Revert "math/*: Improve and expand pkg-descr"

This reverts commit 5f90970e57.

math/R-cran-combinat/pkg-descr

@@ -1,17 +1 @@
-The R-cran-combinat package provides a collection of essential routines
-for combinatorial mathematics within the R environment. Combinatorics is
-a branch of mathematics concerning the study of finite or countable
-discrete structures.
-
-This package offers functions to generate and manipulate various combinatorial
-objects, including permutations, combinations, and partitions. It is
-invaluable for researchers, statisticians, and data scientists who need
-to perform tasks such as:
-
--   Generating all possible orderings of a set of items.
--   Selecting subsets of items without regard to their order.
--   Enumerating ways to divide a set into non-empty subsets.
-
-By providing these fundamental combinatorial tools, R-cran-combinat
-facilitates a wide range of applications in probability, statistics,
-computer science, and experimental design.
+Routines for combinatorics.

math/R-cran-conf.design/pkg-descr

@@ -1,24 +1,2 @@
-The R-cran-conf.design package provides a specialized set of tools
-within the R environment for the construction and manipulation of
-confounded and fractional factorial designs. These experimental designs
-are fundamental in statistics and engineering for efficiently studying
-the effects of multiple factors on an outcome, especially when resources
-are limited.
-
-Confounded designs allow for the study of a large number of factors
-with a smaller number of experimental runs by strategically sacrificing
-information about higher-order interactions. Fractional factorial designs
-are a type of confounded design that uses a fraction of the full factorial
-experiment, making them highly efficient for screening important factors.
-
-This library simplifies the process of setting up and analyzing such
-designs, making it invaluable for:
-
--   Experiment design in industrial and scientific research.
--   Quality improvement and process optimization.
--   Situations where a full factorial experiment is impractical due to
-    cost or time constraints.
-
-By offering these simple yet powerful tools, R-cran-conf.design enables
-researchers and practitioners to conduct more efficient and insightful
-experiments.
+This small library contains a series of simple tools for constructing and
+manipulating confounded and fractional factorial designs.

math/R-cran-cvar/pkg-descr

@@ -1,23 +1,7 @@
-The R-cran-cvar package provides essential tools for risk management,
-enabling the computation of Expected Shortfall (ES) and Value at Risk (VaR).
-ES, also known as Conditional Value at Risk (CVaR), and VaR are key metrics
-used to quantify potential financial losses in portfolios or investments.
-
-This package offers high flexibility, allowing users to compute these
-risk measures from various input types, including:
-
--   Quantile functions
--   Distribution functions
--   Random number generators
--   Probability density functions
-
-It supports virtually any continuous distribution, making it adaptable
-to diverse financial models. The functions are vectorized for efficient
-computation across multiple arguments. The calculations are performed
-directly from their definitions, as detailed by Acerbi and Tasche (2002).
-Additionally, the package includes some support for GARCH (Generalized
-Autoregressive Conditional Heteroskedasticity) models, further enhancing
-its utility for analyzing financial time series volatility.
-
-R-cran-cvar is an invaluable resource for financial analysts, risk managers,
-and quantitative researchers working with R to assess and manage financial risk.
+Compute expected shortfall (ES) and Value at Risk (VaR) from a quantile
+function, distribution function, random number generator or probability density
+function. ES is also known as Conditional Value at Risk (CVaR). Virtually any
+continuous distribution can be specified. The functions are vectorized over the
+arguments. The computations are done directly from the definitions, see e.g.
+Acerbi and Tasche (2002) <doi:10.1111/1468-0300.00091>. Some support for GARCH
+models is provided, as well.

