profile
viewpoint
If you are wondering where the data of this site comes from, please visit https://api.github.com/users/db4/events. GitMemory does not store any data, but only uses NGINX to cache data for a period of time. The idea behind GitMemory is simply to give users a better reading experience.

db4/base 0

Standard library for OCaml

db4/Boost.Application 0

Boost Application Library

db4/conan 0

Conan - The open-source C/C++ package manager

db4/conan-center-index 0

Recipes for the ConanCenter repository

db4/conan-cmake-installer 0

CMake installer as a conan package

db4/conan-cmake_findboost_modular 0

Conan.io recipes containing FindBoost.cmake for Modular Boost packages

db4/conan-cygwin_installer 0

Conan.io recipes for Cygwin Installer

issue commentocaml/opam-repository

fileutils 0.6.3 sources are unavailable

curl -L https://github.com/gildor478/ocaml-fileutils/releases/download/v0.6.3/fileutils-v0.6.3.tbz

works for me. Does it work for you now? (if so i might have been a temporary github issue)

Yes, now it's OK. Indeed, looks like a temporary github issue.

db4

comment created time in 5 days

issue closedocaml/opam-repository

fileutils 0.6.3 sources are unavailable

opam install fails with

[ERROR] The sources of the following couldn't be obtained, aborting:
          - fileutils.0.6.3: curl error code 403

I also tried to download them manually: curl -L https://github.com/gildor478/ocaml-fileutils/releases/download/v0.6.3/fileutils-v0.6.3.tbz

<?xml version="1.0" encoding="UTF-8"?>
<Error><Code>SignatureDoesNotMatch</Code><Message>The request signature we calculated does not match the signature you provided. Check your key and signing method.</Message><AWSAccessKeyId>AKIAIWNJYAX4CSVEH53A</AWSAccessKeyId><StringToSign>AWS4-HMAC-SHA256
...

closed time in 5 days

db4

issue commentocaml/opam-repository

fileutils 0.6.3 sources are unavailable

@gildor478 can you look into this?

db4

comment created time in 6 days

issue openedocaml/opam-repository

fileutils 0.6.3 sources are unavailable

opam install fails with

[ERROR] The sources of the following couldn't be obtained, aborting:
          - fileutils.0.6.3: curl error code 403

I also tried to download them manually: curl -L https://github.com/gildor478/ocaml-fileutils/releases/download/v0.6.3/fileutils-v0.6.3.tbz

<?xml version="1.0" encoding="UTF-8"?>
<Error><Code>SignatureDoesNotMatch</Code><Message>The request signature we calculated does not match the signature you provided. Check your key and signing method.</Message><AWSAccessKeyId>AKIAIWNJYAX4CSVEH53A</AWSAccessKeyId><StringToSign>AWS4-HMAC-SHA256
...

created time in 6 days

issue openedconan-io/conan-center-index

[package] opencv/3.x: debug build with parallel=tbb fails

Package and Environment Details

  • Package Name/Version: opencv/3.4.12
  • Compiler+version: GCC 5
  • Docker image: conanio/gcc5

Steps to reproduce

 VERBOSE=1 conan install opencv/3.4.12@ -s build_type=Debug --build missing -o opencv:shared=True -o opencv:parallel=tbb

