RELOAD!

[rubberduck-vba/Rubberduck] 3 opened issues. 1 closed issue. 12 issue comments.

[Minesweeper] Games Played: 97, Bombs Used: 73, Moves Performed: 12952, New Users: 10

Night pond

[rubberduck-vba/Rubberduck] hdejager starred us

[rubberduck-vba/Rubberduck] eugene-plutoflume starred us

@IvenBach you can edit the last commit using git commit --amend. everything else is generally more involved because you're rewriting parts of history

a rebase -i can allow you to rewrite the commit message

[rubberduck-vba/Rubberduck] daFreeMan created comment on issue #5008: Add ~> Existing Module

Uninstalling 2017 on my works desktop, 2019 proven stable enough for me :-) (Keeping it on the laptop for RD of course :-) )

uh, I don't think RD builds on 2019.

yep, that's why I'm keeping 2017 on the laptop :-)

#ReadFail

remind me when Core3.0 is out of preview

should be september-ish iiuc

I was thinking Tuesday, around 16:30.

;)

I have what appears to be a transient, one-time issue with the parser finishing with no reported error, but a blank CE. There are a ton of NREs in the log.

Closing Excel & reopening seems to restore proper behavior. I've written up an issue and have the log uploaded. Should I submit it or ignore it as a "that's what you get when you leave Excel open for 3 days" problem?

It was also not responding to double clicks on a method in the CE (which lead to the parse which led to that issue). The reboot of Excel seems to have resolved that as well.

@FreeMan sounds like the latter...

a Log would help shedding some light

I'm sure it is due to having Excel open since Monday morning (but that's my usual habit...)

I'll post it (log's already attached). It can get closed if it's determined it's just a transient and there's nothing useful in the log to help track it down.

[rubberduck-vba/Rubberduck] daFreeMan opened issue #5021: Parse leaves "Ready" state, but empty CE

[rubberduck-vba/Rubberduck] daFreeMan added label bug to issue #5021: Parse leaves "Ready" state, but empty CE

It would seriously threaten my computing way of life to admit that leaving things open for weeks (um er I mean days) makes bugginess my fault. We must suppress this idea!!

[rubberduck-vba/Rubberduck] daFreeMan opened issue #5022: Glitch in the SCP matrix

[rubberduck-vba/Rubberduck] daFreeMan added label bug to issue #5022: Glitch in the SCP matrix

@MarkBalhoff heavily guilty of that! I do tend to reboot my work machine every Friday night but only because IT sometimes pushes unannounced patches over the weekend and I hate having my machine rebooted with a ton o' stuff open.

I reboot the home machine only when absolutely necessary.

looks back with horror to the time when it was considered Good Practice™ to reinstall Windows every 3-6 months "just to clean up all the crap"

@FreeMan could you check your VBAProject references for a reference to Rubberduck?

I think that's tripping one of the NREs

big question, is it worth it to upgrade to 2019?

@KySoto I'm not seeing much new

from 2017 that is

the splash screen is pretty

lol

because to be fair, most of my projects are vb.net (since thats what the boss wants)

i mainly wanted it because .net core 3 is likely to be better supported there when it goes GA

is 2017 not going to support the big unified .net?

possibly, but not as well supported iiuc. hacks needed.

today's MS doesn't keep old versions fresh for long...

@Vogel612 ^

@FreeMan yea you might want to shoot that. IIRC unit testing is in a separate TLB

Also, just noticed that @Comintern's lovely new Reference Explorer isn't launching! :( 0_o !!!

@Vogel612 #Bang

hrmm ... something's weird there...

I'm pretty sure that the following line can not throw an NRE

}

I dunno....

[rubberduck-vba/Rubberduck] Vogel612 created comment on issue #5021: Parse leaves "Ready" state, but empty CE

[rubberduck-vba/Rubberduck] Vogel612 edited comment on issue #5021: Parse leaves "Ready" state, but empty CE

fixed a link ...

[rubberduck-vba/Rubberduck] daFreeMan created comment on issue #5021: Parse leaves "Ready" state, but empty CE

@Vogel612 I’ve been doing both for a while now. Rebase -i gives me conflicts often though and I don’t know why.

[rubberduck-vba/Rubberduck] MDoerner created comment on issue #5021: Parse leaves "Ready" state, but empty CE

> QualfiedName should not be able to return null since the return type is a struct.

[rubberduck-vba/Rubberduck] SmileyFtW opened issue #5023: Code inspection result "Object variable 'foo' is assigned without 'Set' keyword"

[rubberduck-vba/Rubberduck] SmileyFtW added label support to issue #5023: Code inspection result "Object variable 'foo' is assigned without 'Set' keyword"

[rubberduck-vba/Rubberduck] retailcoder created comment on issue #5023: Code inspection result "Object variable 'foo' is assigned without 'Set' keyword"

> That should be a Property Set member, but the inspection is confused regardless (a number of issues are open for it), presumably because of the UDT-member nature of foo here.

