Preuve de vie

[Hugo Mercier]

Bonjour,

Je découvre le projet open-udc avec intérêt.
Je n'ai pas encore tout lu en détails.

J'ai cependant déjà une question. Si j'ai bien suivi, tout le système repose sur une distribution équitable d'argent, ce qui nécessite d'identifier de manière unique chaque personne.
Et ceci repose sur la signature de clés GPG avec des personnes déjà dans le système, à intervalle régulier.
Il est mentionné quelque part qu'il faut qu'une partie importante du réseau vérifie ces preuves de vie (il me semble avoir lu 2/3, mais je ne retrouve plus). Comment cela fonctionne t-il dans le détail ? Il faut que la carte d'identité de chaque personne soit vérifié par 2/3 du réseau ?

[Stéphane Laborde]

Tout d'abord il est convient de pr�ciser que "le r�seau" signifie "le
r�seau des membres humains" et pas du tout un quelconque "r�seau de
machines". Le r�seau technique est donc subordonn� au r�seau humain.

Ensuite une fois bien pr�cis� ce point fondamental il convient de
r�fl�chir � la question suivante :

"Quelle la preuve de vie des 64 millions de Fran�ais" ? Y-a-t-il
v�ritablement 64 millions de Fran�ais en vie ? Qui peut l'assurer de
fa�on certaine ? Y-a-t-il 2/3 des Fran�ais ayant v�ritablement v�rifi�
la preuve de vie de chacun des 64 millions pr�tendument "en vie ?

R�fl�chir donc � cette question et aux r�ponses qu'il convient
d'apporter sur comment chacun des 64 millions de Fran�ais est "confiant"
dans le fait qu'il vit bien 64 millions de Fran�ais, permet de
comprendre comment cette confiance peut s'�tablir ou ne pas s'�tablir.

Ce n'est qu'alors qu'on peut v�ritablement comprendre la solution
qu'apporte OpenUDC � cette question.

Le 09/03/2013 11:37, Hugo Mercier a �crit :
> Bonjour,
>
> Je d�couvre le projet open-udc avec int�r�t.
> Je n'ai pas encore tout lu en d�tails.
>
> J'ai cependant d�j� une question. Si j'ai bien suivi, tout le syst�me
> repose sur une distribution �quitable d'argent, ce qui n�cessite
> d'identifier de mani�re unique chaque personne.
> Et ceci repose sur la signature de cl�s GPG avec des personnes d�j�
> dans le syst�me, � intervalle r�gulier.
> Il est mentionn� quelque part qu'il faut qu'une partie importante du
> r�seau v�rifie ces preuves de vie (il me semble avoir lu 2/3, mais je
> ne retrouve plus). Comment cela fonctionne t-il dans le d�tail ? Il
> faut que la carte d'identit� de chaque personne soit v�rifi� par 2/3
> du r�seau ?
> --
> Vous recevez ce message, car vous �tes abonn� au groupe Google
> Groupes open-udc.
> Pour vous d�sabonner de ce groupe et ne plus recevoir d'e-mails le
> concernant, envoyez un e-mail � l'adresse
> open-udc+u...@googlegroups.com.
> Pour plus d'options, visitez le site
> https://groups.google.com/groups/opt_out .
>
>

[Hugo Mercier]

Le samedi 9 mars 2013 12:42:42 UTC+1, Stephane Laborde a écrit :

Tout d'abord il est convient de pr�ciser que "le r�seau" signifie "le
r�seau des membres humains" et pas du tout un quelconque "r�seau de
machines". Le r�seau technique est donc subordonn� au r�seau humain.

Oui, tout à fait.
 

Ensuite une fois bien pr�cis� ce point fondamental il convient de
r�fl�chir � la question suivante :

"Quelle la preuve de vie des 64 millions de Fran�ais" ? Y-a-t-il
v�ritablement 64 millions de Fran�ais en vie ? Qui peut l'assurer de
fa�on certaine ? Y-a-t-il 2/3 des Fran�ais ayant v�ritablement v�rifi�
la preuve de vie de chacun des 64 millions pr�tendument "en vie ?

Je ne sais pas s'il y a exactement 64 millions de français en vie. D'ailleurs, si on voulait compter le nombre de gens qui participent réellement à l'économie française, ça serait sans doute un peu plus (puisque ça comprend les sans papiers justement).
Ensuite, non il n'y a jamais 2/3 des personnes qui vérifient que les autres existent ... pourquoi le ferions nous ? Et si on devait le faire, ce serait très pénible. C'était justement ma question ...
 

R�fl�chir donc � cette question et aux r�ponses qu'il convient
d'apporter sur comment chacun des 64 millions de Fran�ais est "confiant"
dans le fait qu'il vit bien 64 millions de Fran�ais, permet de
comprendre comment cette confiance peut s'�tablir ou ne pas s'�tablir.

Ce n'est qu'alors qu'on peut v�ritablement comprendre la solution
qu'apporte OpenUDC � cette question.

Humm ... merci, mais ça reste énigmatique.
Je vois OpenUDC comme un projet de nouvelle monnaie et d'économie. Il m'a semblé comprendre qu'il n'y avait pas de raison qu'elle se limite au territoire français. Même si je peux être relativement confiant dans le recensement et le processus d'attribution de cartes d'identités dans mon pays, je ne peux rien dire sur les autres pays.
Si le projet de monnaie repose sur la distribution fréquente d'un certain nombre d'unités (de grains), à chaque personne, il faut s'assurer que personne ne puisse se voir attribuer plusieurs fois la répartition, non ? J'ai compris que pour répondre à cette question, on se reposait sur la signature de clés GPG, mais je n'ai pas encore bien compris comment on se protégeait contre la triche.

[jbar]

On Sat, 9 Mar 2013 05:36:59 -0800 (PST)
Hugo Mercier <hmerc...@gmail.com> wrote:

> Humm ... merci, mais ça reste énigmatique.
> Je vois OpenUDC comme un projet de nouvelle monnaie et d'économie. Il m'a
> semblé comprendre qu'il n'y avait pas de raison qu'elle se limite au
> territoire français. Même si je peux être relativement confiant dans le
> recensement et le processus d'attribution de cartes d'identités dans mon
> pays, je ne peux rien dire sur les autres pays.
> Si le projet de monnaie repose sur la distribution fréquente d'un certain
> nombre d'unités (de grains), à chaque personne, il faut s'assurer que
> personne ne puisse se voir attribuer plusieurs fois la répartition, non ?
> J'ai compris que pour répondre à cette question, on se reposait sur la
> signature de clés GPG, mais je n'ai pas encore bien compris comment on se
> protégeait contre la triche.
>

"confiant dans le (...) processus d'attribution de cartes d'identités
dans mon pays" ... relativement en effet !

C'est aussi pour ce relatif (qui l'est encore probablement encore plus
pour d'autres pays que le notre), et par le fait qu'il y des freins, à
la fois psychologique et technologique (solide fiable et grand public),
énormes pour s'appuyer sur la biométrie, que nous n'acceptons plus que
des identifiants (udid2) en clair, et ignorons les identifiants hachés.

Mais pour revenir au coeur de la protection, la toile de confiance
OpenPGP permet juste de relier un certificat OpenPGP à d'autre (par la
signature). Mais ce que l'on pourrait appeler la valeur d'une liaison
est relative à chaque individus, d'ailleurs quand GnuPG génére sa
trustDB, on peut intervenir pour chaque certificat sur la
"force"/"valeur" de sa signature. (La question est : "faites-vous
confiance à ce certificat pour signer d'autre certificat").

On peut même affiner plus précisement une valeur à chaque liaison.
C'est un sujet assez intéressant qui pourrait servir à de nombreuses
thèses et développements, hélas dont l'éxistence est peu consistante
aujourd'hui.

