Data ownership and accessibility.

[Mufferaw]

In response to a recent posting under "The Best Farm Accounting Software" in this discussion group, a suggestion was made that Access is a preferable database to FoxPro. I have responded with my own concerns due to the proprietary nature of accessing data in an Access database versus the non-proprietary and open dbf format. Does anyone else have thoughts on this issue? I have seen reviews of products like Settler Gold and reviewed it myself on occasion. I agree that it is a very good product, but how do we factor in the inherent risks associated with the fact that it uses a proprietary means of data storage (ie Access tables). Given that it is the farmer's data, should he or she not have a choice as to how they access and analyze that data, both now and in the future. Does Settler provide database permissions to get at the data using an off the shelf copy of Access95 or 97? If not, should their clients be demanding those permissions? I look forward to others comments.

[Mufferaw]

Keith, you pose an interesting question. "Open architecture" has often been more successful in the long run - I'm thinking of IBM's original PC vs Apple's proprietory approach. More recently, I believe that Netscape has just "opened up" its Internet browser source code.<BR> With the rapid demise of so much of our non-Microsoft application software, Bill Gates and company are acting very much like IBM in the 1970's and 80's.

[Mufferaw]

The concern is not so much about the software itself, but about file formats and accessibility. For example, if you have an old FoxPro or FoxBase program, or Clipper (eg the Saskatchewan Herbicide Planner) or dBase based program, you will note that you can take the dbf file used to store the data and open it up in WordPerfect, Word, or even Windows notepad. The same pretty much with the data files in older DOS software like CowCHIPS. You will find that all your data is perfectly legible. It is very easy to write a filter to read this data into another piece of software on another system, whether that system is Macintosh, Linux, Unix, MVS, or VMS. Any of us that went through university from the late 1960's on at the very least learned to do this in Fortran. Try the same with an Access .mdb file and you end up with a load of indecipherable gobblygook. The consumer, in this case farmer most likely, has committed him or herself to running a machine with Windows and Access (or whatever Microsoft comes up with for replacements) in 5 years time to access this data regardless of the fact that Windows (or the Microsoft product at that time) may well not be the first choice in operating systems and platform at that point in time. If Settler has not provided the farmer with Access permissions either for the datafiles in Settler Gold, this extends one level further in that the farmer is not only tied to Windows and Access, but also specifically to Settler itself to access their own data. Yes, Microsoft is dominant and behaving much as IBM and company used to. Note that it was that behaviour that turned the market against IBM and is now turning the market against Microsoft and in favour of "open source" products like Linux. That probably more than any other reason why this discussion is more relevant now than it was two years ago. Farmers are investing time and money in information technology. They should be aware as consumers of what they are buying and how they are shaping (or perhaps in this case limiting) their options for the future. This is particularly relevant in the case of Microsoft based products today. One does not need to study Michael Porter's work to understand that though Microsoft's earnings might be very good right now, their business model has huge holes in it that could severely cripple it within the next 3 to 5 years or even put it out of business. Many people point to Linux as the big threat right now. The technically astute will say this is because it is a much superior operating system due to its UNIX heritage. True, Unix is an operating system with a kernel based conceptually in the late 1960s versus Microsoft's 1950's style kernel from a conceptual viewpoint. Hence Linux and Unix operating systems are a generation more advanced and far more powerful and capable than Windows ever will be unless totally rewritten. Windows crashes or freezes up at times on daily basis, leaving a frustrated user with lost or unsaved work. A Unix or Linux based systems can go months, even years without a crash. Even if Windows were totally rewritten, the mathematics which define Unix's underlying architecture go back decades and hundreds of years and hence to achieve the same capabilities as Unix, Windows would basically have to be rewritten as Unix. Where is the value added a consumer will pay for if all Microsoft does is reinvent the wheel? That is unless of course Microsoft has forced the consumer's hand through proprietary means such as the .mdb files. But that is not the real threat to Microsoft and Windows. The real threat is the business model. Microsoft's model relies on closed source code and proprietary standards so that the consumer is locked into buying their products from one generation to the next, since only Microsoft has the ability to make the required upgrades the consumer longs for or needs. To date they have gotten away with this much as IBM used to with proprietary hardware and software as well. If you recall, after IBM dropped the ball and saw PCs commoditized by using open, off the shelf hardware and software in the early 80's for the original PC, they tried to recapture it with their proprietary PS/2 and OS/2 architecture in the late 80's. The consumer rejected this model at the time however. DOS software of the time was still quite open in file formats. This has changed however with the likes of the Access .mdb files and other attempts by Microsoft to corrupt open standards like CORBA, html, http, and Java. If the same business model existed in the automotive industry, then your new GM, Ford, or Chrysler would have the hood welded shut. Every time you wanted an oil change, you would have to take it back to the respective dealer. This is not an accepted business model in the automotive sector and the lightning fast rise of open source products like Linux and Apache is proving that it isn't for software either. A recent industry survey showed that 54% of the servers on the internet were Linux/Apache based and that was growing. Not because they are free (though it helps) or technically superior or easier to use (which Linux isn't), but more so because it puts control back in the hands of the system administators responsible for these sites (ie they can change their own oil). A business reason. All proprietary technologies, including Microsoft's Windows and IIS were losing ground against it the same survey. One might say then that all Microsoft has to do is open their source code up. If they do that, their currently secret APIs (Applications Programming Interfaces) would be open for the world to see. The Linux community would be able to write them into Linux, Unix vendors like Sun and HP into their versions of Unix, and hence everyone would be able to run Windows based programs flawlessly and perhaps better but without Windows. Hardly an appetizing thought to Bill Gates, especially when you view it from the primary distribution channel for Windows; OEMs like Compaq, Dell, HP, etc. Let's see, you are the president of one of these companies faced with two alternatives; You can pay Microsoft $50 to $100 per every machine you sell to include Windows so your customers can run all the Windows software on the market, or you can throw Linux on for free and your consumers can run all the Windows software on the market plus all the Linux software coming onto the market. Which one would you pick? The OEMs start delivering machines with Linux that run both types of software. Next thing developers wish to use the more powerful abilities of Linux over Windows abilities. Not a pretty picture for Bill Gates. If you think Microsoft is blowing smoke in their anti-trust case with US Department of Justice by claiming that Linux is a very real threat to them, think again, and think of the implications down the line. I am not saying that Microsoft is justified in breaking the law. If they are guilty of doing so by abusing current monopoly powers to gain other monopolies or to maintain their existing monopoly (as Access mdb files do), then nail them to the wall. The fact they have to resort to breaking the law and using unethical means of tying consumers to their platform is in and of itself proof of how great a threat projects and open architectures like Linux, Apache, Java and Mozilla (open source Netscape) are to their business model.