[rubberduck-vba/Rubberduck] SmileyFtW created comment on issue #5023: Code inspection result "Object variable 'foo' is assigned without 'Set' keyword"

[rubberduck-vba/Rubberduck] Vogel612 edited issue #5023: Code inspection result "Object variable 'foo' is assigned without 'Set' keyword"

@Duga -.- at least I can fix the formatting now...

[rubberduck-vba/Rubberduck] Vogel612 added label feature-autocomplete to issue #5022: Glitch in the SCP matrix

[rubberduck-vba/Rubberduck] SmileyFtW created comment on issue #4788: Todo explorer organization at the project level, not stop at module level

> Kinda related, but maybe not: When expanding or collapsing the up/down arrows, why does the entire display expand/collapse instead of just the selected node. This implementation seems to be throughout RD. All other implementations I am familiar with don't operate the way RD does. If expanding/collapsing the entire tree is the desired action there is usually an expand/collapse all somewhere. Selecting a single node expands/collapses that single node.

An official date for my Excel Lunch&Learn was scheduled. Lunch is even planned. Here's hoping I don't fubar it up the day of.

Step 1) Install Rubberduck.
Step 2) Does anything hereafter actually matter? You've now got #TheDuck.

for a while, Microsoft has been taking a page out of Apple's playbook --- whenever Office365 (or it Windows 10's doing?) crashes, there is no "application has stopped working blah blah" as there used to be. It simply silently restarts, which is kind of annoying. Am I only one who's peeved by that little change?

@MathieuGuindon Is there anything I need to do for my PR that's failing because of ..\InnoSetup\\Graphics\..? I've already taken long enough to get it to this point.

[rubberduck-vba/Rubberduck] Vogel612 created comment on issue #4788: Todo explorer organization at the project level, not stop at module level

> What you mention in your last comment is already covered under #4849

@IvenBach You could fix the installer script under Rubberduck.Deployment/InnoSetup/Rubberduck.Installer.Build.iss to behave correctly

i'm a bit confused, though - why is this erroring? Doule slashes should not cause problems.

AFAICT the definitions for __INCLUDE__, IncludeDir, BuildDir, GraphicsDir and License need to be adjusted to not have a double backslash. Comparing with OutputDirectory suggests that there always used to be a trailing backslash there

@this the file does exist without double slashes, so either this is a permission issue or an IO error

Double duck checking: How in github.com/rubberduck-vba/Rubberduck/blob/next/… it has #define OutputDirectory SourcePath + "Installers\" while for the others it is ` SourcePath + "\FooBar\"` which includes the path separator?

yea, exactly what I mean.

For my education what is ISS? International Space Station doesn't fit.

hmm.

InnoSetupScript

IIRC, IIS has SourcePath passed in

@this nope

[rubberduck-vba/Rubberduck] SmileyFtW created comment on issue #4788: Todo explorer organization at the project level, not stop at module level

> Thanks!

after_test:
  - cmd: iscc /O "Rubberduck.Deployment\InnoSetup\Rubberduck.Installer.Build.iss"

Time to learn that InnoSetupScript exists. :reading:

Ok. I thought I remembered passing in a variable

How does SourcePath work if it's not defined within that file?

I assume it's the location of the script-file?

that's what I'm trying to remember - either it's built-in

yeah it's built-in

> SourcePath str. Points to the directory where the current script is located, or the My Documents directory if the script has not yet been saved.

jrsoftware.org/ispphelp/index.php?topic=predefinedvars

Once more unto the breach research, dear ducks...

anyways, I'm off for a bit :)

you found the official doc first, good job.

it was bothering me that there was __INCLUDE__ and then SourcePath. Seems horribly inconsistent

@IvenBach I'm thinking I should have had used something like this: jrsoftware.org/ispphelp/index.php?topic=predefinedvars

I don't see a equivalent of FSO.BuildPath in the list of functions, unfortunately. That would have been much more apropos.

I don't want to Duga spam the room but if I don't have any issue with building locally but it barfs on AV... Do I just open a PR solely for editing the ISS and test that way?

@this Am I the only one that read that as an arrow?

it'd need dashes, no? --> or ==>

but I can see how flinging a polished turd could cause it to become unpolished once it hits its target.

@this Not always.

Lots of places leave the stem off.

how do you not confuse with the comparison operator?

@this Or once a new feature comes along and gets stuck on it.

@this In contexts where there isn't a comparison operator. Like menus.

