« first day (260 days earlier)      last day (2920 days later) » 

00:00
RELOAD!
[Cardshifter/Cardshifter] 1 commit. 1 opened issue. 2 issue comments.
[Cardshifter/cardshifter.github.io] 4 commits. 2 issue comments.
[retailcoder/Rubberduck] 4 commits. 1 issue comment.
moahahaha, Cardshifter won today!
But I beat you all!
2
00:20
@SimonAndréForsberg We're going to get quite active next week, preparing for release of v1.2
01:03
> A lot of inspections show up as a "warning". While this may be appropriate for some inspections, we are under-using the "hint", "suggestion" and "error" severity levels.

Is it just because most of the implemented inspections are actually at "warning" level? I think we should classify them better for release 1.2 - let's discuss inspection severity levels here.
3 downvotes on the MSO ad... wtf
40 clicks in 10 days
I feel like removing it. I don't like collecting downvotes, and having the ad go below 6 would be a shame.
Deleted
> 3 downvotes, zero comments. What a lovely constructive community.
01:37
Really downvotes?
Wonder why?
I know why. VBA
> Error is pretty severe as a default setting. It's the kind of thing that a team decides they don't want in their code base. Not exactly the kind of thing we can decide. I'm ok with under utilizing that one.

Otherwise, yes. We need to create a list of all the implemented inspections for the docs anyway. So now is a good time to discuss what levels they should be at.
 
1 hour later…
03:09
@Duga Yeah!
@RubberDuck apparently the guy joined GitHub to star us!
        public virtual void Show()
        {
            try
            {
                _window.Visible = true;
            }
            catch (COMException e)
            {
                // bug: this exception shouldn't be happening. see issue #169.
            }
            catch (NullReferenceException e)
            {
                // bug: this exception shouldn't be happening either. may be related to #169... or not.
            }
        }
03:40
> That readme file needs to get more to the point; let's keep the docs in the wiki, and let the readme be an entry point to the wiki pages - the time spent maintaining the readme file could be spent improving the wiki. - [ ] Move features section to wiki pages The "coming up" section should also be in the wiki, on some what's new in 1.2? page. - [ ] Move coming up section to wiki pages That leaves the intro (which could be revised), a list of links to wiki pages, and then all the bottom...
sections.
03:55
[retailcoder/Rubberduck] retailcoder removed label help-wanted from issue [#198: Remove [ComVisible(false)] attributes sprinkled everywhere](github.com/retailcoder/Rubberduck/issues/198)
[retailcoder/Rubberduck] retailcoder commented on issue [#198: Remove [ComVisible(false)] attributes sprinkled everywhere](github.com/retailcoder/Rubberduck/issues/198)
> !image
R# FTW
    public enum CodeInspectionType
    {
        LanguageOddities,
        MaintainabilityAndReadabilityIssues,
        CodeQualityIssues
    }
04:30
> Indeed. Here's the inspections inventory, with their respective type and default severity level: - ImplicitByRefParameterInspection (CodeQualityIssues) | Warning - ImplicitVariantReturnTypeInspection (CodeQualityIssues) | Suggestion - MultipleDeclarationsInspection (MaintainabilityAndReadabilityIssues) | Suggestion - NonReturningFunctionInspection (CodeQualityIssues) | Warning - ObsoleteCallStatementInspection (LanguageOpportunities) | Warning - ObsoleteCommentSyntaxInspection...
(LanguageOpportunities) | Suggestion - ObsoleteLetStatementInspection (LanguageOpportunities) | Warning - ObsoleteTypeHintInspection (LanguageOpportunities) | Hint - OptionBaseInspection (MaintainabilityAndReadabilityIssues) | Hint - OptionExplicitInspection (CodeQualityIssues) | Warning - VariableNotAssignedInspection (CodeQualiftyIssues) | Hint - VariableTypeNotDeclaredInspection (CodeQualityIssues) | Warning And then there are a few others down the pipeline: -...
CanPassByValInspection (LanguageOpportunities) | Suggestion - ModuleOptionsFirstInspection (MaintainabilityAndReadabilityIssues) | Suggestion - CanPromoteStaticLocalInspection (MaintainabilityAndReadabilityIssues) | Hint - BadNameInspection (MaintainabilityAndReadabilityIssues) | Suggestion - CanExtractGoSubProcedure (MaintainabilityAndReadabilityIssues) | Warning - SetOnlyPropertyInspection (MaintainabilityAndReadabilityIssues) | Warning - AssignedInDeclarationInspection...
(LanguageOpportunities) | Hint - VariantUsageInspection (MaintainabilityAndReadabilityIssues) | Hint - CanCollapseIfBlockInspection (LanguageOpportunities) | Suggestion - CanReduceNestingInspection (MaintainabilityAndReadabilityIssues) | Suggestion - UnusedFunctionReturnValueInspection (CodeQualityIssues) | Suggestion - OnErrorResumeNextInspection (CodeQualityIssues) | Warning - CanExtractRedundantCodeInspection (CodeQualityIssues) | Suggestion
04:51
> removed ComVisibleAttribute usages
[retailcoder/Rubberduck] retailcoder closed issue [#198: Remove [ComVisible(false)] attributes sprinkled everywhere](github.com/retailcoder/Rubberduck/issues/198)
[retailcoder/Rubberduck] retailcoder commented on issue [#198: Remove [ComVisible(false)] attributes sprinkled everywhere](github.com/retailcoder/Rubberduck/issues/198)
> 143488a7 closes this issue.
05:27
> The keyword `Global` from older versions of the language has been replaced with `Public`.

