[RELZ] Wrye Bash -- Thread 64 : IV - Oblivion

» Wed Mar 30, 2011 3:53 am

Next error,when rebuilding bashed patch (CBash)
svn ver. 992

Traceback (most recent call last):  File "Q:\Oblivion Tools\Wrye Bash\Mopy\basher.py", line 5834, in Execute    patchFile.buildPatch(SubProgress(progress,0.1,0.9)) #try to speed this up!  File "Q:\Oblivion Tools\Wrye Bash\Mopy\bosh.py", line 17998, in buildPatch    patcher(self, subProgress)  File "Q:\Oblivion Tools\Wrye Bash\Mopy\bosh.py", line 31769, in finishPatch    reX117 = self.reX117  File "Q:\Oblivion Tools\Wrye Bash\Mopy\bolt.py", line 1736, in setFull    if (1.0*full) == 0: raise ArgumentError('Full must be non-zero!')bolt.ArgumentError: Full must be non-zero!

Something with eyes, race, hair

No Idea how you manage to get that, but fixed. (just curious, but what does your modlist look like when you got that? Did you happen to have every mod unchecked?)

» Wed Mar 30, 2011 1:09 am

Any suggestions? Thank you for your time and for your work on this great tool.

Funny you should raise this - I started seeing the same type of problem on my new installation. Are you installing packages and then running OB while WB is still running? For me, closing WB before running seemed to stop the problem happening, although I have not done any extensive testing to verify that this is the connection or that once done, you can run OB while WB is running.

I am using WB v291.

(quoted from the previous thread)

I'll give it a try. Thanks for your response, I really appreciate it. I do usually keep WB open when I launch the game via Steam since it takes forever to open WB again and when I'm adding and testing mods... it's such a time sink already. I will try your suggestion and see if that helps. Thank you very much for the advice. =)

Strange !
Anyway - till someone else more adept maybe gives some insight could you download the latest beta from the SVN (see OP) and see if still present in 292 ?

(quoted from the previous thread)

Thank you for the suggestion, I will check it out too. =)

» Wed Mar 30, 2011 3:12 am

Got a bit of time out to have a look at r993 - In CBash mode, no errors with my load order after a rebuild of the bashed patch on the face of it (using debug mode).

Noticed if I tick all esps, before clicking yes to auto-deselect recognition that some need merged is not as detailed a report as it is with Non-CBash (When I switched back to non-CBash and tried this again the report of mergeables before auto-deselect seems to have expanded compared with what I usually see there)

After building the bashed pach in CBash mode, all auto de-selects are now correct (plus and dots) with one exception - All Natural filter still comes up as a plus.

In Non-CBAsh mode all plus and dots correct including All Natural filter.

» Wed Mar 30, 2011 11:37 am

No Idea how you manage to get that, but fixed. (just curious, but what does your modlist look like when you got that? Did you happen to have every mod unchecked?)

It's a new setup, I am testing and installing mods. Perhaps some combinations give interesting results.

» Wed Mar 30, 2011 2:55 am

CBash patching issues:

2. Missed water height: http://img695.imageshack.us/img695/6818/missedwaterheight.jpg

The bashed patch should have merged the water height change that's shown for Chorrol Hinterland. The OCR+CH patch contains the proper water height edit, and none of these mods is using a C.Water override.

3. Wrong forms assigned for multiple masters patch: http://img9.imageshack.us/img9/9747/wrongforms.jpg
The cell block in 7,5 should look like this instead: http://img84.imageshack.us/img84/5885/rightforms.jpg

If it helps, the specific mods involved are listed in the second image.

5. Wrong texture form assigned in landscape: http://img580.imageshack.us/img580/6957/wrongtexture.jpg

I don't even begin to know how this happened. Index 4C is a DB sanctuary mod that doesn't even touch landscape records and is well before index 80 in the order.

Ok, more on these:
2 - Looks like CBash isn't scanning World Cells, so that's why it got skipped. I'm trying to figure out how to get it to do it properly
3 ,5 - If you notice, each of these has the Mod ID of 4C, but the rest of the ID is correct. Looks like Bash is changing the Mod Id to == that of the bashed patch, which in these cases it shouldn't (i think, right?). So those ones will be fairly easy to fix, once I figure out which patchers are doing it. Looks like the roads patcher for #5, but not sure which patcher does the info in #3 yet.
1,4 - No idea yet.

