter_download_bandwidth + expected_block_false_positive_bandwidth. Alex
>> has made a JS calculator that allows y'all to explore the affect of
>> tweaking the false positive rate in addition to the following variables:
>> the number of items the wallet is scanning for, the size of the blocks,
>> number of blocks fetched, and the size of the filters themselves. The
>> calculator calculates the expected bandwidth utilization using the CDF of
>> the Geometric Distribution. The calculator can be found here:
>> https://aakselrod.github.io/gcs_calc.html. Alex also has an empirical
>> script he's been running on actual data, and the results seem to match up
>> rather nicely.
>>
>> We we're excited to see that Karl Johan Alm (kallewoof) has done some
>> (rather extensive!) analysis of his own, focusing on a distinct encoding
>> type [5]. I haven't had the time yet to dig into his report yet, but I
>> think I've read enough to extract the key difference in our encodings: his
>> filters use a binomial encoding _directly_ on the filter contents, will we
>> instead create a Golomb-Coded set with the contents being _hashes_ (we use
>> siphash) of the filter items.
>>
>> Using a fixed fp=20, I have some stats detailing the total index size, as
>> well as averages for both mainnet and testnet. For mainnet, using the
>> filter contents as currently described in the BIP (basic + extended), the
>> total size of the index comes out to 6.9GB. The break down is as follows:
>>
>> * total size: 6976047156
>> * total avg: 14997.220622758816
>> * total median: 3801
>> * total max: 79155
>> * regular size: 3117183743
>> * regular avg: 6701.372750217131
>> * regular median: 1734
>> * regular max: 67533
>> * extended size: 3858863413 <(385)%20886-3413>
>> * extended avg: 8295.847872541684
>> * extended median: 2041
>> * extended max: 52508
>>
>> In order to consider the average+median filter sizes in a world worth
>> larger blocks, I also ran the index for testnet:
>>
>> * total size: 2753238530
>> * total avg: 5918.95736054141
>> * total median: 60202
>> * total max: 74983
>> * regular size: 1165148878
>> * regular avg: 2504.856172982827
>> * regular median: 24812
>> * regular max: 64554
>> * extended size: 1588089652
>> * extended avg: 3414.1011875585823
>> * extended median: 35260
>> * extended max: 41731
>>
>> Finally, here are the testnet stats which take into account the increase
>> in the maximum filter size due to segwit's block-size increase. The max
>> filter sizes are a bit larger due to some of the habitual blocks I
>> created last year when testing segwit (transactions with 30k inputs, 30k
>> outputs, etc).
>>
>> * total size: 585087597
>> * total avg: 520.8839608674402
>> * total median: 20
>> * total max: 164598
>> * regular size: 299325029
>> * regular avg: 266.4790836307566
>> * regular median: 13
>> * regular max: 164583
>> * extended size: 285762568
>> * extended avg: 254.4048772366836
>> * extended median: 7
>> * extended max: 127631
>>
>> For those that are interested in the raw data, I've uploaded a CSV file
>> of raw data for each block (mainnet + testnet), which can be found here:
>> * mainnet: (14MB):
>> https://www.dropbox.com/s/4yk2u8dj06njbuv/mainnet-gcs-stats.csv?dl=0
>> * testnet: (25MB):
>> https://www.dropbox.com/s/w7dmmcbocnmjfbo/gcs-stats-testnet.csv?dl=0
>>
>>
>> We look forward to getting feedback from all of y'all!
>>
>> -- Laolu
>>
>>
>> [1]: https://github.com/lightninglabs/neutrino
>> [2]: https://github.com/Roasbeef/btcd/tree/segwit-cbf
>> [3]: https://github.com/Roasbeef/btcutil/tree/gcs/gcs
>> [4]: https://github.com/lightningnetwork/lnd/
>>
>> -- Laolu
>>
>>
[-- Attachment #2: Type: text/html, Size: 10390 bytes --]
^ permalink raw reply [flat|nested] 39+ messages in thread
* Re: [bitcoin-dev] BIP Proposal: Compact Client Side Filtering for Light Clients
2017-06-09 3:59 ` Olaoluwa Osuntokun
@ 2017-11-09 23:44 ` Olaoluwa Osuntokun
0 siblings, 0 replies; 39+ messages in thread
From: Olaoluwa Osuntokun @ 2017-11-09 23:44 UTC (permalink / raw)
To: Arnoud Kouwenhoven - Pukaki Corp via bitcoin-dev
[-- Attachment #1: Type: text/plain, Size: 9911 bytes --]
Hi y'all,
Since my last email we've made a number of changes to the BIP. The changes
made
were driven by the feedback we've received so far in this thread, and also
as a
result of real-world testing using this new proposal as the basis for our
light
weight LN node which powers the demo Lightning desktop application we
recently
released.
A highlight of the changes made between this version and the last follows:
* We've removed the modulus operation in the inner loop when constructing
filters. This has been replaced with an alternative, more efficient
mapping[1] as suggested by gmaxwell and sipa. In our implementation, we
perform the operation in a piece-wise fashion by hand. Alternative
implementations can take advantage of 128-bit arithmetic extensions on
supporting CPU's.
* The txid has been moved from the extended filter to the regular filter.
During out testing of the new light client with our LN node
implementation,
we found that we were able to reduce network traffic as we only need the
extended filter for rare on-chain events.
* We now use the 6th service bit. We realized that the bit we had chosen
prior was already being used to signal support of x-thin block syncing.
To
select this bit number, we ran a scanner on the addrman of our nodes and
also the network to fin da bit that wasn't used widely.
* An error in the BIP that didn't include the public key script of
coinbase
transactions in the filter has been fixed.
* An error in the BIP when constructing the initial "genesis" filter has
been
fixed.
* We no longer use the ProtocolVersion field in the getcfheaders message
or
its response.
* The specification of several newly defined messages were incorrect and
have
been fixed.
* A number of typos spotted by several reviewers have been fixed.
The full commit history of the BIP draft can be found here:
https://github.com/Roasbeef/bips/commits/gcs-bip-draft
At this point, we're ready to make a PR against the official BIP repo and to
request a number to be assigned to our proposal. Thanks to all those that
have
reviewed, and contributed to the proposal!
[1]:
https://lemire.me/blog/2016/06/27/a-fast-alternative-to-the-modulo-reduction/
-- Laolu
On Thu, Jun 8, 2017 at 8:59 PM Olaoluwa Osuntokun <laolu32@gmail•com> wrote:
> Hi y'all,
>
> Thanks for all the comments so far!
>
> I've pushed a series of updates to the text of the BIP repo linked in the
> OP.
> The fixes include: typos, components of the specification which were
> incorrect
> (N is the total number of items, NOT the number of txns in the block), and
> a
> few sections have been clarified.
>
> The latest version also includes a set of test vectors (as CSV files),
> which
> for a series of fp rates (1/2 to 1/2^32) includes (for 6 testnet blocks,
> one of
> which generates a "null" filter):
>
> * The block height
> * The block hash
> * The raw block itself
> * The previous basic+extended filter header
> * The basic+extended filter header for the block
> * The basic+extended filter for the block
>
> The size of the test vectors was too large to include in-line within the
> document, so we put them temporarily in a distinct folder [1]. The code
> used to
> generate the test vectors has also been included.
>
> -- Laolu
>
> [1]: https://github.com/Roasbeef/bips/tree/master/gcs_light_client
>
>
> On Thu, Jun 1, 2017 at 9:49 PM Olaoluwa Osuntokun <laolu32@gmail•com>
> wrote:
>
>> > In order to consider the average+median filter sizes in a world worth
>> larger
>> > blocks, I also ran the index for testnet:
>> >
>> > * total size: 2753238530
>> > * total avg: 5918.95736054141
>> > * total median: 60202
>> > * total max: 74983
>> > * regular size: 1165148878
>> > * regular avg: 2504.856172982827
>> > * regular median: 24812
>> > * regular max: 64554
>> > * extended size: 1588089652
>> > * extended avg: 3414.1011875585823
>> > * extended median: 35260
>> > * extended max: 41731
>> >
>>
>> Oops, realized I made a mistake. These are the stats for Feb 2016 until
>> about a
>> month ago (since height 400k iirc).
