5 This is a production release that fixes a build problem in the FreeBSD base
8 All other users do **NOT** need to upgrade.
12 This is a production release that fixes prompt bugs with editline and readline
13 where the `BC_PROMPT` environment variable was not being respected.
15 This also fixes editline and readline output on `EOF`.
19 This is a production release that fixes a build problem in the FreeBSD base
20 system, as well as a problem in the `en_US` locale. If you don't have problems
21 with either, you do not need to upgrade.
25 This is a production release that adds features and has a few bug fixes.
27 First, support for editline and readline history has been added. To use
28 editline, pass `-e` to `configure.sh`, and to use readline, pass `-r`.
30 Second, history support for Windows has been fixed and re-enabled.
32 Third, command-line options to set `scale`, `ibase`, `obase`, and `seed` were
33 added. This was requested long ago, and I originally disagreed with the idea.
35 Fourth, the manuals had typos and were missing information. That has been fixed.
37 Fifth, the manuals received different formatting to be more readable as
42 This is a production release that fixes this `bc`'s behavior on `^D` to match
47 This is a production release that fixes two bugs in history:
49 * Without prompt, the cursor could not be placed on the first character in a
51 * Home and End key handling in `tmux` was fixed.
53 Any users that do not care about these improvements do not need to upgrade.
57 This is a production release that fixes one bug, a parse error when passing a
58 file to `bc` using `-f` if that file had a multiline comment or string in it.
62 This is a production release that fixes one bug, a segmentation fault if
63 `argv[0]` equals `NULL`.
65 This is not a critical bug; there will be no vulnerability as far as I can tell.
66 There is no need to update if you do not wish to.
70 This is a production release that fixes two parse bugs when in POSIX standard
71 mode. One of these bugs was due to a quirk of the POSIX grammar, and the other
72 was because `bc` was too strict.
76 This is a production release that adds a new feature, fixes some bugs, and adds
77 out-of-source builds and a `pkg-config` file for `bcl`.
79 The new feature is the ability to turn off exiting on expressions. It is also
80 possible to set the default using `configure.sh`. This behavior used to exist
81 with the `BC_EXPR_EXIT` environment variable, which is now used again.
85 * Some possible race conditions with error handling.
86 * Install and uninstall targets for `bcl` did not work.
90 This is a production release that completes a bug fix from `5.1.0`. The bug
91 exists in all versions of `bc`.
93 The bug was that `if` statements without `else` statements would not be handled
94 correctly at the end of files or right before a function definition.
98 This is a production release with some fixes and new features.
100 * Fixed a bug where an `if` statement without an `else` before defining a
101 function caused an error.
102 * Fixed a bug with the `bc` banner and `-q`.
103 * Fixed a bug on Windows where files were not read correctly.
104 * Added a command-line flag (`-z`) to make `bc` and `dc` print leading zeroes on
105 numbers `-1 < x < 1`.
106 * Added four functions to `lib2.bc` (`plz()`, `plznl()`, `pnlz()`, and
107 `pnlznl()`) to allow printing numbers with or without leading zeros, despite
108 the use of `-z` or not.
109 * Added builtin functions to query global state like line length, global stacks,
111 * Added a command-line flag (`-L`) to disable wrapping when printing numbers.
112 * Improved builds on Windows.
116 This is a production release with one fix for a flaky test. If you have not
117 experienced problems with the test suite, you do ***NOT*** need to upgrade.
119 The test was one that tested whether `bc` fails gracefully when it can't
120 allocate memory. Unfortunately, there are cases when Linux and FreeBSD lie and
121 pretend to allocate the memory.
123 The reason they do this is because a lot of programs don't use all of the memory
124 they allocate, so those OS's usually get away with it.
126 However, this `bc` uses all of the memory it allocates (at least at page
127 granularity), so when it tries to use the memory, FreeBSD and Linux kill it.
129 This only happens sometimes, however. Other times (on my machine), they do, in
130 fact, refuse the request.
132 So I changed the test to not test for that because I think the graceful failure
133 code won't really change much.
137 This is a production release with two fixes:
139 * Fix for the build on Mac OSX.
140 * Fix for the build on Android.
142 Users that do not use those platforms do ***NOT*** need to update.
146 This is a major production release with several changes:
148 * Added support for OpenBSD's `pledge()` and `unveil()`.
149 * Fixed print bug where a backslash newline combo was printed even if only one
150 digit was left, something I blindly copied from GNU `bc`, like a fool.
151 * Fixed bugs in the manuals.
152 * Fixed a possible multiplication overflow in power.
153 * Temporary numbers are garbage collected if allocation fails, and the
154 allocation is retried. This is to make `bc` and `dc` more resilient to running
156 * Limited the number of temporary numbers and made the space for them static so
157 that allocating more space for them cannot fail.
158 * Allowed integers with non-zero `scale` to be used with power, places, and
160 * Added greatest common divisor and least common multiple to `lib2.bc`.
161 * Added `SIGQUIT` handling to history.
162 * Added a command to `dc` (`y`) to get the length of register stacks.
163 * Fixed multi-digit bugs in `lib2.bc`.
164 * Removed the no prompt build option.
165 * Created settings that builders can set defaults for and users can set their
166 preferences for. This includes the `bc` banner, resetting on `SIGINT`, TTY
168 * Added history support to Windows.
169 * Fixed bugs with the handling of register names in `dc`.
170 * Fixed bugs with multi-line comments and strings in both calculators.
171 * Added a new error type and message for `dc` when register stacks don't have
173 * Optimized string allocation.
174 * Made `bc` and `dc` UTF-8 capable.
175 * Fixed a bug with `void` functions.
176 * Fixed a misspelled symbol in `bcl`. This is technically a breaking change,
177 which requires this to be `5.0.0`.
178 * Added the ability for users to get the copyright banner back.
179 * Added the ability for users to have `bc` and `dc` quit on `SIGINT`.
180 * Added the ability for users to disable prompt and TTY mode by environment
182 * Added the ability for users to redefine keywords. This is another reason this
184 * Added `dc`'s modular exponentiation and divmod to `bc`.
