« first day (1265 days earlier)      last day (3724 days later) » 

12:06 AM
there are various programming paradigms. the semantics within each paradigm are interchangable
 
user55340
 
user55340
12:21 AM
@WorldEngineer that "how to program" thing - I found it while trying to find a "how to debug" guide... which isn't there, but is the first point in the guide. When showing it to a co-worker he also noted that "how to debug" being the first thing was a very important point.
 
debugging is a vital skill
unless you write code with no bugs
 
user20683
It's -5 for no frontpage btw
 
user55340
12:54 AM
@WorldEngineer Its from 2002... basic html sites were acceptable then... and programmers aren't known for fancy front pages.
 
user20683
@MichaelT har
 
user55340
One important bit there is there's nothing about language. That's not how you become a programmer. And as a beginner, most of your time will be spent debugging.
 
and as an expert, you spend time debugging REALLY weird shit
 
user20683
@MattD like strange edge cases in the compiler
 
those are rare. but yeah.
in 15 years of professional develompent. I've only seen the compiler get it wrong once.
its usually someone forgetting to initalize a variable somewhere
 
1:06 AM
@MattD this morning was usurped by debugging linq2sql having a generated classes property annotated Int when the Db returned a tinyint. Spent all morning trying to figure it before a DBA had time to pull the sproc text for me. (I should really have permissions to do that myself...mumble)
 
hehe
yeah. i hate that
i've been doing some fun EF crazy shit
lots of that kind of stuff
 
1:36 AM
@MattD Yeah, the sproc parsing is pretty shite
generation is probably great though for EF. Hopefully we'll be there with the rearchitecting we're going to do shortly
 
@WorldEngineer Sweet, thanks! I'll try not to screw it up so bad this time :)
 
 
1 hour later…
3:07 AM
Can I get some delete votes on this flame bait question?
7
Q: Is Dogma Driven Development considered harmful?

Tom DignanI have been reflecting lately on the dogma that circulates in the world of software development. Ranging from "goto is evil" to "singleton is an anti-pattern" or "reflection is slow". No concept is harmful in itself. Asserting one of the aforementioned is harmful is similar to saying that "a ha...

After all this time (and an admonishment from a community coordinator), it's still attracting goofy comments from people who want to start an argument. Time to let it slowly sink into the sunset.
 
3:23 AM
@MetaFight is your icon the SOD singer?
 
3:34 AM
can you send it to /dev/null?
 
@GeorgeJester It's Frank Zappa.
 
@metafight Wow! They look similar ;)
@RobertHarvey I noticed you have moderator status. Could help better understand how my recent post: programmers.stackexchange.com/questions/229477/… got moved to a completely different Stack Exchange site?
Is that something only mods can do? Does it effect my profile points when this happens?
 
user55340
@GeorgeJester 3k users can close with a migration for a defined migration path.
 
user55340
For P.SE, our only defined path is to StackOverflow - we don't get enough traffic to any other site that would really make sense.
 
user55340
For anything else, it takes a mod to migrate it.
 
user55340
3:46 AM
When a question is migrated, the up votes are transferred over to the other site (down votes get left behind - because they may be a "you posted this on the wrong site" down vote rather than a "bad content" down vote)
 
user55340
Eventually, migrations that aren't rejected (I don't believe that one will be) are deleted by the system on the origin site and any down votes then disappear.
 
@MetaFight: Is it possible to accelerate the process of developers learning from their design decisions? — Ryan Smith 2 hours ago
 
user55340
@MichaelT thanks for the feedback! So it sounds like I should remove it from U.SE and place it on Stackoverflow. Is that correct? — George Jester 22 secs ago
 
This comment seems completely out of context. Is it an appropriate "chatty" flag?
 
user55340
And Unix.SE is the right place for that question - its not a bash scripting issue... its a permissions issue.
 
user55340
3:49 AM
@MetaFight borderline. Could go either way. Its one of those "if it gets too long, invite them into chat." type things.
 
Hi @MichaelT Thanks for finding me in this chatroom, much easier to discuss here.
 
I feel weird answering his question because it's completely ignoring the chain of comments above it.
 
user55340
(@RobertHarvey would have the expert advice if the question would have been a good target for SO... I happened to have been the one who flagged it for migration to Unix.SE and @WorldEngineer was the one who handled that flag)
 
user55340
@GeorgeJester Discussions are what chat is for.
 
3:50 AM
@MetaFight ignore away, I'm reading them now
 
user55340
For reference, here's the question in question:
 
user55340
1
Q: What is the right file permission for a .pem file to SSH and SCP

George JesterI have tried to SSH into my AWS ubuntu server and copy the directory to my local machine. Throughout the process I experience different file permission errors (noted below). Is there one specific file permission needed for the .pem file that allows me to SSH and SCP? Or do I need to change th...

 
user55340
@GeorgeJester my take on it is that its in the right spot now. And I do tend to watch for the migrations that are good content elsewhere and get sent there.
 
user55340
Noting the ubuntu tag, it could have been migrated that way instead... but it is likely not a ubuntu specific issue.
 
user55340
 
4:00 AM
@MichaelT OK. Thanks a bunch for the feedback. I read the rules at U.SE: http://unix.stackexchange.com/help/on-topic and I'm of the mind that my question does fall in line with 2 their criteria: "Using or administering a *nix desktop or server" and
"The Unix foundation underlying MacOS (but generally not frontend application questions)"
 
user55340
> Applications packaged in *nix distributions (note: being cross-platform does not disqualify)
 
user55340
That one fits well.
 
