M$ made IE6 to be “different” so that their customers would not want to stray to other browsers and other operating systems. While in the short term that worked and propped up market share it is now working against M$ as those same customers cannot move to “7″ and IE8.
“Organizations running IE6 have told Gartner that 40% of their custom-built browser-dependent applications won’t run on IE8, the version packaged with Windows 7. Thus many companies face a tough decision: Either spend time and money to upgrade those applications so that they work in newer browsers, or stick with Windows XP.”
Good. After going to the expense of rewriting all those custom apps, businesses will be able to use any OS and get out of the trap M$ set. Look at it this way. If you just got yourself out of a bear-trap, are you going to step into another close by? Not likely. People and organizations learn from their mistakes.
Think of that the next time you build an application or buy a licence for one.
17450
12755
206
3
2
23895 
11851 
11703 
4626 
4259 
1641 
198 
14 
2 
0 
0 
0 
“Think of that the next time you build an application or buy a licence for one.”
My suspicion is that unless their application can be duplicated by the current state of FOSS, most if not all of these companies in the end will bite the bullet and upgrade their software to the latest Microsoft offering.
It has been my experience that locked-in is not confined to commercial software, especially when one has made a substantial investment in customization, the cost to adapt to the new version is going to be substantial whether commercial software or FOSS is used.
Of course one can choose one’s form of lock-in. The trick is to keep things as open as possible, both the application and the data-structure. One thing you can do with FLOSS that is pretty difficult with non-free software is to contribute the code to the FLOSS community. Then your organization does not need to be the sole supplier. For example, SUN was locked-in to Office. Faced with the prospect of buying 20000 licence they bought StarOffice for less. They also freed the source code so that others could contribute to it although they kept too tight control. If they had given control to an independent body they would have had all the benefit for very little cost, essentially freeing themselves from most of the software development and only being responsible for generating their data and data-structures, documents. A company does not need to bear the whole cost of developing a FLOSS app. They can ensure the viability/vibrance of the application by sharing it and contributing just a share. The benefits then far out-weigh the costs. Lock-in is reduced because at all times the business owns its data-structures and data which they do not need to share.
FLOSS is a great deal for large organizations that way and a gift for smaller organizations who get a free ride. No one is really locked-in because they keep their data in open standard formats.
I am working on a web application to keep attendance here. The cost is mine alone in hours worked but I get the benefit of having a light load for month-end reports until June and the other teachers get the benefit as long as they work here. There are FLOSS apps out there to do this kind of stuff but none can deal with the unique schedule and paper-flow we have at this school… Really. In most schools attendance is kept continually and the register is just tallied up at the end of the month. Here, we do not call the roll daily but deduce it at the end of the month by analyzing the attendance reports of each course published at the end of the month. The result is hours of work with a short deadline, like tomorrow. Further, the schedule is not the same or the same pattern daily so you cannot keep attendance by time-slots but by course so there is lock-in but not in the software. It’s in the way business is done. The attendance has to be recorded by time-slot and course simultaneously. The custom software can easily handle this better than paper. We will keep paper as a back-up but the reports can be generated directly from the database instead of adding rows and columns manually.
My app is written in PASCAL and will be readable and well documented so it is trivial to change in the future if need be. PASCAL is not like C with different versions of the language every other year. It has been pretty well standardized for many years. We just use a tiny subset of MySQL so that should keep on working too. Both FreePascal and MySQL are Debian GNU/Linux packages so they will be easy to maintain. Again, every hour of development time will be paid back per teacher per year so the day or two I spend on it is well worthwhile. Because I use FLOSS I get the benefits of GNU/Linux, MySQL, Apache and Free Pascal at zero cost and yet I get an application that will save us a bundle of stress leaves for a few hours of weekends.
This tool helps the administration, too, as they should be able to get the reports at the close of the day on the relevant month rather than next week. Attendance is vital to schools whose funding is partly set according to enrolment. Automation will cut down on errors and help spot problems in attendance as well. It could also save its costs in paper, too as we used to photocopy our reports and shift them around to all the staff in high school. The data can be merged with report cards as well saving more hours per teacher. We used to edit individual report cards once for each teacher and for each student, taking days. Now teachers will not have to enter the data, only marks and comments. I just need to add a record for the mark. Paper report cards produced from spreadsheets were taking us several days to produce. We should be able to produce mid-term/term reports in a day easily now. It will mostly be a matter of printing, proof-reading and signing.
These guys (ReactOS, a drop in XP FOSS clone) have about 3 years to get it together and if they do they could reap huge marketshare. A real sweet-spot could be opening up between now and 2014 when XP SP3 goes unsupported
http://www.reactos.org/en/about.html
If they do their cleanroom reverse engineering Win32API’s right it will be an extensible, fairly malware resistant (unless the Win32API’s themselves are flawed), structured (maintainable) version of XP.