DIP 1011--extern(delegate)--Preliminary Review Round 1 Begins
The first preliminary review round of DIP 1011, "extern(delegate)", has begun. http://forum.dlang.org/thread/topmfucguenqpucsb...@forum.dlang.org
Re: Hiring D programmers (with cryptography and blockchain knowledge are preferred)
On Friday, 14 July 2017 at 04:47:24 UTC, rikki cattermole wrote: On 14/07/2017 1:41 AM, Vitor Rozsas wrote: Hmmm... But how would a criminal post be deleted or removed? Even if decentralized, govs could forbid the use of it (specially having a server of this social media), considering that it's database could contain child pornography (undeletable child pornography - really bad). But it doesn't matter, I have other projects to invest my money on D. I will post the links soon. Anyway its just silly to go for removal of servers. The overhead is far too great. Our compression algo's can't back it up. You'll be talking terabytes just to store a few 'websites' soon enough. Nah, anyway decentralized != anonymous. Even in Tor there are servers. What we need isn't a decentralized http replacement. What we need is a decentralized dns with some form of filters available. What about a decentralized cryptocurrency (like bitcoin) but instead of the blockchain having transactions recorded on it, it could have something lighter... like a list of coins IDs ("cents", actually) and their owners, like: 1 = Vitor 2 = Vitor 3 = Vitor 4 = Leonard [...] (cents 1, 2, 3 belong to Vitor and cent 4 belong to Leonard, and so on...). A "map" (C++) or associative array (D) with the ID of the cent (key) and owner (value) or opposite, if necessary. Is this even possible? Please consider that I don't know a lot about cryptography. And I have no idea of whether this is feasible with D or with any language at all, so if I'm speaking nonsensic things, just tell me. But I think it would be lighter than store all transactions from the very first one. I just don't know if this is doable; cryptocurrencies have a lot of security measures to avoid transferring money that doesn't exist, money that isn't yours, and so on, to confirm the transaction... Wouldn't it be easier? All the cents should have a first owner though... one that would sell to resellers, or just give away to random people. That way, the blockchain would have a fixed size, forever. Lighter than all that expensive mining stuff (that in my country is impossible - lack of decent machines, pricey electric energy), and the worst part: downloading 100+ GigaBytes because blockchain stores transactions, this database gets bigger and bigger... List of cents is limited; it's fixed. But instead of 18 Million that bitcoin uses, I was thinking of trillions (so the value of the currency is more aligned to fiat currencies that we already know and are used to). So... tell me. Is this possible? *Is this safe*? Is anybody going to be stolen if blockchain stores a map of cent id + owner instead of a bunch of transactions? I know that it would be lighter for CPUs of users and servers as well, and to the disk, that's for sure. Waiting for replies.
Re: Hiring D programmers (with cryptography and blockchain knowledge are preferred)
On Friday, 14 July 2017 at 14:10:29 UTC, Vitor Rozsas wrote: On Friday, 14 July 2017 at 04:47:24 UTC, rikki cattermole wrote: [...] What about a decentralized cryptocurrency (like bitcoin) but instead of the blockchain having transactions recorded on it, it could have something lighter... like a list of coins IDs ("cents", actually) and their owners, like: [...] A transaction would simply change the owner of the cents and not create any message in the database. Change the owner by changing the values (owner) of the keys (cents) being transferred.
Re: Hiring D programmers (with cryptography and blockchain knowledge are preferred)
On 14/07/2017 3:17 PM, Vitor Rozsas wrote: On Friday, 14 July 2017 at 14:10:29 UTC, Vitor Rozsas wrote: On Friday, 14 July 2017 at 04:47:24 UTC, rikki cattermole wrote: [...] What about a decentralized cryptocurrency (like bitcoin) but instead of the blockchain having transactions recorded on it, it could have something lighter... like a list of coins IDs ("cents", actually) and their owners, like: [...] A transaction would simply change the owner of the cents and not create any message in the database. Change the owner by changing the values (owner) of the keys (cents) being transferred. Blockchains work by making the entire history available to be verified and computed against. This is entirely its selling point. So no, a block chain can never be fixed sized. After all, how do you know that X owns the coin and not just some random node trying to corrupt and steal every bodies coins?
Re: Hiring D programmers (with cryptography and blockchain knowledge are preferred)
On Friday, 14 July 2017 at 14:23:49 UTC, rikki cattermole wrote: On 14/07/2017 3:17 PM, Vitor Rozsas wrote: On Friday, 14 July 2017 at 14:10:29 UTC, Vitor Rozsas wrote: [...] A transaction would simply change the owner of the cents and not create any message in the database. Change the owner by changing the values (owner) of the keys (cents) being transferred. Blockchains work by making the entire history available to be verified and computed against. This is entirely its selling point. So no, a block chain can never be fixed sized. After all, how do you know that X owns the coin and not just some random node trying to corrupt and steal every bodies coins? But it's so expensive... Soon, bitcoin will have a whole terabyte to be downloaded... So, no deal? That's sad.
