These forums have been archived and are now read-only.

The new forums are live and can be found at https://forums.eveonline.com/

EVE Information Portal

 
  • Topic is locked indefinitely.
Previous page123Next page
 

Dev blog: In-Space Brackets revisited with Crius

First post
Author
Previous Topic Next Topic
CCP Delegate Zero
C C P
C C P Alliance
#21 - 2014-07-10 11:31:00 UTC
Alexis Nightwish wrote:
CCP Delegate Zero wrote:
We are capping the list at 15 items in both modes, which reflects the pre-existing cap on the bracket listing. To raise this number carries with it some concerns regarding performance and we don’t believe there is a strong case for a higher number as things stand. The balance of feedback was against scrolling in the bounded mode and this again factors in the decision to cap the number at a manageable set of items.


Please explain how this is a performance concern. The client already has the information (name, distance, etc.) of the objects, it's just a matter of displaying it in a new spot.


The gist of why it is a performance concern is that regardless of most information being primed for display, there is still an expense when constructing the bracket list as the picking logic has to iterate over the objects in the scene to confirm whether or not they intersect with the bracket that is at the 'pick point', so to speak. The longer the list to be constructed, the more times the loop has to be iterated to determine whether an object's bracket should be added to the list.

This is a bigger concern the more objects are actually in the scene, which seems to be the use case in question.

In short, the bracket list, by its nature, is not doing quite the same thing as the overview (the obvious other object list to compare it with), and the overview is of course doing its work as well. I am not the UI programmer here and I am giving a rough explanation. But it is based on what programmers tell me and I trust their experience and judgement.

I'll also note that display of UI elements is not free. It's not 'just a matter of displaying it'. Not all information to be displayed in the constructed list is necessarily primed in the exact format to be used. There are always expenses involved and these will necessarily mount up the more display items there are.

Best judgement on our part is that the balance of performance, useful display and typical use cases for the bracket list is still in the 15 item range.

I personally think going much beyond that would call for some deeper design work and likely some optimization deep dives too. We're not addressing that at the moment given that the overview is still the go-to interface element for managing information on large numbers of objects.

Thanks.

CCP Delegate Zero | Content Designer - Writer | @CCPDelegateZero
Sentient Blade
Crisis Atmosphere
Coalition of the Unfortunate
#22 - 2014-07-10 11:56:39 UTC
Hmm, maybe I read that wrong, are you saying that building the list requires something other than O(n) complexity for the picking, prior to ordering?
Crazy KSK
Tsunami Cartel
#23 - 2014-07-10 12:38:08 UTC
the black boxes are certainly nice for people with bad eye sight but for me they are just annoying since they block whatever might be behind it the old system also allowed you to see to which celestial a ship was warping to, I don't see this iteration re-enabling that, (?) which makes this iteration still worse then the old system

Quote CCP Fozzie: ... The days of balance and forget are over.
CCP Delegate Zero
C C P
C C P Alliance
#24 - 2014-07-10 15:04:45 UTC
Sentient Blade wrote:
Hmm, maybe I read that wrong, are you saying that building the list requires something other than O(n) complexity for the picking, prior to ordering?


No, not really. Just making the point that the higher the cap the more likely it is to not reach the cap and therefore check every object in the scene before it stops iterating. As I say, then there is the cost in displaying the information which of course is changing in real time, together with the fact that the cursor is likely moving a lot - which means there will be a lot of tear down and reconstruct going on. It all adds up.

In any event, it's one factor contributing to our judgement as to what works in terms of this feature. It's worth noting that the picking is quite a bit more conservative now so the field of brackets has to be pretty dense and the camera zoomed out some way to get a full list. Which, again, I'd have to say takes it into the overview's territory.

We will keep an eye on things but as I indicated in the blog, we're pretty happy with where it is and will be looking at tooltips proper again as part of our next work

CCP Delegate Zero | Content Designer - Writer | @CCPDelegateZero
Skalle Pande
Teknisk Forlag
#25 - 2014-07-10 22:16:50 UTC
CCP Delegate Zero wrote:
Sentient Blade wrote:
Hmm, maybe I read that wrong, are you saying that building the list requires something other than O(n) complexity for the picking, prior to ordering?


No, not really. Just making the point that the higher the cap the more likely it is to not reach the cap and therefore check every object in the scene before it stops iterating. As I say, then there is the cost in displaying the information which of course is changing in real time, together with the fact that the cursor is likely moving a lot - which means there will be a lot of tear down and reconstruct going on. It all adds up.

In any event, it's one factor contributing to our judgement as to what works in terms of this feature. It's worth noting that the picking is quite a bit more conservative now so the field of brackets has to be pretty dense and the camera zoomed out some way to get a full list. Which, again, I'd have to say takes it into the overview's territory.

