Watch Me Do doesn't take keyboard input

[KBug]

I want to create a Watch me do app that will open safari and log into my router config file. Every thing is good when recording the action but when I run the workflow, Automator opens safari, clicks on the 192.168.1.1 button that I dragged on the bookmarks bar but doesn't input the pass.
I have enabled access for assistive devices from the Universal Access Prefs.

I think Watch Me Do is only taking mouse actions and not recording keyboard input. How can I correct this ?

[#1 Rhapsody]

Are you using Leopard? It is supposed to automatically record keyboard input.

[KBug]

Yes I am using Leopard 10.5.6 but unfortunately Automator doesn't record keyboard input. Any help ??

[ElDiabloConCaca]

Could it be that Automator is indeed passing the keyboard (password) input, but the page simply hasn't loaded yet in your browser?

If I recall correctly, "Watch Me Do" takes what you do as input, but does not (for the most part) retain the pauses between your actions. So, if the mouse correctly clicks the proper bookmark, perhaps a delay of a second or two between that action and the "type the password" action should exist?

Just a shot in the dark... we use screen automation here at work, and have to account for pauses in the automation process while pages load for some of our automated actions.

[KBug]

Here is what is happening : Automator starts, double clicks Safari alias on my desktop, Safari loads, Automator clicks on 192.168.1.1 button on bookmark bar, config page loads, Automator clicks in the pass field and nothing happens from there and after some time the workflow ends. So I guess if it records keyboard input I should see dots (....) in the pass field!! What do you think ??

[ElDiabloConCaca]

I'm thinking so... can you post your script here (removing the router's password, of course, and replacing it with something like "PASSWORD") so that we can test it?

[KBug]

I've attached my workflow

[ElDiabloConCaca]

Try this:

Record your script again, but after "Click the 'Super G' button" and before "Type password", actually click on the password text field on the router's web page.

I recorded a similar app (open Safari, open a page, click on password field with the mouse, type password, press "login" button) and it works flawlessly. The only thing I can think of is that you don't instruct Automator where to type the password -- I realize that when your router's webpage opens that the password field is already highlighted and ready for typing, but perhaps being run under Automator somehow fuggles this up.

When I recorded my script (actually clicking on the password field before typing), there is now an action in the script called "Click the text field" after the "Click 'Super G'" and before the "Type password." This may be what you're missing.

My script looks like this:

Click Safari in the Dock (different, I know)
Click the "D-Link" button
Click the text field.
Type ******* (password)
Click the "Log In" button