The Whiteboard

Benjamin Gruenbaum

11:05

Did anyone here ever actually use spiral?

(The methodology)

gnat

11:21

This question appears to be off-topic because it is about statistical usage data and is somewhat conversational. Consider asking your question on programmers.stackexchange.com — Bohemian 45 mins ago

@Bohemian Please stop using Programmers.SE as your toilet bowl - question has been cross-posted and closed at Programmers - programmers.stackexchange.com/questions/226511/… — gnat 1 min ago

Benjamin Gruenbaum

upd

ConfusedStack

14:07

Hey, so I have a question that I think you guys/gals might be the people to ask, but I think it is off-topic for a post. I want to know what libraries are available (C/C++) to interpret piped input and use it to update state/variables in a program, preferably using some sort of event system. Also if there is a term which refers to this type of thing so I can actually run better searches for solutions. I'm not looking for "best" because that is opinion, rather just existence.

maple_shaft

14:34

@ConfusedStack Yeah that would be offtopic at the main site

user20683

@ConfusedStack Try the C++ room on Stack Overflow's chat

maple_shaft

@WorldEngineer Yeah the C++ room is pretty rude

just be prepared

They have been known to make grown men cry

user20683

@maple_shaft I suppose hanging out on the Bridge has made me jaded.

user55340

16:21

@Shahar Consider that we are the... "craftsmen" of computing. Take a cabinet maker / woodworker. He's got a basic understanding of geometry and such, but it takes a mechanical engineer to compute the weight and stresses on the wood. The woodworker isn't doing that. However, the woodworker probably knows how to join two pieces of wood practically and is familiar with the tools of the trade.

user55340

Likewise, as programmers... there's math back there and we can implement it... but we really aren't math people. We're people who build things, and join parts of libraries together to make them do what we want. Depending on where we work, we may be working on house sized projects, or cabinets... but its about the tools we use and trust the people who wrote the libraries with all the heavy math in them to do it right.

Robert Harvey

The woodworker also will generally have some sort of plans or specifications, in a similar manner to software requirements.

user55340

Likewise, the way a woodworker determines the amount of stress a shelf can take is "keep putting stuff on it until it breaks" and then weigh that. The woodworker may hunt up some math written by some engineer to get a rough ballpark, but isn't doing the math themselves.

Robert Harvey

The plans will specify shelves that are thick enough to hold a certain weight.

Benjamin Gruenbaum

5 hours ago, by Benjamin Gruenbaum

Did anyone here ever actually use spiral?

user55340

16:26

(incidentally, my grandfather was a "wood engineer" - Forest Products Laboratory... as kids we had blocks that he brought home to play with... they were all exactly 2cm x 2cm x 20 cm in size... different woods, and they all had a 'dent' in the end where they were struck to determine the amount of force transmitted through them, deformation of the wood, etc...)

Benjamin Gruenbaum

Trying again :)

user55340

Spiral model

The spiral model is a risk-driven process model generator for software projects. Based on the unique risk patterns of a given project, the spiral model guides a team to adopt elements of one or more process models, such as incremental, waterfall, or evolutionary prototyping. History This model was first described by Barry Boehm in his 1986 paper "A Spiral Model of Software Development and Enhancement". In 1988 Boehm published a similar paper to a wider audience. These papers introduce a diagram that has been reproduced in many subsequent publications discussing the spiral model. [[Imag...

user55340

?

Benjamin Gruenbaum

Yeah, I'm wondering if anyone actually uses it

user55340

Benjamin Gruenbaum

16:28

Yeah, that

user55340

I'd have to say... not formally... I've seen it, it sounds like a good idea... but... well, rarely do I see people actually doing "prototype", "prototype", "product"... it is more often "prototype... it works? ~~prototype~~ its a product."

