GNU Task List ************* This file is updated automatically from `tasks.texi', which was last updated on 04 July 1997. About the GNU Task List *********************** If you did not obtain this file directly from the GNU project and recently, please check for a newer version. You can ftp the task list from any GNU FTP host in directory `/pub/gnu/tasks/'. The task list is available there in several different formats: `tasks.text', `tasks.texi', `tasks.info', and `tasks.dvi'. The GNU HURD task list is also there in file `tasks.hurd'. The task list is also available on the GNU World Wide Web server: `http://www.gnu.ai.mit.edu/prep/tasks_toc.html'. If you start working steadily on a project, please let `gnu@prep' know. We might have information that could help you; we'd also like to send you the GNU coding standards. Because of the natural tendency for most volunteers to write programming tools or programming languages, we have a comparative shortage of applications useful for non-programmer users. Therefore, we ask you to consider writing such a program. Typically, a new program that does a completely new job advances the GNU project, and the free software community, more than an improvement to an existing program. Typically, new features or new programs advance the free software community more, in the long run, than porting existing programs. One reason is that portable new features and programs benefit people on many platforms, not just one. At the same time, there tend to be many volunteers for porting--so your help will be more valuable in other areas, where volunteers are more scarce. Typically, it is more useful to extend a program in functionality than to improve performance. Users who use the new functionality will appreciate it very much, if they use it; but even when they benefit from a performance improvement, they may not consider it very important. Highest Priority **************** This task list mentions a large number of tasks that would be more or less useful. With luck, at least one of them will inspire you to start writing. It's better for you to work on any task that inspires you than not write free software at all. But if you would like to work on what we need most, here is a list of high priority projects. * If you are good at writing documentation, please do. * If you are very good at C programming and interested in kernels, you can help develop the GNU HURD, the kernel for the GNU system. Please have a look at `http://www.gnu.ai.mit.edu/software/hurd/hurd.html', and then get a copy of the latest HURD task list from: * `http://www.gnu.ai.mit.edu/prep/tasks.hurd.html', via the World Wide Web. * `ftp://prep.ai.mit.edu/pub/gnu/tasks/tasks.hurd', via anonymous FTP. * via e-mail. * If you are a Scheme fan, you can help develop Guile. Please have a look at the URL `http://www.gnu.ai.mit.edu:/software/guile/guile.html' and then contact . * Help develop the GNU web browser, E-scape. * Help develop Windows NT emulation for GNU systems--for example, WINE. * Implement the Kermit data transfer protocol. (See below.) * Develop a free compatible replacement for Qt, a graphics toolkit library. * Develop a substitute, which runs on GNU systems, for some very popular or very important application that many non-programmers use on Windows, and which has no comparable free equivalent now. Documentation ************* We very urgently need documentation for some parts of the system that already exist or will exist very soon: * A C reference manual. (RMS made a try at one, which you could start with). * A manual for Ghostscript. * A manual for TCSH. * A manual for PIC (the graphics formatting language). * A manual for Oleo. * A book on how GCC works and why various machine descriptions are written as they are. * A manual for programming X-window applications. * Manuals for various X window managers. * Reference cards for those manuals that don't have them: C Compiler, Make, Texinfo, Termcap, and maybe the C Library. * Many utilities need documentation, including `grep' and others. Unix-Related Projects ********************* * An improved version of the POSIX utility `pax'. There is one on Usenet, but it is said to be poorly written. Talk with `thomas@gnu.ai.mit.edu' about this project. * A `grap' preprocessor program for `troff'. * Various other libraries. * An emulation of SCCS that works using RCS. * Less urgent: `diction', `explain', and `style', or something to do the same kind of job. Compatibility with Unix is not especially important for these programs. Kernel-Related Projects *********************** * An over-the-ethernet debugger stub that will allow the kernel to be debugged from GDB running on another machine. This stub needs its own self-contained implementation of all protocols to be used, since the GNU system will use user processes to implement all but the lowest levels, and the stub won't be able to use those processes. If a simple self-contained implementation of IP and TCP is impractical, it might be necessary to design a new, simple protocol based directly on ethernet. It's not crucial to support high speed or communicating across gateways. It might be possible to use the Mach ethernet driver code, but it would need some changes. * A shared memory X11 server to run under MACH is very desirable. The machine specific parts should be kept well separated. * An implementation of CIFS, the "Common Internet File System," for the HURD. This protocol is an offshoot of SMB. Extensions to Existing GNU Software *********************************** * Enhance GCC. See files `PROJECTS' and `PROBLEMS' in the GCC distribution. * Interface GDB to Guile, so that users can write debugging commands in Scheme. This would also make it possible to write, in Scheme, a graphical interface that uses Tk and is tightly integrated into GDB. * Rewrite GNU `sed' completely, to make it cleaner. * Rewrite Automake and Deja-GNU in Scheme, so they can run in Guile. Right now they are written in Perl and TCL, respectively. There are also other programs, not terribly long, which we would also like to have rewritten in Scheme. * Finish the partially-implemented C interpreter project. * Help with the development of GNUStep, a GNU implementation of the OpenStep specification. * Add features to GNU Make to record the precise rule with which each file was last recompiled; then recompile any file if its rule in the makefile has changed. * Add a few features to GNU `diff', such as handling large input files without reading entire files into core. * An `nroff' macro package to simplify `texi2roff'. * A queueing system for the mailer Smail that groups pending work by destination rather than by original message. This makes it possible to schedule retries coherently for each destination. Talk to `tron@veritas.com' about this. Smail also needs a new chief