OpenMW 0.41.0

Anything related to PR, release planning and any other non-technical idea how to move the project forward should be discussed here.
Locked
User avatar
MiroslavR
Posts: 156
Joined: 12 Feb 2014, 17:45

Re: OpenMW 0.41.0

Post by MiroslavR » 07 Dec 2016, 03:04

These are just warnings, no release blockers:
https://bugs.openmw.org/issues/1032
https://bugs.openmw.org/issues/2771

User avatar
Aussiemon
Posts: 7
Joined: 09 Aug 2016, 00:36

Re: OpenMW 0.41.0

Post by Aussiemon » 07 Dec 2016, 04:47

MiroslavR wrote:These are just warnings, no release blockers:
https://bugs.openmw.org/issues/1032
https://bugs.openmw.org/issues/2771
Figured they were probably benign. Thanks for the heads up.

User avatar
Ace (SWE)
Posts: 838
Joined: 15 Aug 2011, 14:56

Re: OpenMW 0.41.0

Post by Ace (SWE) » 07 Dec 2016, 07:39

Windows RC3 builds then;

32-bit
64-bit

User avatar
raevol
Posts: 3054
Joined: 07 Aug 2011, 01:12
Location: Caldera

Re: OpenMW 0.41.0

Post by raevol » 07 Dec 2016, 10:30

Some stuff I need help on:
Bug #3635: Levelled items in merchants are "re-rolled" (not bug 2952, see inside)
^ what was fixed here?
Bug #3634: An enchanted throwing weapon consumes charges from the stack in your inventory. (0.40.0)
^ I feel like this really isn't that complicated but I am just not parsing what is going on here. Should the thrown weapons be depleting the stack size instead of the charge?
- Fix for weather not updating correctly in Mournhold (Issue 3594)
^ Is this an ok description?
Bug #3576: Minor differences in AI reaction to hostile spell effects
^ see: viewtopic.php?f=2&p=43942#p43942
- Fixed gui elements not working correctly in a certain configuration on fedora (Issue #3575)
^ is this a good way to explain this?
Bug #3564: Editor: openmw-cs verification results
^ what's the issue here? how was it fixed?
Bug #3451: Feril Salmyn corpse isn't where it is supposed to be
^ how was this fixed? Please reference commits or pull requests when you close bugs, at the very least!
Bug #3091: Editor: will not save addon if global variable value type is null
^ how was this fixed? issue comments seem to suggest a multitude of solutions could have been used
- Fixed an issue with the water surface transparency when viewed from certain angles using non-shader water
^ is this forrect for #2716?
- Fixed scripted animation state to be saved, fixes several places where physical obstructions are removed by quest activity: now they will stay removed
^ can someone help me re-word this to include descriptions of the quest events/areas effected by this? I am having a hard time explaining it in a technically correct way, and also describing the effected areas
Feature #3516: Water Walking should give a "can't cast" message and fail when you are too deep
Bug #1138: Casting water walking doesn't move the player out of the water
^^ commit says this moves the character out of the water if there is room, but does it move them from below the depth the original game did? does it display the message it's supposed to when the player shouldn't be able to cast it?

User avatar
Capostrophic
Posts: 717
Joined: 22 Feb 2016, 20:32

Re: OpenMW 0.41.0

Post by Capostrophic » 07 Dec 2016, 11:14

Majority of the misunderstandings can be solved just through a thorough look on the issues pages on the bugtracker, though. There almost always is a "fix commit" on the right.

(off-topic: "Fixed [the issue symptoms]" actually was and is a really bad wording in many cases.)

Bug #3635: the item count wasn't properly tracked; every opening of the trade window levelled goods were re-added to the merchant's inventory.

Bug #3634: throwing weapons used the charge of *all* the projectiles in the stack. Now they only use their own charge -- i.e. the player never witnesses any charge depleting.

Bug #3594: relatively...

Bug #3576: Allofich's reasoning is fair.

Bug #3575: this issue is not unique to Fedora. ID validity checks were incorrect — apparently the ID can be negative.

Bug #3564: the issue is... verification results! "Not local" and "no lore" conditions were reverse for the verificator, and the verificator didn't know about FFFF (no faction) pseudo-faction ID.

Bug #3451: OpenMW used to not like items with invalid enchantment names and disabled their owners or something. With error message, of course. Vanilla ignored them. There should be related "fix" commit in the issue on the bugtracker.

Bug #3091: the issue comments are *8 months apart* from the actual "solved" messages. The related pull request - 1013.

Bug #2716: yep.

- Scripted animation state... nah, no help from me...

Bug #1138 and Feature #3516: Allofich added the message with pull request 1131.

