Next: Hn-Hz Up: Linux Software Encyclopedia Previous: Gn-Gz Contents

Ha-Hm

Last checked or modified: Mar. 9, 1999

Habanero: A project to recast single-user computer software tools as multi-user, collaborative work environments. Habanero is a framework for sharing Java objects with colleagues distributed around the Internet. Already implemented or planned are all the networking facilities, routing, arbitration and synchronization mechanisms needed to accomplish the sharing of state data and key events between collaborator's copies of a software tool.
Several sample applications built using Habanero are available including:

MetRealTime, which provides access to databases and plots of climatological data;

H263 Video Player, an all-Java streaming video player that supports the H263 file format;

Sound Player, an all-Java streaming audio player that supports the GSM file format;

Lattice, a crystallographic tool designed to skeletally view model molecules in 3-D;

Whiteboard, which can display both GIF and JPEG files and has a variety of tools and options;

HDF Viewer, which allows several people to simultaneously view the contents of an HDF file across a network;

Audio Chat, a web phone written in Java;

JavaGraph, a simple 2-D plotting tool based on Xgraph;

Habanero Text Editor, a collaborative editing environment capable of editing any ASCII file;

Chat, a chat program that uses the Habanero object sharing mechanism rather than IRC;

Voting Tool, which can be used to take votes during a collaborative session;

WWW Shared Session, a web browser controller which can be used to collaboratively surf the net;

Color Drawing, a whiteboard applet which can be used to create color drawings;

Visible Human, which permits browsing of the visible human database at the NIH;

and several others with more to come.
The latest version of the Sun Java Developer's Kit (JDK 1.0.2) must be installed for the Habanero tools to work. A version of JDK 1.0.2 is currently (10/3) available for Linux platforms.
[http://www.ncsa.uiuc.edu/SDG/Software/Habanero/]
HADLOP: The HArdware Description Language for Optical Processing is tool and hardware description language (HDL) for the design and simulation of digital 3-D optoelectronic computing systems. It is a simulation system for 3-D optical and optoelectronic circuits on the logical or gate level. Optoelectronic integrated circuits (OEICs) and optical permutation elements (OIMs) are specified by Boolean expressions and by a set of point-to-point interconnections, respectively.
HADLOP is implemented in C++. A key element is an object-oriented class modeling architecture primitives, e.g. masks, smart pixel processing planes, beam splitters and permutation modules. This architecture class provides methods for working with HADLOP in either design or simulation mode. In design mode a CAD editor is used to design an architecture consisting of architecture primitives, while in simulation mode the functional behavior of the architecture primitives is calculated.
The features of HADLOP not well supported in standard electronic design tools include:

parallel computing of 2-D binary matrices with processing layers consisting of identically assembled groups of gates optically connected in the 3rd dimension;

simple modeling of parallel regular optical interconnections like perfect shuffle, butterfly, crossover, etc.;

output of data layers as simulation results and not as an event trace of a collection of signals;

simultaneous simulation of pipelining of processing layers as well as field processing within data layers;

input stimuli consisting of subsequent 2-D data layers brought step by step into the 3-D system under investigation; and

the consideration of polarized light as an additional information channel for the light deflection as it is performed with polarized beam splitters.

Binary distributions of HADLOP are available for several platforms. Since these use the Motif library both statically and dynamically linked versions are available for Linux. Documentation includes a user's manual, a tutorial and some technical papers.
[http://www2.informatik.uni-jena.de/pope/HADLOP/hadlop.html]
[http://metalab.unc.edu/pub/Linux/apps/circuits/]
hal91: See distributions.
HAMR: The Hyperbolic Adaptive Mesh Refinement package is an application code for simulating the Euler equations of gas dynamics in an adaptive mesh refinement (AMR) framework. A high-order, high-resolution upwind shock-capturing Godunov integrator is used to solve the equations. AMR is a method for selectively refining regions of a computational domain (in both space and time) in a block-structured way. Increasingly finer grid patches are overlaid on those parts of a coarser grid where errors are deemed unacceptable. This lets the user avoid the onerous computational requirements of globally refining a grid to achieve a desired accuracy.
A source code distribution of HAMR is available. It is written in C++ and requires the BoxLib package for compilation and use. See the BoxLib entry for further installation information. A user's guide is included in the distribution. See Berger and Colella (1989), Glaz et al. (1987), and Woodward and Colella (1984).
[http://www.nersc.gov/research/CCSE/software/software.html]
Handy: An emulator for the Atari Lynx game system. Ports exist for Windows and Linux Intel.
[http://emu.simplenet.com/lynx/]
Harissa: A Java environment that includes a compiler from java bytecode to C and a Java interpreter. It is primarily aimed at statically configured applications such as the Javac compiler, but is also designed to allow code to be dynamically loaded into an already compiled application. This latter part is accomplished by integrating the bytecode interpreter in the runtime library. Data structures between the Java compiled code and the interpreter are compatible and data allocated by the interpreter do not conflict with data allocated by the compiled code.
The features of Harissa include:

