aboutsummaryrefslogtreecommitdiff
Commit message (Collapse)AuthorAgeFilesLines
* networking: #include <netinet/in.h>Runxi Yu2024-07-278-0/+8
| | | | | IPPROTO_TCP, etc. seem to be defined in netinet/in.h on non-Linux systems, including OpenBSD.
* NickServ FIX command, for cleaning up remnants of problems caused by bugsTest_User2024-07-261-7/+104
|
* CI: Use new example configsRunxi Yu2024-07-264-8/+8
|
* Makefile: Clarify error messages a bitRunxi Yu2024-07-261-6/+5
|
* Makefile: $(error) adds a final period so don't specify oneRunxi Yu2024-07-261-1/+1
|
* Add an .editorconfigRunxi Yu2024-07-261-0/+12
|
* Fix some member ID-related bugsTest_User2024-07-252-16/+92
|
* Makefile and some option changes, and some fixesTest_User2024-07-256-196/+138
|
* InspIRCd v4: Fix erroneous protocol_specific[INSPIRCD3_PROTOCOL]Runxi Yu2024-07-251-2/+2
| | | | | | | | | Commit 86344230a6 fixed some memory leaks in the InspIRCd v3 protocol, and the changes were also ported to the InspIRCd v4 protocol, which erroneously introduced users[n]->protocol_specific[INSPIRCD3_PROTOCOL] into inspircd4.c. Although there are no functional differences, this causes the build to fail when users only enable INSPIRCD4_PROTOCOL in their build options.
* Typo fix: s/seperate/separate/gRunxi Yu2024-07-256-6/+6
|
* Toggleable IPv4 supportTest_User2024-07-244-4/+35
|
* Fix some issues in inspircd3 protocol support, and the parts of inspircd4 ↵Test_User2024-07-243-11/+36
| | | | protocol copied from it
* InspIRCd v4: Fix loop condition in FJOIN memid resetRunxi Yu2024-07-251-1/+1
| | | | | | | | | | | | There was a typo in the InspIRCd v3 code, in the FJOIN handler: The loop condition said n > 0, while n is never changed in the loop. It turns out the the loop condition should have been x > 0. This was fixed for InspIRCd v3 in commit 85716714, but it was not ported to the InspIRCd v4 protocol support code because the v4 code was copied from v3 a while ago. This commit fixes it.
* README: Move project links up, and add man.sr.htRunxi Yu2024-07-251-6/+6
|
* sockaddr is verysane™Test_User2024-07-247-35/+46
|
* YayTest_User2024-07-241-0/+2
|
* Using the proper defines is importantTest_User2024-07-241-2/+2
|
* Fix some bugsTest_User2024-07-242-4/+4
|
* Possibly broken incoming IPv6 supportTest_User2024-07-243-29/+43
|
* Fix cross-protocol netsplit propagationsTest_User2024-07-249-12/+79
|
* README: RTLD_GLOBAL should be in backticksRunxi Yu2024-07-241-1/+1
|
* InspIRCd v4: Don't send or expect 0 hopcount in initial SERVERRunxi Yu2024-07-241-7/+7
| | | | | | | | The SERVER message for directly-linked servers no longer contains the hopcount field. - :36D SERVER irc2.example.com password 0 36D :Example Server + :36D SERVER irc2.example.com password 36D :Example Server
* InspIRCd v4: Support real_username (well, faked)Runxi Yu2024-07-241-10/+16
| | | | | | | | In the 1206 protocol, the UID message now has an extra field for the real username of the user which is being introduced. We don't track this and this is unlikely to be useful for other protocols, so we just ignore it when receiving UID from an InspIRCd 1206 server, and we send the ident instead of the real_username when sending to a 1206 server.
* Initial InspIRCd v4/1206 protocol supportRunxi Yu2024-07-245-1/+2360
|
* This is what was intendedTest_User2024-07-241-1/+1
|
* Probable outgoing IPv6 supportTest_User2024-07-249-40/+84
|
* Fix gnutls_bufferedTest_User2024-07-241-9/+4
|
* haxstring: Add parenthesis to macro argumentsRunxi Yu2024-07-241-2/+2
| | | | | | | | | | | WRITES(fd, *s) where s is a pointer to a struct string will not compile, as the macro expands to write(fd, *s.data, *s.len), which it expects s to be a struct string and attempts to use s.data and s.len as pointers. This is, of course, erroneous. The correct expansion is write(fd, (*s).data, (*s).len); while write(fd, s->data, s->len) is desirable, it is not achievable with a simple macro expansion. In any case, the parenthesis shall be added.
* v4 rejects this otherwise currently, and the reason is fake here anywaysTest_User2024-07-231-1/+1
|
* Fix semaphore codeTest_User2024-07-231-0/+1
|
* Fix SERVER introductionsTest_User2024-07-231-2/+2
|
* CI: Disable Fedora too because runxiyu regularly builds it on FedoraRunxi Yu2024-07-241-5/+6
|
* CI: Fix again, because apparently empty list items are prohibitedRunxi Yu2024-07-241-5/+6
|
* CI: Fix last commitRunxi Yu2024-07-244-9/+9
| | | | | | Apparently it didn't really work because setting the branch list to empty causes it to always run, ignoring the branch. So the current solution is just to comment out the build commands.
* CI: Don't run Debian for now.Runxi Yu2024-07-241-1/+1
| | | | | | The Debian target is extremely slow and probably takes qutie a bit of Codeberg's CI resources. Test_User uses Debian anyway so there's not much point in running the CI.
* InspIRCd v3 raw command for HaxServTest_User2024-07-232-2/+24
|
* Fix InspIRCd v3 PING handlerTest_User2024-07-231-5/+1
|
* InspIRCd v3 supportTest_User2024-07-234-136/+523
|
* Bit of improvement on error handling for networkingTest_User2024-07-225-1/+33
|
* File descriptionsTest_User2024-07-2237-26/+51
|
* Configurable services db pathTest_User2024-07-223-3/+11
|
* CI: Install missing packagesRunxi Yu2024-07-224-4/+4
|
* CI: Use misc.git's configsRunxi Yu2024-07-224-5/+9
|
* CI: Accept the "runxiyu" branch tooRunxi Yu2024-07-224-4/+4
|
* Minimal support for non-semaphore-supporting systemsTest_User2024-07-192-1/+37
|
* Partial nickname enforcement implementedTest_User2024-07-1812-53/+154
|
* Fix REGISTER, add error messagesTest_User2024-07-071-2/+13
|
* UNGROUPTest_User2024-06-261-0/+59
|
* ADDCERT/DELCERTTest_User2024-06-261-5/+108
|
* LISTTest_User2024-06-252-3/+56
|
generated by cgit v1.2.3 (git 2.39.1) at 2024-09-20 02:26:53 +0000