185 * Added the ability to assign strings to variables and array elements and pass
186 them to functions in `bc`.
187 * Added `dc`'s asciify command and stream printing to `bc`.
188 * Added a command to `dc` (`Y`) to get the length of an array.
189 * Added a command to `dc` (`,`) to get the depth of the execution stack.
190 * Added bitwise and, or, xor, left shift, right shift, reverse, left rotate,
191 right rotate, and mod functions to `lib2.bc`.
192 * Added the functions `s2u(x)` and `s2un(x,n)`, to `lib2.bc`.
196 This is a production release that fixes two bugs:
198 1. If no files are used and the first statement on `stdin` is invalid, `scale`
199 would not be set to `20` even if `-l` was used.
200 2. When using history, `bc` failed to respond properly to `SIGSTOP` and
205 This is a production release that only adds one thing: flushing output when it
206 is printed with a print statement.
210 This is a production release with many fixes, a new command-line option, and a
213 * A bug was fixed in `dc`'s `P` command where the item on the stack was *not*
215 * Various bugs in the manuals have been fixed.
216 * A known bug was fixed where history did not interact well with prompts printed
217 by user code without newlines.
218 * A new command-line option, `-R` and `--no-read-prompt` was added to disable
219 just the prompt when using `read()` (`bc`) or `?` (`dc`).
220 * And finally, **official support for Windows was added**.
222 The last item is why this is a major version bump.
224 Currently, only one set of build options (extra math and prompt enabled, history
225 and NLS/locale support disabled, both calculators enabled) is supported on
226 Windows. However, both debug and release builds are supported.
228 In addition, Windows builds are supported for the the library (`bcl`).
230 For more details about how to build on Windows, see the [README][5] or the
235 This is a production release that fixes a small bug.
237 The bug was that output was not flushed before a `read()` call, so prompts
238 without a newline on the end were not flushed before the `read()` call.
240 This is such a tiny bug that users only need to upgrade if they are affected.
244 This is a production release with one tweak and fixes for manuals.
246 The tweak is that `length(0)` returns `1` instead of `0`. In `3.3.1`, I changed
247 it so `length(0.x)`, where `x` could be any number of digits, returned the
248 `scale`, but `length(0)` still returned `0` because I believe that `0` has `0`
251 After request of FreeBSD and considering the arguments of a mathematician,
252 compatibility with other `bc`'s, and the expectations of users, I decided to
255 The fixes for manuals fixed a bug where `--` was rendered as `-`.
259 This is a production release that fixes a divide-by-zero bug in `root()` in the
260 [extended math library][16]. All previous versions with `root()` have the bug.
264 This is a production release that fixes a bug.
266 The bug was in the reporting of number length when the value was 0.
270 This is a production release that changes one behavior and fixes documentation
273 The changed behavior is the treatment of `-e` and `-f` when given through
274 `BC_ENV_ARGS` or `DC_ENV_ARGS`. Now `bc` and `dc` do not exit when those options
275 (or their equivalents) are given through those environment variables. However,
276 `bc` and `dc` still exit when they or their equivalents are given on the
281 This is a production release that removes a small non-portable shell operation
282 in `configure.sh`. This problem was only noticed on OpenBSD, not FreeBSD or
285 Non-OpenBSD users do ***NOT*** need to upgrade, although NetBSD users may also
290 This is a production release that fixes the build on FreeBSD.
292 There was a syntax error in `configure.sh` that the Linux shell did not catch,
293 and FreeBSD depends on the existence of `tests/all.sh`.
295 All users that already upgraded to `3.2.5` should update to this release, with
296 my apologies for the poor release of `3.2.5`. Other users should skip `3.2.5` in
297 favor of this version.
301 This is a production release that fixes several bugs and adds a couple small
304 The two most important bugs were bugs that causes `dc` to access memory
305 out-of-bounds (crash in debug builds). This was found by upgrading to `afl++`
306 from `afl`. Both were caused by a failure to distinguish between the same two
309 Another bug was the failure to put all of the licenses in the `LICENSE.md` file.
311 Third, some warnings by `scan-build` were found and eliminated. This needed one
312 big change: `bc` and `dc` now bail out as fast as possible on fatal errors
313 instead of unwinding the stack.
315 Fourth, the pseudo-random number now attempts to seed itself with `/dev/random`
316 if `/dev/urandom` fails.
318 Finally, this release has a few quality-of-life changes to the build system. The
319 usage should not change at all; the only thing that changed was making sure the
320 `Makefile.in` was written to rebuild properly when headers changed and to not
321 rebuild when not necessary.
325 This is a production release that fixes a warning on `gcc` 6 or older, which
326 does not have an attribute that is used.
328 Users do ***NOT*** need to upgrade if they don't use `gcc` 6 or older.
332 This is a production release that fixes a bug in `gen/strgen.sh`. I recently
333 changed `gen/strgen.c`, but I did not change `gen/strgen.sh`.
335 Users that do not use `gen/strgen.sh` do not need to upgrade.
339 This is a production release that fixes a portability bug in `configure.sh`. The
340 bug was using the GNU `find` extension `-wholename`.
344 This is a production release that has one fix for `bcl(3)`. It is technically
345 not a bug fix since the behavior is undefined, but the `BclNumber`s that
346 `bcl_divmod()` returns will be set to `BCL_ERROR_INVALID_NUM` if there is an
347 error. Previously, they were not set.
351 This is a production release that has one bug fix and a major addition.
353 The bug fix was a missing `auto` variable in the bessel `j()` function in the
356 The major addition is a way to build a version of `bc`'s math code as a library.
357 This is done with the `-a` option to `configure.sh`. The API for the library can
358 be read in `./manuals/bcl.3.md` or `man bcl` once the library is installed with
361 This library was requested by developers before I even finished version 1.0, but
362 I could not figure out how to do it until now.
364 If the library has API breaking changes, the major version of `bc` will be
369 This is a production release that fixes a new warning from Clang 12 for FreeBSD
370 and also removes some possible undefined behavior found by UBSan that compilers
371 did not seem to take advantage of.
373 Users do ***NOT*** need to upgrade, if they do not want to.
377 This is a production release that fixes the Chinese locales (which caused `bc`
378 to crash) and a crash caused by `bc` executing code when it should not have been
381 ***ALL USERS SHOULD UPGRADE.***
385 This is a production release that fixes one bug, changes two behaviors, and
386 removes one environment variable.
388 The bug is like the one in the last release except it applies if files are being
389 executed. I also made the fix more general.
391 The behavior that was changed is that `bc` now exits when given `-e`, `-f`,
392 `--expression` or `--file`. However, if the last one of those is `-f-` (using
393 `stdin` as the file), `bc` does not exit. If `-f-` exists and is not the last of
394 the `-e` and `-f` options (and equivalents), `bc` gives a fatal error and exits.
396 Next, I removed the `BC_EXPR_EXIT` and `DC_EXPR_EXIT` environment variables
397 since their use is not needed with the behavior change.
399 Finally, I made it so `bc` does not print the header, though the `-q` and
400 `--quiet` options were kept for compatibility with GNU `bc`.
404 This is a production release that fixes one minor bug: if `bc` was invoked like
405 the following, it would error:
408 echo "if (1 < 3) 1" | bc
411 Unless users run into this bug, they do not need to upgrade, but it is suggested
416 This is a production release that adds a way to install *all* locales. Users do
417 ***NOT*** need to upgrade.
419 For package maintainers wishing to make use of the change, just pass `-l` to
424 This is a production release that adds two Spanish locales. Users do ***NOT***
425 need to upgrade, unless they want those locales.
429 This is a production release that adjusts one behavior, fixes eight bugs, and
430 improves manpages for FreeBSD. Because this release fixes bugs, **users and
431 package maintainers should update to this version as soon as possible**.
433 The behavior that was adjusted was how code from the `-e` and `-f` arguments
434 (and equivalents) were executed. They used to be executed as one big chunk, but
435 in this release, they are now executed line-by-line.
437 The first bug fix in how output to `stdout` was handled in `SIGINT`. If a
438 `SIGINT` came in, the `stdout` buffer was not correctly flushed. In fact, a
439 clean-up function was not getting called. This release fixes that bug.
441 The second bug is in how `dc` handled input from `stdin`. This affected `bc` as
442 well since it was a mishandling of the `stdin` buffer.
444 The third fixed bug was that `bc` and `dc` could `abort()` (in debug mode) when
445 receiving a `SIGTERM`. This one was a race condition with pushing and popping
446 items onto and out of vectors.
448 The fourth bug fixed was that `bc` could leave extra items on the stack and
449 thus, not properly clean up some memory. (The memory would still get
450 `free()`'ed, but it would not be `free()`'ed when it could have been.)
452 The next two bugs were bugs in `bc`'s parser that caused crashes when executing
455 The last two bugs were crashes in `dc` that resulted from mishandling of
458 The manpage improvement was done by switching from [ronn][20] to [Pandoc][21] to
459 generate manpages. Pandoc generates much cleaner manpages and doesn't leave
460 blank lines where they shouldn't be.
464 This is a production release that adds one new feature: specific manpages.
466 Before this release, `bc` and `dc` only used one manpage each that referred to
467 various build options. This release changes it so there is one manpage set per
468 relevant build type. Each manual only has information about its particular
469 build, and `configure.sh` selects the correct set for install.
473 This is a production release that adds `utf8` locale symlinks and removes an
474 unused `auto` variable from the `ceil()` function in the [extended math
477 Users do ***NOT*** need to update unless they want the locales.
481 This is a production release with two small changes. Users do ***NOT*** need to
482 upgrade to this release; however, if they haven't upgraded to `3.0.0` yet, it
483 may be worthwhile to upgrade to this release.
485 The first change is fixing a compiler warning on FreeBSD with strict warnings
488 The second change is to make the new implementation of `ceil()` in `lib2.bc`
493 *Notes for package maintainers:*
495 *First, the `2.7.0` release series saw a change in the option parsing. This made
496 me change one error message and add a few others. The error message that was
497 changed removed one format specifier. This means that `printf()` will seqfault
498 on old locale files. Unfortunately, `bc` cannot use any locale files except the
499 global ones that are already installed, so it will use the previous ones while
500 running tests during install. **If `bc` segfaults while running arg tests when
501 updating, it is because the global locale files have not been replaced. Make
502 sure to either prevent the test suite from running on update or remove the old
503 locale files before updating.** (Removing the locale files can be done with
504 `make uninstall` or by running the [`locale_uninstall.sh`][22] script.) Once
505 this is done, `bc` should install without problems.*
507 *Second, **the option to build without signal support has been removed**. See
508 below for the reasons why.*
510 This is a production release with some small bug fixes, a few improvements,
511 three major bug fixes, and a complete redesign of `bc`'s error and signal
512 handling. **Users and package maintainers should update to this version as soon
515 The first major bug fix was in how `bc` executed files. Previously, a whole file
516 was parsed before it was executed, but if a function is defined *after* code,
517 especially if the function definition was actually a redefinition, and the code
518 before the definition referred to the previous function, this `bc` would replace
519 the function before executing any code. The fix was to make sure that all code
520 that existed before a function definition was executed.
522 The second major bug fix was in `bc`'s `lib2.bc`. The `ceil()` function had a
523 bug where a `0` in the decimal place after the truncation position, caused it to
524 output the wrong numbers if there was any non-zero digit after.
526 The third major bug is that when passing parameters to functions, if an
527 expression included an array (not an array element) as a parameter, it was
528 accepted, when it should have been rejected. It is now correctly rejected.
530 Beyond that, this `bc` got several improvements that both sped it up, improved
531 the handling of signals, and improved the error handling.
533 First, the requirements for `bc` were pushed back to POSIX 2008. `bc` uses one
534 function, `strdup()`, which is not in POSIX 2001, and it is in the X/Open System
535 Interfaces group 2001. It is, however, in POSIX 2008, and since POSIX 2008 is
536 old enough to be supported anywhere that I care, that should be the requirement.
