Closed gmoro closed 3 years ago
Two of those commits are from me, please fix the authorship. I just copied and paste this, do you really need the authorship?
The approach with
()
has some issues:* There can be multiple hidden networks
Forgot to implement that
* `()` is a valid ESSID
Thats the main problem
* Hidden networks don't broadcast, so they might not show up at all
Hidden networks don't broadcast the SSID but they are always visible for iwlist, with an empty SSID I thought that was not a good enough solution, but now we have something to go from.
So instead I suggest to always add a special entry to enter a SSID manually.
dialogs
supports keys and values in dialogs, so you can use that to distinguish between the ESSIDs and the special option.
I did try to do that initially, but failed, do you have a good example of doing this special entries?
Two of those commits are from me, please fix the authorship. I just copied and paste this,
Copy + paste the entire diff and commit message separately? If you use git cherry-pick
or just work on the branch, it'll keep them with proper author and date.
do you really need the authorship?
Yes please.
The approach with
()
has some issues:* There can be multiple hidden networks
Forgot to implement that
* `()` is a valid ESSID
Thats the main problem
* Hidden networks don't broadcast, so they might not show up at all
Hidden networks don't broadcast the SSID but they are always visible for iwlist, with an empty SSID
I read up on that and indeed, they still send beacons regularly.
I thought that was not a good enough solution, but now we have something to go from.
So instead I suggest to always add a special entry to enter a SSID manually.
dialogs
supports keys and values in dialogs, so you can use that to distinguish between the ESSIDs and the special option.I did try to do that initially, but failed, do you have a good example of doing this special entries?
dialog --no-tags --menu "Choose wisely" 0 0 0 "SSID=SSID 1" "SSID 1" "SSID=SSID 2" "SSID 2" "manual" "Enter SSID manually"
By the way, I need to retab the whole thing, it's a mess of mixed tabs and spaces, I guess should wait until the next release, what do you think?
By the way, I need to retab the whole thing, it's a mess of mixed tabs and spaces, I guess should wait until the next release, what do you think?
IMO that can be done at any point, but in a separate PR. It can be mentioned in the .git-blame-ignore-revs
file
LGTM code wise.
Now
raspberrywifi_get_wlan_networks
never fails, because even whenip link
oriwlist
fail, it has themanual
entry which makes$list
non-empty. Maybe that should be caught explicitly? Though I'm not sure how
I guess that makes the previous behaviour wrong then, because an empty list is a valid thing, there might be no networks in range, we should really start checking return codes for the commands, only way to guarantee
Introduce wifi hidden SSID supports Improve menu appearance fixing bsc#1177188