« first day (732 days earlier)      last day (2448 days later) » 
00:00 - 16:0016:00 - 22:0022:00 - 00:00

Hosch250
Hosch250
4:01 PM
Do you realize we have 899 files for R# to remove unneeded using directives from?
 
Mat's Mug
Mat's Mug
wow
 
Duga
Duga
[Hosch250/Rubberduck] Hosch250 pushed commit ddc772dd to InspectionBugs: General clean-up
> This resolves correctly to an event handler.
 
Hosch250
Hosch250
@Mat'sMug Just two criticals after you merge my rename PR.
 
Duga
Duga
[rubberduck-vba/Rubberduck] build for commit ddc772dd on unknown branch: AppVeyor build succeeded
 
Hosch250
Hosch250
@Mat'sMug Two passing PRs.
 
Duga
Duga
4:14 PM
[rubberduck-vba/Rubberduck] Hosch250 pushed commit a2a3ad0b to next: Fix built-in event discovery
 
Mat's Mug
Mat's Mug
nope. none.
 
Duga
Duga
[rubberduck-vba/Rubberduck] Hosch250 pushed commit 961719ff to next: Remove tests requiring us to load an external library
[rubberduck-vba/Rubberduck] Hosch250 pushed commit c43884f0 to next: Merge branch 'KillFailingTests'
[rubberduck-vba/Rubberduck] Hosch250 pushed commit be9c7b41 to next: Merge branch 'next' of github.com/rubberduck-vba/Rubberduck into InspectionBugs
[rubberduck-vba/Rubberduck] Hosch250 pushed commit ddc772dd to next: General clean-up
 
Hosch250
Hosch250
Can we release tonight?
 
Duga
Duga
Merge pull request #1721 from Hosch250/InspectionBugs

Fix built-in event discovery
 
Mat's Mug
Mat's Mug
hopefully
 
Duga
Duga
4:15 PM
 
Hosch250
Hosch250
The only thing new is the Workbook/Worksheet event thing.
And that could take a long time to get fixed.
 
Duga
Duga
[rubberduck-vba/Rubberduck] Hosch250 pushed commit e9e55460 to next: Close #1506
Merge pull request #1725 from Hosch250/Issue1506

Fix renaming project bugs
 
Mat's Mug
Mat's Mug
I'd like you to take 2-3 hours to take random VBA code on CR and/or SO, and use Rubberduck as much as you can, test everything as a user would, and create issues for everything worth noting, regardless of it being a showstopper or not. if you can quick-fix something, do it. I expect inspections and refactorings to work flawlessly by now.
3
 
Duga
Duga
> We have four sort orders implemented. Closing this as status-declined.
 
Hosch250
Hosch250
@Mat'sMug I don't know how a user would test RD :/
 
Mat's Mug
Mat's Mug
4:18 PM
lol
 
Hosch250
Hosch250
I was just about to tweak the CE again.
1686 and maybe 1616.
 
Zak
Zak
@Hosch250 I'd say leave it with me, but RD has never successfully parsed anything on my computer.
 
Hosch250
Hosch250
@Zak :/
You are running the alpha, right?
 
Duga
Duga
[rubberduck-vba/Rubberduck] build for commit eaa0012c on next: AppVeyor build succeeded
 
Zak
Zak
@Hosch250 yep
 
Hosch250
Hosch250
4:21 PM
That wouldn't cut it.
We are doing final testing for a beta release.
 
Duga
Duga
Merge pull request #1721 from Hosch250/InspectionBugs

Fix built-in event discovery
Merge pull request #1725 from Hosch250/Issue1506

Fix renaming project bugs
[Hosch250/Rubberduck] Hosch250 pushed commit 7a2e04c9 to Issue1686: Close #1868
 
Hosch250
Hosch250
OK, now I'll do that testing.
 
