« first day (1727 days earlier)      last day (3262 days later) » 
00:00 - 18:0018:00 - 00:00

6:01 PM
o_o
 
that's very close to the types of things I like doing nearly entirely across the posting
and I actually have relevant background/etc
 
user55340
37
Q: Better "Flag for migration" interface

MichaelTCurrently, the only way to ask for something to be migrated to a site that isn't part of the standard migration path is through a custom mod note. I currently have two active custom mod notes on Stack Overflow that ask for migration that are not standard targets for SO. I believe the questions ...

 
@MichaelT That's not the one I was thinking of.
It explicitly talked about a check box and a message pointing people to the other site's Help Center.
 
@enderland That is an interesting job post, but won't it end someone's developer career if they're currently a developer?
 
@RobertHarvey perhaps
 
6:07 PM
@RobertHarvey Maybe?
 
I mean, yes it probably would if you want to be in an IDE all day every day
 
I'm at the point now where I don't think I'll even take another obsolete technology programming job again. It's just too hard to prove that you're current.
 
That is a big fear of mine. As much as I love doing things like managing projects or doing process improvement, if I leave the standard SDLC-type requirements, architecture, design, code, test type work, I'll lose effectiveness at managing projects or doing process improvement.
 
No. You won't ever become ineffective at those things. You'll just become obsolete in the eyes of the employer community.
 
@RobertHarvey replied
 
6:10 PM
@RobertHarvey I think I would eventually become ineffective. I'd lose touch with how work happens. It doesn't matter if I write an amazing process, if it doesn't actually work to deliver value...well, it's pointless.
 
@ThomasOwens so my background is in HCI which has a lot of techniques to help you understand what the needs are - this sort of approach to processes is very helpful, too (not just UI design)
 
Ah, c'mon.
-5
Q: Any c# king here? Help!

lol and wow Following is part of my c code(where wrong? Can u help me? :) ): using System; using System.Collections.Generic; using System.Linq; using System.Text; using System.Text.RegularExpressions; public class PinYin { private static string[] getValue = new string[] { ...

 
@enderland Apply for that job.
You'd be better than me at it.
 
@ThomasOwens I'm pretty seriously thinking about it - I'm not sure how long it's been up, but it's absolutely perked my interest and my experience here on SE is actually quite relevant to that job, too
 
Go for it.
 
6:17 PM
Yeah, you should apply.
I'm a good coder, and not a half-bad designer, but I'm completely bereft of product ideas.
 
@RobertHarvey the trick with all UX types of things (or even ideas, in this context) is realizing you don't have to be, as long as you have the ability to understand the needs and pain points of your users - then you just solve them and have now an endless litany of "product ideas"
 
Woo, laptop but no login.
 
@Telastyn Try 12345
 
or "password"
 
6:20 PM
if it's windows try the admin account
 
user41796
@RobertHarvey - did you get an answer to your question in here about globalant and the terms of their offer?
 
I didn't.
 
user41796
Sounded to me like they'll put you in their system but you don't get paid until a client accepts you to work on their projects
 
user41796
So you're "employed" but you aren't necessarily paid.
 
Well, I don't know my username. Password is likely set on first login deal.
 
user41796
6:22 PM
Even with the big consulting firms, some of their clients want to interview the contractor that will end up doing the work
 
Their offer was for an exempt employee. I don't see how they could get away with that. It's not like you can just turn it on and off like a hose.
 
user41796
Benefit of the big firms is they'll pay you even if you aren't actively billing to a client
 
user41796
@RobertHarvey That's not how I read it then. Weird
 
So they make it contingent on interview with the first contractor, then it's salaried after that?
 
user41796
It sounded like they were pushing the risk of you not having a contract to work on (and therefore get paid) fully on to you
 
user41796
6:24 PM
@RobertHarvey Yeah, sounds like it
 
user41796
Being cynical, that likely means you're laid off if the contract you're working on collapses
 
That would seem to be the case.
I checked them out on Glassdoor. The consensus seem to be "Good work, but terrible management." They're probably a pain in the ass to work for, but it's not like I've never done that before.
 
user41796
Almost sounds like a variation on you being a subcontractor to the primary contractor
 
user41796
If it's positive cash flow, that's a good thing. And I would certainly keep looking for something while doing the contract work
 
Well, I guess their safety valve is me interviewing with the client first. The offer is contingent on the interview, so.
 
user41796
6:26 PM
Even when I was in a contract-for-hire position, I kept interviewing in order to make sure I had a backup plan
 
Oh, of course.
 
user41796
Yeah, sounds like they don't start paying you until you're billable (aka client interview is successful). So they've minimized their upfront liability.
 
