« first day (858 days earlier)      last day (2532 days later) » 

11:00 AM
Hello
 
@SimonForsberg 10 queens crashed my computer :O
2
 
Greetings
 
I'm back after hard shut down (actually pressing the power button)
 
@SuperBiasedMan I don't really understand why print(line) and print line print an extra line break, but print(line), doesn't, no; that seems a bit odd
 
11:11 AM
@Phrancis It's a specific construct for Python that the trailing spaces omits the newline that's automatically printed as part of a print call. The brackets don't matter, print line, would have the same result.
 
@Caridorc well that's not optimal
2
 
So there's no explicit just "print" / vs. "println" (like in Java) then @SuperBiasedMan?
 
@SimonForsberg there is some permutation generating code that should be optimized
 
@Phrancis Correct, there's just comma or no comma. I think in Python 3 you can optionally pass a specific parameter of what to append to the end of the line (including passing an empty string) but 2.x just lets you add a comma to change from \n to a space.
 
Ahh OK, definitely good to know, it was driving me a bit crazy there for a while :)
 
11:16 AM
@skiwi Not entirely correct, but close enough.
 
Yeah print in Python 2.x is a bit funny because they made it a statement instead of a function, and that involved some oddness.
 
I enjoyed hacking up my NQueens program.
 
Did I read somewhere that Python compiles to bytecode?
 
If you run an import on a file it compiles to bytecode in a .pyc file, yeah.
But generally people run the actual source files. Compiling is just done on imports for the mild speed boost if you have a lot of modules that have to keep running.
 
Gotcha
Thanks a lot for the review!
 
11:21 AM
@Phrancis So if you want to run the same thing multiple times, make sure to put the heavy stuff in an import instead of in the main file.
It matters in speed, a lot.
 
Does string().format work on other things than print too, like write and such, and generally for any string?
 
My computer is midway a 8096x8096 fractal, is going to come out nice and sharp
 
Yeah, format is a method of all string objects.
 
Great
 
And no problem! If you want to know any of the other stuff format does there's a bunch about it here
 
11:25 AM
Groovy's GString has a construct like that too, sort of, though I like its syntax a bit better... print("// {url}\n// Attribution: {attribution_text}"
 
You can do that too sorta, but it might look redundant as print("// {url}\n// Attribution: {attribution_text}".format(url=url, attribution_text=attribution_text))
 
Ah ok
I'm guessing in that case you could specify the replacement parameters in any order?
 
named parameters ftw,it seems...
 
Oh by the way, I'm curious if you agree or disagree with my assessment of your class and if there was a use for it I didn't think of?
 
11:31 AM
Challenge time...
Does it make sense to try to solve NQueens on the GPU?
 
There is CodeReview, site of StackExchange for that kind of demand. But I think it's too even too broad for them, narrow thing by thing: codereview.stackexchange.comLarme 41 secs ago
 
0
Q: Playing with validation

Please TeachThis code is part of an application page, which lets user select existing or add new members, For email and Username it needs to call web services to see if they already exists then validate respectively according to business logic. I believe this code has a lot of potential to be improved. Plea...

0
Q: The Perfect Logician that used BruteForce against Queens

CaridorcI wrote a Haskell program to solve the N-queens problem by bruteforce. It works and I find it reasonably readable But it is pretty slow: 5 seconds for one solution of 8 queens. 1 minute for one solution for 9 queens. I would like to hear both improvements on readability and performance: im...

 
You should post the most important part of your code to be reviewed at CodeReview — Caridorc 5 secs ago
 
@SuperBiasedMan I agree, a class there was probably not needed
 
Code must be included in the post at CodeReview — Caridorc 19 secs ago
 
11:35 AM
@Vogel612 I'm a glutton for named parameters for some reason, I used them pretty much every occasion I get
Well it's TTGTW
 
@Larme This is not at all too broad for CR actually, we like questions that are open to any kind of improvement. See also A Guide to Code Review for Stack Overflow usersSimon Forsberg 14 secs ago
 
@Phrancis Ok, was just curious because I almost never use classes so I could've easily missed something.
 
