It sure is, and considering the amount of bugs we have in SKYRIM we can consider that they have weak testers.

I'm trying to get you to see that what you're doing is not representative of an actual use case; what you're doing is something the user will never, ever do. And your methodology is not something that needs to be propagated to other scriptwriters. I'm worried less about you and more about anyone else reading this. And strictly speaking, what you think, and what is right, are logically unrelated terms.

And it's not like the cost of using a clean save game is that high. You hit the continue button. What you're putting forth is that scripts have an innate propensity to break when performing the act of loading a save game that has never seen the script in question, which doesn't make sense. If that were true, all of Skyrim would be broken.

Either way, I'm done arguing the point. You either accept, fundamentally, that what you're doing is not an actual use case (forget about the technicalities of the game, save game architecture, script behavior, so on), or you do not. And if you're fine testing your content with invalid use cases, that's your prerogative.

Edit: I would also say that there is nothing wrong with using a COC from menu method to do prototyping or very early testing. My primary point is that that test case should not be the final indicator of the functionality of the script; it has to be tested with a real use case. The more, the better.

I never said it was an use case.

I test like that because I'm always changing my scripts, so its the only option I have for testing. When I reach the use case I will start a game, play a little bit, then add the mod and test. Like I said, different scenarios require different testing.

Maybe that IS the problem, maybe Beth IS COCing and why they keep breaking things every patch!

They have to coc, and if coc is breaking things then they have to fix that same things.

How come it's the only option? I can think of (and use) two different options:

1) In game, I use: ReloadScript
2) I make a "clean save" where I need to load, so for ScenicCarriages, I have a clean save made in the stables right next to the carriage. I load the save and walk over to the driver.

1-Might work
2-Won't work on some script changes/delete/etc

Since I don't want to risk I don't use saves at all, at least until I reach the "user part" you guys are talking about.

Might work - as in the reloadscript functionality doesn't work? (It does, I find it very useful!)

I'd like to undestand you as I don't understand what scenario loading a save game wouldn't work? (I am really curious as I'll be the first to admit there's a lot about modding Skyrim I still need to learn.)

When your script is loaded for the first time on a save game that has never seen it, it fires its OnInit() block and instantiates. To your save game, there is no such thing as a change, delete, or edit. It never existed to start with. For an edit to mean something, it would have had to exist previously. The technical issue you're referring to only applies to changing a script that has already been instantiated in a save game. A clean save will re-instantiate your script each and every time you load it.

In my testing I've found that scripts stick a lot to saves, so, in some cases I was running absolutely NUTS just to find out that everything was already ok 1000 changes to the script ago, so what was ruining what I was seeing in the game? Oh yes, the save! That also provoked many differences between what my users saw and what I saw, because at some point they have a different version of my script in their saves.

After I've learn that, the hard way, I've told myself "NEVER EVER use a save again while testing", so I delete all saves while making my mod, I don't have saves neither I have other mods, my testing scenario is 100% vanilla + my changes.

That said, when I have everything ready for release (eventually...) I will do what you guys are talking about, start a game, play a little bit and save, then add my mod save again and do a little bit testing, but until then no saves whatsoever.

Exactly, we keep talking about the same things

@http://www.gamesas.com/user/617680-chesko/

Offtopic: looking into some of your scripts atm, the array ones.

This thread seems to have wandered off into a discussion about the technicalities of testing. The dialogue problem may or may not have been fixed for ESPs, but that is only tangentially related to the point of this thread.

Does anyone have any evidence that the release of 1.7 has introduced any new incompatibilities with the current CK? Or are any problems being seen just the existing problems which are well known about?

I think all modders would like a friendly heads-up on new problems that they haven't yet stumbled over themselves . If I find any issues that seem brand new I'll try and remember to post them here.

I've not encountered any problems.

Indeed.

It would appear that the update was so minor that the CK update is not required to keep working, and they may not even release one. I took yesterday off (involuntarily ), today I'll be working. If at the end of the day I have any issues, I'll make sure to post them.

Yes, definately there are new problems. I provided a link earlier, but probably got lost in the testing methodology discussion. See this thread for at least one specific new bug: http://www.gamesas.com/index.php?showtopic=1399783. I'm experiencing problems with dialogs that I never had pre 1.7 release.

Thats some bad news Sollar :/

(Link doesn't work)

Thanks! I fixed the link.

Just to update guys, I've added a post to my thread http://www.gamesas.com/topic/1399783-dialogue-no-longer-works/page__pid__21308124__st__30#entry21308124. Basically tested it all and although the cause is not known, just a bug, I have shown what we can do as a work around.

Sometimes. As my earlier post indicates (in your thread), saving and reloading does not always work. In fact, in one situation, it is causing my dialog to show both options for a 0 or 1 condition.

While I was testing, I also found out that the dialogue of my test-quests works just as it's supposed to do, when I start the quest via change location or script event. Maybe you could try this as a workaround.

Thanks for the tip. I'll do some more testing and see what I find.

Yep, I'm getting the same problem, dialogue added by my quest to the carriage driver faction doesn't show up initially, but if I save and re-load it shows up ok.

Many thanks for the workaround tips. I can't help thinking that Story Manager events are very likely the way Bethesda has designed all their add-on quests to start. If so, it's not surprising this fault slipped through. If that turns out to be a reliable way of avoiding the problem, it's just possible we won't see a fix. I really hope not, though!

#DialogueIsTheNewNavmesh

I wouldn't doubt it. There are so many strange bugs reported for this game that go unfixed that something is clearly wrong with their testing methods when even total noob players can find the bugs within an hour of playing normally.

Allowing for the game to coc you from the main menu encourages nothing but bad habits. I can't even fathom how a professional development outfit would ever consider shortcutting ALL of their initialization of the player would be an accepted practice.