I need to do something to get some of these folks off the stick. They all like me and my credentials, but the process always seems to stall for whatever reason. Everyone wants to hire someone, but nobody wants to take the risk.
 
@GlenH7 thanks for that deletion
 
user41796
Probably wouldn't hurt to ask them what their notification period is if the contract ends
 
user41796
6:28 PM
@RobertHarvey For that alone, it's likely worthwhile
 
user41796
@durron597 np. Wasn't that good of a question or answer.
 
@RobertHarvey You could always try offering a short period where you work at reduced salary
 
user41796
roomba should clean it out tonight
 
@GlenH7 It's roomba eligible now... yeah, I didn't want to downvote your good answer to a bad question
 
user41796
@durron597 Many HR departments aren't well setup for the concept of a contract-to-hire
 
6:29 PM
@durron597 I might do that if I really like a particular opportunity, and I can't get them off the fence.
 
"Sorry, we can't create your account, the automated process needs to handle it tonight."
 
"why didn't you create it a week ago?"
 
Well then, I'm going home to "work".
 
@RobertHarvey "Job is paying $75/hr? Okay, pay me $20/hr for one week. If you still like me, we'll keep going at $75/hr for next 6 months"
or something.
 
6:30 PM
They needed my i9 info before they could give me an employee number, and the automated process needs the employee number...
 
@ThomasOwens He's tried that, as far as I know he hasn't gotten a response
 
Stack Exchange tl;dr: "Come back in a year and reapply after you've done some more frontend web development work."
They were very nice about it, though.
 
@RobertHarvey You could probably come back faster than a year if you make a portfolio
 
Ah. Yeah. I haven't even seen a line of HTML/CSS/JS in like...6 or 7 years.
Probably longer.
 
6:32 PM
It's been five years for me. Ergo, the comments about allowing yourself to become obsolete.
ASP.NET MVC work has changed completely. When people say "Full Stack," what they really mean is browser applications with web services.
So ASP.NET MVC has been relegated to the role of creating web services for the browser applications.
 
user41796
@RobertHarvey That's pretty fair. Most CRUD type apps have minimal SQL backing them.
 
my (now somewhat dated) "about" on my careers profile:
> The longer version is, my career interest is more or less a position bridging the gap between software developers and non-developers, given I have a background in development/HCI and nearly always there is significant need for someone interested to facilitate both work together with that background. Most companies however do not seem to find developers interested in such work.
> This results in all sorts of problems, especially given how much easier creating requirements is when you can quickly prototype systems to help users determine their needs. This of course results in a much shorter overall development time and cost when done appropriately.
 
Sounds like you'd be perfect for the product manager SE position.
 
Yeah. Update your Careers profile and apply.
DO IT, @enderland!
 
@enderland Why? Are you not liking your new job?
 
6:37 PM
I'll be sad though if it gets closed before then
@durron597 I'm always open to new opportunities
 
@enderland Still, you just started a new role.
 
@durron597 I expect by the time this process gets anywhere it'll be several months at lesat
 
This question would be quickly closed and likely down voted on Programmers. At it's core, the question is a poll which doesn't work well within the StackExchange Q&A format. — GlenH7 36 secs ago
 
@enderland true
 
6:38 PM
I'm tentatively optimistic, still, the tech stack is boring I think
 
I like the team and work environment, mostly, and keep in mind just interviewing doesn't mean anything about actually taking the job
that description is totally vague
 
(hint: it's a cross post)
 
user41796
@durron597 hint it's already closed
 
@GlenH7 lol
 
6:40 PM
plus I like interviewing and applying for jobs, generally
 
I like being flogged.
 
user41796
Interviewing when you're coming from a position of leverage (aka already have a job and don't necessarily need a new one) is a much different experience than doing so without leverage.
 