>>
>> -- Laolu
>>
>>
>> On Thu, Jun 1, 2017 at 12:01 PM Olaoluwa Osuntokun <laolu32@gmail•com>
>> wrote:
>>
>>> Hi y'all,
>>>
>>> Alex Akselrod and I would like to propose a new light client BIP for
>>> consideration:
>>> *
>>> https://github.com/Roasbeef/bips/blob/master/gcs_light_client.mediawiki
>>>
>>> This BIP proposal describes a concrete specification (along with a
>>> reference implementations[1][2][3]) for the much discussed client-side
>>> filtering reversal of BIP-37. The precise details are described in the
>>> BIP, but as a summary: we've implemented a new light-client mode that
>>> uses
>>> client-side filtering based off of Golomb-Rice coded sets. Full-nodes
>>> maintain an additional index of the chain, and serve this compact filter
>>> (the index) to light clients which request them. Light clients then fetch
>>> these filters, query the locally and _maybe_ fetch the block if a
>>> relevant
>>> item matches. The cool part is that blocks can be fetched from _any_
>>> source, once the light client deems it necessary. Our primary motivation
>>> for this work was enabling a light client mode for lnd[4] in order to
>>> support a more light-weight back end paving the way for the usage of
>>> Lightning on mobile phones and other devices. We've integrated neutrino
>>> as a back end for lnd, and will be making the updated code public very
>>> soon.
>>>
>>> One specific area we'd like feedback on is the parameter selection.
>>> Unlike
>>> BIP-37 which allows clients to dynamically tune their false positive
>>> rate,
>>> our proposal uses a _fixed_ false-positive. Within the document, it's
>>> currently specified as P = 1/2^20. We've done a bit of analysis and
>>> optimization attempting to optimize the following sum:
>>> filter_download_bandwidth + expected_block_false_positive_bandwidth. Alex
>>> has made a JS calculator that allows y'all to explore the affect of
>>> tweaking the false positive rate in addition to the following variables:
>>> the number of items the wallet is scanning for, the size of the blocks,
>>> number of blocks fetched, and the size of the filters themselves. The
>>> calculator calculates the expected bandwidth utilization using the CDF of
>>> the Geometric Distribution. The calculator can be found here:
>>> https://aakselrod.github.io/gcs_calc.html. Alex also has an empirical
>>> script he's been running on actual data, and the results seem to match up
>>> rather nicely.
>>>
>>> We we're excited to see that Karl Johan Alm (kallewoof) has done some
>>> (rather extensive!) analysis of his own, focusing on a distinct encoding
>>> type [5]. I haven't had the time yet to dig into his report yet, but I
>>> think I've read enough to extract the key difference in our encodings:
>>> his
>>> filters use a binomial encoding _directly_ on the filter contents, will
>>> we
>>> instead create a Golomb-Coded set with the contents being _hashes_ (we
>>> use
>>> siphash) of the filter items.
>>>
>>> Using a fixed fp=20, I have some stats detailing the total index size, as
>>> well as averages for both mainnet and testnet. For mainnet, using the
>>> filter contents as currently described in the BIP (basic + extended), the
>>> total size of the index comes out to 6.9GB. The break down is as follows:
>>>
>>> * total size: 6976047156
>>> * total avg: 14997.220622758816
>>> * total median: 3801
>>> * total max: 79155
>>> * regular size: 3117183743
>>> * regular avg: 6701.372750217131
>>> * regular median: 1734
>>> * regular max: 67533
>>> * extended size: 3858863413 <(385)%20886-3413>
>>> * extended avg: 8295.847872541684
>>> * extended median: 2041
>>> * extended max: 52508
>>>
>>> In order to consider the average+median filter sizes in a world worth
>>> larger blocks, I also ran the index for testnet:
>>>
>>> * total size: 2753238530
>>> * total avg: 5918.95736054141
>>> * total median: 60202
>>> * total max: 74983
>>> * regular size: 1165148878
>>> * regular avg: 2504.856172982827
>>> * regular median: 24812
>>> * regular max: 64554
>>> * extended size: 1588089652
>>> * extended avg: 3414.1011875585823
>>> * extended median: 35260
>>> * extended max: 41731
>>>
>>> Finally, here are the testnet stats which take into account the increase
>>> in the maximum filter size due to segwit's block-size increase. The max
>>> filter sizes are a bit larger due to some of the habitual blocks I
>>> created last year when testing segwit (transactions with 30k inputs, 30k
>>> outputs, etc).
>>>
>>> * total size: 585087597
>>> * total avg: 520.8839608674402
>>> * total median: 20
>>> * total max: 164598
>>> * regular size: 299325029
>>> * regular avg: 266.4790836307566
>>> * regular median: 13
>>> * regular max: 164583
>>> * extended size: 285762568
>>> * extended avg: 254.4048772366836
>>> * extended median: 7
>>> * extended max: 127631
>>>
>>> For those that are interested in the raw data, I've uploaded a CSV file
>>> of raw data for each block (mainnet + testnet), which can be found here:
>>> * mainnet: (14MB):
>>> https://www.dropbox.com/s/4yk2u8dj06njbuv/mainnet-gcs-stats.csv?dl=0
>>> * testnet: (25MB):
>>> https://www.dropbox.com/s/w7dmmcbocnmjfbo/gcs-stats-testnet.csv?dl=0
>>>
>>>
>>> We look forward to getting feedback from all of y'all!
>>>
>>> -- Laolu
>>>
>>>
>>> [1]: https://github.com/lightninglabs/neutrino
>>> [2]: https://github.com/Roasbeef/btcd/tree/segwit-cbf
>>> [3]: https://github.com/Roasbeef/btcutil/tree/gcs/gcs
>>> [4]: https://github.com/lightningnetwork/lnd/
>>>
>>> -- Laolu
>>>
>>>
[-- Attachment #2: Type: text/html, Size: 13867 bytes --]
^ permalink raw reply [flat|nested] 39+ messages in thread
* Re: [bitcoin-dev] BIP Proposal: Compact Client Side Filtering for Light Clients
2017-06-01 19:01 [bitcoin-dev] BIP Proposal: Compact Client Side Filtering for Light Clients Olaoluwa Osuntokun
` (2 preceding siblings ...)
2017-06-02 4:49 ` Olaoluwa Osuntokun
@ 2017-06-02 6:00 ` Karl Johan Alm
[not found] ` <CAE0pnx+RRAP269VeWAcxKbrcS9qX4LS8_6nY_js8X5NtQ22t_A@mail.gmail.com>
2017-06-07 21:41 ` Gregory Maxwell
` (2 subsequent siblings)
6 siblings, 1 reply; 39+ messages in thread
From: Karl Johan Alm @ 2017-06-02 6:00 UTC (permalink / raw)
To: Olaoluwa Osuntokun; +Cc: Arnoud Kouwenhoven - Pukaki Corp via bitcoin-dev
Hello,
Really wish I'd known you were working on this a few weeks ago, but
such is life. Hopefully I can provide some useful feedback.
On Fri, Jun 2, 2017 at 4:01 AM, Olaoluwa Osuntokun via bitcoin-dev
<bitcoin-dev@lists•linuxfoundation.org> wrote:
> Full-nodes
> maintain an additional index of the chain, and serve this compact filter
> (the index) to light clients which request them. Light clients then fetch
> these filters, query the locally and _maybe_ fetch the block if a relevant
> item matches.
Is it necessary to maintain the index all the way to the beginning of
the chain? When would clients request "really old digests" and why?
> One specific area we'd like feedback on is the parameter selection. Unlike
> BIP-37 which allows clients to dynamically tune their false positive rate,
> our proposal uses a _fixed_ false-positive. Within the document, it's
> currently specified as P = 1/2^20. We've done a bit of analysis and
> optimization attempting to optimize the following sum:
> filter_download_bandwidth + expected_block_false_positive_bandwidth. Alex
> has made a JS calculator that allows y'all to explore the affect of
> tweaking the false positive rate in addition to the following variables:
> the number of items the wallet is scanning for, the size of the blocks,
> number of blocks fetched, and the size of the filters themselves. The
> calculator calculates the expected bandwidth utilization using the CDF of
> the Geometric Distribution. The calculator can be found here:
> https://aakselrod.github.io/gcs_calc.html. Alex also has an empirical
> script he's been running on actual data, and the results seem to match up
> rather nicely.
