I tend to use URL as my everything URL bucket, which of course I am trying to change. How do you differentiate between ReferenceURL and SourceURL for use in your workflow?
Tom, for me, if I use them at all, ReferenceURL is a reference from a research source, i.e. something I’m citing. SourceURL is associated with watch folder stuff. I don’t use it that often.
Rather than trying to differentiate URL attributes type, I remind myself that every note has every attribute. Moreover, I try to treat every note as a discrete atomic idea, i.e. one idea, one URL, one citation, one media file reference. If in the rare case I need a second URL, I have user attribute $URL2. With this mindset, it makes authoring action code and export code easier as the URL reference is always the same $URL.
Tinderbox defines 4 URL-type attribues:
The first three have definite purposes (see links) so I never use these for my own ‘custom data’. $URL is used for the URLs of dropped links by can also be thought of as a general purpose attribute as if it were a $MyURL (which does not exist).
SourceURL and ReferenceURL should be populated by the processes that use them. So, IMO you shouldn’t be manually adding data there. Better to fix the source and update.
If I need a link for a particular purpose, I make a URL-type user attribute. Some of my research docs have 5-10 URL user attributes for specific defined purposes. This is useful as rather than hiding in text, they can be used in Displayed Attributes to directly open the URL and as specific attributes they can be queried more easily.
Picking up on an inference in the question, don’t worry about not using system attributes. You don’t have to use them. worse using them for a task when they are intended for a specific purpose creates the possibility of invoking that process later on and overwriting your custom data.
I have a lot of user attributes (Tip: add a Description when you create them). It’s much easier to remember what attribute relates to what task/source//data than remember which notes use the same attribute for a different purpose.
Me too! For example, I have one file that has over 400 user attributes.
Thank you both for your explanations, it really helps better understand use cases for me. In my mind, $SourceURL and $ReferenceURL seemed to be so similar, I never thought another source was the intended use. It makes sense.
I must confess, I have not used User Attributes nearly enough in the past. Instead I have been over using $Text. No right or wrong point here, but to your point, your information becomes a LOT more flexible using User Attributes than just using $Name and $Text. I guess over the past six months or so, when I have an insight or context I want to capture, I am now creating its own User Attributes. It seems to help better structure my thinking and pick up where I left off.
Thanks you both of you, PaulW, MarkB and too many others in the forum to mention who have shown me better ways to use tinderbox.
Yup… this is the basis of “meta cognition”, break thing into their atomic elements, discrete values, etc.
Absolutely nothing wrong with using ‘mainly’ $Text. But, each time you have to do
$Text.icontains()—especially with a regex pattern—it might be a moment to reflect on the opportunity to capture that info into metadata, e.g. user attributes. Once there it is much easier to review/calculate over, etc.
They are also great from transforming your information into new insights. They will let you build tables, leverage action code, iterate against your data, create associations, and produce output into various formats, and more.