3162
|
1 <html> |
|
2 <pre> |
2330
|
3 Octave PROJECTS -*- text -*- |
|
4 =============== |
|
5 |
|
6 Check with bug-octave@bevo.che.wisc.edu for a possibly more current |
|
7 copy. Also, if you start working steadily on a project, please let |
|
8 bug-octave@bevo.che.wisc.edu know. We might have information that |
|
9 could help you; we'd also like to send you the GNU coding standards. |
|
10 |
|
11 This list is not exclusive -- there are many other things that might |
|
12 be good projects, but it might instead be something we already have, |
|
13 so check with bug-octave@bevo.che.wisc.edu before you start. |
|
14 |
|
15 --------- |
|
16 Numerical: |
|
17 --------- |
|
18 |
|
19 * Merge control stuff. |
|
20 |
|
21 * Improve logm, and sqrtm. |
|
22 |
|
23 * Improve complex mapper functions. |
|
24 |
|
25 * Make functions like gamma() return the right IEEE Inf or NaN |
|
26 values for extreme args or other undefined cases. |
|
27 |
|
28 * Handle complex values in fread and fwrite. |
|
29 |
|
30 * Support for lp_solve for linear programming problems. |
|
31 |
|
32 * Free QP solver. |
|
33 |
|
34 * Free NLP solver. |
|
35 |
|
36 * Support for sparse matrices. |
|
37 |
|
38 * Fix CollocWt to handle Laguerre polynomials. Make it easy to |
|
39 extend it to other polynomial types. |
|
40 |
3130
|
41 * Make filter faster (perhaps by calling BLAS functions). |
|
42 |
2330
|
43 * Add optional arguments to colloc so that it's not restricted to |
|
44 Legendre polynomials. |
|
45 |
|
46 * Fix eig to also be able to solve the generalized eigenvalue |
|
47 problem, and to solve for eigenvalues and eigenvectors without |
|
48 performing a balancing step first. |
|
49 |
|
50 * Move rand, eye, xpow, xdiv, etc., functions to the matrix classes. |
|
51 |
2477
|
52 * Use octave_allocator for memory management in Array classes once |
|
53 g++ supports static member templates. |
|
54 |
2330
|
55 * Implement the following functions: |
|
56 -- ppval -- cross -- dot |
|
57 |
2789
|
58 * When constructing NLConst (and other) objects, make sure that |
|
59 there are sufficient checks to ensure that the dimensions all |
|
60 conform. |
|
61 |
2439
|
62 * Allow parameters to be passed through the call to fsolve() to the |
2777
|
63 user-supplied function for Matlab compatibility. Don't place an |
2439
|
64 upper limit on the number of arguments. |
|
65 |
2330
|
66 * Check matrix classes for proper handling of empty matrices. |
|
67 |
2602
|
68 * Force all empty matrices to be 0x0 even when other dimensions have |
|
69 been speicified, for compatibility with Matlab, at least when some |
|
70 preference variable is set. |
|
71 |
2789
|
72 * Make operations with empty matrices produce empty matrices, for |
|
73 compatibility with Matlab (but only if Matlab 5 still does things |
|
74 this way). For example: [1, 2] * [] ==> []. |
|
75 |
2799
|
76 * Should any ([]) return [] or 0? What about all ([])? |
|
77 |
2330
|
78 * Improve design of ODE, DAE, classes. |
|
79 |
2439
|
80 * Make it possible to specify a time which dassl and lsode should |
|
81 not integrate past. |
|
82 |
2799
|
83 * Add interface to daspk (new version of dassl). |
|
84 |
2330
|
85 * Extend meaning of .* to include v .* M or M .* v (where v is a |
|
86 column vector with the same number of rows as M) to scale rows of |
|
87 M by elements of v. Similarly, if w is a row vector with as many |
|
88 columns as M, then either w .* M or M .* w scales the columns of |
|
89 M. |
|
90 |
2787
|
91 * Add support for +=, -=, etc. |
|
92 |
2497
|
93 * Given two vectors x and y of length m and n, implement a function |
|
94 outer (x, y, f) that returns an m-by-n matrix with entries |
|
95 f (x(i), y(j)). If f is omitted, multiplication is the default. |
2506
|
96 Should probably work for any vectors, not just if x is a column |