I haven't tried the tool yet, and maybe it will answer some of my questions.
On what data were the simulated wallets on actual data based? How did
false positive rates for wallets with lots of items (pubkeys etc) play
out? Is there a maximum number of items for a wallet before it becomes
too bandwidth costly to use digests?
> We we're excited to see that Karl Johan Alm (kallewoof) has done some
> (rather extensive!) analysis of his own, focusing on a distinct encoding
> type [5]. I haven't had the time yet to dig into his report yet, but I
> think I've read enough to extract the key difference in our encodings: his
> filters use a binomial encoding _directly_ on the filter contents, will we
> instead create a Golomb-Coded set with the contents being _hashes_ (we use
> siphash) of the filter items.
I will definitely try to reproduce my experiments with Golomb-Coded
sets and see what I come up with. It seems like you've got a little
less than half the size of my digests for 1-block digests but I
haven't tried making digests for all blocks (and lots of early blocks
are empty).
On the BIP proposal itself:
In Compact Filter Header Chain, you mention that clients should
download filters from nodes if filter_headers is not identical, and
ban offending nodes. What about temporary forks in the chain? What
about longer forks? In general, I am curious how you will deal with
reorgs and temporary non-consensus related chain splits.
I am also curious if you have considered digests containing multiple
blocks. Retaining a permanent binsearchable record of the entire chain
is obviously too space costly, but keeping the last X blocks as
binsearchable could speed up syncing for clients tremendously, I feel.
It may also be space efficient to ONLY store older digests in chunks
of e.g. 8 blocks. A client syncing up finding a match in an 8-block
chunk would have to grab those 8 blocks, but if it's not recent, that
may be acceptable. It may even be possible to make 4-, 2-, 1-block
digests on demand.
How fast are these to create? Would it make sense to provide digests
on demand in some cases, rather than keeping them around indefinitely?
^ permalink raw reply [flat|nested] 39+ messages in thread* Re: [bitcoin-dev] BIP Proposal: Compact Client Side Filtering for Light Clients
2017-06-01 19:01 [bitcoin-dev] BIP Proposal: Compact Client Side Filtering for Light Clients Olaoluwa Osuntokun
` (3 preceding siblings ...)
2017-06-02 6:00 ` Karl Johan Alm
@ 2017-06-07 21:41 ` Gregory Maxwell
2017-06-09 3:42 ` Olaoluwa Osuntokun
2017-06-08 9:50 ` Tomas
2017-06-19 11:58 ` Andreas Schildbach
6 siblings, 1 reply; 39+ messages in thread
From: Gregory Maxwell @ 2017-06-07 21:41 UTC (permalink / raw)
To: Olaoluwa Osuntokun; +Cc: Arnoud Kouwenhoven - Pukaki Corp via bitcoin-dev
On Thu, Jun 1, 2017 at 7:01 PM, Olaoluwa Osuntokun via bitcoin-dev
<bitcoin-dev@lists•linuxfoundation.org> wrote:
> Hi y'all,
>
> Alex Akselrod and I would like to propose a new light client BIP for
> consideration:
> * https://github.com/Roasbeef/bips/blob/master/gcs_light_client.mediawiki
I see the inner loop of construction and lookup are free of
non-constant divmod. This will result in implementations being
needlessly slow (especially on arm, but even on modern x86_64 a
division is a 90 cycle-ish affair.)
I believe this can be fixed by using this approach
http://lemire.me/blog/2016/06/27/a-fast-alternative-to-the-modulo-reduction/
which has the same non-uniformity as mod but needs only a multiply
and shift.
Otherwise fast implementations will have to implement the code to
compute bit twiddling hack exact division code, which is kind of
complicated. (e.g. via the technique in "{N}-bit Unsigned Division via
{N}-bit Multiply-Add" by Arch D. Robison).
Shouldn't all cases in your spec where you have N=transactions be
n=indexed-outputs? Otherwise, I think your golomb parameter and false
positive rate are wrong.
^ permalink raw reply [flat|nested] 39+ messages in thread* Re: [bitcoin-dev] BIP Proposal: Compact Client Side Filtering for Light Clients
2017-06-07 21:41 ` Gregory Maxwell
@ 2017-06-09 3:42 ` Olaoluwa Osuntokun
2017-06-09 4:47 ` Olaoluwa Osuntokun
0 siblings, 1 reply; 39+ messages in thread
From: Olaoluwa Osuntokun @ 2017-06-09 3:42 UTC (permalink / raw)
To: Gregory Maxwell; +Cc: Arnoud Kouwenhoven - Pukaki Corp via bitcoin-dev
[-- Attachment #1: Type: text/plain, Size: 3122 bytes --]
Gregory wrote:
> I see the inner loop of construction and lookup are free of
> non-constant divmod. This will result in implementations being
> needlessly slow
Ahh, sipa brought this up other day, but I thought he was referring to the
coding loop (which uses a power of 2 divisor/modulus), not the
siphash-then-reduce loop.
> I believe this can be fixed by using this approach
>
http://lemire.me/blog/2016/06/27/a-fast-alternative-to-the-modulo-reduction/
> which has the same non-uniformity as mod but needs only a multiply and
> shift.
Very cool, I wasn't aware of the existence of such a mapping.
Correct me if I'm wrong, but from my interpretation we can't use that
method as described as we need to output 64-bit integers rather than
32-bit integers. A range of 32-bits would be constrain the number of items
we could encode to be ~4096 to ensure that we don't overflow with fp
values such as 20 (which we currently use in our code).
If filter commitment are to be considered for a soft-fork in the future,
then we should definitely optimize the construction of the filters as much
as possible! I'll look into that paper you referenced to get a feel for
just how complex the optimization would be.
> Shouldn't all cases in your spec where you have N=transactions be
> n=indexed-outputs? Otherwise, I think your golomb parameter and false
> positive rate are wrong.
Yep! Nice catch. Our code is correct, but mistake in the spec was an
oversight on my part. I've pushed a commit[1] to the bip repo referenced
in the OP to fix this error.
I've also pushed another commit to explicitly take advantage of the fact
that P is a power-of-two within the coding loop [2].
-- Laolu
[1]:
https://github.com/Roasbeef/bips/commit/bc5c6d6797f3df1c4a44213963ba12e72122163d
[2]:
https://github.com/Roasbeef/bips/commit/578a4e3aa8ec04524c83bfc5d14be1b2660e7f7a
On Wed, Jun 7, 2017 at 2:41 PM Gregory Maxwell <greg@xiph•org> wrote:
> On Thu, Jun 1, 2017 at 7:01 PM, Olaoluwa Osuntokun via bitcoin-dev
> <bitcoin-dev@lists•linuxfoundation.org> wrote:
> > Hi y'all,
> >
> > Alex Akselrod and I would like to propose a new light client BIP for
> > consideration:
> > *
> https://github.com/Roasbeef/bips/blob/master/gcs_light_client.mediawiki
>
> I see the inner loop of construction and lookup are free of
> non-constant divmod. This will result in implementations being
> needlessly slow (especially on arm, but even on modern x86_64 a
> division is a 90 cycle-ish affair.)
>
> I believe this can be fixed by using this approach
>
> http://lemire.me/blog/2016/06/27/a-fast-alternative-to-the-modulo-reduction/
> which has the same non-uniformity as mod but needs only a multiply
> and shift.
>
> Otherwise fast implementations will have to implement the code to
> compute bit twiddling hack exact division code, which is kind of
> complicated. (e.g. via the technique in "{N}-bit Unsigned Division via
> {N}-bit Multiply-Add" by Arch D. Robison).
>
> Shouldn't all cases in your spec where you have N=transactions be
> n=indexed-outputs? Otherwise, I think your golomb parameter and false
> positive rate are wrong.
