FreeOrion

Forums for the FreeOrion project
It is currently Sun Nov 19, 2017 4:49 am

All times are UTC




Post new topic Reply to topic  [ 34 posts ]  Go to page Previous  1, 2, 3  Next
Author Message
 Post subject: Re: 0.4.6 roadmap
PostPosted: Thu Apr 21, 2016 3:01 pm 
Offline
Release Manager, Design
User avatar

Joined: Wed Nov 16, 2011 12:56 pm
Posts: 4241
Location: Sol III
Morlic wrote:
Mandatory:
-Fighters branch needs more adjustments if we put it in.
General consensus so far seems to lean more toward not include the new fighter/carrier mechanic into 0.4.6. Originally I'd wanted to include it, but now I also think we should take the more conservative approach and leave it out.

So, unless anyone objects, I'll flag fighters/carriers as post 0.4.6.
Quote:
High priority:
- The AI is not adapted for supply propagation changes at all. Probably should. Probably also non-trivial, but I have to play more games to actually decide what is good to do and what not. Then again, I can't remember complaints on how the AI sucks at the new system, so potentially not too important and small, simple adjustments are just fine.
Ok, this is important, and definitely needs to be done. Especially given Mats assessment:
MatGB wrote:
It does OK, currently, unless the human player does something to take advantage of the lack of knowledge of how it works, if you get a good supply species in or near their border and/or conquer a smallish world, and then put that world on Logistics, you can fairly easily completely isolate all production centres and then walkover them once you've won eliminated the main fleet, as planets don't even share resources in system if it's not in your colour. They're not awful, but that needs looking at.
That's serious, and needs to be addressed.

We've got two options here: either the AI team manages to adjust the AI to the point where it can handle this properly, or we need to put in a stop-gap solution by adjusting the supply propagation mechanic. The issue here is that if you have stronger supply propagation into a system with enemy colonies than that enemy, you completely overwhelm his supply, and his colonies can't share supplies anymore (not even the colonies in that system among themselves) - and all that without you having any colonies of your own in that system.

The temporary adjustment could be that if you have a colony in a system, no enemy empire can supply propagate into that system and overwhelm your supply, regardless how strong its supply propagation would be into the system - unless such an enemy empire has at least an outpost in the system.

Whatever you guys prefer/can be done within the next two months. Although posting a cheap armed ship in a "supply contested" system should be sufficient to deal with the problem, and that probably shouldn't be to difficult to get the AI to do?
Quote:
- Basic adjustments to new drydock logic if Mat changes more of it...
Yep, that's a must.
Quote:
- Weapon upgrades need some more minor adjustments (should not be too much work)
Ok, will go onto the list too.
Quote:
- The AI research path needs to be checked and adjusted (feel free to make suggestions if you think specific techs are researched too early/late etc.)
Ok. @Everybody: whoever comes accross or notices something of this kind that they think should be addressed, please open a respective issue on github.
Quote:
Low priority:
- I have some mostly finished reworked shipyard logic which works seemingly fine lying around since December. May or may not be included.
Ok, your call. If you think you can finish/wrap it up fairly quickly, I can put it on the list. But only if you can fit it in after taking care of the high priority stuff. In case of doubt better leave it for after 0.4.6.
Quote:
- There is some WIP research rework. Currently doesn't really seem to make viable decisions. I don't know if anyone is working on that at the moment.
Well, then better leave it out.
Quote:
Some things I wouldn't plan to include in the release that were mentioned (unless someone else wants to do it or has already started working on it):
- Stealth/Detection implementation
Agreed. Sounds like a bigger project, and we should not start any more of these before the release.
Quote:
- Improved colonization logic
Depending on what exactly you mean by this and how much effort it would require... to be more precise: if this is about adjusting the AI to better deal with the "new" colonization mechanics (already introduced in the last release, so not really "new" anymore), and it could be done with a moderate amount of effort, then I'd like to see it done.