|
97 vector and y is a row vector. |
2497
|
98 |
2789
|
99 * Make it possible to solve b = L \ x efficiently, either by |
|
100 providing an explicit function call, or by automatically |
|
101 determining that L is triangular. If it is done automatically, |
|
102 provide some means for determining whether Octave has actually |
|
103 detected that the matrix is triangular. |
|
104 |
2799
|
105 * The polyfit function uses the economy QR factorization, but even |
|
106 that can take a lot of time for large datasets. Consider an |
|
107 option to compute the result with inv (A' * A) * A' * y or some |
|
108 other faster method.d Possibly just switch to this method if the |
|
109 dataset is larger than some value. |
|
110 |
3141
|
111 * Make QR more memory efficient for large matrices when not all the |
|
112 columns of Q are required (apparently this is not handled by the |
|
113 lapack code yet). |
|
114 |
2799
|
115 * Consider making the behavior of the / and \ operators for |
|
116 non-square systems compatible with Matlab. |
|
117 |
2330
|
118 -------- |
|
119 Graphics: |
|
120 -------- |
|
121 |
|
122 * Make plotting with plplot work. |
|
123 |
|
124 * Fix interface with gnuplot to wait for gnuplot to output any text |
|
125 before octave returns a prompt. Possible by implementing two |
|
126 way communication between gnuplot and Octave. |
|
127 |
3136
|
128 * Handle gnuplot ranges correctly for parametric modes (accept 3 |
|
129 ranges for 2d plots and 5 ranges for 3d plots). |
|
130 |
2330
|
131 * Make gsave (and possibly gload) work. Implement gsave by having it |
|
132 also alter the plot command to not use temporary files (perhaps |
|
133 with some user-specified template for naming them) and then |
|
134 sending a `save' command to gnuplot. |
|
135 |
|
136 * It would be nice to be able to check that a plot is currently |
|
137 being displayed. |
|
138 |
2422
|
139 * Implement clf, gcf, get, set, orient, print, close, etc. in |
|
140 Matlab-compatible ways. |
2330
|
141 |
|
142 * Make it possible to check the current graphics terminal type. |
|
143 |
2560
|
144 * If possible, pass data to gnuplot without using temporary files. |
|
145 |
2789
|
146 * If using temporary files, delete them when gnuplot exits. |
|
147 |
2560
|
148 * If possible, pass binary data to gnuplot to speed things up. |
|
149 |
2799
|
150 * If using gnuplot, consider setting a smaller default for the |
|
151 `zero' value (e.g., set zero sqrt (realmin) or something). |
|
152 |
2330
|
153 ------- |
|
154 Strings: |
|
155 ------- |
|
156 |
2789
|
157 * Improve performance of string functions, particularly for |
|
158 searching and replacing. |
|
159 |
|
160 * Provide some regex matching functions. |
|
161 |
2330
|
162 * Convert string functions to work on string arrays. |
|
163 |
|
164 * Make find work for strings. |
|
165 |
2378
|
166 * Consider making octave_print_internal() print some sort of text |
|
167 representation for unprintable characters instead of sending them |
|
168 directly to the terminal. (But don't do this for fprintf!) |
|
169 |
|
170 * Consider changing the default value of `string_fill_char' from SPC |
|
171 to NUL. |
|
172 |
2560
|
173 * Consider making ["test", []] ==> "test", for compatibility with |
|
174 Matlab, at least when some set of preferences are set. |
|
175 |
2330
|
176 ---------------- |
|
177 Other Data Types: |
|
178 ---------------- |
|
179 |
|
180 * New types (char, short, etc.). |
|
181 |
|
182 * 3d matrix stuff. |
|
183 |
|
184 * Template functions for mixed-type ops. |
|
185 |
|
186 * Stuff for arithmetic using charMatrix, intMatrix, etc. |
|
187 |
|
188 ------------------------ |
|
189 Graphical User Interface: |
|
190 ------------------------ |
|
191 |
|
192 * In an X11 or other windowing environment, allow the user to pop up |