Logs

<details><summary>Click to expand log</summary>

...
[ 27%] Linking CXX shared library ../../../lib/libopencv_core.so
cd /home/conan/.conan/data/opencv/3.4.12/_/_/build/9a3d20f7cc729548ae41259efb954cbc0d1b7d4d/build_subfolder/source_subfolder/modules/core && /usr/bin/cmake -E cmake_link_script CMakeFiles/opencv_core.dir/link.txt --verbose=1
/usr/bin/g++ -fPIC -m64 -O3 -s -pthread -m64   -fsigned-char -W -Wall -Werror=return-type -Werror=non-virtual-dtor -Werror=address -Werror=sequence-point -Wformat -Werror=format-security -Wmissing-declarations -Wundef -Winit-self -Wpointer-arith -Wshadow -Wsign-promo -Wuninitialized -Winit-self -Wsuggest-override -Wno-delete-non-virtual-dtor -Wno-comment -fdiagnostics-show-option -Wno-long-long -pthread -fomit-frame-pointer -ffunction-sections -fdata-sections  -msse -msse2 -msse3 -fvisibility=hidden -fvisibility-inlines-hidden -g   -DDEBUG -D_DEBUG -m64 -L/home/conan/.opam/4.07.1+local+gcc5+x64+fpic/lib  -m64  -Wl,--gc-sections -Wl,--as-needed -shared -Wl,-soname,libopencv_core.so.3.4 -o ../../../lib/libopencv_core.so.3.4.12 CMakeFiles/opencv_core.dir/src/algorithm.cpp.o CMakeFiles/opencv_core.dir/src/alloc.cpp.o CMakeFiles/opencv_core.dir/src/arithm.cpp.o CMakeFiles/opencv_core.dir/src/arithm.dispatch.cpp.o CMakeFiles/opencv_core.dir/src/array.cpp.o CMakeFiles/opencv_core.dir/src/async.cpp.o CMakeFiles/opencv_core.dir/src/batch_distance.cpp.o CMakeFiles/opencv_core.dir/src/bindings_utils.cpp.o CMakeFiles/opencv_core.dir/src/buffer_area.cpp.o CMakeFiles/opencv_core.dir/src/channels.cpp.o CMakeFiles/opencv_core.dir/src/check.cpp.o CMakeFiles/opencv_core.dir/src/command_line_parser.cpp.o CMakeFiles/opencv_core.dir/src/conjugate_gradient.cpp.o CMakeFiles/opencv_core.dir/src/convert.dispatch.cpp.o CMakeFiles/opencv_core.dir/src/convert_c.cpp.o CMakeFiles/opencv_core.dir/src/convert_scale.dispatch.cpp.o CMakeFiles/opencv_core.dir/src/copy.cpp.o CMakeFiles/opencv_core.dir/src/count_non_zero.dispatch.cpp.o CMakeFiles/opencv_core.dir/src/cuda_gpu_mat.cpp.o CMakeFiles/opencv_core.dir/src/cuda_host_mem.cpp.o CMakeFiles/opencv_core.dir/src/cuda_info.cpp.o CMakeFiles/opencv_core.dir/src/cuda_stream.cpp.o CMakeFiles/opencv_core.dir/src/datastructs.cpp.o CMakeFiles/opencv_core.dir/src/directx.cpp.o CMakeFiles/opencv_core.dir/src/downhill_simplex.cpp.o CMakeFiles/opencv_core.dir/src/dxt.cpp.o CMakeFiles/opencv_core.dir/src/gl_core_3_1.cpp.o CMakeFiles/opencv_core.dir/src/glob.cpp.o CMakeFiles/opencv_core.dir/src/hal_internal.cpp.o CMakeFiles/opencv_core.dir/src/kmeans.cpp.o CMakeFiles/opencv_core.dir/src/lapack.cpp.o CMakeFiles/opencv_core.dir/src/lda.cpp.o CMakeFiles/opencv_core.dir/src/logger.cpp.o CMakeFiles/opencv_core.dir/src/lpsolver.cpp.o CMakeFiles/opencv_core.dir/src/lut.cpp.o CMakeFiles/opencv_core.dir/src/mathfuncs.cpp.o CMakeFiles/opencv_core.dir/src/mathfuncs_core.dispatch.cpp.o CMakeFiles/opencv_core.dir/src/matmul.dispatch.cpp.o