GnuPG attribue des valeurs de confiance, différentes pour chacun, sur
chacun des certificats que l'on connait.
Sans s'attarder sur les calculs par défauts générant ces valeurs, GnuPG
reconnait un certicat comme "valide", s'il est signé par 3 autres
certificats déjà reconnu comme "valide" (localement, par l'instance de
GnuPG utilisé chez chacun).

Les premiers certificats reconnus comme "valide" étant par défault,
celui/ceux dont il connaît la clé privé (donc à priori le/les notre),
et ceux que l'ont a signé nous même.

On comprend donc l'importance des keysigning parties, et pourquoi j'ai
écrit sur wikipedia que "la toile de confiance est relative à un
individu".

Maintenant les fiches de créations (et donc de recensement) sont
validées par des humains (sorte d'admin/recenseur reconnus comme tel
dans la fiche de création précédente) qui ont les compétences technique
pour étudier la toile de confiance et qui doivent se mettre d'accord
(c'est la que la règle, non figée, des 2/3 se retrouve), avant de
proposer une nouvelle fiche de création (et donc de recensement) aux
noeuds OpenUDC.

C'est là (sur eux) que réside à la fois la faille et la sécurité ! ;-)

Le recensement peut donc être vu donc une combinaison, un "merge", à
partir de leurs valeurs de confiance sur les différents certificats de
la toile de confiance.

Pour tricher, la question est donc "peut-on tromper leur (toile de)
confiance" ?
Ces recenseurs s'appuient sur des algos (qui seront j'espère différents)
que l'on pourra bien sûr tromper. Mais rien que pour l'algo par
défauts avec ses paramètres par défauts (ceux de GnuPG sans "tunning"
de l'utilisateur), on devine que c'est très difficile, à moins
probablement d'une corruption généralisé... ("contrat social",
toussa...)

Et quand bien même certains y arriverait (sans corruption généralisé),
peut-on tromper la confiance tout le temps ?

Pour rappel: ne rentre que les "parrainés", le protocole est
ouvert, les noeuds gardent les logs qu'ils veulent,
n'importe qui (de recensé) peut faire tourner son propre noeud.
Donc toutes les données peuvent être accessibles, et à priori : "tout
est lié", et "tout est transparent".
Et donc à postériori : le risque de de se faire démasquer et donc
sanctionner (exclure) est très conséquent.

_Note1_: Le protocole permet que tout le monde puisse être
"recenseur" (votant pour valider une fiche de création) mais ce ne
sera pas possible dans les années à venir, pour cause de limitation
technique.

_Note2_: Il n'existe actuellement pas d'algo pour "merger" les toiles de
confiances relatives, et même si il en existerait, il faudra alors
qu'une majorité (ex. 2/3) s'y mettent d'accord ainsi que sur ses
paramètres...

_Note3_: On pourra très bien imaginer à l'avenir que les "recenseurs"
soient tiré au sort parmis les recensés

--
jbar <jeanjacqu...@gmail.com>

[olivier auber]

Merci jbar, c'est super clair.
Question subsidiaire : que se passe-t-il quand quelqu'un casse sa pipe ?

O

--
Olivier Auber
Evolution, Complexity and COgnition group (ECCO) & Global Brain Institute
Free University of Brussels (VUB) http://ecco.vub.ac.be
Paris +33675038880 / Bruxelles +32492050697
http://perspective-numerique.net
http://twitter.com/#!/OlivierAuber

[Stéphane Laborde]

Le 09/03/2013 14:36, Hugo Mercier a écrit :

Le samedi 9 mars 2013 12:42:42 UTC+1, Stephane Laborde a écrit :

Tout d'abord il est convient de pr�ciser que "le r�seau" signifie "le
r�seau des membres humains" et pas du tout un quelconque "r�seau de
machines". Le r�seau technique est donc subordonn� au r�seau humain.

Oui, tout à fait.
 

Ensuite une fois bien pr�cis� ce point fondamental il convient de
r�fl�chir � la question suivante :

"Quelle la preuve de vie des 64 millions de Fran�ais" ? Y-a-t-il
v�ritablement 64 millions de Fran�ais en vie ? Qui peut l'assurer de
fa�on certaine ? Y-a-t-il 2/3 des Fran�ais ayant v�ritablement v�rifi�
la preuve de vie de chacun des 64 millions pr�tendument "en vie ?

Je ne sais pas s'il y a exactement 64 millions de français en vie. D'ailleurs, si on voulait compter le nombre de gens qui participent réellement à l'économie française, ça serait sans doute un peu plus (puisque ça comprend les sans papiers justement).

Ensuite, non il n'y a jamais 2/3 des personnes qui vérifient que les autres existent ... pourquoi le ferions nous ? Et si on devait le faire, ce serait très pénible. C'était justement ma question ...

Voilà. Plus que pénible, c'est impossible dans le laps de temps d'une vie humaine que de faire cela, d'autant qu'il faudrait le faire chaque année pour tenir compte des morts et des naissances.

R�fl�chir donc � cette question et aux r�ponses qu'il convient
d'apporter sur comment chacun des 64 millions de Fran�ais est "confiant"
dans le fait qu'il vit bien 64 millions de Fran�ais, permet de
comprendre comment cette confiance peut s'�tablir ou ne pas s'�tablir.

Ce n'est qu'alors qu'on peut v�ritablement comprendre la solution
qu'apporte OpenUDC � cette question.

Humm ... merci, mais ça reste énigmatique.
Je vois OpenUDC comme un projet de nouvelle monnaie et d'économie. Il m'a semblé comprendre qu'il n'y avait pas de raison qu'elle se limite au territoire français.

Il n'y a pas de raison qu'il n'y ait pas de raison. Chaque communauté monétaire choisit librement les règles qui la régissent.

S'il s'avère qu'il est en fait impossible ou non-compatible avec la confiance que de savoir s'il y a plus de N membres d'une communauté monétaire en vie, alors on peut ne pas s'étonner qu'une communauté monétaire choisisse de ne pas dépasser N membres.

Encore faut-il comprendre l'essence de ce problème.

Même si je peux être relativement confiant dans le recensement et le processus d'attribution de cartes d'identités dans mon pays, je ne peux rien dire sur les autres pays.
Si le projet de monnaie repose sur la distribution fréquente d'un certain nombre d'unités (de grains), à chaque personne, il faut s'assurer que personne ne puisse se voir attribuer plusieurs fois la répartition, non ? J'ai compris que pour répondre à cette question, on se reposait sur la signature de clés GPG, mais je n'ai pas encore bien compris comment on se protégeait contre la triche.

Triche ou illusion ont les mêmes racines. Si d'aventure un homme croyait qu'il était seul vivant sur terre comme être humain, tous les autres étant des humanoïdes très bien imités, avec qui donc ou avec quoi donc enisagerait-il sérieusement d'établir une monnaie basée sur l'homme ?

La monnaie ayant pour base la confiance, une monnaie établie sur l'homme suppose que les hommes qui l'adoptent ont un degré de confiance suffisant les uns envers les autres, sans quoi la monnaie serait rejetée comme non-fiable relativement à ce envers quoi elle est établie.

D'aucuns ont choisi de faire confiance en un métal inanimé nommé "or", et se sont déclarés floués le jour où ils se sont aperçus que ce qu'ils croyaient être le "gage" de cette monnaie s'avérait être pure illusion, à la fois en tant que pseudo-valeur, et à la fois en tant que quantité "de réserve" et à la fois en tant qu'outil véritablement fiable pour assurer les libertés économiques.

OpenUDC est établie sur des fondements symétriques entre les hommes.

La base de mise en relation des hommes adoptant une telle monnaie est une reconnaissance symétrique basée sur la théorie des graphes.

Est-il possible d'aller plus loin en comprenant le fait que la confiance est l'antinomie parfaite de la défiance ?

Celui qui est empli de défiance envers ses semblables, qui ne fait pas confiance à ses semblables pour lui assurer que tels ou tels hommes sont bien vivants, en qui ou en quoi pourra-t-il jamais avoir confiance ?

