Official Anoncoin chat thread (including history)

[jolskeboll]

https://twitter.com/mikalv/status/606826974418464768

Got this notification from Meeh. I guess you can translate yourselves \,,,/

[SmokingSkull]

https://twitter.com/mikalv/status/606826974418464768

Got this notification from Meeh. I guess you can translate yourselves \,,,/

don't believe that, meeh just talks a lot...

[alacast]

Hi,
I'm trying to compile the develop-themes on debian testing (stretch), and I have some errors related to sse2 as much as I can tell.
I did 2 or 3 hours of research, no luck, so I'm turning to you because I can't acces IRC for the moment

what I did:
After succesfully running autogen.sh and configure, make stops and gives these particular errors (I'm giving interestings chunks of logs only):

Code:

CXX      libanoncoin_common_a-scrypt-sse2.o
scrypt-sse2.cpp: In function ‘void scrypt_1024_1_1_256_sp_sse2(const char*, char*, char*)’:
scrypt-sse2.cpp:124:50: warning: SSE vector return without SSE enabled changes the ABI [-Wpsabi]
    X.i128[k] = _mm_xor_si128(X.i128[k], V[j + k]);
                                                  ^

then many times this same error, repeatedly:

Code:

In file included from scrypt-sse2.cpp:36:0:
/usr/lib/gcc/i586-linux-gnu/4.9/include/emmintrin.h:1282:1: error: inlining failed in call to always_inline ‘__m128i _mm_xor_si128(__m128i, __m128i)’: target specific option mismatch
 _mm_xor_si128 (__m128i __A, __m128i __B)
 ^

and it all finishes like that:

Code:

scrypt-sse2.cpp:92:31: error: called from here
  B[3] = _mm_add_epi32(B[3], X3);
                               ^
Makefile:2674: recipe for target 'libanoncoin_common_a-scrypt-sse2.o' failed
make[2]: *** [libanoncoin_common_a-scrypt-sse2.o] Error 1
make[2]: Leaving directory '/home/user/Code/anoncoin-develop-themes/src'
Makefile:5136: recipe for target 'all-recursive' failed
make[1]: *** [all-recursive] Error 1
make[1]: Leaving directory '/home/user/Code/anoncoin-develop-themes/src'
Makefile:573: recipe for target 'all-recursive' failed
make: *** [all-recursive] Error 1

I tried to add -msse2 flags to the makefile on the CPPFLAGS  CXXFLAGS lines, but none of these work (nothing changed, and I did run "make clean" after every try).
Is it really an sse2 related error? where can I make sure the -msse2 option is passed to the compiler?
And yes my CPU has the sse2 flag (as far as cat /proc/cpuinfo can tell).

Hope anyone of you can help me

[AnonCoinTwitter]

https://twitter.com/mikalv/status/606826974418464768

Got this notification from Meeh. I guess you can translate yourselves \,,,/

don't believe that, meeh just talks a lot...

Google translation.....

It should work fine as far as I know. Has been "offline" in recent months, but should take a look at the project next week.

I have not spoken to Meeh so have no further comment at this time.

[lunokhod2]

Hi,
I'm trying to compile the develop-themes on debian testing (stretch), and I have some errors related to sse2 as much as I can tell.
I did 2 or 3 hours of research, no luck, so I'm turning to you because I can't acces IRC for the moment

what I did:
After succesfully running autogen.sh and configure, make stops and gives these particular errors (I'm giving interestings chunks of logs only):

Code:

CXX      libanoncoin_common_a-scrypt-sse2.o
scrypt-sse2.cpp: In function ‘void scrypt_1024_1_1_256_sp_sse2(const char*, char*, char*)’:
scrypt-sse2.cpp:124:50: warning: SSE vector return without SSE enabled changes the ABI [-Wpsabi]
    X.i128[k] = _mm_xor_si128(X.i128[k], V[j + k]);
                                                  ^

then many times this same error, repeatedly:

Code:

In file included from scrypt-sse2.cpp:36:0:
/usr/lib/gcc/i586-linux-gnu/4.9/include/emmintrin.h:1282:1: error: inlining failed in call to always_inline ‘__m128i _mm_xor_si128(__m128i, __m128i)’: target specific option mismatch
 _mm_xor_si128 (__m128i __A, __m128i __B)
 ^

and it all finishes like that:

Code:

scrypt-sse2.cpp:92:31: error: called from here
  B[3] = _mm_add_epi32(B[3], X3);
                               ^
Makefile:2674: recipe for target 'libanoncoin_common_a-scrypt-sse2.o' failed
make[2]: *** [libanoncoin_common_a-scrypt-sse2.o] Error 1
make[2]: Leaving directory '/home/user/Code/anoncoin-develop-themes/src'
Makefile:5136: recipe for target 'all-recursive' failed
make[1]: *** [all-recursive] Error 1
make[1]: Leaving directory '/home/user/Code/anoncoin-develop-themes/src'
Makefile:573: recipe for target 'all-recursive' failed
make: *** [all-recursive] Error 1

I tried to add -msse2 flags to the makefile on the CPPFLAGS  CXXFLAGS lines, but none of these work (nothing changed, and I did run "make clean" after every try).
Is it really an sse2 related error? where can I make sure the -msse2 option is passed to the compiler?
And yes my CPU has the sse2 flag (as far as cat /proc/cpuinfo can tell).

Hope anyone of you can help me

Did you follow the instructions here: https://wiki.anoncoin.net/How_to_build_Anoncoin_develop, and build everything in the depends directory?

[alacast]

Did you follow the instructions here: https://wiki.anoncoin.net/How_to_build_Anoncoin_develop, and build everything in the depends directory?

wow, thank you lunokhod, I was totally looking in the wrong direction... not building in the depends folder was my error. I needed to "cd depends" after autogen.sh
I'm used to stick to the README and INSTALL files when I'm compiling, and the instructions in the INSTALL file of the anoncoin-developp-themes were:

Code:

./autogen.sh
./configure
make
make install

Next time I'll look more carrefully in the wiki before posting here.

So now I have another building problem:

I ran autogen.sh in the root of the anoncoin-developp-themes uncompressed folder, then ran cd depends, and make.
But after 20 minutes or so, it is stuck at "fetching qt46", and top command shows no cpu activity for that task. I stopped, ran make again, same thing. All dependencies are satisfied by the way.

[Cryptoslave]

Did you follow the instructions here: https://wiki.anoncoin.net/How_to_build_Anoncoin_develop, and build everything in the depends directory?

wow, thank you lunokhod, I was totally looking in the wrong direction... not building in the depends folder was my error. I needed to "cd depends" after autogen.sh
I'm used to stick to the README and INSTALL files when I'm compiling, and the instructions in the INSTALL file of the anoncoin-developp-themes were:

Code:

./autogen.sh
./configure
make
make install

Next time I'll look more carrefully in the wiki before posting here.

So now I have another building problem:

I ran autogen.sh in the root of the anoncoin-developp-themes uncompressed folder, then ran cd depends, and make.
But after 20 minutes or so, it is stuck at "fetching qt46", and top command shows no cpu activity for that task. I stopped, ran make again, same thing. All dependencies are satisfied by the way.

Fetching QT download it from web and this can take alot of time (it is the most long task in the whole building).

Let it run for 30 min - 1 h before becoming worried (take 20 min here)...

[alacast]

...ehrr.... there is still this same problem as before.

I finished to build in the depends directory. Thank you lunokhod and cryptoslave for your fast answers.
Fetching qt46 was long, and fetching bdb was a little too.

And then after that I went back and cleaned the "root" directory anoncoin-develop-themes. I ran 'make clean', deleted the configure file, the makefile and all other files standing there except directories. Then I uncompressed the original develop-themes.zip again to take all the original files needed. As there is no configure script at first, I ran ./autogen.sh in there (not in depends anymore).

and after that I did  accordingly to the wiki tutorial shown by lunokhod (https://wiki.anoncoin.net/How_to_build_Anoncoin_develop)

Code:

./configure --prefix=`pwd`/depends/i686-pc-linux-gnu

And then it's here again, same errors:

Code:

Making all in src
make[1]: Entering directory '/home/user/Code/anoncoin-develop-themes/src'
make[2]: Entering directory '/home/user/Code/anoncoin-develop-themes/src'
  CXX      libanoncoin_common_a-scrypt-sse2.o
scrypt-sse2.cpp: In function ‘void scrypt_1024_1_1_256_sp_sse2(const char*, char*, char*)’:
scrypt-sse2.cpp:124:50: warning: SSE vector return without SSE enabled changes the ABI [-Wpsabi]
    X.i128[k] = _mm_xor_si128(X.i128[k], V[j + k]);
                                                  ^
In file included from scrypt-sse2.cpp:36:0:
/usr/lib/gcc/i586-linux-gnu/4.9/include/emmintrin.h: In function ‘void xor_salsa8_sse2(__m128i*, const __m128i*)’:
/usr/lib/gcc/i586-linux-gnu/4.9/include/emmintrin.h:1282:1: error: inlining failed in call to always_inline ‘__m128i _mm_xor_si128(__m128i, __m128i)’: target specific option mismatch
 _mm_xor_si128 (__m128i __A, __m128i __B)
 ^
scrypt-sse2.cpp:44:39: error: called from here
  X0 = B[0] = _mm_xor_si128(B[0], Bx[0]);
                                       ^
In file included from scrypt-sse2.cpp:36:0:
/usr/lib/gcc/i586-linux-gnu/4.9/include/emmintrin.h:1282:1: error: inlining failed in call to always_inline ‘__m128i _mm_xor_si128(__m128i, __m128i)’: target specific option mismatch
 _mm_xor_si128 (__m128i __A, __m128i __B)
 ^
scrypt-sse2.cpp:45:39: error: called from here
  X1 = B[1] = _mm_xor_si128(B[1], Bx[1]);

