and...we're back

So yeah, its been a bit. Work, low motivation.

Stuff dammit. Don't question me.

So back at the whole ASP.Net MVC thing which, in the considerable interim, has made it to 1.0. Did the whole Nerd Dinner thing which I found interesting - and not just from the framework perspective. Lots of other goodies in there like IoC, mocking, testable patterns.

Anyway, back on to the redesign effort. It would be really easy, and probably considerably faster to just re-implement the app, apply my lessons learned, and be done with the whole thing. Probably be able to knock it out in a month or so. But then I don't know what I would learn and probably skip over some important things.

Like testing.

So I'm making a commitment right here that I will write the tests I didn't last time. I will take the time to have a legitimate go at TDD. I will think first before I go whipping away at new features. The whole point of this exercise is to learn right? And not just the latest hotness but about things that are generally useful no matter what.

But good lord, one 16 tests and only one controller sorta covered? It's gonna be a looong road.

taint

End of Day 2 with the ASP.NET MVC framework.

Taint Rails.

Taint regular ASP.NET.

Taint 1.0 yet - to be fair.

Taint a good developer working here either.

I'm probably being a little hard on myself here.  No, I'm not world-class developer by a long shot but I do pay attention to a few things.  It's an odd thing being caught between two worlds.  I consider myself a solid .NET developer and an OK Rails developer.  

I think the hard part is that when learning something genuinely new (e.g., Rails, .NET the first time 'round, Django) there's a certain expectation of difficulty - I mean, it's not like you've done this before.  But I think that working with ASP.NET MVC represents a certain crossover which doesn't bode well (at least for me) for a non-frustrating learning curve.

So it boils down to this: I know how to work in a .NET environment; I know how to work in a Rails environment.  ASP.NET is a [theoretical] merge of the two.  How could I not be successful?

Turns out, it's pretty easy.

The frustrating part emerged quickly.  For each (forgive the prose, I could break into a [vaguely] witty version of prose in code] but shall not) issue encountered, I could imagine how easily this could be done in one environment (Rails vs. ASP.NET) versus the other -but I'm not in the those environments.

The other part was an unintended side-effect of a rewriting a previously written application.   I'm experiencing an unfortunate non-acceptance of a learning curve (again).  I'm sure that all those now self-evident decisions that I made in my Rails app were not so self-evident at the time.  And probably hard won.  But because this is a .NET app, things should be easy.  I mean, all I have to do is implement the same feature in a language I'm, if anything, more comfortable in.  What could be easier?

And the other, other thing is that this is, above all, a beta version.  I think I picked up Rails at 1.0.  They (DHH et al) had already learned a number of things.  Perhaps not.  At the very least, they'd decided that things were OK enough to slap the 1.0 label against it.

The RC for ASP.NET MVC is coming up shortly (slated for sometime in January 2009) which means that a majority of the articles written are centered around various preview releases.  And because they're preview releases, there's no expectation nor promise of backwards compatibility.  So any number of articles/tips may (and typically do) include code samples based around features no longer in the codebase because, see, it taint 1.0 yet.

More to come.

the notion

Learning in a vacuum is always tough. Say you want to learn some new technology - in my case, the ASP.NET MVC framework. I could go through and just read and watch and probably learn a thing or two. I could build something made up, like a music collection, but the problem with those kind of learning vehicles is that because they're made up, you feel no particular desire to follow through with them to the bitter end. At any point, usually when things get too hard, you can:

• Make it easier by not implementing feature "X" at all


• Dumb down feature "X" to make it easier


• Declare victory and move on

You can do these things because, after all, it's not a real project. It's not for a customer - even myself. Even when it's done, I don't plan on using it.

So the real trick is to build something practical, something you might even use.  Or in my case, rebuild something you use.

We have a system here at work I built using Rails to keep track of projects, requirements, etc.  The problem is that I work for the government and they're not always fans of open source, or all too often, anything new.  So my app sits on our dev environment (which is where it ought to be anyway), but I keep hearing how we're going to need a version of it that will sit on the government network.  And because the .NET framework (which is what we use here 99% of the time) already has a pass from the powers that be, perhaps a should just write a version for .NET?

Historically my answer has been, "I don't feel like it." I can get away with this answer because nobody has pushed on it and the Rails app works just fine on dev.  But it occurred to me the other day that the Rails app would be a perfect candidate for learning the ASP.NET MVC framework.

Practical?  Check.

Defined requirements?  Check.

Plus it is (or should be) a largely 1-1 translation.  Lots of lessons learned from building it the first time plus an interesting comparison on how the two frameworks stack (if you'll pardon the pun) up against one another.

So we'll see what happens.

yer own shit

Everyone hates supporting another's code.  It [someone else's code] is invariably spaghetti, ill-conceived, and poorly executed.  In short, you could have done a far superior job.  But nowhere is this disgust so foul but in the realm of HTML.

Blech.  I hates it.

I am so wildly happy that my present job renders HTML as a means to an end rather than an end in itself.

Hurry back you holiday revelers so that I may go back to my enclave of well-ordered, typed code.

i can haz install?

This screen

was not easy to come by.  Were it not for the whole livelihood thing, I'd probably have given up.  

I'd already had to wait a ridiculous amount of time for the MSDN subscription, so when the DVDs hit my hot little hands I was more than a little excited.  But then, there was saga.

The VS 2008 install went fine on my desktop.  The laptop had issues with copying one of the preliminary files off the DVD so I eventually had to share the DVD drive (yep, I just wrote that) from my desktop to my laptop.  So while that install was going on I switched back to the desktop to install SP1.  And then, the wheels came off.

Permission errors to C:\Config.msi even though I was running as Administrator.  Failing installs that would bomb out at different points with [obviously] different error messages.  Meanwhile I switched gears and tried to install SP1 on my laptop.  One failed install (it ended with one green block to go) and then it worked.  No idea why.

The most frustrating thing about the desktop was the amount of time it took to find out it was failing.  Found some tips on getting the full 3.5 install package, then making sure it was local... none of it working.  A lot of others issues revolved around beta versions or installs of a few other applications/KBs - none of which were applicable to me.  I eventually just gave up and uninstalled anything 2008/3.5 related and started over.

And still a failed install for SP1.  Bit of a low point there.  And then I installed the latest windows installer.  And suddenly (well not really suddenly, the fucking install takes forever and I'm watching and waiting for the green bar to suddenly go in reverse as I'm watching a video about ASP.Net MVC which I can't use because SP1 won't install) it works.  See the screen above for proof.  Was that it?  Was that the key?  I have no idea and even though there's a small part of me that want to know the real why, that just ain't gonna happen.

And now I have the opportunity to be frustrated in a whole new way.