Gotcha.
OK. Now what happens if a moderator at U.SE thinks it does not apply
 
user55340
Its not an 'and' between those... its an or.
 
Do they migrate it back to P.SE?
or to SO
 
user55340
4:01 AM
If they close it, it would get bounced back to P.SE as a rejected migration.
 
@MichaelT good distinction
 
user55340
The only site that I know of that has an 'and' test is Code Review - codereview.stackexchange.com/help/on-topic
 
user55340
> Simply ask yourself the following questions. To be on-topic the answer must be yes to all questions:

Does my question contain code? (Please include the code in the question, not a link to it)
Did I write that code?
Is it actual code from a project rather than pseudo-code or example code?
Do I want the code to be good code, (i.e. not code-golfing, obfuscation, or similar)
To the best of my knowledge, does the code work?
Do I want feedback about any or all facets of the code?
If you answered yes to all the above questions, your question is on-topic for Code Review.
 
user55340
And I'd hate to see the question that tried to be an 'and' condition for our on topic choices. (heh)
 
user55340
> algorithm and data structure concepts
design patterns
developer testing
development methodologies
freelancing and business concerns
quality assurance
software architecture
software engineering
software licensing
 
4:03 AM
lol
yeah that would be intense
 
user55340
I'm trying to write an algorithm to select a design pattern to be tested against existing methodologies. I am doing this as a freelance without a contract, but I need some QA help too. What architecture should I engineer that would match the open source licensing.
 
user55340
But going back to Unix.SE, your question is quite on topic there. Permissions problems, shell scripting, packaged application (ssh & scp) - all good matches.
 
user55340
You've done the research, you are displaying what information you can... I think it will be a question that gets answered there.
 
Gotcha. Thanks for the clarification and pep talk. If it fails there, I suppose I'll post it to SO unless @MetaFight councils otherwise.
 
user55340
4:07 AM
If it bounces back for whatever reason, flag it or come in here and ask (we've got a number of experienced users who can help craft it for whatever issue it was bounced on... and mods who could migrate it)
 
@MichaelT good heads up! Enjoy your night, and thank you for explaining things.
 
4:42 AM
@GlenH7 @MichaelT gamespot.com/articles/…
Re: Console value
 
5:37 AM
Heh wow, score on an answer at +4/-5, that is one seriously divisive answer.
 
 
3 hours later…
9:03 AM
anyone poked around with exception policies before?
 
 
2 hours later…
11:16 AM
maybe this and-clause (slightly rephrased) could find a use in picking appropriate design review questions for Programmers.

Does my question contain design? (Please include the design in the question, not a link to it)
Did I write that design?
Is it actual design from a project rather than pseudo-design or example design?
Do I want the design to be good design, (i.e. not design-golfing, obfuscation, or similar)
To the best of my knowledge, does the design work?
Do I want feedback about any or all facets of the design?
Nov 5 '13 at 16:36, by gnat
@MichaelT maybe design review questions "in pure form" rightfully belong to nowhere, because these have no way to verify research effort. For CR, working code requirement does that - if it doesn't work, this means insufficient effort, reason to close crappy question. For design review, one could just thoughtlessly glue some TLAs and see how it sticks "I am considering SOA+ORM, any thoughts?"
 
11:50 AM
just stumbled upon a couple of meta links from that Soft Rec beta. These guys apparently have learned a lot from NPR fiasco. If they manage to build a community to enforce guidelines like that, it might become an valuable site...
8
Q: What is required for a question to contain "enough information"?

FlykThis is the question edition of What is required for an answer to be high quality? One of the key problems that I see Software Recommendations having, is people asking extremely vague and simple questions that ultimately boil down to one of the two close reasons we see on the rest of the network...

13
A: My instincts are all wrong

Tim PostIt's a pretty simple checklist to ask a good, narrowly-scoped recommendation question. I'll make a rough outline here, I'm currently revising this (which is based on the original ground rules that I posted pre-launch). 1. Straight to the point, succinct title Don't use words like 'best' or 'goo...

64
Q: You're doing it all wrong!

animusonBefore this proposal went into beta, Tim Post laid out some ground rules on Area 51. If you haven't read them yet, I strongly encourage you to do so. The problem, a lot of you guys aren't following them, and it's really bringing down the overall quality of a just-blooming site. It's beginning to ...

7
Q: What is required for an answer to be high quality?

FlykThis is the answer edition of What is required for a question to contain "enough information"? As a continuation of the above, the second key problem that I see Software Recommendations having, is people answering questions in a manner that is extremely low quality and not very useful for future...

I am impressed
 
12:04 PM
this user is likely recently question banned at SO. "Questions" they drop at us now, make me feel that their q-ban here is not too far away: programmers.stackexchange.com/users/120452/abhishek-singh
-1
Q: Stop keyword not to be included into a string on output

abhishek singh<!DOCTYPE html> <html> <head> <script> function countWords(){ s = document.getElementById("inputString").value; var inputstring=s; $wordlist=Array("or","as","to","here"); for ($word in $wordlist) { s = '/\b' +s.replace($wordlist, '')+'\b/'; } s = s.replace($wordlist,'', s); $wor...

