Hello @restertest ,
It is nice but audio is not loud
Iam watching through on my phone speaker
Could you check?
Thanks,
Ramanan
Hello @restertest ,
It is nice but audio is not loud
Iam watching through on my phone speaker
Could you check?
Thanks,
Ramanan
Everyone, please, please stop trying to work AI into everything. This is a good technique, but it has nothing to do with AI, it’s not inspired by AI, it’s not smart and it doesn’t work on every webpage. You just used ChatGPT to write a trivial piece of JavaScript. It’s actually a small subset of a script I use for all sorts of purposes. And I wasn’t the first to write it - it’s very widely used. James Bach uses it in his bookmarklets.
The next issue is that the script does not “spot every clickable button, link, or element”. In fact, it may not spot any of them, depending on how the page is coded.
The original code from ChatGPT only highlights <a>, <button> and <input type=“button”> elements. There are many other clickable elements that it does not highlight, such as <input type=“submit”>, <select>, radio buttons, checkboxes, <details> and <summary> elements etc. It doesn’t highlight any other element types that have JavaScript event handlers, which might look like <div role=“button”>. It looks like you modified the script to pick up some of those, but not all of them.
Also, the logging code from ChatGPT assumes the presence of a JavaScript framework, probably jQuery, so it won’t work if the website under test does not have that framework. You seem to have fixed this, but I can’t see your final code. It’s supposed to be in the video description, but I can’t find it.
This is the problem if you don’t design and write your own tools or at least get them from an authoritative source, and it’s why I now write most of my own. If you’re going to get code from ChatGPT, you’ve got to check what it does, what it doesn’t do, and fix it. And if you are able to do that, you could probably have written the code better in the first place.
The important lesson from this is that testers need to think about what would be useful to them. This is actually the most difficult part - creating the tool is usually a fairly simple mechanical process once you know what you want to build. Almost all testers seem happy to work within the constraints of whatever their tools do, without thinking about how they could do better or faster testing if they had different tools. If you could have a tool that did anything, what would it do? Now go and build it, or get someone else to.
I can confirm that the audio is virtually inaudible, even with a PC and headphones. I had to turn on the transcript to understand what I was seeing.