diff --git a/.github/workflows/main.yml b/.github/workflows/main.yml
index 8a19254..a22e380 100644
--- a/.github/workflows/main.yml
+++ b/.github/workflows/main.yml
@@ -13,12 +13,13 @@ jobs:
   build-and-test:
     strategy:
       matrix:
-        os: [ubuntu-latest, macos-latest]
+        os: [ubuntu-latest, macos-latest, linux-gpu-cuda]
     runs-on: ${{ matrix.os }}
     steps:
     - uses: actions/checkout@v2
     - uses: conda-incubator/setup-miniconda@v2
-      with:        
+      with:
+        miniconda-version: "latest"
         auto-update-conda: true
     - name: Install 
       shell: bash -l {0}
@@ -32,12 +33,15 @@ jobs:
           sudo mv -f /Library/Developer/CommandLineTools /Library/Developer/CommandLineTools.org
         fi
 
+        df -h .
         if [[ "$(uname -s)" == "Linux" ]];
         then
-          conda create --yes -n unifrac -c conda-forge -c bioconda gxx_linux-64 hdf5 mkl-include lz4 zlib hdf5-static libcblas liblapacke make curl
+          conda create -q --yes -n unifrac -c conda-forge -c bioconda gxx_linux-64 hdf5 mkl-include lz4 zlib hdf5-static libcblas liblapacke make curl
         else
-          conda create --yes -n unifrac -c conda-forge -c bioconda clangxx_osx-64 hdf5 mkl-include lz4 hdf5-static libcblas liblapacke make curl
+          conda create -q --yes -n unifrac -c conda-forge -c bioconda clangxx_osx-64 hdf5 mkl-include lz4 hdf5-static libcblas liblapacke make curl
         fi 
+        conda clean --yes -t
+        df -h .
         conda activate unifrac
         echo "$(uname -s)"
         if [[ "$(uname -s)" == "Linux" ]];
@@ -52,19 +56,21 @@ jobs:
         which h5c++
         if [[ "$(uname -s)" == "Linux" ]];
         then
-          ./scripts/install_hpc_sdk.sh
+          ./scripts/install_hpc_sdk.sh >/dev/null 2>/dev/null </dev/null
           source setup_nv_h5.sh
         fi
+        df -h .
         export PERFORMING_CONDA_BUILD=True
         make api && \
         make main && \
         make install && \
         make test_binaries
+        df -h .
         pushd src
         if [[ "$(uname -s)" == "Linux" ]]; 
         then
           rm -f ~/.R/Makevars
-          conda install --yes -c conda-forge r-base
+          conda install -q --yes -c conda-forge r-base
           unset CXXFLAGS
           unset CFLAGS
           unset DEBUG_CXXFLAGS
diff --git a/src/Makefile b/src/Makefile
index 3d12c71..b597cf9 100644
--- a/src/Makefile
+++ b/src/Makefile
@@ -58,9 +58,9 @@ ifndef NOGPU
 	            ACCCPPFLAGS += -ta=tesla:ccall
 		else
 	            ACCCPPFLAGS += -ta=tesla
+		    # optional info
+		    ACCCPPFLAGS += -Minfo=accel
                 endif
-		# optional info
-		ACCCPPFLAGS += -Minfo=accel
 	        LDDFLAGS += -shlib -acc -Bstatic_pgi
 	        EXEFLAGS += -acc -Bstatic_pgi
 	endif
@@ -79,8 +79,10 @@ else
 endif
 
 ifeq (,$(findstring pgi,$(COMPILER)))
+    ifneq ($(PERFORMING_CONDA_BUILD),True)
 	# basically, not gcc
 	CPPFLAGS += -Wextra -Wno-unused-parameter
+    endif
 endif
 
 BLASLIB=-llapacke -lcblas
diff --git a/src/unifrac_internal.cpp b/src/unifrac_internal.cpp
index 1fff6e1..85551dc 100644
--- a/src/unifrac_internal.cpp
+++ b/src/unifrac_internal.cpp
@@ -29,10 +29,11 @@ static int sync_printf(const char *format, ...) {
     va_start(args, format);
 
     pthread_mutex_lock(&printf_mutex);
-    vprintf(format, args);
+    int cnt = vprintf(format, args);
     pthread_mutex_unlock(&printf_mutex);
 
     va_end(args);
+    return cnt;
 }
 
 static void sig_handler(int signo) {
diff --git a/src/unifrac_task.hpp b/src/unifrac_task.hpp
index 091320f..7900059 100644
--- a/src/unifrac_task.hpp
+++ b/src/unifrac_task.hpp
@@ -53,22 +53,23 @@ namespace SUCMP_NM {
 
     public:
       const unsigned int start_idx;
+      const unsigned int stop_idx;
       const unsigned int n_samples;
       const uint64_t  n_samples_r;
       TFloat* const buf;
 
