« first day (1685 days earlier)      last day (1495 days later) » 

7:14 PM
I should probably keep the border around the toggle buttons on the inspection result toolbar when they're checked.
 
7:42 PM
FB:
> I just had the chance to use the Reference Explorer and I AM SO HAPPY. Before it is always a pain to search for the specific reference that I need. But now, because of Rubberduck it is so easy. They continue to redefine the way anyone works with VBA IDE. KUDOS to all the Rubberduck team!
3
 
> Note that this also fails to resolve:

```
Debug.Print [A1]
```

Don't you love dem globally scoped objects wrecking the show?
> Note that this also fails to resolve:

```
Debug.Print [A1]
```

Don't you love dem globally scoped objects wrecking the show?

(to be fair - it at least says it's a `runtime expression` which is better than the example above, but not that much better -- it should resolve into `Excel.Global.Evaluate` or whatever)
 
@Duga "runtime expression" is totally by design
how do we know what the Evaluate method is?
 
Yet it doesn't do that for expression Application.[A1]
well, we need to cache the evaluate member in the ComMember / ComClass
 
Because it thinks it's a bracketed name
Range.[_Default]
 
then resolver can query for either the parent member being accessed to ask if it has an evaluate member, or if there's no member access, query for globally scoped objects and asking if they have any evaluate member.
 
7:56 PM
I think the only way to differentiate is to check whether or not the member is on the interface. If it isn't, then check to see if there's an evaluate member.
 
would be nice to be able to turn [A1] into Range("A1") somehow
 
that should be possible. We can find the evaluate member now
 
It doesn't help that the spec kind of glosses over the semantics of these. I get the impression that someone was like, "crap - how do we handle foreign identifiers?"
 
Blame @spolsky, I guess
 
he did write Wasabi, after all...
 
7:57 PM
and his peer reviewer, some "Gates" guy
 
probably not important.
I don't think he even knew about internet.
 
haha
 
That must be why it was called Internet Explorer. "Huh, what's this? Think I'll look around".
 
Apple saw that and they put on a Safari hat
 
8:00 PM
chasing a fiery fox
 
Netscape had it down. They were navigating the thing.
 
and escaped into the nothingness
@Comintern that was the "wheel" icon! damn... memories...
 
Fell off the edge of the world.
 
and thus the opera sung a mournful chorus.
 
8:03 PM
@this I always though it was just a normal everyday fox, that caught fire looking at IE
 
lol
 
Who knows. Nobody know how the fox came to be on fire but damn, you can't have it running amok, setting fire to the entire jungle!
 
I bet it was after it jumped over some lazy dog
 
lazy brown dog.
 
I was thinking it was defined in fox.js and was run in IE version 8 or something.
 
20 more to go.
 
@Comintern Did your CE PR add any new resource strings?
 
"Treat your code like poetry and take it to the edge of the bare minimum." - ILYO
 
Just a couple - mostly single words if I remember correctly.
 
> I will uninstall other add-ins then test RD and test it on a blank database. Thanks for the input.
 
8:15 PM
Was Dockable a button that got removed? ref 2472
 
Hm, then I have to do German translations again.
I think that was a custom button to demonstrate the issue.
 
^
@M.Doerner I'd hold off until the Inspection Results PR merges. That has a couple more strings.
 
maybe I should try reading the entire issue....
 
Do we want to merge that before the release?
Any further feature will probably delay the release further or introduce bugs that need a hotfix.
 
I don't see why not - nothing really major changed - just how the inspection results are bound and some UI tweaks.
 
8:20 PM
I'd feel better if we had few days of meatbag testings before we release 2.4
 
TBH, the CE shakedown is probably more necessary than the IR.
 
We have already pushed it yet another week.
 
Bonsoir kiddos. When a new standard module is created in Excel, it automatically inserts "Option Explicit". In Access, it also inserts "Option Compare Database". For the latter, does anyone know how it could be changed to "Option Compare Binary" (my prferences) instead..? I can't find anything about it, and no settings for it.
 
just to keep it in POV -
 
I should have kicked AV over lunch.
 
