RELOAD!

[rubberduck-vba/Rubberduck] 2 issue comments.

[Minesweeper] Games Played: 122, Bombs Used: 67, Moves Performed: 17233, New Users: 13

@MathieuGuindon In other words, if I want to be paid from something, don't OSS it.

^

then you can sell it on some market

like that OneNote app

I'd make one, but I've no idea for an app. Like, I'd love to code for my Android phone, but... I don't know what to build, so.. pffft.

I've got a few ideas. Just too tired most of the time.

@MathieuGuindon Are you still here?

Can you please tell me what the current startup project for RD is?

I've got it loading under .NET Core 3.0. I'm going to try to run it.

@Hosch250 Rubberduck.Deployment

Thanks.

sorry, was in another tab

Hitting some errors when I try to run the tests.

hm

Build errors.

Not test errors.

probably related to the sunburst wpf stuff

I took that out.

ATM, I'm thinking it's the WinForms workarounds, maybe?

hm

It looks like it's in one of the lowest-level projects and just cascading up.

But I can't tell for sure.

Second thoughts :)

It's the app.config file.

<?xml version="1.0" encoding="utf-8"?>
<configuration>
  <runtime>
    <assemblyBinding xmlns="urn:schemas-microsoft-com:asm.v1">
      <dependentAssembly>
        <assemblyIdentity name="Castle.Core" publicKeyToken="407dd0808d44fbdc" culture="neutral"/>
        <bindingRedirect oldVersion="0.0.0.0-4.1.1.0" newVersion="4.1.1.0"/>
      </dependentAssembly>
    </assemblyBinding>
  </runtime>
<startup><supportedRuntime version="v4.0" sku=".NETFramework,Version=v4.6" /></startup></configuration>

haha

I'm just going to delete this. I'm pretty sure we don't need it.

The error is a little confusing.

C:\Program Files\dotnet\sdk\3.0.100\Sdks\SMicrosoft.NET.Sdk\Sdk not found. Check that a recent enough .NET Core SDK is installed and/or increase the version specified in global.json.	Rubberduck.Resources

Oh, and we don't need the System.ValueTuple reference anymore.

:+1:

Hmmm. Not quite fixed

Getting close.

I had the SDK name wrong. "SMicrosoft"

Holy carp.

I just almost pushed directly to Next (unless it's Master!).

I hit Cancel just in time.

lol I don't think you could

;-)

I did last year :D

I can merge PRs.

I think I have permissions, LOL.

I'm setting it to my remote.

Also, FWIW, I might be jumping the gun on this a bit, LOL.

no problem!

Some of our referenced projects are still targeting .NET 4.7.2.

That might not work out.

hmm

[rubberduck-vba/Rubberduck] Hosch250 forked us into Hosch250/Rubberduck

Just nuked my fork and re-forked to make sure I have the latest.

probably a good idea lol

OK, now what...

There are no build errors, but there are?

Nothing listed in the output

Nothing listed in the error view.

I can't run the tests--there were build errors pops up.

fun!

The output does say they failed.

But it doesn't give me any error.

somewhere in Rubberduck.Deployment there's an answer

we're kind of hijacking the build process

Is that one referenced by the tests?

no

try right-click /build on RD.Deployment

Then it's not the problem.

I'm trying to run the tests.

Eh, I think I see it.

1>MSBUILD : error : CA0055 : Could not identify platform for 'C:\Users\hosch\source\repos\Rubberduck\RubberduckCodeAnalysis\bin\RubberduckCodeAnalysis.dll'.
1>MSBUILD : error : CA0052 : No targets were selected.

There we go.

ah, so there was a build output :)

Yeah, when I just build one.

interesting

make surevyou build the rubberduckmeta.sln

OK.

that may need adjusting too

like "v" -> " " here or there ;-)

the projects in that solution are built as a part of the rubberduck.sln build

which could be why you get the weird output because meta is failing to build

Yeah, found some issues.

Same errors now, though.

It looks like these errors typically happen when you don't have a "default" target set in the csproj?

Like Debug|x86

need to either restart vs or kill msbuild when u adjust meta then switch to rubberduck.sln

hmmm should have targets iirc

@Hosch250 s/b "any cpu"

Yeah, that's what I'm doing.

maybe sunburst did it weirdly

also, u prolly need to remove workaround sections from core csproj

btw aren’t targets defined in the base csproj?

meta also has its own base csproj as well

github.com/rubberduck-vba/Rubberduck/blob/next/… line 10-11 defines targets

if targets are missing, may need the import

OK, I found more of it.

The Microsoft.Build.* libraries are now nuget packages for Core.

I'm having a bit of trouble finding AppDomainIsolatedTask.

build?

It should be in Microsoft.Build.Utilities.Core.

hmmm why?

only the deployment project needs it

Yeah, it can't find it.

Not sure why.

sorry, no, deployment.build

in meta solution

Yes, that's where I am.

It's not available in core, according to the docs.

docs.microsoft.com/en-us/dotnet/api/…

Try to pick Core 3.0 preview 9.

hm.

It says it's not available and redirects.

not at my comp atm

I just did Task for now.

> Provides the same functionality as Task, but also derives from MarshalByRefObject so that it can be instantiated in its own app domain.

we can try

ITypeLibExporterNotifySink doesn't seem to be available either.

can’t remember why i used app isolated task

hmph.

wait....

can core project consume fx task?

e.g. leave meta as fx, main sln as core?

No.

bummer

Actually, it might be able to at 3.0.

It couldn't before.

as a workaround u can comment this section out

It's basically my bedtime. I'll work on this again later, I guess.

it’s only used when u have midl

ok

OK.

