Forum OpenACS Q&A: when a <formtemplate> form is rendered, all blank elements on any form get a default value other than blank

Collapse
1: when a <formtemplate> form is rendered, all blank elements on any form get a default value other than blank
Posted by Jim Lynch on 08/03/13 01:44 AM
(this might be related to https://openacs.org/forums/message-view?message_id=4065563 and could even be the cause)

so any formbuilder form with missing or ""-valued elements get their value replaced with some single-char default value, and I'm looking for how that happened, including the kernel- or subsite- parameters. Snyone notice anything like this?

Collapse
2: Re: when a <formtemplate> form is rendered, all blank elements on any form get a default value other than blank (response to 1)
Posted by Jim Lynch on 08/03/13 10:50 PM
Weir4der and weirder:

So the system forms also show this bug, and I was trying to get rid of it by removing the single-char default value, and it kept saying "this package (kernel in this case) doesn't have parameters" when I submit the parameters form... is it hosed? I'm considering starting over.

-Jim

Collapse
3: Re: when a <formtemplate> form is rendered, all blank elements on any form get a default value other than blank (response to 2)
Posted by Brian Fenton on 08/04/13 06:52 PM
Sounds hosed Jim. Definitely never seen that before.

Brian

Collapse
4: Re: when a <formtemplate> form is rendered, all blank elements on any form get a default value other than blank (response to 1)
Posted by Jim Lynch on 08/05/13 04:13 AM
Yeah, I'll just start over.

so I first noticed this when I logged in one time, I must have gotten rid of cookies and so it didn't remember my email, so it should have been blank, but it had a char in it.

Then, I tried to reg a new user and the form had the same char on every blank form element, and it put that char in for the username in the db, as it was a hidden form element... when I submitted that, all the form elements which were blank had this same one-char default value, each of which may have gone into the db..., so, weirder and weirder...

I think I'll do the github fork thing and maybe check out 5.8

-Jim