Otherwise (if you're talking about a more general optimization of colonizing logic, and/or that will require some serious effort), agreed, lets keep it out.
Quote:
- Support for some crazy complicated parts such as solar webs. Don't expect it to happen any time soon.
Agreed, lets leave it out.
Quote:
- At least 785 other things which need improvement including pretty much any part of the game and spanning the entire range of complexity and concerning things that were added "recently" as well as things that were surely introduced before 2010.
Well, um , yeah... :lol:

No problem. Things will stay that way until 1.0 I guess... ;)
Quote:
I definitely won't have any time for FO development until end of month though.
I see. Well, with Dilvish back, but still not around that much and probably having to do a lot of catching up, that leaves Cj, and his time is limited too as far as I can see. That means, if we want to release in June, the AI team has basically a month (May) to wrap things up. I can't really judge how much effort all those items above are going to require, your guess is probably far better than mine.

Adjusting the AI to these changes need to be given priority: new supply propagation mechanic, new weapon upgrade mechanic, revised drydock repair rates, adjustment of the AI research path where necessary. Do you think you can get that done until end of May/beginning of June? If you can fit in anything else, fine, but only if you think you can get it done in May.


Top
 Profile  
 
 Post subject: Re: 0.4.6 roadmap
PostPosted: Thu Apr 21, 2016 7:56 pm 
Offline
Psionic Snowflake

Joined: Wed Aug 26, 2015 6:15 pm
Posts: 492
I haven't been using the test versions, so I will looking at switching to the new release when you get it finalized.

This is a bit off the wall, but have there been any changes to Gateway to the Void? Mat mentioned some time back that it was something that needed review. I suspect not, as it doesn't seem like a popular feature but I do actually use it on occasion.


Top
 Profile  
 
 Post subject: Re: 0.4.6 roadmap
PostPosted: Thu Apr 21, 2016 9:09 pm 
Offline
Creative Contributor
User avatar

Joined: Fri Jun 28, 2013 11:45 pm
Posts: 3259
defaultuser wrote:
I haven't been using the test versions, so I will looking at switching to the new release when you get it finalized.

This is a bit off the wall, but have there been any changes to Gateway to the Void? Mat mentioned some time back that it was something that needed review. I suspect not, as it doesn't seem like a popular feature but I do actually use it on occasion.

There have been no changes, it might get reviewed at some point, IIRC the code is outdated and needs cleaning up but there's nothing wrong, in principle, with using it. If you like cowardly tactics ;-) (It really doesn't suit my playstyle, at all, I've tried to use it, but I end up sending a fleet out past it and invading places anyway...

_________________
Mat Bowles

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


Top
 Profile  
 
 Post subject: Re: 0.4.6 roadmap
PostPosted: Thu Apr 21, 2016 11:14 pm 
Offline
Psionic Snowflake

Joined: Wed Aug 26, 2015 6:15 pm
Posts: 492
Where I find Gateway to the Void helpful is for inconvenient Experimentor situations, and the rare occasion when an AI has gotten out ahead of me. I can use gateways rather than fleets to blockage a few checkpoints


Top
 Profile  
 
 Post subject: Re: 0.4.6 roadmap
PostPosted: Fri Apr 22, 2016 2:26 pm 
Offline
Release Manager, Design
User avatar

Joined: Wed Nov 16, 2011 12:56 pm
Posts: 4241
Location: Sol III
I made a first past going over all the currently open issues and PRs on github, and for each item tried to decide on proper assignment to 0.4.6, post 0.4.6 or leaving undecided/optional.

Please review these assignments. You can easily set filters to only show items with milestone "0.4.6", "post 0.4.6", and even "no milestone", which will give you all items which I didn't decide on yet, or decided they can be optional, or don't apply for release milestones because they deal with documentation/"infrastructure"/administrative/whatever issues. Please post your feedback, objections, comments here, so we can decide on adjustments to these assignments. Feedback particularly on the items not assigned to one of the milestones is of course most important (because that can mean a decision has yet to me made).

I will make a second pass where I will review all the feedback already given in the discussion above, and where applicable/reasonable create appropriate issues for all the todo items we've collected and agreed upon so far. In this second pass I will also incorporate the feedback given to my first pass (and this post), as far as we came to any decisions of course.

