6 replies to this topic

[Raven434th]

why do we still have to make a selection for every match for public vs private?? why didn't they include it with assault and other modes??

[Beaching Betty]

BUMP FOR JUSTICE!

[Troutmonkey]

Two buttons
[Play ] [ Private Game]

My preferred option though would be to completely scrap this UI and start again, and let someone who knows what they're doing make it

[gloowa]

Troutmonkey, on 26 July 2014 - 10:44 PM, said:

Two buttons
[Play ] [ Private Game]

This. And grey out private game button when not applicable (no yellow time or whatever this paid thing is called)

Troutmonkey, on 26 July 2014 - 10:44 PM, said:

My preferred option though would be to completely scrap this UI and start again, and let someone who knows what they're doing make it

It's not THAT bad. If just points 2 and 3 were impemented i eould be semi-happy with UI. If they made point 1 a reality, i would say UI is pretty good.

1) Make mechlab look like smurfy - WHOLE GODDAMN MECH is shown (like in mech details view) and you select components from small list on left/right/whatever.
2) Begone trial mech! We do not want your kind in our mechlab!
3) Mechlab ''filter'' (it's double quoted cause filters usually have more than 2 options, but yeah, filter) remembering my last selection and sticking to it until told otherwise, regardles of whether i selected a mech, dropped into battle, restarted the game, had my house flooded or leveled by volcano eruption.

[Sneakypoop]

gloowa, on 27 July 2014 - 11:13 PM, said:

This. And grey out private game button when not applicable (no yellow time or whatever this paid thing is called)


It's not THAT bad. If just points 2 and 3 were impemented i eould be semi-happy with UI. If they made point 1 a reality, i would say UI is pretty good.

I beg to differ. It is THAT bad.
Let's look at some "UI design" best practice guide lines, I did not go too far to find that...
1st result on google search.

Improving the User Experience (by the US gov).

Let's see... those are pretty simple and clear, and to whomever designed an interface, they do ring some bells...

1. Keep the interface simple. No. Really ?
- The whole group management windows is ridiculously complex. One needs a degree to understand the group management. For example, why the **** do I have to invite some total stranger to a "friend list" to be able to get him on his team ?
- The many "confirm" useless messages popping on the screen. Why do you systematically assume all messages absolutely need to be acknowledged by many clicks like the "Find match canceled" message. I mean seriously ? You think I did not notice I'm not in a game. Do I really need to get a click to confirm message box ? Would not just a status message be enough somewhere ?
- As stated, the insanly complex mechlab where 1/3 of the screen is used for a static image that actually gives no usefull info and where you have to go through god know how many click to configure a mech, switching window for every part of the mech... when it's been done on one single window on a website.
- The freaking menus on the left and top that do not work in the same way (same buttons, but not same behavior)...
The UI failed that test.

2. Create consistency and use common UI elements.
There are dark blue buttons, light blue buttons, orange buttons, links, transparent buttons, icon buttons, tabs buttons that work the same ways / some identical buttons that work different ways, links that work like buttons / icons that work as buttons.
Not to mention the 100 "configure" buttons in my mechbay. Why ? Can't I get just ONE configure button ? How ugly is that, by the way. Save the space, make it BIG and easy to read, for example.
But even worse : in order to make the interface better, functionality improvments make the UI even harder to use ! Example ? The drop down menu to choose your game mode requires a click on a pretty difficult to see 5x5 pixel icon... ( come on, how many times did you get the "why can't I select the game modes anymore ?" question, and how many times did you click on the play button instead of the "select mode dropdown menu button". That's not being SIMPLE.)...
And so on and so on....
Complete absence of consistency. Epic fail.


3. Be purposeful in page layout. Let's say this is one is really difficult, due to different displays resolutions (sigh, like that is new). What common UI features are used to adress that difficulty ? Windows ? Moving windows ? Resizable windows ? None. And 1/3 of the display is used to show a robot image that has about... no use.
Then the menu display, the way the save buttons are placed... clearly the twisted mind whom designed the group button (that one is an tiny icon randomly placed on the bottom status bar) illustrates that perfectly. Sometime, it flashes, god knows why, sometime because someone actually wrote a message to you, sometime because someone was recruited to your team (wooow I do need to look, that really needs my attention), sometimes because you recieved a friend invite. Yeah that's a hell of great button placement / design / simplicity. It was designed to be intuitive and simple, no doubt about it.
Yet another fail on my book.