Noticed if I tick all esps, before clicking yes to auto-deselect recognition that some need merged is not as detailed a report as it is with Non-CBash (When I switched back to non-CBash and tried this again the report of mergeables before auto-deselect seems to have expanded compared with what I usually see there)

Don't know what to say here. Every experience I've had it's the other way around.

After building the bashed pach in CBash mode, all auto de-selects are now correct (plus and dots) with one exception - All Natural filter still comes up as a plus.

In Non-CBAsh mode all plus and dots correct including All Natural filter.

This: with CBash, the All Natural filter suddenly is mergeable, so you probably ticked it under the merge patches section. I'd have to defer to Arthmoor/wrinklyninja on if it's safe to merge or not, but my gut feeling is go with the readme and don't merge, just import.

» Wed Mar 30, 2011 12:03 pm

This: with CBash, the All Natural filter suddenly is mergeable, so you probably ticked it under the merge patches section. I'd have to defer to Arthmoor/wrinklyninja on if it's safe to merge or not, but my gut feeling is go with the readme and don't merge, just import.

Ah could be, after switching to CBash I assume all imports are going to be the same and because after switching everything becomes de-selected for importing, I just use the Select All where I would usually not have to de-select any - In that section of imports just the new T.I.E filter usually needs de-selecting (you only need to Import stats for the T.I.E filter) and ....

All Natural would not ordinarily show up there in non-CBash mode http://www.bild.me/bild.php?file=6107137Untitled.jpg

And you are correct All Natural does get placed in Merge Patches in CBash mode http://www.bild.me/bild.php?file=6963315Untitled2.jpg

Here's the report after ticking all in non-CBash http://www.bild.me/bild.php?file=1880966Untitled.jpg

In CBash mode its not quite so comprehensive http://www.bild.me/bild.php?file=9747773Untitled1.jpg

Edit: What the O_o ... Just switched back to non-CBash and reloaded with all selected I am now getting the not so comprehensive list http://www.bild.me/bild.php?file=6035561Untitled3.jpg

Edit2: :facepalm: Maybe if I Mark Mergeable first that would help http://www.bild.me/bild.php?file=1237061Untitled4.jpg

- Thats the difference there I think, Waruddar told us Mark Mergeable would not be necessary with CBash - hence the footnote on this screen of the Pictorial Guide http://www.tesnexus.com/downloads/images/35230-1-1298661323.jpg wordage for that footnote by Waruddar himself http://www.gamesas.com/index.php?/topic/1156281-relz-wrye-bash-thead-58/page__view__findpost__p__16949435

» Tue Mar 29, 2011 10:52 pm

For sure don't let the AN filter get merged. If we need to mark it NoMerge on a future update then we'll do that.

On to a pretty big bug report related to this:

CBash is totally ignoring all imported CELL records from the AN filter. None of the climate data has been entered into the patch other than for vanilla cells. Not one single mod the filter supports had a change imported. I noticed it immediately upon entering the manor at AFK_Weye when the lighting was totally wrong and the raging thunderstorm outside was suddenly silenced. Thinking it might have been an anomaly, I ventured over to Verona Bay to check on someone's complaint about the lighting there and found the same thing. Which prompted me to check the patch and realize nothing got imported.

@alt3rn1ty: How on Earth did Bash end up marking the AN filter with a checkmark? Activating that should have been impossible.