Everything that ends up tagged for the 0.4.6 release will need to be addressed/solved and closed before we can start producing release candidates.


Top
 Profile  
 
 Post subject: Re: 0.4.6 roadmap
PostPosted: Fri Aug 12, 2016 1:27 am 
Offline
Krill Swarm

Joined: Fri Oct 23, 2015 12:54 am
Posts: 14
Sorry for asking here in this thread, but how is 0.4.6 coming along? Is there a planned release date?


Top
 Profile  
 
 Post subject: Re: 0.4.6 roadmap
PostPosted: Fri Aug 12, 2016 9:53 am 
Offline
Release Manager, Design
User avatar

Joined: Wed Nov 16, 2011 12:56 pm
Posts: 4241
Location: Sol III
0.4.6 is coming along a bit bumpy, due to several key devs being completely/partly absent, very busy etc. during the last months. The main holdup being important AI adjustments to various changes/additions having to be addressed, with all AI devs except one inactive. Now Morlic is back, at least has been able to do some work again, so the AI team is up to two now.

I've set another deadline, Sept 25th. We should really get the release done until that date, at the absolute latest, because then I'll be gone for a roughly month, which would cause another massive delay. So I hope very much that we can pull that off, although we are already cutting it close (I mean, it Aug 12th, and we haven't even created the release branch). But I still think we will make it. :D

On the other hand, it's not that we didn't get stuff done these past months. There have been a lot of issues brought up and addressed, so I think we got a lot of ground covered which might allow us to cut the release candidate testing phase short.


Top
 Profile  
 
 Post subject: Re: 0.4.6 roadmap
PostPosted: Mon Aug 15, 2016 4:29 pm 
Offline
Release Manager, Design
User avatar

Joined: Wed Nov 16, 2011 12:56 pm
Posts: 4241
Location: Sol III
Alright, it's Aug 15th, we haven't even managed to get to the point where we can create the release branch, and the list of open issues and PRs that need to be addressed/merged for the 0.4.6 release keeps growing instead of getting shorter. Assuming that we're still firm in our decision to get the release out before my autumn absence, we're running out of time here. I've set a deadline for the release, Sept 25th, which is the absolute latest point at which we have to get the release out, after that I'm gone for a few weeks. Unless someone can take over all the release management stuff for me (which includes producing the builds of course), that would mean another big delay (probably until the end of the year).

In order to meet that deadline, I think we need to get everything wrapped up at least to the point where we can create the release branch and the first release candidates at the end of August. Which is two weeks from now. That's awfully tight for the list of currently open issues and PRs. I don't think we can address all those issues/get all these PRs merged, so we need to start cutting corners, meaning, chosing what is absolutely essential and what we can drop without producing a basically broken release (which would be pointless). I also need everyone to think if there is anything absolutely essential for the release for which no issue/PR has been opened yet, and do so if something comes to their mind.

Please provide your feedback here on the above questions. Geoff, Mat, Marcel, Morlic, Cj, your input is most important as the currently most active core team members. DBenage and LGM-Doyle, as our most active new contributors, your input is also very important. Everyone else (and everyone I've forgot in my confused state of mind) is very welcome to give their opinion too.

Here are my two cent on the absolute essentials:

  • All issues/PRs addressing game breaking bugs (meaning, those which are tagged as "bug" and don't deal with just cosmetic stuff or moderate issues that might be annoying at worst).
  • Issue#632 and PR#814 which addresses it.
  • Issue#630. That one is really critical IMO, and contrary to #632 there is still no PR up addressing it. Morlic, Cj, I don't know if any of you can cook up at least something temporary/a stop-gap solution that is good enough for release. If not we need to fall back on the solution I already mentioned in a previous post above, temporarily change the new supply propagation mechanic to not propagate supply into a system where an enemy empire has colonies, but no ships, and where there are no own colonies. Mat pointed out how that can be exploited too easily against the AI currently, so without any adjustments to address that, we need to adjust the mechanic. Geoff, can that be done reasonably easily?