Duga
Duga
[Hosch250/Rubberduck] Hosch250 deleted branch InspectionBugs
[Hosch250/Rubberduck] Hosch250 deleted branch Issue1506
[Hosch250/Rubberduck] Hosch250 deleted branch KillFailingTests
[Hosch250/Rubberduck] Hosch250 deleted branch Issue1712
[Hosch250/Rubberduck] Hosch250 deleted branch Issue1680
[Hosch250/Rubberduck] Hosch250 deleted branch Parser
[rubberduck-vba/Rubberduck] Hosch250 pushed commit 7a2e04c9 to next: Close #1868
Merge pull request #1726 from Hosch250/Issue1686

Adjust CE node icon position
[rubberduck-vba/Rubberduck] build for commit 7a2e04c9 on next: AppVeyor build cancelled
BUILD FAILURE!
[rubberduck-vba/Rubberduck] build for commit 3e0a5bf1 on next: AppVeyor build succeeded
> Haven't had a chance to re-test this. Hopefully I'll just close this issue tonight. Looks like we're good for a beta then.
 
Mat's Mug
Mat's Mug
@Hosch250 can you try to repro #1644? looks like there's really only 1 critical issue, and it has to do with app teardown
 
Hosch250
Hosch250
OK.
 
Mat's Mug
Mat's Mug
4:30 PM
I mean #1644 is possibly fixed already
 
Duga
Duga
> Reference #1726
 
Hosch250
Hosch250
@Mat'sMug Pasting that code in parses just fine.
Only change made was sticking ptrsafe in.
 
Duga
Duga
[rubberduck-vba/Rubberduck] build for commit 7587f7c5 on next: AppVeyor build succeeded
 
Mat's Mug
Mat's Mug
you have to paste it into a text file, save as "ProgressIndicator.cls" and import as a class module
well I suppose it won't matter, since the attributes are part of the grammar anyway
mind trying, still?
 
Hosch250
Hosch250
I just did that.
Still no crash.
 
Mat's Mug
Mat's Mug
4:35 PM
awesome. close it.
 
Hosch250
Hosch250
The VBE automatically removed that Attributes line in the middle of the file.
It worked even when I stuck that back in, though.
 
Duga
Duga
 
Hosch250
Hosch250
About 1633...
 
Mat's Mug
Mat's Mug
the attributes don't show up in the VBE, but they do when the module is exported.
yeah?
 
Hosch250
Hosch250
I'm not sure what's up there.
That might be fixed too--I don't know.
 
Mat's Mug
Mat's Mug
4:37 PM
I think @Comintern has more than a clue here
 
Duga
Duga
> Can you check this with the latest build to see if it is fixed?
> Can you check this with the latest build to see if it is fixed?
 
Hosch250
Hosch250
Oh, OK.
 
Mat's Mug
Mat's Mug
lunch time
 
Hosch250
Hosch250
See you.
 
Comintern
Comintern
@Hosch250 - I'm pretty sure that #1633 is caused by a parser thread trying to access a COM object after the host application has torn it down.
 
Hosch250
Hosch250
4:41 PM
OK.
 
Duga
Duga
> This is part of the form, as stated by ThunderFrame. However, this will not stay visible even if there is a parse error. Finally, is this an issue, or just to request that this be built into the form if it wasn't already?
 
Mat's Mug
Mat's Mug
@Duga @Hosch250 can you try to see under which circumstances the busy indicator might remain stuck visible? Looks like there's a bug somewhere there
Don't spend all day on it though ;-)
 
Hosch250
Hosch250
@Mat'sMug I haven't found any.
 
Mat's Mug
Mat's Mug
Mkay
 
Hosch250
Hosch250
> Running this code in Excel 2013, the Cc = ... expression throws an "Expression too complex" error.
Wow. 
Cc = 1 - (1 / (2 * n + 1)) * (YP / (YP + K * ((3 * n + 1) * q / (n * WorksheetFunction.Pi * (ID / 2) ^ 3)) ^ n))
Please...
Make that two pleases. One for the OP, one for VBA.
 
Mat's Mug
Mat's Mug
4:59 PM
@Hosch250 where's that from?
 
Hosch250
Hosch250
0
Q: Wrong Data Type in Function

Michael SlivaI'm getting the ever so annoying "A value used in this formula is the wrong data type." error and I can't for the life of me figure out what is causing it. Public Function Cc(q As Double, ID As Double) As Double Dim YP As Double 'lb/100ft^2 Dim Fan600 As Single Dim Fan300 As Single D...

vba excel-vba
 
Mat's Mug
Mat's Mug
I'm not too familiar with VBA. But I do know it has some "interesting" behaviors that aren't present in most other languages. — Chris R. Timmons 1 hour ago
lol
now the question is: does our expression evaluator choke on it?
 
