The Whiteboard

12:06 AM

Programmers Stack Exchange Blog: IT Needs to be a Better Business Partner

Morons on November 08, 2013

Salespeople have a reputation for being just about the worst people to work with when building software. Being in the CRM space myself, this is a constant topic of conversation. Usually the focus is around user adoption, as opposed to mundane complaining about salespeople caring about nothing but their commission checks. But it’s true, that’s all they care about. That’s why when pitching CRM systems to salespeople the pitch always revolves around increased sales\commission. “This will save you time, so you can do more of what you do best, selling!” or “Not only will you have greater visibilit …

psr

12:48 AM

In that post Morons says IT should step up and learn be able to perform the job the user of their system is trying to perform. In my last job where the company was in no way selling software, I spent at least 2 years trying to evangelize the concept of training the programmers on performing other people's jobs (there was a huge disconnect) , and got nowhere. They didn't want to programmers big $ to not program, so the programmers wrote code to do things they didn't understand instead.

jmort253

2:06 AM

@psr Heh, they... as in management. In some organizations, it's the other way around; programmers don't want to learn anything outside of their domain.

Either way, it's sad.

One thing I like about where I work is they encourage that. I sit with the sales and support teams so I see when they hit a pain point with the stuff I'm building, and it's really eye-opening.

Not the same as doing the job, but still helpful.

user20683

@jmort253 I know my long term work as a retail customer service person and my working in receiving and logistics will come in handy down the line.

jmort253

@WorldEngineer It will... I can almost certainly tell you it will. :D

psr

@jmort253 That's really what I was trying to make happen. Some of the users were CPAs and it's delusional to think I could do their job. But it was more delusional to think they could throw requirements over the wall and programmers could maintain the system without spending any time learning about what the CPAs were doing. So strange how hard it can be to pull off.

jmort253

I worked in a call center doing surveys in college

and now I build Voip software for, well, a call center.

So it's like things just nicely fit together...

user20683

@jmort253 I did alumni cold calling years ago

user20683

2:13 AM

never again if I can help it

jmort253

@psr Yeh, it's not like we can become experts, but we can definitely cross the line a little and expand our knowledge.

It can actually be kinda fun to talk to sales and marketing people and be able to keep up with the conversation and sound like you know what you're doing

user20683

yeah

user20683

my sister is looking at working in advertising

user20683

she's a theater tech nut

user20683

and it's a way to actually do what she enjoys and not be broke

jmort253

2:14 AM

sounds like an interesting combination of skills.

for both money, and enjoying the work

user20683

@jmort253 yeah, I'm weird myself, Religion and CS degree and 7 years at a bookstore

user20683

also cover letters are the hardest thing to write

jmort253

lol

prob have to just cut out a lot of stuff.

You remember jonathan heyward?

He posted a question on Workplace about how he couldn't get hired because he had a ton of skills and listed every one of them on his resume/cover letter.

user20683

@jmort253 The name sounds vaguely familiar...

jmort253

He's a self-proclaimed genius.

user20683

2:17 AM

mine is one page

jmort253

You know the types :D

user20683

@jmort253 everyone who complains about Physics.SE...

jmort253

but the question, while it ended up closed, did highlight the importance of staying focused only on pertinent skills.

lol

no comment

lol

but yes....

but it is interesting how, if you're a programmer, you almost do better at your job if youre also somewhat schooled in whatever it is you're building software for.

Building Physics software, it helps to know some physics.

Geology software, you'll prob do good at it if you've spent some time in the geology field.

Business software, same deal....

For instance, we'd never hire a NASA engineer, mostly because she'd get bored writing business apps instead of building the next Mars rover, but she'd also probably not get why sales people don't want to have to click 3 different buttons to create a new lead.

She'd probably complain that they're just whiny babies.

Sort of like this guy who -- gently, mind-you -- demeaned the asker for not reading these horrid Oracle docs: stackoverflow.com/a/19829679/552792

user20683

@jmort253 It's one of the arguments that I've actually considered advocating is that if you are gonna do dev work, either go domain or do Software Engineering. If you are gonna do CS research then go CS but otherwise just take what you need to understand it and move on

jmort253

@WorldEngineer Well, I've had some conversations with devs who didn't have a solid CS background, and it was like trying to teach programming... while also trying to solve critical business problems in a way that wouldn't suck 6 months later when shit started to break. :D

To some extent, I think one can get away with it.