A lot of the stuff not considered as absolutely essential which is close to be addressed/finished/merged anyway will get in of course if the respective PRs get merged before the creation of the release branch. Which will happen once everything we deem absolutely essential has been addressed/merged.

To add a bit urgency, I declare Sept 4th as the absolute deadline for the creation of the release branch and RC1. If we don't meet that deadline, we need to seriously reconsider the release date for 0.4.6 (because Sept 25th will become very difficult to meet in that case). So, in order to avoid that, I appeal to all of you to focus all your efforts on the essential stuff for the next weeks, and lay aside everything else (as much as you might prefer some of that everything else stuff ;)). Unless of course I'm overruled and you guys decide to just delay the release instead of cutting any corners. That's a reasonable option too, so if you'd prefer that, please say so.

Ok, now let your feedback, comments and opinions come! :D


Top
 Profile  
 
 Post subject: Re: 0.4.6 roadmap
PostPosted: Mon Aug 15, 2016 9:07 pm 
Offline
AI Contributor

Joined: Tue Feb 17, 2015 11:54 am
Posts: 224
Vezzra wrote:
Issue#630. That one is really critical IMO, and contrary to #632 there is still no PR up addressing it. Morlic, Cj, I don't know if any of you can cook up at least something temporary/a stop-gap solution that is good enough for release.


I don't know it either. I have started some work on it but I am really low on time at the moment and changes need to be done in parts of the code I am not too familiar with and also in the dark part of the AI code with the 500-1000 line functions spaghetti code...
So, can't really make any promises nor predictions on that front. It all depends on how much time I can find in the upcoming weeks and how merciful the flying spaghetti monster is when trying to unwind its children... I'm trying my best though!

_________________
If I provided any code, scripts or other content here, it's released under GPL 2.0 and CC-BY-SA 3.0


Top
 Profile  
 
 Post subject: Re: 0.4.6 roadmap
PostPosted: Fri Aug 19, 2016 12:27 pm 
Offline
Release Manager, Design
User avatar

Joined: Wed Nov 16, 2011 12:56 pm
Posts: 4241
Location: Sol III
Morlic wrote:
I don't know it either. I have started some work on it but I am really low on time at the moment and changes need to be done in parts of the code I am not too familiar with and also in the dark part of the AI code with the 500-1000 line functions spaghetti code...
No problem, please just take it easy. You've got enough on your plate, I really don't want to put pressure on anyone, I just need to know what I can expect. Your assessment of the situation doesn't sound as if it's going to be easy to adjust the AI in time, so I'd rather propose to change the most critical part of the new mechanic (supply propagating in enemy systems even if I have no ships and no colonies there), assuming that's an easily doable stop-gap solution.

Geoff, how much effort would that be? You're the one who implemented the new supply propagation, do you think changing that part (so that supply does not propagate into a system unless an empire has at least an armed ship or colony there) can be done quickly?


Top
 Profile  
 
 Post subject: Re: 0.4.6 roadmap
PostPosted: Fri Aug 19, 2016 12:57 pm 
Offline
Release Manager, Design
User avatar

Joined: Wed Nov 16, 2011 12:56 pm
Posts: 4241
Location: Sol III
So, there hasn't been any feedback or reply except from Morlic so far. Guys, I really need your input on this, otherwise I just have to go ahead and make all those decisions without that input. Not how I'd prefer to do it, but if you're fine with whatever I decide, so be it. ;)

I'll give it a bit more time (essentially tomorrow). I plan to go over all the issues and PRs tagged for the release again on Sunday and remove the "Release v0.4.6" milestone from any issue and PR I don't regard as absolutely essential. Which, as I want to repeat once again, does not mean they can't go in for 0.4.6. It just means that I won't block proceeding with the release as planned if those issues/PRs aren't addressed/merged in time. So, if there are any particular issues and/or PRs you think should/must remain tagged for the release, speak up now! 8)

To restate the schedule:

  • Deadline for creating the release branch (and subsequently RC1): Sept 4th. That is the absolute latest date, if possible I want to do that as soon as all things tagged for the release have been addressed/merged. I'll give a proper warning in advance, so if anyone is still working on non-essential stuff, they get an opportunity to object and request more time (but not beyond Sept 4th!).
  • Deadline for the final release: Sept 25th. After that I'm in trouble, because I will be on the road. I can't say when I'll have internet access and time to get any stuff done, but most importantly I won't have access to my windows machines, so won't be able to produce any FO builds for Windows. All of which will make release management quite difficult, and those days are supposed to be my vacation, so please give me a break... ;)

Any comments/objections?