 
1:03 PM
voters, please don't forget that since there are 3 crappy questions from this user, you better restrict downvotes to only two of these, to avoid triggering brainless serial reversal script (it would be better for moderator to delete this crap completely off from the site)
 
1:29 PM
So I need to work on my killer interview questions that make everyone but the best and the brightest wash up. Time to read SO and Programming Puzzles?
 
2:22 PM
you guys ever get questions like this?
-4
Q: I am soon to have a 2nd Master degree and no idea what I want to do in life

user3327744I hope someone here can help me with a reasonable advice. If you might need any more info, don't hesitate to ask. I am in my late twenties, soon to have two Master's degrees in both technology and economics from top schools, and no clue what I want to do in life. Both degrees I pursued only bec...

we should migrate it here. I mean it is a person asking about IT stuff! :)
 
user41796
@enderland Doesn't say "as a Programmer" at the end of the title, so it's off-topic here.
 
Don't worry we can edit it!
 
user41796
@gnat He's off the front page now. Only took 2 more votes. :-)
 
user41796
@enderland Oh! In that case, please don't. :-)
 
user41796
Skimming over that TW question, I can't tell if it's self-entitled whinging or trolling. Or perhaps a little of both.
 
2:31 PM
we get a fair number of questions I can't tell whether they are serious or trolling
 
user41796
You guys need a meta question to the tune of "Why was my indulgent, self-entitled whinging rant insta-seleted?"
 
user41796
I'm thinking Oded nailed it with the insta-close. The follow-up comments were pretty funny too
 
user41796
And when does TW officially graduate? I saw the beta tag is still on the site.
 
user41796
2:54 PM
I think ratchet freak won FGITW on this one.
 
user41796
0
A: What is the advantage of wrapping exceptions

GlenH7A compelling principle behind error handling is to handle the error at the layer best suited to deal with the error. Handling an error can involve retrying the operation, signaling the user, or just plain rolling over and dying. Frequently, this means that error handling is taken care of fairly...

 
user55340
@gnat We'll have to have a "this would be software rec, but lacks sufficient information to migrate"
 
user55340
3:17 PM
@enderland That sounds like a good use case for a custom close reason.
 
user55340
(or even a canned one... you know, @enderland if you become a TW mod, you could make that a standard one too...)
 
Alison Sperling on February 19, 2014

In 2013, our Stack Overflow community grew from 21.5 million to 26.9 million monthly visitors from 242 countries around the world. We’re doing a lot to keep growing with the community — we now have localized versions of Careers 2.0 for French and German audiences, we’re developing iOS and Android mobile apps for our entire network, and our first ever localized version of Stack Overflow with the Portuguese site currently in beta. As a way for us to make sure we’re doing the most for our users and community on Stack Overflow, we conduct a survey every year to see what you …

 
user55340
 
user55340
People who fill out surveys know how to search Stack Overflow for answers without asking questions.
 
user41796
@MichaelT Shhhhh.... don't be poking at their sampling bias.
 
user55340
3:24 PM
@GlenH7 I'm fairly sure they are aware of it.
 
user55340
 
user55340
Almost always ~= 10k ... 1k respondents.
 
user41796
Weee. I was part of 18% at the time of the survey. Or maybe it was 12.1%, I dunno
 
user55340
And in other news of other stack overflows...
 
user41796
below 1k looks to be the threshold for 1/2 of the respondents
 
user55340
3:27 PM
 
user55340
Thought that you engineering types might like that.
 
user41796
color me sniped. It's on the reading queue for today
 
3:39 PM
@MichaelT Hahaha. best close reason ever
 
3:50 PM
^-- Approve if you see fit, looks like some low-rep user helping out, lurking
 
user41796
@JimmyHoffa You had the 2nd approve, so it's gone through
 
user41796
edits only require 2 approvals
 
user41796
Or a sufficient rep user to select "improve" and then tweak the edit a little further.
 
user55340
Its only SO where they need more because of robo-reviewers and more edits (good and otherwise)... and many more reviewers too.
 
user41796
4:09 PM
I really don't understand the point behind robo-reviewing
 
user55340
Teh badgez!
 
user55340
When there are so many reviewers on SO, its quite hard to get the badges where the queue is near empty always.
 
user41796
Clearly my priorities are out of whack
 
user41796
although I specifically don't review the first post queue anymore so others have a shot at getting some reviews in
 
user41796
And if / when our close queue stays at a very low volume, I would do the same there. But I'm not holding my breath on that one.
 
4:35 PM
@MichaelT We just discussed this court case in my Software Quality Assurance class. The guys doing the testing were absolutely clueless. removed code from production before testing because they thought "It didn't matter"
 
> The guys doing testing are absolutely clueless.
FTFY
Welcome to industry
 
I just don't see how you remove 1/4 of the code and then test that thinking that it's "close enough"
do you test a car with 3 wheels?
or a 787 with 3 engines?
 
@Ampt a tester would.
Now I have to find the dilbert a colleague at an old job had pinned to his cube..
 
