I would just take out ...OpenSSL...
I'd love to see that. Let me know how far you get!
Topic: Rewritting code and some ideas... (Read 561 times)
I'd love to see that. Let me know how far you get!
I am normally the "If it ain't broke, don't fix it" type, but I do think it would be convenient to have fewer dependencies. I think the client would be better off getting away from Qt and boost. I think using openssl makes sense, but if it weren't needed, that would be nice for me.
I don't know about your project idea of selecting all those variables for perceived happiness but if you want a refactored version of bitcoin there is "libcoin"
https://github.com/libcoin/libcoin
https://github.com/libcoin/libcoin
Nice, this is just what i wanted
Seems that is well documented and well structured. I would just take out "Berkeley DB, Boost and OpenSSL" and put Qt. But having it clean it is not a problem anymore You saved my holidays and a reinvented wheel. Thx
I don't know about your project idea of selecting all those variables for perceived happiness but if you want a refactored version of bitcoin there is "libcoin"
https://github.com/libcoin/libcoin
https://github.com/libcoin/libcoin
There are some alternative bitcoin clients, you may want to take a look.
Hi.
I just wanted ask for opinion and reasons. I guess everybody applied the rule "if it works dont touch it" but looking at the code, i just hate everything i see (i am working with C++ for 8 years already)
I think it is a real problem in bitcoin/litecoin/ppcoin/***coin and i was thinking about work on it... but before starting i wanna save time asking to the ones were checking the code already.
I think it is possible to take out dependencies of bitcoin and delete millions of globals used on it, clarify and simplify lot of things. Using Qt i think there is no sense to use boost and openssl. If i am wrong or u think it would take too much time, just answer "dont try it". Cleaning the app of SHIT would make lot easier to port it to android/iphone, now that Qt is going to work there (it is already)
I just know any static code analyzer would say the whole program is a shit, and all these is just making too complicated something should be easy. All this is because i want to make an experiment adding a learning algorithm to decide reward/difficult/fee
using as input:
reward
difficult
fee
time from last found block
nº transactions
selectable vote applied to transactions/mining so people can decide how happy they are.
And try to maximize this last parameter.
Other idea i didnt see.... but i am quite sure it exists already (or should), is a way to print notes that can be charged to a wallet using the mobile.
thx for your time.
I just wanted ask for opinion and reasons. I guess everybody applied the rule "if it works dont touch it" but looking at the code, i just hate everything i see (i am working with C++ for 8 years already)
I think it is a real problem in bitcoin/litecoin/ppcoin/***coin and i was thinking about work on it... but before starting i wanna save time asking to the ones were checking the code already.
I think it is possible to take out dependencies of bitcoin and delete millions of globals used on it, clarify and simplify lot of things. Using Qt i think there is no sense to use boost and openssl. If i am wrong or u think it would take too much time, just answer "dont try it". Cleaning the app of SHIT would make lot easier to port it to android/iphone, now that Qt is going to work there (it is already)
I just know any static code analyzer would say the whole program is a shit, and all these is just making too complicated something should be easy. All this is because i want to make an experiment adding a learning algorithm to decide reward/difficult/fee
using as input:
reward
difficult
fee
time from last found block
nº transactions
selectable vote applied to transactions/mining so people can decide how happy they are.
And try to maximize this last parameter.
Other idea i didnt see.... but i am quite sure it exists already (or should), is a way to print notes that can be charged to a wallet using the mobile.
thx for your time.
Are you asking : "Should I optimise the code to make porting easier ?"
Who would say no to that
Hi.
I just wanted ask for opinion and reasons. I guess everybody applied the rule "if it works dont touch it" but looking at the code, i just hate everything i see (i am working with C++ for 8 years already)
I think it is a real problem in bitcoin/litecoin/ppcoin/***coin and i was thinking about work on it... but before starting i wanna save time asking to the ones were checking the code already.
I think it is possible to take out dependencies of bitcoin and delete millions of globals used on it, clarify and simplify lot of things. Using Qt i think there is no sense to use boost and openssl. If i am wrong or u think it would take too much time, just answer "dont try it". Cleaning the app of SHIT would make lot easier to port it to android/iphone, now that Qt is going to work there (it is already)
I just know any static code analyzer would say the whole program is a shit, and all these is just making too complicated something should be easy. All this is because i want to make an experiment adding a learning algorithm to decide reward/difficult/fee
using as input:
reward
difficult
fee
time from last found block
nº transactions
selectable vote applied to transactions/mining so people can decide how happy they are.
And try to maximize this last parameter.
Other idea i didnt see.... but i am quite sure it exists already (or should), is a way to print notes that can be charged to a wallet using the mobile.
thx for your time.
I just wanted ask for opinion and reasons. I guess everybody applied the rule "if it works dont touch it" but looking at the code, i just hate everything i see (i am working with C++ for 8 years already)
I think it is a real problem in bitcoin/litecoin/ppcoin/***coin and i was thinking about work on it... but before starting i wanna save time asking to the ones were checking the code already.
I think it is possible to take out dependencies of bitcoin and delete millions of globals used on it, clarify and simplify lot of things. Using Qt i think there is no sense to use boost and openssl. If i am wrong or u think it would take too much time, just answer "dont try it". Cleaning the app of SHIT would make lot easier to port it to android/iphone, now that Qt is going to work there (it is already)
I just know any static code analyzer would say the whole program is a shit, and all these is just making too complicated something should be easy. All this is because i want to make an experiment adding a learning algorithm to decide reward/difficult/fee
using as input:
reward
difficult
fee
time from last found block
nº transactions
selectable vote applied to transactions/mining so people can decide how happy they are.
And try to maximize this last parameter.
Other idea i didnt see.... but i am quite sure it exists already (or should), is a way to print notes that can be charged to a wallet using the mobile.
thx for your time.
Jump to: