Modtool Slowness

[89157Z]

Is there anything to be done to make the mod tools faster? This game has lots of potential for modding. I suspected that well before I'd seen the tools, and I believe that just as much now that I have seen them. But making mods is very difficult right now due to the slowness and unreliability of publishing mods. I wish that were faster. It'd be nice if the interface was faster inside itself too, but that's less critical. If the system worked more quickly, there would already be released mods. I've got one I could release if I could publish it! But so far, I haven't successfully published it.

 

Of course, I'm probably only feeding the Land of Fishes and Wishes by saying this.

[Legate of Mineta]

89157Z;

 

Unfortunately, no- it takes very long to publish a Mod because for every entity in the database the relationships are checked and re-checked with all other entities. It's definitely not efficient as compared to earlier versions of the tools- but there were too many issues with inconsistencies and missing data.

[89157Z]

What about incorporating a 'testmod' option that creates a quick version for testing purposes? I'm worried that the overnight endeavor of publishing a mod will, when I've successfully done it, only cap off with a broken mod because I made some error. Then I've got to go in and fix it, and if I'm not sure what I did, I might end up having to do that publishing step several times.

 

Not to mention that you might want to playtest a mod.

[Legate of Mineta]

89157Z;

 

We've considered it, but I'm not sure we will touch the Mod Tools in Year 1, because of the risks involved for normal merging. We're definitely listening to the feedback for Year 2, however.

[89157Z]

I guess that means I'll just have to sit on my hands.

[Legate of Mineta]

89157Z;

 

I personally suggest sleeping while the mod is building, but...

[89157Z]

The mods crash before completion. I haven't successfully built one yet.

[Legate of Mineta]

89157Z;

 

Ah, out of memory exception? Yeah, there's nothing you can do about that, if you do not have the memory.

[tapsyhapsy]

I've had the publishing running for nearly 12 hours when it gave me an error (something along the line of 'Incorrect object referenced' or whatnot) and *POOF* it was gone. I see two problems here.

 

1) I get absolutely no help figuring out exactly where I have something incorrect referenced, unless you can tell me where 'around 48% the last time I checked' is in the relational database. Please, if it said 'The magic spell broke' it would have been just this useful. That is, not at all.

Plus, I only added two actions and a Background, taking painstaking care to make sure that 'Self' was the target of all and every piece of stuff added. So, what's breaking my mod?

 

2) 12 FREAKING HOURS FOR HALF OF THE PUBLISHING? I'm sitting on a Q6600 here, and while it's not the newest CPU out there, it used to be a top-of-the-line one some time ago. It's still by no means a weak CPU. Runs 99% of games that are out there (barring GPU issues). I even gave it realtime priority on one of the normally unused cores, and let it run amok with the memory too, but we're talking about 12 hours. It used over 2 gigs of memory, which it was welcome to use, since there was some left over and change to run the rest of the PC.

 

How I am supposed to attempt to even created a mod

 

- without at least a simple manual

- with 24 hours+ of testing time for the simplest of modifications

- with unhelpful and useless error messages

 

I used to mod Sins of a Solar empire a lot, and while I admit that the differences between the two games are significant, I'd like this to be asked: How on Earth can anyone think of mod creating without an immediate testing feedback on the effects in has on a game? And I'm not even talking about playing it in the actual game...

 

Oh, and there is really a 'Cancel' button missing to the publishing process.

[Legate of Mineta]

TapsyHapsy;

 

We agree- although it takes less time for us to publish (about 4-5 hours), it's still a very significant time. Unfortunately, it probably will not be improved prior to the Year 2 Tools.

[tapsyhapsy]

TapsyHapsy;

 

We agree- although it takes less time for us to publish (about 4-5 hours), it's still a very significant time. Unfortunately, it probably will not be improved prior to the Year 2 Tools.

 

I've temporarily solved the issue by publishing it as official content. 20% done in 30 minutes, so that should be fast, hopefully no errors to pop up while I get some z's.

 

By the way, if you're looking for feedback and suggestions on modding tools, and noob-friendly modding I'm happy to help ^^

[Legate of Mineta]

TapsyHapsy;

 

Haha, much appreciated.

[retired_deer]

On this subject, would it be possible to have some manually coded alternative? I am thinking about the modding done for Europa Universalis 3, all done in text files and then integrated into the system. Granted, you would need a thoroughly good understanding of the database system utilized by the program. But if there was a text-to-.mdm file converter, could it be done? Also, if that were possible, could the system automatically integrate it into its core data? For instance, say I have a mod in the text file. Could the program rapidly compile the data in a specific text file each time the game started? I am thinking this would allow continuous, immediate modding, rather than waiting hours on end. All hypothetical, however.

[Legate of Mineta]

Retired;

 

It is probably possible, but not nearly as friendly as Paradox titles, unfortunately.

[Zaltys]

Some error checking would probably be a good thing.

I've given up on modding for now. Tried to add two new spells, the first five mods crashed when I attempted to cast those. So I tried chopping them into smaller spells to try to figure out which part is broken. Now the mod crashes whenever I try to load it.

 

Considering that it takes so many hours to publish a mod... I've now spent a week without making any progress.

 

(The first spell was supposed to be a more erratic version of From Nothing, Knowledge: casting it would improve your own skills but also leak over and give a skill step to a completely random student - and possibly to a random familiar. The second spell was supposed to give a large temporary buff to random skills, for the whole clique. What can I say, I like randomness and relying on dumb luck. )

[Blind Gardener]

Just curious here but... what is your big O efficiency for mod-tools? Would it be possible to build a version that 'flags' new or changed entries while you work, and only checks them and their references for inconsistencies and differences in order to publish them?

[Legate of Mineta]

BG;

 

It's possible, and we've debated it internally, but the issue here is the number of connections each entity has with others. To be 'safe' in the absence of a real study and implementation of the rules necessary to insure integrity, we check everything...with the slow results.

[Schwarzbart]

OK that answears why my small change in a single action need now over 1h per 10% (Used the amo as base to create this)

[Schwarzbart]

Do I understand it correct that create Mod creat a mod with the delta betwean the amo and your mod? If so is there a way to make a amo out of the ModbaseX.amm?

[Legate of Mineta]

Schwarzbart;

 

I don't think you can make an AMO, per se- put you can publish an AMM as Official Content, which is a much faster publish.

[Schwarzbart]

I know but I prefer to have my mods so build that the work even with future DLCs and if I want to use one of the new Skills I have a problem with just working with the AMO.

[Legate of Mineta]

Ah yes, that's the disadvantage of it. ;(

[Zangi]

Holy bajeezus. I added one change and published. +6 hours.

[Legate of Mineta]

Zangi;

 

The Year 1 Tools are very, very slow. We (internally) have a new version which is much faster, but we can't release that yet.