In D2578#108522, @Nescio wrote:

Not done: this patch standardizes <Identity> tooltips. <Upgrade> tooltips could be done in a future patch.

My bad.

This would fix all your complains. Improved from the version elexis mentioned before by making the edges not overlap with the slider button.

Actually I'm not sure it would. Also, I agree with elexis those sliders in your screenshot look a bit ugly, though they are clearer.

In rP23430#41084, @Nescio wrote:

Perhaps move the game speed setting to the top, to make things visually more pleasing? This looks a bit ugly to me:

• The sliders are rather narrow and because you have to click on them, this makes them harder to use than the vertically much larger drop-down boxes.

It would be nice if clicking above the slider, rather than exactly on the line, would also work. It shouldn't be too hard to increase the (vertical) sensitive area? And maybe make the arrow keys work with sliders as well?

Also dropdowns require hardcoding the values whereas a slider only has 3 datapoints (min, max, step)

Thanks, I agree that's an improvement.

Gamespeed (1) has labels, so it would remove something, also stepsize other than 1, so I would have added a worse workaround until there was D406.

Perhaps move the game speed setting to the top, to make things visually more pleasing? This looks a bit ugly to me:

• The sliders are rather narrow and because you have to click on them, this makes them harder to use than the vertically much larger drop-down boxes.

It would be nice if clicking above the slider, rather than exactly on the line, would also work. It shouldn't be too hard to increase the (vertical) sensitive area? And maybe make the arrow keys work with sliders as well?

In rP23430#41082, @Nescio wrote:

Then one could at most use a 5 minute step dropdown, but that would also not allow chosing 2 minutes or 38 minutes for example.
It is also easier to pick values with a slider if there are 60 of them.

True; the question is how likely it is people would want 38 minutes, rather than 40.

Also dropdowns require hardcoding the values whereas a slider only has 3 datapoints (min, max, step)

Then one could at most use a 5 minute step dropdown, but that would also not allow chosing 2 minutes or 38 minutes for example.
It is also easier to pick values with a slider if there are 60 of them.

True; the question is how likely it is people would want 38 minutes, rather than 40.

Ceasefire has the main use case at < 5 minutes on nomad.

Then why not reduce it further, to e.g. 30 or 20? (Just curious.)

The number of relics is the number of civs, each relic is spawned at most once.

So each mod that adds civs ought to edit the gui files?

Why did you use it for these four settings, but not others?

This had been discussed as well. It seems not publicly.
There are dropdowns with few choices where the labels are more informative than numbers, for example mapsize "small" is more descriptive than mapsize 192.
For starting resources the upper limit is infinite / 50.000, so using that as a scale wont work well.
For the population limit, maybe, I never said that this should be the final layout, but it was used for the durations first and foremost because those are continuous values selectable.

Understood, I was thinking more of settings with narrower ranges, e.g. game speed (up to 2×) or number of players (up to 8).

The relic and wonder time go from 0 to 60 minutes, ceasefire from 0 to 45; why not 60 as well, for consistency? 60 minutes ceasefire used to be an option.

The more narrow the range, the easier it is to pick values.
I have seen maybe 300 matches with Victory durations and they were usually 50minutes, since for shorter times it was too hard to attack the enemy, as it is easy to turtle up.
Ceasefire has the main use case at < 5 minutes on nomad.
(See also irc discussion with Vladislav linked)

A few minor points:

• The sliders are rather narrow and because you have to click on them, this makes them harder to use than the vertically much larger drop-down boxes.
• The ends are noticeably faded, which is somewhat confusing. (Are the min and max values less valid options?)
• The relic and wonder time go from 0 to 60 minutes, ceasefire from 0 to 45; why not 60 as well, for consistency? 60 minutes ceasefire used to be an option. (Off-topic, why exactly is 13 the maximum number of relics?)
• Why did you use it for these four settings, but not others? Having a drop-down for game speed directly above the ceasefire slider is a bit ugly.

Overall, I'm not entirely convinced sliders are better (more user-friendly) than drop-down lists, though I don't have very strong feelings either way.

Just confirmed that the crash in debug mode of apitrace is not linked to this patch. So only thing to fix is Mac Os.

Fair enough :)

"SB1" and "SB2" appeared exactly once, in the maur.json file, and were obviously placeholders; other civs leave this string empty ("").
Inserting SpecialBuilding is rather meaningless, because the fact that they're “special” (i.e. not available to all civilizations) is the very reason they're listed in the civ files in the first place.

I assume there is no class for special buildings? (SB1 & SB2)

Successful build - Chance fights ever on the side of the prudent.

First proof of concept with stepWidth

From 2017-12-06-QuakeNet-#0ad-dev.log:

20:14 < temple> stepWidth sounds easier

nani posted more review comments in a PM conversation, most importantly motivating me to create the second iteration of the patch, also he tested the patch.

Successful build - Chance fights ever on the side of the prudent.

"SB1", "SB2"

All Armour values got consistent now, feel free to removing trailing 0's from other template values too

D2000 removes trailing zeros from aura files, D2222 improves the civ files.
For templates, maybe later, preferably after D2502 (adjusting wall footprints) is in, because rebasing that would be a lot of work.

In case phabricator can't handle the images (D889#38774), here's a zip with the textures:

Tested. Works as advertised.

Successful build - Chance fights ever on the side of the prudent.

• Remove redundant comment in JSDOC.
• Use correct way of object property notation.

Classic StatusEffect applies.
StatusEffect with modifiers applies modifiers to target.
When effect has duration, it is removed after duration ends.
StatusEffect with interval deals damage correctly to target.
Tooltip with effect for attack is displayed correctly when hovering over attack details.
Tooltip with applied effect is displayed correctly when hovering over effect icon.

Successful build - Chance fights ever on the side of the prudent.

Successful build - Chance fights ever on the side of the prudent.

Inlines.

Initially, I did 0.90 but then they were slower than Heavy Skirmishers, which didn't feel too correct.

That made me wonder, so I looked up the current walk speed numbers:

                      citizen , champion , hero
cavalry javelineer  : ×1.95   , ×2.25    , ×1.9425 (= 1.85 × 1.05) 
cavalry archer      : ×1.95   , ×2.25    , ×1.85
  patch proposal      : ×1.792  , ×2.07    , ×1.85
infantry javelineer : ×1.4    , ×1.75    , ×1
infantry archer     : ×1.1    , ×1.2     , ×1

The differences are especially pronounced for the infantry javelineer. (Maybe something to revisit in a future patch?)
Regardless, I wonder why champions are significantly faster than citizens. They have a lot more armour and health and can't gather (hunt). If anything, I'd expect champions to be slower or at most equally fast.
The current differencea between citizen and champion walk speed go back to rP20230, which introduced relative speeds (multipliers). Prior to that template_unit_cavalry.xml, template_unit_champion_cavalry.xml, and template_unit_hero_cavalry.xml all had a walk speed of 1.65; the commit message says: “No significant value change.”

Successful build - Chance fights ever on the side of the prudent.