538 Second, the BcVm global variable was put into `bss`. This actually slightly
539 reduces the size of the executable from a massive code shrink, and it will stop
540 `bc` from allocating a large set of memory when `bc` starts.
542 Third, the default Karatsuba length was updated from 64 to 32 after making the
543 optimization changes below, since 32 is going to be better than 64 after the
546 Fourth, Spanish translations were added.
548 Fifth, the interpreter received a speedup to make performance on non-math-heavy
549 scripts more competitive with GNU `bc`. While improvements did, in fact, get it
550 much closer (see the [benchmarks][19]), it isn't quite there.
552 There were several things done to speed up the interpreter:
554 First, several small inefficiencies were removed. These inefficiencies included
555 calling the function `bc_vec_pop(v)` twice instead of calling
556 `bc_vec_npop(v, 2)`. They also included an extra function call for checking the
557 size of the stack and checking the size of the stack more than once on several
560 Second, since the current `bc` function is the one that stores constants and
561 strings, the program caches pointers to the current function's vectors of
562 constants and strings to prevent needing to grab the current function in order
563 to grab a constant or a string.
565 Third, `bc` tries to reuse `BcNum`'s (the internal representation of
566 arbitary-precision numbers). If a `BcNum` has the default capacity of
567 `BC_NUM_DEF_SIZE` (32 on 64-bit and 16 on 32-bit) when it is freed, it is added
568 to a list of available `BcNum`'s. And then, when a `BcNum` is allocated with a
569 capacity of `BC_NUM_DEF_SIZE` and any `BcNum`'s exist on the list of reusable
570 ones, one of those ones is grabbed instead.
572 In order to support these changes, the `BC_NUM_DEF_SIZE` was changed. It used to
573 be 16 bytes on all systems, but it was changed to more closely align with the
574 minimum allocation size on Linux, which is either 32 bytes (64-bit musl), 24
575 bytes (64-bit glibc), 16 bytes (32-bit musl), or 12 bytes (32-bit glibc). Since
576 these are the minimum allocation sizes, these are the sizes that would be
577 allocated anyway, making it worth it to just use the whole space, so the value
578 of `BC_NUM_DEF_SIZE` on 64-bit systems was changed to 32 bytes.
580 On top of that, at least on 64-bit, `BC_NUM_DEF_SIZE` supports numbers with
581 either 72 integer digits or 45 integer digits and 27 fractional digits. This
582 should be more than enough for most cases since `bc`'s default `scale` values
583 are 0 or 20, meaning that, by default, it has at most 20 fractional digits. And
584 45 integer digits are *a lot*; it's enough to calculate the amount of mass in
585 the Milky Way galaxy in kilograms. Also, 72 digits is enough to calculate the
586 diameter of the universe in Planck lengths.
588 (For 32-bit, these numbers are either 32 integer digits or 12 integer digits and
589 20 fractional digits. These are also quite big, and going much bigger on a
590 32-bit system seems a little pointless since 12 digits is just under a trillion
591 and 20 fractional digits is still enough for about any use since `10^-20` light
592 years is just under a millimeter.)
594 All of this together means that for ordinary uses, and even uses in scientific
595 work, the default number size will be all that is needed, which means that
596 nearly all, if not all, numbers will be reused, relieving pressure on the system
599 I did several experiments to find the changes that had the most impact,
600 especially with regard to reusing `BcNum`'s. One was putting `BcNum`'s into
601 buckets according to their capacity in powers of 2 up to 512. That performed
602 worse than `bc` did in `2.7.2`. Another was putting any `BcNum` on the reuse
603 list that had a capacity of `BC_NUM_DEF_SIZE * 2` and reusing them for `BcNum`'s
604 that requested `BC_NUM_DEF_SIZE`. This did reduce the amount of time spent, but
605 it also spent a lot of time in the system allocator for an unknown reason. (When
606 using `strace`, a bunch more `brk` calls showed up.) Just reusing `BcNum`'s that
607 had exactly `BC_NUM_DEF_SIZE` capacity spent the smallest amount of time in both
608 user and system time. This makes sense, especially with the changes to make
609 `BC_NUM_DEF_SIZE` bigger on 64-bit systems, since the vast majority of numbers
610 will only ever use numbers with a size less than or equal to `BC_NUM_DEF_SIZE`.
612 Last of all, `bc`'s signal handling underwent a complete redesign. (This is the
613 reason that this version is `3.0.0` and not `2.8.0`.) The change was to move
614 from a polling approach to signal handling to an interrupt-based approach.
616 Previously, every single loop condition had a check for signals. I suspect that
617 this could be expensive when in tight loops.
619 Now, the signal handler just uses `longjmp()` (actually `siglongjmp()`) to start
620 an unwinding of the stack until it is stopped or the stack is unwound to
621 `main()`, which just returns. If `bc` is currently executing code that cannot be
622 safely interrupted (according to POSIX), then signals are "locked." The signal
623 handler checks if the lock is taken, and if it is, it just sets the status to
624 indicate that a signal arrived. Later, when the signal lock is released, the
625 status is checked to see if a signal came in. If so, the stack unwinding starts.
627 This design eliminates polling in favor of maintaining a stack of `jmp_buf`'s.
628 This has its own performance implications, but it gives better interaction. And
629 the cost of pushing and popping a `jmp_buf` in a function is paid at most twice.
630 Most functions do not pay that price, and most of the rest only pay it once.
631 (There are only some 3 functions in `bc` that push and pop a `jmp_buf` twice.)
633 As a side effect of this change, I had to eliminate the use of `stdio.h` in `bc`
634 because `stdio` does not play nice with signals and `longjmp()`. I implemented
635 custom I/O buffer code that takes a fraction of the size. This means that static
636 builds will be smaller, but non-static builds will be bigger, though they will
637 have less linking time.
639 This change is also good because my history implementation was already bypassing
640 `stdio` for good reasons, and unifying the architecture was a win.
