We recently re-wrote an AutoCAD 2010 application to be entirely based on .NET. Our application was using Windows Presentation Foundation for the presentation layer, which lets you nicely separate presentation from business logic using the MVVM pattern. Since this application also had to run in AutoCAD 2009, it had to use the .NET 3.5 version.
It turns out there is a bug in .NET 3.5 (fixed in 4.0) that causes a stack overflow exception when you define two or more default style resources for the same type at the same level. If you use global resource files and window-specific default styles, you may run into this and its not at all obvious what the problem (or the solution) is.
One of the perks of doing software development for small- and medium sized organisations is that you get involved in all aspects of IT. In our projects at ChasmX we have done backup, workstation and router configuration, printer troubleshooting, mail merge support and, of course, database maintenance.
Database administration skills are a must when you develop line-of-business applications, and most of us know our way around the usual suspects: Orace, SQL Server, MySql and sometimes even DB2, Informix or PostgreSQL.
When you specialize in legacy application renewal, you sometimes come across more exotic database varieties. In this article, I’ll share my knowledge of SQLBase , a variety I came across earlier this year.
SQL what now ?
Gupta SQLBase was one of the first relational databases for the PC in the 80s. It has changed owners several times in its 30 years, but currently it seems to target embedded platforms, claiming to be a ‘no maintenance’ RDBMS. But back in the mid to late 90s, SqlBase was still a very popular relational database technology for PC platforms, which is why some of the older legacy apps still use them.
I have been helping a client out with DBA services for an aging 7.6.1 release of SQLBase. To give you an idea of how old that is: the hardware requirements state that you need at least 24MB of RAM available on Windows 95, 98, NT or windows 2000
Recently I was tasked with extracting some data from an Oracle 10gR2 server and converting it to SQL Server 2008. I could not get very much information about the Oracle server other than a sys password and the location of the nightly backup dump files. The server was serving two applications with very little vendor support – but they were working. The icing on the cake was that the the Oracle EM web app was not working – so I was also running a little blind. I didn’t want to break anything so I opted to leave the EM non-functional and I decided to import a recent dump into a fresh scratch Oracle instance. I would do my export to SQL server from there – zero risk of an oops.
The dumps were full system dumps and were tens of GB in size. I needed to know the names of the schemas being backed up so I could selectively restore the schemas to my new Oracle instance. In my messing around I accidentally loaded one of the dumps into a text editor (that was smart enough to deal with large files) and noticed it was mostly text. Keep Reading