Top
 Profile  
 
 Post subject: Re: 0.4.6 roadmap
PostPosted: Fri Aug 19, 2016 11:26 pm 
Offline
Creative Contributor
User avatar

Joined: Fri Jun 28, 2013 11:45 pm
Posts: 3259
The only 'scripting' issues currently are minor, in some cases just reword the stringtable. There are always going to be balance issues,b ut nothing game breaking or that I want to do major work on now as it'd require AI attention as well—it would be nice to get the Hyperspatial Dam fix in if the AI team can look at that PR and confirm if it can just merge.

Re other stuff, the only problem that I think is big enough to block is the AI/supply propagation issue, it is very easy to overwhelm their supply and simply stop them from being able to build/respond to you at the moment and that makes for a less challenging game—I look forward to setting up to do that as a challenge and competing for border systems as it has the potential to be fun, but currently it's just too easy and thus not good for the game.

My only outstanding work is Gateway to the Void, which I might just rework minorly and leave for post release for a proper going over.

The biggest holdup is the AI, and that's something I simply can't help with except in the form of feedback, but if the team want me to test code, etc I'm very happy to do so.

Given the delays, the hope of getting this Release done early and then Fighters before Xmas is pretty much dead, which is a shame, but life got in the way for all of us so can't be helped.

_________________
Mat Bowles

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


Top
 Profile  
 
 Post subject: Re: 0.4.6 roadmap
PostPosted: Sat Aug 20, 2016 8:32 am 
Offline
Programming, Design, Admin
User avatar

Joined: Wed Oct 08, 2003 1:33 am
Posts: 12013
Location: Munich
Vezzra wrote:
Geoff, can that be done reasonably easily?
Does this work?
https://github.com/freeorion/freeorion/pull/896


Top
 Profile  
 
 Post subject: Re: 0.4.6 roadmap
PostPosted: Sat Aug 20, 2016 1:38 pm 
Offline
Creative Contributor
User avatar

Joined: Fri Jun 28, 2013 11:45 pm
Posts: 3259
Appears to do the job for me, but only a very quick load savegame and look around test.

_________________
Mat Bowles

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


Top
 Profile  
 
 Post subject: Re: 0.4.6 roadmap
PostPosted: Sat Aug 20, 2016 7:30 pm 
Offline
Release Manager, Design
User avatar

Joined: Wed Nov 16, 2011 12:56 pm
Posts: 4241
Location: Sol III
Geoff the Medio wrote:
Yes, perfectly.

I did a quick test where I used a save of a 100+ turn game and massively pumped up supply for my starting species in the content scripts. The result showed quite nicely how my supply was able to spread everywhere until it met a system with armed enemy ships and/or enemy colonies. It also showed that the breaking into enemy supply lines with superior supply tactic still works, if your supply can spread into systems between your enemy colonies where he hasn't any forces to stop your supply propagation.

But that's a lot harder than without that patch, so I guess it's good enough to alleviate the problem for the release. Unless Morlic objects and says he has already something ready (or almost ready) to adjust the AI to the current implementation, I vote for merging that fix as it is.

@Morlic: Is that ok with you? Or do you have a strong preference to try to fix the AI for the release without Geoffs fix? Otherwise I'd say better focus on the remaining things concerning the AI, even if they are not that critical, but probably far easier and quicker to fix/adjust.


Top
 Profile  
 
Display posts from previous:  Sort by  
Post new topic Reply to topic  [ 34 posts ]  Go to page Previous  1, 2, 3  Next

All times are UTC


Who is online

Users browsing this forum: No registered users and 1 guest


You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot post attachments in this forum

Search for:
Jump to:  
Powered by phpBB® Forum Software © phpBB Group