642 Another reason for this change is that my `bc` should *always* behave correctly
643 in the presence of signals like `SIGINT`, `SIGTERM`, and `SIGQUIT`. With the
644 addition of my own I/O buffering, I needed to also make sure that the buffers
645 were correctly flushed even when such signals happened.
647 For this reason, I **removed the option to build without signal support**.
649 As a nice side effect of this change, the error handling code could be changed
650 to take advantage of the stack unwinding that signals used. This means that
651 signals and error handling use the same code paths, which means that the stack
652 unwinding is well-tested. (Errors are tested heavily in the test suite.)
654 It also means that functions do not need to return a status code that
655 ***every*** caller needs to check. This eliminated over 100 branches that simply
656 checked return codes and then passed that return code up the stack if necessary.
657 The code bloat savings from this is at least 1700 bytes on `x86_64`, *before*
658 taking into account the extra code from removing `stdio.h`.
662 This is a production release with one major bug fix.
664 The `length()` built-in function can take either a number or an array. If it
665 takes an array, it returns the length of the array. Arrays can be passed by
666 reference. The bug is that the `length()` function would not properly
667 dereference arrays that were references. This is a bug that affects all users.
669 **ALL USERS SHOULD UPDATE `bc`**.
673 This is a production release with fixes for new locales and fixes for compiler
678 This is a production release with a bug fix for Linux, new translations, and new
683 * Option parsing in `BC_ENV_ARGS` was broken on Linux in 2.6.1 because `glibc`'s
684 `getopt_long()` is broken. To get around that, and to support long options on
685 every platform, an adapted version of [`optparse`][17] was added. Now, `bc`
686 does not even use `getopt()`.
687 * Parsing `BC_ENV_ARGS` with quotes now works. It isn't the smartest, but it
688 does the job if there are spaces in file names.
690 The following new languages are supported:
698 All of these translations were generated using [DeepL][18], so improvements are
701 There is only one new feature: **`bc` now has a built-in pseudo-random number
704 The PRNG is seeded, making it useful for applications where
705 `/dev/urandom` does not work because output needs to be reproducible. However,
706 it also uses `/dev/urandom` to seed itself by default, so it will start with a
707 good seed by default.
709 It also outputs 32 bits on 32-bit platforms and 64 bits on 64-bit platforms, far
710 better than the 15 bits of C's `rand()` and `bash`'s `$RANDOM`.
712 In addition, the PRNG can take a bound, and when it gets a bound, it
713 automatically adjusts to remove bias. It can also generate numbers of arbitrary
714 size. (As of the time of release, the largest pseudo-random number generated by
715 this `bc` was generated with a bound of `2^(2^20)`.)
717 ***IMPORTANT: read the [`bc` manual][9] and the [`dc` manual][10] to find out
718 exactly what guarantees the PRNG provides. The underlying implementation is not
719 guaranteed to stay the same, but the guarantees that it provides are guaranteed
720 to stay the same regardless of the implementation.***
722 On top of that, four functions were added to `bc`'s [extended math library][16]
723 to make using the PRNG easier:
725 * `frand(p)`: Generates a number between `[0,1)` to `p` decimal places.
726 * `ifrand(i, p)`: Generates an integer with bound `i` and adds it to `frand(p)`.
727 * `srand(x)`: Randomizes the sign of `x`. In other words, it flips the sign of
728 `x` with probability `0.5`.
729 * `brand()`: Returns a random boolean value (either `0` or `1`).
733 This is a production release with a bug fix for FreeBSD.
735 The bug was that when `bc` was built without long options, it would give a fatal
736 error on every run. This was caused by a mishandling of `optind`.
740 This release is a production release ***with no bugfixes***. If you do not want
741 to upgrade, you don't have to.
743 No source code changed; the only thing that changed was `lib2.bc`.
745 This release adds one function to the [extended math library][16]: `p(x, y)`,
746 which calculates `x` to the power of `y`, whether or not `y` is an integer. (The
747 `^` operator can only accept integer powers.)
749 This release also includes a couple of small tweaks to the [extended math
750 library][16], mostly to fix returning numbers with too high of `scale`.
754 This release is a production release which addresses inconsistencies in the
755 Portuguese locales. No `bc` code was changed.
757 The issues were that the ISO files used different naming, and also that the
758 files that should have been symlinks were not. I did not catch that because
759 GitHub rendered them the exact same way.
763 This release is a production release.
765 No code was changed, but the build system was changed to allow `CFLAGS` to be
766 given to `CC`, like this:
769 CC="gcc -O3 -march=native" ./configure.sh
772 If this happens, the flags are automatically put into `CFLAGS`, and the compiler
773 is set appropriately. In the example above this means that `CC` will be "gcc"
774 and `CFLAGS` will be "-O3 -march=native".
776 This behavior was added to conform to GNU autotools practices.
780 This is a production release which addresses portability concerns discovered
781 in the `bc` build system. No `bc` code was changed.
783 * Support for Solaris SPARC and AIX were added.
784 * Minor documentations edits were performed.
785 * An option for `configure.sh` was added to disable long options if
786 `getopt_long()` is missing.
790 This is a production release with new translations. No code changed.
792 The translations were contributed by [bugcrazy][15], and they are for
793 Portuguese, both Portugal and Brazil locales.
797 This is a production release primarily aimed at improving `dc`.
799 * A couple of copy and paste errors in the [`dc` manual][10] were fixed.
800 * `dc` startup was optimized by making sure it didn't have to set up `bc`-only
802 * The `bc` `&&` and `||` operators were made available to `dc` through the `M`
803 and `m` commands, respectively.
804 * `dc` macros were changed to be tail call-optimized.
806 The last item, tail call optimization, means that if the last thing in a macro
807 is a call to another macro, then the old macro is popped before executing the
808 new macro. This change was made to stop `dc` from consuming more and more memory
809 as macros are executed in a loop.
811 The `q` and `Q` commands still respect the "hidden" macros by way of recording
812 how many macros were removed by tail call optimization.