Shahar

do you guys know how to test my PHP code to make sure it's secure?

from sql injections and such

user55340

@Shahar The thing you want to look for is 'OWASP' - Open Web Application Security Project owasp.org/index.php/Main_Page

user55340

And they've got a whole set of testing tools: owasp.org/index.php/Appendix_A:_Testing_Tools

user55340

One section on there is for just testing sql injection.

Shahar

16:38

Thanks I'll look into that

maple_shaft

16:59

@MichaelT At my company it is worse still. 20 people who have no idea what they are doing go through the project management motions without any understanding or technical ability. Meanwhile because of lack of requirements the 2 developers on the project kind of wing a prototype of some kind on vague information they can garner in the 2-4 hours of status meetings they are forced to go to everyday...

user55340

@maple_shaft Programming methodologies take discipline. Discipline from developers, project managers, and business owners. If any one of them lacks the discipline or fails to buy into the methodology... well, it is a mess.

maple_shaft

... then an analyst discovers the prototype and decides to demonstrate it to a VP without the developers knowledge, then they take credit...

Then the VP decides to set a "goal" to release the "product" in 2 weeks

user55340

That's part of the reason that "fail-agile" where its "just do it" and "waterfail" of "write lots of documents, then just do it" are so common (and cowboy coder - though few admit to that).

maple_shaft

Then the Project Manager seems to misinterpret the goal to "OMFGSPLOSION YOU NEED TO WORK 11039239 HORS OF OVERTIME!!!!!!11one"

Then the developers who don't give a shit do nothing and the project manager gets fired and replaced with three more project managers of equal or lesser incompetence

I will say that for never getting credit I am quite well protected by never getting blame either

user55340

Not getting blamed isn't a bad thing.

Jimmy Hoffa

17:08

([a-z]{3}).*\1

...196 points on number 4

(.{3}).*\1 <-- 200

user55340

@JimmyHoffa Very nice.

maple_shaft

@MichaelT It is as if the executives secretly know that the developers are doing the best that they can and that we save the project nearly all the time through heroic efforts, but they don't want to admit or acknowledge this

user55340

@maple_shaft c2.com/cgi/wiki?HeroicProgramming

Jimmy Hoffa

@maple_shaft Which is why heroic programming should never be done.

user55340

"Commitment Oriented Development"

Jimmy Hoffa

17:15

or at least, not more often than once or twice a year

user55340

And another good read - Cargo Cult Software Engineering

Jimmy Hoffa

and by once or twice I mean once to two weeks a year

maple_shaft

That is an interesting read, but I take offense to the statement that heroic programming should never be done

user55340

@RobertHarvey on your ld thing - ld.so(8) : linux.die.net/man/8/ld.so "ld.so, ld-linux.so* - dynamic linker/loader"

maple_shaft

It is not necessarily overtime and sacrifice that goes into heroic programming

Jimmy Hoffa

17:18

MBAs pay attention to the way things happen and the way money is made, they don't pay any mind to the cost just the result. If they consistently get the result they want at a high cost to their employees, all they know is they consistently achieve the result they're after

maple_shaft

sometimes it is just having a clear view of user needs, technical realities and political obstacles

user55340

> The programs ld.so and ld-linux.so* find and load the shared libraries needed by a program, prepare the program to run, and then run it.

maple_shaft

something that can only be done or figured out through years of experience

user55340

@JimmyHoffa The key question there is "is the employee replaceable at a low enough cost" (something my former employer is having difficulty with now)

user55340

forbes.com/sites/lizryan/2014/02/01/…

Jimmy Hoffa

17:21

@MichaelT Though it may be an important question to ask, again my point is it's not one that even floats onto the radar of management in most orgs

maple_shaft

THe reason for which comes down to human nature

@JimmyHoffa It is an interesting phenomenon

user55340

In many places, the answer is "yes"... or rather "we lose one, we'll hire another one in within 2 weeks and we'll be good"... ok, ignore ramp up time... but the idea that they can hire another programmer in 2 weeks.