@GlenH7 oh man, tell me about it.
 
user41796
Have they gotten back to you?
 
@GlenH7 I should know by the end of the week
 
user41796
6:43 PM
Very cool
 
But even then, it's not the next step, the next step is meeting teams and getting placed
Even if I get a thumbs up I expect I wouldn't start until like, the fall
 
user41796
Are the additional rounds more focused on team fit or technical ability? Or both?
 
@GlenH7 The former
 
@durron597 Did you say who this was with?
 
I think 6 technical interviews is enough, even for Google
 
6:45 PM
@GlenH7 I told my wife worst that happens is they say no, she responded with "no worst is if you get an offer you can't decide if it's worth leaving for" :P
 
@ThomasOwens ^^^
 
@GlenH7 fit, pah.. you should disabuse others of this excuse for H1Bing us all :P
 
@durron597 Oh. It's for Google? Which office?
 
1 min ago, by durron597
But even then, it's not the next step, the next step is meeting teams and getting placed
 
user41796
@enderland Yeah, I've been in that position before
 
6:46 PM
In other words, I have no idea yet
 
user41796
They didn't make the offer sweet enough to offset the 100% travel. In retrospect, definitely the right decision
 
@GlenH7 This was how I unexpectedly ended up leaving a company after 2 months...only did that once, but I left for Microsoft, so it was worth it.
 
course this is putting a lot of things ahead of the conversational horse, so to speak
 
@GlenH7 A guy I play tennis with recently got a new job at a company based in Chicago... he still lives here in Houston. So he works from home... when he's not travelling. He buys and manages property for a large company
so he's travelling almost every week
 
user41796
@JimmyHoffa one-offs don't hurt your resume because you can explain them. Lots of them demonstrate a pattern
 
6:47 PM
the first week, he was like "yes! I get to work from home! I get to take my laptop to the beach!"
 
user41796
@durron597 Lived that life for a while; it's rough
 
now he's like "I never work out anymore, I drink too much (with clients) and feel like shit all the time"
 
@GlenH7 I sincerely believe you suffer from confirmation bias and judge others poorly because of it.
 
@JimmyHoffa ?
 
Not atypical, success breeds success and before you know it all the succesful people are managers and when they look out at folks who've been in bad situations - they can't relate, and decide it must be that person's fault they weren't succesful
 
6:49 PM
I had a 15% travel job when I finished school. Travelling is fun, at first... but going out with your alcoholic clients when you have to actually get stuff done the next day SUCKS.
 
user41796
@JimmyHoffa Could very well be.
 
user41796
@JimmyHoffa I've been in plenty of bad situations too
 
@GlenH7 and it doesn't demonstrate a pattern?
 
user41796
But when I've been in 10 bad situations, there's a single common factor - me.
 
this is why there's a belief that "You've worked in lots of places" equates to "You messed up", as opposed to the very logical reality that perhaps "You ended up in some terrible places" - which is logical as a majority of this industry is pretty terrible.
 
user41796
6:50 PM
@durron597 I was never obligated to after hours with the clients, fortunately
 
user41796
@JimmyHoffa You misread then
 
user41796
"You've worked in lots of places" equates to "you're likely to move on again in a short period of time."
 
@GlenH7 negative
 
@GlenH7 Not only was I obligated, but I was the one who had to do the taking out, so I had to do complicated expense reports when I got back if I wanted to get reimbursed
 
@GlenH7 it equates to "You may have worked some places that were truly terrible"
not always, but often
 
6:51 PM
@durron597 I think part of the thing is that I am not super excited to come to work every day at this point. It's a job to me, it's probably a good job, but it's not something I am super passionate about at all
 
and that's the confirmation bias I refer to
 
user41796
And when I weigh the potential return on invested training, I have to guess at how long they'll stick around
 
@enderland I hear you brotha
 
@GlenH7 see that's dehumanizing; you should think of how you can motivate them to stick around if you want them. Not what is their carcass going to be worth to you.
 
so it's not that I hate my job or even dislike it, but it's not really a passion for me at all
part of that is inexperience
 
user41796
6:53 PM
@JimmyHoffa I'm an engineer. Making cold, misinterpreted as cruel, calculated decisions is what I do. :-)
 
@GlenH7 about people? That's a terrible way to deal with people- quite ineffective. People treated like people tend to be far easier to interact with, and sincerely more effective.
 
user41796
It was a joke....
 
