VBA Rubberducking - 2019-04-02 (page 2 of 2)

« first day (1761 days earlier) ← previous day next day → last day (1419 days later) »

8:09 PM

I've also see where it says it can't build because FooBar.exe isn't found.

you probably need to test this on your home machine

if it builds then you know that your machine really really need to meet Mr. Coffee

Mathieu Guindon

In other news, RD News blog now has as many views in 2019 as it had visitors in 2018

and it's only April

nice!

That means 2018 views are still higher, right?

Mathieu Guindon

user image

@Hosch250 yeah

35,440 views in 2018

8:14 PM

looks like we already beat 2017's visitors

Mathieu Guindon

yup

and are approximately 1/3 into 2018's visitors

Mathieu Guindon

at this rate we'll probably finish 2019 around 50K views... unless the mad exponential growth keeps up, but it looks like it won't

yeah seems a bit more modest

which is fine btw

Mathieu Guindon

...that said I wrote pretty much nothing between Jan and March, other than RD release announcements and maybe a handful of VBA posts

FWIW I edited the "What is Rubberduck" home page recently, and removed the entire "We lose module & member attributes on rewrite" section on the "700 issues?" page

8:19 PM

0

Q: Adding a resource to WPF application causes build error

I'm working through the book Head First C# and consistently have issues when adding resources to a window. This is a 100% repeatable error on any new WPF application I create when adding a new resource. The only way around this is to comment out the resource, build, and uncomment, as detailed in ...

Bring on the answers/solutions

Mathieu Guindon

8:31 PM

@IvenBach fixed the list formatting & xaml syntax highlighting; you should probably specify what version/edition of VS you're using

8:42 PM

.NET Core 3.0 is now scheduled for the second half of this year.

I hope they get to it, because that's what's blocking getting RDs build off of Sunburst

Yep. C# 8 is going out with that, so I have to wait.

@MathieuGuindon Added.

I hope it's not an avalanche of #WorksOnMyMachine leaving me out in the cold.

@Vogel612 i'm a bit disappointed - IIRC, they originally slated it to be in Q1.

yea, that's what I remembered as well

I'm unclear whether 2017 will get it

if it requires 2019... that would be a huge bummer

8:48 PM

@IvenBach WOMM.

(Works On My Machine.)

could be Worms On My Machine, too

Solo devs don't get much experience to acronyms.

user image

That one's not only on it. It's in it too.

hey, now, that's a one happy worm

Because it found your Social, CC, and financial records stored as plaintext.

2

Mathieu Guindon

8:52 PM

@this they were too busy ensuring extensions' menus would all go under an [Extensions] parent-level menu. #Priorities

@MathieuGuindon LOL, different teams.

Mathieu Guindon

meh

No problems on VS 2017 15.9.5. — jsanalytics 2 mins ago

Attempted the steps on VS 2017 15.9.6, no errors — JSteward 15 mins ago

I'm on 2019 latest non-preview.

Mathieu Guindon

and no repro?

Nope.

Mathieu Guindon

8:55 PM

now I wonder what specific version of 2017 I'm running

I have 2017 15.9.10, or something. I can try there too.

Yeah, 15.9.10. 15.9.11 is just a Docker change.

I'm targeting 4.7.2.

FWIW, it would be better to say what you are targeting, @IvenBach, since VS doesn't actually compile, it's MSBuild, and they all reference the installed version.

@IvenBach Stupid idea: have you tried making the class public?

@M.Doerner I didn't have it public, but it's not too bad of an idea.

Just a random thought.

I actually have never done WPF.

Mathieu Guindon

@M.Doerner ha! glad you asked -- it occurred to me as well but I was like "nah can't be that"

:D

should RD warn when a Variant gets assigned a different data type than the previous assignment? something like "The meaning of 'foo' is changing within its declared scope", i.e. "Variable 'foo' is originally assigned a {0:type1}, then subsequently assigned a {1:type2}. Consider introducing a new variable instead."

meh. would only fire for variants, likely assigned to variants

too many false negatives to be really useful

9:13 PM

is Null a different type?

what I really, really want though is a way to say "I want this Variant to be used with only either String or Null but nothing else"....

basically a string?

Mathieu Guindon

what prompted this is the [ab]use of J in this code

nevermind the criminal For l = ... (lowercase L identifier...)

If J Is Null?

that shouldn't work

Mathieu Guindon

> I've tried 'If J is null, if J is nothing, If J is empty" etc

#ReadFail

@Hosch250 Targeting as in what .NET framevork version?