Hosch250
Hosch250
Bug detected.
 
Mat's Mug
Mat's Mug
quick! squish it!
what is it?
 
Duga
Duga
> Removes the entire line, regardless of what else is on the line.
 
Mat's Mug
Mat's Mug
5:04 PM
hmm that sounds vaguely familiar
 
Hosch250
Hosch250
Shouldn't be too hard to fix--just a copy/paste of by variable-remover.
I really, really need to extract that out somewhere.
 
Mat's Mug
Mat's Mug
see #740
 
Hosch250
Hosch250
Not assigned works fine.
I must've fixed that one and missed the other.
 
Mat's Mug
Mat's Mug
wasn't always the case
yep
the cost of duplicate code I guess
 
Hosch250
Hosch250
@Mat'sMug Appears to work.
 
Mat's Mug
Mat's Mug
5:07 PM
nice :)
@awgaya FYI your expression evaluator can handle things VBA chokes on. Well done!!
 
Hosch250
Hosch250
And the Rename selection needs to be overridden. 
var newInstruction = string.Empty;
var newCodeLines = string.IsNullOrEmpty(newInstruction)
    ? string.Empty
    : originalCodeLines.Replace(originalInstruction, newInstruction);
Whoever wrote that bug should be flippin' burgers :P
 
Mat's Mug
Mat's Mug
♪
 
Hosch250
Hosch250
var newCodeLines = originalCodeLines.Replace(originalInstruction, string.Empty);
 
Mat's Mug
Mat's Mug
wait, wat??!
lol!
I blame lack of sleep
 
Hosch250
Hosch250
Doesn't work.
I'm making that one share a quickfix with Not Assigned instead of ID not used.
NVM.
 
Duga
Duga
5:22 PM
[Hosch250/Rubberduck] Hosch250 pushed commit 3e5bc553 to InspectionBugs: Fix Use Meaningful Name selection and Remove Identifier Not Used--Variable quick fix
> …iable quick fix
[rubberduck-vba/Rubberduck] build for commit 3e5bc553 on unknown branch: AppVeyor build failed
BUILD FAILURE!
 
Mat's Mug
Mat's Mug
@Hosch250 that broke two tests/inspections
 
Hosch250
Hosch250
I saw that.
Looking at it now.
 
Mat's Mug
Mat's Mug
protip: run the tests locally before you push :)
 
Hosch250
Hosch250
You know what broke them? Not just wiping the entire line for non-variables.
This entire inspection is so buggy...
 
Mat's Mug
Mat's Mug
lol
 
Hosch250
Hosch250
5:31 PM
Of course, if someone has multiple procedures on a single line, they deserve everything they get, and then some.
 
Mat's Mug
Mat's Mug
5:46 PM
oh wow, a VBA dupe-hammer! stackoverflow.com/q/37685509/1188513
 
Duga
Duga
[Hosch250/Rubberduck] Hosch250 pushed commit 6f1bc8f2 to InspectionBugs: Fix broken tests
[rubberduck-vba/Rubberduck] build for commit 6f1bc8f2 on unknown branch: AppVeyor was unable to build non-mergeable pull request
BUILD FAILURE!
Merge pull request #1726 from Hosch250/Issue1686

Adjust CE node icon position
[Hosch250/Rubberduck] Hosch250 pushed commit e089115e to InspectionBugs: Merge branch 'next' of github.com/rubberduck-vba/Rubberduck into InspectionBugs
 
Hosch250
Hosch250
6:01 PM
TFL.
 
Duga
Duga
[rubberduck-vba/Rubberduck] build for commit e089115e on unknown branch: AppVeyor build succeeded
 
Mat's Mug
Mat's Mug
@Duga first starred repo!
 