S'il a confiance en des machines par exemple, alors il devra adopter une monnaie basée sur des machines, et comprendre qu'il aura une économie avec des machines avec qui échanger.

Celui qui a confiance en l'homme et souhaite échanger avec l'homme doit considérer la confiance en l'homme.

C'est ainsi que la cohérence s'établit.

[Stéphane Laborde]

Le 09/03/2013 19:33, olivier auber a �crit :

> Merci jbar, c'est super clair.
> Question subsidiaire : que se passe-t-il quand quelqu'un casse sa pipe ?
>
> O
>

Les certificats ont une date de p�remption.

Imaginons que cette dur�e soit de 2 ans, et qu'il y ait besoin de 5
signature valides.

Cela signifie 5 signatures valides de moins de 2 ans chacune. la
premi�re qui, arrivant � expiration n'est pas renouvel�e, signifie pour
la communaut� que le membre cesse d'�tre actif.

Quelle que soit la raison, dont celle d'�tre mort bien entendu.

[Hugo Mercier]

Merci pour ces éléments de réponse. Je crois qu'il faut que je regarde concrètement comment se déroule les différentes étapes pour y voir plus clair.

[Hugo Mercier]

Bonjour,

Désolé je déterre un vieux thread, mais je tente toujours de comprendre ...

Le samedi 9 mars 2013 19:23:58 UTC+1, Jbar a écrit :

Ok, effectivement chaque toile de confiance est relative.
 

Maintenant les fiches de créations (et donc de recensement) sont
validées par des humains (sorte d'admin/recenseur reconnus comme tel
dans la fiche de création précédente) qui ont les compétences technique
pour étudier la toile de confiance et qui doivent se mettre d'accord
(c'est la que la règle, non figée, des 2/3 se retrouve), avant de
proposer une nouvelle fiche de création (et donc de recensement) aux
noeuds OpenUDC.

Si je comprends bien : il peut y avoir de nouveaux individus qui entrent dans le réseau à tout moment.
Il faut qu'ils soient "parrainés", i.e. qu'ils soient signés par un certain nombre de personnes déjà membres.
Il y a une étape à chaque distribution du dividende (tous les mois disons) où un ensemble de recenseurs croisent leurs toiles de confiance et valident qui fait partie de la prochaine distribution,
et en particulier si les nouveaux individus font bien partie de cette "toile de confiance mergée".
Si à cette étape les recenseurs redemendaient des preuves d'existence aux nouveaux membres, ça pourrait être trop contraignant en pratique (éloignement géographique par exemple). Donc ils se basent sur leur toile de confiance. C'est ça ?

Donc qu'est-ce qui empêche un certain nombre de personnes de se mettre d'accord pour créer une identité factice. Ils la signent tous, attendent le versement de son dividende et se le partage.

 

C'est là (sur eux) que réside à la fois la faille et la sécurité ! ;-)

Le recensement peut donc être vu donc une combinaison, un "merge", à
partir de leurs valeurs de confiance sur les différents certificats de
la toile de confiance.

Pour tricher, la question est donc "peut-on tromper leur (toile de)
confiance" ?
Ces recenseurs s'appuient sur des algos (qui seront j'espère différents)
que l'on pourra bien sûr tromper. Mais rien que pour l'algo par
défauts avec ses paramètres par défauts (ceux de GnuPG sans "tunning"
de l'utilisateur), on devine que c'est très difficile, à moins
probablement d'une corruption généralisé... ("contrat social",
toussa...)

"On devine que c'est trés difficile". De quels algos s'agit-il exactement ?
 

Et quand bien même certains y arriverait (sans corruption généralisé),
peut-on tromper la confiance tout le temps ?

Non, c'est sûr, le jour où on se fait prendre, il y a de forte chance qu'on soit plus ou moins bani du réseau. Mais avant de se faire prendre, on pourrait en profiter ...
 

Pour rappel: ne rentre que les "parrainés", le protocole est
ouvert, les noeuds gardent les logs qu'ils veulent,
n'importe qui (de recensé) peut faire tourner son propre noeud.
Donc toutes les données peuvent être accessibles, et à priori : "tout
est lié", et "tout est transparent".
Et donc à postériori : le risque de  de se faire démasquer et donc
sanctionner (exclure) est très conséquent.

Cela ne résoud pas il me semble le problème d'une fausse identité. Elle est connue de tous, transparente, mais il est quand même difficile de vérifier qu'elle est fausse sans vérifier tout le monde.
 

_Note1_: Le protocole permet que tout le monde puisse être
"recenseur" (votant pour valider une fiche de création) mais ce ne
sera pas possible dans les années à venir, pour cause de limitation
technique.

Quelle limitation technique ?
 

_Note2_: Il n'existe actuellement pas d'algo pour "merger" les toiles de
confiances relatives, et même si il en existerait, il faudra alors
qu'une majorité (ex. 2/3) s'y mettent d'accord ainsi que sur ses
paramètres...

C'est ce point qui m'intéresse effectivement.

[Stéphane Laborde]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA256

I answer quickly to some points. Please use english.

Le 13/12/2013 18:35, Hugo Mercier a �crit :
> Si � cette �tape les recenseurs redemendaient des preuves
> d'existence aux nouveaux membres, �a pourrait �tre trop
> contraignant en pratique (�loignement g�ographique par exemple).
> Donc ils se basent sur leur toile de confiance. C'est �a ?

Only individuals near other ones can trust they OpenPGP keys.
>
> Donc qu'est-ce qui emp�che un certain nombre de personnes de se
> mettre d'accord pour cr�er une identit� factice. Ils la signent

> tous, attendent le versement de son dividende et se le partage.
>

Only individuals trusted by the community algorithm can recognize a
new one.

>
>
> "On devine que c'est tr�s difficile". De quels algos s'agit-il
> exactement ?

Graph Theory, study OpenPGP WoT graph algorithm by default.
>
>
> Non, c'est s�r, le jour o� on se fait prendre, il y a de forte
> chance qu'on soit plus ou moins bani du r�seau. Mais avant de se

> faire prendre, on pourrait en profiter ...

Your money units are key-signed. So you can individually accept or
reject some money units you don't trust.

So no, you cannot "play" without trust. Not only your identity but
also your money units are signed.
>
> Cela ne r�soud pas il me semble le probl�me d'une fausse identit�.
> Elle est connue de tous, transparente, mais il est quand m�me
> difficile de v�rifier qu'elle est fausse sans v�rifier tout le
> monde.
>

The WoT itself limit the expansion of false money units.

>
>
> Quelle limitation technique ?

All technical improvements are possible, including special OpenUDC WoT
trust algorithm, need to be developped and improved.
>
>
> C'est ce point qui m'int�resse effectivement.
>

You are welcome to participate to OpenUDC 0.4 next version, please
contact jbar for more technical specifications.

> -- OpenUDC aims to provide a open standard for Universal Dividend
> Crypto-Currencies.
>
> homepage: http://openudc.org --- git's home:
> https://github.com/Open-UDC/open-udc.git --- Multi User Chat:
> open...@muc.jappix.com. --- You received this message because you
> are subscribed to the Google Groups "OpenUDC" group. To unsubscribe
> from this group and stop receiving emails from it, send an email to
> openudc+u...@googlegroups.com. For more options, visit
> https://groups.google.com/groups/opt_out.
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.11 (GNU/Linux)
Comment: Using GnuPG with Thunderbird - http://www.enigmail.net/

iQEcBAEBCAAGBQJSq0tWAAoJELdo2ekCUDkj8CEIAJksi/EVzB357amkCqviNdlv
tSWHnuOaNR3frM01ZzqvgJ2Vtl7tLKL9VmvM1AzuGt1XZ1Ra1gA2cvSuwrGf+P6+
XCrVLEGa+txOUVkHc+rozfX5yiwOgqfc5p4lJXp8qWXxqB74naWwdf0ykSdAZWft
K+aoukHRX7iU3WSQBd2fGPH9HOJLPHpwU231V+E0KGamDFmuhCLazPm70ko11zVR
06aSGd+Wlg0+z2NmQKyqZCc716dcapoZtUp6TEMtLYLuLT/V9+QxB1ancXVPM289
fRJtuoHHfX592MF8Vsd0fwPxTQnU8VNMLqW6pz3AD5NE9XrAIyfc+qEdsb4TgME=
=8Bv+
-----END PGP SIGNATURE-----

[Hugo Mercier]

Le 13/12/2013 19:00, St�phane Laborde a �crit :

> I answer quickly to some points. Please use english.
>

Well ... why not.

