#2748 - Tweak how URL caching / automatic linking works
0 guests and 0 members have recently viewed this.
The top 3 point earners from 14th Dec 2025 to 21st Dec 2025.
| PDStig |
|
|
|---|---|---|
| Gabri |
|
|
| sholzy |
|
|
There are no events at this time
One specific scenario is we will skip link parsing if an explicit <a> tag is already in this code. As we take this as a signal that the written URLs have been put there to be displayed literally without parsing.
You shouldn't think of the WYSIWYG as laid on top of Comcode, you should think of it as an alternative to writing Comcode except when you actually want a Comcode tag (in which case you can build that visually).
Automatic link parsing, and some other minor Comcode features, are not guaranteed to work when using the WYSIWYG editor. This is due to a complex set of conditions for what depth/type of Comcode parsing Composr goes into. These conditions are designed to minimalise corruption through incorrect or unnecessary assumptions made by Composr.
You should use the features of the editor to explicitly mark up your content as you want it and think of it as mainly an alternative to writing Comcode, rather than a supplement to it.
So it's not inconsistent, it's just a set of complex rules that do their best to optimise a difficult situation.
I still feel this is a UI / UX issue that can be improved though. We can't assume users will know that if one link is explicitly included, then it disables automatic link parsing. I think there should be another way to do it. For example, instead of automatic link parsing, perhaps make it semi-automatic instead (the user would have to, say, highlight text to inspect and then click a button to automatically parse the URLs to links in the editor prior to submitting. If they don't do this, then URLs remain plain text).
What we could do is make it tell you whether it has done a Comcode scan or not. And also let you explicitly say what kind of scan you want. I'll add a separate issue for that.