| Commit message (Collapse) | Author | Age | Lines |
|---|
| | |
|
| |\
| |
| | |
fix(nightmode): Respect user provided configuration for night mode if provided
|
| | |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
If user sets the cookie or URL query parameter for nightMode, we will
respect that. Note that it is a bit cumbersome to _clear_ this
configuration (afaik?) if it is set.
This is also the case if the user toggles nightmode directly in Gondul
(`n`).
However, this last configuration is not persisted - so on a page reload,
this setting is removed (no need to clear cookies/query params), and
the system preference is again followed.
|
| | | |
|
| |/ |
|
| |
|
|
|
|
| |
Uses the proposal here: https://drafts.csswg.org/mediaqueries-5/#descdef-media-prefers-color-scheme
Seems to be decently supported: https://caniuse.com/?search=prefers-color-scheme
|
| | |
|
| | |
|
| |
|
|
| |
support ipv6 in dhcp
|
| | |
|
| | |
|
| | |
|
| | |
|
| | |
|
| |
|
|
|
|
| |
Caused a problem locally after I had accidentally made tags into an object
instead of an array during testing, but this CAN happen if someone writes {}
instead of [], so better to anticipate it.
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Sort of.
It's two issues at once, related:
1. API: The problem here was that the old way of updating tags simply didn't
use actual JSON, but just sent "'foo','bar','baz'" as a text string. This
seems dumb so I've made it send actual JSON now - just like the "placement"
field. This meant updating the API. And it's not pretty, but it works.
2. Front: I've simplified nms-types a bit (hopefully to provide to simple
methods: either get/set the raw value, or get/set strings. There was a bug
where we sent text-encoded json instead of real json, and the reason
was a confusion between when we're dealing with JSON and when we're
dealing with strings. Now we are explicit.
This makes the nmsEditRow-thing slightly uglier, but it needs to be fixed
properly either way. In the future, we should provide renderers and editors
based on types, e.g.: Placement-editor could start as a general-purpose
JSON-editor, and the "switch reference" should be a drop-down.... etc.
Fixes #202
|
| |
|
|
| |
It creates an extra margin that serves no purpose but to take up space.
|
| | |
|
| | |
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Fixes #183
References #182
It's now possible to add new networks and switches, and it's also possible to
edit networks, but since I haven't exposed it in the UI yet, I can't really
close the bug.
This is also a major step in stripping nms-info-box apart.
Next up is listing existing networks, possibly linking to them. I think that
might be a decent first-step towards a new core nms-info-box since it doesn't
alreay exist.
|
| |
|
|
|
|
|
| |
Also some other random drive-by fixes :D
Fixes #180
References #181
|
| |
|
|
|
|
|
| |
Starting to look seeeexy.
Also killed off some redundant stuff that should be covered by templates
instead now.
|
| | |
|
| |
|
|
| |
Still needs to be tidied up, but we're getting there.
|
| |
|
|
|
|
|
|
|
|
|
| |
The integration is a hack, but a first step. I will eventually remove all the
code in nms-info-box, but I needed a simple test.
This also demonstrates how to add a new switch.... simply hit edit, then change
the name. It aint pretty, and shouldn't work like it does today, but it's
a decent example.
Also, had to fix the backend again now that I actually tested the write-api :D
|
| | |
|
| |
|
|
|
|
| |
Turns out the style-diff was because the old implementation with hard-coded
HTML had indentation == white space, which rendered as white-space between the
elements.... This SHOULD be a CSS fix, but in the meanwhile....
|
| |
|
|
|
| |
The timeout makes things about 1000x more complicated, but adds a really nice
effect too.
|
| | |
|
| | |
|
| |
|
|
| |
Too late, need sleeeezpzpeepzor
|
| |
|
|
|
| |
This is still a bit of a mess, but I'm establishing a new way of doing things,
it'll be awesome, I swear...
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Still not very visible, but the nmsModSwitch() class is starting to look
very interesting.
This introduces a big thing: A type system. So far it is "only" used to
provide generic validation-services for all fields that will be edited, but
the scope can extend further into other uses.
The implementation as it is does shallow validation for IP addresses,
JSON-input (tags,placement), durations/intervals (poll frequency), sysname
references (checks that distro_name is a valid system), and networks
(verifying that mgmt_vlan / traffic_vlan exists). This is expected to
improve over time, and should eventually lead to things like "if it doesn't
exist, then please make it."
As the big comment in nms-ui-switch.js now suggests: This is still just a
proof of concept and while the fundamental concepts of nmsBox seems to work
well, I still feel like I have a lot to iron out in nmsModSwitch, which
really should be generic for networks and switches at the very least.
But we're getting there.
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
nmsBox is a new class for generic HTML-based elements, and will eventually be
used to replace nms-info-box.
The idea is simple: a generic way to deal with containers that are usually
represented in HTML too.
To test it, I've re-implemented the GUI for the oplog. This rewrite didn't
really utilize the new benefits of the framework, but was a small step.
I've also added nms-ui-switch, which isn't exposed anywhere yet and isn't
done, but is a good start.
nms-ui-switch will be the new way to add and edit switches in the future, it
will probably be made more generic over time and thus can be reused for
networks too.
Note how x = nmsUiSwitch() will allow you to do x.row["community"].value
to both get and set the value, and setting will visually update and run
any verifier that will be relevant, and alert the parent. This can then
be used for simple stuff like json-verification, but also for stuff like
auto-complete or whatnot. God only knows.
Obviously I will continue to work on this over the next few days...
|
| |
|
|
| |
Fixes #71
|
| |\
| |
| | |
Cleanup of latencychart
|
| | | |
|
| | | |
|
| | | |
|
| |\ \
| | |
| | | |
Adding CPU map to key 9, moving disco to 0
|
| | | | |
|
| | | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
Fixes #150
Fixe #88
Closes #3
Probably lots of bugs introduced, will fix ASAP.
|
| | | | |
|
| | | | |
|
| | | | |
|
| | | | |
|
| | | |
| | |
| | |
| | |
| | | |
Seeing the day helps immensly, and the rather ugly translate-code avoids
overlap
|
| | |/
|/| |
|
| | | |
|
| |\ \
| | |
| | | |
Set Client ports description to "None configured" if no configured client ports
|