I renamed my question to a more sensible title
 
@CaptainObvious @Caridorc why didn't you add the information about 10 queens? ;)
 
@SimonForsberg added, the queestion remains on-topic if it crashes for too big inputs right?
 
11:45 AM
yup
 
Oh, hai @Caridorc. That question of yours brought me in here
As somebody whose sole question on here was about my solution to a more complex variant of the n-queens problem in Haskell ;)
 
@itsbruce hi
 
By brute force, do you mean "avoiding laziness and really forcing Haskell to create every possible board position and then filtering them?"
Because if so, why the hell?
 
brute force usually implies "pick one at random, see if it works"
 
@itsbruce no, I mean not inference engine
 
11:50 AM
Bwhahahahaha: "I would like to hear both improvements on readability and performance" <---- Use Java! ;-)
 
I generate a stream of all reasonable board and filter them by the isValid function
 
The most significant way to brute the problem is to short-circuit broken boards before you complete them.
 
Quite
 
@rolfl done I use any
That short circuits
maybe fAnd makes things slower though...
 
6
Q: Better or more idiomatic way to apply a sequence of functions in Haskell

itsbruceThis is one of my Haskell solutions to a variation of the N-Queens problem, where the queens can move like knights in addition to their normal moves. It has been optimized somewhat to avoid investigating redundant combinations. place :: Int -> [[Int]] place 0 = [[]] place n = go $ take n $ repe...

 
11:53 AM
@itsbruce does my use of fAnd (functional end) kill short circuting?
 
In the classic n-queens problem, a solution like mine there is about 10 lines long. Curious why you want to do something less performant
I mean, more complex (but also, as it happens, inevitably less performant)
 
@itsbruce my solution is 18 lines of logic
The rest is types, comments, imports and blank lines
 
Meh, despite nhgrif's assertions.... .
3
Can't make heads or tails of that code.
 
@rolf some parts of it are easy type QueenBoard = [[Bool]] A board is a list of lists of booleans
count x = length . filter (==x) the count of an item is the length of items of the list equal to the item
 
@Caridorc Well, said like that, mine's down to 4 or 5. And I still wonder why you prefer to generate all boards when you can simply only generate valid ones. That can only ever be less efficient, even in a strict language.
 
11:58 AM
@itsbruce It is simpler to generate and then filter
and gneration is not totally random
I never put two queens in the same row
@itsbruce By the way allFalseOneTrue length = nub $ permutations ( [True] ++ (replicate (length - 1) False) ) would really benefit from avoiding that call to permutations
 
Well, the simple answer to the performance question is that you're generating a huge number of boards that never need to be considered and there's no way to fix that. Other than, for each queen placed, only generating the next set of legal moves. No mystery there.
 
that ^^^
Still brute force, but a more refined type of brute.... a brute with a glass of wine, not a beer.
4
 
@itsbruce this looks harder, place a queen, put another queen in legal position ... what if there are no legal positions left but you must still place one queen?
 
That can't happen, by the very definition of the n-queens problem
 
@itsbruce what if you plece the first (n -1) badly, leaving no space for the last one?
 
12:13 PM
Ah, talking at cross purposes
 
bye, see you later
 
@Caridorc then you either remove the last placed queen and try to place it elsewhere, or just return an empty set (depending on how your algorithm works)
 
12:30 PM
0
Q: Is there any design pattern for checking resources?

eurekaCurrently, I have a big for..loop with many if..else inside the loop. I just wonder is there any design pattern or other approach to reduce the long code inside the for..loop. Thank in advance. foreach ($workingTimes as $hour => $minutes) { foreach ($minutes as $shift) { $isBookable ...

 
Zak
Ah, this is a nice feeling. I've done enough "Please merge these 2 sets of data together" things now for work that I can write a new macro and have it run completely, first time, pretty much on auto-pilot.
 
@SimonForsberg interesting
 
What Simon said.
 
0
Q: Haskell Applicative idiom?

raxacoricofallapatoriusI'm new to Haskell and am puzzling over how to best express some operations in the most idiomatic and clear way. Currently (there will be more to come) I'm puzzling over <*> (I'm not even sure what to call that). For example, if I have, say f = (^2) g = (+10) as representative functions (in p...