CMakeFiles/opencv_core.dir/src/matrix.cpp.o CMakeFiles/opencv_core.dir/src/matrix_c.cpp.o CMakeFiles/opencv_core.dir/src/matrix_decomp.cpp.o CMakeFiles/opencv_core.dir/src/matrix_expressions.cpp.o CMakeFiles/opencv_core.dir/src/matrix_iterator.cpp.o CMakeFiles/opencv_core.dir/src/matrix_operations.cpp.o CMakeFiles/opencv_core.dir/src/matrix_sparse.cpp.o CMakeFiles/opencv_core.dir/src/matrix_wrap.cpp.o CMakeFiles/opencv_core.dir/src/mean.dispatch.cpp.o CMakeFiles/opencv_core.dir/src/merge.dispatch.cpp.o CMakeFiles/opencv_core.dir/src/minmax.cpp.o CMakeFiles/opencv_core.dir/src/norm.cpp.o CMakeFiles/opencv_core.dir/src/ocl.cpp.o CMakeFiles/opencv_core.dir/src/opencl/runtime/opencl_clamdblas.cpp.o CMakeFiles/opencv_core.dir/src/opencl/runtime/opencl_clamdfft.cpp.o CMakeFiles/opencv_core.dir/src/opencl/runtime/opencl_core.cpp.o CMakeFiles/opencv_core.dir/src/opengl.cpp.o CMakeFiles/opencv_core.dir/src/out.cpp.o CMakeFiles/opencv_core.dir/src/ovx.cpp.o CMakeFiles/opencv_core.dir/src/parallel.cpp.o CMakeFiles/opencv_core.dir/src/parallel_impl.cpp.o CMakeFiles/opencv_core.dir/src/pca.cpp.o CMakeFiles/opencv_core.dir/src/persistence.cpp.o CMakeFiles/opencv_core.dir/src/persistence_base64.cpp.o CMakeFiles/opencv_core.dir/src/persistence_c.cpp.o CMakeFiles/opencv_core.dir/src/persistence_cpp.cpp.o CMakeFiles/opencv_core.dir/src/persistence_json.cpp.o CMakeFiles/opencv_core.dir/src/persistence_types.cpp.o CMakeFiles/opencv_core.dir/src/persistence_xml.cpp.o CMakeFiles/opencv_core.dir/src/persistence_yml.cpp.o CMakeFiles/opencv_core.dir/src/rand.cpp.o CMakeFiles/opencv_core.dir/src/softfloat.cpp.o CMakeFiles/opencv_core.dir/src/split.dispatch.cpp.o CMakeFiles/opencv_core.dir/src/stat.dispatch.cpp.o CMakeFiles/opencv_core.dir/src/stat_c.cpp.o CMakeFiles/opencv_core.dir/src/stl.cpp.o CMakeFiles/opencv_core.dir/src/sum.dispatch.cpp.o CMakeFiles/opencv_core.dir/src/system.cpp.o CMakeFiles/opencv_core.dir/src/tables.cpp.o CMakeFiles/opencv_core.dir/src/trace.cpp.o CMakeFiles/opencv_core.dir/src/types.cpp.o CMakeFiles/opencv_core.dir/src/umatrix.cpp.o CMakeFiles/opencv_core.dir/src/utils/datafile.cpp.o CMakeFiles/opencv_core.dir/src/utils/filesystem.cpp.o CMakeFiles/opencv_core.dir/src/utils/samples.cpp.o CMakeFiles/opencv_core.dir/src/va_intel.cpp.o CMakeFiles/opencv_core.dir/opencl_kernels_core.cpp.o CMakeFiles/opencv_core.dir/arithm.sse4_1.cpp.o CMakeFiles/opencv_core.dir/matmul.sse4_1.cpp.o CMakeFiles/opencv_core.dir/stat.sse4_2.cpp.o CMakeFiles/opencv_core.dir/mathfuncs_core.avx.cpp.o CMakeFiles/opencv_core.dir/mathfuncs_core.avx2.cpp.o CMakeFiles/opencv_core.dir/stat.avx2.cpp.o CMakeFiles/opencv_core.dir/arithm.avx2.cpp.o CMakeFiles/opencv_core.dir/convert.avx2.cpp.o CMakeFiles/opencv_core.dir/convert_scale.avx2.cpp.o CMakeFiles/opencv_core.dir/count_non_zero.avx2.cpp.o CMakeFiles/opencv_core.dir/matmul.avx2.cpp.o CMakeFiles/opencv_core.dir/mean.avx2.cpp.o CMakeFiles/opencv_core.dir/merge.avx2.cpp.o