user55340

My former employer... well... they're down... way down on the programmer count.

maple_shaft

When the housing market was booming sellers had no problem ramping up their asking price

Jimmy Hoffa

Until their approach fails miserably a significant enough portion of the time (terribly failure one in four times isn't enough), they will not alter their approach.

maple_shaft

17:23

however when the market tanked and supply was high, everybody stubbornly refused to drop their rates to be competitive

Nobody wants to be the patsy, the one guy who paid too much/sold too low

even if it means their own financial destruction, people have proven to stubbornly hold out until the very end

Jimmy Hoffa

True

maple_shaft

This is why IMHO the Free Market is a myth

it assumes all actors will act in their rational self interest, people, corporations RARELY do

Jimmy Hoffa

It's too fucking cold outside

Robert Harvey

The free market, like most ideas, is an imperfect one. But just because it doesn't work under every possible circumstance doesn't render the idea completely invalid.

user55340

@JimmyHoffa Move to Atlanta... oh, wait...

maple_shaft

17:27

@RoberyHarvey I look at it like Communism, it works in very small communities where people know each other and get along well

as soon as you start introducing too many people then people start not giving a shit about their fellow man, Eg. free market abuse by grafters and scam artists, Eg. Communistic abuse by freeloaders

Jimmy Hoffa

snow every day this week, single digit highs. Freaky freak shoot freak.

Robert Harvey

Cheaters are built into the fabric of the universe. If you write a computer program that simulates an environment where agents interact with each other and adapt, predators always emerge eventually.

maple_shaft

edited that there for ya ;-)

Robert Harvey

Oh, vagina feathers.

Sassa fras Sassa fras Sassa fras

user55340

@JimmyHoffa at least the highs are above 0.

psr

17:30

@maple_shaft It's more that people usually when heroic programming occurs, management is happy that stuff got done, the programmer is happy to be a hero, and nobody worries about the fact that the heroism was only necessary because something was broken.

2

@maple_shaft Some more modern economics tries to model actual human behavior, instead of simplifying by assuming optimizing Vulcans. It predicts many things a lot more accurately, but under such models efficiency no longer automatically magically happens. (I don't predict any mainstream political party in the U.S. will update to such a model anytime soon).

Jimmy Hoffa

@psr Yeah I've seen this too much so when something's broken I don't just work around it anymore. I run directly into it and either people help me solve it or nothing get's done until they do. I have the nerve to wait them out usually.

@psr Yeah, one of the nobel winners this year won for his modelling like this I recall hearing; basically modelled and showed that the markets are emotional and behave irrationally because of this

maple_shaft

@psr I agree but it is sometimes a necessary evil. I don't like to be the hero, but I know that when I don't, life can sometimes get ... unpleasant... to say the least

Robert Harvey

@JimmyHoffa: I never quite understood the impetus for asking people to flag a mod for a migration. By the time their question is migrated, they've already reposted on the correct site, and now you have Two Problems.

Jimmy Hoffa

@RobertHarvey ?

Robert Harvey

Not getting an answer isn't an excuse for cross posting, flag it asking for migration if you think it should be on another site. To answer your question: This is totally fine and idiomatic. Though I would suggest avoiding the CPS-Nest problem by declaring named functions and using those instead of defining them in-situ. — Jimmy Hoffa 3 mins ago

Jimmy Hoffa

17:35

You mean if you tell them "flag for migration to X.SE" they always just automatically cross-post?

psr

@JimmyHoffa I've heard other people I thought were smart use this tactic. I don't have personal experience to say one way or the other.

Robert Harvey

@JimmyHoffa Not always, but why make them wait? Only stellar questions should get migrated anyway.

Jimmy Hoffa

@RobertHarvey Because mods hate cross-posts - I thought ?

psr

@JimmyHoffa A lot of the recent Nobel prizes are for similar. The traditional model has places where it is full of fail, and this has been the only progress.

