[rt-users] RT 3.6 Custom Field Validation
Stephen Turner
sturner at MIT.EDU
Wed Jan 16 14:39:23 EST 2008
Hello all,
We will soon upgrade from RT 3.4 to 3.6, and I've been looking at the
3.6 custom field validation feature. This has raised several
questions and I'd be interested to hear other people's experiences.
1. The regex format makes me nervous. We have queue admins that
maintain their own CFs and I don't want them to be presented with
regex values. It seems like this will generate a lot of confusion and
questions.
2. The javascript combobox (custom field basics screen) for
maintaining the validation specs works differently in different
browsers. Opera has the most problems, but there are differences
between Firefox, IE, and Safari. For example, IE shows an entry
labeled '-', whereas Firefox does not.
3. There's an entry in the combobox selection list labelled '-'
(value ""). I assumed that this meant "no validation", but selecting
this causes the field's pattern to be set to '-'.
4. Of the supplied formats, "Digits" isn't clear to me. It seems to
allow any combination of digits and '.' characters, e.g.
"234.5656.2.7.23". Is that the intent?
5. "Mandatory" seems misleading. If a field is spec'd as mandatory,
you can still have a ticket with no value in the field (in fact it
has to be that way, otherwise email-generated tickets would be a
problem). You can also do subsequent web updates to the ticket
without supplying a value for the "mandatory" field. Only after you
have first put a value into the field does the "mandatory" validation
kick in (i.e. you can't set the field back to null). So "mandatory"
means "you don't have to enter a value, but once you do, you can't
clear the field", which isn't really mandatory at all.
6. Has anyone created any useful regex's beyond the supplied values?
7. We're leaning towards defining several standard validation
templates and hiding them from the user behind simple descriptions.
The code mods are simple and we would eliminate the exposure of regex
to our users and the awkwardness/inconsistency of the javascript
selection control.
Thanks,
Steve
Stephen Turner
Senior Programmer/Analyst - Client Support Services
MIT Information Services and Technology (IS&T)
More information about the rt-users
mailing list