RELOAD!

[rubberduck-vba/Rubberduck] 2 issue comments.

[Zomis/Games] 5 commits. 302 additions. 108 deletions.

[Minesweeper] New Users: 16, Games Played: 140, Bombs Used: 104, Moves Performed: 17644

I think the issue is not the number of results. It rather looks like there is a race condition in the rendering engine we stumbled upon when adding the scroll viewer.

TBH, we do something that really is not intended. On the grouping grid, we both have grouping and virtualization enabled. This is not a supported scenario for datagrid. Using groupings usually disables virtualization.

The main problem debugging this is that I have never been able to reproduce the issue.

[rubberduck-vba/Rubberduck] MDoerner created comment on issue #5442: MS ACCESS and VBE hung with heavy CPU\MEM use after parse

> @joyfullservice Thank you very much for the link to your repo. I am finally able to reproduce the issue. So, I am finally be able try to figure out what is really going on.

Anyone in contact with Hosch? IINM, he's in the twin city and it's kind of burning.

oh, good question...

...i mean, i laughed. that is awesome, and i gave him +1 on actually self-answering to close the question.

the question itself wasn't posed all that well

@this i've been trying to figure out what happened to him for months... haven't seen him log in here in the longest time

@Cyril AIUI, he deleted his SE account and moved on.

@Cyril he... yeah, ^ this.

oh damn; things either must have been rough for him or he got really fed-up with something ='/

he and a bunch o' others. You've heard of Monica and "the incident"?

from friends?

LOL

No, Monica from one of SE sites.

oh, was thinking about when they ruined her engagement...

oh my, just hit me - Monica from Friends, Monica Lewinsky, and now Monica the Moderator. Monica is not a good name to have.

oh gosh, mind blown

not that familiar with mods, but i believe i recollect that they pretty much scrubbed meta-feedback and all the OG folks jumped ship

@Cyril this is such a great question

Yeah - I didn't know Monica the Mod until the incident and AFAICS, she was pretty level headed but got handed a raw deal because something something

and all mods were a bit peeved about that.

helps me answer my question:

so hosch jumped ship with that?

@this i typically don't think of the name monica without following "lewinski"; so engrained

well not exactly at that point but eventually yes.

yeah, Monica the mod is the anti-Monica Lewinsky but gosh, Monicas seem to have a tendency to end up in news.

Looking at the memory at hex(objptr(Application.VBE.ActiveVBProject)) - lots of interesting pointers, but I have no idea what they represent. Most of the VB6 structs have been decoded, but I can't find anything for VBA...

aside from that convo... everyone holding up? haven't talked to many of you in a month or so? about to jump on a call for the next 2 hours, but will check back after

I need the VBA equivalent of this...

OK considering all things.

Yep not bad. Bit bored, I wish work would unfurlough me

@mansellan I am not sure I follow what you are trying to do. The link is about an image structure?

Yeah, it's the memory layout of various structs. You can use it to get the pcode and all the metadata.

So for example, the typelib API uses that trick to get the embedded TLB. I want to do the same for the project itself

Yeah,

@this glad to hear

in VBA?

IDK if that's a good idea. you'd be doing lot of pointer chasing and VBA make it really hard to do any pointer chasing, especially beyond the first pointer.

not fussed which lang, just spelunking with process hacker atm

Again, why/

If we can reliably decode the meta, we could get a lot of info for RD much quicker than doing a parse run. the compiler has already collected it.

finding event handlers for one thing - not sure if the typelib api gives us those?

Yes

we just need to enable it and consume it from ComProject

Meaning, the code explorer can be modified to run off the typelib API

that'll be nice :-)

I'm basically just doing research atm, see if I can find any gems.

right now, CE uses Declarations which is OK. What we really need is something like mutable Declaration so that CE can create Declarations from TypeLib API + ComProject, and when parsing finishes, to update? upsert? the data.

sounds, uh, "fun"?

@Cyril still really dislike WFH full time. Once I can find an open furniture store, I'll be buying a new desk & chair. I've got an old office desk from the days when typewriters ruled the world and what was a nice desk chair. The chair is failing and it just doesn't fit the desk very well. A day at a time is tolerable. After weeks on end, everything is sore from sitting here 8 hours a day.

Well, possibly an easier option is to just create fake Declarations from TypeLib API/Comproject, then invalid it once the parsing run ends.

How are you doing with your camp chair and card table? (it was something like that, wasn't it?)

that way it won't need to be mutable, but that means more churns.

keeps the GC busy. It gets bored otherwise :-)

yo, boy, take out the trash!

Lol, just loaded VBE7.dll into resource hacker. There's a MIDI file embedded in there. Playing it now, it sounds like a 90s ringtone!

Related: If you ever wonder why MS files seem bloated...

lol

more usefully, there's a couple of TLBs in there

Why is there a MIDI file in the VBE7.DLL???

#BlameMicrosoft

found where class_initialize and class_terminate are defined

so it does uses DISPID_CONSTRUCTOR and DISPID_DESTRUCTOR!

heh, I'll take your word for it, I don't really know how to read those. That's from a hidden TLB called VBInternal.

// Generated .IDL file (by OleWoo)
[
  uuid(6715f050-f076-11cd-9b33-08002b336f14),
  version(6.0)
]
library VBInternal
{
	// Forward declare all types defined in this typelib
	interface IClassModuleEvt
	dispinterface _DClass
	dispinterface _DPropNameOnly
	dispinterface _DPropPublicClassModule
	dispinterface _DPropPrivateClassModule
	interface IVBProjProps
	dispinterface INullExtender
	dispinterface ENullExtender
	[
	  uuid(fcfb3d21-a0fa-1068-a738-08002b3371b5),
	  hidden
	]
	interface IClassModuleEvt : IUnknown#i {

^ there's the whole of it

yeah ThunderFrame mentioned that sometime ago - you can also access it by adding /2 (or hwatever) via the references dialog

ah that makes sense

I wonder what's up with that NullExtender, though

ikr

"let's extend null. It's not big enough!"