|
193 windows for menus and other purposes. A good place to start might |
|
194 be Tk, as long as Tcl is avoided. |
|
195 |
|
196 * Add a way to handle events, like alarms, mouse clicks, etc. |
|
197 |
|
198 ------------ |
|
199 Input/Output: |
|
200 ------------ |
|
201 |
|
202 * Make fread and fwrite work for complex data. Iostreams based |
|
203 versions of these functions would also be nice, and if you are |
|
204 working on them, it would be good to support other size |
|
205 specifications (integer*2, etc.). |
|
206 |
|
207 * Make load and save work for structures. |
|
208 |
|
209 * Make load and save look for <file>.mat if only given <file>. |
|
210 |
2799
|
211 Potential sticky points: |
|
212 |
|
213 - For load, if both foo and foo.mat exist, should it prefer foo |
|
214 or foo.mat? Should the preference depend on the arguments to |
|
215 load? I think it should only prefer .mat files if the |
|
216 -mat-binary option is supplied, or if the file foo.mat exists |
|
217 but the file foo does not. |
|
218 |
|
219 - For save, should it prefer to create foo or foo.mat? Should |
|
220 the preference depend on the arguments to save? Should the |
|
221 default_save_format imply a default preference? I think it |
|
222 should only create .mat files if it is writing Matlab |
|
223 compatible files. |
|
224 |
2330
|
225 * Move some pr-output stuff to liboctave. |
|
226 |
|
227 * Make the cutoff point for changing to packed storage a |
|
228 user-preference variable with default value 8192. |
|
229 |
|
230 * Save image data in binary format to save space. |
|
231 |
2378
|
232 * Make it possible to load other image formats (ppm, pbm, etc. would |
|
233 probably be best since there are already filters to convert to |
|
234 these formats from others.) |
|
235 |
2330
|
236 * Use HDF for binary data. |
|
237 |
2789
|
238 * Make ascii load and save work for Inf and NaN. (This is really a |
|
239 problem with the functions for reading floats in the GNU iolib.) |
2330
|
240 |
|
241 * Complain if there is not enough disk space available (I think |
|
242 there is simply not enough error checking in the code that handles |
|
243 writing data). |
|
244 |
|
245 * Make it possible to tie arbitrary input and output streams |
|
246 together, similar to the way iostreams can be tied together. |
|
247 |
3105
|
248 * Allow comments in number-only data files. |
|
249 |
2330
|
250 ----------- |
|
251 Interpreter: |
|
252 ----------- |
|
253 |
3162
|
254 * Allow customization of the debug prompt. |
|
255 |
|
256 * For the keyboard function, parse return (or quit) more |
|
257 intelligently so that something like |
|
258 |
|
259 debug> x = 1; return |
|
260 |
|
261 will work as expected. |
|
262 |
2601
|
263 * Fix the parser so that |
|
264 |
|
265 function foo () |
|
266 implicit_str_to_num_ok = 1; |
|
267 '#' + 0; |
|
268 endfunction |
|
269 |
|
270 succeeds, even when implicit_str_to_num_ok is 0 at the time the |
|
271 function is parsed. |
|
272 |
3081
|
273 * Fix the parser so that |
|
274 |
|
275 if (expr) 'this is a string' end |
|
276 |
|
277 is parsed as IF expr STRING END. |
|
278 |
2601
|
279 * Consider making x(:) work no matter what the value of |
|
280 do_fortran_indexing. |
|
281 |
2893
|
282 * For indexing operations, allow `$' to indicate the last element. |
|
283 For example, b = a (3:$, 1:$-1). This means b is a, except for |
|
284 its first 2 rows and last column. Note that `$' must be able to |
|
285 appear in arbitrarily complex expressions, so copying the the |
|
286 current implementation of `:' won't work. |
|
287 |
2569
|
288 * If foo.oct and foo.m both exist in the LOADPATH, Octave will |
|
289 always find foo.oct, even if foo.m appears earlier in the list of |
|
290 directories. This should be fixed (in the kpathsearch library) to |