 
12:45 PM
In a functional language, you mostly don't need "What if x isn't the result I want?". You just iterate over the results you do want. Which is why a functional solution to n-queens is elegant.
But without laziness, properly applied, the functional solution generates a huge number of cases that you don't want. You're not being lazy enough, @Caridorc ;)
 
Seems like I would have no problem using functional languages then
 
1:00 PM
@Jamal @200_success @Duga has been updated, should be much less false positives now.
 
Zak
@itsbruce properly applied laziness is a force to be reckoned with
 
Greetings, Programs.
 
Man, that question last night got a lot of good answers so far.
 
Oh, huh. I've impacted ~15k people on CR, but only ~6k on SO. I hope it was a good impact for people.
 
1:16 PM
@SuperBiasedMan Mine is ~10k on CR and ~5k on SO.
@Hosch250 Feature request from discussion last night. Do note, it's not actually a feature request yet, I'm still putting bits and pieces in.
 
What's funny then is I have ~5k on metaSO. I'm comparatively pretty low impact on SO.
 
I also have ~1k on Lifehacks for this answer.
And ~1k on Database Administrators mostly for this question.
I should pimp some of my Database Administrators answers at some point.
 
1
Q: PDO Insert function using PHP

ChrisI have the following PHP function: function insertTest(){ global $db; $field_names = array("id", "first_name", "last_name", "email"); $alias = array(":id", ":fn", ":ln", ":em"); $assignment = array(":id" => "2k39dk38dk2k39dk38dk2k39dk38dk2k39dk38dk", ...

 
Also, bringing this question on SO meta back up, as it's even more relevant now.
 
1:25 PM
I'm largely just on SO and CR. It feels harder to get into the other sites.
 
1
Q: Haskell Applicative idiom?

raxacoricofallapatoriusI'm new to Haskell and am puzzling over how to best express some operations in the most idiomatic and clear way. Currently (there will be more to come) I'm puzzling over <*> (I'm not even sure what to call that). For example, if I have, say f = (^2) g = (+10) as representative functions (in p...

 
I use Database Administrators occasionally, I don't often post there but I do read a lot of questions/answers there.
 
Is this even a CR question? Really seems to be SO or Programmers
 
Though Programmers is pretty useless in my opinion.
 
@itsbruce Looks off topic for being not real code.
 
1:27 PM
@itsbruce At best it's barely on-topic, at worst it's off-topic for "stub code".
 
@EBrown I think it's at worst. Code that works is not the same as real code.
 
@itsbruce I'm writing a custom VTC reason now.
 
And the question explicitly asks for comment on the general use of a technique, with the code only given as an example for the purpose of that discussion
 
Why custom? Doesn't example/stub code cover it?
 
@SuperBiasedMan I use custom reasons when the built-in ones don't quite fit in completely.
Rather than VTC for "example/stub" and then leave a comment explaining more, I'd prefer the VTC reason to have exactly why I VTC'd it.
 
1:30 PM
@Duga that's the spirit!
 
That makes sense. I guess I don't use the custom reasons just because it feels like they're supposed to be special. Like custom flags for mods.
 
@SuperBiasedMan I used "example/stub code" this time, but usually I prefer custom reasons for these situations.
Also, my Repetitive Console Prompt is on the Stack Exchange front page.
 
If this code actually compiles, I would recommend asking on codereview.stackexchange.com, as the possible answers are probably too broad/opinion based for here. — Andy Brown 53 secs ago
@AndyBrown This isn't really on topic for Code Review as it looks like the post just contains example code. Even if that compiles and runs it's off topic as CR is for reviewing real code. — SuperBiasedMan 15 secs ago
 
@itsbruce: God! The things I post there, get moved here, and the things I post here get moved there. Can we just get rid of Code Review? — raxacoricofallapatorius 2 mins ago
 
Hah!
 
1:37 PM
That sounds like a Dr Seuss line to me.
"I cannot post them here or there, I cannot post them anywhere!"
 
Was about to post that here ;)
 
DECLARE @MedInfoNoteTypeId	int = 25100 /* SELECT * FROM NOTE_TYPE WHERE nt_desc LIKE '%ESI%' AND nt_desc LIKE '%MED%'  */
 
Anyone see a problem with the above query ;x
 
@Phrancis No? Unless ESI and MED cannot appear in nt_desc together.
Also, no semicolon after int = 25100.
 
1:43 PM
DECLARE @MedInfoNoteTypeId int = (SELECT * FROM NOTE_TYPE WHERE nt_desc LIKE '%ESI%' AND nt_desc LIKE '%MED%')
^ That would make more sense :)
 
@SimonForsberg I had not, I'll read it now. Thanks!
 
This Mandelbrot generator works really nicely.
 
0
Q: Coredata delete all data in an entity

MwcsMacI have this code that I am use to delete all records from an entity. The question is can it be done better or is already the best? func deleteIncidents() { let appDel: AppDelegate = UIApplication.sharedApplication().delegate as! AppDelegate let context: NSManagedObjectContext = appD...

 
@CaptainObvious "The question is can it be done better or is already the best?" I like that way of framing it
 
@SimonForsberg - Here for a sec (short on time).
 
1:50 PM
@EBrown I gotta check that out just so that I can understand what Mandelbrot is
3
@raxacoricofallapatorius Can you join me for a chat in The 2nd Monitor, Code Review General Chatroom so that I can help you sort things out? — Simon Forsberg ♦ 1 min ago
 
@SimonForsberg It's really fun! :)
 
@SimonForsberg Here.
 
Hey @raxacoricofallapatorius, thanks for dropping by.
The issue with your question is that it looks like example code.
it might perhaps be a small piece of code from a real project of yours, but there's no way for us to tell if that's true or not.
 
@SimonForsberg - It's true.
 
On Code Review, we like context. Your question lacks that a bit.
> (in practice they are more complex, but the key thing here is that they are different and distinct)
> concat $ [(f <$>),(g <$>).tail.reverse] <*> [[1,2,3,4,10]]
it would be helpful if you provided just a bit of information about why you are doing this in your real project
 
1:55 PM
 
@SimonForsberg: But that sums it up. There a are several places in my code where something similar occurs. But they all follow the same pattern. Its an MWE (which we've all been trained by SE to generate).
 
"in practice they are more complex" tells us that you stripped down the code just to share it on Code Review.
Stack Overflow likes minimal examples, Code Review does not.
Have you seen this post on our meta?
46
A: A guide to Code Review for Stack Overflow users

durron597What distinguishes Code Review from Stack Overflow? Different communities have different rules to address different needs. A quick summary: $$ \begin{array}{l|l} \textrm{Code Review} & \textrm{Stack Overflow} \\ \hline \textrm{Open-ended questions} & \textrm{Questions with a specific goal} \\ ...

We often respond to Stack Overflow users who recommend people to post on Code Review by referring to that post
@Duga that edit seems fine
 
@SimonForsberg - Well migrate it and see what happens then. I'm happy with that distinction, but the questions I looked at here to see if mine was appropriate are indistinguishable from what you describe as not appropriate here.
 
You would have more success posting this on codereview.stackexchange.comJeff Noel 56 secs ago
 
@raxacoricofallapatorius I am not confident enough that it would fare well on Stack Overflow in order to migrate it.
overall, I am not a big fan of the migration system, very few moderators are.
 
2:15 PM
@raxacoricofallapatorius This site is for reviewing working code. There needs to be enough context to make a judgement about how appropriate any one method is to the solution
Your question simply had too little context. Almost no context at all. And the primary purpose seemed to be for general discussion of a technique rather than improving a specific piece of code.
 
@SimonForsberg But I – and likely many other Stack Overflow users – are not aware that CodeReview has that requirement. — Richard 20 secs ago
 
Even on StackOverflow - which is well supplied with skilled Haskellers - useful answers would be difficult without more context for your intent. Without that, it's just too open-ended and undefined.
 
@itsbruce @EBrown I got this, thanks
 
@SimonForsberg Fine, I've deleted my comment.
 
@SimonForsberg Not a fan of migration. I never flag for migration.
But some sites are better for some questions, obviously, and telling the OP that gives them the chance to reconsider their question in the terms of that other site. Of course, they may spurn that chance and just repost verbatim.
 
2:22 PM
I don't think it's good that one CR user comments, an SO user replies directly to the CR user, and then two-five other CR users starts replying.
5
 
Aye. I was referring to the earlier conversation, not that second one.
Deleted my comment on the SQL question.
 
That's fine, I'll not bother anymore.
 
I have been considering improving @Duga's filter so that she doesn't post multiple comments from the same question. I think that's a contributing reason to why CR users continue to comment.
 
Does anyone know if there is an easy way to make a new class template in Visual Studio, but only for a specific project?
 
0
Q: AngularJs Directive for common form fields

Matthew FedakI have created a directive for common form fields shared across all the forms in a large AngularJS app I am re-factoring. The app has lots of forms for various data models (user, owners, victims, suppliers, etc). A lot of the forms for these data models contain an address section which is the s...

 
2:36 PM
Is it just me or MongoDB == JSON?
 
@Phrancis Yeah, and that's awesome!!! :p
(Too hyped about NoSql and Json)
 
Oh OK, I don't know why, but I was expecting some "special" data structures or some such thing
 
0
Q: Kind request on code opinion on the Dependency Injection Principle

CeldorI have finished my first project using ASP.NET.MVC. i did not have to wait long to realise I have essentially violate all possible codes of good programming. Well, experience comes with practice. One of the styles I completely missed was Dependency Injection Principle. It was far too late I coul...

 
@Phrancis Well it is special-ish. The idea is to keep every entity as some sort of independant "document".
 
@TopinFrassi Can it store other kinds of data, like files, blobs, tables (financial records and things like that)?
(maybe "tuples" or "arrays" would make more sense than tables in this case)
 
2:51 PM
MongoDB can store anything you can express in JSON
As a substitute for a database, though, it requires you to massively denormalise your data. Lots of duplication
 
@SimonForsberg Perhaps we should only 'appoint' one person on a comment series at once to prevent 3 users hitting it at the same time.
Not sure how that would work though, just a thought.
 
What MongoDB adds to JSON is indexes for fast queries and clustering/replication for resilience
But it is overhyped. There's a price to pay for the simplicity. You should know what you're giving up by tossing away your RDBMS. Too many don't.
 
@itsbruce Enlighten us, what price do you pay when using MongoDB?
 
Once you have this working, I would post it in codereview, I think you'll get a lot of improvement suggestions. — Andy Brown 6 secs ago
 
@Mast The cost imposed by denormalisation, for a start. Potentially very large expansion of the data size.
MongoDB is also slow for updates (rather than simple appending to tables). Very painful if you ever need to update something that would have been one record in an RDBMS table before it was denormalised. But is now duplicated over thousands of "rows"
 
2:58 PM
That may be a problem indeed.
 
The security is also very weak compared to traditional RDBMS
 
Greetings
 
I guess I'll add Boost.Sort to cpp-sort too. The spreadsort is interesting. In the end, I'll have stolen sorting algorithm implementations from at least 4 projects.
 
I've remembered something
 
@Morwenn Nothing wrong with that.
 
3:01 PM
We haven't put up the Community Challenge for this month
 
And while MongoDB clustering is easy copared to RDBMS clustering, that's at (amonst other costs) the price of eventual consistency. Your application has to be designed to cope with that.
 
0
Q: Count words that are longer then a set Length

RiePweI made this functional but not want to know is this a place where IEnumerable can be used ? ( like what else can be done here ?) public static int LongerThanStingCompair(string s, int n) { int counter = 0; // splits based on nextLine break var sCount = s.Split('\n')...

0
Q: getting correct objects from string parameter in factory

Abdul AhmadI've been trying to figure out if theres a way to pass a string to a factory or constructor and create the correct object without having to map the string to an object, or without having a bunch of if/else statements or switch statements. Keep in mind, this is a simple example so I can apply wha...

 
@Mast As long as the licenses are compatible :D
 
@IsmaelMiguel No, we haven't.
 
It should have been put at featured since day 1
 
3:02 PM
@IsmaelMiguel There's no need to either. I'll probably post one for November .
 
(Or how those are handled)
Why no need?
 
@IsmaelMiguel There is no obligation for a community-challenge to be posted every month.
2
Give everyone a chance to keep up.
Feel free to post a question on an old challenge.
UTTT for example.
 
@Mast yeah, I don't think it's practically possible to do that.
 
@Mast I'm not saying every month
I'm saying that this month has been decided a challenge
And it isn't featured
 
@IsmaelMiguel Implement an elevator, which is this month's current challenge. So far there has been no submissions
 
3:04 PM
@IsmaelMiguel Ah, can anyone stick a on it?
 
@IsmaelMiguel the featured tag is usually removed when the challenge starts. it was featured last month
but oh well, me fix
 
@Phrancis I'm not sure :/ I didn't get to work with MongoDb, but I did with other NoSql storage
 
That was what I was talking about
 
fixed
17
Q: September 2015 Community Challenge

Gareth ReesIt's time to choose a community-challenge for September 2015. Post your challenge as an answer to this question. Feel free to resubmit non-winning ideas from previous months. Vote for those answers which interest you. At the end-of-day on Monday, August 31st, the top-voted post will become the ...

 
@SimonForsberg Lacking a more current challenge question, I think it's appropriate.
 
3:06 PM
Ok, so now I can see where your error is. I can now see it fail, and then a fix to make your unit test pass so I have posted an answer. Let me know if that doesn't fix it for you. You can delete all your comments in response to mine (which I have removed) and I have retracted my close vote and down-vote. Once you have this working, I would post it in codereview, I think you'll get a lot of improvement suggestions. — Andy Brown 5 secs ago
 
I'm actually in the process of putting a submission to that challenge now. :P
 
@Mast true. I also noticed that that question hasn't been featured before.
@SuperBiasedMan good!
 
I think I'll drop my current version. Still don't like C#...
 
@Mast :(
 
Anyway, I'll fly tomorrow and won't be back for a while, so perhaps it's getting in late.
 
3:07 PM
@itsbruce That might because your data isn't fit to go in a NoSql DBM, because there are situations where data don't need to be replicated and everything works pretty fine!
 
Zak
3:25 PM
This is encouraging. Work have asked me to design and implement some form of software development framework for my workload :)
So, since I get to build it from scratch, and learn some habits more or less from scratch, what does the perfect-world process of software development look like?
 
@Zak somebody else to bump your ideas off of.
also... version control
and clearly outlined requirements
proper documentation and a document that specifies reqirements that is not code..
in the best case also versioned...
 
Oh, yeah
Delimiting requirements at the very beginning is really important
Listen to Vogel, he's right
 
Monking :-) I wanted to learn Swift, but then I found out the only way to compile it is on Mac. Talk about cutting out a userbase :-(
 
and for christ's sake don't implement stuff that's not in the requirements
3
 
@Zak version control (with semver)
automated builds with various forms of automated tests (unit, e2e, integration)
 
3:31 PM
it's the road to madness, where you have to dispute and check back with what is actually implemented against what the customer thinks he wants now and what he thinks he wanted back then
 
@Vogel612 Everyone needs a rubberduck.
 
@ARedHerring Well, you can install a MacVM
 
horizontal scaling over vertical scaling.
and yes, as @Vogel612 said: implement what's required, nothing more.
 
@ARedHerring Apple in a nutshell.
 
implement the bare minimum because you should be iterating on your development with the customer
focus on conversations instead of large requirements docs.
@IsmaelMiguel $$$
 
3:32 PM
@Zak Don't confuse requirements that are clear and precise from requirements that are actually correct. People writing the requirements aren't necessarily the people who actually know what is needed.
4
 
@ARedHerring Virtual machines are free
 
@Zak Validate everything against the requirements, validate often. Attack pieces of software small part by small part
 
@IsmaelMiguel The ones I can find aren't
My google-fu might not be up to scratch, though.
the tldr is basically the UNIX philosophy: do one thing very well, and make that one thing a small part of the overall functionality. Your application should be composed of many Small Things, rather than one Large Thing
 
@IsmaelMiguel It's kind of complicated to get a Mac's VM
 
@ARedHerring Well. You can download virtualbox, and find a torrent with the dmg file used. Then you will need an iso to boot the dmg file
@TopinFrassi Kind of? It's a gigantic PITA!
 
3:33 PM
@IsmaelMiguel I have virtualbox + vagrant, but it would appear that windows 8 doesn't play well with virtualbox because of the way vritualbox does networking.
 
@Donald.McLean Spoken like a man with experience in that I presume ;)
 
The virtual machines work, but the only way they work is if I switch off all networking outside of virtualbox = no internet.
 
why would you even need a Mac VM??
 
But either you pay 1200€ or you spend hours on it
 
@Vogel612 I want to learn to develop Swift. It doesn't compile on Windows.
 
3:34 PM
@ARedHerring You should set the cards as bridged
 
I appreciate all apps for Swift are iOS only, but I would have liked to have tried out the language before dropping £2k on it.
 
isn't there some online sites for that @nhgrif ?
 
@skiwi I've been working as a paid professional developer for 27 years, and every project I've worked on has delivered. Some of those deliveries have gone better than others...
 
Zak
@Donald.McLean Cool. Do you know anything about organisational-level software development? Obviously, I'm quite a way away from that, but I'd like to be setting up good habits and practices.
 
@Zak source control
and automated testing
 
3:35 PM
and documentation
and requirements
 
you have to make those two very close to your core values as an organisation
 
But, not too much documentation. :p
 
if you don't make them core values, no one will respect them, and no one will do them.
no one is going to do automated tests, or source control - or documentation - if you don't make them feel like that is what your organisation is "about"
(obviously, i'm no runner of an organisation, but that's how I see things from an open source point of view - coughs in cardshifter's HTML client's general direction ;-))
 
@ARedHerring Like having a printer print out documentation every time the automated tests have been run every time a commit has been made?
 
> Like having a printer print out documentation every time the automated tests have been run
So much for saving the planet
 
3:38 PM
Save the coders, not the planet
 
@ARedHerring this is 100000% absolutely true -- and I do "run" an organization.
 
There is not going to be enough paper in the world to make a print-out every time you run automated tests
 
@Zak 1. Communication between your team; 2. Communication with your users; 3. Source code control; 4. Automated testing; 5. Pair program if possible. Documentation can be useful, but only if PROPERLY MAINTAINED and even then they can be ruinously expensive. Requirements are great things, if they are actually correct, but the only way to know if they're correct is by talking with the users.
 
2
Q: Removing this odd while(true) for checking the queue

Laurent TrentinI was reviewing some WPF cone I've written in C# last year. The scenario is the following: I've got a usercontrol that show usermessages and it's docked in the bottom of the application's workspace I've got a ConcurrentQueue that receives message via a IMessageMediator (from Catel, but's irril...

1
Q: Elevator Management System

SuperBiasedManSo here's my attempt at the September community challenge, making an elevator handling system. I'm not really very familiar with OOP, so I'm particularly looking to get feedback on that part of the script even if you're unfamiliar with Python. I'm fairly certain there's times I've mishandled abst...

-1
Q: infix to postfix converter, not outputting right answer

Marco KoivistoI got a school-project where I was given this information: for i=1 to m if c_i is an operand: Transfer c_i to output. if c_i is a left parentheses: Push c_i to tmp. if c_i is a right parentheses: Pop elements from tmp and transfer them to output until a left-parentheses is met. Pop left-parenthe...

0
Q: Optimise Code for KukuKube Game

huynhtridungI developed a Kuku Kube game for iOS. You can download and test here https://github.com/dunghuynh1990/KukuKube-Advanced.git You can review and give some comments about the code, how to optimise or better way of code to use the minimum RAM on iOS device. Currently it takes nearly 30MB for the app...

 
OK, who got me playing with mandelbrots?
 
3:43 PM
@ARedHerring the example code is all .NET / C# specific but the concepts should apply pretty generically
 
@MikeEdenfield Ah okay, I've read that. It's not for me, but more for my colleagues
 
@rolfl EBrown and Caridorc, I'd guess
 
user image
2
 
@rolfl Me. :)
 
@rolfl Nice
 
3:44 PM
@rolfl You should increase the number of colours and blend them better.
I'm working on adding parameters to mine so that you can specify a specific region to render as well.
 
@Duga you got a new feature?
 
@ARedHerring She's going to replace the moderators one day
(In reality: Don't try to do that, some SE site tried once and it got really bad)
 
@Duga editing code, invalidating answer possibly
 
@EBrown Perhaps, but 0.4 seconds for a 2000 limit computation on a 1000 pixel wide image is enough ;-)
 
3:46 PM
@Duga false positive.
 
@Duga The only answer there doesn't touch anything important
 
@skiwi hmm?? which SE site tried that??
 
Who comes back to an answer a year later to add if (foo != bar)???
 
@Vogel612 There's a quite big meta post around somewhere... I don't remember which one though :(
 
s/answer/question/
 
3:47 PM
@rolfl I'm not sure I get what you mean? How long does it take to render 4096x2048 at 1000 iterations?
 
@skiwi are you talking about the comment flag bot?
 
@Vogel612 I think it was a moderator that wrote a bot to automatically close questions with some close reason if the question had some keywords in it
 
@Vogel612 @Duga just says possible answer invalidation. She did right in informing us about this one (therefore it was not a false positive), but it was indeed no answer invalidation.
 
@EBrown 2.9 seconds
 
@rolfl I think that's about how long mine takes.
 
3:48 PM
@ARedHerring yeah, she got that feature last night. see chat.stackexchange.com/transcript/16134?m=23964936#23964936
 
@SimonForsberg hmm.. yea it was a possibly invalidating edit, but if it weren't I'd have said negative :DS
 
@EBrown Sorry, 1.1 seconds, I was running a 2X zoom level.
 
@rolfl What does the generated image look like?
Out of curiosity.
 
 
Ah, you centred yours better than mine.
@rolfl C# or ?
 
3:49 PM
@EBrown Java.
 
That one takes ~2.4s on my PC.
 
Which algorithms are you guys implementing for the Mandelbrots?
 
I'll start moving messages...
 
I'm thinking of posting my code.
 
Or 1.9s if I use all 6 cores (instead of 5)
 
3:51 PM
Good guy @Vogel612
 
looks like a lost cause...
 
@Vogel612 Move one to Nth and one to ____ (or leave it).
@skiwi Escape Time
 
Yup... seen that ... thought "I think I can do better" ;-)
 
0
Q: Gitlab: commenting on full file rather than on diff?

VinceMy understanding is that gitlab allows to comment on code only in the context of: merge requests when looking at commits In both case we can only comment on diffs between two commits. Things are developed here. This makes total sense in the context of continuous integration. Yet, on some occ...

 
3:55 PM
hrmph... it's a pain to sort out 2 conversations + sidebusiness...
 
Zak
@skiwi I think it was over on Database.se
 
@Vogel612 Not to be a pain but I don't think my message was a part of that conversation haha (chat.stackexchange.com/transcript/message/23983345#23983345)
 
@SuperBiasedMan Collateral
 
@Mast Some people just slip through the cracks in the system...
 
@SuperBiasedMan whoops.. well it's kinda hard... 2 conversations, interspersed useful stuff & side business... next time you can try sift through that...
anyways it would be nice to allow for a consistent standard when non-site business should start be moved...
 
3:59 PM
@Vogel612 I mentioned a few minutes prior that i felt the conversation was getting lengthy
 
I think 10 minutes of active conversation is too late already, but that may be a little harsh..
 
I think it's just a gut feeling thing
because some conversation might be lengthy but also relevant to the site (but then maybe it should be meta)
 
we were at like 20 minutes when I started moving, and that was definitely too late.. and took too long.
right, site-business shouldn't be moved away anyways.
 

« first day (858 days earlier)      last day (2532 days later) »