9:21 PM

Yes.

@M.Doerner Explicitly made public. Build error persists.

I'm not actually targeting anything specific. Just trying to get it to build.

@IvenBach Yes you are...

Alt-Enter on the project.

See what framework version you are targeting.

#Words... It's targeting 4.6.1. My guess is that's what was the default when I created the new solution.

@Hosch250 :+1: for hotkey

Cool.

There's still a lot I'm oblivious of that VS does for me.

For my education: changing the framework has what impact on a solution?

I know it can be changed. Why would you want it changed? Or asked another way What is the impetus behind making the change?

9:29 PM

if you're making client applications to be distributed, lower mean it can be installed on more machines

which may not be able to update the .NET framework for whatever reasons

Interesting... as of now the following inspections have been marked experimental:

if you're making web application, you usually want highest, since you have the control of the server and can do what you need so it's less of a dilemma in that scenario.

EmptyCaseBlockInspection
EmptyDoWhileBlockInspection
EmptyElseBlockInspection
EmptyForEachBlockInspection
EmptyForLoopBlockInspection
EmptyIfBlockInspection
EmptyWhileWendBlockInspection

there are no other uses of the experimental attribute

hmmm ...

@this :+1:

hey @this you unbroke experimental types without closing the issue apparently

9:34 PM

I did? I don't remember unbreaking anything....

January 16th 2019, according to the commit

must have been gremlins who stolen my identity. YOu know I break things, not unbreak it.

github.com/rubberduck-vba/Rubberduck/commit/6bef6a09

yep, definitely gremlins.

Not my doing.

whistles

2

[rubberduck-vba/Rubberduck] Vogel612 closed issue #4270: Fix Experimental attribute

[rubberduck-vba/Rubberduck] Vogel612 is playing around with a project card

[rubberduck-vba/Rubberduck] Vogel612 created comment on issue #4270: Fix Experimental attribute

> Fixed as of 6bef6a09ff5707bf1f2decacf762e148daa62c11

[rubberduck-vba/Rubberduck] Vogel612 edited comment on issue #4270: Fix Experimental attribute

> Fixed as of 6bef6a09ff5707bf1f2decacf762e148daa62c11, merged with #4699

Mathieu Guindon

9:37 PM

no experimental feature is listed in the settings dialog though

user image

yea, that seems to be because the registration is incorrect there...

that should be GeneralSettings, right?

Mathieu Guindon

that white text should probably be another color though

oh dear... yea.

Mathieu Guindon

@Vogel612 huh just realized none of these are showing up in inspection settings, or running

so the attribute works (they don't run), but there's no way to enable them

see? I told you I didn't unbreak

9:42 PM

so I "only" need to fix the GeneralSettings then ...

most likely I only fixed the injection thing

yea, the font there is using the HighlightTextBrush...

Mathieu Guindon

thinking experimental inspections are probably better off defaulting to DoNotShow severity, and the [Experimental] attribute simply makes inspection settings warn/confirm when enabling them. thoughts?

alternatively, ditch the whole [Experimental] stuff altogether

i agree with warning.

I wouldn't be upset if we did away with it altogether, though.

maybe we're overcomplicating it -

Mathieu Guindon

^

9:46 PM

why can't they just get a text concatenated (Experimental)

I'm happy to throw them out :)

that is purely an UI thing... right?

[Experimental] was originally intended to deal with more than just Inspections

Mathieu Guindon

cough some refactorings cough

and source control integration

Right - all it needs is a uniform system for adding (Experimental) something

Mathieu Guindon

9:48 PM

and eventually, code panes

yeah

like, the "master switch" for enabling AC features

or rather BlockAC, right?

because SCP-AC is comparatively stable thanks to the extensive test-suite you have

wait, that's something different - you want a single light switch for anything marked [Experimental] as opposed to merely labeling things and leaving it up to users to use it or not?

Hi folks. :)

So... I just did something different than translations :)

Mathieu Guindon

@SonGokussj4 great!