math/R-cran-fracdiff/pkg-descr

@@ -1,20 +1,3 @@
-The R-cran-fracdiff package provides robust functionality for the
-maximum likelihood estimation of parameters in fractionally differenced
-ARIMA(p,d,q) models. These models are a powerful extension of traditional
-ARIMA models, designed to capture long-range dependence in time series data,
-where the 'd' parameter (differencing order) can be a non-integer value.
-
-Fractionally differenced ARIMA models are particularly useful for
-analyzing phenomena that exhibit persistent memory effects, such as:
-
--   Financial time series (e.g., stock prices, volatility)
--   Hydrological data (e.g., river flows, rainfall)
--   Environmental data (e.g., temperature anomalies)
--   Long-memory processes in various scientific and engineering fields
-
-Based on the methodology by Haslett and Raftery (Applied Statistics, 1989),
-this package offers a reliable and statistically sound approach to
-modeling time series with fractional integration. It enables researchers
-and practitioners in R to accurately estimate the parameters of these
-complex models, leading to more precise forecasts and a deeper understanding
-of long-memory processes.
+Maximum likelihood estimation of the parameters of a fractionally
+differenced ARIMA(p,d,q) model (Haslett and Raftery, Appl.Statistics,
+1989).

math/R-cran-gbutils/pkg-descr

@@ -1,17 +1,8 @@
-The R-cran-gbutils package offers general-purpose utilities for numerical
-and statistical computations in R, enhancing flexibility and ease of use.
-
-Key functionalities include:
-
--   **Distribution Analysis**: Plotting density/distribution functions,
-    numerically inverting distributions for quantiles, and simulating
-    real/complex numbers from magnitude/argument distributions.
--   **Polynomial Manipulation**: Creating polynomials from roots
-    (Cartesian or polar form).
--   **Programming Utilities**: Checking for NA identity, counting
-    positional arguments, computing set intersections for multiple sets,
-    identifying unnamed arguments, and graphing S4 classes.
-
-This invaluable toolkit streamlines common tasks in data analysis,
-statistical modeling, and numerical programming, boosting productivity
-and analytical capabilities for R users.
+Plot density and distribution functions with automatic selection of suitable
+regions. Numerically invert (compute quantiles) distribution functions.
+Simulate real and complex numbers from distributions of their magnitude and
+arguments. Optionally, the magnitudes and/or arguments may be fixed in almost
+arbitrary ways. Create polynomials from roots given in Cartesian or polar form.
+Small programming utilities: check if an object is identical to NA, count
+positional arguments in a call, set intersection of more than two sets, check
+if an argument is unnamed, compute the graph of S4 classes in packages.

math/R-cran-magic/pkg-descr

@@ -1,19 +1,11 @@
-The R-cran-magic package provides efficient, vectorized algorithms for
-creating and investigating magic squares and hypercubes. It includes
-functions for manipulating and analyzing multi-dimensional arrays.
-
-Key features:
-
--   **Magic Square Creation**: Methods for generating normal magic
-    squares of any order greater than 2.
--   **Analysis Tools**: Functions for the manipulation and analysis of
-    arbitrarily dimensioned arrays, including numerical verification
-    of magic square properties (e.g., determinant of odd-ordered
-    semimagic squares).
--   **Antimagic Functionality**: Support for antimagic squares and
-    related concepts.
-
-The package aims to be a comprehensive computerized embodiment of magic
-square knowledge, offering direct numerical verification of their
-properties. It is a valuable resource for mathematicians, statisticians,
-and R users interested in combinatorial designs and recreational mathematics.
+A collection of efficient, vectorized algorithms for the creation
+and investigation of magic squares and hypercubes, including a
+variety of functions for the manipulation and analysis of arbitrarily
+dimensioned arrays. The package includes methods for creating normal
+magic squares of any order greater than 2. The ultimate intention
+is for the package to be a computerized embodiment all magic square
+knowledge, including direct numerical verification of properties
+of magic squares (such as recent results on the determinant of
+odd-ordered semimagic squares). Some antimagic functionality is
+included. The package also serves as a rebuttal to the often-heard
+comment "I thought R was just for statistics".