CMakeFiles/opencv_core.dir/split.avx2.cpp.o CMakeFiles/opencv_core.dir/sum.avx2.cpp.o CMakeFiles/opencv_core.dir/matmul.avx512_skx.cpp.o   -L/home/conan/.conan/data/libpng/1.6.37/_/_/package/46c454909ce04fcfe4d444e9e216b9d9fbf98ebb/lib  -L/home/conan/.conan/data/jasper/2.0.25/_/_/package/53a15b73ce69c26f36583fa3856dd04b0057fbec/lib  -L/home/conan/.conan/data/openexr/2.5.5/_/_/package/de41549e84256614563dba7b41e7821e797babf8/lib  -L/home/conan/.conan/data/libtiff/4.2.0/_/_/package/6cbd95515c1d554969c6398e82b0bbc6d2e57799/lib  -L/home/conan/.conan/data/tbb/2020.3/_/_/package/bb6165ab21ac24a96075558fbc4502b91c7447fe/lib  -L/home/conan/.conan/data/zlib/1.2.11/_/_/package/b1b85bfdc2206d81b3ed6a49e78498182bb98ef3/lib  -L/home/conan/.conan/data/libjpeg/9d/_/_/package/b1b85bfdc2206d81b3ed6a49e78498182bb98ef3/lib  -L/home/conan/.conan/data/libdeflate/1.8/_/_/package/b1b85bfdc2206d81b3ed6a49e78498182bb98ef3/lib  -L/home/conan/.conan/data/xz_utils/5.2.5/_/_/package/b1b85bfdc2206d81b3ed6a49e78498182bb98ef3/lib  -L/home/conan/.conan/data/jbig/20160605/_/_/package/d9db921615487161ad127ab3a83447f7a7b94e6d/lib  -L/home/conan/.conan/data/zstd/1.5.0/_/_/package/b1b85bfdc2206d81b3ed6a49e78498182bb98ef3/lib  -L/home/conan/.conan/data/libwebp/1.1.0/_/_/package/44ff56a61aee602b7d6ec9df4c3f7de18222f285/lib  -Wl,-rpath,/home/conan/.conan/data/libpng/1.6.37/_/_/package/46c454909ce04fcfe4d444e9e216b9d9fbf98ebb/lib:/home/conan/.conan/data/jasper/2.0.25/_/_/package/53a15b73ce69c26f36583fa3856dd04b0057fbec/lib:/home/conan/.conan/data/openexr/2.5.5/_/_/package/de41549e84256614563dba7b41e7821e797babf8/lib:/home/conan/.conan/data/libtiff/4.2.0/_/_/package/6cbd95515c1d554969c6398e82b0bbc6d2e57799/lib:/home/conan/.conan/data/tbb/2020.3/_/_/package/bb6165ab21ac24a96075558fbc4502b91c7447fe/lib:/home/conan/.conan/data/zlib/1.2.11/_/_/package/b1b85bfdc2206d81b3ed6a49e78498182bb98ef3/lib:/home/conan/.conan/data/libjpeg/9d/_/_/package/b1b85bfdc2206d81b3ed6a49e78498182bb98ef3/lib:/home/conan/.conan/data/libdeflate/1.8/_/_/package/b1b85bfdc2206d81b3ed6a49e78498182bb98ef3/lib:/home/conan/.conan/data/xz_utils/5.2.5/_/_/package/b1b85bfdc2206d81b3ed6a49e78498182bb98ef3/lib:/home/conan/.conan/data/jbig/20160605/_/_/package/d9db921615487161ad127ab3a83447f7a7b94e6d/lib:/home/conan/.conan/data/zstd/1.5.0/_/_/package/b1b85bfdc2206d81b3ed6a49e78498182bb98ef3/lib:/home/conan/.conan/data/libwebp/1.1.0/_/_/package/44ff56a61aee602b7d6ec9df4c3f7de18222f285/lib::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::: -ldl -lm -lpthread -lrt /home/conan/.conan/data/tbb/2020.3/_/_/package/bb6165ab21ac24a96075558fbc4502b91c7447fe/lib/libtbb_debug.so /home/conan/.conan/data/zlib/1.2.11/_/_/package/b1b85bfdc2206d81b3ed6a49e78498182bb98ef3/lib/libz.a -lm -L
g++: error: missing argument to ‘-L’
source_subfolder/modules/core/CMakeFiles/opencv_core.dir/build.make:1631: recipe for target 'lib/libopencv_core.so.3.4.12' failed
make[2]: *** [lib/libopencv_core.so.3.4.12] Error 1

