Doing a search for an NHibernate-related topic today bought up an ad on Google from Vanatec.

It says "Stop Hibernating; start real O/R mapping - use visual studio - read why!" and links to the following page critical of (N)Hibernate:

http://www.object-relational-mapping.ne ... ibernating

This is good news, because it suggests in the .NET world Hibernate is starting to really gain some traction to. Keep up the good work Sergey and everyone!

N.

OK, I'm in a real bad mood, so I'll bite. :)



With the tools that are already out there, why should the NHibernate crew write one?



First, I'm fine with that - Visual Studio is way over-rated. Second, it's a completely false statement - there is no command-line requirement to use NHibernate. One caveat though, you have to do more than just wiggle your mouse around.



:) First, learn how to spell. Second, HQL has been the standard for almost five years now, and there was E(JB)QL. Hibernate (the Java stuff) is such a powerhouse standard that EJB3 is basically Hibernate with annotations. Secondly, OMG - as in Oh My God, Are You Kidding Me? So, we should all go back to CORBA?



NHibernate has had nullables support for ... ever, basically. Generics are coming along quite nicely, and NHibernate.Generics work just fine in the mean time. OK, so these guys punched out an ORM in 2.0 before NHibernate released a generic version. Way to go! Here's a lolipop for ya.



It's bang-for-the-buck, baby. NHibernate costs zero <insertCurrencyTypeHere>. Your 'product' is $900! And, with Microsoft and Oracle representing the shining examples of commercial software, bugs get fixed in Open Source software exponentially faster.

Which is more expensive:
a) find a bug in a commercial product you depend on, they tell you "it's not a bug, it's a feature" - you're screwed
b) find a bug in an open-source product, notify the dev team and wait or fix it yourself

Time is money - waiting for an open-source team to fix a bug under public scrutiny takes much less time than waiting for Oracle to fix a 4.5 year old security hole. Even if you have to do it yourself, compare that to trying to fix a security hole in Internet Exploder. If you're savvy enough to find a bug, you might be able to fix it yourself. Their argument assumes you (the developer/consumer) are an idiot.

Oh, and by the way, I want to use your product on Linux in Mono - oh, so sorry, you lose.

This is the usual FUD spread by Versant, the did this with Hibernate more than year ago. Seeing your market share plummet because the open source competition is better is no fun apparently. It was actually the reason why we wrote the "Evaluation FAQ", to debunk their lies (IIRC more than 40 in a paper they sent to customers and as Spam).

I stopped reading this after the first item (which is clearly wrong) and after realizing that it was probably written by someone who thinks that the number of ??? used makes a particular point more impressive or less untrue. Wait until they start adding this kind of marketing bullsh*t to NHibernate book comments on Amazon. I wish these clowns would address _real_ weaknesses in NHibernate (or even Hibernate for that matter), so we can fix them. Good news here is that as long as nobody finds anything true or serious, we can just relax and watch the show (for a while). Don't get upset about it, it's a good sign.

As NHibernate obviously becomes a threat to the commercial competition, we might have to do an NHibernate Evaluation FAQ sooner or later.

so.... is there a nhibernate book planned ??? ;-)

Has this "competitor" been around for a long time? It would be interesting to get their product and use reflector with it, then check how much it differs with NHibernate... I would't be surprised if they just modified NHibernate, compiled it all together and changed the name to "OpenAccess"...

i actually believe they are quite big in the java world.
As far as I know, it's a pretty cool product, which uses byte manipulation to do stuff after the compile.

but i've never used it and have probably based above opinion on propaganda from the same ppl that wrote the fud, so feel free to ignore ;-)

The only thing Versant is big on in the Java world is words, although some people still bet their business on their object database products. They have been around with the same old message that "relational" databases are bad for quite a while. I migrated a few JBoss customers from one or the other Versant product to Hibernate in the past years, always quite fundamental problems they had (_real_ scalability, shared data access, backups).

You don't have this "back to the 60s" database stuff in the .NET world, where everything is pretty much stored procedures and SQL.

"With the tools that are already out there, why should the NHibernate crew write one? "


k-dub, I'm new to nhibernate, and have it working well so far.

One tool I'd like is something to generate the hbm.xml either from a class ( which I can gen from my database with another tool )
or straight from the db.

I've looked at object mapper, but it seems overly complex for what I want to do, which is:

Just take existing tables from an existing app that is being updated, make classes where appropriate ( I don't have to touch all the tables for the update ) and use nhibernate for persistence.