>
[-- Attachment #2: Type: text/html, Size: 4614 bytes --]
^ permalink raw reply [flat|nested] 39+ messages in thread* Re: [bitcoin-dev] BIP Proposal: Compact Client Side Filtering for Light Clients
2017-06-09 3:42 ` Olaoluwa Osuntokun
@ 2017-06-09 4:47 ` Olaoluwa Osuntokun
0 siblings, 0 replies; 39+ messages in thread
From: Olaoluwa Osuntokun @ 2017-06-09 4:47 UTC (permalink / raw)
To: Gregory Maxwell; +Cc: Arnoud Kouwenhoven - Pukaki Corp via bitcoin-dev
[-- Attachment #1: Type: text/plain, Size: 3776 bytes --]
> Correct me if I'm wrong, but from my interpretation we can't use that
> method as described as we need to output 64-bit integers rather than
> 32-bit integers.
Had a chat with gmax off-list and came to the realization that the method
_should_ indeed generalize to our case of outputting 64-bit integers.
We'll need to do a bit of bit twiddling to make it work properly. I'll
modify our implementation and report back with some basic benchmarks.
-- Laolu
On Thu, Jun 8, 2017 at 8:42 PM Olaoluwa Osuntokun <laolu32@gmail•com> wrote:
> Gregory wrote:
> > I see the inner loop of construction and lookup are free of
> > non-constant divmod. This will result in implementations being
> > needlessly slow
>
> Ahh, sipa brought this up other day, but I thought he was referring to the
> coding loop (which uses a power of 2 divisor/modulus), not the
> siphash-then-reduce loop.
>
> > I believe this can be fixed by using this approach
> >
> http://lemire.me/blog/2016/06/27/a-fast-alternative-to-the-modulo-reduction/
> > which has the same non-uniformity as mod but needs only a multiply and
> > shift.
>
> Very cool, I wasn't aware of the existence of such a mapping.
>
> Correct me if I'm wrong, but from my interpretation we can't use that
> method as described as we need to output 64-bit integers rather than
> 32-bit integers. A range of 32-bits would be constrain the number of items
> we could encode to be ~4096 to ensure that we don't overflow with fp
> values such as 20 (which we currently use in our code).
>
> If filter commitment are to be considered for a soft-fork in the future,
> then we should definitely optimize the construction of the filters as much
> as possible! I'll look into that paper you referenced to get a feel for
> just how complex the optimization would be.
>
> > Shouldn't all cases in your spec where you have N=transactions be
> > n=indexed-outputs? Otherwise, I think your golomb parameter and false
> > positive rate are wrong.
>
> Yep! Nice catch. Our code is correct, but mistake in the spec was an
> oversight on my part. I've pushed a commit[1] to the bip repo referenced
> in the OP to fix this error.
>
> I've also pushed another commit to explicitly take advantage of the fact
> that P is a power-of-two within the coding loop [2].
>
> -- Laolu
>
> [1]:
> https://github.com/Roasbeef/bips/commit/bc5c6d6797f3df1c4a44213963ba12e72122163d
> [2]:
> https://github.com/Roasbeef/bips/commit/578a4e3aa8ec04524c83bfc5d14be1b2660e7f7a
>
>
> On Wed, Jun 7, 2017 at 2:41 PM Gregory Maxwell <greg@xiph•org> wrote:
>
>> On Thu, Jun 1, 2017 at 7:01 PM, Olaoluwa Osuntokun via bitcoin-dev
>> <bitcoin-dev@lists•linuxfoundation.org> wrote:
>> > Hi y'all,
>> >
>> > Alex Akselrod and I would like to propose a new light client BIP for
>> > consideration:
>> > *
>> https://github.com/Roasbeef/bips/blob/master/gcs_light_client.mediawiki
>>
>> I see the inner loop of construction and lookup are free of
>> non-constant divmod. This will result in implementations being
>> needlessly slow (especially on arm, but even on modern x86_64 a
>> division is a 90 cycle-ish affair.)
>>
>> I believe this can be fixed by using this approach
>>
>> http://lemire.me/blog/2016/06/27/a-fast-alternative-to-the-modulo-reduction/
>> which has the same non-uniformity as mod but needs only a multiply
>> and shift.
>>
>> Otherwise fast implementations will have to implement the code to
>> compute bit twiddling hack exact division code, which is kind of
>> complicated. (e.g. via the technique in "{N}-bit Unsigned Division via
>> {N}-bit Multiply-Add" by Arch D. Robison).
>>
>> Shouldn't all cases in your spec where you have N=transactions be
>> n=indexed-outputs? Otherwise, I think your golomb parameter and false
>> positive rate are wrong.
>>
>
[-- Attachment #2: Type: text/html, Size: 5729 bytes --]
^ permalink raw reply [flat|nested] 39+ messages in thread
* Re: [bitcoin-dev] BIP Proposal: Compact Client Side Filtering for Light Clients
2017-06-01 19:01 [bitcoin-dev] BIP Proposal: Compact Client Side Filtering for Light Clients Olaoluwa Osuntokun
` (4 preceding siblings ...)
2017-06-07 21:41 ` Gregory Maxwell
@ 2017-06-08 9:50 ` Tomas
2017-06-09 3:50 ` Olaoluwa Osuntokun
2017-06-19 11:58 ` Andreas Schildbach
6 siblings, 1 reply; 39+ messages in thread
From: Tomas @ 2017-06-08 9:50 UTC (permalink / raw)
To: bitcoin-dev
[-- Attachment #1: Type: text/plain, Size: 1856 bytes --]
On Thu, Jun 1, 2017, at 21:01, Olaoluwa Osuntokun via bitcoin-dev wrote:> Hi y'all,
>
> Alex Akselrod and I would like to propose a new light client BIP for
> consideration:
> * https://github.com/Roasbeef/bips/blob/master/gcs_light_client.mediawiki>
>
Very interesting.
I would like to consider how this compares to another light client type
with rather different security characteristics where each client would
receive for each transaction in each block,
* The TXID (uncompressed)
* The spent outpoints (with TXIDs compressed)
* The pubkey hash (compressed to reasonable amount of false positives)
A rough estimate would indicate this to be about 2-2.5x as big per block
as your proposal, but comes with rather different security
characteristics, and would not require download since genesis.
The client could verify the TXIDs against the merkle root with a much
stronger (PoW) guarantee compared to the guarantee based on the
assumption of peers being distinct, which your proposal seems to make.
Like your proposal this removes the privacy and processing issues from
server-side filtering, but unlike your proposal retrieval of all txids
in each block can also serve for a basis of fraud proofs and
(disprovable) fraud hints, without resorting to full block downloads.
I don't completely understand the benefit of making the outpoints and
pubkey hashes (weakly) verifiable. These only serve as notifications and
therefore do not seem to introduce an attack vector. Omitting data is
always possible, so receiving data is a prerequisite for verification,
not an assumption that can be made. How could an attacker benefit from
"hiding notifications"?
I think client-side filtering is definitely an important route to
take, but is it worth compressing away the information to verify the
merkle root?
Regards,
Tomas van der Wansem
bitcrust
[-- Attachment #2: Type: text/html, Size: 2537 bytes --]
^ permalink raw reply [flat|nested] 39+ messages in thread* Re: [bitcoin-dev] BIP Proposal: Compact Client Side Filtering for Light Clients
2017-06-08 9:50 ` Tomas
@ 2017-06-09 3:50 ` Olaoluwa Osuntokun
2017-06-09 8:26 ` Tomas
0 siblings, 1 reply; 39+ messages in thread
From: Olaoluwa Osuntokun @ 2017-06-09 3:50 UTC (permalink / raw)
To: Tomas, bitcoin-dev
[-- Attachment #1: Type: text/plain, Size: 3724 bytes --]
Tomas wrote:
> A rough estimate would indicate this to be about 2-2.5x as big per block
> as your proposal, but comes with rather different security
> characteristics, and would not require download since genesis.
Our proposal _doesnt_ require downloading from genesis, if by
"downloading" you mean downloading all the blocks. Clients only need to
sync the block+filter headers, then (if they don't care about historical
blocks), will download filters from their "birthday" onwards.
> The client could verify the TXIDs against the merkle root with a much
> stronger (PoW) guarantee compared to the guarantee based on the assumption
> of peers being distinct, which your proposal seems to make
Our proposal only makes a "one honest peer" assumption, which is the same
as any other operating mode. Also as client still download all the
headers, they're able to verify PoW conformance/work as normal.
