Hi Samuel,

today I got reminded that we have to pass the accessability tests with our UI and was searching if the hamburger meneu design would be in the way of this requirement.  I didn't find any clear statement about it but I got a lot of hits on why to not use hamburger menus:
http://courtneyengle.com/2014/06/24/mobile-responsive-navigation-accessibility/
http://techcrunch.com/2014/05/24/before-the-hamburger-button-kills-you/
https://lmjabreu.com/post/why-and-how-to-avoid-hamburger-menus/

In order to get some feedback or new ideas I contacted a UI-guy I know from a former project and he pointed me to this:
http://www.ibm.com/design/language/framework/visual/introduction

This is in fact reflecting exactly the concept we need.

Please let me know what you think about it.

Thanks, Walter.

On 12.10.2015 16:03, Walter Niklaus wrote:
Hi Samuel,

my comments are from a user point of view, I'm relying on your insights when comes to technical doabilty.

Thanks, Walter.

On 09.10.2015 21:07, Samuel Henrique De Oliveira Guimaraes wrote:

Hi,

 

Although a lot of articles discourages mixing top-level tabs with sidebars, I think in this specific scenario we can make these patterns work but we have to make sure that the users won’t be redirected to a blank landing page or we’ll have to create one.

 

For instance, suppose Containers second-level first item has three children, then when clicking in “Containers” on the top bar, the user should be redirected to the first item from this list (in the third level).

I’ve done some prototypes in the past with similar patterns and for the standard user who navigates on web, this pattern of navigation is confusing. However the average virtualization, cloud, storage and network operator users are usually facing UIs more complex than this.

 

Good point. Makes sense to me.

Between the two proposals I’m voting for proposal 1 but I prefer using Hamburger Button to trigger a floating grouped-menu. This was proposed a few months ago in the ML.

I was in a workshop with some developers from Globo.com (Brazilian news portal) two years ago when they were implementing this menu. The current version is way more polished than the one they presented at the time and I really like how it turned. I’m attaching a screenshot, but you can see it live with animations at http://g1.globo.com/.

 

I like a lot of the aspects from what's implemented on the globo.com page:
  - the menue doesn't take away any real estate from your screen, therefore having an additional level of navigation isn't such a big issue.
  - you can get to your target menue in one click
  - from what I see, it gives you the possibility to turn back on the previous level tabs with just one click.  Did I get that right ?
If all this is right, then I'm perfectly ok with this proposal.

This discussion also brings another question, should new plugins be restricted only to new tabs on the first-level top bar or should we create some sort of “path” in tab-ext.xml? I.e a new plugin can be appended to a specific region on Hosts Dashboard, on the second-level (since Aline suggested moving Peer hosts to that area) or even create a new level on a menu option that previously had no child items.

Our base assumption was that plugins wouldn't be restricted to new tabs only when using the new framework.
Based on that assumption we decided that the functionality which was in the Administration tab in the old design, would now be extending the functionality of the Host tab in the new design.
Ideally a new plugin could be extending functionality on any level, and even on multiple levels.
Here the use cases I'm aware of:
  -  plugin as a new tab --> for example containers may be one
  -  plugin extending an existing tab 
          --> Ginger extending the Host tab which was created by Ginger-common
          --> Aline's example with the peer hosts
  -  plugin extending one or more individual UI-panels with functionality
          --> Gingers390x plugin offering "Enable/Add Network", "Enable/Add SAN Adapter" Action-button on the Networking and Storage screens
               since these are platform specific functions implemented in this specific plugin

Samuel, could we cover all these use-cases via the approach you mentioned as the "sort of path in tab-ext.xml" ? I guess that would be great because it sounds very flexible.
 

 

Regards,

Samuel

               

 

From: Jan Schneider [mailto:schneidj@linux.vnet.ibm.com]
Sent: sexta-feira, 9 de outubro de 2015 13:06
To: Samuel Henrique De Oliveira Guimaraes <samuel.guimaraes@eldorado.org.br>; Aline Manera <alinefm@linux.vnet.ibm.com>
Cc: kimchi-devel@ovirt.org
Subject: Re: [Kimchi-devel] UI Change Request - Navigation

 

Hi Samuel,

I just had a discussion with Walter. We found that we need up to two navigation levels under Host. The third navigation level however is not always required.
Please have a look at the two attached proposals to understand our requirements.