8:22 PM
we have several other PRs that are not really related to the changes in the CE/IR.
I'm inclined to think that CE/IR should be its version, while those other PRs go into version after
 
The other PRs are intentionally on hold for after the release.
 
Exactly.
The question is more - do we want to have IR go into 2.5 with the other PRs?
@spinjector why would you want to do that?
 
IR is mainly bug-fix ATM. I was going to hold off on the stuff I grabbed today until after release.
 
If you do that, your VBA code and your Access queries will start returning different results.
as a default, it's sensible; I can see a need for some specialized module using binary comparison but that would be exception, not the norm.
 
If the PR is supposed to be part of the release, it will have to be merged very soon.
 
8:26 PM
I can't find any documentation as to what "Database" defaults to, nor if it changes between versions or anything like that. When doing text processing, I prefer exact Chr$() binary values instead of some hazy reliance on the code page.
 
ok, if IR is really just bug fixes, I think we should try to merge IR today.
 
The only thing preventing an IR merge is the AV build thing, but I think I just need to sync anything with it now that the base is in next.
 
@spinjector it depends on the Access database's collating order.
 
I would like to avoid to merge some last minute PR that introduces some annoying bugs again.
 
e.g. File -> Options -> Current Database
 
8:27 PM
@thi
@this yea thats first place i looked.
 
@Comintern Will you rebase it on next?
 
sorry, that's actually General
 
@M.Doerner I was going to push some xml comments or something first, and if that doesn't work then I'll rebase it.
 
but it only shows default sort order for a new file
you'd have to use VBA to see the current file's sort order.
Anyway, Option Compare Database will follow the same rules that the database's sort order specify.
 
@Comintern AV is acting up?
 
8:30 PM
Rebasing would lead to a cleaner history.
 
So it's independent of your Windows locale, unlike other VBA projects.
 
@MathieuGuindon Possibly. The build failure output was giving line numbers that didn't make sense.
 
kicked it
 
@this
 
although, from AV it looks like a build error
like, a missing using System.Linq; in CodeInspectionResultsViewModel
 
8:33 PM
@this ok i'll have to just do it manually. i thought there might be a setting or a secret "new module" template somewhere that i hadn't heard of
 
Conceivably, you could write a user-defined template and have rD inject it. The trouble is that it's not fully implemented yet.
 
[rubberduck-vba/Rubberduck] build for commit c9cd8a77 on unknown branch: AppVeyor build failed
BUILD FAILURE!
 
@this ok but i'll wait for further implementation. i have bad luck with betas. lol
 
@Duga wtf, now it's another error, at another nonsensical line
ooh, it's also in a nonsensical "project"
 
I think it's actually the same errors
though I don't remember an error for line 500.
(but may have missed it)
@Comintern suspected it was a problem with the generated code-behind for the XAML file. The error say it's the VM but it could very well be one of the .g.cs behind the view.
 
8:46 PM
no the one before was about ICollectionView.Where, now it's ICollectionView.Remove
 
so AV now has randomizeError() function?
the Where error is still there.
UI\Inspections\InspectionResultsViewModel.cs(495,40): error CS1061: 'ICollectionView' does not contain a definition for 'Where' and no accessible extension method 'Where' accepting a first argument of type 'ICollectionView' could be found (are you missing a using directive or an assembly reference?) [C:\projects\rubberduck\Rubberduck.Core\Rubberduck.Core_jmlei0ly_wpftmp.csproj]
UI\Inspections\InspectionResultsViewModel.cs(498,45): error CS1579: foreach statement cannot operate on variables of type '?' because '?' does not contain a public instance definition for 'GetEnumerator' [C:\projects
 
The one about Remove was there as well.
There was also one about foreach requiring a GetEnumerator.
 
Yes, that's the 2nd one in middle.
 
Gah, I stopped scrolling at the first one I encountered
 
@Comintern When building locally, did you clean and then build?
 
8:54 PM
@M.Doerner unrelated - I was looking at old issues and found #3373 - I thought a merged PR for rewriting would have fixed that?
 
