« first day (1955 days earlier)      last day (62 days later) » 

12:00 AM
RELOAD!
[rubberduck-vba/Rubberduck] 2 commits. 1 closed issue. 4 issue comments. 90 additions. 18 deletions.
[Minesweeper] New Users: 7, Games Played: 39, Bombs Used: 12, Moves Performed: 4626
 
@IvenBach Truly sorry - my sincere condolences
 
@IvenBach my deepest sympathies..
 
12:45 AM
@IvenBach :( I'm very sorry.
 
 
1 hour later…
@Duga oops, wrong branch
 
2:13 AM
[rubberduck-vba/Rubberduck] build for commit a4b574a1 on unknown branch: AppVeyor build succeeded
 
2:27 AM
> @bclothier nah they're going to want the download link... would it be possible to generate a link to the last green release in the body/description of every scripted pre-release?
[rubberduck-vba/Rubberduck] build for commit acdf44ee on next: AppVeyor build succeeded
[rubberduck-vba/Rubberduck] MDoerner pushed commit 8a368f35 to next: Add SuspiciousLetAssignmentInspection
Make ExpandDefaultMember able to fix suspicious Let assignments

This is generally deactivated if one of the sides is unbound.
[rubberduck-vba/Rubberduck] MDoerner pushed commit e98ab4d8 to next: Make UseSetKeywordForObjectAssignmentQuickFix able to deal with Suspicious Let assignments
Stop considering index expressions on Variants as default member accesses

They might as well be array accesses. Now, no references will be generated for these anymore.
Merge pull request #5167 from MDoerner/SuspiciousLetAssignmentInspection

Suspicious Let assignment inspection
> Right? It's WIP?
> The PR can be merged As-Is unless you want to have all the inspections updated/done at one time. I can work with either. If you merge this one, the next PR will simply start with tests beginning with the letter F.
 
2:52 AM
@IvenBach My heartfelt sympathies. As before, deal w/ family before the duck. Take care.
 
> In theory, it should be possible since you are already making the GH calls. The URL would be templatized as thus: `https://github.com/rubberduck-vba/Rubberduck/releases/tag/{name of release}`

Sample:

```
https://github.com/rubberduck-vba/Rubberduck/releases/tag/Prerelease-v2.4.1.5065
https://github.com/rubberduck-vba/Rubberduck/releases/tag/Prerelease-v2.4.1.5061
https://github.com/rubberduck-vba/Rubberduck/releases/tag/v2.4.1
```
[rubberduck-vba/Rubberduck] build for commit 95e95139 on next: AppVeyor build succeeded
 
@Duga total reading fail (was scrolling up from the bottom), saw the other post (which... I had already read before) too late!
 
LOL. Cascading failures.
 
[rubberduck-vba/Rubberduck] retailcoder pushed 13 commits to next (only showing some of them below)
[rubberduck-vba/Rubberduck] bclothier pushed commit 710acb45 to next: Add todo marker and clean up the comments a bit
[rubberduck-vba/Rubberduck] bclothier pushed commit fbcc077d to next: Update tracers and factory
[rubberduck-vba/Rubberduck] bclothier pushed commit 51b2b9c5 to next: Update ComPointer
[rubberduck-vba/Rubberduck] bclothier pushed commit b31892d5 to next: Update RdMarshal class to use conditional method
[rubberduck-vba/Rubberduck] bclothier pushed commit 78f4a824 to next: Merge branch 'next' of github.com/rubberduck-vba/Rubberduck into ReachableDebugCode
[rubberduck-vba/Rubberduck] bclothier pushed commit 2566dd85 to next: Fix missing namespaces in the Disabled attribute that is conditionally compiled.
 
loved the IOCInstaller diff
 
3:04 AM
[rubberduck-vba/Rubberduck] bclothier pushed commit cbaa7fb4 to next: Remove unused configuration. We only really use Debug and Release.
[rubberduck-vba/Rubberduck] bclothier pushed commit 88415fbd to next: Remove multiple conditions, leaving only a single condition to evaluate because conditions are OR'd, not AND'd. Using the unlikely flag will have to suffice. Also remove some str
[rubberduck-vba/Rubberduck] bclothier pushed commit c4f77497 to next: Provide the option of using THRISTY_DUCK to turn on firehose in both places as well the specialized THRISTY_DUCK_WM or THRISTY_DUCK_EVT for each place.
 
that [Conditional] attribute is awesome
 
Merge pull request #5182 from bclothier/ReachableDebugCode

Replace #if DEBUG with conditional methods
 
Thanks! I was feeling a bit too dirty when I put too much of code under conditional compilation, making them unverifiable.
 