a 64-bit compatible virtual machine;

provision of separate configurations with and without garbage collection;

the mixing of compiled binary code and interpreted Java bytecode;

all classes referenced by programs are compiled for each use of the compiler in order to perform whole program optimizations; and

separate compilation is supported in which libraries of classes can be compiled independently and be included in further compilations.

Harissa is currently (3/97) distributed only in binary form for Sun SunOS and Solaris and Linux platforms. To use it you will also have to obtain the Sun java classes from version 1.0.2 of the JDK.
[http://www.irisa.fr/compose/harissa/harissa.html]
HARPA: A 3-D Hamiltonian ray tracing program for acoustic waves in the atmosphere above irregular terrain. This is a large program written in Fortran 77 which is documented in Jones et al. (1986a), a 410 page technical report. Perhaps more later. See als HARPO.
[ftp://www1.etl.noaa.gov/pub/raytracing/harpa/]
HARPO: A 3-D Hamiltonian ray tracing program for acoustic waves in an ocean with an irregular bottom. This is a large program written in Fortran 77 which is documented in Jones et al. (1986b), a 455 page technical report. Perhaps more later. A NEW HARPO program which contains several error correction updates is available in the nharpo subdirectory. Other ancillary programs available in various subdirectories include: CONPLT, a contour plotting program for ray tracing models; EIGEN, a program to process output from ray tracing programs; PROFILE, a profile plotting program for ray tracing models; and PSGRAPH, a plotting program for HARPA, HARPO, EIGEN, PROFILE, and CONPLT.
[ftp://www1.etl.noaa.gov/pub/raytracing/harpo/]
HARPX: A program which extends HARPO or HARPA raypath calculations that have been made for two or three ray loops in a horizontally uniform medium to more loops and correspondingly longer ranges. A rayset file from either of those two programs is input and another rayset file is output. This program is written in Fortran 77 and documented in Jones and Georges (1988).
[ftp://www1.etl.noaa.gov/pub/raytracing/harpx/]
Harvest: An integrated set of tools to gather, extract, organize, search, cache and replicate relevant information across the Web. It can be tailored to digest information in many different formats and offer custom search services on the Web, and makes very efficient use of network traffic, remote servers, and disk space. Harvest consists of several subsystems including:

Gatherer, an efficient and customizable method for collecting indexing information;

Broker, which provides an indexed query interface to gathered information;

Index/Search, a general interface that can handle a variety of search engines;

Replicator, which provides a weakly consistent, replicated wide-area file system;

Object Cache, a hierarchical cache to meet heavy demands on network links and information servers; and

Harvest Object System, the implementation of an object-oriented information access protocol.

A source code distribution is available which (version 1.5) has been compiled on most popular UNIX platforms. See Bowman et al. (1995).
[http://harvest.transarc.com/]
[http://www.tardis.ed.ac.uk/harvest/]
[http://arco.met.fu-berlin.de/~kj/]
Harwell-Boeing Sparse Matrix Collection: A collection of sparse matrices from a wide range of application areas. Release 1 contains 291 matrices of order from 9 to 44609 and includes symmetric, unsymmetric, rectangular, and unassembled matrices. The matrices are available as is an 84 page user's manual in PostScript format.
[ftp://ftp.cerfacs.fr/pub/harwell_boeing/]
HASAS: The HydroAcoustic Signal Analysis System is a modular and scalable system for passive sonar signal analysis that can be used with hardware ranging from the simplest to the most complex available. The HASAS software architecture is divided into input, processing and user interface (UI) categories, with IPC used to provided communcation between categories. The input features include:

sample rate control;

amplifier and filter control;

data conversion;

serving multiple clients;

support for multicast; and

support for encryption.

