« first day (1648 days earlier)      last day (1532 days later) » 

4:00 PM
great, wiki doesn't render them
 
would GH page?
 
no idea
 
Is that specific to commit?
I.e., does it break if the code changes?
 
@Comintern think so
 
Nice
 
4:01 PM
/blob/{hash}/ in the url gives it away
 
Have you ever heard the tale of Darth Plagues the Wise? Suffice to say, it happens often and can be pretty fun to watch. The Star Wars reference was taken from the talk that flipped a character from good to evil and was the most loved film of that trilogy. This mostly had the force of the fact that the hero was a popular villain in the original trilogy. Star Trek DS9 had a fun story where the hero not only joined the Bond villain, but destroyed the world for him, much to the Villain's shock and audience laughs. — hszmv 2 hours ago
LOL
 
4:20 PM
@Duga anyone wants to add anything (ping @Vogel612), feel free to go ahead!
(did I get this [mostly] right?)
 
LGTM
 
hang on ... it might've missed the one you also missed in your PR
AssemblyVersion on Line 28 of Base.csproj is missing
 
gah, thanks
 
4:29 PM
are any of you guys familiar with user alias' in sql server?
 
user alias?
 
yeah. its an old thing from the sql server 2000 days
 
"Credentials"?
 
sorta kinda
 
not quite
 
4:31 PM
so i had to use exec sp_helplogins to actually SEE them
because in SSMS they dont show up
they have a "\" at the beginning of hte username
and i cant reference any users that have a "\" at the start
im attempting to figure out the permissions assigned so i can delete the alias' and make regular users
 
[\name]?
 
\DSEGURA#
is an example
 
yeah try quoting them with square brackets
stab in the dark though, I've no idea - never seen those
 
@KySoto Try: domain\DSEGURA#
 
its not a windows login
its a sql one
but mat's stab in the dark seems to have worked
 
4:35 PM
I know, but they might've used the same idea at some point.
Oh, OK.
 
but, it failed because it doesnt exist or i dont have permission
despite being the SA
 
Probably doesn't exist.
 
-_- and i know it exists since it shows up in sp_helplogins
 
I mean, it probably looked for a different one than what you intended.
 
@KySoto got an account on dba.se?
 
4:36 PM
i specified hte database i KNOW it exists in
 
Like, it looked for "[/DSEGURA#]" instead of "/DSEGURA#".
 
you cant uses quotes on alter user
 
ask on dba.se, they'll know
 
The quotes are just to show what's the data and what's the English :)
 
it says expecting id or quoted id
oh, thats what a quoted id is... nice.
well. bleh, this is an annoying problem
 
4:55 PM
Is the CW default binding transient?
 
no
singleton
that's why it's documented that it's better to specify the lifetime explicitly
Max wrote up about that in the wiki, in case you haven't seen it
 
OK, good. Don't want to leak event handlers.
 
which one?
 
I need the reference persistence provider to listen for IReferences events.
 
hmm, isn't that VbeEvents's responsibility to provide that?
 
5:01 PM
Yes, but I need it in the settings provider to do the file IO when I get an ItemAdded event.
 
ah, not the COM's side.
 
No, managed side. I already got it wired up in VbeEvents.
I also only want to handle each event once... :-)
 
5:51 PM
pop-quiz: how the heck does this ever happen?
That last screenshot does raise eyebrows and depict unexpected behavior - can you confirm that the locals toolwindow isn't showing the value when the tooltip is broken like this? I've deleted my now-irrelevant answer, this is a legit issue. (thanks for edit/clarifying) — Mathieu Guindon 5 mins ago
 
fwiw, there was a similar situation I posted about about tooltip showing Nothing for an uninitialized array
 
the day RD cracks the VBE debugger's shell...... boy the fun we're going to have
 
hmm. Someone downvoted the question.
upvoted since this is a legit question
 
the original post wasn't half as clear
 
ah ok
 
5:58 PM
got 10K+ on SO?
 
lol
try a measly 500 points
 
checks @this' profile... retracts question
 
yeah, not exactly a hotbed of SO activity I am.
unrelated: I got a Revival badge for my CR question. Yay?
 