You know what's OCD? :D

Mathieu Guindon

9:56 PM

@this could be used as whatever - if experimental features are off, IMO they should still show up in the inspection settings, just with the dropdown disabled and a tooltip that says something like "This inspection is experimental, and experimental features are currently disabled; enable experimental features to configure this inspection" -- or something a bit less robotic

looks like [Experimental] is currently being wired-up at IoC level -- which is perfectly fine, but prevents that. Hence, I don't know what the best thing to do would be

conditionally wiring-up experimentals with IoC ensures these features don't interfere in any way with the rest of RD

...I think it needs to remain that way

we're explicitly excluding Experimental types at the IoC level to avoid having "consumers" needing to do their own logic to determine experimental status of things

also to avoid the performance issues

esp. w/ inspections. right?

nah, that doesn't really make that much of a difference

Ok. I won't trouble you :)

So... You know that when you browse through RD's Settings, the "header" (screenshot below) is not in the same position (my light OCD)... So I fixed it? Maybe? Hope so?
It's my first PR apart from translations so...
We'll see :-)

user image

@SonGokussj4 oh, that's cool

Mathieu Guindon

10:00 PM

@SonGokussj4 awesome! bring more!!

[rubberduck-vba/Rubberduck] SonGokussj4 created pull request #4890: balanced settings headers to merge SonGokussj4/Rubberduck/next into rubberduck-vba/Rubberduck/next

> Fixed balancing paddings, margins of the "Settings headers" (don't know how to specify it more). Here's a screenshot :)

http://www.noirgoku.mablog.eu/screenshots/2019-04-02_235744.jpg

Hmm, Ok. I thought I remembered a discussion that the IoC injection would help avoid sucking up resources due to having the types active.

user image

@SonGokussj4 Nice! we always need to keep duck looking all spiffy'n'spic!

Thanks!

Mathieu Guindon

@this the SC panel used to have constructor logic that could throw exceptions; by not wiring it up at all we prevented it from sinking RD :)

10:01 PM

I'm not sure if I did it right... :D So please, look at it thoroughly. But thanks :-)

@Duga @SonGokussj4 the negative margin for the inspection settings seems rather weird, honestly

Mathieu Guindon

@Vogel612 :+1: ...wait there's no checkbox? or is it a checked list of experimental features?

@MathieuGuindon IOW it was done wrong. :D

@MathieuGuindon I still need to fix the checkboxes, actually

there is a control for checkboxes, but the backing property isn't correctly filled

Mathieu Guindon

@Vogel612 should they be in their own tab?

10:03 PM

@Vogel612 You mean before the "fix" or after? I didn't see any negative numbers there?

Mathieu Guindon

or hotkeys. hotkeys should be on their own tab

@SonGokussj4 see the InspectionSettings.xaml Line 279

Mathieu Guindon

with hotkeys in their own tab, exp.features won't look so lost

ttqw

I assume you've been using the graphical designer for this?

@MathieuGuindon yea, they should get their own tab, that's true...

makes for a good up-for-grabs issue that @SonGokussj4 could tackle :)

[rubberduck-vba/Rubberduck] MDoerner commented on Rubberduck.Core/UI/Refactorings/ExtractInterface/ExtractInterfacePresenter.cs of pull request #4851: Refactor refactoring setup

10:10 PM

@Vogel612 I See it now. Yeah, that seems... Weir. I was just changing the Margin things. So that went there automa"gically"?

user image

@SonGokussj4 ya, sometimes the WYSIWYG editor for XAMLs (and the one for MSForms as well) does weird things

generally those editors sometimes do things that people looking at the code can tell to be very wrong

stuff like negative margins on a ScrollView

@MathieuGuindon I found out why the experimental inspections are not shown ...

I didn't know you can do that by some other means. Just wanted to try to fix that and this is the first thing that worked :-)

So cancel pull request and try to change it only through code?

OH i Know what did the negative margin now! (but don't know why...)

before, I didn't see all the elements

user image

nah, just add a commit that fixes the change. The PR will automatically update

So i widened it.

user image

[rubberduck-vba/Rubberduck] retailcoder commented on Rubberduck.Core/UI/Refactorings/ExtractInterface/ExtractInterfacePresenter.cs of pull request #4851: Refactor refactoring setup

10:14 PM

And this added the negative margin.

Mathieu Guindon

Yeah xaml work should never be done with the WYSIWYG designer =)

So:
- i delete the margin
- do git add
- do git commit
- do git push mygitlabfork
and the PR will be automatically updated?

@MathieuGuindon So how do you do "graphics"? if you don't see it?

Mathieu Guindon

The designer does update as the markup changes

@SonGokussj4 yup =)

Okay, another question. How did you insert there the "Import, export" text if you don't see it as seen on the first screenshot?

Mathieu Guindon

By typing the xaml markup?

