1.19m Changelog

*** 1.19m is a BETA patch ***

*** 1.19m is scheduled for release on 4/14/2011 ***

 

Changes:

1. Fixed bug where action entries used for a combat ability would always trigger that combat ability when clicked, even if reassigned a different action when being reused, caused by some entry info not getting cleared out.

2. Fixed bug where dragon breath effect (and any other attached spell effect that isn't a projectile) would play throughout the entire battle once triggered, floating on the tile where it was triggered.

3. Fixed bug where casting a spell from an army with a spellcaster leader on the main map, with the army selected instead of a particular subunit, would cause every unit in that army to cast the spell.

4. When you select an army on the main map, it now only shows the spellcasting action if the leader can cast spells, and if selected, only the leader casts the chosen spell.

5. Fixed bug where, if you finished researching all spells and reached the max spell level, it would start automatically researching non-existent spell levels beyond the max spell level.

6. Fixed bug where the dragon's breath sfx would occasionally not play until the target unit took damage, causing the attack itself to be silent and the attack sfx to linger long after the attack.

7. Map Data now does selective updates of ZOC lines.

8. Improved ZOC performance – huge map save game used to take 30s to update the ZOC at the end of a turn, now takes 14s.  ZOC is no longer cleared every turn and re-calculated; now the objects update the ZOC nodes as necessary, but the graphics and environment do not update until the end of the turn.    It would now be much easier to move the ZOC updating code into a thread so that it could update in realtime.  The ZOC Calculator now tracks which tiles have changed ownership so it only does the end of turn updates if something has actually changed. It still has to re-draw all the lines even if only one sector was affected (I’m working on that).

9. Fixed bug where killing a unit in a tactical battle that had tactical modifier effects from spells on it would cause it's stats to get recalculated using a function with auto-killing logic in it, so that the parent unit would immediately be set to be destroyed, causing bugs like the sovereign dying and immediately ending the game, because no retreat or battle-finishing logic was reached.

10. Fixed a series of population and prestige-related bugs that conspired to cause cities to have negative population and huge power ratings.  The details about changes in City.cpp talks more about the bugs.

11. Fixed crash on closing a game that had a tactical battle in it, caused by a bug in my entry memory changes where the shard listgrid entry always handled its own memory, though the tactical wnd does not manually manage entry memory for its mana and shard display.

12. Memory fixes.

13. Fixed bug where some unit clean-up stuff in the player was not getting called when a unit died, so that the player-wide abilities of dead champions would remain until something reforced an ability value calculation, players would hold onto a pointer to the dead sovereign spouse, and military might and power rating wouldn't be recalculated.

14. The global resource pool in the main UI now reuses entries wherever possible, instead of recreating them for each refresh (performance improvement).

15. Fixed a crash when exiting a game with any shards owned by the local player, caused by the global resource shard entry dxpacks getting deleted too early from having too few refs (the main game wnd was not loading them).

16. Fixed a potential deadlock when a unit is removed from a player, caused by some UI refreshes taking place while the unit critical section was being held onto.

17. Fixed crash when killing an enemy sovereign.

15,638 views 19 replies
Reply #1 Top

1st.

I love it when dev talk sneaks into patch notes, like references to City.cpp.

So... no 1.2 this week? ;(

Reply #2 Top

I don't think we get 1.2 until it's ready.  It's taking a long time to get ready, but there was a lot of work to do.

 

The important thing is getting it right though, even if it's frustrating to not get new toys.

 

Reply #3 Top

i like changelogs that say more than just "memory fixes".  the longer it takes 1.2 to come out, the sweeter 1.3 will be right?  right?

p.s. thanks to mandragoron for finding every little bug and crash out there! 

Reply #4 Top

Then again, If you do the beta test it's like getting a new toy every week

 

(edit)

And indeed Trojasmic,  Mandragoran deserve a really big thanks for all his work

Reply #5 Top

I love reading detailed patchnotes. Those explanations for the bugs are enlightening to a IT noob like me.

 

As for when it's ready, here's what Johann from Paradox (a company with strong after support too) says about them:

"A patch is never late! Nor is it ever early. It arrives precisely when I mean it to do!"

:)

Reply #6 Top

 

  I'm not 100% happy with the state of memory utilization in WoM.  But the developers have been working hard on fixing memory issues.  I wish it was one memory issue that we could fix and everything would be great.  But instead we fix individual memory leaks which may account for a few meg each.  The longer we work the smaller and smaller memory leaks we find (we can profile and see where memory is going) and we are working on getting 1.2 out for everyone because even if it isn't perfect, its much better than 1.1.

  The next step is to look at major changes to the memory system.  But we aren't going to hold 1.2 for that.  That will need some R&D, PoC and other random acronyms to get right.

 

+1 Loading…
Reply #7 Top

I don't think 1.2 is far off now though.

Reply #8 Top

Quoting Derek, reply 6
 

  I'm not 100% happy with the state of memory utilization in WoM.  But the developers have been working hard on fixing memory issues.  I wish it was one memory issue that we could fix and everything would be great.  But instead we fix individual memory leaks which may account for a few meg each.  The longer we work the smaller and smaller memory leaks we find (we can profile and see where memory is going) and we are working on getting 1.2 out for everyone because even if it isn't perfect, its much better than 1.1.