Replace `Global` with `Public`.
 
6 hours later…
11:18
> I think this falls under "Suggestion". It doesn't cause any harm or confusion.
11:33
> * Obolete Call Statement : Suggestion
* Obsolete Type Hint : Suggestion; Maintainability
* Variable Not Assigned : Warning (.Net won't let you even compile in this situation)
* Obsolete Let : Suggestion

* Implicit Variant Return Type & Variable Type Not Declared are basically the same issue. They should both be warnings.

* Unused Function Return Val : Warning

Did I miss anything?
@Mat'sMug Did that break the config? It's okay if it did. Just need to alert users in the release notes if it did.
Also, might just want to bite the bullet and call it Maintainability and be done with it.
> The features should stay in the ReadMe. The read me is the first thing people see when they go to the repo. We shouldn't hide what the project does behind a link. I agree though, we could trim the readme down a bit.
> We're running short on time for this sprint and this is a backward compatible change. Kicking it to 1.3.
> Same as #226. Sprint is almost up and this is backward compatible. Kicking to 1.3.
12:02
> Merge branch 'master' of github.com/retailcoder/Rubberduck into UICleanup
> fixed build; idk why there was a reference to outlook in this file
12:22
hmm... parenthesis in links (needs to be tested)
wait, what? Why did it work for me?
stuff in [links [needs to be fixed]](codereview.stackexchange.com)
there we go
violà
Monking @SimonAndréForsberg!
Looks like I'm getting into multi-threading. I'm frightened.
Fun! Fun! Fun!
lol. Yeah. I don't like writing code I don't understand. =)
12:36
multithreading isn't that complicated
Not when you have years upon years of abstraction to work with.
I wouldn't want to actually work with threads.
at least you C# folks have async await and stuff...
and you do have some kind of ExecutorService stuff as well, right?
No idea. Kind of stumbling in the dark right now.
hopefully @Mat'sMug knows
12:52
I think I might just have it.... we'll see.
> TodoExplorer gets wait cursor while parsing; #171 begins
13:45
Apparently, a sufficient number of puppies can explain any computer science concept. #Multithreaded #programming http://t.co/BFZPzkYQwi
14:01
@RubberDuck Task.Run(() => { ... });
Nope.
Gives errors because it's trying to update the GUI on two threads.
Only the GUI thread can update the GUI from what I can gather from the exception.
14:19
Exactly. Look for (MethodInvoker) casts in the code base
I cast an inline delegate to invoke the GUI update - can't do it directly.
@RubberDuck found it?
14:43
Basically, each thread owns its own stuff, and cannot access objects owned by other threads. So you invoke a delegate on the UI thread, from the background one - and then the UI update runs on the UI thread :)
...at least that's how I understand it!
There's a CR about what I wrote @Mat'sMug. @Duga seems to be down, but I fixed some things.
2
Q: Displaying a wait cursor while we're waiting

