« first day (1698 days earlier)      last day (1482 days later) » 

8:00 PM
lol that's inverted
 
I wouldn't know - as I said, I know precious little Latin. I do know that Latin is far more grammatically complex than English is, so I suppose of course translator is going to faceplant itself.
I guess being a SOV using SVO input doens't help either.
 
yea and the additional case doesn't help
 
It occurred to me that the translation would be also much harder than other translation because you'd have to deal with all technical jargon that likely does not exist in Latin, being a dead language.
Even if all words were successfully latinized, the native Latin speakers, were they alive, would be all "WTF is this"
 
'course
#DidYouKnow, there's Latin translations for the Harry Potter books
 
Oh? No, I didn't.
 
8:11 PM
Makes me wonder how many other software packages have a Latin localization.
@Vogel612 of course! If they can sell a copy, they'll make a translation...
 
but then again, does Harry Potter books really introduce that many new jargons? Probably not....
 
One of my old latin teachers literally had us translate passages of the book
 
wonders what the localization code for Latin is. la-IT?
 
like... "u wot m8?"
 
m8?
 
8:12 PM
mate
@FreeMan va_LT?
 
i was reading it as literally containing m8 in the book's text.
 
@this galleons and sickles and Hufflepuff and Slitherin, oh my!
@Vogel612 "Vatican - Latin"?
 
It's one thing to introduce a name, which likely will stay untranslated or at least latinized, but it's entire another thing to try and translate something like Windows Presentation Framework or Form (in the sense of a Windows dialog that collect inputs) or that kind of thing.
 
huh weird.
Serbian had an explicit locale for Latin characters in Java 7, but the locale was dropped in java 8
 
well, this is fun!
PPT just spewed that.
I've got some code in it & it was opened in the VBE & parsed in RD.
I had previously run the code, but PPT was just sitting there.
 
8:17 PM
wait, why are you using PPT?!?
:D
 
@FreeMan la_VA according to ISO 639-2 and ISO 3166
 
It may have been that I tried opening one PPT to soon after closing another one...
See the end of this message for details on invoking
just-in-time (JIT) debugging instead of this dialog box.

************** Exception Text **************
System.Runtime.InteropServices.InvalidComObjectException: COM object that has been separated from its underlying RCW cannot be used.
   at System.StubHelpers.StubHelpers.GetCOMIPFromRCW(Object objSrc, IntPtr pCPCMD, IntPtr& ppTarget, Boolean& pfNeedsRelease)
   at Microsoft.Vbe.Interop._VBProject.get_References()
   at Rubberduck.VBEditor.SafeComWrappers.VBA.VBProject.get_References() in C:\projects\rubberduck\Rubberduck.VBEditor.VBA\Safe
@this Death. Death by PowerPoint. It's what we do to our clients.
Great Scott!!! That's why they're all leaving us!!
 
I would open a new issue.
 
w/MS or RD? :)
or my boss...
 
I have a hunch it's high time that all dialogs start listening to the terminate event.
RD
but go ahead and open an issue with MS and boss, too. :D
 
8:25 PM
> **Rubberduck version information**
Version 2.4.0.4491
OS: Microsoft Windows NT 10.0.15063.0, x64
Host Product: Microsoft Office 2016 x64
Host Version: 16.0.4266.1001
Host Executable: POWERPNT.EXE

**Description**
I had a PPTM open with a code module opened in the VBE & parsed by RD. I opened a 2nd PPT (without code), closed it and opened a 3rd PPT (again without code). Instead of opening the 3rd PPT, I got the error message.

**To Reproduce**
Steps to reproduce the behavior:
See "D
 
@Vogel612 "Latin_Vatican" got it. That makes sense!
@this done!
 
@Duga yay! 730!!
 
Stars - Issues = 1
 
8:42 PM
"I don't plan on sharing them with anyone...Here are some quotes:..." — maxathousand 3 hours ago
 
TBH, if OP hadn't said "64 year old guy" he could have been referring to me at times.
(f'n Russian).
 
...said the commie worker
 
I didn't say it narrowed it down, comrade.
 
i guess that's what they really meant by "shock brigade"
 
8:51 PM
Actually, once you're outside of earshot of the customer service department, we could probably be mistaken for a gang of sailors.
Huh. This is interesting:
2019-01-29 13:54:21.9474;TRACE-2.4.0.4491;Rubberduck.App;Checking for legacy Smart Indenter settings.;
2019-01-29 13:54:21.9943;TRACE-2.4.0.4491;Rubberduck.UI.CustomComWrappers.Wrapper_IOleClientSite;IOleInPlaceSite::OnPosRectChange() called;
2019-01-29 13:54:22.0290;TRACE-2.4.0.4491;Rubberduck.VBEditor.WindowsApi.SubclassManager;Subclassed hWnd 0x00320ECC as CodePane.;
Note that the logs from #4757 have a resize event hitting before the SubclassManager gets a hold of the window. This might be what was happening in #4722 - I just couldn't get that to happen in that order.
 
Stars - Issues = 2 !!!
 
but in 4722, he didn't get that dialog we see in 4757?
or is that an artifact of machine's debug configuration?
 
No, it crashed because of the bad value in the message pump. The 2 aren't related - it just looks like the log comes closer to meeting the conditions of 4722 than I've ever been able to in repro attempts.
 
@MathieuGuindon heh... we did this in college. Had a Taiwanese (?) prof who taught a Lisp class. We started a translation dictionary of all his Engrish sayings. Plopped it on the university's VAX system. I heard rumors that it was still floating 'round campus 25+ years later!
 
I wonder if I should try to repro it in PPT.
 
8:58 PM
Fortunately, we weren't worried about being PC back then... ;)
TTQW see y'all on the flip side!
 
Wow, I didn't get far enough in that log file. It disconnected ALL THE THINGS.
 
That's why I thought we need to listen to the Terminate event
 
> 2019-01-29 15:15:46.0393;ERROR-2.4.0.4491;Rubberduck.UI.UnitTesting.Commands.AddTestModuleC‌​ommand;System.Runtime.InteropServices.COMException (0x8001010A): The message filter indicated that the application is busy. (Exception from HRESULT: 0x8001010A (RPC_E_SERVERCALL_RETRYLATER))
The VBE is busy?
 
because from my earlier glance, it looks like ComSafe was tearing itself down but codeexplorer is still "where's my junk? I want it now")
yeah that does happen.
you have to retry
 
We don't systemically handle that in the wrappers though.
I'm not sure I care TBH.
You're probably screwed by then anyway.
 
9:02 PM
I once posted a sample of how I handle that.
 
@this It's barfing an the xaml files that have a DesignInstance set
 
2019-01-29 15:15:44.9422;DEBUG-2.4.0.4491;Rubberduck.Parsing.VBA.RubberduckParserState;Project '142030726' was added.;
2019-01-29 15:15:44.9890;DEBUG-2.4.0.4491;Rubberduck.Parsing.VBA.ParseCoordinator;Parsing run started. (thread 30).;
2019-01-29 15:15:44.9890;INFO-2.4.0.4491;Rubberduck.Parsing.VBA.RubberduckParserState;RubberduckParserState (9) is invoking StateChanged (Started);
2019-01-29 15:15:45.0283;DEBUG-2.4.0.4491;Rubberduck.Parsing.VBA.RubberduckParserState;Project 142030726 was renamed.;
That looks like a problem in an EvaluateCanExecute somewhere. You shouldn't be able to do that during a parse.
 
here's the code that dumb chat search won't find for me: chat.stackexchange.com/transcript/14929?m=44222841#44222841
That ensures that VBA eventually gets told about the event being raised.
 
Yeah, except you can't do that on the UI thread because it sleeps the VBE too.
 
@IvenBach using true or false?
 
9:11 PM
When nothing is set.
I thought it was on all files but I was too hasty.
 
confused did you try that w/ the setting?
@Comintern Right. That code sample I posted was from a different project - It had a winforms dialog, though. The code was ran by a background worker, IIRC.
 
I tried adding that element to all csproj files and still get the same errors.
AlwaysCompileMarkupFilesInSeparateDomain being set to true that is.
 
I think we might need to turn the build logging volume up to 11 (Diagnostic).
 
@this Is a background worker a foreign worker that clandestinely contracts out to an employee, rather than a company?
 
Not sure how to fix that one.
<?xml version="1.0" encoding="utf-8"?>
<Project Sdk="Sunburst.NET.Sdk.WPF/1.0.47">
  <PropertyGroup>
    <RootNamespace>Rubberduck</RootNamespace>
    <AssemblyName>Rubberduck.Core</AssemblyName>
    <Title>Rubberduck.Core</Title>
    <Product>Rubberduck.Core</Product>
    <Copyright>Copyright © 2014-2018</Copyright>
    <ProjectGuid>{A1587EAC-7B54-407E-853F-4C7493D0323E}</ProjectGuid>
    <DocumentationFile>bin\Debug\Rubberduck.Core.xml</DocumentationFile>
    <!-- Disable "Missing XML documentation" warning (CS1591) -->
^ Is that the correct location for the element?
@Comintern I'm willing to up it to get this build problem figured out.
WTWTCH What's the worst that could happen...
 
9:33 PM
Fill your HDD with log file?
I'd clean first, then only build Core.
 
It's a redirected desktop. Likely it's actually storing it on the server. Hrm....
:thinks: Fill up the server with 5TB of log files.
Where are the log files actually stored?
:pout: exceed 512kb paste limit...
 
ok, giving in
0
Q: Filtering ComboBox.ItemsSource

Mathieu GuindonI'm trying to make a ComboBox filter its ItemsSource based on a FilterValue string bound to its Text; the control is inside the CellEditingTemplate of a DataGrid, which is itself inside a DataTemplate. The items all implement an ICodeName interface that guarantees there's a Code and a Name proper...

 
15mb text file
 
FML.
OK, pull the bit out where it explains what the issue is and post that.
</ joking> But seriously, can you dropbox it or something (zipped first).
 
Logging reduced down to minimal. If not I'd forget about it.
 
9:46 PM
> 1>DotnetCliToolTargetFramework = netcoreapp2.1
1>DownloadNuGetExe = false
1>EmbeddedWin32Manifest =
1>EmitAssetsLogMessages = true
1>EnableBaseIntermediateOutputPathMismatchWarning = true
1>EnableDefaultCompileItems = true
1>EnableDefaultEmbeddedResourceItems = true
1>EnableDefaultItems = true
1>EnableDefaultNoneItems = true
1>EnableDefaultWpfItems = True
1>EntityDeployDependsOn =
1>EntityDeployIntermediateResourcePath = obj\Debug\net46\edmxResourcesToEmbed\
Oh wait, that's a friggin template for an error.
No wonder this thing is 15MB
 
0
Q: Generating a data validation list and sorting it alphabetically

dwironyThe following code clears 11 columns that contain lists that are used for data validation on another worksheet. This list is constantly changing due to records being added/removed from the source data. Basically I'm checking every value in a column, and if it meets all of my criteria I'm adding i...

 
@Hosch250 LOL. Nope. BackgroundWorker class
 
10:13 PM
Use math, not the Excel object model. Find the last row with data in the sheet, then subtract the last row number of the named range. — Comintern 20 secs ago
Seriously people?
 
@Comintern makes me think of the computer scientist:
 
lol
 
10:29 PM
@IvenBach that's just a discrepancy between project name and location, nothing serious
@FreeMan oddly appropriately it's also lava
 
@this 10!
 
What he ^ said...
 
@IvenBach I don't know. I'd have to ask the computer.
picks up the mouse Computer?
Computer?
No response. Guess it's not working.
@IvenBach yes
 
10:45 PM
@this I'm not answering until I see the implementation of that + operator
the answer could very well be 11
 
LOL
 
LOL
Also depends on whether or not either of them is a Variant...
 
If it's JavaScript, meh. any old answer will do.
 
notes that Mat is extra anxious around JavaScript, jots it down
 
10:52 PM
@this Then I tried it and still get build failure.
 
even with the other value?
 
the other value being false? Extreme duck check.
False --> Fail
True --> Fail
Deleted `AlwaysCompileMarkupFilesInSeparateDomain` element --> Fail
Consistency, I hath found thee.
 
oh well, not the blue pill we need
(or was it red pill?)
 
cosmic rays
 
treatment required: coffee exposure.
 
10:57 PM
Depends on whether or not we're considering RD the matrix.
 
I'm jaded I didn't get super powers from them.
 
well.... Ben Grimms might rather be you.
 
Unless my super power ends up being Sterilized Man "He's shooting blanks!"
 
pretty sure Ben would rather be that than be the thing.
 
Didn't he actually want a wife and kids in the comic?
 