> Le 13/12/2013 18:35, Hugo Mercier a �crit :
>> Si � cette �tape les recenseurs redemendaient des preuves
>> d'existence aux nouveaux membres, �a pourrait �tre trop
>> contraignant en pratique (�loignement g�ographique par exemple).
>> Donc ils se basent sur leur toile de confiance. C'est �a ?
>
> Only individuals near other ones can trust they OpenPGP keys.

What do you mean by "near other ones" ? I guess you mean near from each
other in the web of trust.
So each member has to trust every other members ?

>
>> Donc qu'est-ce qui emp�che un certain nombre de personnes de se
>> mettre d'accord pour cr�er une identit� factice. Ils la signent
>> tous, attendent le versement de son dividende et se le partage.
>
>
> Only individuals trusted by the community algorithm can recognize a
> new one.

Sure. And what if these trusted individuals collude to create a fake
identity ?

>
>
>
>> "On devine que c'est tr�s difficile". De quels algos s'agit-il
>> exactement ?
>
> Graph Theory, study OpenPGP WoT graph algorithm by default.
>

What "graph algorithm" are you referring to ?
A web of trust (from OpenPGP or any other means) express some trust
between nodes (individuals), including between nodes that do not trust
directly each other, but have a common trusted relationship.

But it does not guarantee that each node is unique. The very same user
may have two private keys associated with two identities and be
perfectly trusted and signed by any other valuable and honest member of
the web of trust.
For a "regular" use of OpenPGP web of trust, it does not matter, because
it is essentially a mean of knowing that you are who you claim to be. So
having different "identities" throughout the WoT is not really
desirable. But if having a trusted identity gives now access to money,
then it becomes desirable to have lots of identities.

>
>> Non, c'est s�r, le jour o� on se fait prendre, il y a de forte
>> chance qu'on soit plus ou moins bani du r�seau. Mais avant de se
>> faire prendre, on pourrait en profiter ...
>
> Your money units are key-signed. So you can individually accept or
> reject some money units you don't trust.
>
> So no, you cannot "play" without trust. Not only your identity but
> also your money units are signed.

By who are the coins signed ?
I guess it is by each member when the monthly distribution occurs, right ?
But we are back to square one: what prevents a coin emitted and signed
by a fake identity to be accepetd by other ?

>
>> Cela ne r�soud pas il me semble le probl�me d'une fausse identit�.
>> Elle est connue de tous, transparente, mais il est quand m�me
>> difficile de v�rifier qu'elle est fausse sans v�rifier tout le
>> monde.
>
>
> The WoT itself limit the expansion of false money units.
>

How ?

[Stéphane Laborde]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA256

Le 13/12/2013 23:25, Hugo Mercier a �crit :

>> Only individuals near other ones can trust they OpenPGP keys.
>
> What do you mean by "near other ones" ? I guess you mean near from
> each other in the web of trust. So each member has to trust every
> other members ?
>

No.

>>
>> Only individuals trusted by the community algorithm can recognize
>> a new one.
>
> Sure. And what if these trusted individuals collude to create a
> fake identity ?
>

That fake identity will produce false signed money. If one individual
doesn't trust that false money units he doesn't accept them.

And more when a fake identity is known, trust towards these
individuals fails? They will lose their trust with others, so probably
lose their money units exchange power too.

>>
>
> What "graph algorithm" are you referring to ? A web of trust (from
> OpenPGP or any other means) express some trust between nodes
> (individuals), including between nodes that do not trust directly
> each other, but have a common trusted relationship.
>
> But it does not guarantee that each node is unique. The very same
> user may have two private keys associated with two identities and
> be perfectly trusted and signed by any other valuable and honest
> member of the web of trust. For a "regular" use of OpenPGP web of
> trust, it does not matter, because it is essentially a mean of
> knowing that you are who you claim to be. So having different
> "identities" throughout the WoT is not really desirable. But if
> having a trusted identity gives now access to money, then it
> becomes desirable to have lots of identities.
>

See OpenUDC specification and UDID2 unique information. OpenUDC WoT is
not OpenPGP WoT. It's based on OpenPGP, not equal to.

>
> By who are the coins signed ? I guess it is by each member when the
> monthly distribution occurs, right ?

By the owner yes.

> But we are back to square one: what prevents a coin emitted and
> signed by a fake identity to be accepetd by other ?
>

I1 recognise I3 coins created and signed by I3. He can refuse those
coins because I3 is not in it's own WoT.

I3 is known by I2, and I2 know by I1. I3 can propose I2 an exchange of
coins, and so propose the exchange to I1 finally thru an exchange with
I2, trusted by I3 and I1.

One individual decide who he trusts, and who he doesn't trust, he can
parameter it's soft to do that.

>>
>> The WoT itself limit the expansion of false money units.
>>
>
> How ?

As I explained :

1�) thru renewable trusted individuals able to sign a new one
2�) thru created signed coins by owner (a coin is always signed by
it's individual creator)
3�) thru the ability for any individual to accept / refuse signed
coins following it's own WoT relative distance rules.
4�) The 3�) rule allow I1 and I3 to exchange thru I2 (or a chain Ia,
Ib, Ic...) intermediate trusted members, with which they can exchange
money units.

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.11 (GNU/Linux)
Comment: Using GnuPG with Thunderbird - http://www.enigmail.net/

iQEcBAEBCAAGBQJSq46TAAoJELdo2ekCUDkjFlgIANV4UdLa4kz7ByXNvp75VvXk
I8h715IrWJ8lf4maMLiUBf1ifXvheRPwZ4+eNltSEkH9gIGZZ3zRhb7gU8Zf1n3d
Y+2pRVKNQJrkblNNtnzuuVukLeRaq9KcbnTDRlFVt1CT4t7T3dGq5bFFpR4JGG24
U37YNx0YmqlxiPquP1fbZbxz+KkA8TP/1KiCF/R55wNqzkpwA3ERsBdY8fNMeB3H
PALAHr93MgHC21uNfWtjcJTvI5c2fGtCHDtY7MG9iv04SHo8GmZAeLyHo4p3uWNw
ZObHpDEhNdLG0HTLsYTh9nxuvBubmh4qBcDfq1mIccHjzQhZQZ7M12uWPqe+Ag0=
=/VDc
-----END PGP SIGNATURE-----

[jbar]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA256

I don't have the same patience than Galuel.

Hugo Mercier, your parents didn't teach you that stealing, cheating,
killing, enslaving... are BAD !?

The OpenUDC protocol permit to detect and to "punish" the bad, the
ugly, and the nasty ones. What do you need more ?

A technology respecting the individual freedom, won't permit to give
up any "social contract" (note: I prefer Locke to Rousseau).

And like in the society, if many began to disrespect "the" social
contrat, it creates mess, tensions, wars, executions...

OpenUDC doesn't claim and won't resolve some human issues.

