Frans Bouma's blog

As a comment to my blogpost about the 12 minute long video of LLBLGen Pro with Entity Framework v4.0, Burton Roberts asks:

Today I recorded a video in which I illustrate some of the database-first functionality available in LLBLGen Pro v3.0. LLBLGen Pro v3.0 also supports model-first functionality, which I hope to illustrate in an upcoming video. LLBLGen Pro v3.0 is currently in beta and is scheduled to RTM some time in May 2010. It supports the following frameworks out of the box, with more scheduled to follow in the coming year: LLBLGen Pro RTL (our own o/r mapper framework), Linq to Sql, NHibernate and Entity Framework (v1 and v4).

In the past weeks I've read several articles / blog-posts and other digitally expressed thoughts about relational databases, query systems and how they all suck compared to K/V stores, CQRS, OODBs or whatever Hype of the Day-term. While most of them were simply re-labeling 20+ year old common knowledge, others were pretty stupid and downright sending the (novice) reader the wrong message. With 'wrong' I mean: the conclusions are based on false 'facts', assumptions and hand-waving n==1 pseudo-science.

For the people who know me a little it's no surprise, but in case you didn't know: I love algorithms. I think they're the cornerstone of good software and they should be your first source of wisdom for every piece of software you're creating. This post will show an example of what I mean by that and how easy it is if you have a set of algorithms at your disposal which are solid, proven and correct.

In .NET there's a class called StringComparer. It has some handy helpers, like the InvariantCultureIgnoreCase StringComparer. These classes also implement a method called GetHashCode(string), which produces the hashcode in the scope of the comparer, so if you're calling that method on the InvariantCultureIgnoreCase variant, you get the hashcode for that scope.

Devnology, de Nederlandse developer community die niet gelieerd is aan 1 specifiek platform, heeft z'n 3e podcast nu online gezet, welke volledig bestaat uit een interview met ondergetekende! De podcast duurt ong. een uur.

Below I've linked a short video which demonstrates, among other things, the Quick Model feature of LLBLGen Pro v3.0. Quick Model is a feature which allows the user to specify model elements very quickly using a simple command input system combined with a visual model viewer. The scenario when this feature is ideal is when you're interviewing a Domain expert and you want to store the information you gather in a re-usable way. This feature allows you to do that in such a way that the model is immediately presented to you and the Domain expert (so s/he immediately sees if it's correct or not). Another advantage is that the model is already in your project, so if a developer has to continue with the project, you don't need a translation phase and you don't have to discuss which entities were determined during the interview, they're already in the model. All you need is a little fine tuning perhaps, using the other editors in the LLBLGen Pro designer. As the Quick Model feature is ... quick, you can type while discussing / interviewing, so the interview isn't stalled by you having to perform slow toolbox-jedi-tricks or other slow modeling wizardry.

After I graduated from the HIO Enschede (B.Sc level) in '94 I have worked with a lot of different platforms and environments: from 4GL's like System Builder, uniVerse and Magic to C++ on AIX to Java to Perl on Linux to C# on .NET. All these platforms and environments had one thing in common: their creators were convinced their platform was the best and greatest and easiest to write software with. To some extend, each and every one of them were decent platforms and it was perfectly possible to write software with them though I'll leave the classification whether they were / are the greatest and easiest to the reader. I'll try to make clear below why this dull intro is important.