@GlenH7 We should rename this room to
 
@durron597 of course, we're a bunch of humorless engineers
 
;P anyway, just saying. Nothing against you, I just think that's a confirmation bias from people who haven't been in lots of terrible positions. There's a number of them which all equate to "this person has had failures, it must be there fault, because I haven't had failures". Folks who've never worked in shops that were severely disorganized, or entirely legacy (this person doesn't know the newest thing, he's not keeping current like I always have!)
 
user41796
6:57 PM
@JimmyHoffa I hear what you're saying. And I'd point out that you've got a lot of projected feeling wrapped up within some of those statements. Granted, I'm biased, but I think my perspective is a lot more neutral than what you perceive.
 
user41796
I do know of folk who misinterpret past events and use that to pre-judge the candidate. But I also know of many folk who just look at things as data points and don't judge.
 
user55340
Inspired by a poor p.se question:
 
user41796
That latter group is more interested in predicting likely behavior and potential fit with the company culture.
 
user55340
0
Q: Expand an encoded string

MichaelTThere's the classic run length encoding and decoding. input output a3b2c5 aaabbccccc And that's fairly straight forward and done before. The challenge is to also account for a non-standard behavior when multiple characters precede the run length (a single digit from 0-9). Each character be...

 
7:01 PM
Oh, that's interesting
 
user55340
Heh. I know my regexen.
 
Hey, why is a123 aaa111222 not a12a12a12
can you clarify that bit?
 
user55340
3 is the final digit in the rle block.
 
Levels of regex competency: 1. Regex ninja, 2. Knowz teh regexen, 3. What's regex?
 
user55340
And gets applied to all characters in that block.
 
7:03 PM
@MichaelT right, but if you're repeating the string a12 three times, it would be a12a12a12
Your way is entirely valid, it's just unclear in the problem description outside of your example cases
 
user55340
No, this is the non standard part.
 
@GlenH7 understandable, but we all know past performance is no indicator of future results. So I find it a bit ignorant to try using that rule regardless of it's known failure, when you have other tools for assessing a person (talking to them for instance, like they're a person).
 
@JimmyHoffa it's not a causation of future results, but it definitely has correlation
 
@MichaelT Okay adding that word helps.
 
user55340
Yep @durron597 good edit.
 
7:06 PM
@gyoder this question is a very poor fit for Programmers - it would be quickly voted down and closed over there, see meta.programmers.stackexchange.com/questions/6483/… Recommended reading: What goes on Programmers.SE? A guide for Stack Overflowgnat 59 secs ago
 
@enderland yeah? So if I flip a coin fifty times and it lands on heads every time, I can correlate future results from those past performances?
 
user55340
And yea, it is interesting.
 
@JimmyHoffa coin flips are unrelated (other than if a coin is weighted) but people cannot equally separate themselves from their past life
 
@enderland perhaps they are quite learned from seeing so many ways to fail. Trying to simplify a person to a linear regression is not an effective technique for assessing them.
 
user55340
@durron597 I blame you. Used 30 delete votes and can't delete (vote 3) on the C# king question.
 
7:10 PM
@JimmyHoffa no but the problem isthere is no other way to evaluate someone
if someone has failed at projects for 20 years, I'm not going to expect them to start succeeding in the future
 
@enderland ...you could talk to them. Have them write code in front of you. Ask them about their experiences et al.
@enderland Employer^ everyone who worked there did that. And it was none of their fault. The company organized them for failure, and their clients were basically on a constant never-ending support call with them. The only successes they had was of their contract negotiators that ensured they were well funded regardless of their inability to accomplish anything at all.
 