That should be fixed by a combination of the PR reorganizing rewriting, the PR invalidating inspection results after the next parse and the one making it impossible to rewrite a module after modifying it manually.
 
I had just tested. It still exists.
 
Huh?
How did you reproduce it?
 
I created 2 private subs
run inspections, get 2 inspections for procedure not used.
Run quickfix to add ignore once to the first inspection. That pushes everything below down a line
Run quickfix again on the 2nd procedure
hmm. I could not reproduce it again.
 
9:04 PM
Did you wait for the reparse to finish?
 
no I did not.
went fast as I can
but this time when I try again, it's already disabled
 
so, it's fixed then
 
yeah looks like.
I think the first time was a fluke - there's a Remove Unused Declaration QF that I may have selected by mistake.
because I've tried it few times already and I cannot reproduce.
 
9:09 PM
I guess you're the faster gun in the West, Mat. :)
 
yeah but you're a better gun
:+1: linking the PR
 
In other words, Mat shot this in the hip, but this killed Mat?
:P
 
Better hope this doesn't get infected.
 
I think a shot to the hip in the west would have been debilitating.
mortal, even.
 
@this It would be healable.
 
9:11 PM
with the quack medicine back then?
 
Yeah.
 
where doctors were prescribing morphine and coke?
 
and whiskey
 
The doctors were supposedly actually pretty good at it because they had to do it moderately regularly.
 
Can't forget the whiskey.
 
9:11 PM
You'll understand that I have little faith in the 1800s medicine.
Yeah, by hacking off the limbs.
 
They'd give them whiskey to dull the pain, cut the bullet out, and sew them up.
Not that much medicine involved.
 
> Here's a shot for you, now a shot for me. Time to get cuttin'.
 
Here's the thing though - if it gets infected, not much they can do about it.
 
Heck, I've read stories where people shot through the lung survived.
@this The nice thing it was so clean out there that even TB patients could recover.
 
have you ever seen The Ballad of Buster Scraggs on netflix?
 
9:13 PM
Don't have netflix.
 
Too bad. That was fun one. I think it was 6 shorts in a movie.
 
Sounds like Buster was probably the descendant of the hangman, though, with that name.
 
nah, nothing like that. He was the fastest gun with a lovely voice to go wtih it.
 
Doesn't mean he wasn't...
I read a made-up story once about the fastest gun.
He got quite the reputation, and had to kill a bunch of people just to stay alive, even though he didn't care about it.
So one day he was walking down the street and hears "Reach for it, mister!"
He whips his gun out and turns around, and there was a 10yo kid.
He wasn't gonna shoot the kid, so he turned around and walked away.
The kid shot him in the butt.
 
and?
 
9:17 PM
That was the end of the story.
 
aw, left me hanging back there
 
Anyone else notice how much Geobbels looks like a cartoon mouse in the top one?
> What does the ideal Aryan look like? –– Blond like Hitler, tall like Goebbels, slim like Göring and chaste like Röhm.
 
@Duga s/may/should/
 
> Just for reference, the PR should close #3505.
 
IDK WTH I was doing back there.
 
man, dont you guys love that "Inconsistent database" error you get in access?
 
10:15 PM
when are you getting that particular error?
 
@M.Doerner Sorry, got dragged into a meeting there (apparently I get to learn Salesforce - Yay!). Yes - I was doing a clean build each time I switched between my 2 branches.
 
@Comintern LOL.
 
It was more like, "So, we're migrating our CRM to Salesforce and forgot that we're integrating with our invoicing system. Can you switch all the integrations to Salesforce by the start of March?"
 
My condolences, @Comintern
 
@Comintern Of course!
I can't wait to start! I've just been twiddling my thumbs all month!
Hopefully they don't have a broken sarcasm detector.
 
10:24 PM
I actually told them no chance by start of March, and I'd get back when I knew jack all about Salesforce.
Then we had a discussion about why you do requirement gathering at the start of projects like this instead of during.
 
I'm sure you'll enjoy Java.
 