User avatar
Atahualpa
Posts: 1045
Joined: 09 Feb 2016, 20:03

Re: OpenMW 0.41.0

Post by Atahualpa » 07 Dec 2016, 11:24

raevol wrote:Bug #3634: An enchanted throwing weapon consumes charges from the stack in your inventory. (0.40.0)
^ I feel like this really isn't that complicated but I am just not parsing what is going on here. Should the thrown weapons be depleting the stack size instead of the charge?
When you threw a thrown weapon with an "On Strike" enchantment in previous versions, the charge of the whole stack (i.e. of every single remaining throwing weapon) would be reduced by the charge cost. Now, the remaining weapons are no longer charged for the weapon you just throw in Fargoth's face, i.e., the charge of the remaining stack stays the same.
raevol wrote:- Fix for weather not updating correctly in Mournhold (Issue 3594)
^ Is this an ok description?
I don't know. Looks good to me but the fix affects weather in general.
raevol wrote:- Fixed gui elements not working correctly in a certain configuration on fedora (Issue #3575)
^ is this a good way to explain this?
Wasn't this an error in building MyGUI?
Update: I was wrong. See MiroslavR's post for clarification.
raevol wrote:Bug #3564: Editor: openmw-cs verification results
^ what's the issue here? how was it fixed?
I'd say something like

Code: Select all

- OpenMW-CS: Fixed verification for "Not Local" info condition
- OpenMW-CS: Fixed verification for special faction ID "FFFF"
raevol wrote:Bug #3451: Feril Salmyn corpse isn't where it is supposed to be
^ how was this fixed? Please reference commits or pull requests when you close bugs, at the very least!
See Fixes Bug 3451 - Invalid Enchantment IDs #1008. The way I see it, we changed OpenMW to be more forgiving when an object refers to an enchantment ID which not exists.
raevol wrote:Bug #3091: Editor: will not save addon if global variable value type is null
^ how was this fixed? issue comments seem to suggest a multitude of solutions could have been used
See Fix globals not being initialized with a valid type. #1013 and the corresponding commit.
raevol wrote:- Fixed an issue with the water surface transparency when viewed from certain angles using non-shader water
^ is this forrect for #2716?
Sounds good.
raevol wrote:- Fixed scripted animation state to be saved, fixes several places where physical obstructions are removed by quest activity: now they will stay removed
^ can someone help me re-word this to include descriptions of the quest events/areas effected by this? I am having a hard time explaining it in a technically correct way, and also describing the effected areas
Maybe:

Code: Select all

- Fixed animation state to be saved upon cell-changing or save & load; fixes several traps as well as the metal doors in Sotha Sil
raevol wrote:Feature #3516: Water Walking should give a "can't cast" message and fail when you are too deep
Bug #1138: Casting water walking doesn't move the player out of the water
^^ commit says this moves the character out of the water if there is room, but does it move them from below the depth the original game did? does it display the message it's supposed to when the player shouldn't be able to cast it?
If the character is not too deep, it moves them out of the water. The message is correctly displayed, if you are too deep.
Last edited by Atahualpa on 07 Dec 2016, 18:59, edited 1 time in total.

CMAugust
Posts: 201
Joined: 10 Jan 2016, 00:13

Re: OpenMW 0.41.0

Post by CMAugust » 07 Dec 2016, 12:38

On scripted animation states: before the fix, this is what happens if you revisit the heart chamber after the main quest. The destroyed state is accomplished by an animation, one that OpenMW needed to remember so it would stay destroyed. In Tribunal, this bug could actually block progress in a dungeon.

At least that's how I understand it, feel free to correct me.

Allofich
Posts: 104
Joined: 28 May 2016, 12:50

Re: OpenMW 0.41.0

Post by Allofich » 07 Dec 2016, 13:12

Bug #3576: Minor differences in AI reaction to hostile spell effects
I wrote a response in the linked topic. But anyway, the main issue here could be rewritten like "AI followers counted the friendly-fire magic effects cast on them differently from original game" or something.
Last edited by Allofich on 07 Dec 2016, 13:22, edited 1 time in total.

Allofich
Posts: 104
Joined: 28 May 2016, 12:50

Re: OpenMW 0.41.0

Post by Allofich » 07 Dec 2016, 13:17

CMAugust wrote:On scripted animation states: before the fix, this is what happens if you revisit the heart chamber after the main quest. The destroyed state is accomplished by an animation, one that OpenMW needed to remember so it would stay destroyed. In Tribunal, this bug could actually block progress in a dungeon.

At least that's how I understand it, feel free to correct me.
Yeah I believe you are correct. I would just note that I would try to avoid mentioning any spoiler stuff. There may still be people who haven't played the main quest. I myself haven't played Tribunal or Bloodmoon yet, so I try to avoid where they are being discussed.

User avatar
MiroslavR
Posts: 156
Joined: 12 Feb 2014, 17:45

Re: OpenMW 0.41.0

Post by MiroslavR » 07 Dec 2016, 13:45

Atahualpa wrote:
raevol wrote:- Fixed gui elements not working correctly in a certain configuration on fedora (Issue #3575)
^ is this a good way to explain this?
Wasn't this an error in building MyGUI?
No. I would explain it this way:

Code: Select all

Fixed dialogue/journal links being unclickable on certain platforms due to undefined behavior
There was a nasty bug where object pointers were reinterpreted as signed integers and the game relied on them being positive. There is no telling if they end up positive or negative.

Locked