I've just gotten so used to it, that I usually see an arrow unless it's in a code block or an IDE now.

oh I see what you mean

those usually have a more obtuse angle, though.

True.

@this That's a vivid visual.

Better that than adding some turds to a pan of brownie. ;-)

o/

> serves brownies, victim eats, announcement: You wouldn't take the point the easy way, so have it your way

[rubberduck-vba/Rubberduck] IvenBach created pull request #5024: Edit ISS build file to merge IvenBach/Rubberduck/ISS_Fix into rubberduck-vba/Rubberduck/next

lol

[rubberduck-vba/Rubberduck] SmileyFtW created comment on issue #4996: Member 'XLApp' has a 'VB_VarHelpID' attribute ?

@Duga I have a feeling about this one. It is mostly good. Still a shred of dread about it though.

lol

@this I've not seen that (yet, but we're being updated to o365 this weekend, so I'm sure I will), but I'm betting I'll be irritated by it when I do!

[rubberduck-vba/Rubberduck] build for commit fabd582a on unknown branch: AppVeyor build failed

BUILD FAILURE!

@Duga σ.σ why you do this

Because she can.

[ISPP] Preprocessing.
[ISPP] (25): Include: c:\program files (x86)\inno setup 5;C:\projects\rubberduck\Rubberduck.Deployment\InnoSetup\Includes\.
[ISPP] (26): SourcePath: C:\projects\rubberduck\Rubberduck.Deployment\InnoSetup\.
[ISPP] (27): BuildDir: C:\projects\rubberduck\Rubberduck.Deploymentbin\.
[ISPP] (28): AppName: Rubberduck.
[ISPP] (29): AddinDLL: Rubberduck.dll.
[ISPP] (30): DllFullPath: C:\projects\rubberduck\Rubberduck.Deploymentbin\Rubberduck.dll.
[ISPP] (31): Tlb32bitFullPath: C:\projects\rubberduck\Rubberduck.Deploymentbin\Rubberduck.x32.tlb.

I fixed the `\\` issue though.

Oh... I guess #define BuildDir ExtractFileDir(ExtractFileDir(SourcePath)) + "\bin\" does need \ before bin. :whoopsie:

Yeah - when AppVersion is missing it's because it didn't find the file

[rubberduck-vba/Rubberduck] IvenBach synchronized pull request #5024: Edit ISS build file

[rubberduck-vba/Rubberduck] build for commit fabd582a on unknown branch: 64.23% (target 0%)

@Duga Round#2 START!

I still don't understand why `\\` is even an issue....

Methinks ExtractFileDir(ExtractFileDir(SourcePath)) doesn't end with one.

I'm sure it doesn't. That's why I earlier suggested using AddBackSlash since they didn't provide a equvialent of BuildPath

that way we don't have to care.

Also, note that the sourcepath has backslash but the original include didn't

Include: c:\program files (x86)\inno setup 5;C:\projects\rubberduck\Rubberduck.Deployment\InnoSetup\Includes\.

You link was the same as Vogels.

?

That link was the same as vogel provided for the documentation.

wtf

no

paste fail

this: jrsoftware.org/ispphelp/index.php?topic=predefinedvars

oh ffs

it's not deeplinking

deeplinking?

look, click on the Functions on the left sidebar

then look toward the bottom of list for AddBackslash function

jrsoftware.org/ispphelp/topic_addbackslash.htm

The URL itself doesn't change but it does redirect the browser.

I copied the link directly from hovering over.

yeah. #FunTimes

#Intuitive, too

#TIL deeplinking

if it was deeplinking, the URL on browser would have changed, and I wouldn't have to tell you where to go to get to the same place I was.

I think something happened to the website - BTW. It used to be working but now it seems to be borked.

it used to work just like those old skool Windows help CHm files

you know, with purple books in the leftsidebar?

mkay.

Hey bruglesco!

#define AddBackslash(str S)
#define BuildDir AddBackSlash(ExtractFileDir(ExtractFileDir(SourcePath))) + "bin\"

@this ^ is what you were suggesting?

uh, you don't need the first line

just the 2nd line (and do the same to others where backslashes matters)

mkay. Now I've understood.

[rubberduck-vba/Rubberduck] IvenBach synchronized pull request #5024: Edit ISS build file

huh? I don't see AddBackslash?

also the IncludePath still has a trailing backslash. I'm thinking it might isn't supposed to have one.

[rubberduck-vba/Rubberduck] IvenBach synchronized pull request #5024: Edit ISS build file

:deep-breath:

I may understand git. I've yet to work with it to not :derp: it up.

better but there's still one trailing backslash: #pragma include __INCLUDE__ + ";" + AddBackSlash(SourcePath) + "Includes\"

I think it should be just #pragma include __INCLUDE__ + ";" + AddBackSlash(SourcePath) + "Includes"

Also, not sure if AddBackslash() is available for use in a #pragma instruction but we'll see.