816 This is a production release meant to fix warnings in the Gentoo `ebuild` by
817 making it possible to disable binary stripping. Other users do *not* need to
822 This is a production release. It fixes a bug that caused `-1000000000 < -1` to
823 return `0`. This only happened with negative numbers and only if the value on
824 the left was more negative by a certain amount. That said, this bug *is* a bad
825 bug, and needs to be fixed.
827 **ALL USERS SHOULD UPDATE `bc`**.
831 This is a production release with changes to the build system.
835 This release is a production release. It only has new features and performance
838 1. The performance of `sqrt(x)` was improved.
839 2. The new function `root(x, n)` was added to the extended math library to
840 calculate `n`th roots.
841 3. The new function `cbrt(x)` was added to the extended math library to
842 calculate cube roots.
846 This is a non-critical release; it just changes the build system, and in
849 1. Linked locale files were changed to link to their sources with a relative
851 2. A bug in `configure.sh` that caused long option parsing to fail under `bash`
856 This release is not a critical release.
858 1. A few codes were added to history.
859 2. Multiplication was optimized a bit more.
860 3. Addition and subtraction were both optimized a bit more.
864 This release contains a fix for the test suite made for Linux from Scratch: now
865 the test suite prints `pass` when a test is passed.
867 Other than that, there is no change in this release, so distros and other users
868 do not need to upgrade.
872 This release is a production release.
874 The following bugs were fixed:
876 1. A `dc` bug that caused stack mishandling was fixed.
877 2. A warning on OpenBSD was fixed.
878 3. Bugs in `ctrl+arrow` operations in history were fixed.
879 4. The ability to paste multiple lines in history was added.
880 5. A `bc` bug, mishandling of array arguments to functions, was fixed.
881 6. A crash caused by freeing the wrong pointer was fixed.
882 7. A `dc` bug where strings, in a rare case, were mishandled in parsing was
885 In addition, the following changes were made:
887 1. Division was slightly optimized.
888 2. An option was added to the build to disable printing of prompts.
889 3. The special case of empty arguments is now handled. This is to prevent
890 errors in scripts that end up passing empty arguments.
891 4. A harmless bug was fixed. This bug was that, with the pop instructions
892 (mostly) removed (see below), `bc` would leave extra values on its stack for
893 `void` functions and in a few other cases. These extra items would not
894 affect anything put on the stack and would not cause any sort of crash or
895 even buggy behavior, but they would cause `bc` to take more memory than it
898 On top of the above changes, the following optimizations were added:
900 1. The need for pop instructions in `bc` was removed.
901 2. Extra tests on every iteration of the interpreter loop were removed.
902 3. Updating function and code pointers on every iteration of the interpreter
903 loop was changed to only updating them when necessary.
904 4. Extra assignments to pointers were removed.
906 Altogether, these changes sped up the interpreter by around 2x.
908 ***NOTE***: This is the last release with new features because this `bc` is now
909 considered complete. From now on, only bug fixes and new translations will be
914 This is a production, bug-fix release.
916 Two bugs were fixed in this release:
918 1. A rare and subtle signal handling bug was fixed.
919 2. A misbehavior on `0` to a negative power was fixed.
921 The last bug bears some mentioning.
923 When I originally wrote power, I did not thoroughly check its error cases;
924 instead, I had it check if the first number was `0` and then if so, just return
925 `0`. However, `0` to a negative power means that `1` will be divided by `0`,
928 I caught this, but only after I stopped being cocky. You see, sometime later, I
929 had noticed that GNU `bc` returned an error, correctly, but I thought it was
930 wrong simply because that's not what my `bc` did. I saw it again later and had a
931 double take. I checked for real, finally, and found out that my `bc` was wrong
934 That was bad on me. But the bug was easy to fix, so it is fixed now.
936 There are two other things in this release:
938 1. Subtraction was optimized by [Stefan Eßer][14].
939 2. Division was also optimized, also by Stefan Eßer.
943 This release contains a fix for a possible overflow in the signal handling. I
944 would be surprised if any users ran into it because it would only happen after 2
945 billion (`2^31-1`) `SIGINT`'s, but I saw it and had to fix it.
949 This release contains very few things that will apply to any users.
951 1. A slight bug in `dc`'s interactive mode was fixed.
952 2. A bug in the test suite that was only triggered on NetBSD was fixed.
953 3. **The `-P`/`--no-prompt` option** was added for users that do not want a
955 4. A `make check` target was added as an alias for `make test`.
956 5. `dc` got its own read prompt: `?> `.
960 This release is a production release.
962 This release is also a little different from previous releases. From here on
963 out, I do not plan on adding any more features to this `bc`; I believe that it
964 is complete. However, there may be bug fix releases in the future, if I or any
965 others manage to find bugs.
967 This release has only a few new features:
969 1. `atan2(y, x)` was added to the extended math library as both `a2(y, x)` and
971 2. Locales were fixed.
972 3. A **POSIX shell-compatible script was added as an alternative to compiling
973 `gen/strgen.c`** on a host machine. More details about making the choice
974 between the two can be found by running `./configure.sh --help` or reading
975 the [build manual][13].
976 4. Multiplication was optimized by using **diagonal multiplication**, rather
977 than straight brute force.
978 5. The `locale_install.sh` script was fixed.
979 6. `dc` was given the ability to **use the environment variable
981 7. `dc` was also given the ability to **use the `-i` or `--interactive`**
983 8. Printing the prompt was fixed so that it did not print when it shouldn't.
984 9. Signal handling was fixed.
985 10. **Handling of `SIGTERM` and `SIGQUIT`** was fixed.
986 11. The **built-in functions `maxibase()`, `maxobase()`, and `maxscale()`** (the
987 commands `T`, `U`, `V` in `dc`, respectively) were added to allow scripts to
988 query for the max allowable values of those globals.
989 12. Some incompatibilities with POSIX were fixed.
