RELOAD!

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

[Minesweeper] Games Played: 114, Bombs Used: 62, Moves Performed: 16096, New Users: 9

just making sure - if I do ` var dx = decimal.Parse(x, CultureInfo.InvariantCulture);` this will still work on non-English builds, right? (x would be "1.0", rather than "1,0")

That is my understanding of invariant culture.

thanks; writing unit tests and wnat to make sure I wasn't going to introduce problems

[rubberduck-vba/Rubberduck] bclothier created pull request #5588: Fix PermissiveAssertClass to merge bclothier/Rubberduck/variant into rubberduck-vba/Rubberduck/next

> The PermissiveAssert class is broken. This PR makes it acts using the same semantics as used by VBA by using the VarCmp function to do the comparison. In testing, I found it was necessary to also change the type to handle some comparisons like True = "True". This also refactors the Variants namespace a little bit.

[rubberduck-vba/Rubberduck] MDoerner created pull request #5589: Allow line continuations in precompiler directives to merge MDoerner/Rubberduck/LineContinuationsInPrecompilerDirectives into...

rubberduck-vba/Rubberduck/next

[rubberduck-vba/Rubberduck] bclothier synchronized pull request #5239: Initial version using newest SDK supporting WinForms & WPF

[rubberduck-vba/Rubberduck] build for commit e71328c2 on unknown branch: AppVeyor build succeeded

[rubberduck-vba/Rubberduck] build for commit 3c70bbb6 on unknown branch: AppVeyor build succeeded

[rubberduck-vba/Rubberduck] build for commit 71a1e7f9 on unknown branch: AppVeyor build succeeded

@Duga only took us 11 months....

[rubberduck-vba/Rubberduck] bclothier synchronized pull request #5586: Update InnoSetup installer

[rubberduck-vba/Rubberduck] build for commit f3118f08 on unknown branch: AppVeyor build failed

BUILD FAILURE!

[rubberduck-vba/Rubberduck] bclothier synchronized pull request #5586: Update InnoSetup installer

[rubberduck-vba/Rubberduck] build for commit ad0189c8 on unknown branch: AppVeyor build failed

BUILD FAILURE!

So Hacktoberfest is apparently opt-in now

For the repository maintainers that is.

Because there's been too many spam pull requests

Yeah, saw that and added the "topic", we're in =)

ah, okay. Wasn't able to check from mobile :D

(or rather, wasn't inclined to)

If things go south we can always pull out.

Oh this is so bloody awesome - I got OnKeyPress bindings to work, and I implemented a DecimalKeyValidator and now I have a textbox that only accepts numbers and one decimal separator character - and cancels other keypresses.

And stringformatting works, so I enter "5.5" in that box and tab out and poof it says "$5.50"

Or bind to a Date property and tab out and you get "Saturday, October 3 2020" when you typed 2020-10-03

This feels like ...magic

[rubberduck-vba/Rubberduck] KonH starred us

[rubberduck-vba/Rubberduck] KonH forked us into KonH/Rubberduck

[rubberduck-vba/Rubberduck] KonH created pull request #5590: Add additional checks to ensure that selection is valid to merge KonH/Rubberduck/add_remove_references_window_safety into...

rubberduck-vba/Rubberduck/next

> Related to github.com/rubberduck-vba/Rubberduck/issues/5553

[rubberduck-vba/Rubberduck] KonH created comment on issue #5553: Null Refrence Exception when tabbing in Add/Remove Refrences and related issues.

> @retailcoder Hello! I've added a proposed solution, but unfortunately, I don't have a Windows machine right now and it will be difficult to implement additional unit tests, covers that. I hope it fits your expectation at least initially.

@MathieuGuindon these really make me miss the free time I had for RD. That article is still in my ToRead pile.

Hope we don’t get any fly by night “Acpt dis PR. I wnt free shrt” Opened for Hacktobert.

we can always reject

or as Mat said, opt out

mobile.twitter.com/shitoberfest/status/1311646233128181760

A lot of those makes the consider myself a decent person.