So it's the blind leading the blind?

Why should the #pragma not include a backslash after Includes?

[rubberduck-vba/Rubberduck] build for commit 0ac6779f on unknown branch: AppVeyor build failed

BUILD FAILURE!

you still need a slash for the bin

[ISPP] (27): BuildDir: C:\projects\rubberduck\Rubberduck.Deploymentbin\.
...
[ISPP] (30): DllFullPath: C:\projects\rubberduck\Rubberduck.Deploymentbin\Rubberduck.dll.

hmm. Interesting.

#define BuildDir AddBackSlash(ExtractFileDir(ExtractFileDir(SourcePath))) + "bin\"

That should have had worked.

oh why didn't you give us a BuildPath, Mr. InnoSetup?

huh, something new: appveyor.com/on-premise

[rubberduck-vba/Rubberduck] IvenBach synchronized pull request #5024: Edit ISS build file

[rubberduck-vba/Rubberduck] codecov[bot] created comment on pull request #5024: Edit ISS build file

[rubberduck-vba/Rubberduck] build for commit 0ac6779f on unknown branch: 64.23% (target 0%)

@Duga I think these are the backlogged commits.

yeah it looks like.

The AV also isn't doing anything

just .... pending.

At the very least I have the correction you pointed out to my GH repo.

[rubberduck-vba/Rubberduck] build for commit befb92c5 on unknown branch: AppVeyor build failed

BUILD FAILURE!

@Duga so yeah, extra backslashes weren't really a problem.

I'm at a loss then.

If anything, you've relieved Mat's OCD about the extra backslash. ;-)

You say Mug has OCD mr #BeltsNSuspenders.

If anything prefer erring on the side of caution and have them.

lol

TBH, I never saw anything bad happen because there was an extra backslash.

Any thought as to what's causing failure?

I did look at the files and they are there

Vogels said could be permission or IO issues

I also wondered if git isn't pulling the files but that seems unlikely if I can see them in GH

[rubberduck-vba/Rubberduck] build for commit b3f824d9 on unknown branch: AppVeyor build failed

BUILD FAILURE!

@Duga The revenge of killer double slashes

That doesn't make sense.

I'm missing the obvious. Time for a walk.

I am 99% sure that slashes aren't the issue

A possible test is to omit all extra paths and supply only one file.

[rubberduck-vba/Rubberduck] build for commit dcd0c48f on unknown branch: AppVeyor build failed

BUILD FAILURE!

WizardSmallImageFile={#GraphicsDir}Rubberduck.Duck.Small.55x55.bmp and no other files?

ditto for WizardImageFile (that's the line 82)

Comments begin with ;?

yeah

such is the glorious working of Pascal.

[rubberduck-vba/Rubberduck] IvenBach synchronized pull request #5024: Edit ISS build file

[rubberduck-vba/Rubberduck] codecov[bot] deleted comment on pull request #5024: Edit ISS build file

[rubberduck-vba/Rubberduck] build for commit dcd0c48f on unknown branch: 64.23% (target 0%)

[rubberduck-vba/Rubberduck] MDoerner created comment on issue #5021: Parse leaves "Ready" state, but empty CE

> The problem with the IndentCommand is in the EvaluateCanExecute method, which is triggered regularly. I think the exception only reports the starting line of the predicate. Probably, the NRE is in project.Declaration.ProjectId. This would imply that tge declaration is null.

ugh

the ref to ado borked :(

looks like im going to have to do late binding/.

stupid access 2007 reference crashed

> git status

returns 13 files. Expected 1. O_O

> git diff file1.bas

[rubberduck-vba/Rubberduck] build for commit 4be44bd9 on unknown branch: AppVeyor build succeeded

discovers he typed Dim Report as Worksheet where he had Dim report as Worksheet elsewhere. VBE "helpfully" updated 12 other modules.

sigh...

</FreeMan>

[rubberduck-vba/Rubberduck] codecov[bot] created comment on pull request #5024: Edit ISS build file

[rubberduck-vba/Rubberduck] build for commit 4be44bd9 on unknown branch: 64.23% (target 0%)

@FreeMan DUDE! i totally have that shirt too

@FreeMan these are the kind of gut checks that only experience teaches.

@IvenBach What did you do before it started to fail? I have to say that the commit history of your PR looks a bit strange.

Which PR? I have that are failing because of this.

The most recent PR 5024 was a git fetch upstream next and git merge upstream/next to have my local copy of next sync'd up.

Then branched off and edits to the file in an attempt to address the `\\` issue.