991 In addition, this release is `2.0.0` for a big reason: the internal format for
992 numbers changed. They used to be a `char` array. Now, they are an array of
993 larger integers, packing more decimal digits into each integer. This has
994 delivered ***HUGE*** performance improvements, especially for multiplication,
997 This `bc` should now be the fastest `bc` available, but I may be wrong.
1001 This release contains a fix for a harmless bug (it is harmless in that it still
1002 works, but it just copies extra data) in the [`locale_install.sh`][12] script.
1006 This version contains fixes for the build on Arch Linux.
1010 This release removes the use of `local` in shell scripts because it's not POSIX
1011 shell-compatible, and also updates a man page that should have been updated a
1012 long time ago but was missed.
1016 This release contains some missing locale `*.msg` files.
1020 This release contains a few bug fixes and new French translations.
1024 This release contains a fix for a bug: use of uninitialized data. Such data was
1025 only used when outputting an error message, but I am striving for perfection. As
1026 Michelangelo said, "Trifles make perfection, and perfection is no trifle."
1030 This release contains fixes for OpenBSD.
1034 This release contains bug fixes for some rare bugs.
1038 This is a production release.
1040 There have been several changes since `1.1.0`:
1042 1. The build system had some changes.
1043 2. Locale support has been added. (Patches welcome for translations.)
1044 3. **The ability to turn `ibase`, `obase`, and `scale` into stacks** was added
1045 with the `-g` command-line option. (See the [`bc` manual][9] for more
1047 4. Support for compiling on Mac OSX out of the box was added.
1048 5. The extended math library got `t(x)`, `ceil(x)`, and some aliases.
1049 6. The extended math library also got `r2d(x)` (for converting from radians to
1050 degrees) and `d2r(x)` (for converting from degrees to radians). This is to
1051 allow using degrees with the standard library.
1052 7. Both calculators now accept numbers in **scientific notation**. See the
1053 [`bc` manual][9] and the [`dc` manual][10] for details.
1054 8. Both calculators can **output in either scientific or engineering
1055 notation**. See the [`bc` manual][9] and the [`dc` manual][10] for details.
1056 9. Some inefficiencies were removed.
1057 10. Some bugs were fixed.
1058 11. Some bugs in the extended library were fixed.
1059 12. Some defects from [Coverity Scan][11] were fixed.
1063 This release contains a fix to the build system that allows it to build on older
1064 versions of `glibc`.
1068 This release contains a fix for a bug in the test suite where `bc` tests and
1069 `dc` tests could not be run in parallel.
1073 This release has a fix for a history bug; the down arrow did not work.
1077 This release fixes a bug in the `1.1.0` build system. The source is exactly the
1080 The bug that was fixed was a failure to install if no `EXECSUFFIX` was used.
1084 This is a production release. However, many new features were added since `1.0`.
1086 1. **The build system has been changed** to use a custom, POSIX
1087 shell-compatible configure script ([`configure.sh`][6]) to generate a POSIX
1088 make-compatible `Makefile`, which means that `bc` and `dc` now build out of
1089 the box on any POSIX-compatible system.
1090 2. Out-of-memory and output errors now cause the `bc` to report the error,
1091 clean up, and die, rather than just reporting and trying to continue.
1092 3. **Strings and constants are now garbage collected** when possible.
1093 4. Signal handling and checking has been made more simple and more thorough.
1094 5. `BcGlobals` was refactored into `BcVm` and `BcVm` was made global. Some
1095 procedure names were changed to reflect its difference to everything else.
1096 6. Addition got a speed improvement.
1097 7. Some common code for addition and multiplication was refactored into its own
1099 8. A bug was removed where `dc` could have been selected, but the internal
1100 `#define` that returned `true` for a query about `dc` would not have
1102 9. Useless calls to `bc_num_zero()` were removed.
1103 10. **History support was added.** The history support is based off of a
1104 [UTF-8 aware fork][7] of [`linenoise`][8], which has been customized with
1105 `bc`'s own data structures and signal handling.
1106 11. Generating C source from the math library now removes tabs from the library,
1107 shrinking the size of the executable.
1108 12. The math library was shrunk.
1109 13. Error handling and reporting was improved.
1110 14. Reallocations were reduced by giving access to the request size for each
1112 15. **`abs()` (`b` command for `dc`) was added as a builtin.**
1113 16. Both calculators were tested on FreeBSD.
1114 17. Many obscure parse bugs were fixed.
1115 18. Markdown and man page manuals were added, and the man pages are installed by
1117 19. Executable size was reduced, though the added features probably made the
1118 executable end up bigger.
1119 20. **GNU-style array references were added as a supported feature.**
1120 21. Allocations were reduced.
1121 22. **New operators were added**: `$` (`$` for `dc`), `@` (`@` for `dc`), `@=`,
1122 `<<` (`H` for `dc`), `<<=`, `>>` (`h` for `dc`), and `>>=`. See the
1123 [`bc` manual][9] and the [`dc` manual][10] for more details.
1124 23. **An extended math library was added.** This library contains code that
1125 makes it so I can replace my desktop calculator with this `bc`. See the
1126 [`bc` manual][3] for more details.
1127 24. Support for all capital letters as numbers was added.
1128 25. **Support for GNU-style void functions was added.**
1129 26. A bug fix for improper handling of function parameters was added.
1130 27. Precedence for the or (`||`) operator was changed to match GNU `bc`.
1131 28. `dc` was given an explicit negation command.
1132 29. `dc` was changed to be able to handle strings in arrays.
1134 ## 1.1 Release Candidate 3
1136 This release is the eighth release candidate for 1.1, though it is the third
1137 release candidate meant as a general release candidate. The new code has not
1138 been tested as thoroughly as it should for release.
1140 ## 1.1 Release Candidate 2
1142 This release is the seventh release candidate for 1.1, though it is the second
1143 release candidate meant as a general release candidate. The new code has not
1144 been tested as thoroughly as it should for release.
1146 ## 1.1 FreeBSD Beta 5
1148 This release is the sixth release candidate for 1.1, though it is the fifth
1149 release candidate meant specifically to test if `bc` works on FreeBSD. The new
1150 code has not been tested as thoroughly as it should for release.