Robert Harvey

Migrating marginal questions only to create duplicates is a clusterfuck.

Most of the time, I just close and ask them to repost on the correct site.

user55340

17:37

The advantage of getting people into the habit of flag to migrate means that there's the chance they'll get feedback of why it isn't appropriate.

Jimmy Hoffa

True

@RobertHarvey lots of times the cross-post is as bad an idea as the original question

user55340

16

Q: Better "Flag for migration" interface

Currently, 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 ...

Robert Harvey

Yeah, but at least I don't have to deal with a failed migration, and the target site doesn't have Two Problems.

maple_shaft

starred and pinned

user55340

Find the proper site easily, have a specific callout to migration reason rather than random custom note, call out the points of "help/on-topic" on the interface.

maple_shaft

17:40

It is terribly un intuitive. I am not sure what the answer is but we need something that can be figured out by new users easier

user55340

I need to redo the mockup...

user55340

@oxinabox I like that idea - that information is read into the about page somehow, it would be nice to see it available on flag target page. I'll see about adding it to the mockup at some point in the not distant future (longer than 6-8 hours, but less than 6-8 days). — MichaelT Jan 17 at 15:20

Robert Harvey

Migration is a very precise tool, not a traffic cop. It should only be used for Academy Award quality questions that just happen to be posted to the wrong site.

Everything else should just be closed, and the OP asked to go to the correct site, read the FAQ, and post their question there.

user55340

The thing is a speedy migration or rejection of migration lets the person flagging it get rapid feedback. The quick feedback is key.

Robert Harvey

Migration is never speedy.

user55340

17:45

@RobertHarvey From sites that have a small queue and low question rate, things on the front page that follow an existing path can get migrated before the OP gets impatient and reposts.

Jimmy Hoffa

@maple_shaft Perhaps the auto-searching doodad that shows similar questions as you're typing in a title could search the users previous questions on other sites, if it finds a match it says "Hey - would you like to migrate this question from site X to site Y?"

Robert Harvey

11

A: The business of being a migration path for Stack Overflow

To the Stack Overflowians: Please don't migrate questions to Programmers unless you are familiar with the FAQ, know that the question is on-topic for Programmers, and the question is stellar. By stellar I mean Academy award quality. Ooohh, that better be a good question. If you think the quest...

A variation on "Don't Migrate Crap."

user55340

> This question is more conceptual than code based (comments already asking for code). The 'design patterns' aspect of it suggests that this should be migrated to P.SE. I promise not to cast a close vote on it (I am familiar with what belongs on P.SE - programmers.stackexchange.com/users/40980/michaelt ).

user55340

Unfortunately, I think that one got too old before it the mod looked at the flag - I really think it could have gotten a better answer here.

user55340

3

Q: Object Pool Pattern in Java

So I've implemented my own Object Pool Pattern and it works just fine and as expected. Returning my "Teacher" objects from a list and creating them when there aren't any. My question: The object being returned "Teacher" then needs to be casted into one of its sub classes which is specialised e.g...

java algorithm design-patterns

Jimmy Hoffa

17:49

I keep getting answers that get migrated to SO...stupid SO rep, almost as bad as TW rep...

user55340

Granted, I gave it the best answer I could... and don't mind the SO rep (at least it wasn't MSO rep)

Jimmy Hoffa

At least it's not MSO rep

Robert Harvey

Unicorn dollars.

user55340

Though, I wouldn't mind getting to 10k rep on MSO so that I can see deleted questions there. They might be funny.

Jimmy Hoffa

0

Q: "SOA vs. Modular programming" use cases

What are the use cases for which we would choose "Software-Oriented Architecture", and others "Modular programming" ? Mentioning some personal experiences and examples would be so appreciated. Abstract explanations would be less helpful. Thank you a lot.

soa modules

Software-Oriented-Architecture??