maintainer. * Enhanced cross-reference browsing tools. (We now have something at about the level of `cxref'.) We also could use something like `ctrace'. (Some people are now working on this project.) X Windows Projects ****************** * An emulator for Macintosh graphics calls on top of X Windows. * A music playing and editing system. * High-quality music compression software. (Talk with phr@netcom.com for relevant suggestions.) * A program to play sound distributed in "Real Audio" format. * A program to generate "Real Audio" format from audio input. * A program to edit dance notation (such as labanotation) and display dancers moving on the screen. * Port the Vibrant toolkit to work on X without using Motif. * A program to display and edit Hypercard stacks. * A paint program, supporting both bitmap-oriented operations and component-oriented operations. `xpaint' exists, but isn't very usable. * A vector-based drawing program in the spirit of Adobe Illustrator and Corel Draw. * An interactive 3D modeling utility with rendering/raytracing capabilities. * A program for graphic morphing of scanned photographs. Other Projects ************** If you think of others that should be added, please send them to `gnu@prep.ai.mit.edu'. * A free program for public-key encryption. This program should use the Diffie-Helman algorithm for public key encryption, not the RSA algorithm, because the Diffie-Helman patent in the US will expire in 1997. It should use triple-DES, not IDEA, for block encryption, because IDEA is patented in many countries and the patents will not expire soon. In other respects, it should be like PGP. If the new version of PGP no longer uses RSA and IDEA, compatibility with it may be feasible. There is a suggestion that the Blowfish algorithm might be a good choice for the non-public-key block encryption. I'm not an encryption expert so I don't have an opinion about that. This program needs to be written by someone who is not a US citizen, outside the US, to avoid problems with US export control law. Many people believe that PGP is free software, but that is not actually true. The distribution terms do not allow everyone to use and redistribute it. We used to think that these restrictions had been imposed by the patent holders for the RSA and IDEA algorithms, and that PGP could be free software if these algorithms were replaced by alternatives. But we have since learned that the owners of PGP are themselves unwilling to make it free software. Writing a new program is the only way to get free software for this job. * A free secure telnet program more or less like ssh/sshd. Since this requires a public key encryption algorithm, it is closely related to the preceding item. This program should follow the draft standard for ssh. As always, it cannot implement the RSA algorithm, but must instead support the alternatives that will be patent-free in late 1997. It cannot support IDEA, but can use triple-DES and/or Blowfish or other non-patented alternatives. * Free software for doing secure commercial transactions on the web. This too needs public key encryption. * A free program that can transfer files on a serial line using the same protocol that Kermit uses. * An imitation of Page Maker or Ventura Publisher. * An imitation of `dbase2' or `dbase3' (How dbased!) * A bulletin board system. There are a few free ones, but they don't have all the features that people want in such systems. It would make sense to start with an existing one and add the other features. * A general ledger program, including support for accounts payable, account receivables, payroll, inventory control, order processing, etc. * A teleconferencing program which does the job of CU-SeeMe (which is, alas, not free software). * A program to typeset C code for printing. For ideas on what to do, see the book, Human Factors and Typography for More Readable Programs, Ronald M. Baecker and Aaron Marcus, Addison-Wesley, ISBN 0-201-10745-7 (I don't quite agree with a few of the details they propose.) * A program to convert Microsoft Word documents to text/enriched, TeX, LaTeX, Texinfo, or some other format that free software can edit. * A free replacement for the semi-free Qt library. * Speech-generation programs (there is a program from Brown U that you could improve). * Speech-recognition programs (single-speaker, disconnected speech is sufficient). * More scientific mathematical subroutines. (A clone of SPSS is being written already.) * Statistical tools. * A scientific data collection and processing tool, perhaps something like Scientific Workbench and/or Khoros, * Software to replace card catalogues in libraries. * Grammar and style checking programs. * An implementation of the S language (an interpreted language used for statistics). * A translator from Scheme to C. * A fast emulator for the i386 which works by translating machine instructions into the machine language of the host machine. (Support for emulation of other machines would enhance the program but might make it much more difficult.) * Optical character recognition programs; especially if suitable for scanning documents with multiple fonts and capturing font info as well as character codes. Work is being done on this, but more help is needed. * A program to scan a line drawing and convert it to Postscript. * A program to recognize handwriting. * A pen based interface. * CAD software, such as a vague imitation of Autocad. Compilers for Other Batch Languages *********************************** Volunteers are needed to write parsers/front ends for languages such as Algol 60, Algol 68, PL/I, Cobol, Fortran 90, Eiffel, or whatever, to be used with the code generation phases of the GNU C compiler. You can get the status of the Fortran front end with this command: finger -l fortran@gnu.ai.mit.edu Games and Recreations ********************* Video-oriented games that work with the X window system. * A Doom-compatible display game engine, for running the many free levels people have written for Doom. * Empire (there is a free version but it needs upgrading) * An "empire builder" system that makes it easy to write various kinds of simulation games. * Improve GnuGo, which is not yet very sophisticated. * Imitations of popular video games: - Space war, Asteroids, Pong, Columns. - Defending cities from missiles. - Plane shoots at lots of other planes, tanks, etc. - Wizard fights fanciful monster. - A golf game. - Program a robot by sticking building blocks together, then watch it explore a world. - Biomorph evolution (as in Scientific American and `The Blind Watchmaker'). - A program to display effects of moving at relativistic speeds. * Intriguing screen-saver programs to make interesting pictures. Other such programs that are simply entertaining to watch. For example, an aquarium. We do not need `rogue', as we have `hack'.