|
291 find the first .oct or .m file in the path, and only prefer .oct |
|
292 over .m if both files are in the same directory. |
|
293 |
2330
|
294 * Consider grouping all preference variables in a structure instead |
|
295 of further polluting the namespace. Maybe `Octave_options.xxx'? |
|
296 |
|
297 * Rewrite functions in input.cc to do the right thing in all cases |
|
298 of using readline, forced interactive behavior, echoing of input, |
|
299 etc. |
|
300 |
|
301 * Make it possible to redefine built-in variables and functions. |
|
302 (Fix whatever it is that is happening with clear for built-in |
|
303 variables.) |
|
304 |
|
305 * Make it possible to check if a variable is complex. |
|
306 |
|
307 * Consider making linspace() and logspace() return the value |
|
308 corresponding to the first argument if the number of requested |
|
309 points is 1. |
|
310 |
|
311 * Consider allowing an arbitrary property list to be attached to any |
|
312 variable. This could be a more general way to handle the help |
|
313 string that can currently be added with `document'. |
|
314 |
|
315 * Allow more command line options to be accessible as built-in |
|
316 variables (--echo-commands, etc.). |
|
317 |
|
318 * Allow `octave -c STRING' to execute commands from STRING then |
|
319 exit. |
|
320 |
|
321 * Handle DOS style CRLFs in M files. This should probably be |
|
322 optional. |
|
323 |
|
324 * Make the interpreter run faster. |
|
325 |
|
326 * Make it possible to disable or enable all warnings on an |
|
327 individual basis from the command line or via some built-in |
|
328 structure variable. |
|
329 |
2378
|
330 * Warn about complex comparisons? Could just use double_value() or |
|
331 matrix_value() instead of explicit conversions to real types. |
|
332 |
|
333 * Make warnings also give some indication about the location of the |
|
334 code that triggers the warning. |
|
335 |
|
336 * Warn about complex comparisons? Could just use double_value() or |
|
337 matrix_value() instead of explicit conversions to real types. For |
|
338 this to really be useful, some additional information must be |
|
339 available to point to the location of the code that triggers the |
|
340 warning. |
|
341 |
2330
|
342 * Consider making it possible to have arrays of structures, and some |
|
343 way of indexing them. |
|
344 |
|
345 * Allow arbitrary lower bounds for array indexing. (Watch out for |
|
346 the additional conflict with zero-one style indexing.) |
|
347 |
|
348 * Recursive problems. |
|
349 |
|
350 * Improve the way ignore_function_time_stamp works to allow |
3167
|
351 selecting by individual directories or functions. |
2330
|
352 |
|
353 * Make it possible to ask exist() to only look for certain classes |
|
354 of variables, functions, files, etc. rather than always looking for |
|
355 everything. |
|
356 |
|
357 * Add a command-line option to tell Octave to just do syntax |
|
358 checking and not execute statements. |
|
359 |
|
360 * Is it necessary for do_binary_op and do_unary_op to be friends of |
|
361 the tree_constant class. |
|
362 |
|
363 * Clean up symtab and variable stuff. |
|
364 |
|
365 * Input stream class for parser files -- must manage buffers for |
|
366 flex and context for global variable settings. |
|
367 |
3162
|
368 * make parser do more semantic checking, continue after errors when |
|
369 compiling functions, etc. |
|
370 |
2330
|
371 * Make LEXICAL_ERROR have a value that is the error message for |
|
372 parse_error() to print? |
|
373 |
|
374 * Make it possible to clear dynamically loaded functions. |
|
375 |
|
376 * Add a run-time alias mechanism that would allow things like |
|
377 |
|
378 alias fun function_with_a_very_long_name |
|
379 |
|
380 so that `function_with_a_very_long_name' could be invoked as |