@JimmyHoffa Ok, in that case it may actually be justified, but you get the point. the code wasn't designed to run with 1/4 missing, and that is highly unlikely to happen in the field.
 
@Ampt Also PM's like pushing these things because they don't understand certain things. When we do load tests here, invariably some non-technical person pipes up "Hey, let's turn off all the databases and see how the system behaves" - newsflash, it'll fall the fuck down. That's why you have redundancy for those systems, they're absolutely necessary
 
user41796
4:41 PM
<cue Austin Powers voice> Hey baby, they were LEAN!!! That's all that matters.
 
user41796
The voice is really what makes that comment work...
 
I found it, I feel better now.
I'd print it and pin it in my cube here but I sit too closely to QA and wouldn't want to upset anyone
 
user41796
@JimmyHoffa those gosh-darned cranky engineers. Why can't they just learn to go with the flow of things?
 
user55340
4:57 PM
@JimmyHoffa That is so the QA department at my previous job. They just tested things... "here's a load of 50 users because thats all we can generate and the servers look like they are holding up fine... Oh, we only hit the front page because we don't know how to generate the proper data to get to the parts that actually hit the database either"
 
user55340
I'd send it to a person I used to work with... but the people who would get it the most don't work there anymore either.
 
user55340
(Funny about that)
 
user41796
@MichaelT That's one of the broader meta questions I have long pondered about company culture.
 
user55340
More things for your reading list.
 
user41796
4:59 PM
It seems that a group's culture is relatively fixed and outsiders (those who don't fit in) end up leaving through one means or another
 
user55340
The second one... ghads... its hard to find a big company where that couldn't be written.
 
user41796
I have read the first one before, and I generally agree with what he's saying
 
user55340
> But in my experience, that’s not what happens. Instead, what happens is that the more talented and effective IT engineers are the ones most likely to leave — to evaporate, if you will. They are the ones least likely to put up with the frequent stupidities and workplace problems that plague large organizations; they are also the ones most likely to have other opportunities that they can readily move to.
 
user41796
I wish there were an "easy" way of detecting that before joining an organization. But I think a lot of it is highly specific to the individual and what they can deal with.
 
user41796
5:03 PM
I use a 60 / 40 rule with work - so long as I enjoy 60% of what I do and where I'm at then I can generally deal with the remaining 40%. It's a different way of saying "the grass is always greener on the other side of the fence."
 
user55340
Asking about turnover and team maturity.
 
user55340
Interview the seniors on the team yourself (interviews go both ways).
 
user55340
If you've got a very young team for something that should be old... there's a warning flag that needs some more looking into.
 
user55340
If the seniors on the grumpy / disgruntled side... big red flag.
 
user41796
Those are good points. And I think that swapping jobs before you have to swap is important too.
 
user41796
5:09 PM
Looking back at my prior gig, there were some red flags going in. I didn't have many options at that point though so that forced my hand.
 
@MichaelT my favorite question for company's to broach all of those topics is "Why are you looking to fill this position?" get to find out they've had turnover, or they've had quality issues and they want to get someone more experienced, or they just have a great new project or what not etc.
also "Why do you work here?" asking an engineer
Guys at my current team said straight up "It's not the code, it's not spectacular, there's a fair bit of legacy stuff, but I stay for the people who are friendly, smart, and fun to work with" etc - if they really struggle with the question that's a bit worrisome, or if they just blurt out "The benefits and pay are great" eesh...
 
user41796
I have found "benefits and pay are great" to be a very subjective statement. It's really in comparison to that person's threshold of pain for bureaucratic crap.
 
user55340
My hand was a bit forced that lead me to my previous employer... I had been unemployed for a year (2010) and finding any jobs were hard.
 
@MichaelT this is probably some of the most important and most frequently overlooked part of interviewing. Not to mention, a candidate who is interviewing the employer gives off SO MUCH confidence
 
user55340
Oh the reason I say seniors on the team is thats where the dead sea effect would be most visible.
 
user41796
5:19 PM
That's quite true - they really guide the team's culture by way of inertia tenure
 
user41796
Company culture is important to try and pick up on too. It impacts environmental factors that may end up being outside of anyone's control
 
user41796
@MichaelT - after reading the thermocline article, I was a bit shocked that anyone would dispute the existence of that effect. And then I thought back to a previous PHB who pretended to know everything and was a root cause of the thermocline on a particularly ugly project. That old PHB would have disputed the existence of the thermocline too.
 
user55340
5:36 PM
@GlenH7 I've seen that so visibly at the previous employer... we had debates about where the thermocline was.
 
user41796
There was no debating it at the previous gig. We all knew where it was at.
 
user55340
We knew the manager knew, but we didn't know if the manager was not skilled at expressing this to the director, or if the director didn't understand, or if the director didn't want to understand.
 
user41796
"yes"
 
user55340
And then if the director did understand, did the CIO... and we were fairly sure that the CIO didn't understand, and we knew that even if he did understand he wouldn't tell the owner.
 
user41796
On the flipside, if the executive or upper management layer is receptive then there can be some serious cred to be had by breaking through the thermocline and exposing the faults. Of course, you're gambling that the upper layer will protect you from the intermediate layer.
 
user41796
5:40 PM
But at that point you're playing with fire.
 
user55340
Trying to breakthrough the thermocline is indeed dangerous. You're essentially saying "you don't understand what your managers are saying to you" or "your managers aren't telling you the truth"
 
user41796
5:56 PM
In my situation, it was the latter.
 
@GlenH7 seriously playing with fire. You're gambling that the upper layer isn't of the military-style mind wherein the chain of command is priority number zero, and you never skip a level. Many people see the idea of going over your managers head as simple insubordination and cause for immediate firing.
Don't forget how many companies sue people who privately notify them about security holes in their systems they stumbled across
 
user55340
@GlenH7 When its both, then you've got even bigger problems.
 
user41796
@JimmyHoffa That's what made it fun in my case. PHB had already earned a reputation for manipulating the truth to fit his point of view. And he was dogmatic in following the hierarchy.
 
user41796
By that point, I had paid up my insurance (metaphorically speaking) and had already armed the thermonuclear warhead as a precautionary move.
 
user55340
The percent of "yes" went up as you went up the chain, so it wasn't exactly clear where it switched from "things are shaky" to "things are fine"
 
user41796
6:00 PM
I had consulted with some insiders who knew the exec well. They assured me it would be a relatively safe whistleblowing.
 
@GlenH7 Yeah, that's best time to do it. My last job I actually had my leg out the door and so started telling some really high ups some of the disturbing things management were up to, got one guy demoted (and dinged even further from the exit interview I had provided from what I heard). It was one of those "I wouldn't go back to this job if my entrails were on fire" so I didn't have to worry about bothering management there, and most all of them agreed with me anyway.
But that was a unique situation, there was no thermocline of truth because things were so bad that people at every level above the direct management were all trying to figure out what to do to solve the problem; bringing in folks from different offices around the country to try and come up with ideas for solving the endless quality problems that couldn't be hidden anymore. (When your major clients are fleeing because of defects, VPs can't say "everything's fine!" anymore)
 
user41796
I was identified as the only one or one of a very, very few who were actually moving the project forward. Term'ing me would have guaranteed project failure. Later on, I got one of the funniest compliments I have ever heard from that exec when he told me he was getting tired of hearing my name.
 
user41796
@JimmyHoffa They can keep shouting "everything's fine" as they get flushed down the toilet bowl...
 
@GlenH7 The software was a small bought company's of a larger conglomerate, the overarching people have no ties to or reasons to say everything's fine - it didn't reflect on them. It was an interesting situation, they brought me and some other seniors in because they had quality problems for years that got continually worse, losing large accounts continually for years, the owning company finally said they'd had it and canned the head of the smaller bit and put in a new head who brought her
own management team in. Their first inclination was bringing a few seniors on. But the old management layer was still reporting "Everything's fine!" While the owning company was saying "You're full of shit." basically
 
user55340
"Hiring an emacs programmer would be a good thing... just by editor choice you know they're a full stack developer... emacs database interface, emacs web interface... emacs OS inter... wait, it is an OS"
2
 
6:11 PM
Got to see what happens when a new management team comes in and tries to wrest control from an old management team that had been doing the same thing for literally 30 years. Interesting to see, but not an experience I'm interested in repeating
 
user41796
@JimmyHoffa that's when owning company screams "SHOW ME THE MONEY!"
 
user41796
"You ain't right 'cause you ain't making your numbers."
 
@GlenH7 It wasn't even a matter of making numbers or showing they are doing something, it was a matter of what they showed they weren't doing. They had numerous defects that stopped customers from being able to do basic tasks with the product, and they simply couldn't fix them. They tried and tried for weeks, months, certain customers couldn't use the application. Eventually the customers left - they never fixed the issues. The devs were too junior and the app was too bad and the culture
had too much inertia. Even when they found ways to fix the problems some managers said do not fix those issues because they didn't want to change the code.
Make the numbers? Hell even have the application fucking working would have been an improvement.
 
user41796
woo hoo! I won FGITW
 
user41796
0
A: What do two consecutive minus signs mean in a return statement?

GlenH7Your answer / thinking about the code is close, but not quite right. See this mozilla documentation on arithmetic operations for additional details. the line return --self.newID is using a prefix decrement operator. So it subtracts before returning the value, not after the return. The fir...

 
6:20 PM
@MichaelT yeah, emacs programmer... that sounds like the right approach alright
 
user41796
@JimmyHoffa I love when management develops an emotional attachment to the codebase....
 
user55340
@GlenH7 Just wait until it gets migrated to SO... and you can enjoy those fantasy numbers (though a head start on FGITW too).
 
user41796
@MichaelT Pretty much the only way I earn SO rep.
 
user41796
although my jQuery question pulled in two up votes. Was pleasantly surprised to see that.
 
user55340
Hmm... still debating the flag on this one...
 
user55340
6:34 PM
-2
A: How to teach pointers to beginners in C?

RossCan you tell us a little bit more about the audience? Are they college aged non-CS majors or high schoolers? Whatever the content you wish to cover, if it is a beginner audience you are going to present to, it is probably best to start off with a few slides explaining a schematic of the guts of t...

 
user55340
It hasn't garnered any pile ons, nor any disputes.
 
user41796
@MichaelT I've got 4 up votes on it now. I ought to VTC as OT -> SO now... :-)
 
user41796
@MichaelT I have done my good deed of the day. But I didn't up vote as it doesn't show that much effort or thought.
 
293 rep to programmers going pink. Woot.
 
user41796
Teaser alert! This is the background color: color-hex.com/color/f4eaea
 
user41796
6:48 PM
@MichaelT - I think this shade hints of the dried blood that you had made a request for. I'm half afraid it will be too dark though.
 
user41796
And I looked up javascript fireworks modules - all of them ran horrifically slow on firefox.
 
@GlenH7 Of course they were slow on Firefox, just like the internet
 
user41796
I thought the cloud solved everything though?
 
user55340
@GlenH7 Yea... just look what happened to Venus. ALL in the clouds and all that waste heat from the server farms make it completely inhabitable. If you really want cloud computing, you need to move a bit further away... Saturn I hear isn't too bad.
 
user41796
6:55 PM
stupid formatting rules for chat... grrr
 
user55340
(good author to read... he's got a number of interesting books)
 
user41796
I want to create more time in my schedule for reading. I have quite a few books in the queue that I want to get to
 
user55340
> In the upper cloud layers, with the temperature in the range 100–160 K and pressures extending between 0.5–2 bar,
 
user55340
For Saturn. Its really cold at 1 bar pressure, but not something that we couldn't deal with.
 
user41796
Is that earth bar or local bar?
 
user41796
6:59 PM
And did I give away that I have to deal with measurements with relative units all the time?
 
user55340
Earth bar.
 
Pointers fun and interesting...right, I'll just assume he meant to say beer commercials instead of pointers...
 
user41796
> It's a noble salvage attempt, but it removes a lot of the OP's original work. Some questions simply can't be salvaged.
 
user41796
Can I VTC an internal user's bug report as "unclear on what you're asking"?
 
user55340
@BobCross the fundamental problem with this question (that even your suggested edit didn't fix) is that it is a poll of opinions (thus the too broad) of "how would you approach this topic to make it fun and interesting? Do you have great examples that really help people to understand the concept of pointers?" - this would need to be fundamentally reasked in a different way, preferably posing a problem that the OP faces rather than the general fishing for ideas. — MichaelT 45 secs ago
 
user41796
7:11 PM
@AbijeetPatro - sounds like you would have 3 problems instead of just 1. — GlenH7 9 secs ago
 
Well, I left a message with NewsWatch, asking them to clarify.
 
user41796
@RobertHarvey That one did seem weird
 
user41796
Wasn't clear if it was a PR machine with a news outlet facade or what the case was.
 
I think that's exactly what it is. I'm not thrilled by their sales practices, though.
 
user41796
Does the "Who's Who" book scam still exist?
 
7:15 PM
That was a real thing. Except for the "who gives a shit" factor.
 
user55340
@GlenH7 Fairly sure that, and the poetry and photography scams are still out there.
 
user41796
@RobertHarvey I rounded down to scam because of that factor
 
user55340
I recall the photography one being asked about on photography DeviantArt forums when I was active there.
 
user55340
Now days though, they just have to change their name every time though because of ease of searching.
 
user41796
Every now and then Pandora serves up a dish of wickedly deep, dirty bass. Ah, bliss. Hopefully the full enclosure cans aren't leaking too much noise.
 
user41796
7:20 PM
@MichaelT That's what sticky threads are for - ID'ing the latest scams.
 
@GlenH7 Dubstep?
 
user41796
yeah, that's the current mix
 
Just bumped an entire solution from .NET 4 to 4.5.1, no code changes, built and unit (integration) tests passed. Always nice when upgrading let's you stay current without breaking anything at all (that's obvious anyway)
 
user41796
Drum & Bass will take care of that itch too.
 
user41796
Their Electronica channel has been hit / miss for deep bass.
 
7:21 PM
@GlenH7 Never could get into dubstep. Drum and bass is hit or miss.
 
user41796
@JimmyHoffa Kathleen Dollard has ID'd some hiccups in that upgrade. Just an FYI to research for any hidden gotchas.
 
user55340
@GlenH7 when it comes to photography, there are too many of them out there. Most photo "contests" are rights-grabs where by submitting the photography you also give them the right to reuse it, even if you don't "win". photoattorney.com/…
 
user41796
@ThomasOwens I forget the name of the DJ that hooked me on D&B. He played a set on BBC1's essential mix. It's one of my favorites.
 
@GlenH7 Every upgrade has hiccups, but .NET's been pretty good for a while not doing changes that make terribly huge problems. Some minor consistent code changes are no big deal
 
I have a technical problem that I've been trying to solve for a long time. It involves a data protocol that contains a packetized bit stream contained within outer packets. The outer packets are easy to read, no problem. The inner packets are a nightmare. Because the inner packets are allowed to "float," (they're not the same size) and I'm required to synchronize the bits to an external time base, it's not possible to abstract like I normally would. The tight coupling is a nightmare.
 
user41796
7:24 PM
@MichaelT oh brilliant. So I'm going to run a "contest" to build up my portfolio of stock photos...
 
@GlenH7 Does Pete Tong mix for BBC1?
 
user15026
@MichaelT Things like this are maddening.
 
user41796
@ThomasOwens I don't know that he's done an actual mix in quite a few years, but he used to actively do so yes.
 