Duga
Duga
[rubberduck-vba/Rubberduck] Hosch250 pushed commit 3e5bc553 to next: Fix Use Meaningful Name selection and Remove Identifier Not Used--Variable quick fix
[rubberduck-vba/Rubberduck] Hosch250 pushed commit 6f1bc8f2 to next: Fix broken tests
[rubberduck-vba/Rubberduck] Hosch250 pushed commit e089115e to next: Merge branch 'next' of github.com/rubberduck-vba/Rubberduck into InspectionBugs
Merge pull request #1728 from Hosch250/InspectionBugs

Fix Use Meaningful Name selection and Remove Identifier Not Used--Var…
[rubberduck-vba/Rubberduck] build for commit 4bfc3a19 on next: AppVeyor build succeeded
> Reference #1728
> If comintern is handling this, then I'll unassign myself.
> If comintern is handling this, then I'll unassign myself.
> Is this still an issue?
> What's up here?
> obviously not :wink:
> I'm happy with the way things are right now. I think.
> The "out-of-sync-ness" of token position is by design, since the source string for attribute tokens is not the same source string as the one used for codepane code. What we need is another way of reporting parsing errors in the attributes pass.
> @Hosch250 any ideas about this one?
 
Mat's Mug
Mat's Mug
7:18 PM
crickets
 
Duga
Duga
7:28 PM
> I have not had chance to try it due to a large project that is taking my time. Had the release been updated, if do I will try tomorrow.

-----Original Message-----
From: "Mathieu Guindon" <notifications@github.com>
Sent: ‎07/‎06/‎2016 19:33
To: "rubberduck-vba/Rubberduck" <Rubberduck@noreply.github.com>
Cc: "Talorthain" <anthony_taylor7@btinternet.com>; "Mention" <mention@noreply.github.com>
Subject: Re: [rubberduck-vba/Rubberduck] Code Inspections doesn't open orcrashes VBE (#1237)

Closed #
> @Talorthain thanks! We'll try to issue a beta release tonight (shhh! it's a seekrit!) :wink:
 
Mat's Mug
Mat's Mug
@Hosch250 any update on #1264?
 
Duga
Duga
> @Hosch250 got a repro on this one?
 
Mat's Mug
Mat's Mug
7:46 PM
looks like there are still resolver issues
 
Duga
Duga
 
Mat's Mug
Mat's Mug
ugh. there's so many bugs
 
Duga
Duga
> @ThunderFrame `For` and `For Each` loops *are* treated as assignments. Consider:

Dim i
For i = 1 To 42
Debug.Print i
Next

Local variable `i` is both assigned and referred to here, and that code shouldn't fire up an inspection result for `i` being not used or not assigned.
> @ThunderFrame `For` and `For Each` loops *are* treated as assignments. Consider:

Dim i
For i = 1 To 42
Debug.Print i
Next

Local variable `i` is both assigned and referred to here, and that code shouldn't fire up an inspection result for `i` being not used or not assigned.
> I think I might have forgot to reference this issue... can you get a repro with the latest build?
 
Duga
Duga
8:21 PM
> Well, your last example is invalid VBA, and breaks with Subscript out of range, because foo() is never Dimensioned.

```
Dim foo()
For i = LBound(foo) To UBound(foo)
foo(i) = i
Next i
```

Using the array bounds as for..loop bounds *without* assigning the array (before, or within the loop) is a usage before assignment.

```
'Foo is used but never assigned
Dim foo(6)
Dim bar(6)
For i = LBound(foo) To UBound(foo)
bar(i) = i
Next i
```
> > breaks with Subscript out of range, because foo() is never Dimensioned.

That's the idea: catch runtime error *before* runtime :smiley:
 
Hosch250
Hosch250
@Mat'sMug Catching up now--went driving and took my dog on a walk.
 
Duga
Duga
> Well, in that case it's a Usage of undimensioned array before it's a Usage before assignment
 
Hosch250
Hosch250
Boy, I wish people would include code instead of images...
I'm tempted to create a tag for that and close all the issues.
My other pet peeve is when people include the full stack trace formatted so funky, and 99% of it absolutely useless.
 
Mat's Mug
Mat's Mug
^^ agreed. feel free to edit 'em out.
 