@MathieuGuindon could you kick glowingrunes' build? I think it's a fluke.
 
3:17 AM
oh right I meant to do that earlier
kicked
 
:+1:
 
3:31 AM
[rubberduck-vba/Rubberduck] build for commit bed9ac58 on unknown branch: AppVeyor build succeeded
 
Fluke: Yes.
 
> Thanks again! We appreciate your help to make RD easier to use. Don't hesitate to tackle other issues that piques your interests. :)
[rubberduck-vba/Rubberduck] glowingrunes pushed commit bed9ac58 to next: Added the ClearSearchCommand to the SearchBox ValueContainer.
Merge pull request #5214 from glowingrunes/5137-TextBoxClearOnEsc

Added the ClearSearchCommand to the SearchBox ValueContainer.
 
@Duga why the Ctrl+Esc keybinding though?
 
Oh, I thought it was just esc
Shoot me for being trigger happy.
 
the Ctrl+Esc binding is on the hint container textbox
...which shouldn't be able to receive focus
hm, why does it need to be visible at all?
(not this PR's diff, but..)
 
3:41 AM
how else would we see the hint?
(or my WPF is sorely lacking)
 
seems the control uses 2 textboxes on top of each other (zindex)
 
yeah that's what I thought how it works.
 
but the cursor and focus should only ever be in the topmost box, no?
 
unless there's logic that sends the focus to the control behind
 
like tab navigation?
 
3:44 AM
(but that's WinForm's thinking... IDK if that applies in WPF)
I guess?
 
yeah
I think we might have a UI bug there
 
Of course we do. It's called WPF.
:D
 
@Duga I suppose the bitmaps that aren't displayed wouldn't have been streamed yet
lol, Ctrl+Esc pops the Start menu
 
4:37 AM
[rubberduck-vba/Rubberduck] build for commit 1528e65c on unknown branch: AppVeyor build succeeded
[rubberduck-vba/Rubberduck] bclothier pushed commit 6bef5f63 to next: Plug the WPF leak... maybe.
Merge pull request #5211 from bclothier/PlugWPFLeak

Plug the WPF leak... maybe.
 
#FunFact: We are 7 bugs away from passing below the 10% threshold for open vs. total bugs
2
 
@Duga I doubt you're missing many test cases, but more eyes can't hurt :)
 
4:53 AM
780 open issues. the velocity is impressive.. hacktoberfest should be all year!
3
 
@Duga would removing the ctrl modifier and binding on searchbox rather hintboc have fixed it?
 
the point of esc binding was to make it easy to clear searchbox?
 
the hintbox doesn't need an input binding
it just says "search"
 
no but searchbox does?
 
4:56 AM
yeah
and it's there
but just Esc
 
oh ok thought it was only on hint box
serves me right for being a bit hasty
 
meh.. I rubberstamped the javaesque extract method code way back when
 
#FunFact: 153/780 (19.6%) are issues
 
5:06 AM
that’ll make a good blog post at end of month
 
wait a minute. how do you get 6.8?
 
with a typo in calc.exe
 
lol
 
 
2 hours later…
6:47 AM
> Introduces a command for each annotation; closes #4790.
 
6:59 AM
[rubberduck-vba/Rubberduck] build for commit d8a54ba2 on unknown branch: AppVeyor build succeeded
 
7:38 AM
@IvenBach My sympathies.
 
 
2 hours later…
11:00 AM
Prerelease-v2.4.1.5072 in Word 2016 (Office 365) crashes during first parse without leaving a log to report. Do you want me to log it as a bug or just wait for the next pre-release?
 
11:10 AM
Please open an issue for it. We should never crash on load or on parsing.
 
11:58 AM
> This issue seems like an interesting one to work on for Hacktoberfest. Have we decided exactly how we're going to implement it though?
> **Rubberduck version information**

Version 2.4.1.5071
OS: Microsoft Windows NT 10.0.18362.0, x64
Host Product: Microsoft Office x64
Host Version: 16.0.12026.20320
Host Executable: WINWORD.EXE


**Description**
Rubberduck crashes during the 'Loading' phase of the first parse. There is no log produced. Version 69 was fine, the first and subsequent parses complete without issues. Crashing started in 71 and is repeatable in 72.
 
12:39 PM
 
1:29 PM
good morning
 
> The 69 corresponds to PR #5211. Presumably the fix is not stable and should be reverted
 
@Duga I will be away most of day so won’t be able to revert this until then,
 
> FWIW, I think that the new annotation command base should derive from ComCommandBase because rewriting indirectly does COM access. Looking good so far.
 