</details>

created time in 8 days

issue commentconan-io/conan-package-tools

Custom docker container name

What "feature"? I've just (hopefully) fixed it locally as follows:

--- runner.py.orig	2021-09-15 11:39:01.997868300 +0300
+++ runner.py	2021-09-15 11:39:02.086886100 +0300
@@ -1,3 +1,4 @@
+import uuid
 import os
 import sys
 import subprocess
@@ -271,9 +272,11 @@
                 # Update the downloaded image
                 with self.printer.foldable_output("update conan"):
                     try:
-                        command = '%s docker run %s --name conan_runner ' \
+                        conan_runner = "conan_runner_" + str(uuid.uuid1())
+                        command = '%s docker run %s --name %s ' \
                                   ' %s %s %s "%s"' % (self._sudo_docker_command,
                                                    env_vars_text,
+                                                   conan_runner,
                                                    self._docker_run_options,
                                                    self._docker_image,
                                                    self._docker_shell,
@@ -284,13 +287,14 @@
                             raise Exception("Error updating the image: %s" % command)
                         # Save the image with the updated installed
                         # packages and remove the intermediate container
-                        command = "%s docker commit conan_runner %s" % (self._sudo_docker_command,
+                        command = "%s docker commit %s %s" % (self._sudo_docker_command,
+                                                                        conan_runner,
                                                                         self._docker_image)
                         ret = self._runner(command)
                         if ret != 0:
                             raise Exception("Error commiting the image: %s" % command)
                     finally:
-                        command = "%s docker rm conan_runner" % self._sudo_docker_command
+                        command = "%s docker rm %s" % (self._sudo_docker_command, conan_runner)
                         ret = self._runner(command)
                         if ret != 0:
                             raise Exception("Error removing the temp container: %s" % command)
nicopernas

comment created time in 11 days

issue commentconan-io/conan-package-tools

Custom docker container name

I'm second to this. I'm trying to run two concurrent build jobs in Docker containers on the same Docker host. Each job, in turn, uses ConanMultiPackager(use_docker=True) to create a sibling Docker container for the actual work. The second one fails with

>>  docker run -e CONAN_ARCHS="x86" -e CONAN_CHANNEL="_" -e CONAN_DOCKER_ENTRY_SCRIPT="conan config set general.parallel_download=8 && conan user" -e CONAN_DOCKER_IMAGE="git:4567/docker/images/vs:2017" -e CONAN_DOCKER_RUN_OPTIONS="-e CI_JOB_TOKEN=[MASKED]" -e CONAN_LOCAL_REPO="http://artifactory/artifactory/api/conan/conan-local" -e CONAN_LOGIN_USERNAME="xxxxxxxx" -e CONAN_PASSWORD="xxxxxxxx" -e CONAN_REVISIONS_ENABLED="1" -e CONAN_STABLE_BRANCH_PATTERN="stable/*" -e CONAN_SYSREQUIRES_MODE="enabled" -e CONAN_UPLOAD="http://artifactory/artifactory/api/conan/conan-local@True@conan-local" -e CONAN_UPLOAD_DEPENDENCIES="all" -e CONAN_USERNAME="_" -e CONAN_VISUAL_RUNTIMES="MD,MDd" -e CONAN_VISUAL_VERSIONS="15" -e CONAN_REFERENCE="opencv/4.5.3@_/_" -e CPT_PROFILE="@@include(default)@@@@[settings]@@arch=x86@@build_type=Release@@compiler=Visual Studio@@compiler.runtime=MD@@compiler.version=15@@os=Windows@@[options]@@opencv:shared=True@@opencv:with_ipp=True@@[env]@@@@[build_requires]@@@@" -e CONAN_TEMP_TEST_FOLDER="1" -e CPT_UPLOAD_ENABLED="True" -e CPT_UPLOAD_RETRY="3" -e CPT_UPLOAD_ONLY_RECIPE="True" -e CPT_UPLOAD_FORCE="True" -e CPT_BUILD_POLICY="opencv,outdated" -e CPT_UPLOAD_DEPENDENCIES="all" -e CPT_CONANFILE="recipes\opencv\4.x\conanfile.py" --name conan_runner  -e CI_JOB_TOKEN=[MASKED] git:4567/docker/images/vs:2017 cmd /C " pip install conan_package_tools==0.35.1 --upgrade --no-cache &&  pip install conan==1.40.1 --no-cache"
____________________________________________________________________________________________________
docker: Error response from daemon: Conflict. The container name "/conan_runner" is already in use by container "b87c80b55924efa5a9dd506cfca2f8e7232544d8c8bc6ac88738fd8f6f820c50". You have to remove (or rename) that container to be able to reuse that name.
See 'docker run --help'.

Maybe just add a random suffix to conan_runner?

nicopernas

comment created time in 12 days

issue openedconan-io/conan

[question] Allow omitting channel from a package name?

Since using channels to promote packages is discouraged in favour of repo to repo move I would like to get rid of channels for my private packages. But I would like to retain user field to differ them from everything else. Something like package/1.0.0@user. Looks like it's not possible now - one either has to use package/1.0.0 or package/1.0.0@user/channel. Am I right? If yes, maybe supporting package names with user but without channel makes sense?

created time in a month

issue closedconan-io/conan

[question] Add optional `reference` parameter for `conan info`?

