OLD: s3pe_1005-03-0954

[pljones]

This thread is for people wanting to test at the bleeding edge of development. If you're looking for the stable release, please switch to this thread (http://www.modthesims.info/showthread.php?t=362412).

Any questions regarding the behaviour of this release should be posted as replies below.

Bug reports for this version in this thread please!

Equally as important, post if it's all working perfectly...

PLEASE NOTE

This version contains some deep changes in the code for list handling -- pretty much anywhere there's a list (a collection that can contain a variable number of objects), there is the possibility that something won't update or get written out or something really nasty like that.

Please, please be extra careful in testing this version.

I've left the previous QA version available.

-----------------------
Windows Easy Installer version (http://sourceforge.net/projects/sims3tools/files/s3pe/1005-03-0954/s3pe_1005-03-0954.exe/download)

Unpack-It-Yourself version (http://sourceforge.net/projects/sims3tools/files/s3pe/1005-03-0954/s3pe_1005-03-0954.7z/download)
-----------------------

Change Summary
Fixes to list processing.
TGIBlockListEditor now works.

[pljones]

s3pi_1005-03-0953 -- latest rev: 419

More XML comment changes! http://dino.drealm.info/peter/Help/
[AI]?ResourceIndexEntry: Bring in line with AHandlerElement implementation.
ARCOLBlock: Suppress "Never used" warning on ResourceChanged EventHandler.
TypedValue: Now throws NotImplementedException on Clone() if the underlying object is not cloneable.
AHandlerList<T>: AddRange() and InsertRange() call Add and Insert.
TGIBlockListEditor: Amend MainForm to take a copy of the TGIBlocks to be edited, rather than editing in place.
TGIBlockListEditor: Amend MainForm to use standard modal form conventions (no click handlers needed).
Doc msbuild.bat: Don't complain if the working and output folders don't exist when trying to delete them.
Doc msbuild.bat: Create .htaccess in .\Help from htaccess.txt.

[pljones]

s3pe_1005-03-0954 -- latest rev: 493

BrowserWidget: Consistently cast IResourceIndexEntry to AResourceIndexEntry to access ResourceIndexEntryChanged event handler.

(Also respin for latest s3pi)

[pljones]

PLEASE NOTE

This version contains some deep changes in the code for list handling -- pretty much anywhere there's a list (a collection that can contain a variable number of objects), there is the possibility that something won't update or get written out or something really nasty like that.

Please, please be extra careful in testing this version.

I've left the previous QA version available.

[pljones]

178 downloads and no one reporting anything? I guess that means it's okay, right?

[orangemittens]

It's working perfectly for me. I've changed slots, added new default material thingies, changed FTPT's, and it all worked fine. Thank you for making a lot of this stuff easy enough for me to do it :)

[cmomoney]

I don't know if this is a s3pe problem or not, but all the bookshelves' RSLT give an error:


Error reading resource D3044521:00000000:E9D119091BF95B17
Exception has been thrown by the target of an invocation.
----
Stack trace:
at System.RuntimeMethodHandle._InvokeConstructor(Object[] args, SignatureStruct& signature, IntPtr declaringType)
at System.RuntimeMethodHandle.InvokeConstructor(Object[] args, SignatureStruct signature, RuntimeTypeHandle declaringType)
at System.Reflection.RuntimeConstructorInfo.Invoke(BindingFlags invokeAttr, Binder binder, Object[] parameters, CultureInfo culture)
at s3pi.WrapperDealer.WrapperDealer.WrapperForType(String type, Int32 APIversion, Stream s)
at s3pi.WrapperDealer.WrapperDealer.GetResource(Int32 APIversion, IPackage pkg, IResourceIndexEntry rie, Boolean AlwaysDefault)
at S3PIDemoFE.MainForm.browserWidget1_SelectedResourceChanged(Object sender, ResourceChangedEventArgs e)
----