|
381 `fun'. |
|
382 |
|
383 * What should is_global() return when called for built-in variables? |
|
384 |
|
385 * Allow local changes to variables to be written more compactly than |
|
386 is currently possible with unwind_protect. For example, |
|
387 |
|
388 function f () |
|
389 local prefer_column_vectors = something; |
|
390 ... |
|
391 endfunction |
|
392 |
|
393 would be equivalent to |
|
394 |
|
395 function f () |
|
396 unwind_protect |
|
397 save_prefer_column_vectors = prefer_column_vectors; |
|
398 prefer_column_vectors = something; |
|
399 ... |
|
400 unwind_protect_cleanup |
|
401 prefer_column_vectors = save_prefer_column_vectors; |
|
402 end_unwind_protect |
|
403 endfunction |
|
404 |
|
405 * Fix all function files to check for bogus inputs (wrong number or |
|
406 types of input arguments, wrong number of output arguments). |
|
407 |
2789
|
408 * Reduce the memory and time required to parse very large matrix |
|
409 lists. |
|
410 |
2330
|
411 * Handle options for built-in functions more consistently. |
|
412 |
|
413 * Too much time is spent allocating and freeing memory. What can be |
|
414 done to improve performance? |
|
415 |
|
416 * Error output from Fortran code is ugly. Something should be done to |
|
417 make it look better. |
|
418 |
|
419 * It would be nice if output from the Fortran routines could be |
|
420 passed through the pager. |
|
421 |
|
422 * Attempt to recognize common subexpressions in the parser. |
|
423 |
|
424 * Remove the buffer size limit in octave_read() in input.cc. |
|
425 |
|
426 * Handle arrays with more than two dimensions. |
|
427 |
|
428 * Consider making assignment statements like |
|
429 |
|
430 M (i_idx, j_idx) = RHS |
|
431 |
|
432 produce the value of RHS instead of the value of M. |
|
433 |
|
434 * Consider making it possible to specify an empty matrix with a |
|
435 syntax like [](e1, e2). Of course at least one of the expressions |
|
436 must be zero... |
|
437 |
|
438 * Eliminate force_numeric and make_numeric functions. |
|
439 |
|
440 * Is Matrix::fortran_vec() really necessary? |
2862
|
441 |
2330
|
442 * print_usage() should set error_state in most cases? |
|
443 |
|
444 * Make statements like this |
|
445 |
|
446 foo bar baz |
|
447 |
|
448 turn into function calls like this: |
|
449 |
|
450 foo ("bar", "baz") |
|
451 |
|
452 This is pretty ugly and introduces some conflicts. |
|
453 |
|
454 * Add a command that works like bash's `builtin' command. |
|
455 |
|
456 * Handle comments in parse trees for use with the type command. |
|
457 |
2378
|
458 * Make the type command handle script files too, by just reading and |
|
459 printing them. |
|
460 |
3059
|
461 * The wich and type commands should distinguish between dynamically |
|
462 linked functions and built-in functions. For dynamically linked |
|
463 functions, the location of the .oct file should be displayed. |
|
464 |
2330
|
465 * Clean up eye, eval, feval, keyboard, input, ones, zeros. |
|
466 |
2799
|
467 * Turn off printing when evaluating expressions in an eval() |
|
468 statement, for compatibility with Matlab. |
|
469 |
2330
|
470 * It would be nice to have an interactive debugger. |
|
471 |
2378
|
472 * Make whos report total memory used by variables (and functions?). |
|
473 |
2746
|
474 * Rewrite whos and the symbol_record_info class. Write a built-in |
|
475 function that gives all the basic information, then write who and |
|
476 whos as M-files. |
2439
|
477 |
3125
|
478 * Make whos work for structure elements: |
|
479 |
|
480 prot type rows cols name |
|
481 ==== ==== ==== ==== ==== |
|
482 wd matrix m n struct.x |
|
483 |
2799
|
484 * On systems that support matherr(), make it possible for users to |
|
485 enable the printing of warning messages. |
|
486 |
2862
|