and finishes like that:

Code:

scrypt-sse2.cpp:92:31: error: called from here
  B[3] = _mm_add_epi32(B[3], X3);
                               ^
Makefile:2674: recipe for target 'libanoncoin_common_a-scrypt-sse2.o' failed
make[2]: *** [libanoncoin_common_a-scrypt-sse2.o] Error 1
make[2]: Leaving directory '/home/user/Code/anoncoin-develop-themes/src'
Makefile:5136: recipe for target 'all-recursive' failed
make[1]: *** [all-recursive] Error 1
make[1]: Leaving directory '/home/user/Code/anoncoin-develop-themes/src'
Makefile:573: recipe for target 'all-recursive' failed
make: *** [all-recursive] Error 1

So the only way I see to be sure is to delete everything and restart building clean. will do that in a few hours I think...

[jolskeboll]

https://twitter.com/mikalv/status/606826974418464768

Got this notification from Meeh. I guess you can translate yourselves \,,,/

don't believe that, meeh just talks a lot...

Google translation.....

It should work fine as far as I know. Has been "offline" in recent months, but should take a look at the project next week.

I have not spoken to Meeh so have no further comment at this time.

Maybe, but this is definitively a positive thing. He is alive and seems to be positive about the project anyways. He could have stated that he does not want to be affiliated with Anoncoin anymore too, but he didn't. I think he moved from Norway to Cebu ( http://en.wikipedia.org/wiki/Cebu ) and it probably takes some time to settle down. May be that he eventually returns to the project when life over there gets boring . We have seen coin projects where this has happened before... a developer with respect suddenly returns or takes over a "dead" coin and the value of the coin multi-ten-folds.

Just my thoughts

[lunokhod2]

https://twitter.com/mikalv/status/606826974418464768

Got this notification from Meeh. I guess you can translate yourselves \,,,/

don't believe that, meeh just talks a lot...

Google translation.....

It should work fine as far as I know. Has been "offline" in recent months, but should take a look at the project next week.

I have not spoken to Meeh so have no further comment at this time.

Maybe, but this is definitively a positive thing. He is alive and seems to be positive about the project anyways. He could have stated that he does not want to be affiliated with Anoncoin anymore too, but he didn't. I think he moved from Norway to Cebu ( http://en.wikipedia.org/wiki/Cebu ) and it probably takes some time to settle down. May be that he eventually returns to the project when life over there gets boring . We have seen coin projects where this has happened before... a developer with respect suddenly returns or takes over a "dead" coin and the value of the coin multi-ten-folds.

Just my thoughts

Just to be clear: Anoncoin is not a dead coin. We have a new lead developer, GroundRod, who has brought the code up to bitcoin 0.10 standards, and who has added many new features, such as a innovative PID difficulty retargeting algorithm (see https://wiki.anoncoin.net/Development_schedule). You can download/build the the develop branch on github to see for yourself. We are currently testing the PID algorithm on internal testnets.

But to get back to your original comment: Meeh has not given up on Anoncoin, but even if he did, the project would live on without him.

[jolskeboll]

Just to be clear: Anoncoin is not a dead coin. We have a new lead developer, GroundRod, who has brought the code up to bitcoin 0.10 standards, and who has added many new features, such as a innovative PID difficulty retargeting algorithm (see https://wiki.anoncoin.net/Development_schedule). You can download/build the the develop branch on github to see for yourself. We are currently testing the PID algorithm on internal testnets.

But to get back to your original comment: Meeh has not given up on Anoncoin, but even if he did, the project would live on without him.

Sorry, maybe I was a bit harsh and you are right.