@JimmyHoffa ... but then why did they not leave? it seems a company which sets me up to fail would be on the "leaving job ASAP" list for me
@JimmyHoffa I'm not saying you can categorically group people simply based on that sort of thing
but it can provide insight
 
@enderland many did, but for doing so many of them were looked upon poorly in their interviews and surely passed over for the failures there
@enderland yes, of course it can provide insight, just like horoscopes.
I think people confuse gut feeling with sense of reason a bit much in all of that; so it bugs me.
 
I don't think Glen (or myself) are saying, "your past dictates your future" at all
if you had a situation like that, remotely competent interviewers will want to know, what did you do to help your project succeed and what did you learn
 
@MichaelT Just give me a 6800 rep bounty
 
7:22 PM
Debugging bugs without debugger is hard
 
user55340
@durron597 4k is trusted user on code golf.
 
@MichaelT Hm?
 
beta site ftw!
 
user55340
Oh, here... Yea... No.
 
user55340
7:28 PM
You would also need another 10k to do speedy deletion.
 
@André windbg and code reasoning FTW!
 
@JimmyHoffa window.alert()!
 
0
Q: Fix the Braces, etc

durron597Fix the braces Your mission, should you choose to accept it, is to add the minimum number of parentheses, braces, and brackets to make a given string (containing only parentheses, braces, and brackets) have correct brace matching. Ties of symbols added must be broken by having the maximum distan...

 
@André ugh... yes, that is miserable. Why no debugger? Not used to using a remote debugger?
 
@JimmyHoffa GWT code transliterated to Javascript, legacy code doesn't support superdev that generates sourcemaps, nor old dev mode works.
 
7:30 PM
@durron597 what about [({}{)]
 
@JimmyHoffa [({}{})]
 
There isn't anything to debug, except robot-readable javascript
 
user55340
And @amon does deeper regexen wizardry than I dare.
 
user55340
1
A: Expand an encoded string

amonPerl/Bash 54 chars perl -pe's:([a-z]*\d*)(\d):"\$1=~s/./\$&x$2/egr":egie' It's basically a regex within a regex. And a bit of magic.

 
@durron597 it wasn't mentioned as valid is all; should be an example to clarify. I would have just split at the first closing brace and mirrored from there if it weren't for that scenario
 
7:38 PM
@JimmyHoffa That's not minimal
 
If those people didn't leave the poopy job, and also didn't manage to improve it (or learn to function in it) over the years are to be frowned upon.
Sorry, was afk, late to convo.
Because lots of jobs are bad, and you're setup for failure. Making small successes is enough to sell in an interview.
 
@durron597 ? is it a valid scenario that must be accounted for? Or?
@durron597 don't understand your comment about "minimal"
 
user55340
@durron597 I like how this challenge gets the cjam guys to complain about long solutions at 33 bytes.
 
also quite like the problem. Good one.
@durron597 is my test example of [({}{)] valid input?
 
@JimmyHoffa Yes
 
7:45 PM
ok
 
That is valid input, the only acceptable output is [({}{})].
@JimmyHoffa It's good, but it's basically a duplicate unfortunately :-/
 
user41796
8:07 PM
bye bye delete votes
 
@RoyalAirForce It is not bad or completely off-topic question, just that its abstract design level will probably be a better fit for programmers.stackexchange.com scope. — Eugene Podskal 55 secs ago
 
user15026
I start work tomorrow! I am so excited! (And also freaking out a little, but you know.)
 
@AshleyNunn Say hi to the cows for us.
2
Boy that overexchange is a persistent bugger.
 
user15026
I will pass along your greeting :P
 
user41796
@AshleyNunn At the ISP? Very cool!
 
8:11 PM
@AshleyNunn wow fast? that's awesome
 
user41796
@RobertHarvey There's a lot to be said for tenacity. Listening to the shared wisdom of others is good too
 
@AshleyNunn congratz!
 
@GlenH7 It's just funny how so many software developers will hang their hat on the correct definition of a word.
As if that really matters.
 
user41796
And given the widespread mild misuse of pretty much every term, it is a bit of a fool's errand to find the "one true definition."
 
user41796
8:20 PM
precious. my precious.
 
