aboutsummaryrefslogtreecommitdiffstats
path: root/configure
Commit message (Collapse)AuthorAgeLines
* Write backtrace to /var/lib/bitlbee/crash.log on SIGSEGVdequis2018-03-26-0/+27
| | | | | | | | | | | | | | | | | | | | | | Async-signal-safe code is very restricted (nothing that may call malloc indirectly), so this code tries its best to show meaningful stuff, but the output is still fairly raw. The contents of the log file are: - BITLBEE_VERSION, BITLBEE_CONFIGURE_ARGS - Backtrace as generated by backtrace()/backtrace_symbols_fd() - A small help text explaining how to get more useful symbol names - Memory maps (/proc/self/maps), which also mentions loaded plugins The backtrace() function is a GNU extension, /proc/ is a linux thing. Non-glibc platforms (such as musl) won't show anything, non-linux platforms will skip the memory maps when /proc/self/maps fails to open. I'd like to include timestamps, but I can't find a safe way to format them. Even turning raw unix timestamps to strings is hard. Fun stuff. I used the config directory because it's the only place we can be sure we can write to. The filename is hardcoded for the same reason there are no timestamps.
* configure: Add -Wformat -Werror=format-security to default CFLAGSdequis2018-03-19-2/+2
| | | | | | | | Matches debian build scripts. Also expands the condition to gcc or clang. I'd rather remove this, apparently it exists because some opensolaris compiler didn't have -Wall which seems absurd to me, but whatever.
* Remove old skype plugin. Use the skypeweb purple plugin instead.dequis2018-03-11-13/+0
| | | | | | | | | | RIP This plugin relied on the skype linux client, which doesn't work anymore. During login it just does exit(0) for mysterious reasons. I guess that's the server trying to say that it's not supported. This is officially dead now after a few years of kinda-almost-working.
* Remove MSN. Use the skypeweb purple plugin instead.dequis2018-01-10-11/+0
| | | | | | | | | | | | | | RIP As per http://ismsndeadyet.com/ all versions up to MSNP22 died a few months ago. We had a MSNP21 implementation, bought us two extra years. Implementing MSNP24 is technically possible but also pointless given skypeweb, and the authentication requires some fairly messed up crypto and legacy old-skype-protocol servers. For a long time I tried to reverse a potentially simpler method, and got fairly close, but never completed that. I haven't done any attempts to continue it in the last year, so I'm fine with giving up at this point.
* Add datadir to pkgconfig file and config.hdevelopArtem Savkov2017-08-31-0/+2
| | | | Export datadir through pkgconfig and config.h for later use by plugins.
* configure: Don't require python if docs are already builtdequis2017-07-09-1/+1
|
* cygwin: add -no-undefined on our side to fix issues with mac OSdequis2017-05-29-1/+1
| | | | | | | | | | So cygwin requires -no-undefined and it was supposed to be harmless on platforms other than windows, but mac OS doesn't like it. Instead of adding platform checks to every plugin, i'm going to revert the changes that added -no-undefined to them, and just do what I should have done since the beginning: use the cygwin-specific parts of the pkg-config file to add -no-undefined right next to -lbitlbee
* cygwin: Export a "libbitlbee.dll.a" to allow plugins to link to itdequis2017-05-13-1/+13
|
* configure: respect autotools-like verbose flags for makeTim Harder2017-03-12-0/+13
| | | | | | In other words, running $(make V=1) will force a verbose build while $(make V=0) will force a silent build regardless of the verbose option set by the configure script.
* Add --verbose configure option to control verbose build outputTim Harder2017-03-12-0/+9
| | | | Defaults to disabled to maintain the status quo.
* Include debug symbols in non-debug builds, disable stripping by defaultdequis2016-11-28-2/+2
| | | | | | | | | | | | | | | | With this commit, the difference between debug and non-debug builds is mainly the optimization level and -DDEBUG (which isn't used much) In other words: --debug=0 == CFLAGS="-g -O2 -fno-strict-aliasing" --debug=1 == CFLAGS="-g3 -DDEBUG -O0" And --strip=1 can be used to get rid of the debug symbols. This is closer to the default behavior of autotools. Should have done this forever ago, like back when bitlbee had bugs (lol)
* Remove yahoo (the old protocol). Use funyahoo++ instead.dequis2016-11-19-11/+0
| | | | | | RIP The previous commit already handled the part of telling users.