math/R-cran-nortest/pkg-descr

@@ -1,17 +1 @@
-The R-cran-nortest package provides a suite of five omnibus tests
-for assessing the composite hypothesis of normality in statistical data.
-Normality tests are crucial in statistics to determine if a data set
-is well-modeled by a normal distribution, which is a common assumption
-for many parametric statistical methods.
-
-This package includes implementations of the following widely used tests:
--   Anderson-Darling test
--   Cramer-von Mises test
--   Shapiro-Francia test
--   Lilliefors test (Kolmogorov-Smirnov test with estimated parameters)
--   Pearson chi-square test
-
-These tests are valuable tools for statisticians, researchers, and data
-analysts working with R, enabling them to rigorously evaluate the
-distributional assumptions of their data before applying further
-statistical procedures.
+Five omnibus tests for testing the composite hypothesis of normality.

math/R-cran-quadprog/pkg-descr

@@ -1,18 +1,3 @@
-The R-cran-quadprog package provides an efficient and reliable implementation
-of the dual method by Goldfarb and Idnani (1982, 1983) for solving
-quadratic programming problems.
-
-Quadratic programming is a type of mathematical optimization problem that
-involves minimizing a quadratic objective function subject to linear
-constraints. This package is particularly useful for tasks such as
-portfolio optimization, support vector machines, and other statistical
-modeling applications where such optimization is required.
-
-Specifically, it solves problems of the form:
-minimize   -d'b + 1/2 b'Db
-subject to A'b >= b0
-
-where 'b' is the vector of variables to be optimized, 'd' is a vector,
-'D' is a symmetric positive-definite matrix, 'A' is a matrix, and 'b0'
-is a vector. The routine ensures accurate and robust solutions for
-these types of constrained optimization problems within the R environment.
+This routine implements the dual method of Goldfarb and Idnani
+(1982, 1983) for solving quadratic programming problems of the form
+min(?dT b + 1/2bT Db) with the constraints AT b >= b0.

math/R-cran-qualityTools/pkg-descr

@@ -1,21 +1,10 @@
-The R-cran-qualityTools package provides a comprehensive suite of
-statistical methods essential for Quality Science and Six Sigma
-Quality Management, particularly supporting the Define, Measure,
-Analyze, Improve, and Control (DMAIC) cycle.
+qualityTools: Statistical Methods for Quality Science
 
-Key functionalities include:
-
--   **Distribution Fitting**: Tools for fitting various statistical
-    distributions to data.
--   **Process Capability Analysis**: Calculation of normal and non-normal
-    process capability indices.
--   **Measurement Systems Analysis (MSA)**: Techniques such as gauge
-    capability indices and Gauge Repeatability and Reproducibility (GR&R)
-    studies.
--   **Experimental Design**: Support for factorial and fractional
-    factorial designs.
--   **Response Surface Methods**: Including the use of desirability functions.
-
-This package is an invaluable resource for quality engineers, statisticians,
-and practitioners implementing Six Sigma methodologies, enabling robust
-analysis and improvement of processes.
+Contains methods associated with the Define, Measure, Analyze, Improve and
+Control (i.e. DMAIC) cycle of the Six Sigma Quality Management
+methodology.It covers distribution fitting, normal and non-normal process
+capability indices, techniques for Measurement Systems Analysis especially
+gage capability indices and Gage Repeatability (i.e Gage RR) and
+Reproducibility studies, factorial and fractional factorial designs as
+well as response surface methods including the use of desirability
+functions.

math/algae/pkg-descr

@@ -1,19 +1,4 @@
-Algae is a specialized programming language meticulously designed for
-numerical analysis, particularly adept at tackling complex and large-scale
-computational problems. Developed by the Boeing Company, Algae was
-created to meet the demanding requirements of a fast, versatile, and
-robust tool for advanced engineering and scientific applications.
-
-Its core strengths lie in efficiently handling numerical computations
-involving large systems, making it suitable for:
-
--   Solving differential equations
--   Performing matrix operations
--   Implementing optimization algorithms
--   Simulating complex physical phenomena
-
-With a proven track record of over a decade in aerospace and related
-fields, Algae continues to be a valuable asset for researchers and
-engineers who require a powerful and reliable language for high-performance
-numerical analysis. Its design emphasizes both speed and the ability
-to manage extensive datasets and intricate models.
+Algae is a programming language for numerical analysis.  It was written in
+the Boeing Company to fulfill their need for a fast and versatile tool,
+capable of handling large systems.  Algae has been applied to interesting
+problems in aerospace and related fields for more than a decade.