9:04 PM
What does it mean when a company asks for 5 years of code merging experience? Is this a company that has had trouble with their code merges?
Or 5 years of TFS. Is it really that difficult to learn another source control system?
 
may depend on what the role is, my last team lead did a lot of merging and knowing how to use mercurial or not would be a huge deal there
 
It doesn't sound like a very fun environment to work in. Who is checking out all of these classes and hacking on them simultaneously?
 
@RobertHarvey my team lead kept our branch up to date with all the other team's once they got integrated into the default and release branches
so whenever another team branch was merged, he'd merge those changes into our team branch
 
I worked at a place that had hella merge issues, mostly due to bad developers/code
 
Why are there multiple branches? Why isn't there just one production branch?
 
9:09 PM
Also, I hate how over exchange is trying to code via uml and keywords.
 
@RobertHarvey there was a single production branch, but when you have 200+ people you cna't have everyone pushing to production without some integration testing
the production branch had every other team's changesets merged in and tested/etc
 
@Telastyn He should write some actual code. Then he will be enlightened.
@enderland Ah, now I'm starting to see why companies have CI.
 
Well, he seems to write code, but then worries that it's not "right"
 
A lot of coders seem to worry about that. Being "correct."
 
But only a little code - then argues about it for a month.
 
9:11 PM
@RobertHarvey we had CI too, but when your software goes onto a fairly complex hardware system you need to validate it on the hardware system too - but no team ever did this on their branches actively (due to the amount of time it takes)
 
perhaps we should nudge him towards CodeReview.SE?
 
/me spits.
Correctness is defined in requirements and peer wtfs
 
anyways I'm out of here for the evening
 
G'eve. I will soon as well likely.
 
Is there a way to make this question not awful?
-2
Q: How to do ab testing

Andrew WaltersMany books on startups/kanban strongly advocate the use of AB testing to validate product features. I haven't had any experience with this but it sounds like a great idea for some projects. My question is, how does one go about AB feature testing in practice? Say you want to test how registe...

Some good answers
 
user55340
9:18 PM
The rle challenge is causing challenges for all languages without magic.
 
By magic, do you mean teh regexaz?
 
user55340
4
A: Expand an encoded string

amonPerl/Bash 54 chars perl -pe's:([a-z]*\d*)(\d):"\$1=~s/./\$&x$2/egr":egie' It's basically a regex within a regex. And a bit of magic.

 
user55340
3
A: Expand an encoded string

OptimizerCJam, 33 31 bytes Ughh, lack of regular expressions makes this pretty long... q{LA,s&\:LA,s&>{])~e*[}&L}%)~e* How it works We loop through all characters of the input string and in each iteration, keep track of the last encountered character (starting with an empty character for the first ti...

 
@MichaelT No one has answered mine yet.
 
9:23 PM
@durron597 Fundamentally, it's a product recommendation. Even the self-answer emphasizes that aspect.
 
user55340
Heh. Stupid numbers.
 
@RobertHarvey Downvoting all the answers will make it fit the roomba criteria
 
user55340
4
A: Expand an encoded string

kirbyfan64sosrs, 43 71 chars Dang, this turned long quickly. Stupid numbers... (\d)(\D)/\1 \2 +(\w)(\w+?)(\d)(?= |$)/\1\3 \2\3 (\w)(\d)/(\1)^^(\2) / Original version (did not work with input like 123): +(\D)(\D+)(\d)/\1\3\2\3 (\D)(\d)/(\1)^^(\2) Explanation The first line places spaces between runs, ...

 
9:38 PM
Error 403 - This web app is stopped.

The web app you have attempted to reach is currently stopped and does not accept any requests. Please try to reload the page or visit it again soon.

If you are the Administrator of this web app, please visit the Azure Portal to check why the app is stopped.
Hanselminutes.com
Hmm, looks like it has some sort of watchdog timer/ restart mechanism on it.
 
9:49 PM
14
Q: What should you leave behind for your successors?

Steve EversAssume that you're a sole developer leaving a job. What kind of information/material, outside of the code itself, should you create and leave behind for your replacement? An obvious answer is "whatever you would want at a new job" for sure, but it's been a while since I started a new job, and I ...

 
psr
@JimmyHoffa At the very least it indicates that you leave shitty places, which should be a red flag for shitty places that want you to stay for a long time.
 
