Good news!Last two weeks i was working on GUI.
Probably it was not the best idea, since our project have another important problems (i'll talk about it below), but, since I've already started, I wanted to finish the it.
So, i'm happy to announce that we started testing GUI public betta. This is cross-platform, qt-based, standalone application, that's no needed other proccess to be runned around. But, at the same time, it is designed to let user connect GUI to already running daemon.
As we tested, it works ok under Windows, Linux and MacOS.
I've just merged GUI branch into master and put build instructions on building GUI to website:
http://boolberry.com/howto.html#build_gui (by default GUI is not included into the build since it have more dependencies)
Prebuild binraies for windows uploaded.
Prebuild MacOS binaries will be tomorrow i hope (if anyone of Mac-owners wish to make stand-alone GUI build - you welcome!).
http://boolberry.com/downloads.htmlScreen shot:
Known bugs:
* Many places where html not well-adjustment yet
* Sometimes qt-boolb stay running after window is closed
This GUI use qtWebKit as renderer, so appearance is described via html/javascript. This means that you could experiment with GUI without any knowlage of C++, you doesn't even need to recompile project - just open "html" folder and get fun.
Please, send bug reports to github:
https://github.com/cryptozoidberg/boolberry/issues?labels=bug&state=open or here, in this thread.
We have a lot of todo's for this GUI, here is just a some of it:
1. Implement multi-wallet mode (keep opened few wallets)
2. Add more information on dashboard
3. Add commandline params to work with remote daemon
4. Add aliases support to address (alias name auto-completion, name checking)
5. Add multi recipient option to transfer
6. Add crushdumps collecting
7. Implement colored coin's support both in wallet and in GUI
................
Another good news is that we finally got logo for project:
As i learned before, there are always someone who don't like it, well ... if someone don't like it - there are lot of cryptonote forks where to go
Want to say many thanks to Bitcrea for his wonderful work and cooperation!Now lets talk about GPU.As someone already know, anotherone just guess - GPU miner was implemented for wild keccak.
As I found out, GPU implementation works on enterprise GPU hardware 7 times faster than CPU, they use textures memory cache to get some boost until wild keccak scratchpad is not big enough. (Currently wild keccak scratchpad is about 7MB)
Because size of scratchpad is growing day by day, I expect that the problem is not permanent, but now it hurts this project very much.
Mostly it seen by currency exchange rate, it pressed hard.
I do not underestimate the seriousness of this situation, and I realize that these things can destroy a project despite the good ideas of the coin. Than's why probably was not the best idea to keep making GUI while project was hurting buy GPU-cowboys.
Now I moved my focus on solving this problem and will make every effort to return the project to the way of fair distribution of coins.
One of the possible way out of the situation is to implement support of stratum-pool for BBR.
According to cbuchner1's post, they thinking of publishing their GPU miner with reasonable conditions:
But rejoice, there will be an open sourced Wild Keccak nVidia GPU miner published by us,
as soon as the Boolberry developer(s) provide
a) some stratum capable pool code for Wild Keccak
b) a cpu miner based on cpuminer-multi with some working stratum support
Due to the need for transferring and updating the Wild Keccak scratchpad this
won't be trivial, but it's also not impossible.
once a) and b) are available, it won't take long for us to either integrate this into
ccminer or publish a fork of cpuminer-multi with some nVidia support.
Christian
It would be a good, if Christian is not joking:)
Well, if someone who had experience with stratum pools or cpuminer-multi would like to participate - he is welcome.
Anyway, i'll try to implement it since stratum is desirable technology in anycase.
Is there any another way to solve this problem ?I guess yes, it is. If we gonna stuck with that situation for some reason, i'll do hardfork with changeing PoW algo(just fix details, i'm NOT gonna leave blockchain-based PoW philosophy), and as well i'll change difficulty adjustment algo, since CN-coins have not the best one. But this is plan B, the worst case.
PS: Want to say also many thanks to
dga for caring about simpleminer and pools.
Show Posts
