Jolly good
ill have a go at compiling soon ...
will look for the git - then let you know how successful i am with it ...
#crysx
Ok no problems - let us know
i found the git - and cloned and the first thing i noticed was quite disturbing ...
the code had to have a couple of things done to it before an attempt at a compile was done ...
- the code has all been set to 'x' ecutable ... all of it ... every file and folder ...
- a change in the boost code was required ( rpcrawtransaction.cpp ) ...
- leveldb had to be compiled manually before the main qt compile was to occur ...
- leveldb needed a small code change ( Makefile ) ... though this may be a fedora 23 x64 issue ... i dont have ubuntu to test ...
- qt4 was the only way to compile - as the code is incompatible with qt5 ... as this may also be a fedora 23 x64 issue as well - i dont have this issue with other wallet builds ...
i will now test the daemon build - but cannot see any issue with that if qt now compiles ...
the compile completed after these changes were made - but its concerning that the whole code was set with the executable attribute ... i have no idea why you would want to - or even have a reason for - setting the entire codebase as executable ...
when run - the wallet is running smooth ... quite smooth ... it took a minute to get the first of the nodes - but sync is all good ...
the wallet has been running well - and ive only just shut it down a few minutes ago ...
nice ...
#crysx
thanks firstly for the feedback on your findings.
The code is for ubuntu and has not been tested on fedora or written for it. Can you please give me a reference to the executable attribute as only the binary is executable.
no worries mate ...
below is a copy of the listing of a clean clone of the git ...
[root@horus BATA-SOURCE]# ll
total 4464
-rw-r--r--. 1 root root 10454 Feb 26 17:47 bata.png
-rwxr-xr-x. 1 root root 14145 Feb 26 17:47 bata-qt.pro
-rwxr-xr-x. 1 root root 4499483 Feb 26 17:47 bata-src.zip
drwxr-xr-x. 2 root root 12288 Feb 26 23:55 build
drwxr-xr-x. 13 root root 4096 Feb 26 17:47 contrib
-rwxr-xr-x. 1 root root 1195 Feb 26 17:47 COPYING
drwxr-xr-x. 2 root root 4096 Feb 26 17:47 doc
-rwxr-xr-x. 1 root root 305 Feb 26 17:47 INSTALL
-rwxr-xr-x. 1 root root 2555 Feb 26 17:47 README.md
drwxr-xr-x. 5 root root 4096 Feb 26 17:47 share
drwxr-xr-x. 8 root root 4096 Feb 26 20:51 src
notice the attributes are ALL executable ( actually except for bata.png ) ... and that is all the way through the codebase ...
its easy to fix - but takes a few minutes ...
have you tried to compile it using this code under todays updates? ... does it work? ...
try it - you will be surprised that it doesnt with the latest updates for boost ...
#crysx
I appreciate your assistance.
crysx has submitted some changes to git for fedora which I will look over later today.
no worries mate ...
as explained in the pm a few moments ago ...
a few more changes have been done and all the issues i saw have been changed and corrected ...
- the attributes have been corrected and only those that require it are executable ...
- changes made to the leveldb for compilation under qt4 for fedora 23 x64 ... this 'may' not work for ubuntu - but needs to be tested ...
- removed the source .zip file - as its not required for the compile and is supplied by git on request anyway ...
- changes to the boost code in rpcrawtransaction to eliminate boost compilation issues ...
- changes made to the bata-qt.pro file for auto leveldb compilation ... no more manual compiling of leveldb ...
this compiling call from .pro 'sometimes' causes the windows install to register as a 'false positive' on some antivirus applications ... it is NOT a virus ... just a signature from the files used for compilation ... it doesnt always happen - but it may - depending on what windows compilation versions are used to build the windows qt wallet ...
if the windows qt wallet is built in linux - by cross compiling - it is not likely to show this as a false positive in antivirus apps ... eitherway - its NOT a virus ...
these changes have been pull requested in git ... and i hope they meet your approval for integration into the code ...
tanx again ...
#crysx