10:19 PM
It's not a code or programmer's tools related question, and therefore it's off-topic here. This site is for problems commonly experienced by programmers related to their code. Do you have code written that attempts to implement your solution? — Ken White 42 secs ago
 
@Duga shockingly, that looks like a potentially good Q for PSE, even though the commenter is not recommending migration
 
psr
@durron597 - Um, we're using Angular. Little Googlers jump out of the keyboard and do everything while we just pretend to work.
 
@psr I'd love to see a video of that
 
11:22 PM
@psr Apparently, that's fairly descriptive of Angular's architectural organization.
Also, we get to relearn everything in Version 2. Yay.
 
@amon this is what I was trying to share with you in our recent chat. link
 
@overexchange How long are you going to keep cogitating over that question?
If you're worried about writing the game correctly the first time, don't. You won't get it right the first time. It might take three tries or more. That's true whether or not you actually know the true meaning of "computational object."
 
11:37 PM
it's more important to keep the code well-written and organized in some way so that it's easy to refactor and rearrange later on when you realize you did something wrong (or something you did right became wrong as the requirements changed)
 
user55340
The general workflow of today is with the assumption that you will write something multiple times. Waterfall is dead, long live the waterfall. You don't sit down and write a huge (perfect) design document, and then go from the design document to perfect code. That doesn't happen.
 
user55340
Btw, I hate java.sql.Date.
 
user55340
> Notes on java.sql.Date

(aka "Please do NOT use java.sql.Date, ever!")

Although Jackson supports java.sql.Date, there are known issues with respect to timezone handling, partly due to design of this class. It is recommended that this type is avoided, if possible, and regular java.util.Date (or java.util.Calendar) used instead. If this is not possible, it may be necessary for applications to convert these dates using java.util.Calendar and explicit timezone definition.
 
user55340
ResultSet.getDate gives back a java.sql.Date object. Since java.sql.Date is a subclass of java.util.Date, it fits nicely into the java.util.Date field... and sits there lurking as a bug until you serialize it and you've got the wrong date.
 
11:42 PM
dates are so much harder than they need to be in all languages
 
user55340
@Ixrec Yep. But when its sitting there and previous coders made a special effort to not use it... but forgot to check the return types that they were getting.
 
user55340
So now we are converting it all to Calendar objects.
 
I still can't get over the crash we had in production that only occurred for Brazilian clients 6(?) days after a daylight savings time change equivalent, and the root cause is that there's no 100% timezone-aware way to add/subtract 1 day from a Javascript date
does java.sql.Date do stuff like that?
 
user55340
@Ixrec not really... its how it is seralized that I'm dealing with. The timezone is in there in both (its a subclass) - the difference is that java.util.Date serializes to a number of milliseconds since Jan 1, 1970. java.sql.Date seralizes to the string '2015-05-27'
 
user55340
Which, well, doesn't have any timezone information. So, when you convert it back into a java.util.Date on the other side of the request, it says "oh, this is midnight 2015-05-27 GMT which is 7pm 2015-05-26 CDT" and then you've got a date of 2015-05-26 floating around instead of 27.
 
user55340
11:48 PM
When dealing with things like "when is this person going to be released from prison", such values are very important.
 
I think I have a potentially unjustified bias against any date/time class whose default serialization is not either seconds from the epoch or ISO 8601 format
every other default format I've seen drops information just like what you're describing
 
user55340
So, that was today's bug.
 
user55340
Oh, don't forget that I need to serialize this back into a java.util.Date object to maintain the contract of data objects with client jars.
 
of course
clearly you need a javaDateMultiFactoryTypeAdapter
 
what about a javaDateMultiFactoryManagementAbstractServiceLocationDecorator?
 
user55340
11:57 PM
Yep. Fortunately, the class that does the smashing of one bean to another (that then goes out across the request and gets searlized back to itself) is well unit tested.
 
user55340
... I wrote it.
 
00:00 - 18:0018:00 - 00:00

« first day (1727 days earlier)      last day (3262 days later) »