@RobertHarvey more like unicorn hairballs

user55340

17:51

Btw, @RobertHarvey ... this was something I was going to ask... found as a duplicate and then bountied...

user55340

13

Q: Tag Merge: [fsm] [Finite-State-Machine]

There seems to be at least two tags on Stack Overflow referring to Finite-State machines: finite-state-machine 178 questions tagged Finite State Machine is a mathematical abstraction used as a behaviour model composed of a finite number of states, their transitions and functions to design...

discussion stackoverflow tags retag-request synonym-request

user55340

(its a precursor to trying to get P.SE's finite state tags to try to get closer in line to SO's so they could migrate if need be... failed migration because of misaligned tags is annoying)

Jimmy Hoffa

@MichaelT FS&M is not a finite-state-machine...

user55340

@JimmyHoffa I wonder if the religious SE's have fsm as Flying Spaghetti Monster.

maple_shaft

@JimmyHoffa I had a conversation with @Shog9 a year or two ago about that and he basically said something to the effect that there is a "significant architectural challenge" in doing real time queries across SE site databases

user55340

17:55

@JimmyHoffa and the tag "fsm" ( stackoverflow.com/questions/tagged/fsm ) says: "Acronym for Finite State Machine."

maple_shaft

it would be hard to do Intellisense involving content on other sites

user55340

@maple_shaft Note the time it stalls when doing a migration. I would assume that similar time things are cross database ones also.

user55340

I suspect the databases are separate instances (things run when one site is down) rather than multiple schemas on the same db.

user55340

And there may be a mismatch between schemas as software is rolled out to one site and not another (special things for software recommendation / MSO).

user55340

Thus, pure db links wouldn't be... pleasant and data objects and such would need to know about all versions of the database in the network rather than just their version of the tables.

Jimmy Hoffa

17:59

@maple_shaft which was the answer to our pushes for hot-question algorithm fixes - basically saying that altering the query would require a ton of work because of the challenges involved in coming up with that hot-question list that spans SE sites.

user55340

You've noticed the change in hot questions... for some sessions (I think its session based) I get randomized order and selection for hot questions.

user55340

So things aren't always at the top of the list.

gnat

in The Water Cooler, 8 hours ago, by gnat

trying to imagine how "new reality" feels like for lemming answerers. As usual, they click the sidebar. As usual, they drop their zero effort "meh" answers into the hot question. As usual, they expect a warm wave of similar answers and sympathy upvotes... Oops! Instead, they meet a hard cold surface. "Hello, we've got quality norms here, your post doesn't fit. No, there are no more lemmings around to help you ignore this..." What a disappointment

user55340

@gnat The user experience - both for the incoming and established - is something that SE sometimes seems to miss taking into account... and difficulty guiding.

user55340

The difficulty guiding includes (my favorite MSO topic) migrations - "belongs on XYZ" - repost - closed in two places. Adjusting the ingrained cultural approach on a crowd sourced platform is... difficult.

user55340

18:09

SE had enough trouble with the shift of NPR to P.SE, trying to force such a shift on SO is... impractical (drive by upvotes, meh questions on smaller sites do not get a warm reception).

gnat

@MichaelT lately, I've got a feeling that they figured what they are missing only after a couple of SO questions somehow hit the top of the list

user55340

Possibly. I wonder if it wasn't pain as such, but rather 'enough data to note a disporportate amount of activity on the questions'

Aks

@MichaelT, I've come up with a broad set of points to cover in a system design interview, I'd appreciate some feedback

user55340

It really takes quite a bit to get a question to be painful on SO... typically some MSO activity that causes it to spiral (dueling close / reopen, mod vs non-mod).