1:57 PM
> The 71 corresponds to PR #5211. Presumably the fix is not stable and should be reverted
 
2:09 PM
> To do this, we need to play in the window message pump from where we get the hotkeys and possibly the window message. Start with SubclassManager and its relative and then look at RubberduckHook and Hotkeys
> I just tried to replicate this and failed. I even opened the battelship XLSM and also tried Word but I did not crash.

Can you confirm it's not specific to a certain Word document?
> I just tried to replicate this and failed. I even opened the battelship XLSM and also tried Word but I did not crash. I also tried the release just in case it was a bug that manifested in release only but it worked, too.

Can you confirm it's not specific to a certain Word document?
[rubberduck-vba/Rubberduck] web-flow pushed commit 747b1c8b to revert-5211-PlugWPFLeak: Revert "Plug the WPF leak... maybe."
> Reverts rubberduck-vba/Rubberduck#5211

Provide a release to verify whether this still crashes - see #5217
> I just tried to replicate this and failed. I even opened the battelship XLSM and also tried Word but I did not crash. I also tried the release just in case it was a bug that manifested in release only but it worked, too.

Can you confirm it's not specific to a certain Word document?

I went ahead and created PR #5218 to help us verify in case it's a bug that only manifests on certain environments.
> Yes...the remaining test updates to follow.
> FWIW, my vote would be to merge this PR as-is. Get more of it vetted by our QA team users developers actually running the tests.
> FWIW, my vote would be to merge this PR as-is. Get more of it vetted by our ~QA team~ ~users~ developers actually running the tests.
 
Blargh stupid MD variants
 
> Also, you might want to note that [VBIDE sends a custom message](https://github.com/rubberduck-vba/Rubberduck/wiki/VBE-Window-Messages) on save:

```
WM_USER+3140 wParam:000000E2 = File..Save..
```
[rubberduck-vba/Rubberduck] build for commit 747b1c8b on revert-5211-PlugWPFLeak: AppVeyor build succeeded
> To help us test whether the PR #5211 is truly at fault, you can download this:

https://ci.appveyor.com/api/buildjobs/rmsfcmucy3l36k7x/artifacts/Rubberduck.Deployment%2FInnoSetup%2FInstallers%2FRubberduck.Setup.2.4.1.5075-pre.exe
[rubberduck-vba/Rubberduck] build for commit 747b1c8b on revert-5211-PlugWPFLeak: AppVeyor build succeeded
 
@Duga oh shoot, that wasn't supposed to be merged?
 
@MathieuGuindon it was supposed to be merged, but seems to cause issues that were not seen in development
oh damn, I misread.
yes, that wasn't intended to be merged AIUI
 
> They throw NotImplementedException because.. it's WIP. The idea is that some commands will need to bring up a UI prompt with a specific prompt string (but yes, the duplication will he addressee!); having a command per annotation makes it easy to later have an "Annotate" context menu though.
[rubberduck-vba/Rubberduck] web-flow pushed commit 69375aa2 to revert-5218-revert-5211-PlugWPFLeak: Revert "Revert "Plug the WPF leak... maybe.""
[rubberduck-vba/Rubberduck] build for commit 3dbbe653 on next: AppVeyor build succeeded
 
@Duga huh, had to refresh the page for the "create pr" button to show up
 
3:54 PM
 
4:18 PM
> Sorry for not responding sooner but I was out accompanying my SO on her new electric bike,.

5075 and 5077 still crash.

I reinstalled 5069 but now that is also crashing and I have to go back to 5061 to get stable behaviour.

5067 crashes but has other odd behaviours in that it doen't crash every time, but very occasionally produces an 'Unexpected error' parse message.

I have also done multiple reboots and rebuilds for Microsoft office.
 
note to self: merge-spree no more
 
> Seems that I recently reset Rubberduck settings as no logging was enabled <blush>. Logg from 5077 is attached.
[RubberduckLog.txt](https://github.com/rubberduck-vba/Rubberduck/files/3721950/RubberduckLog.txt)
 
@Duga If we even get to see an exception during the parse, it will only break the parse but not crash the host.
 
@MathieuGuindon Are you OK, if I merge part I of the inspection test refactoring?
 
Sure! I thought it was intended to have more commits coming
 
5:13 PM
I read it as part 1 of 2 from the beginning.
 
> The crash seems to happen in the code loading references. At least that is where the log stops.
Could you tell us which libraries you reference when you experience the crash?
> @BZngr Have you already done more work to push to this PR? Otherwise, I would merge this and you could simply push the rest of the refactoring of inspection tests to a new PR.
 