EDIT: Bug building with CBash

Traceback (most recent call last):  File "C:\Games\Bethesda Softworks\Oblivion\Mopy\basher.py", line 5834, in Execute    patchFile.buildPatch(SubProgress(progress,0.1,0.9)) #try to speed this up!  File "C:\Games\Bethesda Softworks\Oblivion\Mopy\bosh.py", line 17998, in buildPatch    patcher(self, subProgress)  File "C:\Games\Bethesda Softworks\Oblivion\Mopy\bosh.py", line 31907, in finishPatch    subProgress(pstate, _("Assigning random eyes and hairs to npcs missing them...\n"))  File "C:\Games\Bethesda Softworks\Oblivion\Mopy\bolt.py", line 1778, in __call__    self.parent(self.baseFrom+self.scale*state/self.full,message)  File "C:\Games\Bethesda Softworks\Oblivion\Mopy\bolt.py", line 1778, in __call__    self.parent(self.baseFrom+self.scale*state/self.full,message)  File "C:\Games\Bethesda Softworks\Oblivion\Mopy\bolt.py", line 1778, in __call__    self.parent(self.baseFrom+self.scale*state/self.full,message)  File "C:\Games\Bethesda Softworks\Oblivion\Mopy\bolt.py", line 1749, in __call__    self.doProgress(1.0*state/self.full, self.message)  File "C:\Games\Bethesda Softworks\Oblivion\Mopy\balt.py", line 1048, in doProgress    ret = self.dialog.Update(int(state*100))  File "C:\Python26\lib\site-packages\wx-2.8-msw-unicode\wx\_windows.py", line 2971, in Update    return _windows_.ProgressDialog_Update(*args, **kwargs)wx._core.PyAssertionError: C++ assertion "value <= m_maximum" failed at ..\..\src\generic\progdlgg.cpp(337) in wxProgressDialog::Update(): invalid progress value

I did install the WxPython update that was discussed but it's been in place since then and I was able to build the patch before. This is using r993. If it's at all relevant I marked the TIE and AN filters as NoMerge which is the only configuration change I've made to the mods.

» Tue Mar 29, 2011 10:17 pm

For sure don't let the AN filter get merged. If we need to mark it NoMerge on a future update then we'll do that.

On to a pretty big bug report related to this:

CBash is totally ignoring all imported CELL records from the AN filter. None of the climate data has been entered into the patch other than for vanilla cells. Not one single mod the filter supports had a change imported. I noticed it immediately upon entering the manor at AFK_Weye when the lighting was totally wrong and the raging thunderstorm outside was suddenly silenced. Thinking it might have been an anomaly, I ventured over to Verona Bay to check on someone's complaint about the lighting there and found the same thing. Which prompted me to check the patch and realize nothing got imported.

@alt3rn1ty: How on Earth did Bash end up marking the AN filter with a checkmark? Activating that should have been impossible.

EDIT: Bug building with CBash

CBash stuff: ok, I'll try to wrap my brain around that stuff...svcks with PM and Waruddar very busy. They're the ones that really know the bashed patch stuff.

As for activating the filter: activating not impossible, but Bash will give you plenty of warnings when doing it.

Bug report: that was my bad. I used a min where I should have used max. It's fixed now though.

» Wed Mar 30, 2011 3:32 am

Heh. That led to a new error:

Traceback (most recent call last):  File "C:\Games\Bethesda Softworks\Oblivion\Mopy\basher.py", line 5834, in Execute    patchFile.buildPatch(SubProgress(progress,0.1,0.9)) #try to speed this up!  File "C:\Games\Bethesda Softworks\Oblivion\Mopy\bosh.py", line 17998, in buildPatch    patcher(self, subProgress)  File "C:\Games\Bethesda Softworks\Oblivion\Mopy\bosh.py", line 31768, in finishPatch    subProgress.setFull(min(max(ObCollection.LoadOrderMods) * 2,1))TypeError: unsupported operand type(s) for *: 'ObModFile' and 'int'

» Tue Mar 29, 2011 11:30 pm

Heh. That led to a new error:

Traceback (most recent call last):  File "C:\Games\Bethesda Softworks\Oblivion\Mopy\basher.py", line 5834, in Execute    patchFile.buildPatch(SubProgress(progress,0.1,0.9)) #try to speed this up!  File "C:\Games\Bethesda Softworks\Oblivion\Mopy\bosh.py", line 17998, in buildPatch    patcher(self, subProgress)  File "C:\Games\Bethesda Softworks\Oblivion\Mopy\bosh.py", line 31768, in finishPatch    subProgress.setFull(min(max(ObCollection.LoadOrderMods) * 2,1))TypeError: unsupported operand type(s) for *: 'ObModFile' and 'int'

Man I just can't make a coherent python statement to save my life today. Third time's the charm?

» Wed Mar 30, 2011 12:46 pm

Well you at least know what to do with Python. I look at it and eyes glaze over because it's not PHP

3rd time isn't the charm either:

Traceback (most recent call last):  File "C:\Games\Bethesda Softworks\Oblivion\Mopy\basher.py", line 5834, in Execute    patchFile.buildPatch(SubProgress(progress,0.1,0.9)) #try to speed this up!  File "C:\Games\Bethesda Softworks\Oblivion\Mopy\bosh.py", line 17998, in buildPatch    patcher(self, subProgress)  File "C:\Games\Bethesda Softworks\Oblivion\Mopy\bosh.py", line 31768, in finishPatch    subProgress.setFull(max(ObCollection.LoadOrderMods * 2,1))  File "C:\Games\Bethesda Softworks\Oblivion\Mopy\bolt.py", line 1736, in setFull    if (1.0*full) == 0: raise ArgumentError('Full must be non-zero!')TypeError: can't multiply sequence by non-int of type 'float'

» Wed Mar 30, 2011 11:32 am

:brokencomputer: :brokencomputer:

Ok, it should really work this time...

» Wed Mar 30, 2011 5:42 am

Yep, that got it.

» Wed Mar 30, 2011 1:32 pm

I've got a bug in the standard (non-CBash) patch building in both v291 and svn 993. I use Race Balancing Project and Better Redguards. RBP is active, while BR is just selected in the import race data option, and RBP's BR patch is merged. In the resulting patch, the racial skill bonuses for the Redguards are screwed up - they've got two +5 bonuses to restoration, for starters. If I don't merge the RBP/BR patch, the racial bonuses are correct.

I'm also using Robert's Male Body (v5, not that it makes a difference), and I've created my own RBP patch for it. As with the BR patch, if I merge this patch the racial skill bonuses for all of the affected races are screwed. Some of them aren't too bad - the Bosmer gets the bonuses for acrobatics and light armor reversed, for example - but others are completely messed up.

So it seems that if you have something that changes the racial skill bonuses, as RBP does, and then you have a patch for that mod, the bonuses get messed up in the bashed patch. Funny thing is, I'm sure I reported this several versions ago (maybe a year or 18 months ago?) and it got fixed. And I'm fairly certain it was ok in v290.

Edit: OK, so I found a workaround. I added the 'R.Skills' tag to the Robert's & BR patches, and the bonuses were imported into the bashed patch correctly. But it should be working correctly without having to do that - RBP has the 'R.Skills' tag, so the racial skill bonuses should be being picked up and imported into the bashed patch anyway, shouldn't they?

-----
Oh, and another small bug I've noticed in svn 993 (can't see anything that suggests it's been fixed in later svns) - the mod lists don't work. That is, when you select a different mod list, such as the 'Bethesda esms' list, it doesn't change the active mods.

Eloise

» Wed Mar 30, 2011 11:33 am

:whistling: One more request - BAIN Wizards; Could we have the Wizard screen window size sticky (I resize the wizard borders every time I use it now - Just a cosmetic thing but it would be nice if that size stayed put also, hopefully its a simple change)

Edit: I know there are bigger priorities everyone would like to see in 292 concentrated on first so the above is only a request for if you get time to fit it in.

Edit2: Also grabbed r998 after PM snook an update in there, will be testing along with updating Brittainy's BAIN wizard.

Edit3: Will there be a party for r1000? :foodndrink:, venue tbd, place your bets for who will be the committer

» Wed Mar 30, 2011 1:20 am

My money's on Lojack - r999 has landed

» Wed Mar 30, 2011 9:27 am

I've got a bug in the standard (non-CBash) patch building in both v291 and svn 993...

-----
Oh, and another small bug I've noticed in svn 993 (can't see anything that suggests it's been fixed in later svns) - the mod lists don't work. That is, when you select a different mod list, such as the 'Bethesda esms' list, it doesn't change the active mods.

Eloise

Yeah, that sounds like unwanted behavior for the races, added to my growing BP fixing list. Fixed the mod lists though.

:whistling: One more request - BAIN Wizards; Could we have the Wizard screen window size sticky (I resize the wizard borders every time I use it now - Just a cosmetic thing but it would be nice if that size stayed put also, hopefully its a simple change)

I can get window position to stick...size is waay trickier than it should be though. Still trying to figure it out.

Edit3: Will there be a party for r1000? :foodndrink:, venue tbd, place your bets for who will be the committer

My money's on Lojack - r999 has landed

my002 wins! Rounds for everyone!