10:17 PM

[rubberduck-vba/Rubberduck] SonGokussj4 synchronized pull request #4890: balanced settings headers

I would have to see that in real time youtube video. :) If you ever, in future, decide to add a new settings menu or modify something, please, screen-record it and show it to me. I need to see how are people doing this stuff. That would be the best tutorial.

@SonGokussj4 yes

[rubberduck-vba/Rubberduck] build for commit 7ffdf400 on unknown branch: AppVeyor build succeeded

[rubberduck-vba/Rubberduck] codecov[bot] created comment on pull request #4890: balanced settings headers

> # [Codecov](https://codecov.io/gh/rubberduck-vba/Rubberduck/pull/4890?src=pr&el=h1) Report
> Merging [#4890](https://codecov.io/gh/rubberduck-vba/Rubberduck/pull/4890?src=pr&el=desc) into [next](https://codecov.io/gh/rubberduck-vba/Rubberduck/commit/5bb8d818e4b885c772428dc6705a5eaf32774d6b?src=pr&el=desc) will **decrease** coverage by `<.01%`.
> The diff coverage is `n/a`.

```diff
@@ Coverage Diff @@
## next #4890 +/- ##
==========================

[rubberduck-vba/Rubberduck] build for commit 7ffdf400 on unknown branch: 64.53% (target 0%)

@SonGokussj4 I should actually stream working on RD sometime...

but I really need to be careful about what I can show there, because Ireally don't want to accidentally get myself demodded

and checking flags is a reflex by now when I have some downtime, which is really useful for flag response times, but really unhelpful for something like streaming coding

also I think I may have shot VS again ...

~sigh

@Vogel612 YES!!!

10:32 PM

lol my twitch channel still has the stream caption "Programming on Rubberduck Release 2.2.0"

that's how long ago my last attempt was.

@Vogel612 Literally don't know what you talking about here.

Ha :-) What's your twitch?

you get a single guess, actually

@SonGokussj4 I have special privileges on CodeReview.SE that come with a certain set of limitations.

namely I'm not to reveal how those privileges work

and I'm not to show what I can see with these

user image

^^ not currently tested whether it's working

[rubberduck-vba/Rubberduck] build for commit 374e872d on unknown branch: AppVeyor build succeeded

[rubberduck-vba/Rubberduck] codecov[bot] deleted comment on pull request #4890: balanced settings headers

> # [Codecov](https://codecov.io/gh/rubberduck-vba/Rubberduck/pull/4890?src=pr&el=h1) Report
> Merging [#4890](https://codecov.io/gh/rubberduck-vba/Rubberduck/pull/4890?src=pr&el=desc) into [next](https://codecov.io/gh/rubberduck-vba/Rubberduck/commit/5bb8d818e4b885c772428dc6705a5eaf32774d6b?src=pr&el=desc) will **decrease** coverage by `<.01%`.
> The diff coverage is `n/a`.

```diff
@@ Coverage Diff @@
## next #4890 +/- ##
==========================

[rubberduck-vba/Rubberduck] build for commit 374e872d on unknown branch: 64.53% (target 0%)

@Vogel612 Were you streaming while coding RD?

10:47 PM

@IvenBach for something like half an hour

I still have OBS installed

Ping me next time you do.

#metoo

You have a 2 duck an audience waiting.

[rubberduck-vba/Rubberduck] Vogel612 created pull request #4891: Fix experimental attribute usage, see #4270 to merge Vogel612/Rubberduck/fix/experimental-attribute-usage into...

rubberduck-vba/Rubberduck/next

> While #4270 fixed the internal workings of the Experimental attribute, its usage in the Rubberduck codebase was incorrect, resulting in a failure to properly enable the feature switch for types marked as experimental without an associated resource key.

mumble mumble flakey internet mumble 128p webcam mumble mumble copyright mumble

it's weird, currently German authorities are kinda cracking down on streamers alleging that they are something very similar to radio

10:54 PM

Very similar to radio? They want people to pay for watching youtube/twitch streams?????

nah

they want the streamers to pay for a "Rundfunklizenz"

can't even read that.

and I¨m like 300 km from German...

basically if you operate a radio station you need a license to do so

that's a "Rundfunklizenz"

And the "Rundfunkstaatsvertrag" is the law that regulates who needs one

Ah okay. So... They want to put a tax on stream? I mean... Lincence = money to state?

Mathieu Guindon

@Vogel612 wtf?!