math/apc/pkg-descr

@@ -1,20 +1,17 @@
-APC (Auto Payment Calculator) is a simple, Xforms-based graphical
-application designed for the X Window System. It provides a user-friendly
-interface for calculating auto loan payments.
+        the Auto Payment Calculator V1.0 Release
+	Copyright (C) 1997  Eric A. Griff
 
-Users can easily input the principal amount, loan term (in months),
-and interest rate. Upon calculation, it displays the monthly payment,
-as well as the number of weeks and the corresponding weekly payment.
+Auto Payment Calculator is a simple, xforms based, application for
+use under the X-windows system, that calculates auto loan payments.
 
-Key features include:
+It is pretty straight forward. You enter the Principal (Amount),
+Term (in months), and Rate, and then with either [RETURN]
+(or [enter] or whatever your keyboard equivelent is),  (ALT-C), or
+clicking the calculate button; you will have the payment in months,
+as well as number of weeks, and weekly payment.
 
--   **Intuitive Interface**: Built with Xforms for a straightforward
-    graphical user experience.
--   **Loan Calculation**: Quickly determines monthly and weekly payments
-    based on user-provided loan details.
--   **Interactive Input**: Supports keyboard navigation (e.g., Tab, Enter)
-    and mouse interaction for efficient data entry.
-
-APC is a practical utility for individuals needing to quickly estimate
-car loan payments, offering a clear and concise solution within the
-X Window environment.
+You may also [TAB] through the Amount, Term, and Rate, as well as
+hold down ALT and press the character in its Name that is underlined
+to go do that function. As long as all three are filled in, you may
+hit [ENTER] to Calculate right there. This makes it easy to cycle
+quickly through numerous terms, amounts, and rates.

math/aribas/pkg-descr

@@ -1,22 +1,6 @@
-ARIBAS is an interactive interpreter designed for advanced arithmetic,
-offering robust support for both big integer and multi-precision
-floating-point calculations. Its Pascal/Modula-like syntax provides
-a familiar and structured environment for users to perform complex
-mathematical operations.
-
-This powerful tool comes equipped with a rich set of built-in functions
-specifically tailored for algorithmic number theory, including:
-
--   **Number Theoretic Functions**: Greatest Common Divisor (GCD),
-    Jacobi symbol, and continued fraction expansions.
--   **Primality Testing**: Rabin probabilistic prime test for efficient
-    identification of prime numbers.
--   **Integer Factorization Algorithms**:
-    -   Quadratic sieve factorization for general integers.
-    -   Pollard's rho factorization for finding smaller prime factors.
-
-ARIBAS is an invaluable resource for mathematicians, computer scientists,
-and cryptographers who require precise and efficient tools for number
-theoretic research, cryptographic analysis, and other applications
-involving large numbers and complex arithmetic. Its interactive nature
-makes it ideal for experimentation and exploration of numerical properties.
+ARIBAS is an interactive interpreter for big integer arithmetic and
+multi-precision floating point arithmetic with a Pascal/Modula like
+syntax. It has several builtin functions for algorithmic number
+theory like gcd, Jacobi symbol, Rabin probabilistic prime test,
+continued fraction and quadratic sieve factorization, Pollard rho
+factorization, etc.