       UnifracTaskVector(std::vector<double*> &_dm_stripes, const su::task_parameters* _task_p)
       : dm_stripes(_dm_stripes), task_p(_task_p)
-      , start_idx(task_p->start), n_samples(task_p->n_samples)
+      , start_idx(task_p->start), stop_idx(task_p->stop), n_samples(task_p->n_samples)
       , n_samples_r(((n_samples + UNIFRAC_BLOCK-1)/UNIFRAC_BLOCK)*UNIFRAC_BLOCK) // round up
-      , buf((dm_stripes[start_idx]==NULL) ? NULL : new TFloat[n_samples_r*(task_p->stop-start_idx)]) // dm_stripes could be null, in which case keep it null
+      , buf((dm_stripes[start_idx]==NULL) ? NULL : new TFloat[n_samples_r*(stop_idx-start_idx)]) // dm_stripes could be null, in which case keep it null
       {
         TFloat* const ibuf = buf;
         if (ibuf != NULL) {
 #ifdef _OPENACC
-          const uint64_t bufels = n_samples_r * (task_p->stop-start_idx);
+          const uint64_t bufels = n_samples_r * (stop_idx-start_idx);
 #endif
-          for(unsigned int stripe=start_idx; stripe < task_p->stop; stripe++) {
+          for(unsigned int stripe=start_idx; stripe < stop_idx; stripe++) {
              double * dm_stripe = dm_stripes[stripe];
              TFloat * buf_stripe = this->operator[](stripe);
              for(unsigned int j=0; j<n_samples; j++) {
@@ -86,6 +87,9 @@ namespace SUCMP_NM {
         }
       }
 
+      UnifracTaskVector<TFloat>(const UnifracTaskVector<TFloat>& ) = delete;
+      UnifracTaskVector<TFloat>& operator= (const UnifracTaskVector<TFloat>&) = delete;
+
       TFloat * operator[](unsigned int idx) { return buf+((idx-start_idx)*n_samples_r);}
       const TFloat * operator[](unsigned int idx) const { return buf+((idx-start_idx)*n_samples_r);}
 
@@ -95,10 +99,10 @@ namespace SUCMP_NM {
         TFloat* const ibuf = buf;
         if (ibuf != NULL) {
 #ifdef _OPENACC
-          const uint64_t bufels = n_samples_r * (task_p->stop-start_idx); 
+          const uint64_t bufels = n_samples_r * (stop_idx-start_idx); 
 #pragma acc exit data copyout(ibuf[:bufels])
 #endif    
-          for(unsigned int stripe=start_idx; stripe < task_p->stop; stripe++) {
+          for(unsigned int stripe=start_idx; stripe < stop_idx; stripe++) {
              double * dm_stripe = dm_stripes[stripe];
              TFloat * buf_stripe = this->operator[](stripe);
              for(unsigned int j=0; j<n_samples; j++) {
@@ -147,6 +151,9 @@ namespace SUCMP_NM {
         : dm_stripes(baseObj.dm_stripes), dm_stripes_total(baseObj.dm_stripes_total), task_p(baseObj.task_p) {}
         */
 
+        UnifracTaskBase<TFloat,TEmb>(const UnifracTaskBase<TFloat,TEmb>& ) = delete;
+        UnifracTaskBase<TFloat,TEmb>& operator= (const UnifracTaskBase<TFloat,TEmb>&) = delete;
+
         virtual ~UnifracTaskBase()
         {
 #ifdef _OPENACC
@@ -316,6 +323,8 @@ namespace SUCMP_NM {
         , embedded_proportions(_embedded_proportions), max_embs(_max_embs) {}
         */
       
+       UnifracTask<TFloat,TEmb>(const UnifracTask<TFloat,TEmb>& ) = delete;
+       UnifracTask<TFloat,TEmb>& operator= (const UnifracTask<TFloat,TEmb>&) = delete;
 
        virtual ~UnifracTask() {}
 
@@ -346,6 +355,9 @@ namespace SUCMP_NM {
 #pragma acc enter data create(zcheck[:n_samples],sums[:n_samples])
         }
 
+        UnifracUnnormalizedWeightedTask<TFloat>(const UnifracUnnormalizedWeightedTask<TFloat>& ) = delete;
+        UnifracUnnormalizedWeightedTask<TFloat>& operator= (const UnifracUnnormalizedWeightedTask<TFloat>&) = delete;
+
         virtual ~UnifracUnnormalizedWeightedTask()
         {
 #ifdef _OPENACC
@@ -381,6 +393,9 @@ namespace SUCMP_NM {
 #pragma acc enter data create(zcheck[:n_samples],sums[:n_samples])
         }
 
+        UnifracNormalizedWeightedTask<TFloat>(const UnifracNormalizedWeightedTask<TFloat>& ) = delete;
+        UnifracNormalizedWeightedTask<TFloat>& operator= (const UnifracNormalizedWeightedTask<TFloat>&) = delete;
+
         virtual ~UnifracNormalizedWeightedTask()
         {
 #ifdef _OPENACC
@@ -421,6 +436,9 @@ namespace SUCMP_NM {
 #pragma acc enter data create(zcheck[:n_samples],stripe_sums[:n_samples],sums[:bsize])
         }
 
