Hi Kedar,
Kedar Mhaswade wrote:
> Thanks Anissa. But I am sorry, I don't agree.
>
> On the GUI, we don't have to mimic what's there in the schema (which
> is confusing, IMO).
>
> I was *genuinely* confused when I wanted to turn on the security on
> admin-listener. I think we have regressed here from V2 without any
> apparent gain because in V2, there was a simple check-box to make
> admin-listener (or any listener) secure. I can bet you'll get some
> frowns if we don't make this easy to understand and configure.
>
> Remember, end user does not care if security-enabled attribute is
> on a <network-listener> element or on a <protocol> element. All they
> want to do is to secure the admin-listener, mostly without having
> to do the RTFM. Telling them, yeah, there's a way, but you'll need to
> know and understand this new indirection called "protocols", is a step
> backwards.
I mentioned we will improve by providing a link to the protocol so user
can set any other attributes besides security. I am not against the
idea of providing the checkbox for security in the network listener
page, but there is some difficulty that we need to think about although
technically it is easy.
Since the user can switch to another protocol for this listener, and if
there is also a checkbox for the security, it is very ambiguous if
the user means to switch to use another protocol AND at the same time,
want this protocol to have security enabled. (listener.gif attached).
Well, this is UI design that our team need to think about. Just want
to let you know that it is not as straight forward as you assume. But
we will do something about this.
>
> Regarding HCI guidelines -- well, I just don't get them. BTW,
> this was not how we did it for V2. Have the guidelines changed?
> Labeling an unchecked box as "Enabled" is simply counter-intuitive.
It definitely has been like this for many release.
>
> Also, my point regarding turning off the feed was -- there should
> be an obvious way to tell admin console that it does not try to connect
> to the Internet. We have been there as you know and my question was
> whether you had a simple flag somewhere that does that. (Also, note that
> there is an annoying issue of authenticating proxy servers).
>
Yes, understand. We need to do something regarding network issue.
thanks
Anissa.
> -Kedar
>
> Anissa Lam wrote:
>>
>> Hi Kedar,
>>
>> Please see comments inline.
>>
>> Kedar Mhaswade wrote:
>>> First off, great job on console! I like the GlassFish News Section.
>>>
>>> (I hope there is a way to turn off the Internet access though).
>> Most of those is RSS feed, so without internet access, you don't see
>> much.
>>>
>>> Hope these are simple questions on GlassFish v3 (latest workspace
>>> build):
>>>
>>> 1- Does anyone know how to turn on "security" on an http-listener? I
>>> see
>>> no check-box like interface when I do: Network Config -> Network
>>> Listeners
>>> ->(Choose a listener) -> View the property sheet.
>>>
>> The security enable attribute is not with the Network Listener, it
>> is specified in the Protocol that is associated with the listener.
>> We can improve this by adding a link to the protocol page in the
>> Listener page itself as there maybe other attributes of the protocol
>> that one may want to change besides security.
>> The easiest is click on the Protocol link in the Network Listeners
>> table (attached listerns-table.gif)
>>
>> My pet peeve is the name of the Protocol, why a protocal has the
>> name of "listener" ?? this is too confusing.
>>
>>> It's hard to find the admin listener. Can we do something about it?
>>> (like -- Is it possible to provide a link to network listeners
>>> from HTTP
>>> Service configuration page?)
>> Yes, everything is possible, almost, given the time :)
>>>
>>> 2- This is general observation with check-boxes. Regardless of
>>> status of these
>>> check-boxes, i.e. regardless whether they are checked or
>>> unchecked, I keep
>>> on seeing a label named "Enabled" (yes, that's right, even when
>>> the check-box
>>> is unchecked, it reads "Enabled").
>>> I think it is wrong and at times annoys me ;). Can we just remove
>>> the label?
>> No. Sorry ;-)
>> This is in accordance to our HCI-Admin guideline. Specified in
>> section 7.2.1.1 Checkbox. Please see attached image (checkbox.gif)
>>> If not, can we change it to obey the visual status of the check-box?
>> Thats is a label, nothing to do with the actual status of that
>> attribute.
>>>
>>> (See attached image that illustrates the point).
>> Maybe i am biased, but it is very clear to me that Enabled is not
>> checked, so, that particular attribute is disabled. :)
>>
>> Anissa.
>>>
>>> Regards,
>>> Kedar
>>>
>>> ------------------------------------------------------------------------
>>>
>>>
>>> ------------------------------------------------------------------------
>>>
>>>
>>> ---------------------------------------------------------------------
>>> To unsubscribe, e-mail: admin-unsubscribe_at_glassfish.dev.java.net
>>> For additional commands, e-mail: admin-help_at_glassfish.dev.java.net
>>>
>>
>> ------------------------------------------------------------------------
>>
>>
>> ------------------------------------------------------------------------
>>
>>
>> ------------------------------------------------------------------------
>>
>> ---------------------------------------------------------------------
>> To unsubscribe, e-mail: admin-unsubscribe_at_glassfish.dev.java.net
>> For additional commands, e-mail: admin-help_at_glassfish.dev.java.net
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: admin-unsubscribe_at_glassfish.dev.java.net
> For additional commands, e-mail: admin-help_at_glassfish.dev.java.net
>