OK, so yesterday I got my first pay cheque for an entire month of 
employment. (Last month was only for 3 weeks, not 4.) With Christmas 
just around the corner, I'm really loving the 50% pay rise right about 
now. ;-)

It may be too early to say, but the WTF quotient of my new employer 
appears to be significantly lower than my previous one. I mean, there's 
still some minor brokenness there. But it's not like the last bunch of 
losers I worked for, where the company was run by deranged howler monkeys.

I've learned a couple of things so far. First, in almost two months, 
I've yet to write a single line of C# code. I have, however, written 
more Bash scripts than can possibly be healthy for my sanity. On Friday 
I discovered that "sudo" and "su -c" do not, in fact, do the same thing. 
(!!) I could have saved myself a lot of time if I'd known that sooner...

I have also learned enough about Vi that I can use it successfully 
approximately 80% of the time. (The other 20% of the time I open a file, 
start typing, and end up putting Vi into some weird mode that there's no 
way out of. Or I press escape when I'm already *in* command mode, which 
does something very odd.)

I have also learned that there's a thing called "Spotify", who's purpose 
appears to be to allow my boss to quickly locate terrible music with 
which to troll us all with. (Or just to replay Gangnam Style for the 
twenty-ninth time today...) Supposedly this is somehow legal. Also, I 
now know the words to The Troll Song.

Last week I brought in a lemon Victoria sponge. One senior member of 
staff remarked that it was worth hiring me for the cake alone. I hear 
rumours that my contract has been amended, making cake a condition of my 
continued employment.

(Some statistics: At my last place, I brought a cake, and with 23 people 
in the building, the cake lasted 3 days. A brought a similar cake to my 
current place of work, and it vanished in UNDER THREE HOURS. Less than 
10 people work here.)



Of course, The Real WTF is hiring. With unemployment at historic record 
levels, we literally cannot find *anybody* with the skill-set we need. 
Personally I find this baffling. The market must be /flooded/ with 
experienced programmers, and yet we keep interviewing people who can't 
program their way out of a paper bag.

Here's a challenge - write me a C# function (or even just some 
psuedocode) which will take a string containing space-delimited positive 
integers, and return them in ascending order.

We have yet to interview anybody who can actually accomplish this 
Herculean task. My personal favourite is the guy who, 25 minutes into 
the task (??!), decided to add a comment explaining what the function 
does. Because, hey, if you can't work out how to write the code, at 
least look like you know how to type, amIright?

(The best attempt featured a guy who managed to write a bubble-sort 
implementation but couldn't work out how to do the actual string splitting.)

We're using an online site where we can watch people type while we're on 
the phone to them, so save wasting our time with face-to-face 
interviews. Apparently the last guy we interviewed, the guys could hear 
typing but not see anything on the screen, and then suddenly big chunks 
of text would appear... It's /almost/ as if the guy was desperately 
Googling the code while he was on the phone.

It seems employment agents aren't just useless for employees; they seem 
to be pretty worthless for employers too. We suspect they're not sending 
us the good candidates because we're not paying enough. (Small company, 
small budgets. You know the drill.) One guy submitted a CV written in 
Comic Sans. My boss immediately refused to read it, and emailed the 
agent explaining that such a CV cannot possibly be taken seriously. He 
even helpfully included this little flowchart:

http://catbird.tumblr.com/image/216293561

The agent replied by taking the CV, changing the font, and emailing it 
back saying "there, I fixed the problem; hope that helps".

For real.

A professional job agent actually did this.

Of course, I've seen first-hand that there are *a lot* of people who 
can't program, and never will. But I would have thought that with an 
ocean of people looking for work, it shouldn't be too hard to find the 
minority who can. I WAS WRONG! >_<

Post a reply to this message

Orchid Win7 v1 <voi### [at] devnull> wrote:
> Here's a challenge - write me a C# function (or even just some 
> psuedocode) which will take a string containing space-delimited positive 
> integers, and return them in ascending order.

If it were C++, I could write a few lines of code that does that, although
with the caveat that I would first have to ask what the return type should
be and whether there should be any kind of error-checking or not (and if
yes, how errors should be notified to the calling code.) Space-delimited
integers in a string happens to be one of the easiest things to parse in C++.
(It would be much more complicated if the syntax were more complex.)

I assume there's an at least as easy way of doing it in C#, but since
I don't know the language I would fail the test.

-- 
                                                          - Warp

Post a reply to this message

On 02/12/2012 03:06 PM, Warp wrote:
> Orchid Win7 v1<voi### [at] devnull>  wrote:
>> Here's a challenge - write me a C# function (or even just some
>> psuedocode) which will take a string containing space-delimited positive
>> integers, and return them in ascending order.
>
> If it were C++, I could write a few lines of code that does that,

Or, indeed, any half-sane programming language you care to use, really...

> although
> with the caveat that I would first have to ask what the return type should
> be

This is unspecified. Another space-delimited string would be fine, a 
list or array of integers would be fine, really any "appropriate" 
container is fine. What we're really looking for is

1) You can figure out how to split a string and parse it into integers.
2) You can sort the result.
3) You can return it to the caller.

Not forgetting item 0 ("you realise that this is what you need to do to 
solve the overall problem"). For example, sorting the ASCII substrings 
won't give the correct answer, because the integers may have different 
numbers of digits. So you actually have to parse them. (I /suppose/ 
padding the substrings with zeros would achieve the same thing... but 
parsing an integer isn't exactly hard.)

> and whether there should be any kind of error-checking or not (and if
> yes, how errors should be notified to the calling code.)

That would certainly add bonus points. But then, about the only error I 
can think of is if the string contains characters other than spaces and 
digits. Returning null would probably be fine.

> Space-delimited
> integers in a string happens to be one of the easiest things to parse in C++.
> (It would be much more complicated if the syntax were more complex.)

The matter is similar in C# - a standard library function performs this 
exact task. Which, presumably, is why it was chosen.

> I assume there's an at least as easy way of doing it in C#, but since
> I don't know the language I would fail the test.

At this point, we're not even asking for correct class or method names. 
Just something that looks like vaguely valid C# syntax and has logic 
which isn't obviously gibberish.

This is how easy we're making it, and nobody has yet managed to pull it off.

I weep.

Post a reply to this message

On 02/12/2012 03:06 PM, Warp wrote:
> Orchid Win7 v1<voi### [at] devnull>  wrote:
>> Here's a challenge - write me a C# function (or even just some
>> psuedocode) which will take a string containing space-delimited positive
>> integers, and return them in ascending order.
>
> If it were C++, I could write a few lines of code that does that

Haskell:

   foobar :: String -> [Int]
   foobar = sort . map read . words


C#:

   public List<Int> foobar(String s)
   {
     string[] words = s.Split(' ');
     List<Int> list = new List<Int>();
     foreach (String word in words) {list.add(int.Parse(word));}
     list.Sort();
     return list;
   }


Not exactly taxing, eh?

Post a reply to this message

Orchid Win7 v1 <voi### [at] devnull> wrote:
> 1) You can figure out how to split a string and parse it into integers.

In this particular case I wouldn't split it at all, but parse the ints
directly from the string (which is quite easy to do since they are
whitespace-separated.)

Incidentally, there aren't any "string splitting" functions in C++ per se
(unless you count strtok() as being one.) If you literally need to create
a bunch of substrings from a given strings, there's no easier way than to
simply write the loop explicitly. (As much as it pains me to admit it,
parsing is not the forte of the C++ standard library, and it sometimes can
be a burden.)

In this case, however, it's just easier to parse the ints directly from the
input string, with no splitting or tokenization. (Basically there would be
two ways of doing that: Either use a stringstream, or use std::strtol() in
a loop. With a stringstream you could get fancy and use an input iterator
and call a standard algorithm to read the ints, thus eliding having to write
an explicit loop. If you want efficienty, though, std::strtol() is the
better choice, but it's more low-level.)

> 2) You can sort the result.

I can't imagine a programming language that doesn't offer sorting directly.
Even standard C offers such a function, and that's saying quite a lot.
In most languages it should be a one-liner (except in C, of course.)

If using the standard sorting function is disallowed for the purpose of
the exercise, then it becomes quite an academic question in itself. It
becomes a question of whether it needs to be efficient or short. A simple
insertion sort is a four-liner (I count each main element of the algorithm
as one line of code), but it's slow with large amounts of data. Faster
algorithms are harder to implement.

> > I assume there's an at least as easy way of doing it in C#, but since
> > I don't know the language I would fail the test.

> At this point, we're not even asking for correct class or method names. 
> Just something that looks like vaguely valid C# syntax and has logic 
> which isn't obviously gibberish.

I don't know enough C# to make it even look like it were. Unless C#
resembles a lot C++ (which it probably does.) :)

-- 
                                                          - Warp

Post a reply to this message

Orchid Win7 v1 <voi### [at] devnull> wrote:
> Haskell:

>    foobar :: String -> [Int]
>    foobar = sort . map read . words


> C#:

>    public List<Int> foobar(String s)
>    {
>      string[] words = s.Split(' ');
>      List<Int> list = new List<Int>();
>      foreach (String word in words) {list.add(int.Parse(word));}
>      list.Sort();
>      return list;
>    }

In C++ I would probably write it like:

    std::vector<int> foobar(const std::string& s)
    {
        std::istringstream is(s);
        std::vector<int> result;
        int value;
        while(is >> value) result.push_back(value);
        std::sort(result.begin(), result.end());
        return result;
    }

-- 
                                                          - Warp

Post a reply to this message

On 02/12/2012 04:21 PM, Warp wrote:

> In C++ I would probably write it like:
>
>      std::vector<int>  foobar(const std::string&  s)
>      {
>          std::istringstream is(s);
>          std::vector<int>  result;
>          int value;
>          while(is>>  value) result.push_back(value);
>          std::sort(result.begin(), result.end());
>          return result;
>      }

...and in much less than 25 minutes. Clearly you're better than any of 
the people we've interviewed. (But we knew that.)

It's also conspicuous that, unlike any of the people we interviewed, 
you're currently employed. :-P

Post a reply to this message

On 02/12/2012 04:11 PM, Warp wrote:
> Orchid Win7 v1<voi### [at] devnull>  wrote:
>> 1) You can figure out how to split a string and parse it into integers.
>
> In this particular case I wouldn't split it at all, but parse the ints
> directly from the string (which is quite easy to do since they are
> whitespace-separated.)

Two months of Bash scripting. That means we do this kind of thing all 
the time. It makes me twitch slightly when I think about it too much.

>> 2) You can sort the result.
>
> I can't imagine a programming language that doesn't offer sorting directly.
> Even standard C offers such a function, and that's saying quite a lot.

Like I say, one guy did start trying to write a bubble-sort by hand. (Of 
course, nobody else I work with sees anything wrong with that.) It does 
amuse me that somebody could know what a bubble-sort is, but /not/ know 
how to split a string, parse some integers and sort the result...

Of course, any sane person would use the built-in sorting facilities. 
Off the top of my head, the only languages I can think of that lack such 
a thing are PostScript, Tcl and BASIC. And who the hell would be using 
any of those? (OK, PostScript is used quite a lot - but *not* as a 
general-purpose programming language!)

>> At this point, we're not even asking for correct class or method names.
>> Just something that looks like vaguely valid C# syntax and has logic
>> which isn't obviously gibberish.
>
> I don't know enough C# to make it even look like it were. Unless C#
> resembles a lot C++ (which it probably does.) :)

Well, yeah, like Java it bears a strong superficial resemblance.



It's weird... I'm in a room full of programmers, and I seem to be the 
only person who knows anything at all about computer science. One day my 
boss says to me "I can't believe you don't know who J-Lo is". I replied 
"I can't believe you don't know the difference between a binary search 
tree and a binary heap tree". Every single human in the room stopped 
what they were doing, turned around and stared at me as if I'd just gone 
stark, raving mad. "What the *hell* are you talking about?" was my boss' 
reply.

Sometimes I weep.

Then again, in all honesty, how often do you need to implement a BST by 
yourself? Not very often...

Post a reply to this message

On Sun, 02 Dec 2012 12:30:30 +0000, Orchid Win7 v1 wrote:

> It may be too early to say, but the WTF quotient of my new employer
> appears to be significantly lower than my previous one. I mean, there's
> still some minor brokenness there. But it's not like the last bunch of
> losers I worked for, where the company was run by deranged howler
> monkeys.

It's good that you are seeing that what many of us told you for so long 
is in fact true - not all employers are as odd as the one you just 
left. :)

> I have also learned that there's a thing called "Spotify", who's purpose
> appears to be to allow my boss to quickly locate terrible music with
> which to troll us all with. (Or just to replay Gangnam Style for the
> twenty-ninth time today...) Supposedly this is somehow legal. Also, I
> now know the words to The Troll Song.

Spotify is great, subscribed recently myself - very large catalog of 
music, not all of it is as annoying as Gangnam Style or "other terrible 
music."

And yes, it is legal - you pay a monthly fee and get access to what's in 
their catalog.  Think Netflix or Lovefilm for music.

Jim

Post a reply to this message

>> It may be too early to say, but the WTF quotient of my new employer
>> appears to be significantly lower than my previous one. I mean, there's
>> still some minor brokenness there. But it's not like the last bunch of
>> losers I worked for, where the company was run by deranged howler
>> monkeys.
>
> It's good that you are seeing that what many of us told you for so long
> is in fact true - not all employers are as odd as the one you just
> left. :)

Or rather, I had assumed that I was working for *the only* broken 
company. But now it appears that each company is broken, it's just a 
question of how seriously or trivially.

> Spotify is great, subscribed recently myself - very large catalog of
> music, not all of it is as annoying as Gangnam Style or "other terrible
> music."

I think my boss just likes trolling people, BTH. ;-)

> And yes, it is legal - you pay a monthly fee and get access to what's in
> their catalog.

But not Pink Floyd or The Beatles, apparently... And yet, at the same 
time, unknown bands like The Senti-Mentals are there. Very odd.

> Think Netflix or Lovefilm for music.

Ah yes - two other services that I don't yet understand...

Post a reply to this message

On 12/2/2012 6:30 AM, Orchid Win7 v1 wrote:

> (Some statistics: At my last place, I brought a cake, and with 23 people
> in the building, the cake lasted 3 days. A brought a similar cake to my
> current place of work, and it vanished in UNDER THREE HOURS. Less than
> 10 people work here.)

A plague of locusts descends upon everything left in our coffee bar. 
Except for the bacon-flavored jelly beans.  Those lasted a while.

> Of course, The Real WTF is hiring. With unemployment at historic record
> levels, we literally cannot find *anybody* with the skill-set we need.
> Personally I find this baffling. The market must be /flooded/ with
> experienced programmers, and yet we keep interviewing people who can't
> program their way out of a paper bag.

What I noticed that that when the dot-com bubble crashed, a whole lot of 
firms became very selective in their hiring.  They wanted people who had 
experience with a specific environment or database.

> We have yet to interview anybody who can actually accomplish this
> Herculean task. My personal favourite is the guy who, 25 minutes into
> the task (??!), decided to add a comment explaining what the function
> does. Because, hey, if you can't work out how to write the code, at
> least look like you know how to type, amIright?

My company asked about the languages I knew, but I don't recall them 
testing knowledge of specific languages during the testing for the job.

> Of course, I've seen first-hand that there are *a lot* of people who
> can't program, and never will. But I would have thought that with an
> ocean of people looking for work, it shouldn't be too hard to find the
> minority who can. I WAS WRONG! >_<

The hard part is finding people who like coding enough that they won't 
get tired of a 40-hour week spent coding, and who are actually good 
enough at it that their code is worth hiring them for.

Regards,
John

Post a reply to this message

On 03/12/2012 02:17 AM, John VanSickle wrote:
> What I noticed that that when the dot-com bubble crashed, a whole lot of
> firms became very selective in their hiring. They wanted people who had
> experience with a specific environment or database.

Ah, so /that/ is where that came from? "Yes, I've been programming for 
20 years, writing code morning, noon and night, in half a dozen 
radically dissimilar languages. But no, I haven't actually used the one 
specific one you're asking about. Can I still be hired?" Most places 
immediately say no.

> My company asked about the languages I knew, but I don't recall them
> testing knowledge of specific languages during the testing for the job.

Several of the people we tested claim to have many, many years of C# 
experience. And yet none of them could write trivial programs like 
sorting integers or counting words or printing prime numbers.

> The hard part is finding people who like coding enough that they won't
> get tired of a 40-hour week spent coding, and who are actually good
> enough at it that their code is worth hiring them for.

IMHO, there are two kinds of coders:

The ones who go to college, take a Java course because they think 
there's money in it, learn the bare minimum to pass, get a job, and then 
completely stop learning. They write code from 9 to 5, go home and stop 
thinking about coding. /Everybody/ I went to uni with seemingly fell 
into that category. They all /hated/ coding, but it was on the syllabus, 
and they believed there was big money to be made with it.

On the other hand, there are people like me. I presume that doesn't 
require much further description. There are people who actually /like/ 
coding, are passionate about it, and voluntarily do it all day just for 
the /fun/ of it.

The latter are the people you want to hire.

One of the interview questions we use is "what hardware is in your PC?" 
If you're a computer enthusiast, you quite probably built your own PC, 
and even if you didn't, you'll almost certainly know if it's an Intel 
Core 2 or an AMD Athlon or whatever, how much RAM is has, and so forth. 
It's really very, very easy to tell the difference between somebody 
who's memorised a list and somebody who knows what they're talking about.

Knowing what hardware you have is of course in no way directly relevant 
to the job - but it's a nice way to gauge how much computer knowledge 
somebody has.

Post a reply to this message

> Of course, The Real WTF is hiring. With unemployment at historic record
> levels, we literally cannot find *anybody* with the skill-set we need.
> Personally I find this baffling. The market must be /flooded/ with
> experienced programmers, and yet we keep interviewing people who can't
> program their way out of a paper bag.

Hehe welcome to the real world. We have exactly the same problems with 
hiring Engineers. Can you describe what this circuit does? No, ok what 
about this one? No, ok what does this *component* do ... ermmm not 
exactly, have you every used an oscilloscope? No, ok we'll be in touch, 
thanks bye.

> Here's a challenge - write me a C# function (or even just some
> psuedocode) which will take a string containing space-delimited positive
> integers, and return them in ascending order.

I would have written almost exactly the same as you posted, but would 
have probably written this line instead (no reason other than I didn't 
know Int.Parse existed, and I don't normally use {} for single line 
loops...).

foreach (String word in words)
     list.add(Convert.ToInt32(word));

> We're using an online site where we can watch people type while we're on
> the phone to them, so save wasting our time with face-to-face
> interviews. Apparently the last guy we interviewed, the guys could hear
> typing but not see anything on the screen, and then suddenly big chunks
> of text would appear... It's /almost/ as if the guy was desperately
> Googling the code while he was on the phone.

Or typing it into the IDE then copying & pasting over, that's what I 
would probably do, as I pretty much rely on the auto-complete stuff to 
get code right first time (or even remember the method names correctly).

> Of course, I've seen first-hand that there are *a lot* of people who
> can't program, and never will. But I would have thought that with an
> ocean of people looking for work, it shouldn't be too hard to find the
> minority who can. I WAS WRONG! >_<

 From what was discussed here before (sorry don't remember who with) it 
seems recruitment agencies for software are very poor compared to ones 
for engineering for some reason. Seems like there is a market for a 
decent software recruitment agency that actually employs people who know 
a tiny bit how to program so can weed out people like you describe.

Also as a software company you should be able to knock up a quick 
website that has programming exercises (like the one above) that the 
candidate has to fill in within a set time, design them so you can't 
just google the answer. Send the link to people with a unique login 
before you even invite them to interview...

Post a reply to this message

> I've learned a couple of things so far. First, in almost two months,
> I've yet to

...install Thunderbird.

Post a reply to this message

Orchid Win7 v1 <voi### [at] devnull> wrote:
> It's also conspicuous that, unlike any of the people we interviewed, 
> you're currently employed. :-P

I have had the great fortune of never having to work with someone who
isn't a competent programmer. My first job was at the university, in the
deparment of computer science, so you might imagine the competence of
the average programmer working there. (In my current job I basically work
alone on the code. We have another developer but he develops unrelated
projects. Nevertheless, he's quite competent himself too.)

-- 
                                                          - Warp

Post a reply to this message

On Sun, 02 Dec 2012 21:40:55 +0000, Orchid Win7 v1 wrote:

>>> It may be too early to say, but the WTF quotient of my new employer
>>> appears to be significantly lower than my previous one. I mean,
>>> there's still some minor brokenness there. But it's not like the last
>>> bunch of losers I worked for, where the company was run by deranged
>>> howler monkeys.
>>
>> It's good that you are seeing that what many of us told you for so long
>> is in fact true - not all employers are as odd as the one you just
>> left. :)
> 
> Or rather, I had assumed that I was working for *the only* broken
> company. But now it appears that each company is broken, it's just a
> question of how seriously or trivially.

All companies face challenges, certainly.  People are involved.

>> Spotify is great, subscribed recently myself - very large catalog of
>> music, not all of it is as annoying as Gangnam Style or "other terrible
>> music."
> 
> I think my boss just likes trolling people, BTH. ;-)
> 
>> And yes, it is legal - you pay a monthly fee and get access to what's
>> in their catalog.
> 
> But not Pink Floyd or The Beatles, apparently... And yet, at the same
> time, unknown bands like The Senti-Mentals are there. Very odd.

It's called "licensing fees", and some of the very popular bands (and 
their labels) want extortionate fees to be included in the catalog.

But I did find some Floyd on Slacker, another similar service that I also 
subscribe too.

> 
>> Think Netflix or Lovefilm for music.
> 
> Ah yes - two other services that I don't yet understand...

What's to understand?  I want to watch a movie, it's in their catalog, I 
stream it to my PS3 and watch it.  I pay a monthly fee and can stream as 
much as I want.

Jim

Post a reply to this message

Orchid Win7 v1 <voi### [at] devnull> wrote:
> > Think Netflix or Lovefilm for music.

> Ah yes - two other services that I don't yet understand...

You don't understand why people would pay a moderate sum of money in
order to be able to watch movies or listen to music? (And *legally* at
that...)

-- 
                                                          - Warp

Post a reply to this message

On 03/12/2012 12:20 PM, scott wrote:
>> I've learned a couple of things so far. First, in almost two months,
>> I've yet to
>
> ...install Thunderbird.

At works? Are you nuts?? They have aggressive network surveillance 
systems. It would probably be grounds for immediate termination. :-P

Post a reply to this message

On 3/12/2012 2:54 AM, Orchid Win7 v1 wrote:
> On 02/12/2012 03:06 PM, Warp wrote:
>> Orchid Win7 v1<voi### [at] devnull>  wrote:
>>> Here's a challenge - write me a C# function (or even just some
>>> psuedocode) which will take a string containing space-delimited positive
>>> integers, and return them in ascending order.
>>
>> If it were C++, I could write a few lines of code that does that
>
> Haskell:
>
>    foobar :: String -> [Int]
>    foobar = sort . map read . words
>
>
> C#:
>
>    public List<Int> foobar(String s)
>    {
>      string[] words = s.Split(' ');
>      List<Int> list = new List<Int>();
>      foreach (String word in words) {list.add(int.Parse(word));}
>      list.Sort();
>      return list;
>    }
>
>
> Not exactly taxing, eh?

Of course there are many ways to approach this in C# and most other 
languages.  Using LINQ and and working in decimal datatype without error 
handling:

private List<decimal> ParseAndSortValuesFromString1(string text)
{
	return text.Split(' ').Select(s => Decimal.Parse(s)).OrderBy(s => 
s).ToList();
}

Post a reply to this message

>> Not exactly taxing, eh?
>
> Of course there are many ways to approach this in C# and most other
> languages. Using LINQ and and working in decimal datatype without error
> handling:
>
> private List<decimal> ParseAndSortValuesFromString1(string text)
> {
> return text.Split(' ').Select(s => Decimal.Parse(s)).OrderBy(s =>
> s).ToList();
> }

According to one of the guys I work with, "LINQ is great and everything, 
but it makes the code extremely slow". Does anybody have any actual 
/facts/ with which to confirm or refute this? I would have expected 
there to be a fairly minimal speed difference...

Post a reply to this message

On Tue 04/12/12 19:09, Orchid Win7 v1 wrote:
>>> Not exactly taxing, eh?
>>
>> Of course there are many ways to approach this in C# and most other
>> languages. Using LINQ and and working in decimal datatype without error
>> handling:
>>
>> private List<decimal> ParseAndSortValuesFromString1(string text)
>> {
>> return text.Split(' ').Select(s => Decimal.Parse(s)).OrderBy(s =>
>> s).ToList();
>> }
>
> According to one of the guys I work with, "LINQ is great and everything,
> but it makes the code extremely slow". Does anybody have any actual
> /facts/ with which to confirm or refute this? I would have expected
> there to be a fairly minimal speed difference...

LINQ has a certain amount of overhead, as to whether that significantly 
impacts the speed of your code obviously depends on the exact code and 
data you're dealing with. Even in your example I suspect the speed 
differential is highly dependent on the size of the list.

Post a reply to this message

Orchid Win7 v1 <voi### [at] devnull> wrote:
>    public List<Int> foobar(String s)

Btw, why return a list?

I know that memory allocation in C# is probably faster than it is
in C++, but it still feels odd to use a list instead of something more
memory-efficient (which C# most certainly ought to support) and faster
to access. What possible advantage could there be in using a list of
ints?

Are you transferring your customs of using lists from Haskell to C#?

-- 
                                                          - Warp

Post a reply to this message

Le 2012-12-03 12:15, Warp a écrit :
> Orchid Win7 v1 <voi### [at] devnull> wrote:
>>> Think Netflix or Lovefilm for music.
>
>> Ah yes - two other services that I don't yet understand...
>
> You don't understand why people would pay a moderate sum of money in
> order to be able to watch movies or listen to music? (And *legally* at
> that...)
>
No, he doesn't understand how a service can offer near-HD resolution 
over the Internet.  Despite the fact that tv providers have had no 
problem doing 60 channels of the same thing over the same wires for the 
last 30 years (10 years in the case of HD).

He also complains of not knowing what "cable tv" is, despite the fact 
that he admitted his parents used to have it.

You, know, regular Andy stuff.

;)

-- 
/*Francois Labreque*/#local a=x+y;#local b=x+a;#local c=a+b;#macro P(F//
/*    flabreque    */L)polygon{5,F,F+z,L+z,L,F pigment{rgb 9}}#end union
/*        @        */{P(0,a)P(a,b)P(b,c)P(2*a,2*b)P(2*b,b+c)P(b+c,<2,3>)
/*   gmail.com     */}camera{orthographic location<6,1.25,-6>look_at a }

Post a reply to this message

On 05/12/2012 02:22 PM, Warp wrote:
> Orchid Win7 v1<voi### [at] devnull>  wrote:
>>     public List<Int>  foobar(String s)
>
> Btw, why return a list?
>
> I know that memory allocation in C# is probably faster than it is
> in C++, but it still feels odd to use a list instead of something more
> memory-efficient (which C# most certainly ought to support) and faster
> to access. What possible advantage could there be in using a list of
> ints?
>
> Are you transferring your customs of using lists from Haskell to C#?

The C# "List" class actually represents a growable array. So it has O(1) 
index and O(1) append until the underlying array needs to be enlarged 
(which is obviously O(n) time).

The "LinkedList" class represents a (double-linked) list, with the usual 
performance characteristics you would expect for that container.

As best as I can tell, List is pretty much the default container to use 
unless you need something more specific.

You could also use a normal array. But that isn't resizable (which 
doesn't matter in this particular case).

Post a reply to this message

>> You don't understand why people would pay a moderate sum of money in
>> order to be able to watch movies or listen to music?
>>
> No, he doesn't understand how a service can offer near-HD resolution
> over the Internet.

Providing data over the Internet isn't hard. But providing HD video *in 
realtime* would seem difficult given that people don't have 
megabit-speed Internet access yet.

> Despite the fact that tv providers have had no
> problem doing 60 channels of the same thing over the same wires for the
> last 30 years (10 years in the case of HD).

Digital TV isn't 30 years old. :-P

Besides, it's no secret that a simplex broadcast channel is much easier 
to set up than a full-duplex communications channel.

Post a reply to this message

On 03/12/2012 02:46 PM, Warp wrote:
> I have had the great fortune of never having to work with someone who
> isn't a competent programmer.

I've only met people who are very bad with computers. When I went to 
university, I seemed to know more than most of the lectures I was paying 
to teach me, which is always a little bit worrying.

Bafflingly, all the people in my class appeared to hate computers, which 
begs the question "why the hell are you here??"

This job is my first experience of working with people who have a clue. 
And even they think I'm a bit strange for having all this CS knowledge...

Post a reply to this message

>> According to one of the guys I work with, "LINQ is great and everything,
>> but it makes the code extremely slow". Does anybody have any actual
>> /facts/ with which to confirm or refute this? I would have expected
>> there to be a fairly minimal speed difference...
>
> LINQ has a certain amount of overhead, as to whether that significantly
> impacts the speed of your code obviously depends on the exact code and
> data you're dealing with. Even in your example I suspect the speed
> differential is highly dependent on the size of the list.

I would expect that for this example question, you would be very 
hard-pressed to construct a program inefficient enough to be considered 
"too slow". ;-)

You could perhaps write the program in the Iota calculus, running in an 
Iota interpreter written in Bash script. That might perhaps be slow 
enough...

Post a reply to this message

On 03/12/2012 08:50 AM, scott wrote:
>> Of course, The Real WTF is hiring. With unemployment at historic record
>> levels, we literally cannot find *anybody* with the skill-set we need.
>> Personally I find this baffling. The market must be /flooded/ with
>> experienced programmers, and yet we keep interviewing people who can't
>> program their way out of a paper bag.
>
> Hehe welcome to the real world. We have exactly the same problems with
> hiring Engineers. Can you describe what this circuit does? No, ok what
> about this one? No, ok what does this *component* do ... ermmm not
> exactly, have you every used an oscilloscope? No, ok we'll be in touch,
> thanks bye.

This is what baffles me.

Consider the following conversation...

Mr Fail: Hi. I've got 20 years' experience as a brick layer. Are you hiring?

Boss: Oh really? Well, can you just quickly tell me, is that wall over 
there stretcher-bond of Flemish-bond?

Mr Fail: Uh... well, I really specialised in /internal/ walls.

Boss: Great! So tell me, at what height should the internal 
damp-proofing course be run?

Mr Fail: Well, I, er, mostly worked in tropical countries where damp 
isn't a problem.

Boss: Cool. So how many runs of bricks between tie-in wraps?

Mr Fail: Tie-in wraps?

Boss: Goodbye.

The thing is, *nobody* thinks they can just walk in and pretend to know 
how to be a brick layer. Because it's ****ing obvious that it'll take 
the people interviewing you about 11 seconds to figure out that you know 
nothing about anything.

So *why* the hell does this constantly happen in computing?!? >_<

>> It's /almost/ as if the guy was desperately
>> Googling the code while he was on the phone.
>
> Or typing it into the IDE then copying & pasting over, that's what I
> would probably do, as I pretty much rely on the auto-complete stuff to
> get code right first time (or even remember the method names correctly).

Yeah, but I'm fairly sure even VS doesn't write entire blocks of code 
for you. Google searching does. ;-)

> From what was discussed here before (sorry don't remember who with) it
> seems recruitment agencies for software are very poor compared to ones
> for engineering for some reason. Seems like there is a market for a
> decent software recruitment agency that actually employs people who know
> a tiny bit how to program so can weed out people like you describe.

I'm currently unsure as to whether we're just not paying the agents 
enough money, or whether there's nobody out there to recruit.

(As a small company, we have very little money to spend. And the owners 
are the type of people who see job agents as a totally unnecessary 
expense and want to spend the absolute minimum possible on them.)

With such high unemployment, you would THINK there's loads of people out 
there... but we're beginning to suspect that maybe all the /competent/ 
people get snapped up faster than we can find them...

Post a reply to this message

On 05/12/2012 9:23 PM, Orchid Win7 v1 wrote:
> Providing data over the Internet isn't hard. But providing HD video *in
> realtime* would seem difficult given that people don't have
> megabit-speed Internet access yet.


WTF!

-- 
Regards
     Stephen

Post a reply to this message

On Wed, 05 Dec 2012 21:23:24 +0000, Orchid Win7 v1 wrote:

>>> You don't understand why people would pay a moderate sum of money in
>>> order to be able to watch movies or listen to music?
>>>
>> No, he doesn't understand how a service can offer near-HD resolution
>> over the Internet.
> 
> Providing data over the Internet isn't hard. But providing HD video *in
> realtime* would seem difficult given that people don't have
> megabit-speed Internet access yet.

m-/

That is where you're wrong, Andy.  Even my pokey internet connection is 3 
Mbps.  More than enough to handle sufficient data to do a medium quality 
HD picture on a 10' screen.

>> Despite the fact that tv providers have had no problem doing 60
>> channels of the same thing over the same wires for the last 30 years
>> (10 years in the case of HD).
> 
> Digital TV isn't 30 years old. :-P

No, but cable TV is close to that.  We had cable when I was a teenager, 
and easily had 120 channels at that time (that would be in the 80's).

That's not really classified information, either.

Jim

Post a reply to this message

On Wed, 05 Dec 2012 21:36:09 +0000, Orchid Win7 v1 wrote:

> So *why* the hell does this constantly happen in computing?!? >_<

Because often times the managers hiring computing professionals do not 
have a background in computing.

That's less and less common now, but when I started in IT, it was very 
common, and a lot of people talked volumes of BS about it and got hired.

Jim

Post a reply to this message

On 12/5/2012 1:27 PM, Orchid Win7 v1 wrote:
>
> I would expect that for this example question, you would be very
> hard-pressed to construct a program inefficient enough to be considered
> "too slow". ;-)

That's because you're not thinking like someone who has no idea what 
they're doing.

Take for instance the following algorithm, which I *tried* to make mimic 
the sort of answer a clueless person might stumble and grope their way 
to after hours of trial and error.  I couldn't manage make the coding 
style quite bad enough to be realistic, but you'll get the idea.  I did 
at least manage to put in a bug or two which ends up working due to 
blind luck.

example of use:

sortInts("5000 353 1  145 145  145 354 23 16564")


the (horrible horrible) code:



def sortInts(S):
	i = 0
	count = 0
	while count < countSpaces(S)+1:
		n = countInStr(i, S)
		if n > 0:
			for j in range(n):
				print i
				count += 1
		i += 1

def countInStr(n, S):
	S = " " + S + " "
	count = 0
	for i in range(len(S)):
		for j in range(i,len(S)):
			if S[i-1:j+1] == " " + str(n) + " ":
				count += 1
	return count
	
def countSpaces(S):
	count = 0
	wasLastSpace = False
	for i in range(len(S)):
		if S[i] == " ":
			if wasLastSpace == False:
				count += 1
				wasLastSpace = True
		else:
			wasLastSpace = False
	return count

Post a reply to this message

Orchid Win7 v1 <voi### [at] devnull> wrote:
> The C# "List" class actually represents a growable array. So it has O(1) 
> index and O(1) append until the underlying array needs to be enlarged 
> (which is obviously O(n) time).

Why is C# using the wrong terminology? (Or at the very least, really
misleading?)

What that sounds more like is a dynamic array, not a list.

-- 
                                                          - Warp

Post a reply to this message

Orchid Win7 v1 <voi### [at] devnull> wrote:
> Providing data over the Internet isn't hard. But providing HD video *in 
> realtime* would seem difficult given that people don't have 
> megabit-speed Internet access yet.

I think you are confusing the word "people" with the word "I".

-- 
                                                          - Warp

Post a reply to this message

Le 2012-12-05 16:23, Orchid Win7 v1 a écrit :
>>> You don't understand why people would pay a moderate sum of money in
>>> order to be able to watch movies or listen to music?
>>>
>> No, he doesn't understand how a service can offer near-HD resolution
>> over the Internet.
>
> Providing data over the Internet isn't hard. But providing HD video *in
> realtime* would seem difficult given that people don't have
> megabit-speed Internet access yet.

I currently have 10Mbps downstream/1.5Mbps upstream.  And both my 
provider and its direct competitor offer better packages.

>
>> Despite the fact that tv providers have had no
>> problem doing 60 channels of the same thing over the same wires for the
>> last 30 years (10 years in the case of HD).
>
> Digital TV isn't 30 years old. :-P

That's why I said 10 years for HD.  And besides, digital channels take 
the same bandwidth spectrum as analog tv channels.

>
> Besides, it's no secret that a simplex broadcast channel is much easier
> to set up than a full-duplex communications channel.

That's why it's asymmetrical.  And also why most ISPs frown upon you 
hosting servers out of your basement, since you would be hogging the 
upstream channel.

-- 
/*Francois Labreque*/#local a=x+y;#local b=x+a;#local c=a+b;#macro P(F//
/*    flabreque    */L)polygon{5,F,F+z,L+z,L,F pigment{rgb 9}}#end union
/*        @        */{P(0,a)P(a,b)P(b,c)P(2*a,2*b)P(2*b,b+c)P(b+c,<2,3>)
/*   gmail.com     */}camera{orthographic location<6,1.25,-6>look_at a }

Post a reply to this message

Le 2012-12-05 16:36, Orchid Win7 v1 a écrit :
> The thing is, *nobody* thinks they can just walk in and pretend to know
> how to be a brick layer. Because it's ****ing obvious that it'll take
> the people interviewing you about 11 seconds to figure out that you know
> nothing about anything.
>
> So *why* the hell does this constantly happen in computing?!? >_<
>

Because a lot of people don't even know what they don't know.

Afew years ago, the customer I was working for had their own cafeteria. 
  At one point, they decided to outsource the services to an outside 
supplier.  The cafeteria cooks, being members of the company's union, 
were given the choice of either accepting a non-union job at the new 
supplier (at a lower salary, of course), or put themselves on the union 
avaialability list in the hope that another job would become available 
in another dept.

One of the cooks asked me one day:  "You're in IT, right?  Do you think 
I could apply in your dept?  I'm pretty good with my computer at home..."

Me: "Do you know the difference between the IP address classes?"
Him:  [quizzical dog look]
Me: (Ok networking is not for him) "Have you ever played with Linux?"
Him:  [quizzical dog look]
Me: (Ok, hmm... ) "Do you have a home network?"
Him: Huh... No.  But I customized the hell out of my MySpace page.
Me: "I'll get back to you."

He was really disappointed when I told him we were also an outside 
supplier so he couldn't simply waltz in an bump one of the lower 
seniority guys.

There was also the girl who wrote in her presentation letter that she 
was a networking expert, (her cv mentioned her only experience as having 
set up a 10 PC internet cafe), and that she knew both versions of 
Windows inside and out (Win 98 *and* Win2000... we're talking serious IT 
Geekage, here!)

-- 
/*Francois Labreque*/#local a=x+y;#local b=x+a;#local c=a+b;#macro P(F//
/*    flabreque    */L)polygon{5,F,F+z,L+z,L,F pigment{rgb 9}}#end union
/*        @        */{P(0,a)P(a,b)P(b,c)P(2*a,2*b)P(2*b,b+c)P(b+c,<2,3>)
/*   gmail.com     */}camera{orthographic location<6,1.25,-6>look_at a }

Post a reply to this message

Francois Labreque <fla### [at] videotronca> wrote:
> There was also the girl who wrote in her presentation letter that she 
> was a networking expert, (her cv mentioned her only experience as having 
> set up a 10 PC internet cafe), and that she knew both versions of 
> Windows inside and out (Win 98 *and* Win2000... we're talking serious IT 
> Geekage, here!)

OTOH, the reverse also is problematic. In other words, if you are
completely and absolutely *honest* about your expertise, you might not
get even a job interview.

For instance, if I were seeking for a job that requires PHP skills, and
I would write honestly that I have only rudimentary experience of said
language, they would most probably not even interview me, instead opting
for someone who *claims* to have extensive knowledge and experience.

However, if I outright lied and told them that I have a lot of experience
on PHP, and I got the job, I'm pretty certain that I would learn the
tidbits quite quickly. (The language itself is really close to C++, and
basically the only thing you need to learn is the few syntactical and
semantical oddities and the different libraries.)

-- 
                                                          - Warp

Post a reply to this message

On Thu, 06 Dec 2012 09:37:04 -0500, Francois Labreque wrote:

> That's why I said 10 years for HD.  And besides, digital channels take
> the same bandwidth spectrum as analog tv channels.

IIRC, in the US at least, digital takes less spectrum.

Jim

Post a reply to this message

On 06/12/2012 07:54 AM, Warp wrote:
> Orchid Win7 v1<voi### [at] devnull>  wrote:
>> The C# "List" class actually represents a growable array. So it has O(1)
>> index and O(1) append until the underlying array needs to be enlarged
>> (which is obviously O(n) time).
>
> Why is C# using the wrong terminology? (Or at the very least, really
> misleading?)
>
> What that sounds more like is a dynamic array, not a list.

Yeah, I don't really know.

Java has a /interface/ called "list", which represents "a list of 
things". It then provides several different implementations of this 
interface, including "LinkedList" and "ArrayList".

C# seems to have copied this terminology, having an IList interface and 
a List class which implements it. But it's certainly nothing to do with 
linked lists (which don't implement IList at all, and hence don't even 
*offer* element indexing as a build-in method).

Go sue the designers. :-P

Post a reply to this message

On 06/12/2012 03:48 PM, Warp wrote:
> OTOH, the reverse also is problematic. In other words, if you are
> completely and absolutely *honest* about your expertise, you might not
> get even a job interview.

This is the problem. The market is so utterly flooded with "IT experts" 
who claim to know everything about everything, that if you write on your 
CV what you *actually* know about, nobody will even look at it. That's 
utterly messed up.

Seriously, how many people apply to work as an engine designer when they 
don't know anything about physics?

Post a reply to this message

On 06/12/2012 02:57 PM, Francois Labreque wrote:
> Le 2012-12-05 16:36, Orchid Win7 v1 a écrit :
>> The thing is, *nobody* thinks they can just walk in and pretend to know
>> how to be a brick layer. Because it's ****ing obvious that it'll take
>> the people interviewing you about 11 seconds to figure out that you know
>> nothing about anything.
>>
>> So *why* the hell does this constantly happen in computing?!? >_<
>
> Because a lot of people don't even know what they don't know.

Sure. But why is that peculiar to IT?

Nobody out there thinks they could totally draw up the blueprints for a 
suspension bridge and have it actually work. Yet people think they can 
write commands to make a computer perform a complex task and it'll be 
fine. WTF is up with that?

> Me: "Do you know the difference between the IP address classes?"
> Him: [quizzical dog look]
> Me: (Ok networking is not for him) "Have you ever played with Linux?"
> Him: [quizzical dog look]
> Me: (Ok, hmm... ) "Do you have a home network?"
> Him: Huh... No. But I customized the hell out of my MySpace page.
> Me: "I'll get back to you."

Yeah, this is pretty standard. Everybody understands that designing 
(say) an aeroplane is really hard. And yet people think that designing 
computer software is somehow trivially easy...

This probably isn't helped by the following fact: If you pay somebody to 
build a skyscraper, and they actually build a small wooden hut, you know 
that you did not get what you paid for. If you pay somebody to build an 
enterprise-class data management engine and they actually give you an 
Excel spreadsheet and an instruction manual, you might not necessarily 
realise that something is wrong - and neither might they...

Post a reply to this message

On Thu, 06 Dec 2012 20:54:43 +0000, Orchid Win7 v1 wrote:

> Seriously, how many people apply to work as an engine designer when they
> don't know anything about physics?

It's a little different when the results aren't tangible, I find.  As in 
something physical that someone can look at and test.  When something is 
non-physical, I find people tend to think it's easier than it is to go 
from concept to implementation.

Jim

Post a reply to this message

> Why is C# using the wrong terminology? (Or at the very least, really
> misleading?)

According to Wikipedia:

"Lists are typically implemented either as linked lists (either singly 
or doubly linked) or as arrays, usually variable length or dynamic arrays."

Post a reply to this message

> The thing is, *nobody* thinks they can just walk in and pretend to know
> how to be a brick layer. Because it's ****ing obvious that it'll take
> the people interviewing you about 11 seconds to figure out that you know
> nothing about anything.

But seriously that is what it was like with the Engineer guy who we 
interviewed. After that we decided to always do a very short telephone 
interview first (even if they lived round the corner) rather than 
wasting several man-hours of interviewing.

> So *why* the hell does this constantly happen in computing?!? >_<

To be fair the guy I mentioned applied directly, the ones through the 
agency were always pretty much there. It just strikes me as odd that you 
can't find a decent agency to find the right people.

>> Or typing it into the IDE then copying & pasting over, that's what I
>> would probably do, as I pretty much rely on the auto-complete stuff to
>> get code right first time (or even remember the method names correctly).
>
> Yeah, but I'm fairly sure even VS doesn't write entire blocks of code
> for you. Google searching does. ;-)

I meant I would write the block in VS and then coyp&past the whole block 
to the browser window.

> I'm currently unsure as to whether we're just not paying the agents
> enough money, or whether there's nobody out there to recruit.

There's always good people out there to recruit, you just need to reach 
them (monster or similar is easy nowadays, or a decent agent) and they 
must want to work for your company (location, salary, if they are 
interested in what you do, etc).

> (As a small company, we have very little money to spend. And the owners
> are the type of people who see job agents as a totally unnecessary
> expense and want to spend the absolute minimum possible on them.)

That's why then. Advertise on monster and ask them a few coding 
questions as part of their application on monster (you can set it up to 
do that and email you their answers with their cv). Sure some will try 
and google it then copy&paste, but you should be creative enough to 
think of something that can't be googled.

Post a reply to this message

> Providing data over the Internet isn't hard. But providing HD video *in
> realtime* would seem difficult given that people don't have
> megabit-speed Internet access yet.

LOL that comment would hardly make sense even a decade ago!

Post a reply to this message

>> Why is C# using the wrong terminology? (Or at the very least, really
>> misleading?)
>
> Yeah, I don't really know.
>
> Java has a /interface/ called "list", which represents "a list of
> things". It then provides several different implementations of this
> interface, including "LinkedList" and "ArrayList".
>
> C# seems to have copied this terminology, having an IList interface and
> a List class which implements it. But it's certainly nothing to do with
> linked lists (which don't implement IList at all, and hence don't even
> *offer* element indexing as a build-in method).

It's worse...

C# actually *has* an ArrayList class. But that was written before they 
added generics. (I.e., an ArrayList holds only Objects.) So when they 
added generics, ArrayList became just List.

Yuck.

Post a reply to this message

>> (As a small company, we have very little money to spend. And the owners
>> are the type of people who see job agents as a totally unnecessary
>> expense and want to spend the absolute minimum possible on them.)
>
> That's why then.

Really? Is that all it is?

> Advertise on monster

In my limited experience, Monster is an utter waste of time for 
computing jobs. (The whole "there are no jobs in MK" came from my using 
Monster to try to find work. Once I switched to a *real* jobs site, I 
found plenty of stuff to apply for...)

Post a reply to this message

>>> (As a small company, we have very little money to spend. And the owners
>>> are the type of people who see job agents as a totally unnecessary
>>> expense and want to spend the absolute minimum possible on them.)
>>
>> That's why then.
>
> Really? Is that all it is?

Peanuts and monkeys. IME not all recruitment agencies are created equal...

Post a reply to this message

INTERVIEWER: "Hi. Welcome to ACME Software Corporation. We're a cutting edge
software design company. Tell me, do you have state-of-the-art programming
skills?"

APPLICANT (Kenneth): "Oh sure, lots of skills and experience."

INTERVIEWER: "I see. Such as...?"

APPLICANT: "Six years writing super-complex code in POV-Ray, using the Scene
Description Language! Oh, and Fortran IV, of course."

INTERVIEWER: " (...???...)  I see. Well, just a few questions. Are you familiar
with classes?"

APPLICANT: " Yes, yes, I took some of those in High School and College."

INTERVIEWER: " (....?!?!?!....) Ahem. OK. So, do you even know what a for-loop
is?"

APPLICANT: "Is that like a #while loop?"

INTERVIEWER: "Hmm. So, moving right along...How about any experience with
Python, C#, Lisp, C++ or .NET?"

APPLICANT: "Um, yeah, I've heard of those. And don't forget BASIC!"

INTERVIEWER: "No, no, of course not...well...uh..."

APPLICANT: "So when do I start?!"

Post a reply to this message

Le 2012-12-06 12:32, Jim Henderson a écrit :
> On Thu, 06 Dec 2012 09:37:04 -0500, Francois Labreque wrote:
>
>> That's why I said 10 years for HD.  And besides, digital channels take
>> the same bandwidth spectrum as analog tv channels.
>
> IIRC, in the US at least, digital takes less spectrum.
>

I thought analog channels were 6MHz wide as well, but couldn't find any 
actual numbers after a 3.5 nanosecond wiki search.


-- 
/*Francois Labreque*/#local a=x+y;#local b=x+a;#local c=a+b;#macro P(F//
/*    flabreque    */L)polygon{5,F,F+z,L+z,L,F pigment{rgb 9}}#end union
/*        @        */{P(0,a)P(a,b)P(b,c)P(2*a,2*b)P(2*b,b+c)P(b+c,<2,3>)
/*   gmail.com     */}camera{orthographic location<6,1.25,-6>look_at a }

Post a reply to this message