10:57 PM

The problem with making any sort of general statement there is that every state (Bundesland) in Germany can make their own rules about that

Mathieu Guindon

So you put up a vid on YT, you're good. Stream live it instead, and you need a license?

but generally speaking streamers are required to tell the relevant authorities that they are a "Rundfunkprogramm" under the criteria of the law. Depending on the size of their audience they may be required to pay a one-time-fee to obtain a license

Streamers fall under that law because professional (and semi-professional) streamers have a schedule and they comment on what they are doing (which counts as a type pf journalistic work)

Mathieu Guindon

I want a "Round Funk Program" t-shirt

:D :D :D

@MathieuGuindon "Rundfunk" = "radio&|tv"

as opposed to "Richtfunk"

Mathieu Guindon

11:00 PM

funky

Rundfunk: -> Send radio signal from antenna in every direction (rundherum)
Richtfunk: -> Send radio signal from antenna in a narrowly defined direction (richtung)

@Vogel612 confirmed working

Mathieu Guindon

Woot!

PR is already open.

nothing really much to do, thanks to this doing the legwork

also I gave up writing docs halfway

Mathieu Guindon

We need a build warning for types with an experimental attribute, lest we forget they're there (like these inspections)

I think

nah. I removed the no-arg ctor for the ExperimentalAttribute

Mathieu Guindon

11:04 PM

how does that remind us what types have the attribute?

it doesn't, but it makes sure we have them display in the GeneralSettingsView

Mathieu Guindon

Wait I'm missing something... do they... get individually listed in the settings?

Without a resource key [Experimental] is basically equivalent to [Disabled]

@MathieuGuindon no. Their resource key is displayed and used for determining their exclusion status

Mathieu Guindon

get off your phone, mug, and just review the darn PR

[Experimental("EmtpyBlockInspections")] -> Belongs to EmptyBlockInspections group

11:09 PM

@Vogel612 wut...

[rubberduck-vba/Rubberduck] build for commit 2bf697c0 on unknown branch: AppVeyor build succeeded

[rubberduck-vba/Rubberduck] codecov[bot] created comment on pull request #4891: Fix experimental attribute usage, see #4270

> # [Codecov](https://codecov.io/gh/rubberduck-vba/Rubberduck/pull/4891?src=pr&el=h1) Report
> Merging [#4891](https://codecov.io/gh/rubberduck-vba/Rubberduck/pull/4891?src=pr&el=desc) into [next](https://codecov.io/gh/rubberduck-vba/Rubberduck/commit/5bb8d818e4b885c772428dc6705a5eaf32774d6b?src=pr&el=desc) will **decrease** coverage by `<.01%`.
> The diff coverage is `50%`.

```diff
@@ Coverage Diff @@
## next #4891 +/- ##
==========================

[rubberduck-vba/Rubberduck] build for commit 2bf697c0 on unknown branch: 64.53% (target 0%)

Mathieu Guindon

11:29 PM

This is why I couldn’t be happier that nearly none of my code survives in @rubberduckvba. It evolved beyond those early iterations. https://twitter.com/mfeathers/status/1113181804188454912

Tweeted by Rubberduck203 on April 2, 2019 at 10:52 PM

That isn't the mindset of all coders?

> He be my 30+ year old baby troglodite hack. Still there and ain't going anywhere.

ya, some people can be possessive about their code

Mathieu Guindon

Makes me wonder how much of his code is actually still running

AFAICT: none

Mathieu Guindon

^

11:44 PM

the last few lines that still were around when I last checked about a year or even longer ago should be gone by now

Tweet him "Correction: None of your code remains. Congratulations on writing self documenting replaceable code."

yea... 500 LoC left in July 2018

@IvenBach you had 4.7kLoC at time of checking those stats ;)

No way. I have only had a few small commits.

Link?

Why would it be gone now? That was after we removed SC which was the majority of his work, IIRC?

@Rubberduck203 @rubberduckvba soo .. out of 183kLoC in *.cs files in current next, you are the last one that modified 500. First and Scone place are pretty close to one another, but the multiple names make it a bit harder to see that @retailcoder is currently "in the lead"

Tweeted by vogel612 on July 9, 2018 at 10:33 AM in reply to Rubberduck203

« first day (1761 days earlier) ← previous day next day → last day (1419 days later) »

Transcript for

Apr '192

VBA Rubberducking

This chat has moved to Discord: discord.gg/MYX9RECenJ

antlr com-interop parsing rubberduck vb6 vba whiskey-tango-foxtrot

join this room
about this room

00:00

06:00

12:00

18:00

all times are UTC

site design / logo © 2024 Stack Exchange Inc; legal

mobile