He has a radio show in Boston. I think he does BBC1, too. I think it was him who did a trap set on the radio.
 
user41796
I haven't been following BBC1 as much lately due to logistical hiccups. So I haven't been following Tong as much either. Interesting that he's running a show in Boston. Then again, Tiesto has been hitting the US pretty hard for gigs.
 
user41796
7:26 PM
@RobertHarvey languages at play?
 
user41796
Can you pad the floating inner packets? Fluff them out to a guaranteed fixed size?
 
@RobertHarvey snif snif I smell...a parsing problem... :)
 
@RobertHarvey Pad the packets. Don't let them float.
 
What is tightly coupled?
 
7:28 PM
@GlenH7 That's a great question. The only padding allowed are some zero bytes to get it to a 32 bit boundary.
 
@GlenH7 It's an iHeart station, so you can stream it anywhere in the US. I think his time is on...Saturday nights? I'd have to check the station schedule. It is supposed to be the first (and only completely) EDM radio station in the US.
 
@JimmyHoffa Normally I'd just convert the inner packets to a bit stream, but because I have to correlate with a time base in the outer packets, I still need access to the outer packet reader.
 
user41796
RTOS or other hard time constraints?
 
No, it's not real-time. It does have to be reasonably fast, though (the data files are extremely large.
 
@RobertHarvey Do you need to stream? Can you create discrete messages?
 
user41796
7:29 PM
@ThomasOwens Phoenix was running one for a while back in 2008 / 2009, afaik
 
@RobertHarvey So the outer packet has a variable which is necessary in the inner packet?
 
@JimmyHoffa Yes.
 
user41796
@RobertHarvey I would consider wrapping the floating bits within their own fixed sized message block and possibly put a header on the block to indicate actual size.
 
@RobertHarvey So you need a binding environment with that outer variable in it that knows how to handle the inner packet. OuterPacket.Parse(Func<DateTime,....> bindingEnvironment) should do
 
user41796
In a previous life, we would write out objects that way. We had perhaps 3 layers total. The lowest layer wrote out it's headers most frequently, and then the upper layers less frequently.
 
7:31 PM
then your outerpacket parse takes the inner packet parser and executes it, giving it the timestamp it needs
 
At the moment, I've created a Correlated Frame to contain the necessary bits. The problem isn't so much encapsulating it. The problem is that the inner bits are stored as Little Endian, but the data is Big Endian. So I have to rotate the inner packet on 32 bit boundaries, and then stick it into a bit buffer.
 
user41796
so the lowest layer would own a 1k block, the next layer a 256k block and then upper layer some ##MB block for example
 
To make things even more fun, the packet is not guaranteed to start on a byte boundary. :)
So it's very difficult to get a referent for the start of the inner packet.
 
user41796
@RobertHarvey Oy...
 
@RobertHarvey outerpacket parser should find the boundary, pad it, and hand the DateTime + byte stream to the inner packet parser, inner packet parser can do the endianness swizzling (blech)
 
user41796
7:33 PM
@JimmyHoffa yeah, that sounds about right
 
user41796
Or force all transfers to be done as ___-endian and make the clients / senders adapt
 
@RobertHarvey That part is easy. Sync words.
 
The data already has sync words in it.
That's a whole nother fun story. You're allowed a certain amount of slop in the sync words. :)
 
@RobertHarvey Isn't the sync word a constant value?
 
user41796
@JimmyHoffa - looks like Kathleen's primary worry was that 4.5.1 updates in place on top of 4. Likely not an issue in your case.
 
7:35 PM
@ThomasOwens Yes, unless you're allowed a certain number of bits in error.
 
user41796
@RobertHarvey Oy y vay....
 
Got that solved with a Hamming Distance.
It does make for some interesting problems if you have to navigate backwards, though.
 
@RobertHarvey Oh. I was unaware you were also introducing errors in the transmission protocol.
 
Ba da boom.
 
Or at least had the ability. I was thinking like something over TCP where your data is guaranteed in order, complete, and correct.
 
7:36 PM
I wish. Almost forgot about that. There are multiple data streams, and the packets in each stream have a sequence number.
 
psr
@MichaelT What's the point of being a senior if you can't be grumpy /disgruntled? It's like a programmer's version of a social security check.
 
@psr sounds like you should ask a question about it. "how to check if a senior programmer is grumpy"
 
Anyway, I've got a Correlated Frame object that holds the inner packet data. The only problem I'm still trying to solve is counting the number of bits between the start of the outer packet and the start of the inner packet, so that I can correlate the outer packet time with the inner packet time.
It's actually all nice and tidy, except for that one little detail.
It requires knowledge of the outer packet, and maybe the one before that. And the one before that.
 
user41796
@RobertHarvey Can you stash a timestamp in that correlated frame object header?
 
It's already there.
I just need to figure out how to calculate it.
And I don't think it can be done without tracking several outer packets.
 
user55340
7:39 PM
@psr There's grump / disgruntled and there's "grumpy because I can't get a job elsewhere and don't like this one"
 
user41796
@RobertHarvey doesn't sound like you have any other options
 
user55340
If its the later, that means that the company is likely suffering from the dead sea effect - the more capable people are leaving at a faster rate than they are getting hired.
 