> I don't completely understand the benefit of making the outpoints and
> pubkey hashes (weakly) verifiable. These only serve as notifications and
> therefore do not seem to introduce an attack vector.
Not sure what you mean by this. Care to elaborate?
> I think client-side filtering is definitely an important route to take,
> but is it worth compressing away the information to verify the merkle
> root?
That's not the case with our proposal. Clients get the _entire_ block (if
they need it), so they can verify the merkle root as normal. Unless one of
us is misinterpreting the other here.
-- Laolu
On Thu, Jun 8, 2017 at 6:34 AM Tomas via bitcoin-dev <
bitcoin-dev@lists•linuxfoundation.org> wrote:
> On Thu, Jun 1, 2017, at 21:01, Olaoluwa Osuntokun via bitcoin-dev wrote:
>
> Hi y'all,
>
> Alex Akselrod and I would like to propose a new light client BIP for
> consideration:
> *
> https://github.com/Roasbeef/bips/blob/master/gcs_light_client.mediawiki
>
>
>
> Very interesting.
>
> I would like to consider how this compares to another light client type
> with rather different security characteristics where each client would
> receive for each transaction in each block,
>
> * The TXID (uncompressed)
> * The spent outpoints (with TXIDs compressed)
> * The pubkey hash (compressed to reasonable amount of false positives)
>
> A rough estimate would indicate this to be about 2-2.5x as big per block
> as your proposal, but comes with rather different security characteristics,
> and would not require download since genesis.
>
> The client could verify the TXIDs against the merkle root with a much
> stronger (PoW) guarantee compared to the guarantee based on the assumption
> of peers being distinct, which your proposal seems to make. Like your
> proposal this removes the privacy and processing issues from server-side
> filtering, but unlike your proposal retrieval of all txids in each block
> can also serve for a basis of fraud proofs and (disprovable) fraud hints,
> without resorting to full block downloads.
>
> I don't completely understand the benefit of making the outpoints and
> pubkey hashes (weakly) verifiable. These only serve as notifications and
> therefore do not seem to introduce an attack vector. Omitting data is
> always possible, so receiving data is a prerequisite for verification, not
> an assumption that can be made. How could an attacker benefit from "hiding
> notifications"?
>
> I think client-side filtering is definitely an important route to take,
> but is it worth compressing away the information to verify the merkle root?
>
> Regards,
> Tomas van der Wansem
> bitcrust
>
> _______________________________________________
> bitcoin-dev mailing list
> bitcoin-dev@lists•linuxfoundation.org
> https://lists.linuxfoundation.org/mailman/listinfo/bitcoin-dev
>
[-- Attachment #2: Type: text/html, Size: 5240 bytes --]
^ permalink raw reply [flat|nested] 39+ messages in thread
* Re: [bitcoin-dev] BIP Proposal: Compact Client Side Filtering for Light Clients
2017-06-09 3:50 ` Olaoluwa Osuntokun
@ 2017-06-09 8:26 ` Tomas
0 siblings, 0 replies; 39+ messages in thread
From: Tomas @ 2017-06-09 8:26 UTC (permalink / raw)
To: Olaoluwa Osuntokun, bitcoin-dev
On Fri, Jun 9, 2017, at 05:50, Olaoluwa Osuntokun wrote:
> Tomas wrote:
> > I don't completely understand the benefit of making the outpoints and
> > pubkey hashes (weakly) verifiable. These only serve as notifications and
> > therefore do not seem to introduce an attack vector.
>
> Not sure what you mean by this. Care to elaborate?
>
I will rephrase. The BIP reads:
> Additionally, Full nodes can nearly undetectably lie by omission causing a denial of service which can
lead to undesirable failure modes in applications whose safety
critically relies on responding to certain
on-chain events.
I understand that the compact header chain is used to mitigate against
this, but I am unsure about the use
cases and trade-offs.
For a normal wallet, the only thing I can imagine an attacker could do
is pretending a transaction did not confirm
yet, causing nuisance.
An application critically depending on knowing what happens on-chain
surely is better off downloading
the TXIDs, providing PoW security? Gaining knowledge of incoming TXIDs
is nicely solved the payment protocol.
Are there enough use cases that critically depend on pub key hashes
being used on-chain, to make the compact header chain worth its costs?
Regards,
Tomas
^ permalink raw reply [flat|nested] 39+ messages in thread
* Re: [bitcoin-dev] BIP Proposal: Compact Client Side Filtering for Light Clients
2017-06-01 19:01 [bitcoin-dev] BIP Proposal: Compact Client Side Filtering for Light Clients Olaoluwa Osuntokun
` (5 preceding siblings ...)
2017-06-08 9:50 ` Tomas
@ 2017-06-19 11:58 ` Andreas Schildbach
2017-06-19 12:26 ` bfd
6 siblings, 1 reply; 39+ messages in thread
From: Andreas Schildbach @ 2017-06-19 11:58 UTC (permalink / raw)
To: bitcoin-dev
I'm not sure if this has been brought up elsewhere in this thread.
This proposal doesn't seem to be a complete replacement of BIP37: It
doesn't provide a filter for unconfirmed transactions like BIP37 does.
That means that most light clients will continue to use BIP37 even if
they may use this BIP as a supplement. Otherwise users would not get
timely notification of incoming payments any more.
On 06/01/2017 09:01 PM, Olaoluwa Osuntokun via bitcoin-dev wrote:
> Hi y'all,
>
> Alex Akselrod and I would like to propose a new light client BIP for
> consideration:
> * https://github.com/Roasbeef/bips/blob/master/gcs_light_client.mediawiki
> [...]
^ permalink raw reply [flat|nested] 39+ messages in thread* Re: [bitcoin-dev] BIP Proposal: Compact Client Side Filtering for Light Clients
2017-06-19 11:58 ` Andreas Schildbach
@ 2017-06-19 12:26 ` bfd
2017-06-19 15:15 ` Tom Zander
` (2 more replies)
0 siblings, 3 replies; 39+ messages in thread
From: bfd @ 2017-06-19 12:26 UTC (permalink / raw)
To: Andreas Schildbach; +Cc: bitcoin-dev
Several times. It's been debated if unconfirmed transactions are
necessary, methods of doing more private filtering have been suggested,
along with simply not filtering unconfirmed transactions at all. My
collected data suggests that there is very little use of BIP37 at
present, based on incoming connections to nodes I know end up in the DNS
seed responses (no "SPV" clients do their own peer management).
On 2017-06-19 12:58, Andreas Schildbach via bitcoin-dev wrote:
> I'm not sure if this has been brought up elsewhere in this thread.
>
> This proposal doesn't seem to be a complete replacement of BIP37: It
> doesn't provide a filter for unconfirmed transactions like BIP37 does.
>
> That means that most light clients will continue to use BIP37 even if
> they may use this BIP as a supplement. Otherwise users would not get
> timely notification of incoming payments any more.
>
>
> On 06/01/2017 09:01 PM, Olaoluwa Osuntokun via bitcoin-dev wrote:
>> Hi y'all,
>>
>> Alex Akselrod and I would like to propose a new light client BIP for
>> consideration:
>> *
>> https://github.com/Roasbeef/bips/blob/master/gcs_light_client.mediawiki
>> [...]
>
> _______________________________________________
> bitcoin-dev mailing list
> bitcoin-dev@lists•linuxfoundation.org
> https://lists.linuxfoundation.org/mailman/listinfo/bitcoin-dev
^ permalink raw reply [flat|nested] 39+ messages in thread* Re: [bitcoin-dev] BIP Proposal: Compact Client Side Filtering for Light Clients
2017-06-19 12:26 ` bfd
@ 2017-06-19 15:15 ` Tom Zander
2017-06-19 15:49 ` Jonas Schnelli
2017-06-19 15:43 ` Andreas Schildbach
2017-06-19 22:41 ` Gregory Maxwell
2 siblings, 1 reply; 39+ messages in thread
From: Tom Zander @ 2017-06-19 15:15 UTC (permalink / raw)
To: bitcoin-dev
On Monday, 19 June 2017 14:26:46 CEST bfd--- via bitcoin-dev wrote:
> It's been debated if [filtering of] unconfirmed transactions are
> necessary,
Why would it not be needed? Any SPV client (when used as a payment-receiver)
requires this from a simple usability point of view.