Aks

1) Go over requirements and write down broad categories of tasks
For e.g, if you're designing Twitter, you'd need
1) Display Tweets
2) Display a timeline
3) Build a social graph
4) Build a search index

2) Give an approximate server architecture that includes web servers, caching servers and db servers
3) Describe the data model you'd come up with for the db(users, tweets etc)
4) Describe the cache, what it would store, eviction policy etc.
5) Scalability and security, describe master/slave dbs, shards, etc.

user55340

18:12

@Aks Umm... context? (trying to remember where this came from)

Aks

oh, haha.

I asked you about system design interviews about a week ago

designing youtube etc.

I had a post on programmers and you directed me here

user55340

Ok... that makes some more sense... I vaguely recognize it now.

user55340

And yep, that looks like a good starting point for the "design twitter" type interview question. Breaking down functionality and then the design choices for each bit of functionality.

user55340

Likely the interviewer may then drill down into the various parts that they're most interested in... the architecture or the data model or caching...

user55340

But the point is that you've thought about how to take a large problem and start breaking it into the bits where you can think about it and discuss various choices.

user55340

18:16

You could apply a similar approach to youtube, or yahoo, or any other very large piece of software.

Aks

great! Thanks!

psr

So, I'm reading Working Effectively With Legacy Code (belatedly). Interesting book. The idea is that to work to a better design you need to make tactical changes that 1. don't break existing code and 2. allow you to write tests. In the course of that you sacrifice, ironically, better design. Then you clean up design once you have test coverage.

user55340

(just pulling up the amazon page...)

user55340

amazon.com/Working-Effectively-Legacy-Michael-Feathers/dp/…

psr

It's interesting thinking about the techniques with a dynamic language vs. static. For static, the compiler will help you find all the X that reference Y, which is extremely helpful. On the other hand, a huge goal is to be able to replace a real call with a test call, and that's usual easier for a dynamic language without major refactoring.

But I can imagine in, say, Ruby, you could never be sure your change is non-breaking.

user55340

18:20

amazon.com/Refactoring-Improving-Design-Existing-Code/dp/…

user55340

Another good one to look at.

psr

@MichaelT I've read that one. Bought it, actually.

user55340

And very true about ruby refactoring - never being sure if something is breaking or not.

user55340

Dynamic languages really require test coverage to make sure everything still works.

psr

@MichaelT So imagine if you inherit a dynamic code base with no tests, and no design for testing. What do you do?

user55340

18:23

@psr Write tests for everything you touch... preferably before you touch it too deeply. Put the framework for testing into place.

user55340

Though, honestly, many times the business wants you to start working on fixing things, not writing things that don't have any immediately obvious business value.

psr

Anyway, much of the book is devoted to, often ugly, hacks to getting things testable, and discussion of pros and cons of each. Many are pretty language specific.

@MichaelT At one point the author mentions that he tries (as a consultant-coach) to get teams to agree to put any changes under test first. He claims in one iteration the teams usually come around, even though week 1 is pretty brutal.

I suppose he has motivation to toot his own horn with such claims though.

user55340

(Safari books has it... its added to my favorites now...)

user55340

Another book to look at - Quality Code: Software Testing Principles, Practices, and Patterns

user55340

Chapter 15 is about introducing testing into a legacy javascript application.

user55340

18:28

First paragraph of chapter 15:

user55340

> Michael Feathers defines legacy code as code without tests [WEwLC]. JavaScript and other dynamic languages benefit from tests by exercising a wider range of the dynamic behaviors that sometimes make them behave in unanticipated ways. Open-source projects benefit from tests so that multiple contributors can make changes without each having to be intimately familiar with all behaviors.

Oded

19:47

So, I go and get the new realtime questions on the Programmers front page. Click through to something that I know will be terrible and bam closed by 5 community members!

user20683

@Oded Welcome to my world

user20683

@maple_shaft Raspberry Pi is on a boat

user55340

@Oded You see, its a challenge to get our votes in before you...

user55340

@WorldEngineer Boat programming redux!