We will keep an eye on things but as I indicated in the blog, we're pretty happy with where it is and will be looking at tooltips proper again as part of our next work

Sounds like a very sensible explanation - and decision - to me :-)
Gaijin Lanis
Gallente Federation
#26 - 2014-07-11 00:42:38 UTC
That's... nice.

I mean the original system with a drop shadow or even a single pixel worth of black border on text would've been ideal, but this is definitely a step in the right direction to correcting the botched tooltip update.

The above was written and posted with nothing but love in my heart for all.
Logicycle
Stimulus
Rote Kapelle
#27 - 2014-07-11 03:35:47 UTC
This looks good. Nice works, go Team Fluffy Unicorn.
Mag's
Azn Empire
#28 - 2014-07-11 11:32:04 UTC
As long as you give me options in regards to tooltips, then I will like the changes.

These options should include ways that make me able to completely delay certain types of tips, that really are not required with experienced players.

Thank you for admitting your mistakes. Blink

Destination SkillQueue:- It's like assuming the Lions will ignore you in the Savannah, if you're small, fat and look helpless.
Pic'n dor
Sebiestor Tribe
Minmatar Republic
#29 - 2014-07-12 08:20:25 UTC
Sometimes you got a huge list of object in the bracket.. Sometimes, more that you can actually display.. Can you make a scrolling feature in these brackets ?

Same issue sometimes with the overview selection. The list may contain more items that the game can display..

COUCOU TOUCHE TOUCHE
Marian Devers
Rage and Terror
Against ALL Authorities
#30 - 2014-07-12 09:02:04 UTC
So whats the plan when a battleship warps to Asteroid belt #16?
Sable Moran
Moran Light Industries
#31 - 2014-07-12 18:14:05 UTC
Mag's wrote:
These options should include ways that make me able to completely delay certain types of tips, that really are not required with experienced players.


I would like to emphasise this. When it comes to the UI one size really does not fit for everyone. A noob has different requirements for it than a vet. Likewise scout, FC, ratter, miner, logi all need the UI configured differently. Please help make their work easier.

Look into the customising work done by the audio team, that's what I would like to see for the UI too.

Sable's Ammo Shop at Alentene V - Moon 4 - Duvolle Labs Factory. Hybrid charges, Projectile ammo, Missiles, Drones, Ships, Need'em? We have'em, at affordable prices. Pop in at our Ammo Shop in sunny Alentene.
Galmas
United System's Commonwealth
#32 - 2014-07-13 18:04:25 UTC
Will these pop-up thingies be available in the system map as well now? which actually is the most usfull place for it anyway since it would fix the problem with not being able to get to the context menu of some moons when there are lots of moons on a planet.

speeking of system map weakness, any plan to give us the general in space context menu (containing the bookmarks) as we have in the normal in space view?

Regards
Gal
Ax Pym
Doomheim
#33 - 2014-07-14 11:34:53 UTC
the compact bracket list looks hot. i couldn't test it on Singularity. I hope it sizzles, like the last release.

axpymlives.tumblr.com
Ammzi
Dreddit
Test Alliance Please Ignore
#34 - 2014-07-14 13:41:03 UTC  |  Edited by: Ammzi
I can't believe this thread is not getting more feedback as it is one of the most fundamental aspects of playing EVE online: Brackets in Space. Blink

I want to emphasize the need for a longer bracket list enabled/custom max limit as it sometimes hides extremely valuable information. Here's an example:

Unless you are already familiar with the system layout then knowing whether there is a gate on a certain planet or not is up to the mercy of the bracket list to display this when a target warps towards a planet cluster.
I am warping onto gate A in an interceptor as a target ship decloaks and warps away before I get a lock on it. I saw it warp towards a planet cluster - of course since I am in an interceptor or similar "fast tackle" I have enabled moons on my bracket list with the short cut ALT+SHIFT+X. This means there is dozens of moons on top of that planet the target warped towards and therefore I can't see whether there is a gate or no if I am not familiar with the system. So I end up having to move my mouse away, disable the moons, hover back up, confirm whether there's a stargate or whether he went to a planet and that valuable 5 seconds wasted means he could get away in either case.
That's a convoluted case, but I have had it happen multiple times where I vaguely remember "Isn't there usually a stargate over there?".