487 * Make it possible to mark variables and functions as read-only. |
|
488 |
3060
|
489 * Provide a built-in function for applying a scalar function to an |
|
490 array. Be sure to note in the manual that this is not the |
|
491 preferred way to write a function that can handle vector/matrix |
|
492 arguments because there is a significant overhead for function |
|
493 calls. If you are really looking to make a function work for |
|
494 vector/matrix arguments and you want it to run fast, you should |
|
495 write it in terms of the existing vector/matrix operators as much |
|
496 as possible. |
|
497 |
3092
|
498 * Make it possible to write a function that gets a reference to a |
|
499 matrix in memory and change one or more elements without |
|
500 generating a second copy of the data. |
|
501 |
2330
|
502 ------- |
|
503 History: |
|
504 ------- |
|
505 |
|
506 * Add an option to allow saving input from script files in the |
|
507 history list. |
|
508 |
3092
|
509 * The history command should accept two numeric arguments to |
|
510 indicate a range of history entries to display, save or read. |
|
511 |
2862
|
512 * Add an option to include information about the Octave session in |
|
513 the history list. Possibly a time/date stamp and the current |
|
514 Octave line number, appended as a comment (users should probably |
|
515 be able to control the format). |
|
516 |
2330
|
517 * Fix history problems -- core dump if multiple processes are |
|
518 writing to the same history file? |
|
519 |
|
520 ------------------------------ |
|
521 Configuration and Installation: |
|
522 ------------------------------ |
|
523 |
2473
|
524 * Handle USE_READLINE so that --enable-readline will work. |
2330
|
525 |
|
526 * Make Octave as independent of the particular readline version as |
|
527 possible. |
|
528 |
2473
|
529 * Add an --enable-pathsearch option to configure to make it possible |
|
530 to configure and run without kpathsea. |
|
531 |
2811
|
532 * Make configure take more defaults from the environment. It should |
|
533 definitely pay attention to CFLAGS, CXXFLAGS, FFLAGS, etc. |
2469
|
534 |
2330
|
535 * Should --enable-lite-kernel imply --enable-shared? |
|
536 |
3069
|
537 * Should --enable-dl imply --enable-shared? |
|
538 |
2330
|
539 * Make it possible to configure without readline. |
|
540 |
|
541 * Makefile changes: |
|
542 -- eliminate for loops |
|
543 -- define shell commands or eliminate them |
|
544 -- verify distclean |
|
545 -- consolidate targets |
|
546 |
|
547 * Make it possible to configure so that installed binaries and |
|
548 shared libraries are stripped. |
|
549 |
|
550 * Make installation of the static libraries optional. |
|
551 |
3069
|
552 * Create a docs-only distribution? |
|
553 |
2330
|
554 ------------------------------ |
|
555 Documentation and On-Line Help: |
|
556 ------------------------------ |
|
557 |
|
558 * Document new features. |
|
559 -- history-search-{back,for}ward. |
|
560 -- try/catch. |
|
561 -- Other stuff mentioned in the NEWS file. |
|
562 |
|
563 * Improve the Texinfo Documentation for the interpreter. It would |
|
564 be useful to have lots more examples, to not have so many forward |
|
565 references, and to not have very many simple lists of functions. |
|
566 |
|
567 * The docs should mention something about efficiency and that using |
|
568 array operations is almost always a good idea for speed. |
|
569 |
|
570 * Texinfo documentation for the C++ classes. |
|
571 |
|
572 * Support multiple info files, perhaps allowing one or more in each |
|
573 directory in the LOADPATH, so that local collections of M-files |
|
574 could be documented with Info. |
|
575 |
|
576 * Figure out a good way to have functions and variables documented |
|
577 in only one place. |
|
578 |
|