But there's a lot I learned in CS that just changed the way I think about problem solving.

user20683

2:28 AM

@jmort253 fair enough, I still think I'd have done better with Math/CS than Pure CS

jmort253

If I had to do it all over again I don't think I'd skip the upper division CS classes, even though one of my prof's basically told us after CS 1010 that we could go get a $50k a year job.....

user20683

I'd not trade Analysis of Algorithms or Programming Language Concepts for anything

jmort253

@WorldEngineer Depends on what you're doing I guess.

user20683

I would trade my Blender class I was forced to take because scheduling

jmort253

And with Math... I have a lot of respect for Math folks.... they're able to wrap their heads around complex stuff.

I've worked with engineers, like Mechanical Engineers who somehow ended up as programmers.

They could only think linearly....

they couldn't cut through a recursive algorithm or even write code that could be reused or scaled.

not to say that's every ME.......

just my experience with the few I've encountered.....

I don't wanna stereotype anyone.....

sometimes people just plain suck at their jobs and it wouldn't have mattered what their background was...

So I can't disagree with your earlier statement.... take someone smart with the right combination of hunger for knowledge, and it really doesn't matter what they were educated in.

user20683

2:33 AM

@jmort253 yeah

jmort253

We had a guy with no degree working with us who could sling some code. It was beautiful code

I should add open-mindedness to the list.

user20683

@jmort253 @JimmyHoffa has no degree and can sling code like few I've ever seen.

user20683

yeah, I've few linguistic prejudices

jmort253

If you didn't have a CS background and are humble, you'll do better than if you don't have a CS background and think CS stufff is a waste of time :D

user20683

whatever it takes to get the job done

user20683

2:35 AM

I'm actually intrigued by the idea of using APL or similar to do Image Processing work

jmort253

or maybe 3D printing? :D

user20683

@jmort253 it's funny, 3D printing as a thing holds 0 appeal for me

user20683

I like working with my hands when it comes to making physical stuff.

user20683

too much the artist

jmort253

I like the idea because it brings us closer to the replicators on Star Trek.

user20683

2:38 AM

yeah

jmort253

and reducing dependency on manufacturing.

user20683

I like the idea of being able to make stuff out of just like waste plastic. Just chop it up and sinter.

jmort253

you're your own 3D printer

user20683

@jmort253 yeah, I suspect eventually it will be synthetic microbes that "grow" the parts.

user20683

you just drop a data chip into a vat and out comes you thing a while later

TRiG

11:42 AM

So, the blog is clearly in need of a proofreader. I mean, ouch. Solduses slanting in the wrong direction, using a noun where you should have a verb (setup != set up), and all sorts of other problems. I'll try to go through it properly tomorrow.

gnat

12:58 PM

simple math: 23 answers, 16 answers, 15 answers, 17 answers -- hotness score just doesn't discard answers when voting evidence indicates that these are not good data points — gnat 4 hours ago

Math guys easily exploit the bug in hotness formula. 2% penalty doesn't help, unsurprisingly

gentle 2% penalty for same-site-questions is easily overruled by stuffing more answers into these, as demonstrated by recent example of 4 math questions occupying top of collider. Dropping tiny 2% when QScore is blindly multiplied by AnswerCount straight up to 10 answers looks futile — gnat 2 hours ago

9

Q: Normalize the rankings for the "hot questions" list

Not sure if this is being done already, but... There seems to be a problem with the "hot questions" list, in that it almost always has a lot of questions from the Math.SE site. I imagine two possibilities: All SE users are really into their Cauchy sequences and algebreic theory The Math.SE use...

gnat

1:18 PM

would be interesting to estimate how Math questions would rank at collider if formula would discard answers when voting evidence indicates that these are not good data points -- ie those "having score less than (TopAnswerScore/10-1)". Per my observations, Math guys indeed upvote a lot, but they don't tend to upvote low quality answers (why would they?) — gnat 5 mins ago

gnat

2:37 PM

funny hack for posts you recommend to delete in LQ queue - prior to voting, cast a VLQ flag on that post and voila, number of "Look Good" votes required to drop it out, will increase - giving it more chance to get enough recommend deletion votes. No risk hack - the worst case scenario is you'll get a disputed flag, not a big deal

number of "Looks Good" responses equivalent to the number of flags -- it would be helpful to add an explanation for why "Recommend Deletion" votes from the queue don't follow this approach. It is not obvious why 2 flags (potentially from low rep users) raise the bar higher than 1 flag + 5 deletion votes from 1K+ users (I can ask this in a separate question if you prefer a dedicated post to clarify this). To avoid misunderstanding, I don't challenge this decision, merely wonder about the reasons for it — gnat 13 mins ago

user20683

3:14 PM

@TRiG I do what I can for the blog but I'm only one guy. I corrected a good chunk of the original draft but there are things I miss. This is one of the largest problems with the blog aside from drumming up content. I'm a reasonably good writer but my editing skills are underdeveloped.