Stack Exchange to the rescue...
 
I'm sure I'll figure out how to avoid it and just provide a web hook or something.
 
A sad story - we used to have a old client who had relationship with their client -- their IT used SalesForce and wanted to integrate the client's web API with their SalesForce. Guess who got to spend time reading Salesforce documentation to show the SalesForce developer how to use the web API?
 
Is it just me, or is that a bad omen when the Salesforce SE link doesn't even render correctly.
 
10:30 PM
Nope, bad omen.
 
That's what I thought. All I'm missing is the black cat doing laps of my office.
 
time to stockpile
 
apparently SE designers overlooked chat one-boxing when they decided the colors... or whoever implemented chat one-boxing for SE sites overlooked the design team's love of white foregrounds
 
lol
 
well, it's kind of cloudy soooo....
 
10:34 PM
Why is the developer documentation written in marketing speak?
 
redflags++;
 
just tell them that Salesforce was written by Lenin and Stalin's gulags
and that it's used in bloody conflicts for sales of diamonds.
and is a part of vast right-wing conspiracy.
whatever.
 
Hmm... This might not be that bad after all. I can probably just run a web service on our pre-Wordpress web server.
 
@MathieuGuindon i'm not clear how to reproduce this: 3472 - I think it's stale and should be close but I want to be sure about reproduction steps.
 
> This should be fixed by upcoming PR #4686
 
10:39 PM
I was never able to come up with clear repro steps for that. should probably be closed with no-repro if it's not happening
 
might as well. it's very old, too
 
meh, it's in the 3Ks
 
[rubberduck-vba/Rubberduck] retailcoder did something with some project card
 
#NotSoOld
 
dang, forgot about #263
> retailcoder self-assigned this on 23 Mar 2015
 
> We have no idea how to reproduce so we'll optimistically hope that it was fixed. 🦄
 
once again, I was too slow, and reopened it by mistake.
 
10:41 PM
[rubberduck-vba/Rubberduck] bclothier did something with some project card
 
#3472 is not consistently opened or closed.
 
It's a Schrödinger issue.
 
^ now that's a blast from past.
 
so the oldest issue is #13
 
10:45 PM
I think I did that wrong earlier today. I closed like 2 issues and ended up with 22 assigned to myself.
 
yeah. For extract method.
 
EM is just a prerequesite for it :)
 
Right, still.
 
#16 could be implemented now
 
10:46 PM
need to have less shiny things to distract with.
 
Ironically, it's the rate of new issue openings that prevent those super old ones from dragging the average close time down.
 
Actually, those shouldn't.
 
nah, issues older than 6 months aren't counted
 
AIUI, the stats only consider those flagged as bug.
and that, too
 
10:47 PM
^ also
 
Ah, nice.
 
That's why we only have 12% and ~9 days
though if you consider the totals, we have ~25% open
and definitely not ~9 days
 
Oldest bug is only #2294
 
10:49 PM
yep
 
@Comintern pretty sure it's fixed though
 
Vogel pointed out at one point that for the total amount of issues, bugs constitute only ~42%.
 
there's a lot of inspection ideas
 
Yeah, I was thinking of going through the old issues, including 2294 to re-test and update those issues.
That's why I was looking at them earlier.
 
10:51 PM
@MathieuGuindon Yep. And about half of them would be pretty easy to implement.
Some of these old bugs are fairly trivial fixes.
 
and the other half mind-blowingly hard
 
While it's cool to have lot of enhancement suggestions, it's nice to keep the bug count down (or up if you want to count as kill count).
 
geez so much stuff closed
 
Well, not really. I think we just updated more than closing.
 
10:54 PM
for the record: I like that the issues are being cleaned up :)
 
#SpringCleaning
(for some values of "spring")
 
Sweet!
 
now the oldest is #2299, and it's possibly Access-specific
if the "export all" command doesn't insert extraneous () parentheses at the end of an Access form's code-behind, it can be closed
 
Is that an export thing, or source control?
 

« first day (1685 days earlier)      last day (1495 days later) »