579 * Improve help messages for operators and keywords in help.cc. |
|
580 |
|
581 * Make index entries more consistent to improve behavior of `help -i'. |
|
582 |
|
583 * Make `help -i' try to find a whole word match first. |
|
584 |
|
585 * Allow help for local additions to be accessible with help -i. |
|
586 |
|
587 * Make the arrow keys work in the info reader. |
|
588 |
|
589 * Update Info to be based on the current Texinfo release. |
|
590 |
2378
|
591 * Should info/terminal.c include definitions for PC, BC, UP, and |
|
592 ospeed on all systems? |
|
593 |
3130
|
594 * Clean up help stuff. |
|
595 |
2330
|
596 * Demo files. |
|
597 |
|
598 * As the number of m-files with octave grows perhaps a 'Contents.m' |
|
599 file for each toolbox (directory) would be appropriate so one |
|
600 knows exactly what functions are in a toolbox with a quick look. |
|
601 It would be best to generate information for each function directly |
|
602 from the M-files, so that the information doesn't have to be |
|
603 duplicated, and will remain current if the M-files change. It |
|
604 would also be best to do as much of this as possible in an M-file, |
|
605 though I wouldn't mind adding some basic support for listing the |
|
606 names of all the directories in the LOADPATH, and the names of all |
|
607 the M-files in a given directory if that is needed. |
|
608 |
2787
|
609 Also make it possible to recursively search for Contents files: |
|
610 |
|
611 help dir -- Contents from dir |
|
612 help dir// -- Contents from dir and all its subdirectories |
|
613 help dir1/dir2 -- Contents from dir2 which is under dir1 |
|
614 |
2330
|
615 * Some sort of blurb (2-3 pages) describing Octave in a reasonably |
|
616 coherent way, where to get it etc., perhaps formatted pretty, |
|
617 i.e. not just text. Maybe start with the latest Announce file. |
|
618 |
|
619 ----- |
|
620 Tests: |
|
621 ----- |
|
622 |
|
623 * Improved set of tests: |
|
624 |
|
625 -- Tests for various functions. Would be nice to have a test file |
|
626 corresponding to every function. |
|
627 |
|
628 -- Tests for element by element operators: |
|
629 + - .* ./ .\ .^ | & < <= == >= > != ! |
|
630 |
|
631 -- Tests for boolean operators: && || |
|
632 |
|
633 -- Tests for other operators: * / \ ' .' |
|
634 |
|
635 -- Tests from bug reports. |
|
636 |
|
637 -- Tests for indexed assignment. Need to consider the following: |
|
638 o fortran-style indexing |
|
639 o zero-one indexing |
|
640 o assignment of empty matrix as well as values |
|
641 o resizing |
|
642 |
|
643 * Tests for all internal functions. |
|
644 |
|
645 ----------- |
|
646 Programming: |
|
647 ----------- |
|
648 |
|
649 * Move toward using more stuff from standard C++ library. |
|
650 |
|
651 * More C++/Fortran cleanups. |
|
652 |
|
653 * It is likely that there are still some memory leaks. Hunt then down |
|
654 and plug them. |
|
655 |
2475
|
656 * Better error messages for missing operators? |
|
657 |
|
658 * Eliminate duplicate enums in pt-exp.cc, pt-const.cc, and ov.cc. |
|
659 |
|
660 * Handle octave_print_internal() stuff at the liboctave level. Then |
|
661 the octave_value classes could just call on the print() methods |
|
662 for the underlying classes. |
|
663 |
|
664 * As much as possible, eliminate explicit checks for the types of |
|
665 octave_value objects so that user-defined types will automatically |
|
666 do the right thing in more cases. |
|
667 |
2330
|
668 * Only include config.h in files that actually need it, instead of |
|
669 including it in every .cc file. Unfortunately, this might not be |
|
670 so easy to figure out. |
|
671 |
|
672 * GNU coding standards: |
|
673 |
|
674 -- Add a `Makefile' target to the Makefiles. |
|
675 -- Comments on #else and #endif preprocessor commands. |
|
