Argolance wrote:I didn't see, directly from Frisbee interface
(while it is obvious using BK/dougal GUIs) how to:
- - give the preference to a wireless network when several available?
- load another driver if the "wrong" one is automatically loaded at startup or when the Windows one is required (using ndiswrapper)?
I don't think Jemimah addressed either of those issues. The preference seems like something I can try to add to frisbee. But the driver item is best left to SNS. The selection would be a one-time change. With the new frisbee, there is no problem switching among the network managers. The betas seemed to assume Frisbee is exclusive.
Now, back to the internationalization: I have done more fine tuning of the GUI windows to add more of Argolance's recommendations, although sometimes in a different way, particularly, moving the Exit button outside of the tabs (so it is active when wireless is greyed out), shortening of more button labels and re-organizing some of the buttons to put only 2 per line and be pleasing to the eye.
I attach a test version of what I hope to release, as well as screenies of the 4 windows. Note that the Exit button is gone from the secondary windows, requiring that they be closed with the X button. Argolance and anyone, please report any remaining issues/problems, so I can fix them right away. Thanks.
Richard
UPDATE 2/12/2012: Re-uploaded frisbee (after download #12), which is the same as before except for 3 fixes to reduce or eliminate annoying wpa_supplicant messages when there is no wifi hardware installed. They also fix a hang of frisbee while it waits for wpa_supplicant to time out (if it ever does) when there is no wifi interface to be found.
EDIT: I re-loaded frisbee-1.0 after download #13, to fix a potential problem for internationalization. A test was made in "connection-check" for the text, "Connected to", which is only part of the message being tested. If a translation put the interface ID before the translation of "Connected to", the test would fail, resulting in an inappropriate message that wpa_supplicant failed in some way. I added true/false return status from the current-status function, true meaning "connected". Therefore, apparently, that message, as well as "IP Address: ...", was not internationalized; now they are.
Please verify I did no harm, by ensuring you can connect to a wireless network. Thanks.
Richard
UPDATE 2/13/2013: Re-uploaded again, to add one more internationalized text string and change 2 icons, per Argolance's recommendations.
UPDATE 2/19/2013: Re-uploaded 20130213 without the .pot file and without the dependency on wag_profiles. No functional changes.
UPDATE 2/20/2013: Uploaded version with corrections for the eval_gettext function, to restore connection status and other text involving variables. Reported by peebee -- thank you.
EDIT: Re-uploaded 20130220 with further corrections: "Manage Saved Networks" "SSID:" initialization to avoid residue and change "Reset DHCP" to "Restart DHCP", to be consistent with "Restart Networks".
UPDATE 2/27/2013: Uploaded version 20130227 to reduce the size of the ever-growing wpa_supplicant log. The Diagnostics dialog now has a checkbox to enable debug-level logging, defaulted to unchecked. Previous frisbees always log at the debug (-d) level, creating a large and growing file in /tmp (which is not infinite in capacity). That would be a "showstopper" for long-running puppy sessions, thus this last-minute fix. Also eliminated output of a few remaining wpa_supplicant warning messages.
EDIT: Uploaded 20130228 after download #4 to add compatibility with slacko (in the pinstall script), but to improve a test. There is no functional difference from 20130227, which is only vulnerable if, unlikely, there are multiple matches for "/usr/local/lib/X11/mini-icons/networkdead-eth*.*"
EDIT: I re-uploaded the last 20130228 version of frisbee, to reset the download count, in the hope that earlier downloaders will want to ensure they have the latest. I had to make incremental updates that downloaders #5 & #6 do not have.
UPDATE 3/4/2013: Uploaded 20130304 version to correct a failure found in testing on racy and wary, reported by mavrothal and peebee. The problem was an "-f" option passed to wpa_supplicant that is now being rejected, causing wpa_supplicant to not start, at all. The mystery is that "-f" seemed to work in precise, although it apparently does no longer. Please test this on precise and racy soon, so I can offer it for beta2. Thank you for focusing my attention on this issue.