Re: Hiring D programmers (with cryptography and blockchain knowledge are preferred)
On Friday, 14 July 2017 at 14:37:56 UTC, Vitor Rozsas wrote: On Friday, 14 July 2017 at 14:23:49 UTC, rikki cattermole wrote: On 14/07/2017 3:17 PM, Vitor Rozsas wrote: On Friday, 14 July 2017 at 14:10:29 UTC, Vitor Rozsas wrote: [...] A transaction would simply change the owner of the cents and not create any message in the database. Change the owner by changing the values (owner) of the keys (cents) being transferred. Blockchains work by making the entire history available to be verified and computed against. This is entirely its selling point. So no, a block chain can never be fixed sized. After all, how do you know that X owns the coin and not just some random node trying to corrupt and steal every bodies coins? But it's so expensive... Soon, bitcoin will have a whole terabyte to be downloaded... So, no deal? That's sad. I was thinking of doing something big... An operating system? With everything that a user could desire of? Big projects are necessary, for D to get more... known (and therefore, more used). A phone OS (ARM)? I heard that DMD will accept ARM very soon. While that doesn't happen, there are other compilers for ARM... So a phone OS could be interesting. A phone OS written in D. There are kernels done in D already. But I don't think they are as complete as Linux's for example.
Re: Hiring D programmers (with cryptography and blockchain knowledge are preferred)
On 14/07/2017 4:01 PM, Vitor Rozsas wrote: On Friday, 14 July 2017 at 14:37:56 UTC, Vitor Rozsas wrote: On Friday, 14 July 2017 at 14:23:49 UTC, rikki cattermole wrote: On 14/07/2017 3:17 PM, Vitor Rozsas wrote: On Friday, 14 July 2017 at 14:10:29 UTC, Vitor Rozsas wrote: [...] A transaction would simply change the owner of the cents and not create any message in the database. Change the owner by changing the values (owner) of the keys (cents) being transferred. Blockchains work by making the entire history available to be verified and computed against. This is entirely its selling point. So no, a block chain can never be fixed sized. After all, how do you know that X owns the coin and not just some random node trying to corrupt and steal every bodies coins? But it's so expensive... Soon, bitcoin will have a whole terabyte to be downloaded... So, no deal? That's sad. I was thinking of doing something big... An operating system? With everything that a user could desire of? Big projects are necessary, for D to get more... known (and therefore, more used). A phone OS (ARM)? I heard that DMD will accept ARM very soon. While that doesn't happen, there are other compilers for ARM... So a phone OS could be interesting. A phone OS written in D. There are kernels done in D already. But I don't think they are as complete as Linux's for example. Scale your dreams back a bit. If you want to put money into anything here is my list: 1. Get shared libraries 100% working, with clear articles on how to use it, on every platform. 2. std.experimental.color, it seems Manu is stalled 3. std.experimental.image, ok ok, this one is mine but its about time we get the damn interfaces standardized in Phobos! 4. GUI widget rendering library, GUI widget != GUI toolkit (think just rendering a single control), basically handles rendering + accessibility. These are not big projects, but they would go a long way to getting any thing really "cool". Keep in mind, money talks. Getting the right people available to work on a project full time for a few months would really speed a lot of this stuff up. There is also other stuff like a bidi library, font rasterizer and other more useful infrastructure projects that actually has a chance of being completed. Depending on how far you want to go, my list alone is a few years worth of work for a 3-5 people team.
Re: Hiring D programmers (with cryptography and blockchain knowledge are preferred)
On Friday, 14 July 2017 at 15:21:09 UTC, rikki cattermole wrote: [...] Scale your dreams back a bit. If you want to put money into anything here is my list: 1. Get shared libraries 100% working, with clear articles on how to use it, on every platform. 2. std.experimental.color, it seems Manu is stalled 3. std.experimental.image, ok ok, this one is mine but its about time we get the damn interfaces standardized in Phobos! 4. GUI widget rendering library, GUI widget != GUI toolkit (think just rendering a single control), basically handles rendering + accessibility. These are not big projects, but they would go a long way to getting any thing really "cool". Keep in mind, money talks. Getting the right people available to work on a project full time for a few months would really speed a lot of this stuff up. There is also other stuff like a bidi library, font rasterizer and other more useful infrastructure projects that actually has a chance of being completed. Depending on how far you want to go, my list alone is a few years worth of work for a 3-5 people team. What about the following idea: We extend DUB (the website) with a direct link for a paypal donation for every package. If a certain minimum amount is in, the D Foundation organizes the funding of the ongoing development? Regards mt.