* configure: improve error message on missing pkg-configdequis2016-07-25-1/+7
|
* configure: add 'arch' variable again, needed for portabilitydequis2016-05-25-0/+2
| | | | | | It was removed by 2e78f75, but it only needed to remove the variable that was written to config.h, not the one that was used internally by configure.
* Remove the ARCH / CPU defines.Jelmer Vernooij2016-05-15-7/+0
| | | | | | | These only reflect on what arch/cpu bitlbee was built, not on which it is running. This makes the Debian package unreproducible. See e.g. https://tests.reproducible-builds.org/rb-pkg/testing/i386/bitlbee.html
* ldap authentication backendDennis Kaarsemaker2016-03-25-0/+13
| | | | | | We only support the openldap scheme for now, with users that are posixAccounts. Moreover, as the plugin cannot be configured directly, you must configure libldap correctly in /etc/openldap/ldap.conf
* Linux pam authentication backendDennis Kaarsemaker2016-03-25-0/+15
| | | | This backend authenticates users against pam.
* Authentication: scaffolding for multiple authentication backendsDennis Kaarsemaker2016-03-25-0/+5
| | | | | | | | | | | | | | | | | | | | | | | Instead of always putting users passwords in XML files, allow site admins to configure a different authentication method to integrate authentication with other systems. This doesn't add any authentication backends yet, merely the scaffolding. Notably: - Password checking and loading/removing from storage has been decoupled. A new auth_check_pass function is used to check passwords. It does check against the configured storage first, but will handle the authentication backends as well. The XML storage merely signals that a user's password should be checked using an authentication backend. - If unknown-to-bitlbee users identify using an authentication backend, they are automatically registered. - If an authentication backend is used, that fact is stored in the XML file, the password is not. Passwords are also stored unencrypted in this case, as the password used to encrypt them can change underneath us. - configure and Makefile changes for the backend objects
* Show ./configure args in bitlbee -V, config.h and Makefile.settingsdequis2015-11-21-0/+7
|
* Add missing space in configureAndré Glüpker2015-10-08-4/+4
|
* Improved cross compiler support.Guillermo A. Amaral2015-08-31-4/+27
| | | | | | | | | * Added a configure option for sysroot. * Prevent pkgconfig environment variables to be clobbered. * Prevent PATH environment variable pollution. * Backward compatible. Signed-off-by: Guillermo A. Amaral <g@maral.me>
* configure: use pkg-config for libotrdequis2015-08-07-22/+14
|
* Replace `backticks` with $(something better) in configure script.Wilmer van der Gaast2015-07-29-26/+26
|
* configure: allow specifying location of python executabledequis2015-06-04-1/+7
| | | | | Because openbsd/netbsd don't have a 'python' in the PATH unless the user creates the symlink explicitly.
* configure: bsd sed doesn't support \+ without -r, use * insteaddequis2015-05-31-1/+1
| | | | | | | | | Tested in openbsd 5.6, which does have sed -r (unlike Mac OS X) but it needs -r to recognize + at all. The previous commit did fix enough to make it 'work' with the git describe output as version string, the only problem was that the sed didn't match in bsds and mac os x.
* configure: 'sed -r' doesn't work in mac os xdequis2015-05-31-2/+2
|
* configure: handle 'git describe' failures when no tags are reachabledequis2015-05-31-2/+7
| | | | Just make up some fake describe output based on the real version
* configure: workaround for branch names with slashes in them...dequis2015-05-30-1/+1
|
* Include prebuilt skyped manpage, remove asciidoc dependencydequis2015-05-17-10/+0
|
* configure: Replace xmlto/xsltproc checks with a python>=2.5 checkdequis2015-05-17-8/+5
| | | | | | | | It actually checks that xml.etree.ElementTree can be imported, which is good enough. The script works fine in 2.5 (tested with centos 6), and doesn't work in 2.4 but we don't support centos 5 anymore so no problem there.
* configure: fix spoofed BITLBEE_VERSION env variable handlingdequis2015-05-06-0/+7
|
* configure: change version number format, add --dump-versiondequis2015-05-03-25/+28
|
* configure: Indentation fixesdequis2015-02-23-21/+21
|
* Add --doc= configure option to disable helpfile generationdequis2015-02-23-8/+27
| | | | | | | | | | | | | It will get disabled automatically if the deps (xmlto and xsltproc) are missing Also added checks for asciidoc (a2x) for the skype plugin, which needs it for the skyped man page, and will also get that disabled if --doc=0 is passed or if asciidoc isn't installed. This should keep those ugly deps under control for now. I'd like to replace them at some point with something less dumb.