Duga
Duga
> No idea. I thought @autoboosh was doing something about this.
> No, I haven't. I'll think about it and see if I come up with something.
> I'll work around the image this time, but please include the raw code as well as images in the future.
 
Mat's Mug
Mat's Mug
8:39 PM
@Hosch250 release notes for the beta will have to include the list of broken inspections under "known issues"
 
Hosch250
Hosch250
I don't know if any are "broken".
And some are working, but other things aren't.
 
Mat's Mug
Mat's Mug
foo = Range("A1") ' should be implicit reference to ActiveSheet
 
Hosch250
Hosch250
Like, Param Not Used doesn't work for Workbook_ event parameters because we don't resolve WorkbookEvent_ members right.
 
Mat's Mug
Mat's Mug
right
 
Hosch250
Hosch250
So, I can't exactly report that as an inspection not working, or can I?
@Mat'sMug Sure.
I thought I read there was something resolving to Range and triggering that incorrectly?
@Mat'sMug Nope.
 
Mat's Mug
Mat's Mug
8:43 PM
so that's one broken inspection
 
Hosch250
Hosch250
How do I trigger ActiveWorkbook?
@Mat'sMug I think it is the resolver, actually.
 
Mat's Mug
Mat's Mug
foo = Sheets("Sheet1")
@Hosch250 could very very well be
 
Hosch250
Hosch250
It picks up plenty of Range and Sheet and Cell declarations.
None of them have any references.
 
Mat's Mug
Mat's Mug
yeah, these shitz are everywhere in the frakkin' Excel object model
 
Hosch250
Hosch250
Looks like it is the resolver for Active Workbook too.
 
Mat's Mug
Mat's Mug
8:50 PM
makes sense
 
Duga
Duga
> No repro. Did you not refresh the parser after you added the call?
> @Hosch250 I think one of your recent commits might have fixed this.
> This issue should be resolved now--at least on the inspection side. The fix will not work for Workbook_ and Worksheet_ events because we are currently having trouble resolving them correctly.
 
Hosch250
Hosch250
@Mat'sMug Where are Class_Initialize and Class_Terminate defined?
I have an inkling that if we could load that, we could get a lot of good information.
 
Mat's Mug
Mat's Mug
ClassModuleEvt, within the VBA runtime library (which we're not loading)
 
Hosch250
Hosch250
Why not?
Would the VBA runtime library potentially contain information about what is a workbook?
 
Mat's Mug
Mat's Mug
nope
 
Hosch250
Hosch250
8:57 PM
OK.
 
Mat's Mug
Mat's Mug
that's Excel territory
 
Hosch250
Hosch250
Well, I'm sure I won't find anything useful, but I'm going to go exploring.
 
Rossco
Rossco
9:08 PM
Morning all.
 
Mat's Mug
Mat's Mug
hi!
 
Rossco
Rossco
What's the discussion around the Class_Initialize/Class_Terminate methods?
 
Hosch250
Hosch250
@Rossco RD is having trouble with events.
 
Rossco
Rossco
ah OK. Listening to events or understanding event expressions?
 
Hosch250
Hosch250
Resolving event handlers
 
Rossco
Rossco
9:12 PM
Ok, so understanding that the Class_Initialize is an event handler?
 
Hosch250
Hosch250
Yes.
We hard-coded that, but that isn't going to cut it for everything.
For example, we are having the same trouble with Worksheet/Workbook/Document/etc event handlers.
 
Vogel612
Vogel612
@Mat'sMug you may want to start prioritizing stuff more fine-grained
 
Rossco
Rossco
Yeah the class terms are actually defined in the VBA language specification as keywords
 
Vogel612
Vogel612
We're at 260 open issues
 
Mat's Mug
Mat's Mug
gosh, and 140 have no milestone
 
Vogel612
Vogel612
9:15 PM
That's not a real problem.
 
Mat's Mug
Mat's Mug
55 under v2.0
 
Vogel612
Vogel612
Also I just read the conversation at #1264 where nobody knew who was actually working on this
 
Rossco
Rossco
@Hosch250 yeah you get to deal with the following chestnut: 
Private WithEvents app As Application

Private Sub app_NewWorkbook(ByVal Wb As Workbook)
    ' he he he...
End Sub
 
Vogel612
Vogel612
Assignees could help. It's becoming a real management task, eh?
 