Revival would be for an answer IIRC
oh, it's your selfie!
got +2 on an answer posted to a question asked 30+ days ago
 
yeah.
 
6:01 PM
get +5 and that's necromancer :)
 
didn't think that would be such an impossible question to answer.
 
Huh. Did the breakpoint get inserted after the string was freed?
 
I doubt it
 
The only other idea I have on that one is that the compiler optimized away the assignment.
 
lol that would be a first!
 
6:03 PM
IKR?
 
wait VB6 does do that on compilation, right?
you might be onto something there
 
Yes, I wondered the same thing but that's to native machine code
 
I think so. It's a setting IIR.
 
I don't think VBA compiles to that.... right?
 
it compiles to p-code AFAIK
does that mean it doesn't do any compiler optimizations?
 
6:05 PM
Right. In VB6, you had a choice of compiling to p-code or to native machine code.
Sorry that'd be a Wayne question.
 
I keep getting told that VBA doesn't, but I'm not entirely certain.
 
"optimize for speed" or "optimize for size floppy disk install"
 
I'm sure there has to be some optimizations
 
I can't imaging there wouldn't be - document size was a big thing when VBA was introduced.
 
p-code was "optimize for speed", wasn't it
 
6:07 PM
I thought that was a separate setting
e.g. you had 2 settings of whether to compile to native code or to p-code, then whether to optimize for size/speed?
 
That has the ring of familiarity.
 
no, not exactly
 
eh, can't recall
 
 
6:10 PM
so yes they were 2 separate settings but optimizations are only available w/ native code
 
"Favor Pentium Pro(tm)" lol
 
IKR?
 
"no, my code needs to run on a 386"
 
back in the days, hardware mattered a lot more.
 
This glitch is worrisome, RD devs are looking into what could possibly be the cause (we'll need to understand the ins & outs of it if we ever want to enhance the VBE debugger!). One hypothesis is that the assignment might have been optimized away in the p-code, but not in the interpreted source. — Mathieu Guindon 59 secs ago
 
6:11 PM
remember how big deal MMX was?
 
that and "MMX-ready", yeah
and 4x CD-R drives
 
"Advanced Optimizations"?
Not sure I'm buying that one.
 
@MathieuGuindon yet so impractical
@Comintern read: unsafe
Turn off the bound checking! We don't need no steenkin' bound checking! Giddy-yap! Full speed ahead!
 
I always read that as an alternate label for the obiquitous "Advanced" button
 
I was thinking more along the lines of "Use i387 math coprocessor (experimental)".
 
6:15 PM
@this Damn the torpedos!
 
I wonder where VBA puts its .pdb "symbolic debug info" data
 
Did VB6 even use a .pdb?
 
Isn't that part of the pcode?
 
IDK
"Create symbolic debug info" rings that bell though
 
i'm just not sure if that actually is in same format as today's .pdb file is
 
6:17 PM
Did VS6 support attaching a debugger to a running process?
 
uh.... IDK.
 
I don't think so
 
I was still picking my nose back then.
 
I should ask over on Retro Computing SE... :-P
 
6:18 PM
^
 
OK, so WTH am I getting an invalid cast on icpc.FindConnectionPoint(ref g, out _icp);?
 
wait, isn't that screenshot from VB6, while the question is about VBA? Similar but different ducks...
 
what interface are you using?
 
that's... a nice signature you got here
 
You can blame me for that siggy, I think.
 
6:20 PM
Is there not an IConnectionPointContainer for VBA.References? It's apparently trying to return something.
 
wait, you created a SafeEventedComWrapper around the reference source interface, right?
 
there is one... wait isn't that already done somewhere?
the parser is already kicking in when you add/remove refs
 
I don't think it was ever implemented but my memory is... a filthy liar.
 
something is already listening to these events
 
I'm getting a Microsoft.Vbe.Interop._dispReferencesEvents for the type parameter.
 