[alacast]

Just to let others know, I tried again to build the develop-themes branch on Debian (9) stretch x86 (and I'm not in a VM) but this time I deleted every anoncoin directories and started all new. I did follow the instructions and did exactly what is told here https://wiki.anoncoin.net/How_to_build_Anoncoin_develop , except that I did 'git checkout develop-themes' in that case.
Unfortunately the last 'make' command still ends with the scrypt-sse2 error that I already posted, and no anoncoin programs are build.

So my next step is to install debian 8 (Jessie) in a VM and try again.
Sorry to flood here. Should I fill a bug report somewhere? can it be a problem with the chipset of my laptop?

[K1773R]

Just to let others know, I tried again to build the develop-themes branch on Debian (9) stretch x86 (and I'm not in a VM) but this time I deleted every anoncoin directories and started all new. I did follow the instructions and did exactly what is told here https://wiki.anoncoin.net/How_to_build_Anoncoin_develop , except that I did 'git checkout develop-themes' in that case.
Unfortunately the last 'make' command still ends with the scrypt-sse2 error that I already posted, and no anoncoin programs are build.

So my next step is to install debian 8 (Jessie) in a VM and try again.
Sorry to flood here. Should I fill a bug report somewhere? can it be a problem with the chipset of my laptop?

Compiling has nothing to do with chipsets.

You can open a Issue on Github with all details you have.

[SmokingSkull]

Just to be clear: Anoncoin is not a dead coin. We have a new lead developer, GroundRod, who has brought the code up to bitcoin 0.10 standards, and who has added many new features, such as a innovative PID difficulty retargeting algorithm (see https://wiki.anoncoin.net/Development_schedule). You can download/build the the develop branch on github to see for yourself. We are currently testing the PID algorithm on internal testnets.

But to get back to your original comment: Meeh has not given up on Anoncoin, but even if he did, the project would live on without him.

Sorry, maybe I was a bit harsh and you are right.

Development branch hasn't been altered since February 2015.
We don't see GroundRod coding^^ or getting the code up to 0.1 standards

not blame - just saying

[lunokhod2]

Just to be clear: Anoncoin is not a dead coin. We have a new lead developer, GroundRod, who has brought the code up to bitcoin 0.10 standards, and who has added many new features, such as a innovative PID difficulty retargeting algorithm (see https://wiki.anoncoin.net/Development_schedule). You can download/build the the develop branch on github to see for yourself. We are currently testing the PID algorithm on internal testnets.

But to get back to your original comment: Meeh has not given up on Anoncoin, but even if he did, the project would live on without him.

Sorry, maybe I was a bit harsh and you are right.

Development branch hasn't been altered since February 2015.
We don't see GroundRod coding^^ or getting the code up to 0.1 standards

not blame - just saying

You are right in not trusting us. All I can say is that if you look at your Anoncoin peers, you will see about three that are running 0.9.6.1. We are holding back on pushing the post 0.9.4 code to github as some of our developments, including real block hashes and PID retargetting, could be easily copied.

Of course, it would be easy to make some clients with fake version numbers if we wanted to pump and dump.

[entertheabyss]

Just to be clear: Anoncoin is not a dead coin. We have a new lead developer, GroundRod, who has brought the code up to bitcoin 0.10 standards, and who has added many new features, such as a innovative PID difficulty retargeting algorithm (see https://wiki.anoncoin.net/Development_schedule). You can download/build the the develop branch on github to see for yourself. We are currently testing the PID algorithm on internal testnets.

But to get back to your original comment: Meeh has not given up on Anoncoin, but even if he did, the project would live on without him.

Sorry, maybe I was a bit harsh and you are right.

Development branch hasn't been altered since February 2015.
We don't see GroundRod coding^^ or getting the code up to 0.1 standards

not blame - just saying

i have tested the new 9.6.1 client and can confirm it does exist and is has lots of improvements from the last public release. I cant say that im enthusiastic about how secretive the current anoncoin leaders are about it but progress is being made.

[entertheabyss]

Just to be clear: Anoncoin is not a dead coin. We have a new lead developer, GroundRod, who has brought the code up to bitcoin 0.10 standards, and who has added many new features, such as a innovative PID difficulty retargeting algorithm (see https://wiki.anoncoin.net/Development_schedule). You can download/build the the develop branch on github to see for yourself. We are currently testing the PID algorithm on internal testnets.

But to get back to your original comment: Meeh has not given up on Anoncoin, but even if he did, the project would live on without him.

Sorry, maybe I was a bit harsh and you are right.

Development branch hasn't been altered since February 2015.
We don't see GroundRod coding^^ or getting the code up to 0.1 standards

not blame - just saying

You are right in not trusting us. All I can say is that if you look at your Anoncoin peers, you will see about three that are running 0.9.6.1. We are holding back on pushing the post 0.9.4 code to github as some of our developments, including real block hashes and PID retargetting, could be easily copied.