The processing features of HASAS include:

beam forming with support for many array shapes;

multiple direction finding and locating algorithms for continuous and transient signals;

multiple spectral analysis algorithms, e.g. spectrogram, LOFAR and DEMON;

multiple transient analysis algorithms;

multiple detectors including neural networks;

variable filters and 3-D processing for audio output;

variable interpolation and decimation;

multiple array data combining;

data history;

support for multicast and data encyrption; and

support for special processor architectures.

The user interface features include:

optimized graphical and audio output using special hardware features;

2- and 3-D displays and audio output;

support for multiple monitors;

simultaneously connecting to multiple servers;

saving graphical output to a file or a printer;

saving audio output to a file; and

creating audio and data CDs.

[http://www.pp.clinet.fi/~visitor/]
hashjava: A Java library which alters symbols in a set of class files, i.e. an obfuscator. It includes two applications which obfuscate files and reduce their size. The first is an easy to use wizard which workds well for simple applets, and the second a more powerful application which uses a configuration file to offer precise control over what and how symbols are obfuscated.
[http://www.sbktech.org/hashjava_old.html]
[http://www.meurrens.org/ip-Links/Java/codeEngineering/blackDown/hashjava.html]
Haskell: A general purpose, purely functional programming language. Haskell programs are composed solely of functions, with computation proceeding by replacing functional expressions with their values. Haskell is based on lambda calculus and is named for the logician Haskell B. Curry whose work provided much of the basis for the language. See Bird (1998b), Davie (1992) and Thompson (1996).
Implementations of Haskell and other Haskell-related software include:

Glasgow Haskell Compiler

HBC

Hugs

nhc13

[http://haskell.org/]

Fudgets

A GUI toolkit for the Haskell language that also has the capability to create client/server applications that communicate via the Internet.
[http://www.cs.chalmers.se/ComputingScience/Research/Functional/Fudgets/]

Green Card

A foreign function interface preprocessor for Haskell that simplifies the task of interfacing Haskell programs to external libraries. This is usually applied to interface with C libraries. Source code and binary distributions are available, with one of the latter being for Linux Intel.
[http://www.dcs.gla.ac.uk/fp/software/green-card/]

Haggis

A GUI framework for the Haskell language running under X11. This is used with the Glasgow Haskell Compiler. The features of Haggis include:

a multi-threaded, concurrent user interface framework;

an extensible framework for building interfaces;

an almost complete set of common user interface abstractions;

a simple model of structured graphics;

an abstract view of the underlying window system; and

a modern functional I/O model.

A source code version is available.
[http://www.dcs.gla.ac.uk/fp/software/haggis/]
Hawkeye: A complete Internet/Intranet server suite which implements the most important Internet protocols for information interchange. Its functionalities make it an integrated web, mail, news, file, and chat server, a feat accomplished by letting all the services operate on a single SQL database. Creating a user in the Hawkeye database allows that user to access all the services. Hawkeye is also easy to install and configure, with configuration possible using a web browser.
The Hawkeye web server supports the HTTP/1.0 command set along with most important HTTP/1.1 features such as persistent connections and cache control commands. A CGI/1.1 gateway supports interactive web applications. The server also supports parsed HTML pages which allow the placement of user-dependent information inside web pages, e.g. you can create pages which allow users to read their mail or news. The SMTP/POP3 server allows users to exchange mail with the rest of the Internet. Every account has a unique email address from which users can send mail using parsed HTML pages or via other popular Internet mail software. Incoming mail is handled by POP3, a standard for storing mail until users download it. Discussion groups created using parsed HTML pages can be made available via the NNTP which is supported by most Internet news software packages. The entire RFC-977 command set is supported along with several modern extensions. The FTP server operates on the same directory structure as the Web server, allowing files to be downloaded either via FTP or HTTP. In addition to creating directories in which files can be stored, Hawkeye has an advanced, full-indexed file database where users can store information about files. The chat capabilities include an arbitrary number of different chat rooms along with restricting access to certain user groups. The Hawkeye server allows the system administrator to perform every detail of configuration with a web interface which is a GUI to the SQL database.
A source code distribution of Hawkeye is available as is a binary for Linux Intel platforms. It was developed on a Linux box so compilation shouldn't present much difficulty. A user's guide is available in HTML format. Hawkeye is freely available for private and educational use.
[http://hawkeye.net/]
HBC: The Chalmers Haskell-B Compiler is a full implementation of Haskell 1.3 along with some extensions including:

local existential and universal polymorphism via data types;

state monad with local state threads;

flexible calls to C within the IO monad; and

several HBC-specific libraries not found in the standard.