RubberDuckI come from VB6 where everything is single threaded, so I've never written a lick of multi-threaded code before. I just added a wait cursor to one of our GUI's by stumbling through the docs, but I'm unsure I've done it the best way. Did I? Control is a Winform view and the code below is located ...

15:06
@RubberDuck Cool!
$5 the first answer suggests async/await ;)
You might be right.
(mine would.. no time now :(
That seems unnecessary though. The result assignment makes it wait. idk. Like the post says, I've never written a lick of multi-threaded code in my life.
 
5 hours later…
20:36
I've gone down the rabbit hole and come back around to that com exception again. Properly disposing of things causes it... Somehow. But I did find the null exception in dockable presenter base.
I can't commit. I can't fix it. I don't know what to do.
Yes I do.
Merge the last sync and grab a beer. It was from before I started this mess.
Or... Stash my changes and make sure it was stable first. Yeah. That's a good idea @Ducky.
21:25
It's not me. Master has the same issue.
21:35
Oh wow, I borked the extract method refactoring.. while refactoring :(
Fixed.
Oh, and inspections say "variable is never assigned" ...for a Const lol
I must have been more tired than I thought I was
Same issue in 1.1 too.
21:52
> Still an issue. The try/catch doesn't correct the behavior.
I was able to fix the null reference exception by creating a new toolwindow if it's null.
I still don't know why it goes null to begin with, but it does have something to do with closing the window via the x button.

Interestingly enough, the Test Explorer does not suffer from this issue. If we can figure out why it works, maybe we can figure out why there's an issue with TodoExplorer and CodeExplorer.
the test explorer is... not using a base presenter :(
we need to investigate what's happening to the window host in both scenarios
Gotta keep track. This has been driving me nuts for months.
the presenter is written though, it's just not used yet
selfie:
0
A: Wait, is that variable ever assigned?

Mat's MugExtract a whole class out of this code, call it IdentifierUsageInspector, and then have your identifier usage inspections use that class. Keep the methods simple, straightforward, and to the point, and the result will be (much) cleaner code. I boiled the GetInspectionResults implementation down...

 
1 hour later…
23:01
> fixed a bug introduced in 31a695c4 - an IEnumerable<QualifiedContext<TContext>> can't be filtered with .OfType<TContext> directly, must select the context first.
and boom
> fixed a bug where VariableNotAssignedInspection would complain about... constants never being assigned to.
> Cleaned up VariableNotAssignedInspection
> removed unused "using" statements
> added generic type constraint on TContext - method was extending every single type out there...
@Duga talk about micro-commits!
> added Equals and GetHashCode overrides for QualifiedContext<TContext>
> fixed/reimplemented VariableNotUsedInspection
voilà!
23:26
        public override IDictionary<string, Action<VBE>> GetQuickFixes()
        {
            // don't bother implementing this without implementing a ChangeSignatureRefactoring
            return new Dictionary<string, Action<VBE>>
            {
                //{"Remove unused parameter", RemoveUnusedParameter}
            };
        }
> fixed bug in UnusedLocals with usage.MemberName equality check, and implemented ParameterNotUsedInspection.
23:56
> added IndentifierUsageInspector.UnassignedParameters()
> Implemented ParameterCanBeByValInspection; removed "Pass parameter by value" quickfix for ImplicitByRefParameterInspectionResult, since that inspection doens't know whether parameter is assigned or not.

« first day (260 days earlier)      last day (2920 days later) »