1152 ## 1.1 FreeBSD Beta 4
1154 This release is the fifth release candidate for 1.1, though it is the fourth
1155 release candidate meant specifically to test if `bc` works on FreeBSD. The new
1156 code has not been tested as thoroughly as it should for release.
1158 ## 1.1 FreeBSD Beta 3
1160 This release is the fourth release candidate for 1.1, though it is the third
1161 release candidate meant specifically to test if `bc` works on FreeBSD. The new
1162 code has not been tested as thoroughly as it should for release.
1164 ## 1.1 FreeBSD Beta 2
1166 This release is the third release candidate for 1.1, though it is the second
1167 release candidate meant specifically to test if `bc` works on FreeBSD. The new
1168 code has not been tested as thoroughly as it should for release.
1170 ## 1.1 FreeBSD Beta 1
1172 This release is the second release candidate for 1.1, though it is meant
1173 specifically to test if `bc` works on FreeBSD. The new code has not been tested as
1174 thoroughly as it should for release.
1176 ## 1.1 Release Candidate 1
1178 This is the first release candidate for 1.1. The new code has not been tested as
1179 thoroughly as it should for release.
1183 This is the first non-beta release. `bc` is ready for production use.
1185 As such, a lot has changed since 0.5.
1187 1. `dc` has been added. It has been tested even more thoroughly than `bc` was
1188 for `0.5`. It does not have the `!` command, and for security reasons, it
1189 never will, so it is complete.
1190 2. `bc` has been more thoroughly tested. An entire section of the test suite
1191 (for both programs) has been added to test for errors.
1192 3. A prompt (`>>> `) has been added for interactive mode, making it easier to
1193 see inputs and outputs.
1194 4. Interrupt handling has been improved, including elimination of race
1195 conditions (as much as possible).
1196 5. MinGW and [Windows Subsystem for Linux][1] support has been added (see
1197 [xstatic][2] for binaries).
1198 6. Memory leaks and errors have been eliminated (as far as ASan and Valgrind
1200 7. Crashes have been eliminated (as far as [afl][3] can tell).
1201 8. Karatsuba multiplication was added (and thoroughly) tested, speeding up
1202 multiplication and power by orders of magnitude.
1203 9. Performance was further enhanced by using a "divmod" function to reduce
1204 redundant divisions and by removing superfluous `memset()` calls.
1205 10. To switch between Karatsuba and `O(n^2)` multiplication, the config variable
1206 `BC_NUM_KARATSUBA_LEN` was added. It is set to a sane default, but the
1207 optimal number can be found with [`karatsuba.py`][4] (requires Python 3)
1208 and then configured through `make`.
1209 11. The random math test generator script was changed to Python 3 and improved.
1210 `bc` and `dc` have together been run through 30+ million random tests.
1211 12. All known math bugs have been fixed, including out of control memory
1212 allocations in `sine` and `cosine` (that was actually a parse bug), certain
1213 cases of infinite loop on square root, and slight inaccuracies (as much as
1214 possible; see the [README][5]) in transcendental functions.
1215 13. Parsing has been fixed as much as possible.
1216 14. Test coverage was improved to 94.8%. The only paths not covered are ones
1217 that happen when `malloc()` or `realloc()` fails.
1218 15. An extension to get the length of an array was added.
1219 16. The boolean not (`!`) had its precedence change to match negation.
1220 17. Data input was hardened.
1221 18. `bc` was made fully compliant with POSIX when the `-s` flag is used or
1222 `POSIXLY_CORRECT` is defined.
1223 19. Error handling was improved.
1224 20. `bc` now checks that files it is given are not directories.
1226 ## 1.0 Release Candidate 7
1228 This is the seventh release candidate for 1.0. It fixes a few bugs in 1.0
1229 Release Candidate 6.
1231 ## 1.0 Release Candidate 6
1233 This is the sixth release candidate for 1.0. It fixes a few bugs in 1.0 Release
1236 ## 1.0 Release Candidate 5
1238 This is the fifth release candidate for 1.0. It fixes a few bugs in 1.0 Release
1241 ## 1.0 Release Candidate 4
1243 This is the fourth release candidate for 1.0. It fixes a few bugs in 1.0 Release
1246 ## 1.0 Release Candidate 3
1248 This is the third release candidate for 1.0. It fixes a few bugs in 1.0 Release
1251 ## 1.0 Release Candidate 2
1253 This is the second release candidate for 1.0. It fixes a few bugs in 1.0 Release
1256 ## 1.0 Release Candidate 1
1258 This is the first Release Candidate for 1.0. `bc` is complete, with `dc`, but it
1263 This beta release completes more features, but it is still not complete nor
1264 tested as thoroughly as necessary.
1268 This beta release fixes a few bugs in 0.4.
1272 This is a beta release. It does not have the complete set of features, and it is
1273 not thoroughly tested.
1275 [1]: https://docs.microsoft.com/en-us/windows/wsl/install-win10
1276 [2]: https://pkg.musl.cc/bc/
1277 [3]: http://lcamtuf.coredump.cx/afl/
1278 [4]: ./scripts/karatsuba.py
1281 [7]: https://github.com/rain-1/linenoise-mob
1282 [8]: https://github.com/antirez/linenoise
1283 [9]: ./manuals/bc/A.1.md
1284 [10]: ./manuals/dc/A.1.md
1285 [11]: https://scan.coverity.com/projects/gavinhoward-bc
1286 [12]: ./scripts/locale_install.sh
1287 [13]: ./manuals/build.md
1288 [14]: https://github.com/stesser
1289 [15]: https://github.com/bugcrazy
1290 [16]: ./manuals/bc/A.1.md#extended-library
1291 [17]: https://github.com/skeeto/optparse
1292 [18]: https://www.deepl.com/translator
1293 [19]: ./manuals/benchmarks.md
1294 [20]: https://github.com/apjanke/ronn-ng
1295 [21]: https://pandoc.org/
1296 [22]: ./scripts/locale_uninstall.sh