6:21 PM
makes sense
 
Vbe.Interop => VBA, right, not VB6?
 
Oh, yeah - it was mostly wired up - I just made the call to AttachEvents.
@this Yes - that's determined by the SCW set in use.
 
Ok. The current code shows this: public interface IReferences : ISafeComWrapper, IComCollection<IReference>, IEquatable<IReferences>
 
It already inherits from the evented wrapper - I just corrected the interface.
 
6:24 PM
Thought so, ok
I wonder if there's another interface without the underscore?
 
I'm getting this GUID at the call site: {0002e118-0000-0000-c000-000000000046}. Wonder if that's correct.
 
e.g. should be dispReferencesEvents, not _dispReferencesEvent have to check the OleView for that.
I remember running into something similar because they dumbly had multiple interfaces with similar name for... reasons.
 
That interop is still a dependency, is it?
 
shoudl be yeah
 
I could just pop that open and find the GUID.
 
6:28 PM
:+1:
 
Does anyone have any quick takes on how much they like or don't like Google Apps Script?
 
aka javascript
'nuf said :)
 
Oops never mind then.
 
Did you really need to ask that? You should know my opinion on JS by now.
 
lol
wth are these getting wired-up anywhere?
 
6:30 PM
@Hosch250 Sorry, unintentional troll there!
 
@MathieuGuindon They weren't before, I need them though.
 
@MathieuGuindon Doesn't mean anything if they aren't dispatched via the VbeEvents
 
Though I'm happy because Javascript is routinely popular in the SO Dev survey per my memory.
 
they're harmful then
 
when I implemented that class, I set it up so that all COM events funnel through the VbeEvents.
 
6:30 PM
I still think it's funny that error manifests itself as a VBE messagebox.
 
Read: a main shutoff valve
 
bwahaha
 
that's... interesting.
 
6:31 PM
IKR?
 
that's how deep RD is into the VBE's guts - we get the VBE to think it's the source of our bugs :)
 
Pretty sure I didn't get that when I was working that out
 
It's calling QueryInterface internally.
When it doesn't get it, the COM object throws on the cast.
 
of course it has to.
 
It's just throwing into VBE-space and not RD-space.
 
6:33 PM
it's strange though because my memory (again, a stinking liar) is that I did get a proper VS error.
but that might been me with debugger attached or something, IDR.
 
I only get it in VBE if the VS debugger isn't attached.
 
Prolly why then
 
#CrossingTheStreams
 
Let's put Zuul back here he belongs.
 
6:34 PM
@MathieuGuindon not sure if good or bad...
 
it's in the same vein as "Excel has stopped working...", just less critical
apparently the VBE does have error handling in some places
what's funny is that it uses its "run-time error" msgbox for it
> Run-time? But I'm not running any VBA!
 
SCP Issue: Type Private Function IsReportRequested(| get Private Function IsReportRequested((). Happens pretty consistently. Delete the ((), type ( and get (|)
Yay!
 
ugh dammit
 
FYI, did I mention that all Manning MEAPs are half-off today?
 
Hmmm... GUID is correct.
[
  uuid(0002E118-0000-0000-C000-000000000046),
  helpcontext(0x0001873e)
]
dispinterface _dispReferencesEvents {
    properties:
    methods:
        [id(0x00000001), helpcontext(0x0001873f)]
        void ItemAdded([in] Reference* Reference);
        [id(0x00000002), helpcontext(0x00018740)]
        void ItemRemoved([in] Reference* Reference);
};
 
6:39 PM
@Hosch250 someone checked the "Yes! Sign me up to receive weekly spam mail with offers and discounts!" checkbox
 
No, I check the daily offer every day.
And I've been doing the Countdown to 2019. I won a book in it yesterday.
 
@Comintern and the References coclass implements that as the source interface?
 
Every day in December, you answer a question. 2 people of everyone who got it right wins an ebook of their choice.
 