--
Tom Zander
Blog: https://zander.github.io
Vlog: https://vimeo.com/channels/tomscryptochannel
^ permalink raw reply [flat|nested] 39+ messages in thread
* Re: [bitcoin-dev] BIP Proposal: Compact Client Side Filtering for Light Clients
2017-06-19 15:15 ` Tom Zander
@ 2017-06-19 15:49 ` Jonas Schnelli
2017-06-19 15:59 ` Andreas Schildbach
2017-06-19 16:07 ` Tom Zander
0 siblings, 2 replies; 39+ messages in thread
From: Jonas Schnelli @ 2017-06-19 15:49 UTC (permalink / raw)
To: Tom Zander; +Cc: bitcoin-dev
[-- Attachment #1: Type: text/plain, Size: 534 bytes --]
Hi
> On Monday, 19 June 2017 14:26:46 CEST bfd--- via bitcoin-dev wrote:
>> It's been debated if [filtering of] unconfirmed transactions are
>> necessary,
>
> Why would it not be needed? Any SPV client (when used as a payment-receiver)
> requires this from a simple usability point of view.
I think many users would be willing ...
a) … to trade higher privacy (using client side filtering) for not having the „incoming transaction“ feature
b) – if they want 0-conf – to fetch all inved transactions
/jonas
[-- Attachment #2: Message signed with OpenPGP --]
[-- Type: application/pgp-signature, Size: 833 bytes --]
^ permalink raw reply [flat|nested] 39+ messages in thread
* Re: [bitcoin-dev] BIP Proposal: Compact Client Side Filtering for Light Clients
2017-06-19 15:49 ` Jonas Schnelli
@ 2017-06-19 15:59 ` Andreas Schildbach
2017-06-19 16:22 ` Jonas Schnelli
2017-06-19 16:07 ` Tom Zander
1 sibling, 1 reply; 39+ messages in thread
From: Andreas Schildbach @ 2017-06-19 15:59 UTC (permalink / raw)
To: bitcoin-dev
On 06/19/2017 05:49 PM, Jonas Schnelli via bitcoin-dev wrote:
>>> It's been debated if [filtering of] unconfirmed transactions are
>>> necessary,
>>
>> Why would it not be needed? Any SPV client (when used as a payment-receiver)
>> requires this from a simple usability point of view.
>
>
> I think many users would be willing ...
> a) … to trade higher privacy (using client side filtering) for not having the „incoming transaction“ feature
> b) – if they want 0-conf – to fetch all inved transactions
Another number: I'm answering dozens of support inquiries about
delayed/missing transactions per day. Over the 7 years of Bitcoin
Wallet's existence, I estimate about 50000 inquiries.
On the other hand, I remember only 1 (one) inquiry about the privacy
problems of BIP37 (or privacy at all).
From a regular user's point of view, privacy is non-issue. Sure,
everyone would take it for free, but certainly not if it a) delays
incoming payments or b) quickly eats up your traffic quota.
^ permalink raw reply [flat|nested] 39+ messages in thread
* Re: [bitcoin-dev] BIP Proposal: Compact Client Side Filtering for Light Clients
2017-06-19 15:59 ` Andreas Schildbach
@ 2017-06-19 16:22 ` Jonas Schnelli
2017-06-19 16:36 ` adiabat
2017-06-20 7:03 ` Eric Voskuil
0 siblings, 2 replies; 39+ messages in thread
From: Jonas Schnelli @ 2017-06-19 16:22 UTC (permalink / raw)
To: Andreas Schildbach; +Cc: bitcoin-dev
[-- Attachment #1: Type: text/plain, Size: 1362 bytes --]
>
> On the other hand, I remember only 1 (one) inquiry about the privacy
> problems of BIP37 (or privacy at all).
IMO privacy its something developers should make sure users have it.
Also, I think, todays SPV wallets should make users more aware of the possible privacy implications.
Do users know, if they pay for a good in a shop while consuming the shops WIFI, that the shop-owner as well as the ISP can use that data to combine it with the user profile (and ~ALL FUTURE purchases you do with the same wallet IN ANY LOCATION online or in-person)?
Do users know, that ISPs (cellular; including Google) can completely link the used Bitcoin wallet (again: all purchase including future ones) with the to the ISP well known user profile including credit-card data and may sell the Bitcoin data to any other data mining company?
If you use BIP37, you basically give your transaction history (_ALL TRANSACTIONS_ including transactions in future) to everyone.
>
> From a regular user's point of view, privacy is non-issue. Sure,
> everyone would take it for free, but certainly not if it a) delays
> incoming payments or b) quickly eats up your traffic quota.
This may be true because they are not aware of the ramification and I don’t think client side filtering is a drop-in replacement for todays, smartphone SPV-model.
/jonas
[-- Attachment #2: Message signed with OpenPGP --]
[-- Type: application/pgp-signature, Size: 833 bytes --]
^ permalink raw reply [flat|nested] 39+ messages in thread
* Re: [bitcoin-dev] BIP Proposal: Compact Client Side Filtering for Light Clients
2017-06-19 16:22 ` Jonas Schnelli
@ 2017-06-19 16:36 ` adiabat
2017-06-19 20:49 ` Andreas Schildbach
2017-06-20 7:03 ` Eric Voskuil
1 sibling, 1 reply; 39+ messages in thread
From: adiabat @ 2017-06-19 16:36 UTC (permalink / raw)
To: Bitcoin Protocol Discussion
This has been brought up several times in the past, and I agree with
Jonas' comments about users being unaware of the privacy losses due to
BIP37. One thing also mentioned before but not int he current thread
is that the entire concept of SPV is not applicable to unconfirmed
transactions. SPV uses the fact that miners have committed to a
transaction with work to give the user an assurance that the
transaction is valid; if the transaction were invalid, it would be
costly for the miner to include it in a block with valid work.
Transactions in the mempool have no such assurance, and are costlessly
forgeable by anyone, including your ISP. I wasn't involved in any
debate over BIP37 when it was being written up, so I don't know how
mempool filtering got in, but it never made any sense to me. The fact
that lots of lite clients are using this is a problem as it gives
false assurance to users that there is a valid but yet-to-be-confirmed
transaction sending them money.
-Tadge
^ permalink raw reply [flat|nested] 39+ messages in thread
* Re: [bitcoin-dev] BIP Proposal: Compact Client Side Filtering for Light Clients
2017-06-19 16:36 ` adiabat
@ 2017-06-19 20:49 ` Andreas Schildbach
0 siblings, 0 replies; 39+ messages in thread
From: Andreas Schildbach @ 2017-06-19 20:49 UTC (permalink / raw)
To: bitcoin-dev
Most SPV wallets make it quite clear that unconfirmed transactions are
just that.
On 06/19/2017 06:36 PM, adiabat via bitcoin-dev wrote:
> This has been brought up several times in the past, and I agree with
> Jonas' comments about users being unaware of the privacy losses due to
> BIP37. One thing also mentioned before but not int he current thread
> is that the entire concept of SPV is not applicable to unconfirmed
> transactions. SPV uses the fact that miners have committed to a
> transaction with work to give the user an assurance that the
> transaction is valid; if the transaction were invalid, it would be
> costly for the miner to include it in a block with valid work.
>
> Transactions in the mempool have no such assurance, and are costlessly
> forgeable by anyone, including your ISP. I wasn't involved in any
> debate over BIP37 when it was being written up, so I don't know how
> mempool filtering got in, but it never made any sense to me. The fact
> that lots of lite clients are using this is a problem as it gives
> false assurance to users that there is a valid but yet-to-be-confirmed
> transaction sending them money.
>
> -Tadge
>
^ permalink raw reply [flat|nested] 39+ messages in thread
* Re: [bitcoin-dev] BIP Proposal: Compact Client Side Filtering for Light Clients
2017-06-19 16:22 ` Jonas Schnelli
2017-06-19 16:36 ` adiabat
@ 2017-06-20 7:03 ` Eric Voskuil
1 sibling, 0 replies; 39+ messages in thread
From: Eric Voskuil @ 2017-06-20 7:03 UTC (permalink / raw)
To: Jonas Schnelli; +Cc: bitcoin-dev, Andreas Schildbach
The reason that BIP37 presents a long list of problems is that it is a client-server scenario wedged into a peer-to-peer network. The only possible excuse for this design was implementation shortcut.