  The next step is to look at major changes to the memory system.  But we aren't going to hold 1.2 for that.  That will need some R&D, PoC and other random acronyms to get right.

 
End of Derek's quote

 

Awesome to hear that, least your going in the right direction ;)

Reply #9 Top

Many thanks to Brad, Derek SD team, (Troj and Reinbad for the shout out), and other folks like Harpo, Fatnonfree, etc! The painful journeys are oftentimes the most rewarding. That's one hell of a change log up there - awesome.  

Onwards to 1.3!

Reply #10 Top

Quoting Derek, reply 6
 

  I'm not 100% happy with the state of memory utilization in WoM.  But the developers have been working hard on fixing memory issues.  I wish it was one memory issue that we could fix and everything would be great.  But instead we fix individual memory leaks which may account for a few meg each.  The longer we work the smaller and smaller memory leaks we find (we can profile and see where memory is going) and we are working on getting 1.2 out for everyone because even if it isn't perfect, its much better than 1.1.

  The next step is to look at major changes to the memory system.  But we aren't going to hold 1.2 for that.  That will need some R&D, PoC and other random acronyms to get right.

 
End of Derek's quote

Understand on my end and am cool with the decision to move forward. As long as the OOM's and bugs still get priority going forward I'm easy. I personally am impatient to get to 1.2 and 1.3 so I can play with the mods again. Just my humble opinion. :grin:

Reply #11 Top

I like the change log, it's nice and long, not just memory fixes but much much more. Thanks for the shout out AlLanMandragoran, most of these fixes are yours!

I really hope that when we move on to 1.3 more stuff is added to the game than in listed. Like I said in that post magic needs much more love. As the game is now I almost never use it. All you need is battle axes leather armor and you can walk through almost everyone until much later on.

Having said that I know what I'll be doing to night.

 

P.S.

What is R&D and PoC?

Reply #12 Top

Quoting FatNonFree, reply 11
What is R&D and PoC?
End of FatNonFree's quote

R&D stands for Rage & Destruction, and PoC is really just too colorful to spell out if there are children in the room.

Reply #13 Top

It's live get it now!

Reply #14 Top

Quoting FatNonFree, reply 11
What is R&D and PoC?
End of FatNonFree's quote

 

R&D- Research and Design, the time spent checking out what the memory system would be like, how it would work and how long it would take to implement

PoC- Proof of Concept, the trial to make sure the proposed system is meeting the goals we have for it

We really don't talk in acronyms, I was just being goofy.

Reply #15 Top

Quoting Derek, reply 14
R&D- Research and Design, the time spent checking out what the memory system would be like, how it would work and how long it would take to implement

PoC- Proof of Concept, the trial to make sure the proposed system is meeting the goals we have for it

We really don't talk in acronyms, I was just being goofy.
End of Derek's quote

Yeah, it means that too, in some circles.

Reply #16 Top

11. Fixed crash on closing a game that had a tactical battle in it, caused by a bug in my entry memory changes where the shard listgrid entry always handled its own memory, though the tactical wnd does not manually manage entry memory for its mana and shard display.
End of quote

 

I like the "my entry memory changes" bit.  The same thing happens to me at my software development company.  The marketing department will ask me whats new in the upcoming service pack, and so I'll respond with something like "I fixed this, but  blah blah..." and somehow  it ends up in the newletter to our customers verbatim.

 

Actually the whole fix is a bit of a "what the?", but it sounds like progress! 

Reply #17 Top

this sounds like a verry interesting set of patches.

thanks to all the team

harpo

 

Reply #18 Top


7. Map Data now does selective updates of ZOC lines.

8. Improved ZOC performance – huge map save game used to take 30s to update the ZOC at the end of a turn, now takes 14s.  ZOC is no longer cleared every turn and re-calculated; now the objects update the ZOC nodes as necessary, but the graphics and environment do not update until the end of the turn.    It would now be much easier to move the ZOC updating code into a thread so that it could update in realtime.  The ZOC Calculator now tracks which tiles have changed ownership so it only does the end of turn updates if something has actually changed. It still has to re-draw all the lines even if only one sector was affected (I’m working on that).
End of quote

One thing I've noted related to the DX errors I continue to receive. Oftentimes (about 75%) instead of a DX CTD error (25%), the game will freeze up on the transition between cloth and 3d maps. It will then zoom all the way into a 3d tile and I'll get the windows warning that my device driver has recovered and the game will continue on. The ZOC lines will disappear but ending the turn will redraw them. In this patch when the game recovers, the ZOC lines do not redraw. Thus, I'll have a game that doesn't CTD (good) but the ZOC lines are missing (bad). It doesn't appear subsequent end turns will redraw the lines so if I want the lines back I have to exit the game and reload. The next time this happens I'll take a screenshot, save, and provide the debug file.

Reply #19 Top

I started a thread about the pseudo-DX errors where the display driver recovers (but again, it throws the ZOC's all out of whack): https://forums.elementalgame.com/407744.

Note the zoc lines are gone post driver recovery: