It is a general purpose scripting language.

An object-oriented general purpose scripting language with a number of features that makes system administration easier.

One example is DSC. It is a scripting language that can use the DSC *platform* to make sure that target systems are all configured the same way, albeit each with different parameters.

Another example is workflows. Wake me up when bash or python can start a script that can survive system restarts and pick up and continue from where it was when the system restarted, complete with state, variables etc.

Nah. In PowerShell:

iwr "http://server/path/to/file.zip" -outf "${env:TEMP}/file.zip"

You can try to make it look more verbose. But expect to be called out on it.

You can accomplish almost any task with fewer keystrokes in PowerShell compared to bash - and the resulting command will still be more readable.

Desired State Configuration (DSC) that FTFA was about, is definitely one such thing that PowerShell has that python has not. DSC is a *declarative* description of the configuration you want for a target system. You should think more in line of Chef or Puppet than Python. PowerShell DSC for Linux actually *uses* Python.

The idea is that you use PowerShell to define a data structure (much like a Ruby hash) that describes the configuration of the node. DSC will itself resolve dependencies. If you require a feature DSC will ensure that the feature is installed - much like a package manager - but it actually interacts with the package manager. What package managers do not do is to configure the products once they are installed. This could be connection strings, IP addresses, user accounts.

PowerShell DSC for Linux has "resources" for file system, user accounts, text file content, package managers (Yum, Apt, Zypper), scripts, daemons, ssh keys and more. You use those resources to describe how you want a system to look - like a Chef recipe. The resource description can be parameterized (it is just a PowerShell function and can take parameters like PS functions) so that the same resource description can be used for multiple targets with slightly different values.

Once applied, DSC will ensure that the target is set up so that it matches the target. From there on it can also report on drift (e.g. more users created, files deleted/changed etc) and can warn about it and automatically bring the node back to the desired state (undoing the drift).

How about if the unix app allowed only the long form option names - but allowed them to be abbreviated as long as the abbreviation was unambigous? (That's what powershell does)

That why we still code in assembler and don't use those modern touch screens. Oh wait... (lalalalalalalal! -- fingers in ears, eyes firmly closed)

Yes, there is no concept of SUID/setuid on Windows. So there's no sudo "happy go lucky".

No, PowerShell is aimed at admins who want *robust* scripts - both the ad-hoc ones they whip up as well as the ones they choose to save. String parsing is extremely brittle, and most bash shell scipters do it the insecure and brittle way because it is easier.

String parsing is often thrown off if presented with unusual characters in file names, if executed in locales where dates and numbers are both generated and parsed different, etc.

OMI is available on Linux. Read the FTFA

You're the incompetent one. There's is no "IDE in a shell". The ISE *is* the shell - much like if you did bash scripting from emacs. The difference is that the ISE will provide you with intellisense (automatic suggestions), help, syntax highlighting, snippets, multiple script panels, integrated source-level debugging (complete with breakpoints, variable inspection etc) and even a command "builder".

Most *nix shells have good documentation. PowerShell has good documentation as well. All of the cmdlets have syntax descriptions (automatically generated from metadata), description and multiple examples. In PowerShell even user-defined functions, cmdlets and script files can have the same level of documentation. Comment based help (look it up) makes it super easy to document scripts and functions. And the auto-generated syntax diagrams and parameter descriptions also work for your own script files.

PowerShells command syntax is extremely consistent. Cmdlets are *always* of the verb-noun form, and there are only about 40 or so standard "approved" verbs. Parsing of the command parameters is the responsibility of the *shell* not of the commands like in *nix shells. Hence, all commands follow the same convention with no strange outliers like e.g. dd. Parameter names are always "long" - but can be abbreviated as long as the abbreviation is unambigous.

No, I can't. I started out and I sucked. I got better eventually through experience. In order for it to be truly bimodal, people have to start in either camp A or camp B and end in the same camp they started in. Because if you transition from one to another over time, any point in time will capture a group of people in between the modes. Now, you can argue that people don't spend much time in between those modes but you haven't presented any evidence for that. What's more likely is you have geocities coders on one tail and John Carmack/Linus Torvolds on the other tail. And in between are people like the presenter and I. And since I'm not instantaneously going from bad to good, the reality of the situation is most likely some degree of a normal curve filled with people trying to get better at programming or even just getting better though spending lots of time doing it and learning a little along the way.

For all your attacks on the presenter, your argument of a bi-modal distribution sounds more flawed to me. I would love to see your study and hear your argument.

To be fair, you can spend a great deal of time talking about something and make progress on the issue without solving it.

For example the current metrics are abysmal so it's worth explaining why they're abysmal. I just was able to delete several thousand lines of JavaScript from one of my projects after a data model change (through code reuse and generalization) -- yet I increased functionality. My manager was confused and thought it was a bad thing to get rid of code like that ... it was absolute dopamine bliss to me while he felt like our production was being put in reverse. KLOC is a terrible metric. But yet we still need to waste a lot of breath explaining why it's a terrible metric.

Another reason to waste a lot of time talking about a problem without reaching an answer is to elaborate on what the known unknowns are and speculate about the unknown unknowns. Indeed, the point of this article seemed to be to advertise the existence of unknown unknowns to "recruiters, venture capitalists, and others who are actually determining who gets brought into the community."

Perhaps ... but that would imply that one does not transition over time from one hump to the next or if they do, it's like flipping a light switch. When I read this I assumed that he was talking only about people who know how to program and not "the average person mixed in with programmers."

But you just said there's no way to measure this ... how could he have collected data? What data set could have satiated us? The answer is quite obvious and such collection would have been a larger fool's errand than the original article's content.

Maybe because Net Applications is the only counter that tries to correct for known skewed sampling. Net Applications uses CIA internet usage data (how much of the population in each country has access to the Internet) to estimate absolute numbers for each country based on the measures distribution and the "Internet" population number. Net Applications is perfectly honest and upfront about this.

The other counters just report whatever stats has been collected. They also are perfectly honest and upfront about this.

Both correcting and not correcting may leave errors. Be your own judge.

But there's a perfectly good explanation as to *why* the numbers seem not to agree: They do not even claim to illustrate the same thing. Net Applications tries to create a number for "true" global distribution (and risk errors), the others do not even claim to compute such a number. In theory you could take the numbers from, say statcounter, by country and extrapolate the absolute number per country, sum them up by browser and calculate a number similar to net app. Could be interesting to see.

Also, be aware that there is also great popential for skewed demographics between the counters, not to mention the fact that Net Applications tries to measure unique visitors (discarding repeat visitors within a month) while most of the others just report page impressions. If for instance users of Chrome are more active on the 'net than users of IE, chrome would have a bigger share of page impressions than they would of unique visitors. There is no "right" in this: It all depends on the question you ask: If the Q is "which browser is the most popular?" you would look at unique visitors. If the Q is "which browser is used the most?" you would look at page impressions.

Maybe because they use the *least bad* methodology. The others do not even *pretend* to estimate global usage. They may report what *they* see of usage globally, but none of them claim to know how many users there are in each contry.