Exception has been thrown by the target of an invocation.
----
Stack trace:
at System.RuntimeMethodHandle._InvokeConstructor(Object[] args, SignatureStruct& signature, IntPtr declaringType)
at System.RuntimeMethodHandle.InvokeConstructor(Object[] args, SignatureStruct signature, RuntimeTypeHandle declaringType)
at System.Reflection.RuntimeConstructorInfo.Invoke(BindingFlags invokeAttr, Binder binder, Object[] parameters, CultureInfo culture)
at s3pi.GenericRCOLResource.GenericRCOLResourceHandler.RCOLDealer(Int32 APIversion, EventHandler handler, UInt32 type, Stream s)
at s3pi.GenericRCOLResource.GenericRCOLResource.ChunkEntryList..ctor(Int32 requestedApiVersion, EventHandler handler, Stream s, TGIBlock[] chunks, RCOLIndexEntry[] index)
at s3pi.GenericRCOLResource.GenericRCOLResource.Parse(Stream s)
at s3pi.GenericRCOLResource.GenericRCOLResource..ctor(Int32 APIversion, Stream s)
----

Expected zero, read 0x00000001 at 0x00000104

The only RSLT that gives this error other than the bookshelves is the phoneTableModern.

[Inge Jones]

Thanks for the report, will investigate.

[pljones]

Errors like "Expected zero, read 0x00000001 at 0x00000104" mean the Wiki says a valid file has a particular value at a particular location and the file being read did not have this value. Either the file is wrong or the wiki is wrong. Either situation is possible. However, if the file is in use and working as expected, it's more likely that the wiki is wrong! ;) In this case, it means we'll need to analyse the resources in question carefully to see if there are any other deviations from the wiki.

That's a long version of "Thanks for the report, will investigate" :lol:

[Inge Jones]

Actually I have just discovered we've had this or similar reported before. The treasure chest also has a slot that doesn't work in sp3e. The result of that, now I remember, is that in spite of spending many hours trying to analyse the alternative format, we were unable to create any meaningful algorithm so it was decided we would simply allow s3pe to error with dignity when it met one of these.

What I will try next, now that TSR Workshop allows slot editing, is see how that handles these objects. If it manages ok, I will ask the developers for any advice.

[cmomoney]

TSRW has an error with them also. The culprit may be the kinematic slot. There are no hash names and the matrix is all wrong. I put in (what I assumed was) the right bone name and matrix and the error went away(in TSRW).

[Inge Jones]

Ok I'll try to discuss it with Micke on Monday

[Tokfarbrorn]

I´ve had a look at the RSLT entries for the treasure chests, bookshelves and the modern phone all discovered that the "DWORD 0" after each section of slot entries, referenced here http://sims2wiki.info/wiki.php?title=Sims_3:0xD3044521 , isn´t always 0. For some entries the value is 1 and after that DWORD 7 follows 7 more floats. So this could either be a counter or a flag of some kind indicating that there are more data.

Im not completely sure that it really is 7 floats, it could also be 2 dwords and then 3 floats a 2 more dwords, or whatever.

I will update the wiki with this info.

/pommes a.k.a micke

[pljones]

You have far too many names ;). Thanks for posting the updates. I've incorporated them into the version of s3pi/s3pe Inge has for testing.

Now, you've written it like the field is boolean (and I made that explicit)... I've coded it like that but used a list that can hold zero or one entries, just in case we later find that it is actually a count. I could easily take the limit off s3pi but I don't want to risk breakage elsewhere, so I'll leave it in place till we find otherwise.

[Tokfarbrorn]

You have far too many names ;). Thanks for posting the updates. I've incorporated them into the version of s3pi/s3pe Inge has for testing.



Yeah, I know, i think pommes was either taken or I just had a blackout when registering \=


Now, you've written it like the field is boolean (and I made that explicit)... I've coded it like that but used a list that can hold zero or one entries, just in case we later find that it is actually a count. I could easily take the limit off s3pi but I don't want to risk breakage elsewhere, so I'll leave it in place till we find otherwise.

Yep, Since Im unsure if that really is a counter I treat it as a boolean until, like you said, we find it to be something else.
I´ll make a batch run and see if I can find some RSLT´s with another value than 1.

/pommes

[pljones]

Thanks -- s3pi/s3pe with the fixed RSLT in the new QA thread.