Thoughts on latest release...

Describe your experience with the latest version of FreeOrion to help us improve it.
Forum rules
Always mention the exact version of FreeOrion you are testing.

When reporting an issue regarding the AI, if possible provide the relevant AI log file and a save game file that demonstrates the issue.
Message
Author
User avatar
EricF
Space Dragon
Posts: 315
Joined: Fri Jul 29, 2016 10:12 am

Re: Thoughts on latest release...

#16 Post by EricF » Wed Oct 26, 2016 6:29 pm

Logs after loading save
AI_1.log
(3.71 KiB) Downloaded 12 times
AI_2.log
(45.68 KiB) Downloaded 7 times
AI_3.log
(158.28 KiB) Downloaded 7 times

User avatar
EricF
Space Dragon
Posts: 315
Joined: Fri Jul 29, 2016 10:12 am

Re: Thoughts on latest release...

#17 Post by EricF » Wed Oct 26, 2016 6:30 pm

Logs after loading save
AI_4.log
(173.62 KiB) Downloaded 7 times
freeorion.log
(26.06 KiB) Downloaded 6 times
freeoriond.log
(27.21 KiB) Downloaded 6 times

User avatar
EricF
Space Dragon
Posts: 315
Joined: Fri Jul 29, 2016 10:12 am

Re: Thoughts on latest release...

#18 Post by EricF » Wed Oct 26, 2016 6:50 pm

I loaded at turn 130. Turns 131 to 133 took about 10 to 15 seconds to finish, max.
Then processing turn 133 to 134 it took about 2 minutes.
Then processing turn 134 to 135 it took 3 minutes and 26 seconds.

Logs follow for turn 135
AI_1.log
(689.33 KiB) Downloaded 14 times
AI_2.log
(640.24 KiB) Downloaded 15 times
AI_3.log
(889.15 KiB) Downloaded 12 times

User avatar
EricF
Space Dragon
Posts: 315
Joined: Fri Jul 29, 2016 10:12 am

Re: Thoughts on latest release...

#19 Post by EricF » Wed Oct 26, 2016 6:53 pm

More logs for turn 135
AI_4.log
(884.42 KiB) Downloaded 7 times
freeorion.log
(151.02 KiB) Downloaded 7 times
freeoriond.log
(335.09 KiB) Downloaded 10 times
So, to be clear, yes turns went by fast after just loading, but after just four or five turns, things really slowed down.
Hope this helps.

ovarwa
Pupating Mass
Posts: 92
Joined: Thu Dec 11, 2014 3:22 am

Re: Thoughts on latest release...

#20 Post by ovarwa » Wed Oct 26, 2016 9:44 pm

(It may be that the only change needed for shields is to have the better shield techs automatically upgrade the defense grid)

User avatar
Vezzra
Release Manager, Design
Posts: 4686
Joined: Wed Nov 16, 2011 12:56 pm
Location: Sol III

Re: Thoughts on latest release...

#21 Post by Vezzra » Sat Oct 29, 2016 10:04 pm

Regarding shields being too expensive (particularly the research costs): back when Mat toned down shields, he reduced the shields stats and build costs, but (IIRC) left the original research costs unchanged. As these were balanced for the original stats, they are now too high of course, and should be reduced. I guess Mat never got around to that...

defaultuser
Vacuum Dragon
Posts: 583
Joined: Wed Aug 26, 2015 6:15 pm

Re: Thoughts on latest release...

#22 Post by defaultuser » Sat Oct 29, 2016 10:09 pm

Vezzra wrote:I guess Mat never got around to that...
Cut his salary in half.

User avatar
Vezzra
Release Manager, Design
Posts: 4686
Joined: Wed Nov 16, 2011 12:56 pm
Location: Sol III

Re: Thoughts on latest release...

#23 Post by Vezzra » Sat Oct 29, 2016 10:17 pm

defaultuser wrote:
Vezzra wrote:I guess Mat never got around to that...
Cut his salary in half.
Oh, harsh! :twisted:

User avatar
MatGB
Creative Contributor
Posts: 3310
Joined: Fri Jun 28, 2013 11:45 pm

Re: Thoughts on latest release...

#24 Post by MatGB » Sun Oct 30, 2016 12:10 am

OK, so instead of making, um, zero, I now make, um, zero? Is that right, I was never that good at maths.

But yeah, I rarely if ever touched research costs when balancing in game effects, I was balancing cost/benefit purely in terms of PP/time.

This was partially a confidence thing as it was fairly early in the process when Is tarted doing parts and partially because, well, the entire tech tree is horrifically unbalanced in terms of cost/benefit, especially in regard to hulls and parts, I'd not know whether to compare things to stuff currently undercosted or overcosted.