Kind regards
Jan and Walter



On 10/09/2015 04:36 PM, Samuel Henrique De Oliveira Guimaraes wrote:

Hi Jan,

 

I’m not sure if I’m following you. Did you mean changing the second-level navigation to tabs instead of panel areas? If so should we merge Hosts and Admin tab in one page and collapse the panels like old-ui behavior (maybe keeping the dashboard statistics on top unchanged) but with new-ui styles OR add two new tabs to the toolbar?

 

I think the concept of accordion/collapsible elements works for Admin, but I don’t see it working for the panels on Hosts tab, unless we group Basic Information, Repositories and Debug Reports in one single collapsible area.

 

I also discussed with Aline if we should move Peer Hosts from top to the toolbar and make it a plugin. Here’s a mockup I did with Chrome:

 

.... deleted images(Walter Niklaus)

 


 

 

The reason why I think all panels in these pages can’t be merged in one tab is because if all panels were collapsed, the page height would be around 3757 pixels. If admin items were hidden, it would be around 2108 tall. I believe we would have to set the accordions to hide other items when one item is collapsed.  They don’t seem that tall in the PDF because the images were resized.

 

I believe we have to study this following a “mobile-first” approach because once we all agree with the new design, even if we don’t include the mobile design in 2.0 release, we’ll have to keep some markup ready for responsive design to receive the new styles. I think this new tab design would need some updates in tab-ext.xml and the JS files before 2.0 release, assuming other people or teams would start developing new plugins from 2.0 and on. If we update how the tabs are built in the UI from 2.0 to 2.1 for instance, that may turn against us with retro compatibility issues.

 

.... deleted image(Walter Niklaus)

 

Regards,

Samuel

 

 

From: kimchi-devel-bounces@ovirt.org [mailto:kimchi-devel-bounces@ovirt.org] On Behalf Of Jan Schneider
Sent: quinta-feira, 8 de outubro de 2015 13:10
To: Aline Manera
<alinefm@linux.vnet.ibm.com>
Cc:
kimchi-devel@ovirt.org
Subject: [Kimchi-devel] UI Change Request - Navigation

 

Hello Aline,

I refer to the latest version of the
User Interface Design Specification - Kimchi, 2014-12-23 (aka UI Design Spec)
which defines the following structure of functionalities:

Host (second level navigation via panel areas)
   Performance (System Statistics)
   Basic Information
   Repositories
   Debug Report
   Software Updates

Guests
    no second level functionalities

Templates
    no second level functionalities

Storage
    no second level functionalities

Networks
    no second level functionalities

Administration (second level navigation via collapse/expand)
   Firmware Update
   SEP Configuration
   Power Options
   Configuration Backup
   Network Configuration
   SAN Adapters
   Sensor Monitor



Problem Statement

We already decided to move all Administration functionalities to Host (currently not updated in the UI Design Spec).

We are currently facing the following problems:
1) Host now contains 12 second level functionalities, all other (Guests, Templates, ...) none.
     We need to introduce a second level navigation for Host other than collapse/expand
2) The navigation bar elements Storage and Network (refering to Virtualization) also exist in the Host context.
     This might confuse the user.



Proposal

The described problems can be solved with the following changes:

1) Introducing a second level navigation
2) Changing the structure of functionalities as follows:

Host
   Performance (System Statistics)
   Basic Information
   Repositories
   Debug Report
   Software Updates
   Firmware Update
   SEP Configuration
   Power Options
   Configuration Backup
   Network Configuration
   SAN Adapters
   Sensor Monitor

Virtualization
   Guests
   Templates
   Storage
   Networks

Containers (future extension)
   to be defined



Let's start a discussion on this.

Kind regards
Jan

 



_______________________________________________
Kimchi-devel mailing list
Kimchi-devel@ovirt.org
http://lists.ovirt.org/mailman/listinfo/kimchi-devel



_______________________________________________
Kimchi-devel mailing list
Kimchi-devel@ovirt.org
http://lists.ovirt.org/mailman/listinfo/kimchi-devel