Re: Hiring D programmers (with cryptography and blockchain knowledge are preferred)
On Thursday, 13 July 2017 at 05:18:40 UTC, wigy wrote: On Wednesday, 12 July 2017 at 20:11:06 UTC, Vitor Rozsas wrote: So... suggestions... Centralized? Decentralized? I think the centralized wouldn't fit in any country. It would certainly contain pedophile posts... and any sane country would shut down the servers immediately... So... DEcentralized? Hi! I do not think the debate you have with yourself is decentralized vs centralized. You are thinking about moderated vs unmoderated. One is a technical structure, the other is a social one. We got used to have moderated channels in media and unmoderated channels in person. Now the problem we are facing is that we use these social media platforms for replacing "in person" communications with friend and family. And the owners of these platforms are still treating it as "media" that they should moderate. But this is not so black-and-white still. When i am talking to my mother-in-law who has different political biases than me, I moderate *myself* not to bring up topics that would just divide us, because I love her enough to tolerate her opinions. What happens is that we have many social circles in which we have different topics and ethical norms. This is in our nature and that is fine. Football fans ventilate their emotions at the game, but they would not use the same language in their workplace. So what I see is that a social media platform should be decentralized to avoid influence from its owner. It should be divided into many communities. And each community should be able to downvote content that is not tolerated in those circles. And downvoted content should be also available by others, it should just take more actions to peek into that and convince yourself that it was indeed something inapt for that community. In the digital world, everything seems to be black and white. But social behaviors are more subtle than that. It is easy to create a total dictatorial system like facebook, and it is also easy to create a total anarchist system like Silk Road. And our goal is to create a system that is similar to in-real-life communication, which is neither completely free, nor completely controlled. You cannot build that system on top of a centralized architecture where a government can just ask for all data including a order to keep that secret. People never trusted their inner thoughts or family conversation onto the government. And they should not. This answer is brilliant...it comes out of understanding...deep thinking Now...personally I don't think this social media platform potential hype will last. Very soon it will get out of hand. It means Facebook, etc. that doesn't offer any quantifiable needed value will die if they don't innovate out of the social media realm. Presure from governments and users needs will facilitate the death of the social media market. But one thing is certain. Security is becoming a problem. Monies are becoming digital. There is a rising need for securing digital value without sacrificing convenience. Our systems today are not designed for that. Sooner or later those patches we are making to our systems temporary will get exhusted. Its either an innovative use of blockchain-like systems or a secure-from-scratch sandboxing system. Blockchain seem interesting for D.
Re: Munich D Meetup July 2017
Bump the thread, the next Munich D Meetup is getting closer. Dragos On Monday, 3 July 2017 at 18:23:27 UTC, Dragos Carp wrote: Hi all, On 18 July, we will have our next Munich meetup. Mario will give a talk with the title "Avoiding the Big Ball of Mud". As usual before and after the talk we will also have good conversations with pizza and drinks. Please RSVP on: https://www.meetup.com/de-DE/Munich-D-Programmers/events/241264180/ Thanks, Dragos
Re: Hiring D programmers (with cryptography and blockchain knowledge are preferred)
On 14/07/2017 7:10 PM, Martin Tschierschke wrote: On Friday, 14 July 2017 at 15:21:09 UTC, rikki cattermole wrote: [...] Scale your dreams back a bit. If you want to put money into anything here is my list: 1. Get shared libraries 100% working, with clear articles on how to use it, on every platform. 2. std.experimental.color, it seems Manu is stalled 3. std.experimental.image, ok ok, this one is mine but its about time we get the damn interfaces standardized in Phobos! 4. GUI widget rendering library, GUI widget != GUI toolkit (think just rendering a single control), basically handles rendering + accessibility. These are not big projects, but they would go a long way to getting any thing really "cool". Keep in mind, money talks. Getting the right people available to work on a project full time for a few months would really speed a lot of this stuff up. There is also other stuff like a bidi library, font rasterizer and other more useful infrastructure projects that actually has a chance of being completed. Depending on how far you want to go, my list alone is a few years worth of work for a 3-5 people team. What about the following idea: We extend DUB (the website) with a direct link for a paypal donation for every package. If a certain minimum amount is in, the D Foundation organizes the funding of the ongoing development? Regards mt. That is a service, which D Foundation shouldn't be offering. Anyway, they don't have enough money to do that. Somebody needs to come in and foot the bill, otherwise this type of infrastructure code is all on peoples own time and desire.