I'm trying to use conan info to topologically sort Conan packages by their dependencies in my build scripts. conan info <path to conanfile.py> works well until it hits a recipe without version containing something like if tools.Version(self.version) < "1.9":. Maybe adding an optional reference parameter to conan info makes sense?

closed time in a month

db4

issue commentconan-io/conan

[question] Add optional `reference` parameter for `conan info`?

Except python_requires that are needed to evaluate recipes, conan export do not analyze or compute the dependency graph, so it can be done in any random order, for all recipes in your repo.

I see, thanks a lot. Since additional parameters are already planned for Conan 2.0, I'm closing the issue.

db4

comment created time in a month

issue commentconan-io/conan

[question] Add optional `reference` parameter for `conan info`?

I have a private git repo with a bunch of internal recipes (with a folder structure very similar to CCI). On each push to the repo CI builds packages for the changed recipes. Since they depend on each other, the order in which they are built is important . Also, I only want to build the packages whose recipes have been changed (and several recipes may be changed in one commit). So CI first runs conan info for every recipe, combines all dependencies into one graph, topologically sorts it, checks if the necessary binaries are available, and uses Conan Package Tools to build the missing ones.

Have you considered doing a single conan info over a conanfile.txt that just declares the set of dependencies (first level, the transitive ones will be automatically added. It might require some (I think small) amount of manual updates to that conanfile.txt, but it might be worth, as it can simplify a lot the whole process.

But then I need to conan export all recipes first during CI, right? And if done in the wrong order, some conan export commands will fail.

A different issue is if you want to do that for different versions of the same package, and your dependency graph is not really a dependency graph (couldn't be used in a single consumer), then the problem in general is more challenging. We propose the "products" approach with lockfiles, and the lockfile bundles can provide topologically sorted build orders for combinations of products in which different versions of the same package are involved. Maybe something worth checking if you still havent. Still need the concept of "products".

Do you mean https://docs.conan.io/en/latest/versioning/lockfiles/bundle.html?

Again, lockfiles require all recipes to be exported first. Looks like what I actually need is a list of direct dependencies for a given recipe (dependencies names are enough, no full reference resolution is necessary). Can it be obtained somehow (via Conan API?) having only recipe's conanfile.py, reference, and the list of options/settings?

db4

comment created time in a month

issue commentconan-io/conan

[question] Add optional `reference` parameter for `conan info`?

What are you exactly trying to achieve? Why needing to call info on conan packages that you don't control (it seems those are ConanCenter packages, that typically do not define the version in the recipe, but on command line?)

I have a private git repo with a bunch of internal recipes (with a folder structure very similar to CCI). On each push to the repo CI builds packages for the changed recipes. Since they depend on each other, the order in which they are built is important . Also, I only want to build the packages whose recipes have been changed (and several recipes may be changed in one commit). So CI first runs conan info for every recipe, combines all dependencies into one graph, topologically sorts it, checks if the necessary binaries are available, and uses Conan Package Tools to build the missing ones.

This may be not very optimal. When conan info fails, I retry with several iterations of conan export/conan info to build the full dependency graph. Maybe there is a better way? Unfortunately, CCI Jenkins scripts are closed, so it's impossible to borrow some ideas from there...

db4

comment created time in a month

issue openedconan-io/conan

[question] Add optional `reference` paramener for `conan info`?

I'm trying to use conan info to topologically sort Conan packages by their dependencies in my build scripts. conan info <path to conanfile.py> works well until it hits a recipe without version containing something like if tools.Version(self.version) < "1.9":. Maybe adding an optional reference parameter to conan info makes sense?

created time in a month

Pull request review commentconan-io/conan-center-index

opencv: add Intel IPP support

 def requirements(self):             self.requires("eigen/3.3.9")         if self.options.parallel == "tbb":             self.requires("tbb/2020.3")+        if self.options.with_ipp and self.options.ipp_package:+            self.requires(str(self.options.ipp_package))

