GitLab

Due to commit before this I also noticed non-nodes were being put into
the database.
There were over a 100 in my database alone.
These are most likely non-node devices and will never be able to be
mapped.
Initially, I opted for a "save and recheck" style of polling, in the
hopes that something was just missed and it would get picked up later.
This is not the case and over time I think can see what is happening.

added a check in get-map-info.php to try and skip these devices if we
reach one, and due to that change, the numbers of shown and not show
node and links at the bottom right of the map is more accuate now (I
hope).

73

Some nodes returned a blank lat,lon and grid_square differently than
others that were also blank.

This was causing issues when writing to the SQL database.
get-map-info would throw this error when it would happen:
"Could not send to SQL. Incorrect double value: '' for column 'lat' at
row 1"

Fix:
Any blank lat || lon now get converted to a single "0".
Any blank grid_square now gets converted to a NULL.

it used /bin/sh before and for some reason just recently developed a
problem running get-map-info on my RPi.
Using /bin/sh caused an SQL error, /bin/bash doesnt...
Not sure why yet.

There is no longer a user-files.tar.gz file!

The user must remove the "-default" from the file extension, save the
file and then make their own changes. I have seen many other programs
use this and it seems to make sense.
The README has been updated to reflect this change.
Other ideas are appreciated still! :)

With the exception of user-settings.ini, the code will check for the
existance of the non "-default" files and use those instead of the
default ones.
Also if the user does not create a user-settings.ini file of their own
an error will be thrown telling them such.

Fixed the Popup for "total number of link" in the attributon bar.

Some new values were added to the meshmap.css file in order to make it
easier for the user to make customizations to the page if they want.
More of these will probably be added in the future.

"markdown" is a pain! I should have just written a webpage! :)

Updated to README.md file and made it much nicer. Logo updated in repo. favicon.ico added. Started the change to the -default files for the user editable files

please test and let me know.

I will update (and create) and changelog file quite soon.

Enjoy!

trying to make it nicer with the auto-html generation...

also removed the previously created README.html

browser.

(also testing README.html generation)

Don't run your code through an HTML5 validator if you can help it! :)

There is no reliable internet server that provides both HTTP and HTTPS
access.

general cleanup.

not.

There is a way to have javascript give us some indication if the clients
browser has access to the greater internet or not. This would allow the
map webpage to either pull it's local resources or to use resources from
the internet.
For 1 you can get better Map layers if the client has internet access,
and why pull data over the mesh if we do not have to?

Added "ini_set('memory_limit', '128M') to get-map-info and map_display.
If you have a large network, you may need to increase the amount of
memory needed. 128M should be more than enough in most cases, unless you
have a "Mega Mesh" like N2MH does. :)

Added "last_seen" info to the station popups (map_functions.inc)

Also added (hopefully) a catch in get-map-info to find a "stalled" node
polling run.
Sometimes the node polling portion of the script never completes and it
never gets a chance to unset the "currently_running" bit in the
database.
There is now a check so that if the "currently_running" bit is 1 and the
time the script last polled the node is greater than 3 * the
"node_polling_interval" (from the ini file), then the script will just
go out and poll the nodes as normal.
If it completes, it will unset the "currently_running" bit and will
update the timestamp.

These changes have all been suggested by N2MH. Thanks Mark! I finally
got to it!