Note: Now we may suspect you and some of your friend to be the bad,
ugly and nasty ones ... ;-)

On Fri, 13 Dec 2013 23:47:49 +0100
Stéphane Laborde <gal...@glibre.org> wrote:

> -----BEGIN PGP SIGNED MESSAGE-----
> Hash: SHA256
>

> --
> OpenUDC aims to provide a open standard for Universal Dividend Crypto-Currencies.
>
> homepage: http://openudc.org --- git's home: https://github.com/Open-UDC/open-udc.git --- Multi User Chat: open...@muc.jappix.com.
> ---
> You received this message because you are subscribed to the Google Groups "OpenUDC" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to openudc+u...@googlegroups.com.
> For more options, visit https://groups.google.com/groups/opt_out.

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.12 (GNU/Linux)

iQEcBAEBCAAGBQJSrB5aAAoJEICx309/5mldPqUH/iTI/Yh8cbeG8IG1Ht9NsnJs
XmbhLhSbsThWnt/DsgBMnOXC/cbWHkL0tRtw3M4mlqqn1Y5dJUR57+NmrPfvlexx
Q5pAt/l9dIKFOolVxzfk+SWhrWih9cGq48OVqCw9pQTiLuDEq7fkcRa6ovJIUhuf
OyuspmjD/6pKcdGcU6w7k/Cts1puDAO5hjrye+3l8G902XY4c2SYnaU4+3vJPmF8
vB1zKSgaXiclI+5z4j8/x3c6H6HWQr+kIOBwpTCqiPtg1uqDKdEN4uE+dGg3e1YL
WjcpEB6DJiSmzdaY351WDMzrpTzBPrN4OthWUMrwxte/Ieo6tj8MFXG9Q4xfdM4=
=6NF0
-----END PGP SIGNATURE-----

[Hugo Mercier]

Le 13/12/2013 23:47, St�phane Laborde a �crit :

> Le 13/12/2013 23:25, Hugo Mercier a �crit :
>

>> What "graph algorithm" are you referring to ? A web of trust (from
>> OpenPGP or any other means) express some trust between nodes
>> (individuals), including between nodes that do not trust directly
>> each other, but have a common trusted relationship.
>
>> But it does not guarantee that each node is unique. The very same
>> user may have two private keys associated with two identities and
>> be perfectly trusted and signed by any other valuable and honest
>> member of the web of trust. For a "regular" use of OpenPGP web of
>> trust, it does not matter, because it is essentially a mean of
>> knowing that you are who you claim to be. So having different
>> "identities" throughout the WoT is not really desirable. But if
>> having a trusted identity gives now access to money, then it
>> becomes desirable to have lots of identities.
>
> See OpenUDC specification and UDID2 unique information. OpenUDC WoT is
> not OpenPGP WoT. It's based on OpenPGP, not equal to.
>

Ok, so you added a field containing a birth date and a place of birth
for each individual. Indeed, it lowers the probability of someone being
signed by two different "branches" of the WoT graph.
But it still does not prevent some people to create a bunch of fake
identities.

>> But we are back to square one: what prevents a coin emitted and
>> signed by a fake identity to be accepetd by other ?
>
>
> I1 recognise I3 coins created and signed by I3. He can refuse those
> coins because I3 is not in it's own WoT.
>
> I3 is known by I2, and I2 know by I1. I3 can propose I2 an exchange of
> coins, and so propose the exchange to I1 finally thru an exchange with
> I2, trusted by I3 and I1.
>
> One individual decide who he trusts, and who he doesn't trust, he can
> parameter it's soft to do that.
>

Ok, so now it makes a bit more sense ...
If I understand correctly, this is a crucial point. It means if I1 does
not trust I3, it asks I2 to be some kind of warranty for him. Then a
transaction can happen between I3 and I1, through I2.
If I3 tries to abuse the system, he actually abuses I2, not the whole
network ...

Is the ability to find a "trust path", in an automated way, part of
OpenUDC protocol ?

I find it very close to what the Ripple network does with IOU: anyone
can emit its own currency and payment are resolved by a path finding in
a trust network and IOU exchange process.

Then, now I am confused about why you need a keysigning / UDID process :)
Could you just let people emit their own monetary units ? If every one
has a "direct" trust circle and only deals with it, and eventually with
a chain of trust circle hops, it still seem to work ...

[Hugo Mercier]

Le 14/12/2013 10:01, jbar a �crit :

>
> I don't have the same patience than Galuel.
>
> Hugo Mercier, your parents didn't teach you that stealing, cheating,
> killing, enslaving... are BAD !?

... I am not sure to understand your reaction.

Jean Jacques Brucker, didn't your parents teach you that not all human
beings are perfect angels and that some of them will try to profit from
others, even if you're a perfect gentleman ?

Trying to understand how a (computer) system is tolerant to attacks is a
way to gets confidence in it.
If nobody explained how it is hard for an attacker to crack a private
OpenPGP key, would you use it ?
And if it takes part in some form of money creation, understanding how
it is robust to fraud is a way to give value to the money (i.e. having
lots of people using it).

[Stéphane Laborde]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA256

Le 15/12/2013 11:24, Hugo Mercier a �crit :

>
> Ok, so you added a field containing a birth date and a place of
> birth for each individual. Indeed, it lowers the probability of
> someone being signed by two different "branches" of the WoT graph.
> But it still does not prevent some people to create a bunch of
> fake identities.
>

You do not prevent, because OpenUDC is a trust network, not a mistrust
one. So first, you trust, with some information allowing every
individual to control, and secondly if it appears you lose trust
towards some false individuals, including the individual who sign
them, then you lose trust regarding those humans.

So you don't start from mistrust as bitcoin does, avoiding any
reference to humans, but refering only machines, to go eventually to
some hypothetical trust between humans, you start from trust between
humans to go eventually to mistrust.

Causality comes here from the opposite point.

Because Relative Money Theory fundamentaly supposes economy is
something between humans, not at all something between machines,
matterial, or any other specific non-human value.

>>
>
> Ok, so now it makes a bit more sense ... If I understand correctly,
> this is a crucial point. It means if I1 does not trust I3, it asks
> I2 to be some kind of warranty for him. Then a transaction can
> happen between I3 and I1, through I2. If I3 tries to abuse the
> system, he actually abuses I2, not the whole network ...

That's a correct interpretation, yes (can be I3 or I1 who "ask" to I2).

>
> Is the ability to find a "trust path", in an automated way, part
> of OpenUDC protocol ?
>

Absolutely, there is such a trust path by default. But you can also
determine your own "trust path algorithm", is something fully
independant. Why ? Because you always has 100% information of who
created one money unit, and you can so always accept or refuse it for
the reason you decide to trust the individual or not.

> I find it very close to what the Ripple network does with IOU:
> anyone can emit its own currency and payment are resolved by a path
> finding in a trust network and IOU exchange process.
>

Ripple has no understanding at all of Universal Dividend as a
space-time symetrical money system, the only solution for a freedom
money system between humans in space (symetry at "t") and in time
(symetry between individuals separated by generations).

You cannot compare a freedom money system with a not-freedom one.

> Then, now I am confused about why you need a keysigning / UDID
> process :) Could you just let people emit their own monetary units
> ? If every one has a "direct" trust circle and only deals with it,
> and eventually with a chain of trust circle hops, it still seem to
> work ...

Free Money system is based on Time-Space symetry, that is to say no
individual create more money than others in space (at the moment "t")
and no individual now or in the future will created less or more money
units relatively to monetary mass during it's life expectancy.

So individual money creation of a free money is an average UD(t) =
c*M(t)/N(t) where c=ln(ev:2)/(ev:2) M(t) the monetary mass and N(t)
the number of members.

So you need the same database to share the money, and the same
database to share information about the individuals.

It's the same contract betweent N(t) individuals with t from 2014 to
2094, from 2015 to 2095, from 2016 to 2096, as well as the same
contract from 2400 to 2480.

So no, you cannot share the same contract with doing what you want
alone. You accept to enter, or you do not enter. You are free to share
a freedom money system or not.