@MathieuGuindon I got more than enough free time to deal with obvious spam PRs that come in

Cool! I don't think we'll get much spam, we're big but not so much as to attract that kind of crap. I think/hope.

Having learned enough C# and coding principles to submit valid a PR and correct it based on feedback.

@IvenBach lol did you ever doubt that?!

You attracted me to this pond #GrainOfSalt

[rubberduck-vba/Rubberduck] build for commit 3079f21d on unknown branch: AppVeyor build succeeded

@MathieuGuindon not really. Especially after seeing how people act towards one another the last several years.

Self deprecating humor is just ingrained really deep.

I think I got binding update triggers and data validation solidly done now, just needs a bunch of unit tests because I suck at TDD. Then I need to fix what I broke with the ErrorFormatter logic (more tests), and then I think VBA pretty much has a MVVM framework.

[rubberduck-vba/Rubberduck] bclothier synchronized pull request #5586: Update InnoSetup installer

@MathieuGuindon make me really want to make something that can take VBA code and make a DLL out of it, for host-agnosticity

Esp. considering that you might not necessarily want to step into the MVVM code

#TomorrowsProblems

@MathieuGuindon if i was that kind of shitty person, I'd probably want to go after a smallish repo. The bigger/more popular repo is, the more formalized process they have in which would make it very difficult for me to fly in some one-character PR

true

(on all counts)

ugh error formatting is annoying, even with MVVM. I wonder how hard it would be to implement some ValidationErrorAdorner that would dynamically add the required controls instead of requiring them at design-time...

[rubberduck-vba/Rubberduck] bclothier removed label Status: Blocked by other issue from pull request #5239: Initial version using newest SDK supporting WinForms & WPF

[rubberduck-vba/Rubberduck] build for commit 20f00f90 on unknown branch: AppVeyor build failed

BUILD FAILURE!

[rubberduck-vba/Rubberduck] bclothier synchronized pull request #5586: Update InnoSetup installer

[rubberduck-vba/Rubberduck] build for commit 6e6539b2 on unknown branch: AppVeyor build failed

BUILD FAILURE!

Got an email from Christian Buse raising flags about interfaces in VBA apparently contributing to corruption of the VBA project

I really wonder how a VBA project ever gets that large

without tons of redundant code, that is

oof

FWIW I'm using a fair number of interfaces myself, and I haven't experienced those specific symptoms of "never compiled", but I have seen some strange crashes...

@MathieuGuindon 40kLoC is not that much TBH

FWIW, in Access (and this might be Access-specific limitation), a VBA project can only load 1000 modules (including any document modules).

The worst thing about that is that it's not a hard limit; you can have more than 1,000 modules but then you get weird problems. In that project I worked on where they had more than 1000, code would start to fail to find the method's definition

considering the purpose that VBA was intended to serve, that's theoretically more than enough...

@Vogel612 for a VBA project, it's quite huge

@this I think the real reason is the debugger's edit-and-continue feature rewriting chunks of p-code on the fly, "randomly" corrupting the storage stream - perhaps Implements statements exacerbate the problem or somehow make it more apparent, but it is already known & established that edit-and-continue does corrupt breakpoints.

assuming no document module Implements anything

well, i have a number of access forms implementing interfaces

I put up an "answer" with my thoughts

@this you know what, I'm thinking the "don't make document modules implement interfaces" recommendation might just be Excel-specific (I don't VBA much in other hosts)

[rubberduck-vba/Rubberduck] retailcoder merged pull request #5590: Add additional checks to ensure that selection is valid from KonH/Rubberduck/add_remove_references_window_safety into...

rubberduck-vba/Rubberduck/next

[rubberduck-vba/Rubberduck] KonH pushed commit 3079f21d to next: Add additional checks to ensure that selection is valid

[rubberduck-vba/Rubberduck] web-flow pushed commit 4797188e to next

[rubberduck-vba/Rubberduck] retailcoder commented on Rubberduck.VBEEditor/Variants/VariantComparer.cs of pull request #5588: Fix PermissiveAssertClass