user55340

@WorldEngineer Codereview for blogs!

user55340

Gah... when I see things like this I wonder if the poster cared about what they were writing at all...

user55340

> THNKS iN ADvance FoR your COmments

user55340

Its either that, or a cat was playing with the shift key while they were typing.

user20683

@TRiG Also what is a "Soldus"?

TRiG

3:39 PM

@WorldEngineer Sorry, I meant solidus, by which I meant virgule.

@MichaelT bash.org/?100798

user55340

3:56 PM

> (mike): HEH, THIs Is AMuSINg
(mike): I Got a WiNAMP pLUgiN THAt BLInKS THE KEYboARD ledS tO THE MUSIc
(mike): BUT IT acTUALLY turNs THe CAPsLOcK On AND oFF iNSTEad OF JuST the LIGHt

3

user55340

@TRiG Gah... that would be amusing/annoying rather quickly.

Jimmy Hoffa

@MichaelT I can just picture my cat laying next to the keyboard while I type, just pawing the shift key because he's not being paid attention to

That's a surprisingly plausible cause

Jimmy Hoffa

4:21 PM

@jozefg looking closer at your reify code, and it looks like you only count it as a record if there's 1 type constructor the way I read it, you aren't counting it as a record if there's multiple record constructors in the type?

user55340

0

Q: How to deal with requirement changes in agile development model?

I know that in agile requirement changes should not only be planned for but also embraced. But I still don't know agile how to handle these changes.

agile

Jimmy Hoffa

Any reason for that? You could check length cs > 0, and can you explain this bit to me: cs >>= constrFields where cs is all of the constructors, so are you getting the permutation of fields for all of cs?

user55340

This is a straight forward question that has a fair bit of possibility to generate a bit of activity.

user55340

(but, is also a duplicate...)

Jimmy Hoffa

@jozefg can't you all (((flip >) 1) . length) $ fmap cs constrFields to then allow multiple type constructors calling it a record when all constructors have multiple fields?

(or does it only need to be > 0 ? I'm not familiar with those Data functions you're using there sorry)

user55340

4:40 PM

{trolling} @GlenH7 i18n keys should be done in a language where there is little change. Like Latin.{/trolling}

user41796

@MichaelT I was trying to think of something smarmy to say about greek, but gave up

user55340

Oh! I know! i18n keys should be done in emoticons!

user55340

emojidick.com

user41796

I have been playing words with friends with one of my brothers. He dropped a latin word in there that completely erased the 50 point lead I had. Was so mad

user55340

user41796

4:43 PM

I didn't have to do a lot of i18n work, but I did a fair share. Handling that in C could get challenging

user55340

user41796

> Emoji-Dick doesn't seem very interesting

user41796

Kind of amazing that they translated the book in just under 44 days

user55340

Mechanical turk on amazon. They had LOTS of people working on it.

user41796

granted, those are full 24 hour days, not working days. But 132 working days ain't bad either

user55340

4:48 PM

800 people doing 3.795M seconds of work with $0.05 / translation and $0.02 / vote / translation.

user55340

So... the thing is, to have that be the keys for translation.

user55340

page.🚂🚦="train schedule"

user41796

@MichaelT unfortunately, my browser (firefox) only shows the "this is a unicode character but we're not telling you which one" symbol

jozefg

@JimmyHoffa Because it becomes ambiguous which one you want to reify too very quickly. By 1.0 I plan to have a recordTAt :: Int -> Maybe RecordT and have RecordT contain the information about constructors

user55340

4:55 PM

See, that way we can get around the native language of the translation service and help those 20 somethings who only know how to type in such a way.

user55340

This can be extended to contractor code too! class 💩ThirdPartyCode - just add the 'pile of poo' as the prefix to all externally written code to make it easy to do code reviews on.

user41796

@TRiG - just now reading the recent blog post. I see what you mean regarding editing.