676 -- Change error message format to match standards everywhere. |
|
677 |
|
678 * Use STL stuff instead of libg++ lists, maps, and stacks. |
|
679 |
|
680 * Eliminate more global variables. |
|
681 |
|
682 * Encapsulate readline in a class. Include interface to stuff like |
|
683 blink_matching_paren. |
|
684 |
|
685 * Encapsulate resource stuff in a class. |
|
686 |
|
687 * Move procstream to liboctave. |
|
688 |
|
689 * Replace more C-style vectors with Array<T> stuff. |
|
690 |
|
691 * Use references and classes in more places. |
|
692 |
|
693 * Share more code among the various *_options functions. |
|
694 |
|
695 ------------- |
|
696 Miscellaneous: |
|
697 ------------- |
|
698 |
|
699 * Implement some functions for interprocess communication: bind, |
|
700 accept, connect, gethostbyname, etc. |
|
701 |
2454
|
702 * The installation process should also install octave.el. This |
|
703 needs to detect the appropriate Emacs binary to use to |
|
704 byte-compile the .el file. Following GNU Emacs philosophy, |
|
705 installation would be into $(prefix)/share/emacs/site-lisp by |
|
706 default, but it should be selectable. |
|
707 |
2330
|
708 * The ability to transparently handle very large files: |
|
709 |
|
710 Juhana K Kouhia <kouhia@nic.funet.fi> wrote: |
|
711 |
|
712 If I have a one-dimensional signal data with the size 400 |
|
713 Mbytes, then what are my choices to operate with it: |
|
714 |
|
715 * I have to split the data |
|
716 * Octave has a virtual memory on its own and I don't have to |
|
717 worry about the splitting. |
|
718 |
|
719 If I split the data, then my easily programmed processing |
|
720 programs will become hard to program. |
|
721 |
|
722 If possible, I would like to have the virtual memory system in |
|
723 Octave i.e. the all big files, the user see as one big array or |
|
724 such. There could be several user selectable models to do the |
|
725 virtual memory depending on what kind of data the user have (1d, |
|
726 2d) and in what order they are processed (stream or random |
|
727 access). |
|
728 |
|
729 Perhaps this can be done entirely with a library of M-files. |
|
730 |
3136
|
731 * An interface to gdb. |
|
732 |
|
733 Michael Smolsky <fnsiguc@weizmann.weizmann.ac.il> wrote: |
|
734 |
|
735 I was thinking about a tool, which could be very useful for me |
|
736 in my numerical simulation work. It is an interconnection |
|
737 between gdb and octave. We are often managing very large arrays |
|
738 of data in our fortran or c codes, which might be studied with |
|
739 the help of octave at the algorithm development stages. Assume |
|
740 you're coding, say, wave equation. And want to debug the |
|
741 code. It would be great to pick some array from the memory of |
|
742 the code you're develloping, fft it and see the image as a |
|
743 log-log plot of the spectral density. I'm facing similar |
|
744 problems now. To avoid high c-development cost, I develop in |
|
745 matlab/octave, and then rewrite into c. It might be so much |
|
746 easier, if I could off-load a c array right from the debugger |
|
747 into octave, study it, and, perhaps, change some [many] values |
|
748 with a convenient matlab/octave syntax, similar to |
|
749 a(:,50:250)=zeros(100,200), and then store it back into the |
|
750 memory of my c code. |
|
751 |
2330
|
752 * Add a function like strptime() which is the opposite of |
|
753 strftime(). A C version is apparently in recent releases of the |
|
754 Linux C library. |
|
755 |
2789
|
756 * Add a definition to lgrind so that it supports Octave. |
|
757 (See http://www.tex.ac.uk/tex-archive/support/lgrind/ for more |
|
758 information about lgrind.) |
|
759 |
2330
|
760 ------ |
|
761 Always: |
|
762 ------ |
|
763 |
|
764 * Squash bugs. |
|
765 |
|
766 --30-- |
3162
|
767 </pre> |
|
768 </html> |