[rubberduck-vba/Rubberduck] rubberduck-vba-releasebot created draft prerelease Prerelease-v2.5.1.5580

[rubberduck-vba/Rubberduck] rubberduck-vba-releasebot prereleased draft prerelease Prerelease-v2.5.1.5580

[rubberduck-vba/Rubberduck] rubberduck-vba-releasebot published prerelease Prerelease-v2.5.1.5580

[rubberduck-vba/Rubberduck] rubberduck-vba-releasebot created tag Prerelease-v2.5.1.5580

[rubberduck-vba/Rubberduck] build for commit 4797188e on next: AppVeyor build succeeded

[rubberduck-vba/Rubberduck] retailcoder closed issue #5294: #If...Then...#Else directive not recognized

[rubberduck-vba/Rubberduck] retailcoder merged pull request #5589: Allow line continuations in precompiler directives from MDoerner/Rubberduck/LineContinuationsInPrecompilerDirectives into...

rubberduck-vba/Rubberduck/next

[rubberduck-vba/Rubberduck] MDoerner pushed commit 3c70bbb6 to next: Allow line continuations in precompiler directives

[rubberduck-vba/Rubberduck] web-flow pushed commit 9b85c6f7 to next

[rubberduck-vba/Rubberduck] rubberduck-vba-releasebot prereleased draft prerelease Prerelease-v2.5.1.5581

[rubberduck-vba/Rubberduck] rubberduck-vba-releasebot created draft prerelease Prerelease-v2.5.1.5581

[rubberduck-vba/Rubberduck] rubberduck-vba-releasebot published prerelease Prerelease-v2.5.1.5581

[rubberduck-vba/Rubberduck] rubberduck-vba-releasebot created tag Prerelease-v2.5.1.5581

[rubberduck-vba/Rubberduck] build for commit 9b85c6f7 on next: AppVeyor build succeeded

@M.Doerner mind if I merge Imh0t3b's translation PR first (you said there'd be merge conflicts with yours)?

@Duga @MathieuGuindon that requires the assert to know what module it is in when executing and pass it to the comparer. who should tell assert class that?

looks like a TestMethod can be extended to store the needed info, but would the assert neec to read it via a global object??

Hm but the test engine already knows what module it's running tests for no?

Oh shoot that's right, the Assert call is out of that context

We could make IAssert stateful (pukes a little)

uhh... yea no...

lol

that's an especially terrible idea in conjunction with randomized test runs

yeah

I'd say it's better to document the behaviour and leave overriding it to the user using stuff like Assert.AssertTrue expected = actual

@MathieuGuindon If @Vogel612 is OK with the current state of the translation PR, you can go ahead and merge it. Anyway, merging my super type PR is not really urgent.

@Vogel612 does that work even if the result is 1, not true?

@this hell if I know?

@M.Doerner @MathieuGuindon I left some PR comments and TBH at least some of them are blockers

none of them have been really addressed, so I'm against merging the PR pending the fixes I outlined

@Vogel612 ok. my worry is because VBA is really loosey goosey that it is easy to accidentally do an implict conversion and get a numerical result instead of a boolean

it's mostly minor stuff, following from not cleaning up the grammar around the wording change

@this that honestly sounds entirely plausible, but I'm not even remotely up to speed on VBA anymore

anyway, i wouldn’t make assert stateful. if we are doing it, probably need a static global object that tje comparer can use to set its defaults

similar akin to cultureinfo but for vba, i guess. i would rather do this as a separate pr, tbh

making Assert stateful is a terrible idea ...

=)

VBA 31 char limit for class names makes following best guidelines difficult. Sufffixing with Collection removes 10 chars.

Disemvoweling because of limitations Collection->Clxn

[rubberduck-vba/Rubberduck] bclothier commented on Rubberduck.VBEEditor/Variants/VariantComparer.cs of pull request #5588: Fix PermissiveAssertClass

[rubberduck-vba/Rubberduck] bclothier synchronized pull request #5586: Update InnoSetup installer

[rubberduck-vba/Rubberduck] build for commit 26876082 on unknown branch: AppVeyor build failed

BUILD FAILURE!