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.11 (GNU/Linux)
Comment: Using GnuPG with Thunderbird - http://www.enigmail.net/

iQEcBAEBCAAGBQJSrZDUAAoJELdo2ekCUDkj5IoH/jjEpWnFESImzi54UGEpSNlJ
tzJIkp1wGajYe9A66tulbkt5a+G00L1HH3CUMfRPzw1+qqIaqTqazlPyIF+IgW6Z
gMHIp2NsHn6l84FjRfoEX8nMXTQVBkZqTcA0j59s5i1JsQs1NkN0WDsZ0tA6AGe8
405ks5ia11LkRX05qB1PfaSiArhHfz27rFq5S5yZ0nouALsDEvOw8o+Z6tyDbCSz
6ITMnhYARgeAWgt4NC9efmx70Vmasn8yS5pt0t/JF9fKUL0GgiBw5gZ4kKZZzKrf
YZDHfNAOt83/GY8YRBXoAUMYDGQXADowcAxmPceQ2mf2Wh9IBHjxZsafWff9JkE=
=ww1w
-----END PGP SIGNATURE-----

[jbar]

I don't care about kindness, but smartness.
Apparent kindness is often time just a consequence of smartness.

The fact that many people aren't smart enough to consider others as
themselves (and maybe smarter) is not a technical issue.

(But for sure : the more of smart people, the less we need legislation
and state.)

I also only consider as "forbidden" what I temporary forbid myself, and
I forbid myself to abuse from other only because I also don't want to
be abused.

Concerning software, the question isn't "what is "forbidden", what's
not ?", but only "what is possible/impossible ?". And once you are a
developer and moreover when using free software, everything become
possible (it is then just a question of time or efforts).

So you may sign a false certificate, as you may for example injure your
neighbour, just mind the consequences.

And I really want that some dumb people try it, so that more others
understand as dumb they were.

(It "simply" may destroy all the confidence other could lend to
your(s) (crypto-)identity(ies), which are related and should affect at
the end your own body...).

I now invite you to start using OpenPGP, to read the
currents specifications in
https://github.com/Open-UDC/open-udc/tree/master/docs , and to come in
our next developer's meeting in April.

L.F. to see you.

---

On Sun, 15 Dec 2013 11:24:27 +0100
Hugo Mercier <hmerc...@gmail.com> wrote:

[Hugo Mercier]

Le 15/12/2013 12:21, St�phane Laborde a �crit :

> You do not prevent, because OpenUDC is a trust network, not a mistrust
> one. So first, you trust, with some information allowing every
> individual to control, and secondly if it appears you lose trust
> towards some false individuals, including the individual who sign
> them, then you lose trust regarding those humans.
>
> So you don't start from mistrust as bitcoin does, avoiding any
> reference to humans, but refering only machines, to go eventually to
> some hypothetical trust between humans, you start from trust between
> humans to go eventually to mistrust.
>
> Causality comes here from the opposite point.
>
> Because Relative Money Theory fundamentaly supposes economy is
> something between humans, not at all something between machines,
> matterial, or any other specific non-human value.

Yeah ... I already read this somewhere. But ... that is a question of
interpretation. Bitcoins are used and made by humans for humans. OpenUDC
uses (or plan to use) machines to work.

I think what brings "mistrust first" is the fact bitcoin (or gold, or
cash) is only considered as a good. You could also consider money as a
debt, where it is then more "social" than "physical". Historically, we
have been using both of them, with different relative importance,
through history and culture. None of the two aspects is perfect.

>
>>>
>
>> Ok, so now it makes a bit more sense ... If I understand correctly,
>> this is a crucial point. It means if I1 does not trust I3, it asks
>> I2 to be some kind of warranty for him. Then a transaction can
>> happen between I3 and I1, through I2. If I3 tries to abuse the
>> system, he actually abuses I2, not the whole network ...
>
> That's a correct interpretation, yes (can be I3 or I1 who "ask" to I2).
>
>
>> Is the ability to find a "trust path", in an automated way, part
>> of OpenUDC protocol ?
>
>
> Absolutely, there is such a trust path by default. But you can also
> determine your own "trust path algorithm", is something fully
> independant. Why ? Because you always has 100% information of who
> created one money unit, and you can so always accept or refuse it for
> the reason you decide to trust the individual or not.
>

Ok, I may have been unclear here. I was looking for a way to make the
chained transaction I3->I2->I1 triggered automatically when I3 wants to
pay I1.

>> I find it very close to what the Ripple network does with IOU:
>> anyone can emit its own currency and payment are resolved by a path
>> finding in a trust network and IOU exchange process.
>
>
> Ripple has no understanding at all of Universal Dividend as a
> space-time symetrical money system, the only solution for a freedom
> money system between humans in space (symetry at "t") and in time
> (symetry between individuals separated by generations).
>
> You cannot compare a freedom money system with a not-freedom one.

I think Ripple is agnostic to what money "system" you use. It could be
one with a fixed amount of monetary units (bitcoin or any forks),
national currencies, a LETS, or ... probably a universal dividend. It is
"just" a public distributed ledger which registers trust networks
between accounts (expressed in terms of credit lines of any currency you
like) and allows to transfer credits (and exchange currencies), with a
consensus mechanism that do not need to spend lots of power on mining.

Rules that people from a given community agree to follow on could be
implemented by external means (scripts on the public ledger for instance)

>
>> Then, now I am confused about why you need a keysigning / UDID
>> process :) Could you just let people emit their own monetary units
>> ? If every one has a "direct" trust circle and only deals with it,
>> and eventually with a chain of trust circle hops, it still seem to
>> work ...
>
> Free Money system is based on Time-Space symetry, that is to say no
> individual create more money than others in space (at the moment "t")
> and no individual now or in the future will created less or more money
> units relatively to monetary mass during it's life expectancy.
>
> So individual money creation of a free money is an average UD(t) =
> c*M(t)/N(t) where c=ln(ev:2)/(ev:2) M(t) the monetary mass and N(t)
> the number of members.

Right, this is what brings me here ... this is an interesting view, not
sure yet what problem it is trying to solve, but ... interesting :D

If the amount of monetray units someone is "allowed" to create depends
on the number of members of a community, and if you can choose not to
recognise money creation of someone else, it means the number of members
part of the community is relative to each member ?

>
> So you need the same database to share the money, and the same
> database to share information about the individuals.
>
> It's the same contract betweent N(t) individuals with t from 2014 to
> 2094, from 2015 to 2095, from 2016 to 2096, as well as the same
> contract from 2400 to 2480.
>
> So no, you cannot share the same contract with doing what you want
> alone. You accept to enter, or you do not enter. You are free to share
> a freedom money system or not.

Ok ...
There is something in the contrast between how you need to share
something globally for individual membership and how you could act on
your own for money creation that is still unclear for me ... I probably
have to sleep on it ...

[Hugo Mercier]

Le 15/12/2013 16:49, jbar a écrit :
>
> I don't care about kindness, but smartness.
> Apparent kindness is often time just a consequence of smartness.
>
> The fact that many people aren't smart enough to consider others as
> themselves (and maybe smarter) is not a technical issue.
>
> (But for sure : the more of smart people, the less we need legislation
> and state.)
>
> I also only consider as "forbidden" what I temporary forbid myself, and
> I forbid myself to abuse from other only because I also don't want to
> be abused.

It seems to be a good philosophy of life ... so what ? :)

>
> Concerning software, the question isn't "what is "forbidden", what's
> not ?", but only "what is possible/impossible ?". And once you are a
> developer and moreover when using free software, everything become
> possible (it is then just a question of time or efforts).
>
> So you may sign a false certificate, as you may for example injure your
> neighbour, just mind the consequences.
>
> And I really want that some dumb people try it, so that more others
> understand as dumb they were.

Not sure to follow you : you want dumb people to try what ?