EDIT: Finally got the wizard size thing to work. Wizards will now remember their size and position.

» Wed Mar 30, 2011 4:07 am

my002 wins! Rounds for everyone!

EDIT: Finally got the wizard size thing to work. Wizards will now remember their size and position.

:D Cheers myk002 mines a double

Quit r1001 after messing with the wizard, reload, launch wizard - Window position and size remembered http://www.bild.me/bild.php?file=5674579Untitled.jpg

Thank you

» Wed Mar 30, 2011 11:54 am

:D Cheers myk002 mines a double

Quit r1001 after messing with the wizard, reload, launch wizard - Window position and size remembered http://www.bild.me/bild.php?file=5674579Untitled.jpg

Thank you

"That's a cute screenshot" she says.
...anyway updating your wizard I guess.... That's the default screen, Yes,No?

» Wed Mar 30, 2011 1:02 am

"That's a cute screenshot" she says.
...anyway updating your wizard I guess.... That's the default screen, Yes,No?

Yes, very much WIP at the moment, http://www.bild.me/bild.php?file=7441275Untitled.jpg - Still have a zip full of options to crack open yet ( hence the question the other day reference how many sub-packages can BAIN handle

)

» Tue Mar 29, 2011 10:05 pm

- Still have a zip full of options to crack open yet ( hence the question the other day reference how many sub-packages can BAIN handle

)

Well, in theory, It should be able to handle anything you can throw at it, up to your memory limit.
Just don't get stupid with X folders x Y Folders x Z folders.
These things can multiply quickly.

note: Algebra always annoyed me for how simple examples where explained. What is X? What is Y? And why the heck am I looking for Z? Give me a 'rubber ducky', 'train headed for Chicago', and 'destination/arrival time' and only then everything can be peachy :read:

EDIT: that is a trig question...for what it is worth...final exam...define V

» Wed Mar 30, 2011 4:36 am

my002 wins! Rounds for everyone!

LOL :celebration:

Now back to omods - no hangups no glitches but :

---------------------------OMOD Extraction Complete---------------------------The following OMODs failed to extract.  This could be a file IO error, or an unsupported OMOD format: * 5 Illumination Within Revived 0.9.2.omod * Natural_Faces_v95_OMOD-1965.omod---------------------------ΟΚ   ---------------------------

They must be rather ancient but maybe interested in seeing why
http://www.tesnexus.com/downloads/file.php?id=4955
http://www.tesnexus.com/downloads/file.php?id=1965

edit : 1004
Forgot to mention that another 20 or so unpacked just fine (check my previous post)

» Wed Mar 30, 2011 1:49 pm

edit : 1004
Forgot to mention that another 20 or so unpacked just fine (check my previous post)

Has anyone that still uses OMODs been brave enough to test toss the whole folder yet....

» Wed Mar 30, 2011 10:38 am

Now back to omods - no hangups no glitches but :

---------------------------OMOD Extraction Complete---------------------------The following OMODs failed to extract.  This could be a file IO error, or an unsupported OMOD format: * 5 Illumination Within Revived 0.9.2.omod * Natural_Faces_v95_OMOD-1965.omod---------------------------ΟΚ   ---------------------------

They must be rather ancient but maybe interested in seeing why
http://www.tesnexus.com/downloads/file.php?id=4955
http://www.tesnexus.com/downloads/file.php?id=1965

edit : 1004
Forgot to mention that another 20 or so unpacked just fine (check my previous post)

Thanks for those. I had messed up my code for very old omod formats, fixed now.

» Wed Mar 30, 2011 4:56 am

my002 wins! Rounds for everyone!

Haha I totally didn't notice when I checked that in : ) For those that are interested, revision 1000 was a http://oblivionworks.svn.sourceforge.net/viewvc/oblivionworks/Programs/Wrye%20Bash/experimental/bait/doc/func_spec_view.txt for the work I'm doing for the installers tab. It's basically an explanation of all the user-visible functionality. It could possibly be used as a first cut at a users guide, perhaps written by someone better at writing than I am : P

edit: added link

edit2: here's a http://oblivionworks.svn.sourceforge.net/viewvc/oblivionworks/Programs/Wrye%20Bash/experimental/bait/doc/gui_mockup.png so the func spec makes some sense : )