To be honest, on balance, I think shields are at the right price. It's virtually everything else that's too cheap ;-)

Also, Geoff didn't balance the shield research costs when he introduced the current damage/shot reduction system so they weren't balanced for the old stats either: that was IIRC one of the first new things introduced as I started hanging out on the forums and not just playing every so often.

The only actual signficant change I've made to research cost was the Nanorobotic Hull, because it was always overcosted compared to similar stuff and I toned it down substantially near the end of the last cycle. We need to basically redo virtually the entire tech tree but, um, that's a big task. So it might be easier to do it peacemeal, but given virtually everything changes when Influence comes in properly it's probably best to only tweak the most egregiously out of place costs then.
Mat Bowles

Any code or patches in anything posted here is released under the CC and GPL licences in use for the FO project.

dbenage-cx
Programmer
Posts: 389
Joined: Sun Feb 14, 2016 12:08 am

Re: Thoughts on latest release...

#25 Post by dbenage-cx » Sun Oct 30, 2016 2:21 am

I'd love a keyboard shortcut to close the civilopedia.
Geoff added this in late July. It is mapped to F1 by default and works in map, production and research windows.

ovarwa
Pupating Mass
Posts: 92
Joined: Thu Dec 11, 2014 3:22 am

Re: Thoughts on latest release...

#26 Post by ovarwa » Mon Oct 31, 2016 8:51 pm

Thoughts from playing more, especially regarding the UI:

* The AI is pretty incapable of using (reaching?) the upper half of the tech tree. Collapse a star system into oblivion? Build and exploit starlanes? Etc.

* I found a use for some probably overcosted and underwhelming tech: If you're Gyasche, better weapons mean bigger penalties, one tech advance worth: For most players, a basic Death Ray is 15 and a Laser-4 is 11 but for you it is 10 and 9 damage. Yuck! But this does not apply to the big Cannon. It also does not apply to the bonuses you get from using Organic Ships and lasers and the right tech: It's expensive, but 3 lasers and the special are cheaper than 3 Death Rays; 2 lasers+ is competitive. Most species would prefer the Death Rays (and on better hulls), but not you, not for a long time! And you get it earlier. The Cannon is the long pole, regardless of what hull type you choose (other than Solar), so even though Self-Gravitating Hulls grant an earlier Core Slot than Sentient Hulls, that doesn't matter.

* Every planet should have an icon next to its name in its system window that when you click on it, you go to the production screen for that planet with a *single click*. Or even just clicking on the name would be fine.

* All those little windows close on ESC, but they are not ordered. They should be, latest opened is first closed. This would solve lots of annoyances, such as not being able to close the pedia from the research screen, and losing the sitrep screen when you click on a planet mentioned in the sitrep and then click ESC....

* Does the AI understand stealth?

* The problem with being able to ignore most weapon and hull paths is that there is eventually nothing else left to research when you have 1000rp/turn (and it takes surprisingly few planets to get there!), because the useful techs are tightly constrained by turns. I suppose it might be possible to plan for this, in theory, but that won't happen because other techs are absolutely necessary in the early game. I suspect this isn't by design, and is probably a corner case of not caring overmuch for Death Rays or hulls beyond Sentient.

* Robotic species with broad adaptibility are a very big So What, because Exobots like barren planets just fine. (Robot Interface Shields are pretty eh.)

* I find that I prefer the Organic hulls to asteroids... movement is king. If I'm 1 turn from an enemy I can do things even if he is 1 turn from me; once I'm 2 turns away, my defending force is pinned. A small difference in movement can often have a huge impact. (But I'd still rather have an energy hull.)

* I wish it were easier to understand what an AI player can and cannot see, in terms of stealth. If we can see detection range, why not more? (Although I could also understand changing things to make detection range secret. Still, both human and AI players alike are more likely to use and defend against stealth if the abilities were more visible.)

Anyway,

Ken

User avatar
Geoff the Medio
Programming, Design, Admin
Posts: 12290
Joined: Wed Oct 08, 2003 1:33 am
Location: Munich

Re: Thoughts on latest release...

#27 Post by Geoff the Medio » Mon Oct 31, 2016 8:55 pm

ovarwa wrote:* Every planet should have an icon next to its name in its system window that when you click on it, you go to the production screen for that planet with a *single click*.
Is double-clicking really so much more trouble...?

ovarwa
Pupating Mass
Posts: 92
Joined: Thu Dec 11, 2014 3:22 am

Re: Thoughts on latest release...