> I now invite you to start using OpenPGP, to read the
> currents specifications in
> https://github.com/Open-UDC/open-udc/tree/master/docs , and to come in
> our next developer's meeting in April.
>
> L.F. to see you.

I will have another look at current specs.

[Stéphane Laborde]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA256

Le 15/12/2013 18:29, Hugo Mercier a �crit :

>
> Ok, I may have been unclear here. I was looking for a way to make
> the chained transaction I3->I2->I1 triggered automatically when I3
> wants to pay I1.

That's a freedom impossibility. I2 can refuse of course, because I2 is
absolutely free.

>
> I think Ripple is agnostic to what money "system" you use. It could
> be one with a fixed amount of monetary units (bitcoin or any
> forks), national currencies, a LETS, or ... probably a universal
> dividend. It is "just" a public distributed ledger which registers
> trust networks between accounts (expressed in terms of credit lines
> of any currency you like) and allows to transfer credits (and
> exchange currencies), with a consensus mechanism that do not need
> to spend lots of power on mining.
>

Yes Ripple is interesting tool, but not a freedom money. It's
something else. We don't care here in OpenUDC about that, it's fully
independant, and different thing.

>
> Right, this is what brings me here ... this is an interesting view,
> not sure yet what problem it is trying to solve, but ...
> interesting :D
>

Symetry (liberty) between men(t1) and men(t1+t2). No generation(t1)
has any right to create more money than other generation(t2).

Erase economical time crisis due to asymetrical time based money
creation that ignore human die at any time and other human are borned
at the same time, and at any time.

> If the amount of monetray units someone is "allowed" to create
> depends on the number of members of a community, and if you can
> choose not to recognise money creation of someone else, it means
> the number of members part of the community is relative to each
> member ?
>

Very bad question.

Universal Dividend rule can be also UD(t+1) = UD(t)*(1+c) not
depending of N(t) in OpenUDC graph. It has some problem relatively to
money, and some good aspects also.

Other rules are possible to calculate UD(t) that are also compatible
with Relative Money theory. All of them are very close, but conduct to
some variations when you consider different N(t) possibilies.

OpenUDC is independant or Universal Dividend rule you choose to
initiate the money. So it doesn't matter at all.

>
> Ok ... There is something in the contrast between how you need to
> share something globally for individual membership and how you
> could act on your own for money creation that is still unclear for
> me ... I probably have to sleep on it ...
>

Perhaps I just gave you a perspective to think about it.

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.11 (GNU/Linux)
Comment: Using GnuPG with Thunderbird - http://www.enigmail.net/

iQEcBAEBCAAGBQJSreq+AAoJELdo2ekCUDkjLOoH/A4hXpot3rbdm9gR4J7RhgYu
F35MqKZLtGh8TiByV2I9DEFBnehbs44IvhiouvtahqQavHU3Cna0TNmQZOkDml9C
co7epAKTBH6vQa7iiNqOYdaYgC7iLtsrTe6AdRlNvxOwgtozAwEf53Y8bjqUb8EM
YzGikGNDmzbKI9g3y3Iq6SfGJbb6dS2lHmcz+0kSGUZBSnuvWcmU/97+OLo/pbYU
MwcNG8AMiZ5b+I1llHeXV4HYpmyW2ASIT3Iyv/w85CvSXLF9szyZNK+4fIFv5FKw
4Htj2Buswatq1uRFYv7jMg+A2u9zQxiKgTvkcCyo/G/SUPg4Dsti5AtfD2mP1YQ=
=bLdz
-----END PGP SIGNATURE-----

[Cédric Moreau]

Le dimanche 15 décembre 2013 18:29:50 UTC+1, Hugo Mercier a écrit :

Le 15/12/2013 12:21, St�phane Laborde a �crit :

Ok ...
There is something in the contrast between how you need to share
something globally for individual membership and how you could act on
your own for money creation that is still unclear for me ... I probably
have to sleep on it ...

What you share is a common agreement, an agreement on who may issue new money and how much. This is a contract.

What you do on your own is issuing new money according to this contract, if you want to follow it.

[Hugo Mercier]

Le 15/12/2013 18:45, St�phane Laborde a �crit :

> Le 15/12/2013 18:29, Hugo Mercier a �crit :
>
>> Ok, I may have been unclear here. I was looking for a way to make
>> the chained transaction I3->I2->I1 triggered automatically when I3
>> wants to pay I1.
>
> That's a freedom impossibility. I2 can refuse of course, because I2 is
> absolutely free.
>

Sure, but it means every chained transactions has to be setup manually.
Which is not always very convenient.
There could be a way for I2 to declare he is ok (until he is not) to
"pass" transactions from I3 to I1 (and/or the opposite). It is a matter
of implementation.

>
> Yes Ripple is interesting tool, but not a freedom money. It's
> something else. We don't care here in OpenUDC about that, it's fully
> independant, and different thing.

Yes. But OpenUDC still lacks a working implementation if I am correct. I
was wondering if Ripple (or other transaction frameworks if any) could
be used to setup a prototype implementation on top of it. That could
ease dissemination and testing of the universal dividend concept.

>
>> Right, this is what brings me here ... this is an interesting view,
>> not sure yet what problem it is trying to solve, but ...
>> interesting :D
>
>
> Symetry (liberty) between men(t1) and men(t1+t2). No generation(t1)
> has any right to create more money than other generation(t2).
>
> Erase economical time crisis due to asymetrical time based money
> creation that ignore human die at any time and other human are borned
> at the same time, and at any time.

What about ownerships that you can give to your children ? Is it a
source of temporal asymetry ?

>
>> If the amount of monetray units someone is "allowed" to create
>> depends on the number of members of a community, and if you can
>> choose not to recognise money creation of someone else, it means
>> the number of members part of the community is relative to each
>> member ?
>
>
> Very bad question.

You're welcome ...

>
> Universal Dividend rule can be also UD(t+1) = UD(t)*(1+c) not
> depending of N(t) in OpenUDC graph. It has some problem relatively to
> money, and some good aspects also.
>
> Other rules are possible to calculate UD(t) that are also compatible
> with Relative Money theory. All of them are very close, but conduct to
> some variations when you consider different N(t) possibilies.
>
> OpenUDC is independant or Universal Dividend rule you choose to
> initiate the money. So it doesn't matter at all.
>

There are other rules, ok.

[Hugo Mercier]

Le 15/12/2013 20:55, C�dric Moreau a �crit :
>
>
> Le dimanche 15 d�cembre 2013 18:29:50 UTC+1, Hugo Mercier a �crit :
>
> Le 15/12/2013 12:21, St�phane Laborde a �crit :

>
> Ok ...
> There is something in the contrast between how you need to share
> something globally for individual membership and how you could act on
> your own for money creation that is still unclear for me ... I probably
> have to sleep on it ...
>
>
> What you share is a common agreement, an agreement on who may issue new
> money and how much. This is a contract.

How do you reach decentralized consensus on such a common agreement ?

[Cédric Moreau]

2013/12/16 Hugo Mercier <hmerc...@gmail.com>

Le 15/12/2013 20:55, Cédric Moreau a écrit :

>
>
> Le dimanche 15 décembre 2013 18:29:50 UTC+1, Hugo Mercier a écrit :
>

>     Le 15/12/2013 12:21, St�phane Laborde a �crit :

>
>     Ok ...
>     There is something in the contrast between how you need to share
>     something globally for individual membership and how you could act on
>     your own for money creation that is still unclear for me ... I probably
>     have to sleep on it ...
>
>
> What you share is a common agreement, an agreement on who may issue new
> money and how much. This is a contract.

How do you reach decentralized consensus on such a common agreement ?

If I remember correctly, OpenUDC aims at doing this spreading each individual signature to the whole network.

Then, each one is aware of what was signed, by whom.

Of course, this is still relative point of view. But an officialy displayed one (so anyone can check it, at any moment for any node).