"with what?"

"null, of course!"

MOAR NULL!

@Duga I just tested a bit and the inspection results window is not the problem in this case; it is the scrollviewer on the test explorer window.

what?

even if it's not showing?

Yes

It does not really matter whether the windows are showing because they are always there.

The VBE only hides windows and never closes them.

BTW, I think there are no tests at all.

That also causes a NRE somewhere.

Hm, the test explorer still had the request bring into view handler swallowing the event.

Let's see what happens if I remove that.

it's true that VBE doesn't do away with the windows when closed but it doesn't mean that we have to have it hot if it's closed. We still control the WPF and thus can disable that.

Hm, after removing the handler, it works fine.

[rubberduck-vba/Rubberduck] MDoerner created pull request #5506: Remove handler for RequestBringIntoView from TestExplorer to merge MDoerner/Rubberduck/RemoveBringIntoViewHandlerFromTestExplorer into...

rubberduck-vba/Rubberduck/next

> This PR is another stab at resolving issue #5442 Not swallowing that event in the test explorer made the sample project provided in the issue stop freezing for me. (In PR #5482, it had already been removed from the inspection results window, but not from the test explorer as it had been added in a different PR than the one introducing the issue, probably a later one.) I hope this change also works for other people. Once we know that it really works in the field, we might reintroduce a...

handler, but using a better scheme than swallowing the event to handle the original problem for which the handler has been added.

[rubberduck-vba/Rubberduck] build for commit eeb0a86c on unknown branch: AppVeyor build succeeded

[rubberduck-vba/Rubberduck] build for commit eeb0a86c on unknown branch: 61.14% (target 0.00%)

[rubberduck-vba/Rubberduck] codecov[bot] created comment on pull request #5506: Remove handler for RequestBringIntoView from TestExplorer

@FreeMan i completely hear you... i've been using a nightstand to hold both monitors (stacked upon the laptop and docking station)with a lawnchair to sit in (i'm on chair #2 after the first one died)

even if not ideal, at least it's work... i have a few friends who are going out of head and mind with nothing that entertains them consistently for 3 months

[rubberduck-vba/Rubberduck] retailcoder merged pull request #5506: Remove handler for RequestBringIntoView from TestExplorer from MDoerner/Rubberduck/RemoveBringIntoViewHandlerFromTestExplorer into...

rubberduck-vba/Rubberduck/next

[rubberduck-vba/Rubberduck] MDoerner pushed commit eeb0a86c to next

[rubberduck-vba/Rubberduck] web-flow pushed commit 37bd1bb1 to next

even if not ideal, at least it's work... i have a few friends who are going out of head and mind with nothing that entertains them consistently for 3 months

[rubberduck-vba/Rubberduck] rubberduck-vba-releasebot prereleased draft prerelease Prerelease-v2.5.0.5457

[rubberduck-vba/Rubberduck] rubberduck-vba-releasebot created draft prerelease Prerelease-v2.5.0.5457

[rubberduck-vba/Rubberduck] rubberduck-vba-releasebot published prerelease Prerelease-v2.5.0.5457

[rubberduck-vba/Rubberduck] rubberduck-vba-releasebot created tag Prerelease-v2.5.0.5457

[rubberduck-vba/Rubberduck] build for commit 37bd1bb1 on next: AppVeyor build succeeded

@Cyril I lol the moment I read the title.

[rubberduck-vba/Rubberduck] MDoerner created comment on issue #5442: MS ACCESS and VBE hung with heavy CPU\MEM use after parse

[rubberduck-vba/Rubberduck] jan-becka starred us

[rubberduck-vba/Rubberduck] ubidev created comment on issue #5442: MS ACCESS and VBE hung with heavy CPU\MEM use after parse

[rubberduck-vba/Rubberduck] ubidev edited comment on issue #5442: MS ACCESS and VBE hung with heavy CPU\MEM use after parse

I just had some WTF while trying to do a unit test. Kept giving me object or with not set stuff on my asserts. The problem magically went away when I did a parse....

So question is: Shouldn't the run test check if the parse is dirty and tell you to parse before running your tests?

[rubberduck-vba/examples] web-flow pushed commit b13f9a79 to master: Create builder folder

[rubberduck-vba/examples] web-flow pushed commit bc229a6b to master: Delete builder

oops

@HackSlash hm.. it.. does ..that's how it discovers unit tests. or maybe that's just when you "run all"

[rubberduck-vba/examples] web-flow pushed commit efbe88f0 to master: Create readme.md

[rubberduck-vba/examples] web-flow pushed commit 1e0fdc98 to master: Add files via upload

[rubberduck-vba/examples] web-flow pushed commit 8b3e5628 to master: Update readme.md

@MathieuGuindon I was right clicking on a single test and running that one test over and over again. I couldn't figure what I was doing wrong.

Not sure how parsing can affect the outcome of a test though. Is something weird going on with the Assert object instance?

Yeah, it was like the Assert object wasn't initialized because I got object errors on all methods of Assert.

Oh, you have a note about this in the unit testing page

This is looking like PEBKAC

[rubberduck-vba/Rubberduck] ubidev created comment on issue #5442: MS ACCESS and VBE hung with heavy CPU\MEM use after parse