Mat's Mug
Mat's Mug
yeah
feels like all I'm doing is management and PR already
 
Hosch250
Hosch250
9:16 PM
@Rossco I think we resolve that.
 
Mat's Mug
Mat's Mug
I can't remember last time I fixed an issue myself
 
Hosch250
Hosch250
@Mat'sMug @Comintern This might actually cut it: fernandof.wordpress.com/2008/02/05/…
So, first we need to get the interfaces from the host.
 
Comintern
Comintern
9:33 PM
@Hosch250 - What object are you going to get an IUnknown from?
 
Mat's Mug
Mat's Mug
how do you get the __ComObject representing the actual worksheet?
 
Hosch250
Hosch250
@Comintern The VBComponent.
 
Mat's Mug
Mat's Mug
the type is known, that's a VBComponent
 
Hosch250
Hosch250
Oh.
 
Comintern
Comintern
I think I tried that - it doesn't inherit from the document base type. I'm still not 100% sure how it implements interfaces if it isn't a supertype.
 
Mat's Mug
Mat's Mug
9:35 PM
basically we've just inherited Excel's spaghetti mess
 
Comintern
Comintern
IKR
Could be worse - it could be Word's spaghetti mess.
 
Mat's Mug
Mat's Mug
and quite possibly no one at MS would be able to help us
 
Vogel612
Vogel612
actually... the spaghetti mess of Excel, Access, Word, Outlook, Notes and Powerpoint, as well as the spaghetti mess that goes by COM and the spaghetti mess of all VBE Hosts ever
 
Mat's Mug
Mat's Mug
@Comintern oh, better (?) - the whole Office suite's spaghetti mess
@Vogel612 yeah that :)
 
Comintern
Comintern
At least the Access spaghetti is served with a bunch of DAO sauce.
 
Vogel612
Vogel612
9:37 PM
don't forget the ADODB sauce that's in there somewhere, too
and add a bit of RegEx seasoning
 
Mat's Mug
Mat's Mug
I'm discouraged. so many bugs, so little time, so many hosts, so many libraries, so many interfaces, ...
perhaps Rubberduck is just outright too ambitious?
 
Vogel612
Vogel612
@Mat'sMug just release a beta, and get yourself a shot of motivation and user-gratefulness
also don't overwork yourself. You almost have a second full-time job with this :)
not to mention you're a mod, too
 
Mat's Mug
Mat's Mug
..and a slew of new issues
 
Hosch250
Hosch250
I think what we need is to make this like the unit tests for now.
 
Mat's Mug
Mat's Mug
aye, ...being a mod isn't the hard part ;-)
@Hosch250 meaning?
 
Hosch250
Hosch250
9:41 PM
Because this is so host-specific, we need to figure it out in a host-specific way.
That would make it quite easy, actually.
 
Mat's Mug
Mat's Mug
which brings us back to the plug-in architecture
 
Hosch250
Hosch250
For example, only workbooks implement property A, so we know that is a workbook.
 
Vogel612
Vogel612
@Mat'sMug sounds like 3.0?
 
Hosch250
Hosch250
Same for worksheets.
 
Mat's Mug
Mat's Mug
because going down that avenue without a plug-in architecture will become a huge mess
yeah
 
Hosch250
Hosch250
9:42 PM
Yep, we can go full-plugin in 3.0, but we need something before our final release.
 
Mat's Mug
Mat's Mug
I don't see the end of it. everything is blurry, I feel like utter shit.
TTQW I guess
 
Hosch250
Hosch250
I think we've gone too long without a stable release.
We bit off way more than we could chew with 2.0, but the end is in sight.
Been a good learning experience too.
And just think of all those critical issues that were closed in the past few days.
OK, which fire should I fight now?
 
Vogel612
Vogel612
any criticals left?
 
Hosch250
Hosch250
One.
Comintern is working on that.
 
Duga
Duga
9:57 PM
> I'd have to say this is kind of obsolete.
 
00:00 - 16:0016:00 - 22:0022:00 - 00:00

« first day (732 days earlier)      last day (2448 days later) »