Ah, looks like IPP is now a part of Intel oneAPI and a commercial license is no longer needed. Well, then OpenCV recipe could use oneAPI Conan package (alas, not existing yet)

db4

comment created time in a month

PullRequestReviewEvent

Pull request review commentconan-io/conan-center-index

opencv: add Intel IPP support

 def requirements(self):             self.requires("eigen/3.3.9")         if self.options.parallel == "tbb":             self.requires("tbb/2020.3")+        if self.options.with_ipp and self.options.ipp_package:+            self.requires(str(self.options.ipp_package))

Well, I need to use a private Conan IPP package internally (IPP is licensed by my company from Intel so I cannot make it public). It's done with -o opencv:with_ipp=True -o opencv:ipp_package=ipp/2018.3@my_company/stable. But others can use the special IPP version for OpenCV leaving ipp_package option set to None

db4

comment created time in a month

PullRequestReviewEvent

PR opened conan-io/conan-center-index

opencv: add Intel IPP support

library name and version: opencv/4.5.3

This PR adds Intel IPP support for OpenCV via with_ipp and ipp_package options. If ipp_package is set to None (default value), the special version of IPP is used that OpenCV downloads during the build.

The recipe modifications can be suboptimal so any hints on how to do this better are very welcome.


  • [x] I've read the guidelines for contributing.
  • [x] I've followed the PEP8 style guides for Python code in the recipes.
  • [x] I've used the latest Conan client version.
  • [x] I've tried at least one configuration locally with the conan-center hook activated.
+108 -63

0 comment

3 changed files

pr created time in a month

push eventdb4/conan-center-index

Dmitry Bely

commit sha 656d11e1c83fe279cadf7e77c65216a5f26e6fe0

Add opencv/4.5.3

view details

Dmitry Bely

commit sha f3b1a7ca02924fb2e880ce75b9e17e140cbf0045

opencv: add Intel IPP support

view details

push time in a month

create barnchdb4/conan-center-index

branch : opencv-ipp

created branch time in a month

issue commentconan-io/wishlist

Intel MKL

The Intel Simplified Software License states: You may use and redistribute the software (the "Software"), without modification ... Is it a modification to repackage the software into a Conan package?

Of course not. Otherwise, one won't be able to redistribute Intel DLLs with his/her application (even without taking into account the fact that well credible anaconda.org has already repackaged them)

What happens when a new version is released and it is not available on anaconda.org?

We'll wait until it's available.

kylemacfarlan

comment created time in 2 months

issue commentconan-io/wishlist

Intel MKL

@bredej I mean creating the "official" Conan package based on anaconda.org MKL binaries. What's wrong with it and what's the alternative?

I would prefer an official Conan package supported by Intel and it's legal team.

Of course, I also do. But I doubt "Intel and its legal team" are aware of our wishes.

kylemacfarlan

comment created time in 2 months

issue commentconan-io/wishlist

Intel MKL

I've found a group of recipes https://github.com/shellshocked2003/mkl-include https://github.com/shellshocked2003/mkl-static https://github.com/shellshocked2003/mkl-shared that packages MKL using anaconda.org binaries. Looks like a way to go.

kylemacfarlan

comment created time in 2 months

issue commentconan-io/conan

[question] Internal Artifactory server silenly spies on me???

@memsharded I was too emotional, sorry for that. Please change the issue title to something more appropriate :)

db4

comment created time in 2 months

issue commentconan-io/conan

[question] Internal Artifactory server silenly spies on me???

Another possibility is that you set up a virtual and remote repositories in Artifactory, and didn't filter (exclude) your package queries.

Yes, I've set up a remote bincrafters repo on that server that I planned to use for faster downloads. But I actually do not use it. But I'm querying my internal packages from conancenter/bincrafters first during CI. Looks like these queries are not only cached but also made public? Maybe this leak is related to my conan client, not the server?

In that case, the package names (and only the packages names) might be requested to Bintray via the virtual repo. Do you have such a setup?

Not only names, but package ids as well.

db4

comment created time in 2 months

issue commentconan-io/conan

[question] Internal Artifactory server silenly spies on me???

More info: the tree contains index.json files (fortunately only them) with Source Path attribute starting from bintray-conan-cache.... But I never uploaded those recipes/packages to bintray.

db4

comment created time in 2 months

issue openedconan-io/conan

[bug] Internal Artifactory server silenly spies on me???

I have a JFrog Artifactory Community Edition for C/C++ instance inside my LAN. The server has no public IP and is used only internally. Now browsing https://bincrafters.jfrog.io/ui/repos/tree/General/public-conan I see a tree of my private recipes/packages. Why???

created time in 2 months