Recently I've generally implemented MySQL data layer for my private sandbox  MP forum module VZ-Forum based on  YAF- fully functional forum with subforums, private messaging and so on. The native data layer was in MSSQL, so finally it'll work with it with small efforts. 

The major (but without any doubts the most silly) current problem is that it gives out sequence of unicode caracters than it comes to non-latin alphabets in posts.

The second problem is in triggers, as MySQL doesn't support creating them without SUPER privelege before 5.1.16 or so. It makes some things like statistics, or UserForums update work not as smoothly as expected. In other aspects it works fine on my development server.

The third problem it's not tested on Mono, so it'll need some tricky code jobs, but as far I can't see any possible problems in the crazy code to cancel it. If I  find time, I'll possibly share a link to Sneak Preview Version here in about 3 days as a test option for MySQL users, and the one for MSSQL users in the nearest future. But my current workload can postpone it.

1)Database porting is completly impossible for Sqlite only. For a success  it is desirable, that a database supported stored procedures and workable transactions handling. 90% of the plethora of the database objects are very simple.

2)I've written a couple of sample skins for techmania (by default) and orangesunset.

3) Of course,  mojoPortal cannot handle all the complex access rights for  users. So the product  uses it's own bit-based access mask system. Any good forum has a complex access rights managment system.   If you option write your own for MP modules, a full integration is possible. It uses curently a double users count. If you restructure mp_users, mp_modulesettings  tables, it will be possible to integrate it more closely in future. Anyway, before I adapt a MS SQL data layer it's imposible to change anything.

4)  Multiple sites are not supported at the moment,  but this is not a problem. Currently boards (modules in mp termonology) can be created for 1 any site, but I guess it's not a challenge to implement them for many ones, the only condition -   module numbers must differ. Else it requires some serious changes in the database.

I hope that link for the source will be available in this post tomorrow UTC and you can see all himself.

  Now there are problems with statistics updates, because a lack of  not SUPER triggers support in production My SQL. But except this it works fine, thought I have no doubts that it is awfully buggy. 

The forum module looks great!  I've got 5 friends who want to run mojoPortal once mono works properly with the svn version of mojoportal, and all of them liked it except for the default forum module.   (No offense Joe, its good, just as much as people are used to)     I might be able to help you out with mono testing if you're looking for testers.

I don't have much time for development (busy job gets in the way, I never get my personal projects done) but I can help you track down the bugs.

I just installed vzforum into a fresh mojoportal installation and on rev 3 when you are in the test forum and hit New Topic it comes up with a 'Table 'mojoportal.vzfrm_smiley' doesn't exist' error.  Checking mysql the error is true, the table doesn't exist.

Great work!   I can't wait to run the stable version once we get there.

Hi, Richard.

As to letter casing I had some troubles with it too. But now it works. I wrote all info in a file. Unfortunataly, I don't have time to change some things to adapt it to Mono. I didn't use the same DB in Linux and Mono, as was focused on scripts.  I solve the problem as soon as I can, if it really currently exists.

There is a lot of current problems with mono, i.e.

1)Mono doesn't understand things like Src="../Modules/sss.ascx", so full path should be used, but I will try and find a clever workaround for this.  Mono doesn't convert data types as .NET Framework as I see.  

2)Then I wrote datalayer I changed some casting  in code like (int) myvar to Convert.ToInt32(myvar). 

3)The thing uses  object data type as DataRow type entries for current control info. I used cookies as a temporary decision to remember module ids it's not a good solution but it gives a possibility to make many instances on many pages right now, but not many instances on a one page. It costs and I don't think this is a priority. 

4) It has some bugs. But I don't think they  can currently damage tables data.

5) I guess that on the weekend I find some time to wipe out some more bugs and you can play around with it in .NET Framework.

6) It's difficult to spot problem with Mono, as I can't trace some bugs source so simply as in VS. So it takes a lot of time. Presently I've temporarily migrated from Gentoo to kUbuntu 8.04. I guess it takes some time to launch it with Apache again - I'am afraid I completly forgot all Debian things. 

This was intended  only as a Sneak Preview. So then it can be used in Windows as an alternative I will write at once. But I don't intend make releases for Windows at all - only svn will be avalable. It's not a joke to support a complex thing like this for multiple platforms and DBs for community sake. I make it for myself and intend it for Linux only. I don't think that I can find time to implement it for Mono even next month, but in June I expect some time to focus on it.  Then it works with Mono I'll probably finish  MS SQL data layer, it takes only 4  large SPs and some minor table changes. 

Regards, bob. 

I've merged some changes to the forum svn. Now it can be used with Mono and MySQL too. Keep  in mind, that you use it at your own risk - it's only for testing purposes . No support, liabilities  or fitness for a purpose. You can download and compile it as any mojoPortal Module. I can't see any bugs different from the .NET version.  

Currently you can't use local avatars, can store attachments  in database  only.  Enjoy, if you can...

 Oh, thank you. I didn't consider it as a project, but SF simply is a good place to share a useless code. It has a very small target of Mono users and at the same time mojoPortal users with MySQL, plus a tiny share of MySQL + .NET Framework freaks. I put it to give some boost to MP on Linux, as Windows aand their SQL Server users can easily integrate YAF in MP and mostly all the features will work just fine. Of cause it requires some lines of extra code. 

Hi, Richard.

I know about the problem. It comes from mysql devs in latest versions . I've seemingly found a workaround for this. The problem is that I should rewrite all the SPs where I used prepared statements. @_start - this signals about the problems. Now I've completly swamped in work and  my apartment grand renovation and can't think much about the project. Currently I'm waiting for YAF 1.9.3 release and, possibly, MySql 5.1. There is a problem as currently triggers require SUPER privilege, which is no good for a production version.  It's not possible to implement the existing data layer easily without them.

I guess, in a month, I find time to play around with the code. There is a need to comletely abstract data layer in future to make it  easy to implement any data layer. Anyway, first I implement and test mysql data layer for 1.9.3 port and Mono. After this I'll change all the old scripts for YAF 1.9.1.x based version. 

Best regards, bob. 

Hi, Joe.

I guess, that my forum will simply be an option. This forum is a very complex and finicky beast and there'll be no possibillity to debug and support it on an acceptable level if it's too closely integrated to the mojoPortal. So your native forum is needed anyway. The more, the better. I launched a test server some months ago, but it was not in English. This winter I make it publicly available with English sample forum integrated in your portal so you will hopefully have a chance to see it in action. 

Best regards, bob.