The "Please Stop Using 1.0" Discussion

[Warboy1982]

i was referring to the mods themselves.

[volutar]

Ok. I stop understanding that too.

New thread tittle: Stop using nightlies
[/irony]

Completed mods: finished mods working with 1.0 and updated version compatible with nightly
Experimental mods: unfinished WIP mods working with 1.0 or/and updated version compatible with nightly
Mod archive: finished mods working with 1.0 and not working with nightly (without updated version)

[Falko]

if nobody makes/uses mods with nightly
some errors in the new code are not discovered
trivial example: vanilla mapscript had no removeblock command only if somebody plays around with that it can be discovered that the command is/was not working correctly
without the early testing some functionality is not tested properly and bugs stay in the next milestone
(if a mod uses the rulset correctly but does "more" than vanilla and runs into a bug its still a bug even if its not a problem for vanilla | i do not talk about limitation like exactly 10 alienitemlevels/month or misformed rulesets)

[volutar]

That's exactly my irony was about. If modmakers aren't catching up with the nightly and players not using/testing them in their LPs or just play, then devs won't fix bugs which will disallow "updated" mod work with it after next milestone. Devs don't test your mods. And won't fix engine to fit them better. Devs are care only about vanilla rulesets. Everything else is up to you.
Don't want to update mods? Don't want to play updated mods and nightly? Expect huge bugs with new milestone then. And don't complain.

What "stability" then you're expecting from the "stable"?

[Yankes]

I think of 2 solution for current situation based on what ETA of TFTD:

1) if its less than 1 month, wait for it.
2) if its longer than 1 month, create new milestone based on today master branch.

This is probably most beneficial for whole community. Modders can stop wild ride after nightlies, Warboy (Destroyer of savesmods ) can continue this job and user have bug fix version form the start (not like download 1.0->find bug->go to forum->download nightly).

I can volunteered to help push case 1).

[Hobbes]

I've been keeping an eye on this thread and from my previous experience with the nightlies here's what I think:

1) Saying that modders help up the development process by keeping their mods updated and reporting nightly bugs is largely wishful thinking. Modders are not playtesters, so trying to use them as such will merely frustrate both developers and modders.

2) It is also irrealistic to expect that most modders keep up with the changes introduced by the nightlies. Again, modders shouldn't be expected to have the same zeal or availability of developers, since they will usually quickly move to their other gaming interests after they've finished their projects.

3) Most, if not nearly all currently existing mods will become unplayable on the nightlies until 2.0 gets released, and a lot will never be updated afterwards. A modder can drop by and update once in a while but you really need a lot of free time to be constantly updating a mod to be compatible with the latest nightly. 

4) When OpenXcom 1.0 was released I remember a video review where the journalist said something that surprised me and it was like: "there's a ton of mods for this game now. I guess everyone was holding to their mods until the game got released". I now strongly suspect the same will happen when 2.0 gets released because of all the reasons above.

Personally, after my experiences in the past days, I'm going to pick a nightly that is "stable" and I'll stick with it while developing my current mod project. It's simply not productive for me to be constantly interrupted from creating new stuff because of the need to fix things that were working perfectly until the latest nightly got released. And if a player wants to use my mods with the latest nightly, well he/she will need to wait until 2.0 or I post an update.

Meanwhile, if playtesting is really needed, I'd suggest at some point to place a call on the forums and elsewhere for playtesters among the players. People whose sole responsability is to keep track of the nightlies and check the vanilla game (not the mods) for bugs. I'm sure a few people would volunteer to help.

Developers do their thing their way, modders and playtesters the same. And hopefully we all meet again at the end of the road.

[ivandogovich]

I've been keeping an eye on this thread and from my previous experience with the nightlies here's what I think:

1) Saying that modders help up the development process by keeping their mods updated and reporting nightly bugs is largely wishful thinking. Modders are not playtesters, so trying to use them as such will merely frustrate both developers and modders.

2) It is also irrealistic to expect that most modders keep up with the changes introduced by the nightlies. Again, modders shouldn't be expected to have the same zeal or availability of developers, since they will usually quickly move to their other gaming interests after they've finished their projects.

3) Most, if not nearly all currently existing mods will become unplayable on the nightlies until 2.0 gets released, and a lot will never be updated afterwards. A modder can drop by and update once in a while but you really need a lot of free time to be constantly updating a mod to be compatible with the latest nightly. 

4) When OpenXcom 1.0 was released I remember a video review where the journalist said something that surprised me and it was like: "there's a ton of mods for this game now. I guess everyone was holding to their mods until the game got released". I now strongly suspect the same will happen when 2.0 gets released because of all the reasons above.

