“Wouldn’t it be more effective to do a Silverlight talk that happens to show a useful use of DI, YAGNI, etc. Jeffrey Palermo comes to mind as someone that did an ASP.Net MVC talk and littered it with examples of Composition, DI, IoC, etc. Thats exactly the approach we need more of.”

Yes, good point, I think that would be fantastic. Makes a lot more sense now.

Again not at all advocating watering down of content. But altering how it is introduced.

I’m advocating we get back to the roots of why these principles came into existence in the first place. Focusing on the concepts not as theoretical benefits with abstract examples but as _actual_solutions_to_pain_points_ .

If we are only talking to ourselves whats the point? Great now I’m a better dev, but now I have to hire other people because I’m too busy, let me just throw a crap load of acronyms at them and if they don’t just “see” why this matters well I’ll just get another wave of people and repeat the process, over and over again.

If you cannot show “the masses” how this benefits them without concrete real benefits, then whats the point of the principles? We absolutely need better ways of approaching those without the knowledge than we often currently exercise.

At my local user group we’ve had about 5 or 6 SOLID oriented talks since I’ve started and honestly most of them were fantastic and even I who’s done one of the talks learned something from each of the others.

But we get lower attendance for those normally than when we do talks on Silverlight or some other MS technology. Wouldn’t it be more effective to do a Silverlight talk that happens to show a useful use of DI, YAGNI, etc. Jeffrey Palermo comes to mind as someone that did an ASP.Net MVC talk and littered it with examples of Composition, DI, IoC, etc. Thats exactly the approach we need more of.

Finally, I think we have enough idealism in the community to not have to worry about being too close to the least common denominator. We have a long way to go before thats actually a concern.

I fullly agree that there is statistically quite a few people making good money while outputing lousy software, the problem is many of those people don’t care to improve what they are doing. I really don’t think those people are worth preaching too, because frankly they’ll never read a Alt.NET site anyways. You don’t end up reading Los Techies, Code Better, et al by accident, you find them initially by searching Google for solutions to problems, but you keep reading when you find what they are saying make sense. Who you want to speark to is the younger developer who doesn’t really no any better than what Microsoft has told them, but who wants to improve their craft. In that vein, I think Los Techies does a great job in highlighting the “basics” on your main page, something other Alt.NET sites should follow. Also, improved examples would probably help too (I find most of the SOLID examples out there incredibly contrived, trivial and not very “real worldish”).

Think of it as trying to pickup a girl a little bit out of your league at a bar, you really only have one chance to impress her before she just puts up the “okay, stop hitting on me” wall. Alt.NET sites might really only have one chance to capture the imagination of what is commonly referred as a “5:01 Develpoer”, but often miss the opportunity because the blogs are getting filled with, for lack of a better word, crap. Bicking Blog posts or long winded entries talking about fringe conditions are a huge turn off to most people.

As for the Ayende situation, having a civilized discussion about the queries part of his “solution” within the comments of the actual blog posting would have been good, instead it turned into an excuse to write a pissing contest blog about abandoning the Repository Pattern on another website. Basically, it turns into something alienating to all but a small portion of the already converted, the majority of people really don’t care to see a geek fight.

Again, I think Los Techies is definitely on the better side at introducing Alt.NET type philosophies to the masses, and I think its’s the other websites that are lagging behind.

There is some truth to what you say, but the sad reality is there are statistically a lot more people coding making good money without using DRY< YAGNI than there are that are.

We need to do what we can to show them they’re shooting themselves in the foot, if they really feel that DRY isn’t worth it, then increase their workload (even if it means increasing your own) so they have to feel the pain of their slow way of doing things. You’re right sometimes even that won’t work, but honestly we should be in the trenches trying to make gains where we can. Sometimes that involves getting help from on high, but ideally not.

Also side note Ayende was also advocating a Query Object for complicated queries rather than a full repository which he felt was too coupled, so not only was he advocating simplicity where it makes sense (via ISession), but for the truly complicated queries, dropping them all in one place leads to a mess to maintain and test.

Sorry Len but I feel that the attitude you’re advocating is the exact thing I’m trying to prevent.

I do not think its “dumbing down” to be careful with dropping acronyms _all_the_time_ particularly with the uninitiated, or to focus on the actual _solution_ that these principles provide, instead of just assuming people will care when you say “you’re doing wrong try TDD instead”.

I’m saying its high time we starting refocusing ourselves to solving a problem instead of living in an ivory tower where we say to ourselves “well at least we do it right”.

From a self interest point of view its high time we do what we can to decrease the amount of crappy code that is out there, and that we have an easier time hiring colleagues that have no concept in their being a different way to code.

Obviously one of us isn’t going to do it all by themselves so everyone that does have a clue, should not be hesitant or unwelcome of the concept to changing their _own_ approach .

I don’t think I’ve heard a developer say that testing is bad, but I have heard the traditional arguments against testing, such as “it will take to long”, “we’ll test after the system is complete”, etc. We all know that those reasons are unfounded. People may just be to ignorant to do anything about it, since they’ve never done it, hence it’s a big scary monster nobody wants to touch.

I agree with you that those principles has nothing to do with agile, like many of the things in DDD has nothing to do with DDD. It’s plain old good OO. It may take a bit of self discipline at first to start using those principles, but once you’re used to it, you’ll never go back.

I might be painting a darker picture than it really is, and I’m hoping I am doing so, because I really want people to learn and be open minded. However, I get a nagging feeling about it, so time will tell.
I realize that I’m not be able to teach everybody everything, but if I’m able to nudge them in the right direction, and they start questioning things then I’m quite happy.

I’m giving the talk tomorrow after work. I can tell you how it went hehe.

dumbing stuff down so the masses can write bad software is microsoft’s job

To be honest, DRY, YAGNI and SOLID have nothing to do with Agile. If developers can’t see the benefits of not repeating themselves or not wasting time building something that might never be used, well then I’m afraid there isn’t much hope for them (from a learning standpoint).

That might sound harsh, but it’s the reality of our profession. In our profession, whether we are qualified or not, we make too much money especially considering there is no real way to quantify success or failure.

People like that, espeically those who are entrenched in a position that they’ve occupied for years, need direct instructions from above them on the food chain to change. I’m not saying it’s not possible (because it is), but it can be daunting to try and get people to change from the status quo.

Good luck though.

A perfect example is a few months back when in his own blog Ayende/Oren said that he feels that the Repository Pattern is over used. For small ASP.NET MVC projects using NHibernate, he thought it’d be much eaiser to just use ISession in the Controllers, why bother with all that added complexity when NHIbernate can do it for you.

Well, you’d think that he’d set Martin Fowler’s pants on fire and kicked Eric Evans square in the nutsack because people freaked. Of course, considering everyone and their uncle in Alt.Net has a blog, it turned into a “Read my blog to find out why Ayende is wrong!” sort of thing. Maybe we was “wrong”, but It was very off putting.

I see a lot of petty bickerings about things that shouldn’t even really matter.

Also, for what it’s worth, I doubt there are many “brilliant, talented and overall effective” C# developers out there who think SOLID, testing, et al is dumb. As a matter of fact, I assume most good developers adhered to the SOLID principles without even knowing it because they just make sense. Unit testing, even if they don’t do it, I’ve never heard a good developer argue that it was bad (just that it was time consuming).