Message8565
Willam said:
> This is properly brilliant!
Thanks. It was about 2 days of work (20 hours). Mostly javascript stuff
that I cribbed from the Mozilla developer site demo. I have a few updates for
styling etc. and am tracking it via a fossil website. I'll send the link
to the site via separate email.
> > William is your tracker a jina based tracker (using jinja2 template)?
> We actually transitioned to Jinja just a handful of weeks back
> (primarily out of developer comfort). Though, from a
> quick glace, this looks like a dead simple TAL to Jinja/Bootstrap
> migration, so we'll happily try it out in our instance regardless.
It should be relatively straight forward. Maybe a good jobs for an LLM to take
a first crack at if you use them. Using the Roundup DeepWiki
(https://deepwiki.com/roundup-tracker/roundup) might also be helpful.
I noticed w/ deepwiki it sometimes dropped attributes. So make sure to keep all
the attributes in your jinja translation. The javascript and ability to
color the columns or collapsed columns by property and property value need
the attributes.
> I'm happy to contribute any work we put into this,
> for example, the Jinja/Bootstrap port.
That would be great. My personal rule is not to add anything to the distributed
templates unless I can maintain parity between TAL and jinja2 implementations. If
your jinja2 reworking has the same functionality as the TAL, maybe I can add it
to the distribution.
Also a few things to note:
* the user has to have rest enabled and you have to run 2.0 or newer of Roundup
(2.5 or greater is preferred (2.6 is current)).
* the current implementation does not work on tablets/touch devices.
You can see the kanban layout but not manipulate it.
The https://github.com/drag-drop-touch-js/dragdroptouch polyfill
might be a solution.
* the HTML script and style tags don't include the CSP nonce. If you aren't
using content security policies you don't need to worry about it.
Adding: tal:attributes="nonce request/client/client_nonce" (or jinja
equivalent) to the script and style tags will fix it.
* depending on how fast the backend runs, you can drop a card and have a small
delay while the backend is updated or returns an error.
If an error is returned, the error is displayed using alert() and the card doesn't
replace the placeholder.
If the update succeeds, a second request is done to get an updated parameter (etag)
that is needed if the card is dragged again. Then the card is replaces the placeholder.
So this adds to the delay.
I have a patch to the roundup core code so the tag is returned with the response
that changes the backend data. This will reduce but not eliminate the delay.
For a first pass, I think this is acceptable.
Have a great day and thanks for being willing to try this out. |
|
| Date |
User |
Action |
Args |
| 2026-10-22 16:40:35 | rouilj | set | messageid: <1792687235.01.0.0761316848901.issue2551363@roundup-tracker.org> |
| 2026-10-22 16:40:35 | rouilj | set | recipients:
+ rouilj, wilsj |
| 2026-10-22 16:40:34 | rouilj | link | issue2551363 messages |
| 2026-10-22 16:40:34 | rouilj | create | |
|