Of course, it would be easy to make some clients with fake version numbers if we wanted to pump and dump.

As i have stated before I think we should push the current 9.6.1 code to a public development branch. More transparency would be refreshing,

[K1773R]

Just to be clear: Anoncoin is not a dead coin. We have a new lead developer, GroundRod, who has brought the code up to bitcoin 0.10 standards, and who has added many new features, such as a innovative PID difficulty retargeting algorithm (see https://wiki.anoncoin.net/Development_schedule). You can download/build the the develop branch on github to see for yourself. We are currently testing the PID algorithm on internal testnets.

But to get back to your original comment: Meeh has not given up on Anoncoin, but even if he did, the project would live on without him.

Sorry, maybe I was a bit harsh and you are right.

Development branch hasn't been altered since February 2015.
We don't see GroundRod coding^^ or getting the code up to 0.1 standards

not blame - just saying

You are right in not trusting us. All I can say is that if you look at your Anoncoin peers, you will see about three that are running 0.9.6.1. We are holding back on pushing the post 0.9.4 code to github as some of our developments, including real block hashes and PID retargetting, could be easily copied.

Of course, it would be easy to make some clients with fake version numbers if we wanted to pump and dump.

As i have stated before I think we should push the current 9.6.1 code to a public development branch. More transparency would be refreshing,

Im against this. A lot of people still run the oldest working release! If we release this now, people will get it and stop bothering getting the finished new version.

[entertheabyss]

Just to be clear: Anoncoin is not a dead coin. We have a new lead developer, GroundRod, who has brought the code up to bitcoin 0.10 standards, and who has added many new features, such as a innovative PID difficulty retargeting algorithm (see https://wiki.anoncoin.net/Development_schedule). You can download/build the the develop branch on github to see for yourself. We are currently testing the PID algorithm on internal testnets.

But to get back to your original comment: Meeh has not given up on Anoncoin, but even if he did, the project would live on without him.

Sorry, maybe I was a bit harsh and you are right.

Development branch hasn't been altered since February 2015.
We don't see GroundRod coding^^ or getting the code up to 0.1 standards

not blame - just saying

You are right in not trusting us. All I can say is that if you look at your Anoncoin peers, you will see about three that are running 0.9.6.1. We are holding back on pushing the post 0.9.4 code to github as some of our developments, including real block hashes and PID retargetting, could be easily copied.

Of course, it would be easy to make some clients with fake version numbers if we wanted to pump and dump.

As i have stated before I think we should push the current 9.6.1 code to a public development branch. More transparency would be refreshing,

Im against this. A lot of people still run the oldest working release! If we release this now, people will get it and stop bothering getting the finished new version.

We could let everyone know that it is unstable. There hasn't been a lot of activity on GitHub in several months so I can empathize with the frustration within the community. what If we released screenshots of the new wallet? I think this could go a long way towards alleviating some of the concerns about progress members of the community may have.

[K1773R]

Just to be clear: Anoncoin is not a dead coin. We have a new lead developer, GroundRod, who has brought the code up to bitcoin 0.10 standards, and who has added many new features, such as a innovative PID difficulty retargeting algorithm (see https://wiki.anoncoin.net/Development_schedule). You can download/build the the develop branch on github to see for yourself. We are currently testing the PID algorithm on internal testnets.

But to get back to your original comment: Meeh has not given up on Anoncoin, but even if he did, the project would live on without him.

Sorry, maybe I was a bit harsh and you are right.

Development branch hasn't been altered since February 2015.
We don't see GroundRod coding^^ or getting the code up to 0.1 standards

not blame - just saying

You are right in not trusting us. All I can say is that if you look at your Anoncoin peers, you will see about three that are running 0.9.6.1. We are holding back on pushing the post 0.9.4 code to github as some of our developments, including real block hashes and PID retargetting, could be easily copied.

Of course, it would be easy to make some clients with fake version numbers if we wanted to pump and dump.

As i have stated before I think we should push the current 9.6.1 code to a public development branch. More transparency would be refreshing,

Im against this. A lot of people still run the oldest working release! If we release this now, people will get it and stop bothering getting the finished new version.

We could let everyone know that it is unstable. There hasn't been a lot of activity on GitHub in several months so I can empathize with the frustration within the community. what If we released screenshots of the new wallet? I think this could go a long way towards alleviating some of the concerns about progress members of the community may have.

Aslong the beta/WIP code isnt released, im fine with it.