#28 Post by ovarwa » Mon Oct 31, 2016 9:58 pm

Hi,
Geoff the Medio wrote:
ovarwa wrote:* Every planet should have an icon next to its name in its system window that when you click on it, you go to the production screen for that planet with a *single click*.
Is double-clicking really so much more trouble...?
Yes. It doesn't seem like it should be. It does seem like I'm being silly about it. (And I should step back from words like 'should': I don't expect you to do any of this! Not like I'm paying you.... )

But think of it this way: Many of your players either have RSI or are well on their way to it, in part because they are playing computer games! :D/20

So cutting the number of clicks *in half* for very common activities is extremely beneficial. (The icon next to the name helps make the option obvious; when there are very few options, having an icon for each next to the them cuts down on clicks even more, and improves speed. Consider an outpost: There is very little one can do there, so an icon next to the name to turn it into a colony and others for the very few buildings that can be built there is a massive game speedup.)

Think of this another way too: Consistency in a UI is wonderful. Pretty much anything else you click on in FreeOrion activates on a single click. So why should this be an exception?

One final thing to consider: The production screen should be retired. It currently serves two functions, better served by functions on the main screen or little popups:

* Tell a planet to produce something. This is best done from the planet.

* Reorder the global production queue. This is best done from the queue window, which can be popped up like a system or fleet window.

These 2 operations are often done separately.

The research screen probably still needs to be its own thing, because the tree takes up so much space and kind of replaces the main map.

BTW, while here some other useful (IMO :) ) UI tweaks:

* Obsolete a ship from the planet production list. A player is most likely to decide "I don't want to see that anymore" when about to choose what next to produce.

* Select a bunch of ships and obsolete them at once, or maybe just an "obsolete everything except the very basic troop and colony ships" option.

So, in combination with LIFO windows, producing something involves clicking on a planet, either from the objects window or the planet's system window. That brings up the planet's production/status windowlet. A single click or keystroke shortcut can bring up the production queue if needed, whether or not the planet's production window is open.

Right-clicking on an item in the global queue currently allows the item to be moved around in the queue or deleted; one more option that could be added is to open up the planet's production&status window.

ESC closes these things in the order they were opened. A close-all shortcut closes all the windows.

Just me.

Anyway,

Ken

User avatar
MatGB
Creative Contributor
Posts: 3310
Joined: Fri Jun 28, 2013 11:45 pm

Re: Thoughts on latest release...

#29 Post by MatGB » Mon Oct 31, 2016 10:47 pm

ovarwa wrote:Thoughts from playing more, especially regarding the UI:

* The AI is pretty incapable of using (reaching?) the upper half of the tech tree. Collapse a star system into oblivion? Build and exploit starlanes? Etc.
I've seen the AI collapse red stars to make black holes, but the black hole collapser and starlane bore stuff are very new features that haven't had a proper balance pass yet let alone have the AI setup to use them: the latter will be awhile for the AI as understanding of basic topography is currently very limited anyway.
* I found a use for some probably overcosted and underwhelming tech: If you're Gyasche, better weapons mean bigger penalties, one tech advance worth: For most players, a basic Death Ray is 15 and a Laser-4 is 11 but for you it is 10 and 9 damage. Yuck! But this does not apply to the big Cannon. It also does not apply to the bonuses you get from using Organic Ships and lasers and the right tech: It's expensive, but 3 lasers and the special are cheaper than 3 Death Rays; 2 lasers+ is competitive. Most species would prefer the Death Rays (and on better hulls), but not you, not for a long time! And you get it earlier. The Cannon is the long pole, regardless of what hull type you choose (other than Solar), so even though Self-Gravitating Hulls grant an earlier Core Slot than Sentient Hulls, that doesn't matter.
Yup.

When I changed the refinement bonuses from being %ages to be flat numeric changes it was mostly for balance reasons as Misiorla with Death Ray 4s were ridiculously overpowered and Mu Ursh/Eaxaw were close to that. A side benefit is that the different weapons can be more/less worth the cost depending on species meaning that Bad Weapons species can make different tech tree choices to their advantage.

Re the SAM, you're right, it's probably, currently, too expensive, but I haven't tried to balance it cost/benefit yet as we've discussed adding some extra 'bombard' features to it, it's a big enough gun to be a planet killer ;-)

Re the Solar Concentrator, it is, deliberately, too expensive at the moment, it's roughly balanced at about 50 pp production cost not 80pp as it currently is. It will go down in cost as/when the AI is able to take it into account when a human is using it, and possibly go down a bit again when the AI is also able to use it. But the AI team are busy working on far bigger and more important things and we agreed updating them for this, given it's a non-simple task, was a very low priority for now.
--
snip UI stuff I can't help with
--
* Does the AI understand stealth?
Sort of, but not very well. It knows that you can have hidden stuff, and is aware when it's got hidden stuff, and tries to remember if any given system may or may not have hidden stuff in, but it does all of this at a very basic level. It definitely doesn't try to play stealth as a strategy, nor does it currently give a higher priority to detection techs if it goes up against someone using stealth.