Source and binary distributions are available, with the latter including one for Linux Intel.
[http://www.cs.chalmers.se/~augustss/hbc.html]
HBF: The Han Binary Format is a specification developed for sharing Han character font information across incompatible bitmap file formats.
[http://umunhum.stanford.edu/~lee/chicomp/HBF_TR.html]
HCL: The Hilbert Class Library is a collection of C++ classes that apply object-oriented programming principles to implement mathematical objects such as vectors and linear operators. It allows the generic part of a program (e.g. optimization algorithms, linear solvers, etc.) to be written independently of the implementations of the data structures, simulators and other complex, application-specific details. A central design goal of HCL was to avoid legislating coordinate-based representations and other idiosyncrasies that usually make optimization codes unusable to those working in application areas. The obtains the ability to treat very large out-of-core data sets as vector objects.
The HCL consists of four categories of classes:

core classes that define the basic mathematical objects useful for specifying and solving optimization problems;

algorithm classes that implement algorithms for solving various optimization and linear algebra problems;

tool classes that allow the construction of certain standard constructs out of objects derived from the abstract base classes; and

standard concrete classes derived from the core classes, e.g. vector and matrix classes.

The algorithm classes include:

HCL_LinearSolver, base class for linear system solution methods;

HCL_PCG, conjugate gradient algorithm for SPD linear systems (with or without preconditioning);

HCL_ItEigSolvers, implicitly restarted Lanczos and Arnoldi methods (after ARPACK);

HCL_LineSearch, a base class for linear search that implements several such methods;

HCL_TrustRegionSolvers, base class for trust region solvers;

HCL_MatTR, implements the More-Hebden algorithm for the trust region problem;

HCL_UMin, base class for unconstrained minimization methods;

HCL_UMinNLCG, implements unconstrained minimization by nonlinear conjugate gradient algorithm; and

HCL_CMinAL, implements the augmented Lagrangian constrained minimization method.

The tool classes include:

HCL_Adjoint, creates adjoint and normal operators;

HCL_CompLinearOp, creates compositions of linear operators;

HCL_LinCombLinearOp, creates linear combinations of linear operators;

HCL_QuadFcnl, creates a quadratic form from a linear operator;

HCL_GenericProductVector, creates a product vector from factors;

HCL_BlockBiLinearOp, creates a bilinear operatoar on a product space in block form;

Table, a concrete associative array class;

HCL_lmbfgs, implements a limited memory BFGS Hessian approximation; and

HCL_Arch, defines various machine-dependent constants (stolen brazenly from LAPACK.

A source code distribution is available under the GPL as is documentation including a class manual, tutorials, and various technical papers.
[http://www.trip.caam.rice.edu/txt/hcldoc/html/index.html]
HCP: The Human Communications Protocol is a protocol designed to handle any form of communications a human might want to use over a computer network. It is designed to handle text, audio and video communications as well as communications that are both realtime or message based (e.g. IRC for the former and email for the latter). It can also handle conference-style communications with more than two peole communicating at the same time.
The HCP implementation is written in Java with the package consisting of a client and a server. The server implements the protocol and whatever else is needed to satisfy its requirements. The client implements code to get the transferred data, e.g. code that receives text from the keyboard, sound from the sound card, video from the camera, etc. This is done using modules with a defined interface that includes at least one input stream and one output stream. A source code distribution is available.
[http://205.241.209.107/HCP/about.html]
HDF: The Hierarchical Data Format is a multi-object file format that facilitates the transfer of various types of data between machines and operating systems. It allows self-definitions of data content and is easily extensible for future enhancements or compatibility with other standard formats. The latest version of HDF supports the complete NetCDF interface.
The HDF distribution contains several utility programs including:

fp2hdf, which converts 2- and 3-D floating point data sets into HDF SDS;

hdf24to8, which converts 24-bit raster images to HDF 8-bit images;

hdf2jpeg, which converts HDF raster images to JPEG images;

hdf8to24, which converts an 8-bit image into a 24-bit image;

hdfcomp, which re-compresses an 8-bit raster HDF file;

hdfed, an HDF file editor;

hdfls, which lists basic information about an HDF file;

hdfpack, which compacts an HDF file;

hdftopal, which extracts a palette from an HDF file;