math/arpack++/pkg-descr

@@ -1,18 +1,4 @@
-ARPACK++ provides an object-oriented C++ interface to ARPACK (ARnoldi
-PACKage), a widely used Fortran library for solving large-scale
-eigenvalue problems. This wrapper allows C++ developers to leverage
-ARPACK's power within a modern programming paradigm.
-
-ARPACK is known for efficiently computing a few eigenvalues and
-eigenvectors of large, sparse matrices, making it vital in quantum
-mechanics, structural engineering, and data analysis. ARPACK++ retains
-the original Fortran package's strengths:
-
--   **Full Capability**: Access to all ARPACK functionalities for
-    various eigenvalue problems.
--   **High Performance**: Maintains computational speed and efficiency.
--   **Exceptional Accuracy**: Delivers precise numerical results.
--   **Low Memory Requirements**: Optimized for large matrices.
-
-By integrating ARPACK's robust numerical algorithms with C++ flexibility,
-ARPACK++ offers a powerful solution for complex eigenvalue computations.
+ARPACK++ is a collection of classes that offers c++ programmers an interface
+to ARPACK. It preserves the full capability, performance, accuracy and low
+memory requirements of the FORTRAN package, but takes advantage of the C++
+object-oriented programming environment.

math/atlas/pkg-descr

@@ -1,21 +1,18 @@
-ATLAS (Automatically Tuned Linear Algebra Software) is a high-performance
-software library for numerical linear algebra. It focuses on applying
-empirical optimization techniques to deliver portable and efficient
-performance across diverse hardware architectures.
+The ATLAS (Automatically Tuned Linear Algebra Software) project is an ongoing
+research effort focusing on applying empirical techniques in order to provide
+portable performance.  At present, it provides C and Fortran77 interfaces to
+a portable, efficient BLAS implementation, as well as enhanced versions of a
+few routines from LAPACK.  To link with ATLAS shared libraries:
 
-ATLAS provides optimized implementations of:
-
--   **BLAS (Basic Linear Algebra Subprograms)**: Offers C and Fortran77
-    interfaces for Level 1, 2, and 3 BLAS routines, crucial for vector,
-    matrix-vector, and matrix-matrix operations. Both serial (thread-safe)
-    and multi-threaded versions are available.
--   **LAPACK (Linear Algebra Package)**: Includes enhanced versions of
-    key LAPACK routines, providing efficient solutions for problems
-    like solving systems of linear equations, eigenvalue problems, and
-    singular value decomposition.
-
-The project's core strength lies in its ability to automatically tune
-itself to the specific characteristics of the underlying hardware during
-installation, ensuring optimal performance. ATLAS is an invaluable
-resource for scientific computing, engineering simulations, and any
-application requiring fast and reliable linear algebra computations.
+Serial (thread-safe) Fortran77 BLAS:
+	-lf77blas
+Multi-threaded Fortran77 BLAS:
+	-lptf77blas
+Serial (thread-safe) C BLAS:
+	-lcblas
+Multi-threaded C BLAS:
+	-lptcblas
+ATLAS-enhanced LAPACK, serial (thread-safe) interface:
+	-lalapack -lf77blas -lcblas
+ATLAS-enhanced LAPACK, multi-threaded interface:
+	-lalapack -lptf77blas -lptcblas

math/blacs/pkg-descr

@@ -1,23 +1,5 @@
-The BLACS (Basic Linear Algebra Communication Subprograms) library is a
-fundamental component for high-performance parallel computing, specifically
-designed to facilitate linear algebra operations on distributed memory
-platforms. It provides a standardized and efficient message passing
-interface tailored for numerical linear algebra algorithms.
-
-BLACS enables the communication and synchronization of data between
-processors in a parallel computing environment, which is crucial for
-implementing scalable versions of dense linear algebra routines. This
-makes it an essential building block for:
-
--   **Distributed Linear Algebra Libraries**: Such as ScaLAPACK, which
-    relies on BLACS for inter-processor communication.
--   **Scientific Simulations**: Large-scale computations in physics,
-    engineering, and other fields that require solving complex linear
-    systems or eigenvalue problems across multiple nodes.
--   **High-Performance Computing (HPC)**: Optimizing numerical workloads
-    on clusters and supercomputers.
-
-By offering a uniform and efficient communication layer, BLACS allows
-developers to write portable and high-performing parallel linear algebra
-code, ensuring that numerical applications can effectively utilize the
-power of distributed memory architectures.
+The BLACS (Basic Linear Algebra Communication Subprograms)
+project is an ongoing investigation whose purpose is to create
+a linear algebra oriented message passing interface
+that may be implemented efficiently and uniformly across
+a large range of distributed memory platforms.