As this thread and others demonstrate, reproducing this design flaw will not eliminate the problems. The fact that there are many wallets dependent upon it is an unfortunate consequence of the original sin, but is not likely to last. There is no rationale for node operators to support wallets apart from their own. As a node implementer interested in privacy, security and scalability, I would never waste the time to code BIP37, or and client-server feature into the P2P protocol, especially one that delegates some aspect of validation.
Other nodes (servers) provide independent, securable, client-server interfaces. Many of these are made available as community servers for use at no charge. They could also provide mechanisms for operator payment without polluting the P2P network.
However as a community we should be working toward full node wallets. A secured personal node/server can support remote mobile wallets with security, privacy and no wasted bandwidth. And if we avoid counterproductive increases in blockchain growth rate, full nodes will eventually be able to run on mobile platforms with no difficulty whatsoever. A wallet that delegates full validation to node operators is just another centralization pressure that we do not need.
e
On Jun 19, 2017, at 6:22 PM, Jonas Schnelli via bitcoin-dev <bitcoin-dev@lists•linuxfoundation.org> wrote:
>>
>> On the other hand, I remember only 1 (one) inquiry about the privacy
>> problems of BIP37 (or privacy at all).
>
> IMO privacy its something developers should make sure users have it.
> Also, I think, todays SPV wallets should make users more aware of the possible privacy implications.
>
> Do users know, if they pay for a good in a shop while consuming the shops WIFI, that the shop-owner as well as the ISP can use that data to combine it with the user profile (and ~ALL FUTURE purchases you do with the same wallet IN ANY LOCATION online or in-person)?
>
> Do users know, that ISPs (cellular; including Google) can completely link the used Bitcoin wallet (again: all purchase including future ones) with the to the ISP well known user profile including credit-card data and may sell the Bitcoin data to any other data mining company?
>
> If you use BIP37, you basically give your transaction history (_ALL TRANSACTIONS_ including transactions in future) to everyone.
>
>
>>
>> From a regular user's point of view, privacy is non-issue. Sure,
>> everyone would take it for free, but certainly not if it a) delays
>> incoming payments or b) quickly eats up your traffic quota.
>
> This may be true because they are not aware of the ramification and I don’t think client side filtering is a drop-in replacement for todays, smartphone SPV-model.
>
>
> /jonas
> _______________________________________________
> bitcoin-dev mailing list
> bitcoin-dev@lists•linuxfoundation.org
> https://lists.linuxfoundation.org/mailman/listinfo/bitcoin-dev
^ permalink raw reply [flat|nested] 39+ messages in thread
* Re: [bitcoin-dev] BIP Proposal: Compact Client Side Filtering for Light Clients
2017-06-19 15:49 ` Jonas Schnelli
2017-06-19 15:59 ` Andreas Schildbach
@ 2017-06-19 16:07 ` Tom Zander
2017-06-19 16:30 ` Jonas Schnelli
1 sibling, 1 reply; 39+ messages in thread
From: Tom Zander @ 2017-06-19 16:07 UTC (permalink / raw)
To: bitcoin-dev
On Monday, 19 June 2017 17:49:59 CEST Jonas Schnelli wrote:
> Hi
>
> > On Monday, 19 June 2017 14:26:46 CEST bfd--- via bitcoin-dev wrote:
> >> It's been debated if [filtering of] unconfirmed transactions are
> >> necessary,
> >
> > Why would it not be needed? Any SPV client (when used as a
> > payment-receiver) requires this from a simple usability point of view.
>
> I think many users would be willing ...
> a) … to trade higher privacy (using client side filtering) for not having
> the „incoming transaction“ feature b) – if they want 0-conf – to fetch
> all inved transactions
You seem to misunderstand the usecase.
If you send me a transaction, both of use are using our phones, then I need
to be able to have immediate feedback on the transaction being broadcast on
the network.
This is not about zero-conf, this is simple seeing what is happening while
it is happening.
Additionally, when the transaction that is meant for my wallet is broadcast,
I want my SPV wallet to parse and check the actual transaction.
It is not just to see that *something* was actually send, but also to be
able to see how much is being paid to me. Maybe If the transaction is marked
as RBF-able, etc.
Really basic usability: provide information to your users when you can,
should they want to, and by default on.
--
Tom Zander
Blog: https://zander.github.io
Vlog: https://vimeo.com/channels/tomscryptochannel
^ permalink raw reply [flat|nested] 39+ messages in thread
* Re: [bitcoin-dev] BIP Proposal: Compact Client Side Filtering for Light Clients
2017-06-19 16:07 ` Tom Zander
@ 2017-06-19 16:30 ` Jonas Schnelli
2017-06-19 16:38 ` Tom Zander
0 siblings, 1 reply; 39+ messages in thread
From: Jonas Schnelli @ 2017-06-19 16:30 UTC (permalink / raw)
To: Tom Zander; +Cc: bitcoin-dev
[-- Attachment #1: Type: text/plain, Size: 1520 bytes --]
>>
>> I think many users would be willing ...
>> a) … to trade higher privacy (using client side filtering) for not having
>> the „incoming transaction“ feature b) – if they want 0-conf – to fetch
>> all inved transactions
>
> You seem to misunderstand the usecase.
> If you send me a transaction, both of use are using our phones, then I need
> to be able to have immediate feedback on the transaction being broadcast on
> the network.
> This is not about zero-conf, this is simple seeing what is happening while
> it is happening.
>
> Additionally, when the transaction that is meant for my wallet is broadcast,
> I want my SPV wallet to parse and check the actual transaction.
> It is not just to see that *something* was actually send, but also to be
> able to see how much is being paid to me. Maybe If the transaction is marked
> as RBF-able, etc.
>
> Really basic usability: provide information to your users when you can,
> should they want to, and by default on.
I see this use case.
But I did receive bank wire transfers for the last decades without _immediately_ knowing that someone sent funds to me.
I personally would ALWAYS trade the higher bandwidth consumption (300MB mempool filtering) or slower notification time (maybe ~1h) for preserving privacy.
I agree, there are use cases where you want immediate notification, those use cases could probably be solved by not trowing away privacy („parsing“ all transactions and running in the background).
/jonas
[-- Attachment #2: Message signed with OpenPGP --]
[-- Type: application/pgp-signature, Size: 833 bytes --]
^ permalink raw reply [flat|nested] 39+ messages in thread
* Re: [bitcoin-dev] BIP Proposal: Compact Client Side Filtering for Light Clients
2017-06-19 12:26 ` bfd
2017-06-19 15:15 ` Tom Zander
@ 2017-06-19 15:43 ` Andreas Schildbach
2017-06-19 16:10 ` Jonas Schnelli
2017-06-19 22:41 ` Gregory Maxwell
2 siblings, 1 reply; 39+ messages in thread
From: Andreas Schildbach @ 2017-06-19 15:43 UTC (permalink / raw)
To: bitcoin-dev
Just to give you a number: based on the statistics of the Bitcoin Wallet
app there are at least 2 million wallets depending on BIP37. Not all
would need instant notification but based on the daily support enquiries
instant notificaton is the most asked property of Bitcoin.
On 06/19/2017 02:26 PM, bfd--- via bitcoin-dev wrote:
> Several times. It's been debated if unconfirmed transactions are
> necessary, methods of doing more private filtering have been suggested,
> along with simply not filtering unconfirmed transactions at all. My
> collected data suggests that there is very little use of BIP37 at
> present, based on incoming connections to nodes I know end up in the DNS
> seed responses (no "SPV" clients do their own peer management).
>
>
> On 2017-06-19 12:58, Andreas Schildbach via bitcoin-dev wrote:
>> I'm not sure if this has been brought up elsewhere in this thread.
>>
>> This proposal doesn't seem to be a complete replacement of BIP37: It
>> doesn't provide a filter for unconfirmed transactions like BIP37 does.
>>
>> That means that most light clients will continue to use BIP37 even if
>> they may use this BIP as a supplement. Otherwise users would not get
>> timely notification of incoming payments any more.