hdftor8, which extracts 8-bit rsater images and palettes from an HDF file;

hdfunpac, which unpacks an HDF file by exporting the scientific data elements to external object elements;

hdp, which provides quick and general information about all objects in an HDF file;

jpeg2hdf, wh ich converts JPEG images to HDF raster images;

ncdump, which generates an ASCII representation of a NetCDF file for input into ncgen;

ncgen, which converts a CDL description of a NetCDF file into a NetCDF file or generates C or Fortran code required to create it;

paltohdf, which converts a raw palette to HDF;

r8tohdf, which converts 8-bit raster images to HDF;

ristosds, which converts a series of raster images HDF files into a single 3-D SDS HDF file;

vcompat, which converts HDF VSET V1.0 files to V2.0;

vmake, which creates VSETs; and

vshow, which dumps VSETs into an HDF file.

A source code distribution of the HDF package is available as are binaries for many machines including Linux Intel (a.out and ELF). An extensive user's guide is available in several formats. Various ancillary packages are available and are described below.
[http://hdf.ncsa.uiuc.edu/]

AMR-HDF

A set of routines to extend the SDS interface to store a grid hierarchy produced by Adaptive Mesh Refinement (AMR) codes in the HDF format. The entire grid hierarchy is flattened so it can be stored in a single file, and each grid in the hierarchy can be uniquely identified by its level, grid ID, and timestep so the grids can also be randomly accessed in the flattened file.
[http://zeus.ncsa.uiuc.edu/~jshalf/IO/AMRlibrary.html]

HDFNOW

A library that supports HDF for a network of workstations (NOW) running the MPICH version of the MPI message passing library. The features in this Distributed Scientific Datasets (DSD) interface include I/O speedup, a self-describing distributed file layout, asynchronous I/O access by compute processes, a chunking scheme for distributed scientific data sets, ease of data access, and additional HDF commands for distributed environments.
The DSD routines include: DSDstart, which opens the distributed HDF files; DSDnametoindex, which uses the name of a DSD to determine the index assigned to it; DSDselect, which opens an existing DSD for access; DSDcreate, which creates a DSD; DSDreaddata, which reads a hyperslab of data for a DSD; DSDwritedata, which writes a hyperslab of data for a DSD; DSDendaccess, which closes a DSD; and DSDend, which closes distributed HDF files. See also the PANDA project.
A source code version of the HDFNOW interface is available. The supported platforms are HP-UX, SGI IRIX, Linux Intel, and Sun SunOS. Additional software requirements are HDF release 4.1r1 and MPICH 1.1.0. The software is documented in some technical reports available in PostScript format.
[http://hdf.ncsa.uiuc.edu/Parallel_HDF/HDFNOW/hdfnow.html]

HDF-Java

A Java-based implementation of an HDF viewer. It can display the structure of an HDF file using different icons for the object types, display information about objects, display a subsampled or full resolution image for 2-D arrays, support animation for the first dimension of 3-D arrays, display a spreadsheet for SDS and GR objects, make a basic scaled or interpolated image for highlighted data sets, and much more. The source code, written in Java, is available as is the documentation for the package in HTML format.
[http://hdf.ncsa.uiuc.edu/java-hdf-html/]

h5utils

A collection of utilities for working with HDF 5 files including:

h5topng, extracts a 2-D slice from an HDF5 files and creates a corresponding PNG image;

h5totxt, extracts a 2-D slice and outputs comma-delimited text files;

h5tov5d, converts multidimensional datasets to Vis5D format;

h5read, a plug-in for Octave for importing 2-D slices from HDF5 files.

[http://ab-initio.mit.edu/mpb/hdf.html]

SDS

A set of wrapper routines for the HDF multifile (SD) interface that allow files to be referred to by name instead of file handle. In addition to simplifying the multifile interface, SDS also allows more file descriptors to be opened than are allowed by system-defined limits.
[http://zeus.ncsa.uiuc.edu/~jshalf/IO/SDSlibrary.html]
HDLC: The High-level Data Link Control is a group of protocols or rules for transmitting data between network nodes. It operates at the Data Link Layer of the OSI reference model which organizes data into units, sends the units to a destination that verifies their arrival, and also manages the rate at which data is sent. HDLC is based on an earlier IBM protocol called Synchronous Data Link Control (SDLC) which was widely used in mainframe environments, with what was essentially SDLC now known as Normal Response Mode in HDLC. In NMR a primary station (e.g. the mainframe) sends data to local and/or remote secondary stations in what is called a multidrop or multipoint network. Several variations of HDLC are or have been used including:

NRM, the Normal Response Mode used for multipoint networks with SDLC;

LAP, the Link Access Procedure used in early X.25 implementations;

LAPB, the Link Access Procedure, Balanced used in current X.25 implemtations;

LAPD, the Link Access Procedure for the ISDN D channel used for that and also for frame relay; and

LAPM, the Link Access Procedure for Modems used with error-correcting modems.
Hdrug: A graphical user environment for natural language processing in Prolog which can be used to develop logic grammars, parsers, and generators for natural languages. Hdrug comes with a selection of grammars including categorical, tree adjoining, unification, extraposition, and definite clause grammars. Each of the grammars comes with a set of parsers including Earley-like, left-corner, and head-driver parsers, with some grammars having variants of the head-driven generator. The package facilitates the easy comparison of different parsers and generators, and has extensive capabilities for compiling feature equations into Prolog terms. The interactive capabilities include parsing sentences, generating sentences from logical form representations, viewing parse trees, changing a parser version on the fly, comparing the results of parsing the same sentences with different parsers, and much more. The output capabilities include onscreen graphics, trees in LaTeX and Prolog, feature structures, Prolog terms, statistical information graphs, and tables of statistical information in LaTeX.
A source code distribution of Hdrug is available. It is written in Sicstus Prolog and Tcl/Tk and uses ProTcl as an interface between them. It is documented in a user's manual available in several formats as well as in some technical reports in PostScript format.
[http://odur.let.rug.nl/~vannoord/hdrug/]
Heart: A redundant, distributed cluster technology that provides a framework for a high availability implementation. Heart performs the duties of exchanging and updating the status of all nodes within a cluster so a crash at any node will be immediately noticed and appropriate action taken. Specifically, this exchanges heartbeats among all nodes and keeps a distributed database with this information, noticing when the heartbeat from one node stops and calling a script to cope with the problem. It was designed with Web and database servers in mind and IP-Takeover as the action taken, although it is sufficiently general to extend to other modes. Note: As of mid-2000 this project has been subsumed as part of the High-Availability Project and renamed Heartbeat.
[http://www.lemuria.org/Heart/]
heitml: A full-featured programming language which is completely backward compatible with HTML source code. Heitml makes it easy to tailor Web pages to specific features and Web browsers without having to develop and maintain separate versions of the pages. It makes it possible to design entire suites of Web pages using a single design template that draws text from databases or from text files.
Heitml is a server-based extension of HTML which generates HTML code. The features include: heitml-defined tags, procedures and functions which encourage a modular approach to HTML programming; the heitml functions are implemented as a monolithic language; database capabilities which are fully integrated into the language; a syntax similar to that of HTML; an alternative syntax that resembles Java and C; support for parameter passing and both global and local variables; session variables that hold state between Web pages; procedures that can be called recursively; conditional statements such as if and while; an include tag that allows source code to be assembled from various files; portability among a wide variety of database systems; and more.
The heitml package is available in binary form for Windows NT and 95, Linux Intel, FreeBSD, and Sun SunOS platforms. The database systems with which it will work include mSQL, PostgreSQL, and some commercial database systems. A source kit is also available which enables other database systems to be linked with heitml. The heitml binary is available as either a module for the Apache web server, a CGI executable, or as a static executable identical to the CGI version although called with different parameters. Documentation is available in HTML form on the heitml site.
[http://www.h-e-i.de/heitml/]
hendrix: A package for finding common patterns in data, e.g. if given 10 strings it can find a substring common to all in a process known as multiple sequence alignment in the molecular biology community. The purpose of this is to find patterns that seem to exist more frequently than chance occurrence, a process useful for what are now known as general data mining tasks. Hendrix can handle both the forward and inverse problems, with the former involving finding examples of a given model in data and the latter involving the determination of a model given a large amount of data. This would typically involve (in molecular biology) first aligning a stack of sequences to train the model and then using the model to search through an unknown sequience for regions similar to the model. A source code distribution of hendrix is available. It is documented in a technical report available in PostScript format.
[http://laslo.eecs.uic.edu/~bkao/hendrix/]
Hermes: A library mostly concerned with converting between all pixel formats, with most conversions done quickly via special routines. It is intended for use in graphics libraries or directly in graphics programs. Hermes is designed to do three things:

handle palettes;

convert surfaces from one pixel format to another including converting most common formats quickly with special converters, using generic converters for uncommon formats, stretching surfaces during conversion if possible, and dithering upon request; and

clearing surfaces to a specific color.

