Fix fact sorting for uptime

[jgrammen-agilitypr]

an attempted fix that doesnt appear to work, and may have consequences for other pages

[coveralls]

Coverage remained the same at 82.252% when pulling e66d691 on jgrammen-agilitypr:master into 974c51a on voxpupuli:master.

[coveralls]

Coverage remained the same at 82.252% when pulling e66d691 on jgrammen-agilitypr:master into 974c51a on voxpupuli:master.

[gdubicki]

Hi @jgrammen-agilitypr ! I think that you almost have it working.

I would just add an if to apply the code that you have added here but only if the fact name is "uptime".

To do that you could add an extra boolean parameter to the datatable_init macro, like a "natural_time_delta_sort", false by default.

And the value of this flag could be passed from the def fact().

[codecov]

Codecov Report

Merging #591 (8d38167) into master (c34f65b) will decrease coverage by 0.05%.
The diff coverage is 66.66%.

@@            Coverage Diff             @@
##           master     #591      +/-   ##
==========================================
- Coverage   82.77%   82.72%   -0.06%     
==========================================
  Files          12       12              
  Lines         900      903       +3     
==========================================
+ Hits          745      747       +2     
- Misses        155      156       +1     

Impacted Files	Coverage Δ
puppetboard/app.py	78.30% <66.66%> (-0.06%)	⬇️

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update c34f65b...8d38167. Read the comment docs.

[jgrammen-agilitypr]

ok, I have rebased on voxpupuli:master
I have added the script to the static folder for offline mode.
the datatable_init flag has been created and is being used in the template
but I could not figure out the change tot he def fact() function, some more guidance on what do do here would be appreciated

[gdubicki]

Thanks @jgrammen-agilitypr !

I meant that I imagine that we want to enable the natural time delta sorting only for the specific facts, like "uptime".

So in the def fact() (see the code here) we should pass an extra parameter like natural_time_delta_sort to the 'fact.html' template. The value of natural_time_delta_sort should be boolean saying the the fact is one of those that are natural-time-delta-sortable.

And then we can pass this parameter as a value for natural_time_delta_sort in the datatable_init macro call here.

[jgrammen-agilitypr]

ok, I made more adjustments, moved the natural_time_delta_sort around. and made an attempt at modifying the fact function.

[gdubicki]

Almost there @jgrammen-agilitypr, check out my comments please.

[jgrammen-agilitypr]

I made my best attempt at correcting your 2 comments. Thank you for your patience with me as I struggle through this.

[gdubicki]

I made it work with my commit, @jgrammen-agilitypr. Please check out its commit message to know what I changed. 🙂

Please just rebase and in my opinion then it will be ready to merge. ☺️

(We also think that in the long term we should rather switch to DataTables plugins to:

• detect the type of our values, f.e. IP address,
• sort the columns based on the detected type, f.e. IP addresses numerically.
  Then we won't have to have almost anything in the Python code. I am just worried that DataTables type detection & sorting plugins don't go in hand with custom renders - see that we had to read the raw value to sort with from the "title" field.

Update: we should switch to https://datatables.net/manual/data/orthogonal-data.)

[jgrammen-agilitypr]

My branch has been rebased on the 'voxpupuli:master' branch. If you would prefer that I abandon this work and and make some attempt at datatables orthogonal data, just let me know

[gdubicki]

I prefer to be pragmatic, so in my opinion if we have a working feature here and the code is ok, then we should merge, release it and let ourselves and the community already use what you did.

If you have the time and will to start working on a more mature and universal solution then please go ahead and to it in a separate PR.

But I am not the only (de facto) maintainer here - what are your opinions, @bastelfreak / @mterzo / @corey-hammerton ?