You are imagining things.

The experienced monks are just good in reading RAT titles, and telling if it's interesting by judging the participants. At least that's what I do.

> without having to continuously re-roll the Perlmonks website scraping,

There are no triggers that you can set like in an email filter.

But you don't exactly need website scrapping if you parse the XML feeds.

So indeed you could poll some meta XML tickers and XML views of posts like ?displaytype=xml;node_id=1196487 to parse if your name "StevieB" or keywords were mentioned.

Updates

Since there are also some RSS Feeds you might find a ready to use tool to get alarms for your keywords.

Concerning CSS, there are many classes available to highlight special monks in RAT View.

I'm only using it to highlight me and hiding others.

But that's not what you're after, I'm afraid.

"Question is, is there a way, without having to continuously re-roll the Perlmonks website scraping, to see if a username or phrase or something is published, to recognize this? I'm definitely no web expert, but I can work my way with JS/jQuery/CSS to a degree, but searching is usually required."

What you're asking for sounds like some sort of user configurable watch list, which does not exist, and would probably be better developed as an enhancement to the site, rather than a third party site scraping tool.

I think investment into a keyword heuristics should come first.

I started once hacking a spam filter for this site based on Tf-idf and this method should lead to pretty good keywords categorizations too.

Unfortunately working for food and preparing talks for conferences doesn't leave me with too much spare time for the monastery.

Cheers Rolf
_{(addicted to the Perl Programming Language and ☆☆☆☆ :)
Je suis Charlie!}

Supersearch http://www.perlmonks.org/index.pl?node_id=3989;BIT=stevieb;dy=01;Wi

or

Google sitesearch https://www.google.com/advanced_search?hl=en&as_sitesearch=www.perlmonks.org

To meet your requirements.

Not for this but I use them to strong effect for a few other things; shopping and PR stuff relevant to me. Google indexes PM very quickly at least some of the time. A couple years ago I searched for something to help me answer a question and the second or third Google hit was that question. Node was like 2 minutes old.

its real simple, visit newest nodes, open each question/root node in a new tab, read, repeat

Perhaps it's not practical from a time spent perspective to follow this approach, depending on your free time, how frequently you visit. I can understand some use cases or where having this automated would be useful.

automating search terms for google site search with a 24 h window should be sufficient.

a link could be added to one of the nodelets.

see Re: CSS/JS if a username is mentioned (search engines)

Cheers Rolf
_{(addicted to the Perl Programming Language and ☆☆☆☆ :)
Je suis Charlie!}

Huh? Maybe on braille or text to speech or ??? The time it takes for a node to load is all i need to filter for interest, whether im reading or skimming

You can read n root nodes at a time in each respective section with the corresponding slower page load time

but keyword nodelet has been rejected forever and always, As a lifetime member i am offended by talk of it