There is also support for multiple color spaces, alpha blending, and plugin support on systems that support dynamic libraries. All functions are implemented using an internal caching system. Handwritten assembler routines are available on X86 or MMX CPUs.
[http://www.clanlib.org/hermes/]
Hermes Group: The trials and tribulations of a high energy physics group called (you guessed it) Hermes in installing and using the Linux operating system for performing tasks in the aforementioned field of study. This is interesting and even moreso if you have access to the CERNLIB software suite.
[http://www-hermes.desy.de/ww/linux-hermes.html]
Hesiod: A name server which provides naming for services and data objects in a distributed network environment. Hesiod can serve as a replacement for databases which have heretofore had to be duplicated on each workstation and provides a flexible mechanism to supply new information as the need arises. It is based on the Berkeley Internet Domain Name Server called BIND. Hesiod provides a name service for use by workstations on a network. It doesn't address the problems of centralized management and distribution of such information. That is the domain of the Athena Service Management System (SMS) called Moira which maintains and distributes information to each of the Hesiod name servers.
The Hesiod package is available as source code. It can be compiled and installed on most UNIX systems using the supplied configure file. It is documented in a technical report in PostScript format as well as in a set of man pages. See also Kerberos, Moira, and Zephyr.
[ftp://athena-dist.mit.edu/pub/ATHENA/]
[http://sunsite.unc.edu/pub/Linux/system/network/]
HEVEA: A program that translates LaTeX into HTML. The input language is a fairly complete subset of LaTeX2e and the output HTML 3.2. Exotic symbols are translated into their analogues in the browser symbol fonts using the nonstandard FACE attribute of the FONT tag. HEVEA understands LaTeX macro definitions and simple user style files are mostly understood. It can be customized via writing LaTeX code. This is written in Objective Caml. Both source and binary distributions are available, with one of the latter being for Linux Intel. A user's manual is available in the expected formats.
[http://para.inria.fr/~maranget/hevea/]
hfs_fs: A loadable kernel module which implements the Macintosh HFS (Hierarchical File System) under Linux. The ultimate goal is to be able to read and write HFS disks under Linux as if they were native to it. Presently (8/97) the hfs_fs module can read and write HFS filesystems on 1.44M floppies, hard disks, CD-ROMS, Zip drives, and any other media supported by the hardware drivers in Linux. This means that 800K floppies cannot be used since they are only fully supported by system software. See also netatalk, hfsutils, CAP, and MacGate.
[http://www-sccm.stanford.edu/~hargrove/HFS/]
hfsutils: A package for manipulating HFS (Hierarchical File System) volumes from UNIX systems. HFS is the native volume format used on modern Macintosh computers. The hfsutils package includes: several command-line programs (e.g. hformat, hmount, hls, hcopy, etc.) of the same form and intended to be used in the same way as those in the mtools package; an X Window based front-end xhfs for browsing and copying files; a Tcl/Tk interface which offers a scriptable HFS shell which is more efficient than the command-line programs and allows for greater extensibility; and a C library for low-level access to disk volumes which can be linked with other programs to allow them to manipulate Mac files in their native format. Support is included for the manipulation of volumes of virtually any size and on any medium, and media with more than one HFS partition are supported.
A source code version of hfsutils is available. It is written in C and can be easily compiled and used. The Tcl/Tk interface requires Tcl 7.6 and Tk 4.2. The programs are documented in man pages.
[http://www.mars.org/home/rob/proj/hfs/]
hftpd: A lightweight FTP server that performs well on low-end hardware and scales well on high-end machines. It uses a heavily threaded model that takes full advantage of the Linux kernel.
[http://www.zabbo.net/hftpd/]
hiermap: A packages of utilities that take raw hierarchical data and create a PostScript map showing the hierarchical relationships. More than one page is printed if the map is large enough. Adjustable parameters include font size, line size, paper size, size of non-printable border for the printer, number of copies, orientation, and aspect ratio. Hiermap can be used to build a map of any domain name hierarchy on the Internet or any other kind of hierarchical information. The input file format is well-documented.
A source code distribution of hiermap is available. It is written mostly in Perl, requiring version 4.036 or newer. Usage also requires some sort of interactive communication with PostScript, which can be handled with Ghostscript or via TCP/IP with a networked printer using Expect. It is documented in an ASCII text file.
[http://www.domtools.com/dns/hiermap.shtml]
High-Availability Project: The goal of this project is to provide a high-availability clustering solution for Linux which promotes reliability, availability, and serviceability (RAS) through a community development effort.
[http://www.linux-ha.org/]
HINV: A hardware inventory script for Linux boxen that can identify CPUs, memory size, IDE devices, SCSI devices, keyboards, serial and parallel ports, Ethernet controllers, and IDE and SCSI controllers.
[http://www.brownnut.com/hinv.htm]
Histo-Scope Widget Set: A collection of six Motif widgets for graphing and plotting. These feature the high performance and screen stability needed for animation with interactive controls for direct manipulation via the mouse. Most of the widgets have some support for color, although not all have the ability to color arbitrary elements. They are all dependent upon Motif.
Also included in the package is NPlot, a tool for plotting columnar data from text files which is built using the Histo-Scope n-tuple interface. Both this and the individual widgets can be used to produce interactive graphs and plots including multi-variable graphs, 2- and 3-D scatter plots, 1- and 2-D histograms, and several types of specialty plots including adaptive histograms and cell plots. Interactive capabilities include scaling, zooming, and panning by dragging on axis scales and other sensitive areas. The 3-D widgets can be rotated using the mouse as a hand on a virtual track ball sphere surrounding the plot. The plots can also be combined with animation sliders to reflect additional variables or to re-bin histograms.
The source code for the Widget Set is available and is usable on platforms with an ANSI C compiler and the Motif library. The Widgets are documented in a user's manual included in the distribution in PostScript format.
[http://www-pat.fnal.gov/nirvana/histo.html]
[file://ftp.fnal.gov/pub/histoscope/]
Hive: A Java toolkit for building distributed systems. Hive is a distributed agents platform for building applications by networking local system resources. It provides ad-hoc agent interaction, ontologies of agent capabilities, mobile agents, and a graphical interface to the distributed system. The Hive architecture consists of three components:

decentralized cells analogous to web servers;

a set of local resources called shadows that encapsulate capabilities such as a screen display or a digital camera; and

agents that use local resources and communicate with each other.

In analogy to a conventional operating system, a cell is a kernel, shadows are device drivers, and agents are processes. Java source and binary versions are available under the GPL.
[http://www.hivecell.net/]
HMMER: The Hidden Markov Model ER software is a computational biology package which performs sensitive database searching and multiple sequence alignment using hidden Markov models. Hidden Markov models are statistical models of the primary structure consensus of a sequence family. They use position-specific scores for amino acids or nucleotides and for opening and extending an insertion or deletion as opposed to traditional pairwise-alignment methods which use position-independent scoring parameters. They have a formal probabilistic basis in Bayesian theory which allows them to do things which more heuristic methods can't, e.g. training from unaligned sequences.
The programs in the HMMER package include:

hmmt, which builts an HMM from initially unaligned training sequences;

hmmb, which builds an HMM from a multiple alignment;

hmma, which calculates a multiple sequence alignment using an existing HMM;

hmmls, which searches a sequence database for local matches to an HMM;

hmms, which searches a sequence database for global matches to an HMM;

hmmsw, which performs Smith/Waterman searching of a sequence database with an HMM, looking for local matches of part of the sequence against part of the HMM;

hmmfs, which searches for multiple fragments in long sequences and reports an optimal non-overlapping set of local matches between the HMM and the sequence;

hmme, which generates likely sequences stochastically from an HMM; and

hmm-convert, which converts HMM files between ASCII and binary formats.

The HMMER software is available both as source code and in binary format for SGI IRIX, Linux Intel, and Sun Solaris platforms. A user's manual is available in HTML format. See Eddy (1996).
[http://genome.wustl.edu/eddy/hmm.html]
HMML: The Hypertext Markup Macro Language is a macro language designed for preprocessing HTML. This can be used to improve the consistency of a set of HTML documents, generate live documents on demand, and makes writing HTML documents quicker and easier.
[http://www.cs.york.ac.uk/~james/hmml.html]

Next: Hn-Hz Up: Linux Software Encyclopedia Previous: Gn-Gz Contents

Manbreaker Crag 2001-03-08