Getting this fixed is something planned, but we need to also do some tweaks to stealth to mkae it a more interesting/viable strategy as well: improving the AIs understanding of stealth and detection techs in priority setting is definitely, for example, a higher priority than the solar concentrator and terrain effects in general (partially as stealthed fighter carriers are lethal against them when they're introduced to master, coming soon we hope).
* The problem with being able to ignore most weapon and hull paths is that there is eventually nothing else left to research when you have 1000rp/turn (and it takes surprisingly few planets to get there!), because the useful techs are tightly constrained by turns. I suppose it might be possible to plan for this, in theory, but that won't happen because other techs are absolutely necessary in the early game. I suspect this isn't by design, and is probably a corner case of not caring overmuch for Death Rays or hulls beyond Sentient.
The tech tree is currently not at all balanced, some of the higher end techs are too cheap, you can get a stupidly high tech output near the end, etc.

Basically, yes, you're 100% correct here and rebalancing it all is planned: we have a fairly major project in the works introducing a feature called Influence, this will change the balance of pretty much everything cost/benefit wise and recosting the tech tree taking into account the effect Influence has is going to happen after that.
* Robotic species with broad adaptibility are a very big So What, because Exobots like barren planets just fine. (Robot Interface Shields are pretty eh.)
Again, sort of, yes, I think the techs that adapt you to planets are too powerful in many ways meaning being merely Adequate is more than enough, this is again something I want to balance but is a lower priority than the tech tree.

Ideally the ability to colonise a Poor world then terraform it to Good should be a significant boost for a species over having far fewer worlds you can initially colonise, but currently it's not as good as I think it ought to be.
* I find that I prefer the Organic hulls to asteroids... movement is king. If I'm 1 turn from an enemy I can do things even if he is 1 turn from me; once I'm 2 turns away, my defending force is pinned. A small difference in movement can often have a huge impact. (But I'd still rather have an energy hull.)
Agree completely, I far prefer a fast, fragile hull that makes me thing over a slow, strong hull that's harder to destroy, but that's a personal preference, I have used asteroid hulls very effectively to win, and in many respects if you've got a good system with a Belt in early enough they're a VERY powerful line, but they don't suit my playstyle.
* I wish it were easier to understand what an AI player can and cannot see, in terms of stealth. If we can see detection range, why not more? (Although I could also understand changing things to make detection range secret. Still, both human and AI players alike are more likely to use and defend against stealth if the abilities were more visible.)
The UI here does need some work, but again it's not an incredibly high priority. Relatively recent feature, on the Empires window, each empire has a column showing their detection strength. I have mine set to display at a size just big enough to show detection strength and in the bottom left of my screen, where the fleet window opens but doesn't cover the strengths, so I can easily see which colour has what detection.

If you know what stealth value your stuff has (which is displayed fairly well and you can easily work it out if you know your tech/designs) then you can tell what they can see, but it requires paying attention in an unintuitive way.

Given I love playing with stealth I would love it if someone suggested and/or implemented an easy UI method of telling you what your opponents can see of your stuff as it is fairly important. But I haven't come up with any ideas myself and I wouldn't be able to do the code even if I did.

Again, thank you for this feedback, you've given me an idea ;-)
Mat Bowles

Any code or patches in anything posted here is released under the CC and GPL licences in use for the FO project.

ovarwa
Pupating Mass
Posts: 92
Joined: Thu Dec 11, 2014 3:22 am

Re: Thoughts on latest release...

#30 Post by ovarwa » Mon Oct 31, 2016 11:08 pm

The techs that adapt you to planets are among the best in the game. I think. I'm new at this. :)

Both Terraforming and Gaia Transformation are underwhelming, except perhaps on sparse maps: Why spend industry to modify your planet when you can spend industry to find or (better) conquer another?

Two crappy planets are better than one good planet for most of the game.

---

I liked Asteroids a lot 2 years ago: Their cheapness and durability was sufficient for me to utterly not care about the speed, though I noticed it. Now that that they have been nerfed... not so much.

I suspect that it would serve the game to balance all four hull trees, so that all had useful ships at the very beginning (right now, there's only robotic) and all at the very end (solar > titanic > scattered > sentient; the last two are closer to midgame).

Yes, wait for Fighters :)

One interesting way to do this might be to move stuff from non-hull branches of the tech tree to the hull branches. This already happens with Organic Hulls: By the time I reach Sentient Hull, I automatically get some good growth techs. Doing the same for some of the other techs, or creating new techs of that kind and scatter them along the 'hull' tech... Or maybe not: It would just mean that everyone would have to research all the hulls and then only use the best.

Anyway,

Ken

Post Reply