I don't have anything against changing to a new UI. Its always going to be a big change though. To make it easier I think they need to have a setting which lets individual users switch over, or back, and not a server wide setting. I know you can open the new UI from the tree, but it only works from the treeview, which means you rarely ever launch the new ui if you have the old UI activated. This makes moving over more painful. How about a single button on the main CC screen - New view/Old View?! c'mon make life easier for people who actually want to learn to use the new stuff.
The new UI suffers from being very very slow. The old UI was pretty slow, but due to it loading a *ton* of useless info every time you opened a computer. The new UI should be way faster as its loading a lot less, but it just crawls.
The main functions which people tend to use are still far to slow to get to - launch scripts, type commands into the shell, check script execution, view errors. The new UI needs extra widgets which can be configured to display summary information on the main page. Its not like they did a lot of work on the actual UI - its simply the standard dev express libraries, but with a load of the functionality disabled.
Version 2 may be more useable but at the moment, the basic - open computer, check alerts, run a script, check result, investigate why script failed or reported a problem - is just too slow.
The UI was obviously developed by someone who had a wide screen which was not very tall. They also never left their desks. The evidence is that the computerview only displays 2 rows and does not dynamically resize. I have two monitors, both will display 3 rows easily. However my laptop will not display 5 tiles wide. If i connect to my desktop from a client or from my laptop, when I open computerview I can only see a few of the tiles and I can't resize the layout. I can create new views, but changing view is sloooow.
Apart from computerview, the new layout looks good but the new versions are all lacking simple things which should really be there. As an example, the patching tool has had a lot of work go into it but theres one simple thing which is a PITA. If you need to block a patch on a single machine or a single client you have to create an entire system based on new dedicated groups just to achieve this. One patch, ignore it on this machine... it shouldn't be so difficult because its something which is frequently needed.
Apart from that, the patch manager still needs a lot of work on the refresh handling. Its highly confusing to use as the various panels don't refresh correctly, sometimes you click on something and everything is ok, sometimes what is displayed doesn't correspond with what you have clicked through. The refresh button seems to be the solution provided by the devs to correct this problem, but its a mess as you never trust what you are looking at.
Standards and Health is a mess. The scroll bar business is appalling. Its unusable as the ui jumps all over the place. It also displays all clients - no regard paid to the service plans, so I just see a big list of clients. We have a few we don't maintain. We have some where we only maintain the servers but have the agent installed on a few workstations just for remote access. They shouldn't be displayed.
No margin is allowed for exceptions. As an example, we still have a couple of clients who have old 2003 machines which run LOB systems which can't move for the moment. These machines drag down the statistics for the whole client.
I can't have an exception for servers with no AV - we don't put AVs on all servers, but I can't create a profile for servers with no AV, so the stats once again come up as bad. Our own Automate server shows up with low compliance as it has no AV, high memory useage and cpu spikes - so its a perfectly normal database server but it drags down our stats.
I can't have an exception for virtualization servers. If I have a 1Tb disk on my VM host and it it 95% full with vhdx files I really don't care. Its not going to expand so I have no problem. These machines drag down the stats as well.
If a VM host is using 60 of 64Gb of memory it shows up as a problem when its really just optimized use of available memory. 2 VMs with 30Gb memory on a 64Gb host will drag down the stats.
I can't have an exception for SQL servers. SQL server uses all the memory it can get, so it provokes low memory alerts. These machines drag down the stats but there isn't a problem. Quite the opposite, they're running exactly as they should.
On the workstation front, simple things like the HP_Recovery and HP_Tools partitions, which are small partitions which come as factory installed, but which are almost 100% full, completely throw the stats as they show up as 'secondary drive usage of 100% exceeded...'
I could go on, and on, and on.
-- Ian Murphy