| Commit message (Collapse) | Author | Age | Files | Lines |
|
|
|
| |
Obscure PHP issue with mixed variable types.
|
| |
|
| |
|
|
|
|
|
|
| |
Do not call conf_mount_ro just before call write_config(),
write_config() calls conf_mount_(rw|ro), and if filesystem is already
mounted in rw state, it'll only increase the refcounter
|
|
|
|
| |
Destination port fields now behave as required.
|
|
|
|
|
| |
Javascript logic changged for source port display
"Advanced" button now has toggle acton and its state survives a protocol change.
|
|
|
|
|
|
|
| |
Notes:
The addGlobal() method is applicable to a form, not a section
'hidden' elements should not specify a label
Form_Groups with no input elements currently break the framework. Use Form_StaticText instead.
|
|\ |
|
| | |
|
| | |
|
| |
| |
| |
| | |
This work was reverted then re-applied without the (accidental) tab/space screw-up
|
| |
| |
| |
| |
| | |
Accidentally broke spaces/tabs
This reverts commit 910b77ad1682740e481df29fc2d9c528a13c2da4.
|
| | |
|
| | |
|
| | |
|
| |
| |
| |
| | |
See also: commit:5372d26d9d25d751d16865ed9d46869d3b0ec5e1
|
|\ \
| | |
| | |
| | | |
itshari-bootstrap
|
| | | |
|
| |\ \ |
|
| | | | |
|
| |\ \ \ |
|
| | | | | |
|
| | | | | |
|
| | | | | |
|
| | | | | |
|
|\ \ \ \ \ |
|
| | | | | |
| | | | | |
| | | | | |
| | | | | | |
Cleaner version of https://github.com/pfsense/pfsense/pull/1846
|
|\ \ \ \ \ \ |
|
| |/ / / / /
| | | | | |
| | | | | |
| | | | | | |
to improve the English.
Resubmit of PR https://github.com/pfsense/pfsense/pull/1847
|
|\ \ \ \ \ \ |
|
| |/ / / / /
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | | |
I believe these should reference the [0] array element of pptp_local and pptp_remote. The code in interfaces.php only presents the [0] (first) array element to the user for editing.
The current code also does not preserve any additional pptp_local, pptp_subnet or pptp_remote entries in the coma-separated list in the config. That is redmine bug #4568.
I am currently integrating the fix for Redmine #4568 Preserve MLPPP settings when saving interface settings. The fix for that bug will add more code to preserve additional pptp_local, pptp_subnet or pptp_remote entries.
For now it would be good to have this little fix to the existing code - then the fix for redmine bug #4568 will sit nicely on top of it.
|
| | | | | | |
|
| | | | | | |
|
| | | | | |
| | | | | |
| | | | | |
| | | | | | |
Direction value was not being saved in config
|
| | | | | |
| | | | | |
| | | | | |
| | | | | | |
hideInput -> hideCheckbox
|
| | | | | |
| | | | | |
| | | | | |
| | | | | | |
serversdisabled list filtered to remove blank entry
|
| | | | | |
| | | | | |
| | | | | |
| | | | | | |
JS revised to allow subnet bits on CARP and IP Alias
|
| | | | | |
| | | | | |
| | | | | |
| | | | | | |
Resolver GET/POST error
|
|/ / / / /
| | | | |
| | | | |
| | | | | |
JS added to provide DHCP preset functionality
|
|\ \ \ \ \ |
|
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | | |
DHCP6
This has been broken ever since commited in 420aa48
As noted on https://redmine.pfsense.org/issues/3670 - the get_interface_ipv6() function in /etc/inc/interfaces.php incorrectly identifies the interface as the physical hardware interface. As a result, no global IPv6 address can be found (empty $ifcfgipv6) - https://redmine.pfsense.org/issues/3556
|
| | | | | | |
|
| | | | | |
| | | | | |
| | | | | |
| | | | | | |
Credit for idea to Phil Davis
|
|\ \ \ \ \ \ |
|
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | | |
The notices are currently keyed by the Unix time stamp second. If
file_notice() is called more than once in the same second, then the
previous notice is overwritten. Only the last notice in any second
actually ends up in /tmp/notices and thus on the webGUI flashing
display.
The $queuekey value seems to be used to work out the actual time of the
event for display, so it is a bit tricky to change that to a
finer-grained key - callers of these functions are expecting an array
key that is the Unix time in seconds, so all calls to this stuff would
have to be examined and adjusted...
The workaround here is to increment the key if the existing key is
already in use. This allows all notices in the same second to be saved,
but the time of each one will be an incrementing seconds counter, even
though the events all happened in the same second. Maybe that is a
reasonable/practical workaround for now? Given that there should not be
too many notices filed at the same second.
This is a resubmit of PR #1782 integrated with the current master.
|
| | | | | | | |
|
| | | | | | | |
|
| | | | | | | |
|
| | | | | | | |
|
| | | | | | | |
|