« first day (1270 days earlier)      last day (658 days later) » 
00:00 - 22:0022:00 - 00:00

12:00 AM
RELOAD!
[rubberduck-vba/Rubberduck] 2 commits. 2 closed issues. 5 issue comments. 111 additions. 6 deletions.
[Zomis/flexgame-server] 3 commits. 578 additions. 10 deletions.
 
 
2 hours later…
Much you have to learn still.
However, you are learning :)
No serious concerns there, except with your understanding about the differences between:
public bool FooProp { get; } = true;
and
public bool FooProp => true;
The first is evaluated once. The second is evaluated every single time it is called.
So, if you did this:
public int RandomNumber { get; } = (new Random()).NextInt();
You'd get the same number every time you called it.
If you did this:
public int RandomNumber => (new Random()).NextInt();
You'd get a new number every time.
(Well, it would probably be the same if you called it twice in the same second, but...)
Anyway, time to wash the dishes.
 
@Hosch250 The only way I'll learn is by failing.
The first one initializes it to the value.
The second calls it every time.
Not what I thought at first.
 
2:25 AM
Hmm. Interesting. This throws an inspection from the R#....
        switch (currentContext)
        {
            case VBAParser.FunctionStmtContext stmt:
                stmtContext = stmt;
                break;
            case VBAParser.SubStmtContext stmt:
                stmtContext = stmt;
                break;
            case VBAParser.PropertyGetStmtContext stmt:
                stmtContext = stmt;
                break;
            case VBAParser.PropertyLetStmtContext stmt:
                stmtContext = stmt;
                break;
            case VBAParser.PropertySetStmtContext stmt:
sez the loop variable isn't changed in the loop. Apparently not smort enough to see that I'm climbing the tree with the currentContext = currentContext.Parent
 
2:45 AM
@Hosch250 Please question me on anything change I propose. It'll make sure I can adequately explain my reasoning for doing the change.
 
3:07 AM
@this That a do/while loop?
Are you sure it isn't a plain while loop?
@IvenBach Nah, only way you'll learn is by succeeding.
 
I tend to find failure on my way to success.
 
Also, bedtime here. Got to get up at 6:30 tomorrow--holiday is over.
@IvenBach So do we all. Don't worry about it.
 
Thanks for bringing it to my attention. The initialization difference.
 
Yeah. I thought it was the same thing for a month or two too.
 
@Hosch250, no do; just a while at the end.
 
3:10 AM
Then R# is right.
LOL.
You have a block followed by an infinite while loop...
 
oh jesus
 
LOL.
If you want it to execute 1-N times, you do this:
do {
    ...
} while (condition);
 
it seemed off when I saw that inspection
gotcha.
 
@IvenBach Remember that ^
You don't see that construct a whole lot, but you need to know it when you see it.
 
and now the inspection is gone. Good job, #R. Bad job, @this.
 
3:12 AM
do { ... } while will always execute at least once. Without fail.
 
felt a bit weird saying do while since while (without a do) is valid in C#. Go figure.
Ayup, @IvenBach.
 
while (fooTest) { ... } may execute 0 times. It's dependent on the fooTest value.
Oh, I was responding to Hosch.
 
I know. I'm confirming anyway.
Also, #TIL RE: the initialization difference.
 
I'd read it before. At a high level I understood it. Now while coding I'll be a lot more aware of it.
 
3:20 AM
:+1: for running unit tests to keep yourself from looking like a duckhead...
 
[rubberduck-vba/Rubberduck] build for commit 5531e449 on unknown branch: AppVeyor build succeeded
 
@Hosch250 public int SomeNumber { get { return _fooNumber; } } can be replaced with public int SomeNumber => _fooNumber; because there is no backing field that's generated.
It's only grabbing the number and that's why it can be replaced with lambda syntax.
 
> A document module like `ThisWorkbook` has a hidden/implied WithEvents `Workbook` object named `Workbook` (although it is not addressable as `Workbook`, but rather as `Me`), against which, typically, all of the workbook events are wired up with a `Workbook_` prefix.

Likewise, a document module like `Sheet1` has a hidden/implied WithEvents `Worksheet` object named `Worksheet` (although it is not addressable as `Worksheet`, but rather as `Me`), against which the events of the worksheet are wir
 
3:44 AM
For github.com/rubberduck-vba/Rubberduck/blob/next/… can documentation, helpFile, and helpContext be discarded with the _ (underscore)? blogs.msdn.microsoft.com/dotnet/2017/03/09/…
Based on what I see, yes. I'm often mistaken and need further education.
 
looks to me yes
 
Ok. Made the change.
 
4:00 AM
_ is basically when you don't care about the output. So therefore, if you have no code that later needs it, it's usually safe to change to _ so you can cut on the variable declarations
 
That way there's no second guessing 'do we need this or not?' later on.
That was what I had understood. Made sure my comprehension was accurate.
 
yeah. going to turn in. GN!
 
Night.
 
4:19 AM
Rubberduck.Parsing.VBA has a couple methods constructors that are prompting for protected constructors. This falls out because the class itself is public abstract class ... and as such can't directly be instantiated but needs to be derived from.
Is this correct?
 
4:42 AM
Duck check: are LINQ queries preferred or is it best to leave them an non-LINQ? I'm guessing this falls into #ItDepends.
 
 
9 hours later…
1:34 PM
Hopefully someone has some kind of solution for this. I need to make a class that maps an entire sub-directory (archived reports) and I need the filenames so that I can match them against a pattern. I got all of that working just fine, but it takes 43 seconds to load the filenames (a retrieval for a single report otherwise takes about 48 seconds, so this would nearly double the time cost). The Scripting.File 'File.Name' accessor seems to be the issue. Any clue why? Or any workaround?
 
does the folder have a lot of files?
because the way Windows is, the more files there are in a given folder, the more expensive it becomes to enumerate the files - it's not a linear scale. If that is the case, you should consider dividing into further subfolders, perhaps by years & month for instance.
If the files are kept on a remote network drive, then you might want to consider using Windows indexing service and execute SQL query to get back file names.
 
1:55 PM
@IvenBach Yep--calculated properties can be replaced with it. Readonly properties can't.
@IvenBach Yep, it really falls into #ItDepends
That one has so many variables it isn't funny.
If it's a hotpath, avoid Linq.
Linq adds some significant overhead.
That's why we pulled it out of the parsing/resolving paths completely.
 
> The installer says that RD is only copyright through 2016. It should probably say 2017 by now, and in just a couple of weeks should read 2018.

Is it possible to get that part to use YEAR(currentDate)? :)

![image](https://user-images.githubusercontent.com/11889733/33270011-97ced3a4-d350-11e7-84cf-ee5eaf396aa5.png)

Note: the 'About' box correctly reads 2017, so I'm guessing that it is picking up the current year.
 
For inspections, we pulled it out of a few to improve performance.
 
Hey guys, if any of you have time to give me any feedback on my MVP attempt I'd be super grateful. codereview.stackexchange.com/q/181394/115693
 
@FreeMan Yes, we can totally do DateTime.Now.Year.
Oh, the installer? Yeah, that should be automagic too.
 
@Hosch250 there you go!
 
1:58 PM
That's something else, though, not C#.
 
And thanks for teaching me about MVP in the first place! It's gonna make a massive difference to a project I'm working on
 
so when it's year 34732, and an archeologist manages to load it, it'll say Copyright 2015-34732, even though it's long ago defunct. :p
 
Ugh. The copyright's year is hard-coded in the About too...
 
@this I dont have control over the directory, otherwise I would just keep them organized and wouldnt need a solution like this. The files are in a network drive, but I dont know if they are remote or not. Any examples for the indexing service I can test?
 
note that the indexing service has to be running on the computer that holds the drive
which means you need admin access to configure that machine
otherwise, that solution is no-go
 
2:05 PM
0
Q: Simple MVP Name-Finder

CallumDAI've recently learnt about the Model-View-Presenter architectural pattern and I wanted to give it a go with a simple VBA project. The aim is to retrieve a person's name from a UserForm. The userform looks like this: A user should select a team, which updates the name field > Select a name b...

 
Yeah, that wont be possible then. For anything I write I have to assume limited access.
 
Hmm. @BrandonBarney it looks like they changed the way they did it. So what I knew is outdated, too. FWIW here's the documentation anyway
Ok, one thing to check, though.
 
@this You say that as though it's a problem...
 
If you load up the cmd or ps, then do a dir/ls, how long does it takes to enumerate it all? is it equally slow to bring up a single file name somewhere in the middle of the name in the cmd/ps?
@FreeMan and how do you intend to defend the copyright of something that's no longer maintained, all contributors are long ago bones'n'dust? :)
 
what would be the CMD command for a dir on all sub-files as well?
 
2:09 PM
@BrandonBarney Is is possible for you to build your own index? (comes in late to the discussion, has no real idea of what's going on...) Maybe something that you can refresh in the background of your app and keep stored in a SQL server somewhere?
 
@BrandonBarney ask Google
 
@FreeMan Anything I create has to be internal to the project and will basically be instantiated for each call.
 
@FreeMan These file-paths could theoretically change at-will without me knowing, so I can't hold anything in memory for too long.
 
@BrandonBarney comes in late to the discussion, has no real idea of what's going on... it was a thought.
 
2:11 PM
@FreeMan No worries :)
 
@FreeMan if you can truly can reject reality and substitute your own, then you should have no problem flying off a cliff instead of falling.
</killjoy>
 
46
Q: List all files and directories in a directory + subdirectories

derp_in_mouthI want to list every file and directory contained in a directory and subdirectories of that directory. If I chose C:\ as the directory, the program would get every the name of every file and folder on the hard drive that it had access to. A list might look like fd\1.txt fd\2.txt fd\a\ fd\b\ fd...

 
@this lol
 
Plenty fast--I've done it ^^
Even on large(ish) directories.
 
@Hosch250 Thats C# not VBA :)
I'm not evolved yet
 
2:12 PM
That's because C# is smarter and better than VBA in every way.
And you'd better start evolving if you want to stay relevant :P
 
Great! Now go to Redmond and tell them to put C# in Office!
 
Mat's been telling me all about VSTO.
 
Meh.
 
I wish I could.
 
not that terribly impressed with VSTO.
 
2:13 PM
<~ Not that terribly impressed with VBA :P
 
it brings more problem to the table than it solves, IMO.
 
Just like VBA!
 
Baloney.
 
anyone have any idea why the VBE opens all code windows each time you launch it? I had 4 modules open on Tuesday (when I was last working), with them arranged all nice & neat like I wanted 'em, and now, everything's open with windows scattered everywhere! (Usually, I work full screen and it doesn't matter, but on occasion I want them windowed). haz a sad
 
If all you want is some functions, VBA has very very low barrier to entry.
C# need to give that or nobody wants it in their Excel spreadsheet.
with VSTO or even a DLL, you're forcing a plateau
 
2:15 PM
@this which, sadly, leads to some very very low-bar code, but you get what you pay for...
2
 
I'd rather have a higher barrier to entry and get better code.
 
@FreeMan, FWIW, I just close them all, then compile and save. I believe when you save VBA project, it remembers all windows setting but it's quirky. If it's compiled, saving it again doesn't update it or something.
Higher barrier to entry != better code
Correlation != causation, I am sure you all know that.
 
@this No, but it implies it :P
 
and you have to keep ROI in mind. When there's money involved, they don't have the luxury of climbing a plateau, much less installing anything.
I'd rather that they invest in an programming environment that made it easy to write high-bar code.
 
And that's why they get no return--they have no investment.
 
2:18 PM
Unfortunately true.
MSFT can be pretty pigheaded.
"Nobody writes in VBA code anymore! We don't need to support it!" ..... right. They tried to take it out from Mac Office
 
FWIW VBA has solved a ton of problems, and automated reports that otherwise took us hours or days. We are still finding the potential of VBA, so I would have a hard time convincing them "I need C# and 6 months of practice"
 
oh whoops it's back in the next version. Gee, I wonder why....
 
LOL, you need to get the 6 months of practice first, and then say "C# would solve this problem better and faster."
 
and beggars would be riding for last 6 months
 
Exactly, I dont have the ability to go to my boss and say "I need to learn C# so I can figure out which language would be better."
Heck, convincing him Python was better than R was a battle in itself.
 
2:21 PM
That's a debatable statement right there :P
 
IMPOV, majority of solutions in place came from non-programmers. That's why we have this business in first place.
But the solutions would have never happened if they always required a $10K starting price tag and 6 months gestation.
 
@Hosch250 Its only somewhat debatable. Many data scientists are pushing Python, even though R led the pack for quite a bit. For me, I won the argument by showing what I could do quickly and efficiently in Python, that he couldnt manage in R.
 
Yeah, data scientists are split on the two, last I heard.
 
They still are, theres pro's and con's for each
 
And I'm sure someone who knows R properly can do what you were doing in Python.
 
2:23 PM
FWIW, R is easier for a non-programmer to learn
Oh certainly they can
 
Are you suggesting he should have got you a search engine instead of Python? :P
 
I am at a dead end here. dir "\\some\file\path" /s /b *.xlsx should list all excel files in the directory, and it does, but then starts mapping my local drive as well.
@Hosch250 I am not sure what you mean. I enjoy learning Python plenty. He's the one who doesn't, but he is also a non-programmer.
And Python can certainly still be self-taught, but some of the elements are harder for someone new to the language to understand.
Not to mention RStudio makes it stupid simple to see what is happening with your variables.
 
I was jokingly hinting that you should have learned R better instead of Python :P
 
Oh, I had no desire to learn R, it wasnt ability it was interest.
I learned it for a bit and made a few machine learning projects, but I didnt see the value in it for the long term.
And I do get the subtle "Aren't you a non-programmer?" there.
 
TFS is down, and has been for half an hour.
 
2:30 PM
tell them to get rid of tfs and use git.
 
I have been.
 
@this interesting. thx for the feedback. I've just generally found the VBE to be pretty quirky. Maybe with the coming RD editor environment, RD will save currently open modules & window positions. hint hint
 
Looks like Git would be cheaper than TFS if we went through GitHub, anyway...
By $9/user/month
 
hmmm. interesting. Something to keep in mind. We currently use VSTS.
 
And maybe by $11/user/month, if we just got the team version instead of the whole business version.
 
Of course, you could just get git server for free
 
2:57 PM
Eh, GitHub would be easier to set up.
Especially since half the top devs don't like it too much, and they'd be involved in setting it up.
 
Yeah. I wouldn't want to be in charge of managing a git server, either.
so technically it's really a tradeoff between paying someone else to manage it vs. paying yourself to manage it.
 
theoretically a git server is not a setup in the first place...
it's just an open ssh port
 
Goes to show you how much I know. :) I know it's technically client-client, but I imagined there'd be additional steps to ensure that this repository is designated as the master of all others' repositories.
 
Not really. They are just set up with chained remotes.
 
3:21 PM
Still not resolved...
Been down for 1.5 hours at least (that's when I came in).
 
@this repository designations are local names
repositories don't have names.
 
Double-checked, I think I read the pricing wrong :(
GitHub is per user/month. VSTS is just per month.
 
hm. pretty sure it's user/month for VSTS.
I have to pay extra if I add a new guy to the service.
 
Oh, OK.
 
but it did came with 5 spots included at the base
6th and above, I pay a bit extra (don't remember off the cuff)
 
3:24 PM
better than remote desktop client licenses ...
 
Also, in some cases if they already have MSDN, I don't have to pay
 
those only come in packs of five
 
> Do note that this isn't equivalent, however....

Select Case True
Case MyConnection Is Nothing, MyConnection.State = adStateClosed
'Initialize the ADODB.Connection
End Select

Rephrasing this to an `If` as shown:

If MyConnection Is Nothing Or MyConnection.State = adStateClosed Then
'Initalize ADODB.Connection....
End If

Will result in a runtime error because the `Or` is not short-circuited as was in the `Select Case` expression.

With that
 
3:46 PM
> We'd have to start out with only doing it for cases with only one expression to match on.
 
@this Thanks for pointing me in the right direction. This ultimately ended up doing the trick:
Function GetFileList(ByVal FolderLocation As String)
    Dim WindowsShell As WshShell
    Set WindowsShell = New WshShell

    Dim Command As String
    Command = "cmd /c Dir """ & FolderLocation & """ /s /b /a-d"

    GetFileList = Filter(Split(WindowsShell.Exec(Command).StdOut.ReadAll, vbCrLf), ".")
End Function
 
3:58 PM
Yeah. FSO is convenient but performant it is not.
Curious - how fast is that trick?
 
4:13 PM
.10 seconds
Way better than 43
 
Anyone know the commitstrip with the guy hanging by one arm answering the phone?
I kind of need it for work.
 
The only thing I dont like is that it creates a command line window, but thats more of a nuisance than anything.
 
We just had that moment here, lol.
TFS is still down, 2.5 hours later.
 
curious why is it down? hosted or on-prem?
 
On-prem.
DB is apparently full.
I'm just going to set up a local Git.
 
4:25 PM
@BrandonBarney yeah, glad that worked out.
@Hosch250 that might be a good demo of why they should be using git. They wouldn't have to muck with a db - it's obvious nobody's administrating it since if it were, they'd have gotten an alert if db is getting too fat for the server.
 
same thing. whether it's hard drive or it's the db files, there should have been alerts for it.
 
 
1 hour later…
5:35 PM
Hosch250 has made a change to the feeds posted into this room
 
TFS is back up since about 30 minutes ago.
It was down for pushing 3 hours (I'd over estimated a half hour up above).
Well, 3.25 hours at least, since I came in, I guess.
 
@Hosch250 So a general rule of thumb for me would be to not use it until I'm fully grok the implications?
 
Feel free to use it.
Just, if there is an obvious place to use it that we aren't using it in, ask.
Most places, it's perfectly fine.
 
@BrandonBarney I know of an Open Source project that would help you learn C#.
@Hosch250 If I can understand the LINQ syntax I'll put them in. I don't want to suggest something only because 'R# told me is bettah'.
 
5:51 PM
No.
If we aren't using Linq, we probably don't want to be.
RD uses Linq more heavily than a lot of projects already.
Especially in the inspections and the parser/resolver we need to be careful with Linq.
Other than those two (large) places, feel free to use it.
 
@Hosch250 Speed is the main constraint for not using LINQ, for these alt least?
 
Pretty much.
 
Ok. Ill try and keep that in mind.
I better understand why you want the foreach and not LINQ if it's faster.
 
6:56 PM
huh... I didn't know we could get off-site RSS feeds in the ticker. Interesting, @Hosch250
 
It came in for you?
 
I had one earlier that wasn't SO.
 
The Commit Strip one I just added?
 
It had a blue icon. I dismissed it without thinking.
 
7:19 PM
Yeah, I saw 4 from Commit Strip, @Hosch250. I almost dismissed them, thinking they were from SU or SO, but realized they were very short, nothing like question text and did the hover-over.
 
7:55 PM
What is that fancy code pattern that allows you to create one array off of multiple ranges at once (equal sizes)? Or am I imagining something?
Nvm, it appears I am indeed imagining things. I can Union two ranges, but I cant make an array with them.
 
If they are the same column size you can Union then Copy to a destination and then use that destination to get it into an array.
 
That would likely decrease performance instead of boosting it.
For some context, I have a class that loads an entire sheet into an array, and then filters and outputs it. It runs just fine on ~20,000 records (about 20 seconds start to finish).
On ~80,000 records it takes about 80 seconds, huge perfomance decrease
 
You just wanted to do the filtering before loading it?
 
I was hoping to break up the pull into multiple chunks, but I need the headers in the array.
 
I don't follow. 20 seconds on 20K => 80 seconds on 80K is great.
 
8:02 PM
Nope, I want the filters to happen in memory.
It would be great, except previously the same operation on 80k records would take about 50 seconds.
And I know 30 seconds doesnt seem like much, but this could easily run 12 times in a row, and those minutes add up.
 
@this Seems linear to me. 20 = 20k and 80 = 80k.
 
Hmm. Interesting. One usually don't expect less-than-linear increase.
Yes, that's my point - if it's linear, then it's generally as good as you can get.
 
I'll run the actual numbers and see, but it isnt directly linear.
it has to do with RAM, the larger load is consuming close to all of the RAM, and thus a large performance decrease at the limit.
 
if it's less than linear, then there has to be indexing schema at play to cut down on the search scope.
more likely, I know nothing about Excel OM. :)
 
I don't have enough experience to give definitive help.
 
8:07 PM
No worries, I'll keep running tests and see if I can track down the issue elsewhere. The load process seems to be hogging less power now, so the problem may be elsewhere (or I may have fixed it by dropping the records once I got the filtered array).
Oh well, the time cost will have to do for now. Its at 92 seconds, with 52ish seconds due to the load and then output of the array. More coffee time I guess.
 
Step away from it for a bit. Mull it over and ideas for improvement may come to you.
 
Or you could post a CR post
 
I might. I want to get more of it done and then upload it to GitHub so I have somewhere to link to. I have an idea of how I could get smaller chunks of the tables, but I dont know if it will be effective.
 
8:40 PM
@this Wrong.
On the one hand, you could improve the process for processing a single record, and it would still be linear.
On the other hand, it might mean you are processing them in sequence and you should be doing it in parallel, or something.
(Which isn't possible with VBA, unless you can offload it to a DB or something, I know.)
 
I'm sorry but I think you made my point?
 
@Hosch250 Would be a shame to see you step back, but I do understand the need to broaden skills. FWIW, AvalonEdit/Intellisense will presumably be quite different from the parsing/analysis side of RD, and your work on Roslyn.
 
Linear is generally the best possible for raw reads, with no indexing schema or data structure to help things.
 
@ThunderFrame I'll probably still be around in the chat.
@this Often, yes. But just because a process expands in linear time doesn't mean it is optimal :)
For example, I took a for-loop at work and drastically cut the time of each iteration. It was linear before and after my work.
Essentially, I want to be able to work on small experimental projects without feeling I should be working on RD.
 
@Hosch250 Just the steepness of the slope was decreased?
 
8:53 PM
Agreed.
 
@IvenBach Basically.
 
Makin sure me comprehenderins' be good. :+1:
 
Not the least of which is setting up a site/business teaching people how to program.
 
I still look forward to being your guinnea pig.
 
For example, right now, I'm prompting Malachi with an interview project, and I'm horrified at how much he has to learn for being on CR this long.
@IvenBach You have been.
You are pretty far along now.
A couple years hard work, and I'll be willing to recommend you as a strong junior C# dev.
 
8:56 PM
You overestimate this pig, greatly.
 
No, you underestimate yourself.
It's been about 3/4 of a year since you started?
 
I just need to put into practice what I've previously read and connect the dots.
Feb was when I first found the light of RD, yes.
 
Well, a lot of the easy work has been done on RD.
Want to collaborate on a couple easy console apps to get yourself experience?
It'll basically be you doing it and me jumping up and down yelling :P
2
 
I think that'd help. It'd turn the tables on you and make you into Mug.
Having someone eager to learn and making mistakes.
 
Or really, I should have you write something and post it on CR.
It really helps to get lots of different opinions. And I could explain the answers to you, which was more than I had half the time :P
 
9:03 PM
I'm starting to understand the general idea of answers. Nuances and minutiae are still very elusive.
 
It'll come.
 
But when sensei?
 
Soon.
Lots of reading and programming lies between you and comprehension.
Practice is your only limiting factor now.
 
It's a hockey stick.
 
And reviews. Get reviews and practice more.
 
9:08 PM
That I will.
 
OK, project #1. Write a hangman.
 
Between now and the new year I will fix 2 issues I have my sights on in RD.
 
Cool.
But remember what I said about RD being big. It's best to learn on small projects--you can push the limits without breaking anything serious.
 
I want Mug overwhelmed by my PR.
 
And its more complex than it is big.
@IvenBach He just might be. In a different way :)
 
9:10 PM
Haha.
 
Although, that's getting less likely now.
 
I want the whelming to be because of what it fixes, not by the bad content.
 
I think it would be hard to overwhelm Mat, unless you suddenly PRed a fully-working tearable code panes that work in debug mode, lol.
 
Nope. You can have overwhelming or underwhelming, but no whelming. Ain't a werd.
 
@IvenBach In my experience, whelms are best done medium - neither over nor under
 
9:12 PM
^ Thunder got my joke.
 
Welp, call Oxford and Merriam and berate them for not making it a word. :p
 
No time. I've got RD to work on and Hangman to code.
 
This CodeProject (registration required to download) has a really nasty VBA parser - granted it was written in 2003, but OMG, RD's parser is so much more robust.
 
Given that the author wrote it to convert to C#, it likely wasn't meant to be robust.
I wonder if it's even better than the one that came with Visual Studio originally (do they still support VB6 conversions? I imagine not...)
 
@Mat'sMug/@M.Doerner I've got a "Module of Member 'Class1' has a description attribute but no annotation inspection... But the attribute is on the Event Foo (not the class) -is that possibly because the attribute parser fails to understand an attribute for an Event?
 
9:28 PM
have you compared the output to how VBA would have done it?
 
9:42 PM
@Hosch250 you should cut the "junior" there.
 
Sure :)
Well, do you know any company that doesn't advertise for a dev between junior and senior?
 
FWIW the classification into "junior", nothing and "senior" is IMO not tenable in it's current practice
 
LOL, agreed.
 
because it does not differentiate between the monk who made 10 mistakes, a thousand times each and the one that made 10000 mistakes
 
TBH I never actually saw a job posting that asks for a junior anything, much less a office boy.
 
9:45 PM
a junior dev is a dev that needs a lot of guidance.
nothing more nothing less.
 
Those days, "junior devs" will be titled something silly like "Ninja superstar level 5"
 
and @IvenBach you're easily fitting that description.
 
and seniors get to be "Program Unicorns"
 
OK, I'll qualify it as a position where he isn't the smartest in the room :)
 
it's hard work to be the smartest in the room
 
9:47 PM
Give or take another year doing C# dev full-time, and he'll easily be among the smartest in the room.
 
usually nobody always fits that description anyways
 
Except Jon Skeet
 
I remember how nhgrif was very happy about having stolen a checkmark from skeet
 
Me too.
 
I wonder. Would it be viable for medium-sized companies to hire a coder for ... say three months as an evaluation?
 
9:50 PM
That's literally what happened to me.
 
That's what we usually do at my company
 
90-day contract-to-hire position.
 
here we call it "probationary period" but same idea.
 
I got the position because they said they thought I would go far, even though I didn't have a lot of experience.
 
@this actually Germany has probation periods by employment law
but that's not what I'm talking about.
 
9:51 PM
France should take a lesson from them.
 
The problem I see is that many companies go for "acceptable is good enough for now"
 
@Vogel612, and I believe probationary period is almost universal. I don't see lot of companies not having one.
Anyway, finding a good talent is very hard
 
Maybe their economy would get a boost and they would stop just hiring people they know and start hiring people they view as "higher risk", such as minorities...
 
the french?
 
more often, I find that I end up firing the bright star because they can't work in a team.
 
9:52 PM
@this some people just don't really work well in a team
that's not terrible per se, but it's a problem if you must use them in a team
 
Yeah, and that's OK. But that's the setting we have, so yes, that's a problem for me.
I'm more likely to hire someone who is willing to learn and adapt even if they have less experience than someone who has 20 years doing everything his way.
 
00:00 - 22:0022:00 - 00:00

« first day (1270 days earlier)      last day (658 days later) »