* Use glib's GChecksum for md5/sha1dequis2015-01-31-1/+1
| | | | | | | | | | | | | | | | | | | This changes behavior slightly: - md5_init()/sha1_init() allocate a GChecksum - md5_finish()/sha1_finish() close and free() it - md5_digest_keep() was added (no sha1 equivalent needed) And yes, glib has this concept of "closing" the GChecksum, which means it can't be used anymore after g_checksum_get_digest(). jabber_cache_add() actually seems to need to do that to generate some random-ish values, so i kept that working by adding a md5_digest_keep() function that copies the GChecksum before it gets closed GChecksum was introduced in glib 2.16, so the configure script version was bumped. We were already depending on glib 2.16 accidentally (some post-3.2.2 code uses GHashTableIter)
* Add git specific stuff!dequis2015-01-25-0/+6
| | | | | | | | - A few build system tweaks - A fancy README.md - s/.bzrignore/.gitignore/ - .travis.yml for travis-ci integration - doc/git-bzr-rev-map for historical purposes
* Fix compiler warnings on Cygwin and Mac OS X.Jason Copenhaver2015-01-16-0/+8
| | | | | | | | * Don't use PIE/PIC on Cygwin/Darwin unless specified as these platforms don't support it. * Cleanup warnings for 'make check' build. * Fix the type issue for getsockopt calls. * Fix enum warnings in Yahoo libs on Mac OS X.
* configure: --asan=1 parameter for AddressSanitizerdequis2014-11-24-1/+14
| | | | | | | | Requires gcc >=4.8 or clang >=3.1 AddressSanitizer (ASan) is a fast memory error detector. See also: https://code.google.com/p/address-sanitizer/wiki/AddressSanitizer
* RIP native win32 support (use cygwin instead)dequis2014-09-27-3/+1
| | | | It has been broken for a very long time and nobody cared about it.
* Add -lgcrypt when building with OTR built-indequis2014-09-27-1/+1
| | | | Needed when building without gnutls (which already depends on gcrypt)
* fix latest HAS_NAMESER code on cygwin, mac and openbsdJason Copenhaver2014-07-24-27/+13
|
* srv_lookup: Portability fixes, handle compressed responsesjcopenha2014-07-24-1/+112
| | | | | | | | | | srv_lookup works on cygwin and openbsd now. Provide ns_initparse, friends, and types where they aren't provided by platform. Use dn_expandname instead of custom parser so compressed DNS responses are handled correctly.
* Allow LDFLAGS to be provided in the environment and consistently use $(LDFLAGS)Jelmer Vernooij2014-07-20-0/+2
| | | | | when linking (not partially linking).
* Fix version in pkg-config file, also change bzr version format slightlydequis2014-06-26-2/+6
| | | | | | | | | Old bzr version format: bzr-1234 New bzr version format: 3.2.1-bzr-1234 Version spoofing with the BITLBEE_VERSION also changed slightly - it does not need quotes around the version number anymore for it to compile correctly.
* configure: Let the user enable both built-in protocols and purpledequis2014-06-26-10/+15
| | | | | | | | | | There's no real risk in enabling these, since protocols/purple/purple.c already avoids adding duplicate protocols in purple_initmodule(): /* If we already have this one (as a native module), don't add a libpurple duplicate. */ if( find_protocol( prot->info->id ) ) continue;
* configure: Add spaces after <<EOF heredoc markers for claritydequis2014-06-26-9/+9
| | | | I mean, they confuse the vim bash syntax highlighter
* pkgconfig: Plugindir fix for bitlbee.pc (Ticket #1085)jgeboski2014-02-27-0/+2
|
* Update json-parser code to git rev 11a80f389769d1f66ca7fbe64ad89c82d3ba4ba9.Wilmer van der Gaast2014-02-13-1/+1
| | | | | | Few BitlBee-specific diffs now. Annoyingly need to link to libm now for the use of the function pow() since the lib now does its own number parsing...
* Fix building of debug binaries (partially broken by changeset:devel,253).Wilmer van der Gaast2013-12-20-2/+1
|
* Ugly: Bogus dependency on libotr2-dev, but adding a warning to the configureWilmer van der Gaast2013-12-07-0/+6
| | | | | | | | | script that this actually won't work. This because I can't figure out how to make a source dependency conditional on whether a certain binary packages is going to be built. It's kind of okay since libotr2-dev is unavailable on new systems anyway.
generated by cgit v1.2.3 (git 2.25.1) at 2025-08-14 08:36:16 +0000