Balder's History

Balder Technology Group was founded in June 1995 (originally under the name Ivy Software), and remains a corporation registered in the State of Washington and owned by David Goebel.

David worked in the Microsoft® Windows NT® base group from September 1990 through June 1995, principally on Windows NT file system development. He was a key contributor to Windows NT 3.1 (the most difficult march), 3.5, and 3.51. As a contractor to Microsoft he added important functionality to fastfat.sys for Windows NT 4.0 and helped design the Single Instance Store functionality in Windows 200 Datacenter Edition. David was a speaker at the Windows NT Installable File System Conference held October '94 in Seattle.
David Presenting At IFS Conference
(David Goebel presenting at the IFS conference, reproduced by permission from Microsoft)

His talk on the complex subject of File System Synchronization was among the most useful according to the participant surveys, and won hands down as the scariest talk. The latter was unintentional but stemmed from the explicit examples given during the talk of how easy it was to introduce deadlocks into the system that escape detection in all but the most stressed scenarios. FS/Cc/Mm interactions & synchronization is NOT black magic, but it does require careful attention to certain sets of rules to get it right.

Since leaving Microsoft, David has pursued several over-neglected projects such as his indie record label Ivy Records, long distance bike touring, and just having more of a life in general. Along with these non-software interests, David has been consulting and coding for several customers since June '95 (Microsoft among them).

Balder's Philosophy

David gave the second half of his twenties developing and debugging Windows NT (which became just "Windows" and after Windows XP). The sleeping bag in his Microsoft office saw frequent use (thankfully there was a shower in the Windows NT building), and he became quite an authority on Marriott's catering menu. He is proud of the team's accomplishment and now to further validate the sacrifice of his youth to Microsoft, wants to do what he can to help Windows NT attain its deserved dominance of both the desktop and the mission critical computing markets.

David wants to help vendors write "good citizen" kernel mode code (specifically file systems and filter drivers) that adds value to Windows without compromising any system integrity. Windows as a system is only as robust as the least robust kernel mode component. The three most important aspects of kernel mode code are quality, quality, and quality. If you have a similar view towards your project, then David may be able to help you achieve your goal.

If, on the other hand, your goal is just to get something that "kinda works" to market as soon as possible or to add a useless/dangerous encumbrance to Windows because there is a market of naïve people who will buy it (Windows "cachers" fall into this category), then there are others out there that can assist you.