[Stéphane Laborde]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA256

Le 16/12/2013 13:02, Hugo Mercier a �crit :

>
> Sure, but it means every chained transactions has to be setup
> manually. Which is not always very convenient. There could be a way
> for I2 to declare he is ok (until he is not) to "pass" transactions
> from I3 to I1 (and/or the opposite). It is a matter of
> implementation.
>

It is individual WoT trust rules. OpenUDC provide a standard one by
default, but you can always choose your own. OpenUDC cannot oppose you
to accept or refuse some coins units.

If you rule is ok with I2,I4,I7,...,In rule, it can work. If it is
not, it will not. It's not part of OpenUDC core system, it's something
you can add on it.

>
> Yes. But OpenUDC still lacks a working implementation if I am
> correct. I was wondering if Ripple (or other transaction frameworks
> if any) could be used to setup a prototype implementation on top of
> it. That could ease dissemination and testing of the universal
> dividend concept.
>

Ask this to Ripple developpers, not OpenUDC ones.

>
> What about ownerships that you can give to your children ? Is it a
> source of temporal asymetry ?
>

No. You should read the Relative Money Theory first. Temporal Symetry
or Asymetry judgment concern money creation, not money ownership.

>
>
> There are other rules, ok.
>

"One OpenUDC money community can choose another Universal Dividend rule"

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.11 (GNU/Linux)
Comment: Using GnuPG with Thunderbird - http://www.enigmail.net/

iQEcBAEBCAAGBQJSrvdWAAoJELdo2ekCUDkj4TEIANDRcZ4iIri4sG36lJzLsrpH
TtY6tEtGPTf+b65pXYTpU285pJ9sNdlrXSsmj+VOrXMY53OXdxlHpk9t5rRTrx/i
wjQLkTIK4ylbhSHEAc4PtAKXb6xIyGH1+w6RsCMJ0QjYdMsgBpVy00e0qt2m8Aop
eysr5ultA9UTMBnJbdAlP/H5iY4n6wTq/Hzsz4GJmGpcY6IIXtkCPxzZyEkknXos
ytNeUEZAIJ9B9GS3WNX3hJznlX4daFmISOP6a1JRTW2y5V7SPn8sL8Fm6zkjZaed
sx8pu8Qv7LGK9o8PiTe32pgFiF7WXWRHNsNDVBI2+r5KcbCm5abhOWGf90GjWtI=
=tMH+
-----END PGP SIGNATURE-----

[Stéphane Laborde]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA256

Le 16/12/2013 13:19, C�dric Moreau a �crit :
> 2013/12/16 Hugo Mercier <hmerc...@gmail.com
> <mailto:hmerc...@gmail.com>>

>
> How do you reach decentralized consensus on such a common agreement
> ?
>
>
> If I remember correctly, OpenUDC aims at doing this spreading each
> individual signature to the whole network. Then, each one is aware
> of what was signed, by whom.
>
> Of course, this is still relative point of view. But an officialy
> displayed one (so anyone can check it, at any moment for any
> node).
>

Cedric say it correctly, you can also interprate it as :

When I choose bitcoin money system, I implicitly accept Bitcoin Money
Creation Rules

When I choose an OpenUDC money system, I implicitly accept this
OpenUDC community Universal Dividend Rules.

So you are creator of money units, but following the rules, not
following your own view.

But in fact you follow your own view because you accepted the rules !

So you cannot interprate a sentence if you don't define the conceptual
referential where you say it. Do you say "I choose" before accepting
the rules ? => yes. Is it a personnal choice concerning individual
money creation before accepting the rules ? yes.

And after having accepted the rules, can I still choose ? You cannot
inside the system.

So i'm not free after accepting the rules ? Yes you are still free
inside the system, you can accept to exchange or not, or refuse /
accept money units. You are also free outside the system because you
can join other money systems without any obligation.

So as you can understand individual freedom mean different things in
space and in time, following your own acts.

> -- OpenUDC aims to provide a open standard for Universal Dividend
> Crypto-Currencies.
>
> homepage: http://openudc.org --- git's home:
> https://github.com/Open-UDC/open-udc.git --- Multi User Chat:
> open...@muc.jappix.com. --- You received this message because you
> are subscribed to the Google Groups "OpenUDC" group. To unsubscribe
> from this group and stop receiving emails from it, send an email to
> openudc+u...@googlegroups.com. For more options, visit
> https://groups.google.com/groups/opt_out.

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.11 (GNU/Linux)
Comment: Using GnuPG with Thunderbird - http://www.enigmail.net/

iQEcBAEBCAAGBQJSrvm5AAoJELdo2ekCUDkjlP4IAM2GzQZFpF7QlNk9rsLvFFsT
6tez3+jOquU1mpnFIo7SC7cj9gIjseUHbNFHZZxhIKqi/1UxQkDI40H5hMsHmFSo
+aBZsZMAkGYlNveuBfke7eUm28kAkVg5rw6/elX3jYm2DBD90YOOcNCD3SXTSXPu
jhPzHFG5KXkF7XqCR8VQp5KWumk2BpSSlApJ8nocr9Kp0Z+kM27H/ZsuKqMenlmp
AdWJpexKo1iCfEQAVYJvYJwy4D+v5GUP4oLvW/2E29UR3yrbr9tr/tWVneFozael
f2E5+hIUxp44ZjhdKmNgng3yImXI91nSTT7HQ9dvF4PhLCjkz0Sgr1TBdAK1soM=
=pFuJ
-----END PGP SIGNATURE-----

[Hugo Mercier]

Le 16/12/2013 13:51, St�phane Laborde a �crit :

>> Yes. But OpenUDC still lacks a working implementation if I am
>> correct. I was wondering if Ripple (or other transaction frameworks
>> if any) could be used to setup a prototype implementation on top of
>> it. That could ease dissemination and testing of the universal
>> dividend concept.
>
>
> Ask this to Ripple developpers, not OpenUDC ones.

LOL. Seriously ???

I don't understand how you guys work ... I give up ...

[Stéphane Laborde]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA256

Le 16/12/2013 17:50, Hugo Mercier a �crit :

> Le 16/12/2013 13:51, St�phane Laborde a �crit :
>
>>> Yes. But OpenUDC still lacks a working implementation if I am
>>> correct. I was wondering if Ripple (or other transaction
>>> frameworks if any) could be used to setup a prototype
>>> implementation on top of it. That could ease dissemination and
>>> testing of the universal dividend concept.
>>
>>
>> Ask this to Ripple developpers, not OpenUDC ones.
>
> LOL. Seriously ???

Yes seriously. Ripple is not at all a freedom money system. Ripple is
a fixed based money system not based on humans(x,y,t).

You should study what is a freedom money system for individuals by
studying Relative Money Theory

>
> I don't understand how you guys work
>

We don't care about not-free money systems.

That's how we "work".

> ... I give up ...

You are free.

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.11 (GNU/Linux)
Comment: Using GnuPG with Thunderbird - http://www.enigmail.net/

iQEcBAEBCAAGBQJSrzgjAAoJELdo2ekCUDkjKSIH/ikHosrzcghguIoY5ALXiRqG
7aJ9OBvKX4LEnptiUu3geib4huGc6qNY+PowC5whixRrBrja/MkzOOyb+xPuId4+
35UZtc7HrBRiX6kyH7WHQeUOFnpRAjp2BHqmcjKh6Q5cfNr0SZryFjU6Yp9hyikA
eHxRefL/lX9nTt6k5aT+uvyCnxwzgiBPQqoyXKf4x+BIDfS320f++NwM1NsElZyf
fivUnMm9dYsYHTCouvHxSwdpk/JmhU7upV5pX8J/1hhURzbQLeLYJcQNoF2gLKOM
Ytp+Ek4upQs9Ih2U/3wzCHUyRZj/c4uXcbKVJkPx5QMiI2fMxEcKvT7a4kTseVg=
=ErUa
-----END PGP SIGNATURE-----