Jimmy Hoffa

20:53

@MichaelT how do I say in regex - "everything that doesn't match: (.)(.)\2\1"

because that matches all the wrong things and none of the right...

user20683

@JimmyHoffa There you are, I got something interesting for ya

user20683

careers.stackoverflow.com/jobs/43305/…

user20683

var youShould = from c
in "3%.$@9/52@2%35-%@4/@./3,!#+%23 !2#526%N#/-"
select (char)(c^3<<5);

user20683

This is LINQ I believe.

Jimmy Hoffa

@WorldEngineer Yes. takes each character, treats it as it's integral value doing the mathematical operations - which implicitly convert it to int or byte or whatever, and then forcibly get's recast to char so it can go back out as a char[] which can easily become a string (presumably the desired output)

Oh that's in that careers link

user20683

21:01

aye

amon

@JimmyHoffa which regex engine are you using? Either test that that regex doesn't match, or use a negative lookahead: /(?:(?!(.)(.)\2\1).)*/, which will match anything up to that forbidden pattern or the end of string

Jimmy Hoffa

@WorldEngineer that's "comprehensive" syntax - as in, C#'s List Comprehensions. You can translate it to haskell list comprehensions if you recognize it right; in haskell that would be: [(char)(c^3<<5) | c <- "3%.$@9/52@2%35-%@4/@./3,!#+%23 !2#526%N#/-"]

I fucking hate C#'s comprehensive syntax to be sure.

It's all desugared of course by the compiler - thus why I say C# comprehensive syntax because it's a part of the C# compiler which was built to desugar to using the LINQ functions which would be: "3%.$@9/52@2%35-%@4/@./3,!#+%23 !2#526%N#/-".Select(c => (char)(c^3<<5));

You could tell them it's LINQ in C#, better to say it's C# comprehensive synta which desugars to use LINQ - where LINQ is a collection of functions in the standard library

user20683

or python -> [str(c**3<<5) for c in "3%.$@9/52@2%35-%@4/@./3,!#+%23 !2#526%N#/-"]

user55340

@amon regex.alf.nu

user20683

Python has real issues with that list

user55340

21:14

Gah... sometimes auto correct makes for fun when you have a typo.

user55340

> This is a dead link to someone's code in a link only answer. Please delegate the answer.

user55340

@amon have you just realized that all your time for the next ${duration to solve} will be sucked away?

amon

@MichaelT Err what? I don't quite understand what you're saying without a bit more context…

user55340

@amon I noticed the link got stared... I wondered if you had starred it and since you haven't said anything in a bit where trying to construct a score of regexes to solve them all.

amon

@MichaelT That's indeed my star, and I'm currently at #10. It's an interesting challenge.

user55340

21:28

I haven't gotten that far... #10 is one of those "hmm" when I think about an old FSA that recall from college about "divisible by 3"

user55340

In binary:

user55340

(incidentally the s# has # = n % 3 --- 7 = 111 -> s1, 7 % 3 = 1)

Oded

21:52

programmers.stackexchange.com/questions/226600/… (+10k link)

user55340

Hmm - neat: mbostock.github.io/protovis/ex/jobs.html

user55340

@Oded It also goes to the "don't ask" help center bit: "I would like to participate in a discussion about ______"

user55340

22:24

Ohh... Anna is a developer now! ( meta.stackoverflow.com/posts/99341/revisions had me go hunting to see the "what? huh? why?" )

user55340

> Developer at Stack Exchange, working on the Q&A engine that powers our sites.

Formerly a community manager and elected moderator on Stack Overflow/Programmers Stack Exchange. I like cats, computers, and video games... pretty much in that order.

user55340

22:43

I feel dirty with this code... but I really don't want an exception.

user55340

if(!h.optString("couchId", "").equals("")) {
    ...
}

user55340

The getString approach throws an exception if its missing. json.org/javadoc/org/json/…

user55340

22:57

@Oded Looking back at that... he should have been pinging Eric rather than Jon.

Transcript for