math/blocksolve95/pkg-descr

@@ -1,15 +1,13 @@
-BlockSolve95 is a scalable parallel software library designed for the
-efficient solution of large, sparse linear systems. It is particularly
-optimized for problems arising from physical models, especially those
-with multiple degrees of freedom at each node (e.g., finite element
-methods in structural engineering).
+BlockSolve95 is a scalable parallel software library primarily intended for the
+solution of sparse linear systems that arise from physical models, especially
+problems involving multiple degrees of freedom at each node. For example, when
+the finite element method is used to solve practical problems in structural
+engineering, each node typically has two to five degrees of freedom;
+BlockSolve95 is designed to take advantage of problems with this type of local
+structure. BlockSolve95 is also reasonably efficient for problems that have
+only one degree of freedom associated with each node, such as the three-
+dimensional Poisson problem.
 
-The library effectively handles problems with this local structure,
-while also remaining reasonably efficient for systems with a single
-degree of freedom per node (e.g., three-dimensional Poisson problems).
-
-BlockSolve95 is a general-purpose solver, requiring only that matrices
-are sparse and symmetric in structure (though not necessarily in value).
-It provides a robust solution for complex scientific and engineering
-simulations that demand high-performance parallel computation for
-large sparse linear systems.
+BlockSolve95 is general purpose; we do not require that the matrices have any
+particular structure other than being sparse and being symmetric in structure
+(but not necessarily in value).

math/brial/pkg-descr

@@ -1,22 +1,11 @@
-BRiAl (Boolean Rings and Algebra) is a powerful C++ library for
-computations with polynomials over Boolean rings, serving as the
-successor to PolyBoRi. It provides high-level data types and efficient
-algorithms for symbolic computation in this specialized algebraic domain.
+BRiAl is the successor to PolyBoRi.
 
-Key features include:
-
--   **High-level Data Types**: For Boolean polynomials, monomials,
-    exponent vectors, and related algebraic structures.
--   **Binary Decision Diagrams (BDDs)**: Utilizes BDDs as the internal
-    storage type for polynomial structures, enabling efficient
-    representation and manipulation.
--   **Python Interface**: Offers a convenient Python binding, allowing
-    for parsing complex polynomial systems and implementing sophisticated
-    strategies for Grobner basis computation.
--   **Grobner Basis Computation**: Provides a robust and powerful
-    reference implementation for Grobner basis algorithms, essential
-    for solving systems of polynomial equations.
-
-BRiAl is an invaluable tool for researchers and developers in areas
-such as cryptography, coding theory, formal verification, and computer
-algebra, where efficient manipulation of Boolean polynomials is critical.
+The core of PolyBoRi is a C++ library, which provides high-level data
+types for Boolean polynomials and monomials, exponent vectors, as well
+as for the underlying polynomial rings and subsets of the powerset of
+the Boolean variables. As a unique approach, binary decision diagrams
+are used as internal storage type for polynomial structures. On top of
+this C++-library we provide a Python interface. This allows parsing of
+complex polynomial systems, as well as sophisticated and extendable
+strategies for Groebner base computation. PolyBoRi features a powerful
+reference implementation for Groebner basis computation.

math/clblas/pkg-descr

@@ -1,21 +1,11 @@
-clBLAS is a high-performance software library that provides optimized
-BLAS (Basic Linear Algebra Subprograms) functions implemented in OpenCL.
-BLAS routines are fundamental building blocks for numerical linear algebra,
-widely used in scientific computing, engineering, and data analysis.
+clBLAS
 