4. Strategically use color and texture. Total fail. I mean really, this interface is difficult to ready ( writting white on gray buttons / white on light orange buttons, what a greaaat idea). Not to mention the "texts" hardly readable in 1920x1080 due to the size font.
What about the "nearly invisible" checkboxes (really , deep dark gray on black with a 1pixel grayish bevel...Seriously ?) (example in the 'settings' windows...). Wrong usage all over the place. And there are SO many examples all over the UI.
Just select a mech in your mechlab, and read it's info in the buttom... the info titles are writen in white on a white smoke moving background. What a good usage of texture / color / contrast is that ? One might argue that the smoke disappears for 10 secs every 30 secs.
Some buttons even have the exact same behavior, but do not the same display. How strategic is that ?

5. Use typography to create hierarchy and clarity. Let's see.. SETTINGS / GAME / AUDIO / VIDEO / KEYBOARD / CONTROLLER SAVE / RESET / CANCEL. _____Advanced options _____ (I would use the color but I can't) Figure it out yourself.
Some menus have a title, some don't.... Again, that's an epic fail.

6. Make sure that the system communicates what’s happening. Yes that one is about properly done. Good job at that. It's basic, but not always being used, even by those whom made this the rule of thumb (explains a lot about microsoft success, even though they tend to fail at following this simple rule nowadays).
So that one is a big EXCELLENT design.

7. Think about the defaults. The whole thing is so obviously wrong and takes sooo long to fix. I mean how long did it take before we actually could mute the UI itself, for instance ?
Whomever has used a phone in his life knows it's freaking annoying when it beeps everytime one hits a single key. Yet they did not think about this default on their own. No really deep thinking here.
Criticizing once work is a hard thing to do, but it was not a concern during UI 2 development.
I bet they just turn the sound off on the computer when they design the UI, but while testing the sound actually works.
Fellow developers would go crazy otherwise. Guess what : users too.
But not a concern, right, since they are not using it / have the opportunity to turn it of (hey it's their own program).
I dare the designer(s) tell me that it did not happen and they hve been developping in a "beep" when hovering over any item environment with speakers on.
Therefore I guess, as an example : it's a failure. No deep thinking here.

Edited by Sneakypop, 28 July 2014 - 03:11 AM.

[gloowa]

Sneakypop, on 28 July 2014 - 03:08 AM, said:

(...)

Honestly, tl;dr, but as a programmer and administrator of some big servers i tell you: MWO UI is NOT that bad. Trust me. But, i fully acknowledge that this GUI does not follow best practices.

The main issue IMO, is that the main design approach is ass-first. What we should be doing in UI is:

1) After start we are in mechbay and select a mech from filterable/orderable/etc list.
2) After mech selection we are transferred into mechlab, where we can customize everything (loudout, modules, skills, all other stuff that is mech-specific) without having to select same mech over and over again in every aspect of customization.

3) Click Play.

[Sneakypoop]

gloowa, on 28 July 2014 - 04:40 AM, said:

I agree : MANY UIs are waaaaay worse.
At least it "works" : the functionality behind the presentation is about right.
Which is better than many softwares projects this size can achieve.
I guess that could qualify MWO as a good software with an ok UI compared to the industry standards.

Although I judge of an item quality for what I believe can be done rather than what is usually done.
In other terms, it's not because a million people are doing it wrong that they are any more right about it.
It's not because it's the way we've always done it that it is the way it should be done either.

Therefore I stand with my statement : UI 2.0 is not good at all.
It is very possible (and I am pretty certain not that hard) to make it WAY better, even to make it really good.

Besides, the description you made implies a re-font of the current UI ("a smurfy-like mechlab", you said).
And yes, for now it seems doing things right is not happening UI-wise.
Fixing the UI part is not optional in such a business : it's the first thing players use, and past the first match most players go "this is a solo player game", not massive -- and pass to the next game...just because they can and other games have a much better interface.
Neglecting the UI is the worst thing developer teams can do to themselves.
And that's what I see here.

Again I stand with the statement I made : the current User Interface is very wrong.