Sorta breaks my bit stream architecture.
 
user41796
From what I'm reading, your transmission is essentially real time and you have to cope with noise in the line.
 
Yes, essentially.
 
user41796
7:40 PM
and you need to do time synchronization based upon the bits rolling off the stream
 
Exactly.
It gets better. :)
 
user41796
wait, there's more?!?!
 
The time in the outer packet is not guaranteed to be correct. >_<
 
user41796
That's epically hilarious
 
It is.
 
user41796
7:42 PM
I would take a hint from the chaos theory playbook and realize that your signal is going to be inherently noisy and there's nothing you can do to eliminate all of the noise
 
psr
@MichaelT I feel you aren't giving enough credit to "grumpy on general principle".
 
user41796
So with that said, the receiver could have multiple threads running each thread's best guess at what the time should be based upon the bit tracking that's been going on
 
user55340
@psr Thats where you need to interview them to determine if they are skilled and capable, or are the "I've claimed my domain and no one will take it from me"
 
user41796
"Each" bit kind of represents a decision point and each decision point spawns a new thread to handle that particular case
 
user41796
At some point, you'll get convergence which should reset the number of running threads
 
7:44 PM
I think I'm starting to understand why scope creep has been such a problem.
 
user41796
@RobertHarvey Um, perhaps.... :-)
 
user41796
But it sounds like you've done everything you can reasonably do within the stream itself. Now you need to tweak the receiver so it can accurately guess which path that particular stream of bits followed.
 
I'm worried less about fixing the time than I am correlating the time with the inner packets. Fixing the outer time is actually a solved problem. :)
What I think I'm going to have to do is split the bit buffer into smaller buffers... one for each inner packet.
 
user41796
Are the inner packets sequential?
 
user41796
and no concerns with repeat inner packets?
 
7:46 PM
For purposes of this conversation, yes. :)
They are resequenced before they go into this part of the software.
I haven't figured out what to do about large dropouts yet. I think they will either be zero-filled, or hold last value.
 
user41796
If the inner packets have their own timestamp then you could generally ignore the outer timestamp and just use it for loose correlation
 
user41796
@RobertHarvey We use both approaches based upon what the client prefers
 
user41796
But we pick an approach and just stick with it.
 
user41796
Alternately, you can stick a "known bad" value in there if your data range supports it
 
@GlenH7 Right, basically the Correlated Frame serves that function. It holds a time stamp for each inner packet.
It sounds like you've worked with telemetry systems before. Familiar with Chapter 10?
The problem isn't so much calculating the time, it's figuring out how many bits have passed between the start of the outer packet and the start of the inner packet, without causing the bit buffer to explode (it has to be periodically resized, or I run out of memory).
 
user41796
7:50 PM
@RobertHarvey I just mentally circled back to needing the outer packet (and perhaps outer time?) in order to find the correct offsets for the inner packet...
 
And I thought some of my problems were insane.
 
user55340
@GlenH7 use recursion!
 
user41796
@MichaelT but only if you put the data section in harms way of your stack growth
 
user55340
(@ThomasOwens did you see the infographic I dropped in chat earlier?)
 
user55340
4 hours ago, by MichaelT
user image
 
user55340
7:52 PM
4 hours ago, by MichaelT
Full article http://embeddedgurus.com/state-space/2014/02/are-we-shooting-ourselves-in-the-fo‌​ot-with-stack-overflow/
 
user41796
@RobertHarvey I'm still toying with the idea of parallelizing de-ciphering the outer packet. Not strictly necessary, I realize, but it seems an easy way of hitting all the possible combinations in a short period of time. But if you're memory constrained that may be counter-productive.
 
user41796
Buy more RAM for the box? :-)
 
Hmm, no. But you did give me an idea.
There is another mode that these packets can be written in that is synced to the outer packet, and which does already contain the necessary timestamps. I could just convert the packets on-the fly to the better format.
 
user41796
I don't think I'm explaining the parallel approach very well. What I think I heard is there are N possible ways to decipher the outer packet depending upon what is believed to be the number of bits read and the previous packets. But there should be a correct answer. So you hammer all of those N possible ways at once to determine the correct answer and cause all of the other processing threads to reconcile
 
user41796
@RobertHarvey sounds simpler than my suggestion. :-)
 
7:56 PM
The problem really comes down to normalizing the input stream into something a bit more... sane.
 
user41796
Metaphor for Life there
 
And if I converted the packets to the normalized format, I wouldn't need a bit stream anymore (well, only during the conversion process, anyway).
 
user41796
Life == normalizing the input stream to something a bit more sane. :-)
 
user41796
@RobertHarvey overhead from the conversion is greatly outweighed by the development and processing time involved with the alternatives
 
Why aren't I already doing that? ... The Correlated Frame object does that (in an even more generic form), but it happens too far downstream to avoid the high coupling.
I think that's what the problem is.
It would be nice if I weren't spinning so many plates.
...
 
user41796
7:59 PM
@RobertHarvey You certainly sound like you're battling a lot of low level stuff. Sidestepping that is an elegant solution.
 
user55340
@RobertHarvey Add a variable gravity field for fun...
 
Thanks for the conversation. I really did need to talk about it with someone, and all of the folks here have their asses in their own alligators.
 

« first day (1265 days earlier)      last day (3724 days later) »