11:00 PM
> When compiling using the MIDL branch, MIDL will additionally generate C files -- usually in the format of `<filename>_i.c` and `<filename>.h` and possibly more.

Because the build tasks drop them in the same output folder (simply out of pure laziness), the installer then proceeds to pick those up and package them. Which is... odd.

We probably should designate a temporary folder for the MIDL to unload its junk, so we can only copy the TLB files and delete the folder as part of the build pr
[rubberduck-vba/Rubberduck] bclothier is bored so why not move a project card
 
@IvenBach I can't remember if he already had a wife but yes, it was a big deal for him because everyone but him can basically be "normal human" but he's permanently.... thingified.
@Vogel612 do you want to pick up the 4758?
 
11:23 PM
@all, would like to get a consensus on how to go forward with merging PRs that has been laying around. We have 4072, 4686 and 4699 ready to be merged, I think. Other are WIP/blocked. Both 4686 and 4699 are much smaller in the scope and thus has less chance of breaking something. 4072 needs lot of field testing before it ever could make into a green release. Thoughts?
 
None of those have much overlap in scope. I'm not thinking there will be much interaction between them.
 
Nope, I don't expect so, but being a Chicken Little .... checks sky one more time to be sure, I don't want to do a mega-merge-fest then have a messy issue that's a result of combination of things.
 
Oh, I know. I'm just not seeing it. If 4699 or 4686 take the wheels off, it's not going to be in the refactoring dialogs.
4699 would be much more spectacular than that.
 
yeah considering that it's a part of the startup
Actually keep in mind both 4699 and 4072 modify the IoC.
 
If the IoC fails, that qualifies as spectacular too.
(and creates a message box about the size of Iven's last build log)
 
11:37 PM
yeah. I love how they don't let you see the buttons, too.
 
Or the last 1000 lines of the message.
 
why would this not get picked up?
(this is after parsing, btw)
 
Gimme a sec, lemme check the condition.
 
woah, after i opened the node the icon updated
 
ClassModuleDeclaration.IsInterface must be returning false.
Lazy evaluation?
 
11:42 PM
i'm 99% sure I saw it working correctly when it was loaded via opening
 
This is the relevent LoC
            if (component.Declaration is ClassModuleDeclaration classModule && classModule.IsInterface)
            {
                return InterfaceIcon;
            }
 
but then I was working away and I noticed it wasn't interface anymore.
oh wait, I might have a repro
let me verify quickly.
 
Which points to... public bool IsInterface => _subtypes.Count > 0;
Note that it isn't checking the annotation at all.
 
no, nothing to do with that, I think.
here's the steps to reproduce.
1) open a folder that contains an interface and maybe a class, too.
2) right-click the folder, Add module, add a class module
3) after it finishes parsing the new module's added, but the interfaces lose their interface icons.
4) click on the interface node to expand it. It get its icon back.
 
Probably just missing an OnPropertyChange().
 
11:47 PM
just tested in a different folder that only had an interface. That reverted the icon, too.
and this time that was for standard module, so yeah, not specific to what was being added; just adding anything will cause the loss of interface icon.
Interestingly, removing the newly added module will restore the icon.
 
Hmm... that should be calling OnPropertyChange("Declaration") when it recycles.
Now I'm wondering if the icon doesn't get set if the node isn't visible.
Go ahead and open an issue for it. If it is the visibility effecting the change notification, that's easy to sort out.
 
@this noticed that too when I took the shots for the blog post
 
Also tied to the collapsed state?
 
didn't notice that
was toying with adding/removing the @Interface annotation and reparsing
so... the missing annotation had nothing to do with the icon switching back to 'class module'??
 
Yeah, the annotation isn't currently checked - it's only considered an interface if it's being implemented.
This is the binding:
                        <Image.Source>
                            <MultiBinding Converter="{StaticResource NodeToIcon}">
                                <MultiBinding.Bindings>
                                    <Binding RelativeSource="{RelativeSource Self}" Path="DataContext"/>
                                    <Binding Path="Declaration" />
                                    <Binding Path="IsExpanded" />
                                </MultiBinding.Bindings>
                            </MultiBinding>
 
11:58 PM
hmm. it probably should account for it.
 
It should re-evaluate when the expanded state is changed, but if it isn't visible when the declaration notification arrives, that might be getting suppressed.
 

« first day (1698 days earlier)      last day (1482 days later) »