-The primary goal of clBLAS is to empower developers to leverage the
-performance and power efficiency of heterogeneous computing environments.
-It achieves this by:
+a software library containing BLAS functions written in OpenCL
 
--   **OpenCL Integration**: Directly utilizes OpenCL interfaces, allowing
-    users full control over OpenCL state management for maximum
-    performance and flexibility.
--   **Optimized Kernel Generation**: Automatically generates and enqueues
-    optimized OpenCL kernels, freeing users from the complex task of
-    writing, optimizing, and maintaining kernel code.
-
-clBLAS is an invaluable resource for developers and researchers who need
-to accelerate their linear algebra workloads by harnessing the parallel
-processing capabilities of GPUs and other OpenCL-compatible devices.
-It streamlines the development of high-performance computing applications
-by providing a robust and efficient foundation for numerical operations.
+The primary goal of clBLAS is to make it easier for developers to utilize the
+inherent performance and power efficiency benefits of heterogeneous computing.
+clBLAS interfaces do not hide nor wrap OpenCL interfaces, but rather leaves
+OpenCL state management to the control of the user to allow for maximum
+performance and flexibility. The clBLAS library does generate and enqueue
+optimized OpenCL kernels, relieving the user from the task of writing,
+optimizing and maintaining kernel code themselves.

math/clblast/pkg-descr

@@ -1,22 +1,2 @@
-CLBlast is a cutting-edge, lightweight, and highly performant OpenCL
-BLAS (Basic Linear Algebra Subprograms) library. It provides efficient
-and accelerated linear algebra computations on OpenCL-compatible devices.
-
-BLAS routines are fundamental building blocks for numerical algorithms
-in scientific computing, machine learning, and data analysis. CLBlast
-leverages OpenCL to offload these tasks to GPUs and other accelerators,
-significantly speeding up applications.
-
-Key features and benefits:
-
--   **Modern Design**: Built with contemporary OpenCL practices for
-    optimal performance.
--   **Lightweight Footprint**: Minimizes overhead for diverse systems.
--   **High Performance**: Achieves superior execution speeds through
-    careful optimization.
--   **Tunable**: Allows fine-grained control to extract maximum
-    performance from specific hardware (Intel, AMD, NVIDIA accelerators).
-
-CLBlast is an invaluable resource for developers and researchers seeking
-to accelerate numerical workloads by harnessing parallel processing
-capabilities of modern hardware through OpenCL.
+Modern, lightweight, performant and tunable OpenCL BLAS library. Tuned for
+Intel, AMD, and NVIDIA accelerators.

math/clfft/pkg-descr

@@ -1,24 +1,7 @@
-clFFT is a high-performance software library providing optimized Fast
-Fourier Transform (FFT) functions implemented in OpenCL. The FFT is a
-fundamental algorithm in digital signal processing and numerical analysis,
-used for tasks such as spectral analysis, image processing, and solving
-partial differential equations.
+clFFT
 
-Leveraging the OpenCL framework, clFFT enables efficient computation
-of FFTs on a wide range of parallel processing devices. Its key features
-include:
+a software library containing FFT functions written in OpenCL
 
--   **GPU Acceleration**: Primarily designed to harness the power of
-    Graphics Processing Units (GPUs) for significant speedups in FFT
-    computations.
--   **CPU Support**: Also supports execution on Central Processing Units
-    (CPUs), which is beneficial for debugging, development, and
-    heterogeneous computing environments where a mix of device types
-    is utilized.
--   **OpenCL Standard**: Adheres to the OpenCL standard, ensuring
-    portability across different hardware vendors and platforms.
-
-clFFT is an invaluable resource for developers and researchers who need
-to perform fast and efficient Fourier transforms on large datasets,
-particularly in applications that can benefit from the parallel
-processing capabilities of modern GPUs and multi-core CPUs.
+clFFT is a software library containing FFT functions written in OpenCL. In
+addition to GPU devices, the libraries also support running on CPU devices to
+facilitate debugging and heterogeneous programming.