> RD .4291

SCP is consistently inserting an extra `(` when defining a method signature:

Public Sub Foo(|

Yields

Public Sub Foo((|)

a single backspace from that cursor location removes the `()` leaving

Public Sub Foo(|

Which is where the vanilla VBE would have left me in the first place.
 
At the end, 1 person of everyone who got the most right get every 2019 MEAP.
 
6:41 PM
@this Yep
 
@FreeMan I can repro, but there's also a way to make it not do that... somehow... unclear... lovely whack-a-mole feature.
 
coclass References {
    [default] interface _References;
    [default, source] dispinterface _dispReferences_Events;
 
with 2nd _?
_dispReferences_Events?
 
@Hosch250 @Hosch250
 
4 hours ago, by FreeMan
flags @Hosch250's posts for spam
 
Not _dispReferencesEvents?
 
OHAI dispinterface _dispReferences_Events with the extra underscore.
 
6:43 PM
Bingo
That's why
use that interface instead
17 mins ago, by this
I remember running into something similar because they dumbly had multiple interfaces with similar name for... reasons.
 
@MathieuGuindon happy to help :)
 
@FreeMan Cool. I want to make sure everyone who wants a good book can get it cheap!
 
I swear that feature will <power-Gandalf voice> BEND TO MY WILL
2
 
@Hosch250 think Manning can give Mat a free iron gauntlet?
 
@Hosch250 :D
 
6:45 PM
@this As far as I'm aware, they aren't in the iron-working business or in the clothing industry.
Mat's in the clothing industry, though, so...
 
Maybe if we ask real nice, they'll iron his guantlet...
 
@this ...and the Interop doesn't expose that. Do you remember how you handled that before?
 
would dynamic help?
 
you're kidding.
Not sure we can use dynamic - we need the actual interface
VBE
@Comintern do they both have the same guid?
 
@this god no. VBE extensibility is a stupid useless pile of unreliable half-exposed goo as far as events are concerned.
 
6:47 PM
Ummm... they can't?
 
in my case, the interface I actually needed was already exposed, so it wasn't.
 
Unique
 
I'm still trying to work out why you got the GUID for a different interface in the first place
 
@MathieuGuindon read that as dynamite...
 
@this That makes 2 of us.
 
6:48 PM
that might also help... in some ways.
 
There's like 3 of them.
 
IKR?
i'm sure that was during some intern's bong time.
 
"...49 ways of doing it right..." -somebody 'round here
 
Event sinking into the bong water?
 
6:49 PM
anyway, if you don't have the interface, you'll have to import it.
 
Well f*ck.
 
that shouldn't be hard
 
This was supposed to be the easy part though.
 
just create an interface that's compatible, decorate it with the ComImport attribute and provide the same GUID of whatever that 2-underscore thingee is
i'm sure you even can copy'n'paste from the 1-underscore thingee
and just change the guid but the ComImport is important.
then add some colorful commentary about how interns at microsoft are hitting the bong a bit too hard back in '98.
 
Oh, I have plenty of colorful commentary.
 
6:52 PM
remember a certain black duck thinks we need more comment. Might as well provide some entertainment.
 
I've heard some things about black sheep that makes me think that if black ducks are any relative, we may not want them hanging around our project.
 
nah, referring to the company named BlackDuck -- the one that's hosting the openhub project
 
I know :P
 
:facepalm:
 
Don't mind me. I'm Captain Obvious, you know.
 
6:55 PM
Running the build clean fixed it.
 
wut?
 
Your guess is as good as mine.
 
so it works with the one-underscore interface?
 
No, it picks up the hidden 2 underscore interface.
 
oh not too bad
at least it's making sense in its own insane way.
 
@Comintern interns are still hitting the bong?
 
lol
 
Not sure. They may have moved on to something beyond the bong.
 
@this thanks, Cap'n Obvious, I'm Cap'n Oblivious. (#TIL)
 
6:59 PM
OK, down to 2 items:
Insert the command in more locations (i.e. RD context menus, RD add-in menu, etc.).
Add custom icons for projects?
 

« first day (1648 days earlier)      last day (1532 days later) »