Personally, after my experiences in the past days, I'm going to pick a nightly that is "stable" and I'll stick with it while developing my current mod project. It's simply not productive for me to be constantly interrupted from creating new stuff because of the need to fix things that were working perfectly until the latest nightly got released. And if a player wants to use my mods with the latest nightly, well he/she will need to wait until 2.0 or I post an update.

Meanwhile, if playtesting is really needed, I'd suggest at some point to place a call on the forums and elsewhere for playtesters among the players. People whose sole responsability is to keep track of the nightlies and check the vanilla game (not the mods) for bugs. I'm sure a few people would volunteer to help.

Developers do their thing their way, modders and playtesters the same. And hopefully we all meet again at the end of the road.

Well said, Hobbes. +1

[Gifty]

Hear, hear!

[Meridian]

+10

[volutar]

If you read me accurately you would notice that I said:
1) developers are worried about only vanilla. And no tests concerning mods are done.
2) modmakers have to test what are they doing but the main tests are done by players/lpers which shouldn't ignore latest versions. Those who are still on v1.0 and old mods are not helping project a bit. And shouldn't expect these mods working with 1.1.
3) milestone is the most buggy version at the moment and nightly of 6 months after 1.0 release is much much better. Keeping mods actual for 1 month old nightly is hundred times better than 1.0. 1.0 is essentially the worst case here. It's not better than 0.9 in terms of any usefulness.

The main appeal is about stop using oxc1.0 and mods for it. And start to catching up the nightly. Because 1.1 is nigh and without proper attention from modmakers and modplayers it may become modless version, which will require another iteration to become stable.

Basically consider current nightly as prerelease and aim to support of it.
No one will do that for you.

[Meridian]

1) agreed, that is how it should be
2) you obviously don't want to listen to any of our arguments, so I will not discuss this topic anymore
3) agreed

1.1 will definitely not be modless... on the contrary, once it is released, people will start updating their mods and create new mods too... whether you like it or not, it is AFTER the release of 1.1 when the people will become more active... heck, I may even try to make my own mod, but definitely not before a milestone is released

PS: this is my last post in this thread

[volutar]

2. I didn't hear any viable argument advocating the fact that some modmakers are living in the past (continuing making mods for 0.9/1.0) and by that provoking people not to help testing 1.1 prerelese. Instead of helping project to be released bugless they are distracting players, dragging their attention to outdated stuff.

[Falko]

what happens if there is no modmaking/testing for nightlies
the problem is there is no bugfixing for a release version (1.0, 1.1,...)
and if huge crowds of modmakers make new mods only if 1.1 is out they will encounter bugs
these bugs will be fixed but the fixes are now only in these scary nightlies

i do not say all modders/player should switch to nightlies
i have no numbers but i think if >20% of player base play with nightlies that would be good enough

IMHO if you publish a mod that works only with nightly put it in the thread title

also as a suggestion for the devs: before milestone release make a release candidate
define a nightly "1.1 RC1" approx. a month before release with the intention "there should not be (significant/any) ruleset changes until release" so more modder feel confident using this RC to test and find bugs for the next release

[Meridian]

I feel like I am talking to a wall 

Have I EVER said anything about bugs? Hint: no, I haven't!

I am talking about stability. And stable does not mean "bug free". Just google the word stable and you will find a definition close to this: "stable = not likely to change; firmly established". Every software has bugs and even OpenXcom version 99.0 will have bugs. That's life.

Developers can live and work with a very low degree of stability.
Beta-testers the same.
Modders need at least moderate degree of stability.
Normal users (incl. LPers) need a very high degree of stability.

By definition, version 1.0 is stable (even if it is full of bugs).
By definition, nightly build is unstable.... because it CHANGES every day.

PS: I am not a native English speaker... if I have a severe language problem in this post, please let me know... because that is the only thing I can think of, which causes our misunderstandings

[volutar]

Stable in terms of software means not causing crashes or other breaking stuff.
"Stability" you refer is another kind of thing. It's often called "feature freeze".
As soon as oxc in on its road to tftd it can't be feature frozen. If community decide it can choose particular nightly as the RC without any devs involvement and simply use it as base for both mod making and LPing. No one is stoping people from deciding that.
It's just matter of making public download point and perhaps even custom installer if you want to.
Milestone release actually means NOTHING in terms of project development. Just another installer for the current nightly. The problem with OpenXcom release schedule IMHO is not having official prerelease or RC versions. It really confuses and makes real "stability" some theoretical term.