@Duga Better ask than be sorry.
 
5:33 PM
@Duga could it be the COM collector blowing up with one of these non-standard libs? it happened before...
 
Which PR changed something about the collector?
 
None that I know of, but maybe one of these libs is doing something we aren't handling
oh wait it does work fine with another build
?
wtf
 
I have a possble candidate. The PR about removing conditional compilation did more than it advertised.
It also activated generating CM projects for user projects, which really should have happened as a separate PR to enhance tracability of errors happening after this move.
 
> @MDoerner My most recent Sync now updates tests A thru H. My preference would be to merge this PR and follow up with an additional PR(s) as well.
 
5:50 PM
The typeLib API does stuff that can crash the host immediately.
 
6:32 PM
[rubberduck-vba/Rubberduck] Vogel612 pushed 29 commits to next (only showing some of them below)
[rubberduck-vba/Rubberduck] BZngr pushed commit d338d441 to next: Modified to use InspectionTestsBase
[rubberduck-vba/Rubberduck] BZngr pushed commit 5279d909 to next: Modified to use InspectionTestsBase
[rubberduck-vba/Rubberduck] BZngr pushed commit 94dff304 to next: Modified to use InspectionTestsBase
[rubberduck-vba/Rubberduck] BZngr pushed commit 63fca903 to next: Modified to Use InspectionTestsBase
[rubberduck-vba/Rubberduck] BZngr pushed commit 92b5d8db to next: Merge branch 'rubberduck-vba/next' into 5191_AthruE
[rubberduck-vba/Rubberduck] BZngr pushed commit 9d641383 to next: Modified to use InspectionTestsBase
[rubberduck-vba/Rubberduck] BZngr pushed commit 88a7b8f9 to next: Modified to use InspectionTestsBase
[rubberduck-vba/Rubberduck] BZngr pushed commit 64ef9658 to next: Modified to use InspectionTestsBase
[rubberduck-vba/Rubberduck] BZngr pushed commit 294442b8 to next: Merge branch 'rubberduck-vba/next' into 5191_AthruE
Merge pull request #5205 from BZngr/5191_AthruE

Incorporate InspectionTestsBase progress - part 1
 
 
2 hours later…
9:10 PM
Awwww yeah baby! I found Milton’s stapler. Mom had it.
 
10:08 PM
I am currently wondering a bit how the code for RegexSearchAndReplace for blocks works.
Am I right that there are no tests for this and that this is basically not validated to work as expected?
 
it's considered so utterly experimental that we never really announced it even exists, I think
 
        private List<RegexSearchResult> SearchCurrentBlock(string searchPattern)
        {
            var declarationTypes = new[]
                    {
                        DeclarationType.Event,
                        DeclarationType.Function,
                        DeclarationType.Procedure,
                        DeclarationType.PropertyGet,
                        DeclarationType.PropertyLet,
                        DeclarationType.PropertySet
                    };

            using (var pane = _vbe.ActiveCodePane)
I think the context labled block is exactly nevre a blockContext.
FindTarget returns the selected declaration with a valid declaration type.
Here, this will either be an event or a moduleBodyElement.
My problem is that I want to delete FindTarget so that I have to replace it with the appropriate alternaive.
However, here I have no idea what to do.
 
10:30 PM
> **Rubberduck version information**

Version 2.4.1.5081
OS: Microsoft Windows NT 10.0.18362.0, x64
Host Product: Microsoft Office x86
Host Version: 16.0.11328.20420
Host Executable: EXCEL.EXE

**Description**
Accessing the value of a Cell using either `MyWorksheet.Cells(1, 1).Value = "Foo"` or `MyWorksheet.Range.Cells(1, 1).Value = "Foo"` triggers the below Code Quality Issue result (see screenshot).

Is this intended behaviour and if so, what is the best way to use `Cells` without t
 
11:22 PM
> This result is legitimate. `Cells` is not a parameterized member of `Excel.Range`. Instead `rng.Cells(1,1)` expands to `rng.Cells.[_Default](1,1)`, which is basically `rng.Cells.Item(1,1)`.
In `worksheet.Range.Cells(1,1)` the call to `Cells` is acually superfluous.

The intended function of `Cells` is to turn the enumeration behaviour, e.g. in a `For Each` loop, into that of a two-dimensional array of single cells. This is the counterpart to `Columns` and `Rows`, which turn the enumeratio
 

« first day (1955 days earlier)      last day (62 days later) »