+        UnifracUnweightedTask<TFloat>(const UnifracUnweightedTask<TFloat>& ) = delete;
+        UnifracUnweightedTask<TFloat>& operator= (const UnifracUnweightedTask<TFloat>&) = delete;
+
         virtual ~UnifracUnweightedTask()
         {
 #ifdef _OPENACC
@@ -450,6 +468,9 @@ namespace SUCMP_NM {
         UnifracGeneralizedTask(std::vector<double*> &_dm_stripes, std::vector<double*> &_dm_stripes_total, unsigned int _max_embs, const su::task_parameters* _task_p)
         : UnifracTask<TFloat,TFloat>(_dm_stripes,_dm_stripes_total,_max_embs,_task_p) {}
 
+        UnifracGeneralizedTask<TFloat>(const UnifracGeneralizedTask<TFloat>& ) = delete;
+        UnifracGeneralizedTask<TFloat>& operator= (const UnifracGeneralizedTask<TFloat>&) = delete;
+
         virtual void run(unsigned int filled_embs, const TFloat * __restrict__ length) {_run(filled_embs, length);}
 
         void _run(unsigned int filled_embs, const TFloat * __restrict__ length);
@@ -516,6 +537,8 @@ namespace SUCMP_NM {
         , embedded_proportions(_embedded_proportions), embedded_counts(initialize_embedded<TFloat>()), sample_total_counts(_sample_total_counts), max_embs(_max_embs) {}
         */
 
+       UnifracVawTask<TFloat,TEmb>(const UnifracVawTask<TFloat,TEmb>& ) = delete;
+       UnifracVawTask<TFloat,TEmb>& operator= (const UnifracVawTask<TFloat,TEmb>&) = delete;
 
        virtual ~UnifracVawTask() 
        {
@@ -554,6 +577,9 @@ namespace SUCMP_NM {
                     unsigned int _max_embs, const su::task_parameters* _task_p)
         : UnifracVawTask<TFloat,TFloat>(_dm_stripes,_dm_stripes_total,_sample_total_counts,_max_embs,_task_p) {}
 
+        UnifracVawUnnormalizedWeightedTask<TFloat>(const UnifracVawUnnormalizedWeightedTask<TFloat>& ) = delete;
+        UnifracVawUnnormalizedWeightedTask<TFloat>& operator= (const UnifracVawUnnormalizedWeightedTask<TFloat>&) = delete;
+
         virtual void run(unsigned int filled_embs, const TFloat * __restrict__ length) {_run(filled_embs, length);}
 
         void _run(unsigned int filled_embs, const TFloat * __restrict__ length);
@@ -566,6 +592,9 @@ namespace SUCMP_NM {
                     unsigned int _max_embs, const su::task_parameters* _task_p)
         : UnifracVawTask<TFloat,TFloat>(_dm_stripes,_dm_stripes_total,_sample_total_counts,_max_embs,_task_p) {}
 
+        UnifracVawNormalizedWeightedTask<TFloat>(const UnifracVawNormalizedWeightedTask<TFloat>& ) = delete;
+        UnifracVawNormalizedWeightedTask<TFloat>& operator= (const UnifracVawNormalizedWeightedTask<TFloat>&) = delete;
+
         virtual void run(unsigned int filled_embs, const TFloat * __restrict__ length) {_run(filled_embs, length);}
 
         void _run(unsigned int filled_embs, const TFloat * __restrict__ length);
@@ -578,6 +607,9 @@ namespace SUCMP_NM {
                     unsigned int _max_embs, const su::task_parameters* _task_p)
         : UnifracVawTask<TFloat,uint32_t>(_dm_stripes,_dm_stripes_total,_sample_total_counts,_max_embs,_task_p) {}
 
+        UnifracVawUnweightedTask<TFloat>(const UnifracVawUnweightedTask<TFloat>& ) = delete;
+        UnifracVawUnweightedTask<TFloat>& operator= (const UnifracVawUnweightedTask<TFloat>&) = delete;
+
         virtual void run(unsigned int filled_embs, const TFloat * __restrict__ length) {_run(filled_embs, length);}
 
         void _run(unsigned int filled_embs, const TFloat * __restrict__ length);
@@ -590,6 +622,9 @@ namespace SUCMP_NM {
                     unsigned int _max_embs, const su::task_parameters* _task_p)
         : UnifracVawTask<TFloat,TFloat>(_dm_stripes,_dm_stripes_total,_sample_total_counts,_max_embs,_task_p) {}
 
+        UnifracVawGeneralizedTask<TFloat>(const UnifracVawGeneralizedTask<TFloat>& ) = delete;
+        UnifracVawGeneralizedTask<TFloat>& operator= (const UnifracVawGeneralizedTask<TFloat>&) = delete;
+
         virtual void run(unsigned int filled_embs, const TFloat * __restrict__ length) {_run(filled_embs, length);}
 
         void _run(unsigned int filled_embs, const TFloat * __restrict__ length);