Jimmy Hoffa

5:16 PM

@jozefg That's fair, but there's a proper solution

Hear me out: Match up the constructor lexicographically, and then you do the permutation. If there's one type constructor that matches thats all that will float through; a list of one, but if there's 4 you'll end up with a list of all 4 reified. It's up to the consumer to solve ambiguities at that point

you're already working in the list monad (I think?) when you're doing the cs >>= constrFields, so just follow the list monad through to it's conclusion filtering down to possibilities and let consumers of your code filter further

Jimmy Hoffa

5:30 PM

Make sense or am I spouting incoherent babble again?

(I didn't ask you @MichaelT, though maybe I should, you write Perl which means you're an expert in incoherent babble)

gnat

free spam flags, and (not surprisingly) in a close-worthy question...

-2

A: What pricing model has more benefits in mobile app stores: free or paid?

Very nice article.. thanks for the effort... eBook Conersion

user55340

Can't even spell "Conversion" right though.

user55340

@JimmyHoffa Been setting up couchDB on my local machine... and documenting how.

user55340

There's an issue I need to fix thats an internal one... its not "fix" so much as for some databases, we only have production. And that makes me sad when I want to test a change that requires some data to change.

jozefg

@JimmyHoffa I suppose you could do this, but this would change all function signatures to the corresponding type wrapped in a list.. which strikes me as icky. I'd rather implement it so that you specify what sort of reflection you're interested in and we'll go with that. It'd be trivial to implement a combinator taking [RecordT a] -> Map String Dynamic -> [a]

gnat

6:14 PM

I'm leeet

gnat

13.3k 9 45 81

java testing issue-tracking project-management terminology documentation agile

> 13337

user41796

@gnat You beat me to the punch on advertising the spam flags - I just protected the question

gnat

6:34 PM

@GlenH7 unfortunately, protection doesn't always save questions fro getting low quality answers...

-1

A: What makes C so popular in the age of OOP?

Because C has a huge user base. Yes it is a bit of a catch-22, but when I ask a question about C over on StackOverflow, I get the answer almost instantly. The same question about Python could take hours to get answered. With respect to C++, it's IMO more complicated to learn. Furthermore, after ...

user41796

@gnat The protection was to fend off future spam. Just seemed like a question that had the right words to attract spammers like honey.

user41796

@gnat Note that the user had all of 10 points earned on the site, so was able to bypass the protection on that Q. Kinda funny.

gnat

@GlenH7 yeah. I understand the difference. In some cases, the only protection against low quality are downvotes ("answer" I ranted about is a pure meh, not even remotely flaggable)

user41796

definitely meh. I was tempted to leave a comment to the effect of "and how does this add anything beyond what was already answered?!?!"

gnat

@GlenH7 so did I. Until I discovered that there's simple hole in the reasoning there...

did you compare asking SO questions for C# or Java? — gnat 13 mins ago

for the purpose of user base size, guy compares C against Python. For purpose of simplicity, against C++. How convenient

Straw man

A straw man or straw person, also known in the UK as an Aunt Sally, is a common type of argument and is an informal fallacy based on misrepresentation of an opponent's position. The misrepresentation can be intentional or accidental. It is possible to unintentionally misrepresent an opponent's argument by failing to understand it in the first place and honestly communicating what one (wrongly) thinks is the actual argument. This can lead to opposition believing the creation of a straw man argument to be intended, when it may or may not be. To "attack a straw man" is to create the illusi...

user41796

6:48 PM

Kind of a shame that Q went CW back in august. Oh well.

user41796

19

Q: How do I prove or disprove "god" objects are wrong?

Problem Summary: Long story short, I inherited a code base and an development team I am not allowed to replace and the use of God Objects is a big issue. Going forward, I want to have us re-factor things but I am getting push-back from the teams who want to do everything with God Objects "becau...

object-oriented class object srp

user41796

is my reading for the day. For one of the areas I own, I have ended up with a god object. I'm not convinced it's a bad thing though

gnat

my reading of the day is this Editor Manifesto...

3

A: Is it necessary to edit out the phrase "Good luck" from answers?

Let's Build an Encyclopedia The goal of Stack Exchange is to form an encyclopedia of knowledge that's helpful to future visitors for years to come, using a Questions and Answers format, and every time you post a question or an answer, you're planting a seed that helps grow that knowledge. It's i...

Transcript for