LOL:

> Sure must be nice being Canadian because whatever you do, especially if it's plum stupid, people will just assume you're American.

gfycat.com/inferioryellowamericancreamdraft

lol

[zspitz/ANTLR4ParseTreeVisualizer] zspitz created draft release 0.1.0.28

[zspitz/ANTLR4ParseTreeVisualizer] zspitz published release 0.1.0.28

[zspitz/ANTLR4ParseTreeVisualizer] zspitz released release 0.1.0.28

Anyone know how to do this?

@MathieuGuindon (in response to this)

Nice!

[zspitz/ANTLR4ParseTreeVisualizer] zspitz created draft release 0.1.0.29

[zspitz/ANTLR4ParseTreeVisualizer] zspitz published release 0.1.0.29

[zspitz/ANTLR4ParseTreeVisualizer] zspitz released release 0.1.0.29

morning all.

on a totally unrelated note, I just noticed this on the GitHub page:

> Note: Rubberduck is not a JetBrains product. JetBrains does not contribute and is not affiliated to the Rubberduck project in any way.

Is that officially approved JB wording? If not, they do contribute by giving free software...

:+1:

@FreeMan that's pretty standard wording to avoid getting into muddy waters around stealing endorsement

OK. Just the random thoughts of a brain not quite fully stimulated by the morning dose of caffeine.

@Hosch250 re that OSS and getting paid thing: it's a damn shame that "just do OSS on the side" is good advice nowadays still.

because that means OSS is mainly done by students who got too much time and most likely white dudes that earn enough to not have to worry about money

Agree.

And the other thing is, I don't even agree with their argument.

Especially with the highly non-technical userbase nowadays.

If I built some productivity app that MS didn't already beat me to, I could easily open-source it and sell it on the Windows/Apple/Google stores.

Devs could play with it all they want (and nobody would, unless they wanted to modify it), but most people would just go through the store because it's easier.

Kind of like, say, ANTLR.

Anyone could modify it--but who's going to when it already just works?

Or, say, for example, MS's markdown parser. I found a bug with parsing table cells that contained escaped | tokens. I PRed a fix that was accepted, because it was OSS, and easier for me to just PR it than to try to maintain my own version.

Of course, you get into gray area with that kind of thing with paid software, because people would make money off your changes.

You could, for example, have a policy of sending swag out to contributors (stickers, shirts, etc.) for changes.

[zspitz/ANTLR4ParseTreeVisualizer] zspitz created draft release 0.1.0.30

[zspitz/ANTLR4ParseTreeVisualizer] zspitz released release 0.1.0.30

[zspitz/ANTLR4ParseTreeVisualizer] zspitz published release 0.1.0.30

in SQL, Format('', '000-000-0000') --> --. How do I get it to give me a blank string? i.e. I want a "properly" (for those asking for the data) formatted string or completely blank

[zspitz/ANTLR4ParseTreeVisualizer] zspitz edited release 0.1.0.30

normally you simply don't format them. One possible way is FORMAT(NULLIF(@foo, ''), '000-000-0000')

Nothing about this project is "normal" they want them formatted with dashes or totally blank. I'll give that a shot.

eh... that gives me NULL, they don't want that, then want blank.

take 4...

You really shouldn't do that in SQL, I think....

that's a presentation concern

I'm extracting the data to a CSV to transfer to a 3rd party. Seems they're kinda finicky.

if you have SQL spitting out ''s for "no or invalid data", how do you know it's really an invalid data vs. legitimately a '' (e.g. missing data for example)

I'm not trying to store it in the table that way

within SQL?

Anyway, if you must do it all in SQL, I guess ISNULL(FORMAT(NULLIF(...)...), '') will do the job.

SELECT From SQL table to CSV for transfer via FTP to external party.

well, who's doing the actual writing of CSV? Access? SSIS?

I'm writing the query, we've got another guy (who I haven't met) who promises to schedule it and export the results to the actual CSV then FTP it. He's already set up the process for this vendor, I just need to give him the query to run

well, there's your problem: COALESCE(DayPhone, NightPhone, CellPhone) does not return NULL because I'm not getting NULL, I'm getting blanks from my XML->table load process

#BangsHeadInANotMetalWay

if you bang your head on a metal desk, is it metal again?

prolly

[zspitz/ANTLR4ParseTreeVisualizer] zspitz created draft release 0.1.0.31

[zspitz/ANTLR4ParseTreeVisualizer] zspitz released release 0.1.0.31

[zspitz/ANTLR4ParseTreeVisualizer] zspitz published release 0.1.0.31

im doing it!

im making an internal C# application

from a powershell script i wrote...

[zspitz/ANTLR4ParseTreeVisualizer] zspitz created draft release 0.1.0.32

[zspitz/ANTLR4ParseTreeVisualizer] zspitz published release 0.1.0.32

[zspitz/ANTLR4ParseTreeVisualizer] zspitz released release 0.1.0.32

noice, @KySoto!!

[zspitz/ANTLR4ParseTreeVisualizer] zspitz edited release 0.1.0.32

annnd im already screwing up in the syntax

lol

looks like var is my savior for getting a variable length array

I don't think arrays can be variable length?

You want a System.Collections.Generic.List<T> for that.

Pretty sure that C# doesn't have variable-length arrays. If you extend the array, it makes a new one behind the scenes.

The make-field-readonly thing I did just caught a field I forgot to make readonly :)

Probably the first time :)

First time for a field I just implemented, I mean. I normally remember to do that.

:D

ok, so you guys are my second hope

trying to reference System.Data.SqlClient for .net 3.5

canna find it

some say nuget package, but thats for .net 4.7

it just works in vb.net

many WTF's

@KySoto You probably need to import an earlier version of the nuget package.