>>
>>
>> On 06/01/2017 09:01 PM, Olaoluwa Osuntokun via bitcoin-dev wrote:
>>> Hi y'all,
>>>
>>> Alex Akselrod and I would like to propose a new light client BIP for
>>> consideration:
>>> *
>>> https://github.com/Roasbeef/bips/blob/master/gcs_light_client.mediawiki
>>> [...]
>>
>> _______________________________________________
>> bitcoin-dev mailing list
>> bitcoin-dev@lists•linuxfoundation.org
>> https://lists.linuxfoundation.org/mailman/listinfo/bitcoin-dev
^ permalink raw reply [flat|nested] 39+ messages in thread
* Re: [bitcoin-dev] BIP Proposal: Compact Client Side Filtering for Light Clients
2017-06-19 15:43 ` Andreas Schildbach
@ 2017-06-19 16:10 ` Jonas Schnelli
0 siblings, 0 replies; 39+ messages in thread
From: Jonas Schnelli @ 2017-06-19 16:10 UTC (permalink / raw)
To: Andreas Schildbach; +Cc: bitcoin-dev
[-- Attachment #1: Type: text/plain, Size: 1119 bytes --]
> Just to give you a number: based on the statistics of the Bitcoin Wallet
> app there are at least 2 million wallets depending on BIP37. Not all
> would need instant notification but based on the daily support enquiries
> instant notificaton is the most asked property of Bitcoin.
Yes. Users probably like this feature and client side filtering is not a drop-in replacement for BIP37.
We should also consider:
BIP37 works, because node-operators are willing to offer that service for free (which maybe change over time).
BIP37 consumes plenty of horsepower (disk/cpu) from nodes. Filtering a couple of days of blocks (assume 1000+) eats lots of resources for something, that has no direct long-term value for Bitcoin (the filters data is unique and will be "thrown away“ [can’t be used by other peers]). Same applies for mempool (filtering mempool of a couple of hundred of mb each time the HD gap limit has been exceeded or the app gets sent to the foreground again).
Purely relying on the availability of BIP37 seems fragile to me and start to explore other ways is very reasonable.
/jonas
[-- Attachment #2: Message signed with OpenPGP --]
[-- Type: application/pgp-signature, Size: 833 bytes --]
^ permalink raw reply [flat|nested] 39+ messages in thread
* Re: [bitcoin-dev] BIP Proposal: Compact Client Side Filtering for Light Clients
2017-06-19 12:26 ` bfd
2017-06-19 15:15 ` Tom Zander
2017-06-19 15:43 ` Andreas Schildbach
@ 2017-06-19 22:41 ` Gregory Maxwell
2017-06-20 9:52 ` Tom Zander
2 siblings, 1 reply; 39+ messages in thread
From: Gregory Maxwell @ 2017-06-19 22:41 UTC (permalink / raw)
To: Bitcoin Dev
On Mon, Jun 19, 2017 at 12:26 PM, bfd--- via bitcoin-dev
<bitcoin-dev@lists•linuxfoundation.org> wrote:
> Several times. It's been debated if unconfirmed transactions are necessary,
> methods of doing more private filtering have been suggested, along with
> simply not filtering unconfirmed transactions at all. My collected data
> suggests that there is very little use of BIP37 at present, based on
> incoming connections to nodes I know end up in the DNS seed responses (no
> "SPV" clients do their own peer management).
Sending just the output addresses of each transaction would use about
1 kilobit/s of data. Sending the entire transactions would use
~14kbit/sec data. These don't seem to be a unsustainable tremendous
amount of data to use while an application is running.
Doubly so for SPV wallets which are highly vulnerable to unconfirmed
transactions, and many which last I heard testing reports on became
pretty severely corrupted once given a fake transaction.
Can someone make a case why saving no more than those figures would
justify the near total loss of privacy that filtering gives?
"Because they already do it" isn't a good argument when talking about
a new protocol feature; things which already do BIP37 will presumably
continue to already do BIP37.
^ permalink raw reply [flat|nested] 39+ messages in thread
* Re: [bitcoin-dev] BIP Proposal: Compact Client Side Filtering for Light Clients
2017-06-19 22:41 ` Gregory Maxwell
@ 2017-06-20 9:52 ` Tom Zander
2017-06-20 13:08 ` bfd
0 siblings, 1 reply; 39+ messages in thread
From: Tom Zander @ 2017-06-20 9:52 UTC (permalink / raw)
To: bitcoin-dev
On Tuesday, 20 June 2017 00:41:49 CEST Gregory Maxwell via bitcoin-dev
wrote:
> Can someone make a case why saving no more than those figures would
> justify the near total loss of privacy that filtering gives?
First, your figures are wrong and also fall out of the sky with no
justification. Can’t debunk something that is pure garbage.
Second, stating that a bloom filter is a "total loss of privacy" is equally
baseless and doesn’t need debunking.
> "Because they already do it" isn't a good argument when talking about
> a new protocol feature; things which already do BIP37 will presumably
> continue to already do BIP37.
I think you just made the case for completely rejecting this proposal based
on the fact that nobody will use it, BIP37 already exists.
Not sure if I agree with that, improvements are always useful and we should
be able to come up with replacements.
But arguing against a feature you don’t like, especiallyh one used by
millions every day, is a sad way to stiffle innovation, Greg.
--
Tom Zander
Blog: https://zander.github.io
Vlog: https://vimeo.com/channels/tomscryptochannel
^ permalink raw reply [flat|nested] 39+ messages in thread
* Re: [bitcoin-dev] BIP Proposal: Compact Client Side Filtering for Light Clients
2017-06-20 9:52 ` Tom Zander
@ 2017-06-20 13:08 ` bfd
2017-06-20 17:20 ` Adam Back
0 siblings, 1 reply; 39+ messages in thread
From: bfd @ 2017-06-20 13:08 UTC (permalink / raw)
To: Tom Zander; +Cc: bitcoin-dev
On 2017-06-20 12:52, Tom Zander via bitcoin-dev wrote:
> Second, stating that a bloom filter is a "total loss of privacy" is
> equally
> baseless and doesn’t need debunking.
>
"On the Privacy Provisions of Bloom Filters in Lightweight Bitcoin
Clients"
> We show analytically and empirically that the reliance on Bloom filters
> within existing SPV clients leaks considerable information about the
> addresses of Bitcoin users. Our results show that an SPV client who
> uses a modest number of Bitcoin addresses (e.g., < 20) risks revealing
> almost all of his addresses.
https://eprint.iacr.org/2014/763.pdf
^ permalink raw reply [flat|nested] 39+ messages in thread
* Re: [bitcoin-dev] BIP Proposal: Compact Client Side Filtering for Light Clients
2017-06-20 13:08 ` bfd
@ 2017-06-20 17:20 ` Adam Back
0 siblings, 0 replies; 39+ messages in thread
From: Adam Back @ 2017-06-20 17:20 UTC (permalink / raw)
To: bfd; +Cc: Bitcoin Dev
Also Jonas Nick gave a fairly comprehensive presentation on privacy
leaks in bitcoin protocol including SPV and bloom query problem
specifics:
https://www.youtube.com/watch?v=HScK4pkDNds
Adam
On 20 June 2017 at 14:08, bfd--- via bitcoin-dev
<bitcoin-dev@lists•linuxfoundation.org> wrote:
> On 2017-06-20 12:52, Tom Zander via bitcoin-dev wrote:
>>
>> Second, stating that a bloom filter is a "total loss of privacy" is
>> equally
>> baseless and doesn’t need debunking.
>>
>
> "On the Privacy Provisions of Bloom Filters in Lightweight Bitcoin Clients"
>
>> We show analytically and empirically that the reliance on Bloom filters
>> within existing SPV clients leaks considerable information about the
>> addresses of Bitcoin users. Our results show that an SPV client who uses a
>> modest number of Bitcoin addresses (e.g., < 20) risks revealing almost all
>> of his addresses.
>
>
> https://eprint.iacr.org/2014/763.pdf
>
> _______________________________________________
> bitcoin-dev mailing list
> bitcoin-dev@lists•linuxfoundation.org
> https://lists.linuxfoundation.org/mailman/listinfo/bitcoin-dev
^ permalink raw reply [flat|nested] 39+ messages in thread