What is much more "game breaking" is when you are leading a bomber squad and 200 km off you a 30 man hostile frigate fleet just killed someone and their wreck is there providing you a perfect warp in for your bombers, but since the 30 frigates are all clustered up on the wreck I am not able to get the wreck to show up on the bracket list and warp my squad. Therefore as I finally either get the wreck to show up after enough zooming or forced to probe the frigates it is too late and the frigates got away. :(

This goes for any case where someone has to warp X fleet on top of Y fleet given a singular Z warp object, but can't because of 15 bracket list ceiling. When you are in warp range or further away, zooming in to get a higher resolution of the field to reduce the number of brackets so you can isolate that one object is difficult and time demanding. :(


Edit: So conclusively if I was able to scroll and enable a higher ceiling on bracket list that wold be very useful. That or some intelligent selection of brackets, but I'd assume that's more costly to have running.
Alexis Nightwish
#35 - 2014-07-14 23:02:59 UTC
CCP Delegate Zero wrote:
Alexis Nightwish wrote:
CCP Delegate Zero wrote:
We are capping the list at 15 items in both modes, which reflects the pre-existing cap on the bracket listing. To raise this number carries with it some concerns regarding performance and we don’t believe there is a strong case for a higher number as things stand. The balance of feedback was against scrolling in the bounded mode and this again factors in the decision to cap the number at a manageable set of items.


Please explain how this is a performance concern. The client already has the information (name, distance, etc.) of the objects, it's just a matter of displaying it in a new spot.


The gist of why it is a performance concern is that regardless of most information being primed for display, there is still an expense when constructing the bracket list as the picking logic has to iterate over the objects in the scene to confirm whether or not they intersect with the bracket that is at the 'pick point', so to speak. The longer the list to be constructed, the more times the loop has to be iterated to determine whether an object's bracket should be added to the list.

This is a bigger concern the more objects are actually in the scene, which seems to be the use case in question.

In short, the bracket list, by its nature, is not doing quite the same thing as the overview (the obvious other object list to compare it with), and the overview is of course doing its work as well. I am not the UI programmer here and I am giving a rough explanation. But it is based on what programmers tell me and I trust their experience and judgement.

I'll also note that display of UI elements is not free. It's not 'just a matter of displaying it'. Not all information to be displayed in the constructed list is necessarily primed in the exact format to be used. There are always expenses involved and these will necessarily mount up the more display items there are.

Best judgement on our part is that the balance of performance, useful display and typical use cases for the bracket list is still in the 15 item range.

I personally think going much beyond that would call for some deeper design work and likely some optimization deep dives too. We're not addressing that at the moment given that the overview is still the go-to interface element for managing information on large numbers of objects.

Thanks.

So if I'm understanding this correctly, the list is built server-side? Because that's the only way this makes sense.

If the bracket list was built client-side then the 'use case' is whatever the individual player feels is best, not some arbitrary number you guys decide upon. We're not playing EVE on the Xbox. Our PCs are not equivalent in power. What works for one is inadequate for another.

Man you guys are stubborn on this whole UI/tooltips issue. Makes me feel like you're making the game for you and not us. :(

CCP approaches problems in one of two ways: nudge or cludge

EVE Online's "I win!" Button

Fixing bombs, not the bombers
Oraac Ensor
#36 - 2014-07-17 08:33:13 UTC
Quote:
As the in-space brackets are a completely different feature to tooltips per se, we have been concentrating initially on improving this key part of the UI. Now that this has been done, we will look at improving the experience with tooltips in later releases.

But it has NOT been done!

I still see an immersion-breaking black box, albeit now containing only the icons. Why is any sort of black box element necessary?

The screen-edge boxes are fine.
Dread Nanana
Doomheim
#37 - 2014-07-23 06:32:22 UTC
Ammzi wrote:


What is much more "game breaking" is when you are leading a bomber squad and 200 km off you a 30 man hostile frigate fleet just killed someone and their wreck is there providing you a perfect warp in for your bombers, but since the 30 frigates are all clustered up on the wreck I am not able to get the wreck to show up on the bracket list and warp my squad. Therefore as I finally either get the wreck to show up after enough zooming or forced to probe the frigates it is too late and the frigates got away. :(


Well, when you say leading, you mean S-boxing bombers, right?

Anyway, why don't you yet know about bracket profiles? Those existed for years and years now.
Vorpheus
Three Sword Inc
#38 - 2014-07-23 08:30:17 UTC
i liked the bracketlist of pre-crius better. It was smaller.
Arec Bardwin
#39 - 2014-07-24 23:37:32 UTC
I really like the new bracket list! Just a few things:

- an option to be able to scroll would be very nice
- an option to change max number of displayed items
- an option to scale the bracket list (independently of the rest of the UI)
Rommiee
Mercury Inc.
#40 - 2014-07-26 19:33:17 UTC
Rommiee wrote:
Quote:
Now that this has been done, we will look at improving the experience with tooltips in later releases. In that regard we are looking at configuring tooltip behavior by category, which we think would allow for you to tune how tooltips display based on the relevance of the information to you.


FINALLY !!!

After countless mentions of this, the idea has finally sunk in.

It’s only been 2 months since this disaster started, let’s see if we can get this done by Xmas (2014 preferably).



So how is this coming along ?

Or is it true to past form and just a throwaway line to keep people happy while it slowly slides into a black hole never to be seen again ?
Previous page123Next page