As a newbie who was trying to make myself marketable, it felt like a bit of a hamster wheel, but then through the user group, I met .NET pros who felt exactly the same way.
Most .NET developers learn what they need to know for the project at hand. A year ago, I worked on a fairly cutting-edge EF Code-First n-tier project, with knockout.js on the front end, using ASP.NET MVC web API. I am currently extending a crufty ASP.NET 2.0 webforms app.
Adaptability is an important skill for developers. Also, each iteration of the Microsoft data access stack feels like an improvement over the earlier ones, so it's not an unbearable burden to learn it.
And at the end of the day, the greatest irony was that when I finally broke into the market, it was a Java-based developer role on the basis of my JavaScript. So all my years of frantically spinning the .NET wheel was for naught.
Hiring someone for a Java job, based on their JavaScript knowledge is fairly unconventional (to put it politely). Are you certain that the general programming knowledge you acquired through "spinning the wheel" wasn't a factor?
I never stopped (and never do stop) learning as much as I could, and I apologise if I gave the impression I was only learning .NET stuff. I was primarily coding in Python, but got into JS to write stealth code to automate a horrendously manual part of my day job in a government contact centre.
To clarify for you on the hire, I was hired for the JS, but expected to learn Java pretty quickly.
> Adaptability is an important skill for developers.
For sure.
Same here. I design databases, and MS kept bringing new data access methods every three years; I never could wrap my head around the ADO/DAO difference, but that did not matter for long because both were replaced by .Net
I sincerely wondered if this was not a deliberate tactic by Microsoft : while developpers are busy learning their newest "next great thing", they can't invest time in anything else.
Meanwhile I learned how to use an OSS lamp stack and have been happily using it for the last 10 years. Very stable and reliable, I could not be happier.
Bingo!
Either Microsoft is incompetent or promoting/retirement of APIs is more profitable.
mmmh... I thought about this too.
However, MS is anything but stupid marketing wise, I find. So I dismissed incompetence.
I find that Java's biggest problem (stability/stagnation depending on who you ask) is also the JVM development ecosystem's benefit. There's not one body driving adoption of technologies by deciding what's supported and what's not, so technology choice is less dictated by economic imperatives.
Linux, Apache, Mod_perl, Postgresql
All very pleasant to work with (1) and stable; mod_perl has a rather steep learning curve, but worth it for web work. Very powerful _and_ low cost.
( 1 : if you like Perl, obviously )
[Edit] asterik in note caused weird formatting; replaced with 1
WPF and Silverlight share almost the XAML stack with little differences. One just needs to know those differences.
LINQ-to-SQL and Entity Framework are usually used together and also share common concepts.
Other vendors also keep changing their stacks, it is the job of a developer to keep up with changes and learn to quickly pick up what is relevant for a given project.
Sure, on Windows Phone I'm forced to use LINQ, and I've written some really crazy code to hide it from my app, but at least my app is fast.
So, I joined a .NET user group, I installed VS, I started learning.
Problem was, MS was churning out the "next great thing" faster than I could learn the last great thing. While I was getting my head around WinForms, WPF came out. I started using WPF, then Silverlight arrived, which was like WPF, just a subset of it, and therefore different enough to be problematic.
But that wasn't too bad. It was the 'blessed' data layer stuff that threw me. When I started it was DataSets etc. Then it was LINQ-to-SQL, then it was very much not LINQ-to-SQL, but instead was Entity Framework.
As a newbie who was trying to make myself marketable, it felt like a bit of a hamster wheel, but then through the user group, I met .NET pros who felt exactly the same way.
And at the end of the day, the greatest irony was that when I finally broke into the market, it was a Java-based developer role on the basis of my JavaScript. So all my years of frantically spinning the .NET wheel was for naught.
I still do love F# though.
In hindsight, the thing that strikes me about .NET development is that most businesses using it really want the whole stack to be from one provider whom they trust - and as a result, third party alternatives never really get a look-in if an MS or MS blessed technology such as JSON.net (http://james.newtonking.com/json) can do the same job.
That is a massive sweeping generalisation of course, and only applicable to the limits of my experience in a small city in a small country.