[Mufferaw]

I was the one who initially brought up the issue of Access vs FoxPro so I feel that I should respond. It has been my experience that applications based on FoxPro have been somewhat unstable. As Phil Anderson mentioned in another post this may not be the problem of FP itself but rather faulty programming. One of reasons for setting permissions in any database program is to ensure data integrity. I'm not privy to the source code for Settler's program, however, I'm sure that the permissions have been set at the design level. To change these you would have to go back to the orginal development to alter the permissions. Off the shelf Access 95/97 would not allow you to do this. <b> After having said that I don't believe that farm managers are really interested in gaining permissions access to their database tables. What they are mainly interested in is a stable program that is designed to do the job promised. They are definitely looking for integrated packages that will allow them tie in all production aspects of their farm businesses with financial and marketing components. These managers are producers not programmers.

[Mufferaw]

My point is not with respect to FoxPro versus Access but rather data formats and accessibility. As you are probably aware, Access can use dbf,csv, and other formats as well as its native container mdb format for data storage. Besides FoxPro, Clipper, dBase, and a number of other tools use dbf files. Farmers may not and should not have to be programmers, but they are consumers, and as such have a right to choice, including the choice 1 year, 2 years, or 5 years out to choose a product other than Settler Gold (Note that I am only using Settler Gold for illustration here, not ecause I have any particular bias for against this particular product). They also have a right to work with their accountants without their accountants necessarily also having to licence Settler's product for access to the data. Even if those accountants do have Settler Gold, they may or their clients (farmers) may require more advanced types of analysis not supported in Settler Gold but rather by custom apps in an Excel or Lotus 1-2-3 spreadsheet or a sophistcated analysis package like SAS. How then can this analysis be performed if the the data is stuck in an inaccessible proprietary file format? Farmers want business solutions and information to better manage their operations. They do not want technology getting in the way. Proprietary data formats have the potential to do just that, especially on a weak platform like MS Windows. If, as you say, farmers desire integrated packages that will allow production, financial, and marketing to be tied to each other for decision support, proprietary formats are not going to help them versus any open format, but they certainly have the potential to inhibit them (ie. all the risk is downside).

[Mufferaw]

I think farmers are caught between a rock and a hard place when it comes to integrated solutions and ODBC (Open Data Base Connectivity) standards. The fact of the matter is that so few farm managers demand the level of sophisticated analysis that you describe that it becomes cost prohibitive for a software developer to build an application that they can market profitably. <br> I agree with what your saying in terms of the ODBC issue being important. Applications such as accounting packages really need a seamless, easy way of exporting data into other packages even if the integrated solution is not available. That way a farm manager can develop his or her own system similar to the way a producer may piece together a production system utilizing different lines of machinery. A further aspect that compounds the exporting of data is the managers' own chart of accounts. Have you ever seen a chart of accounts that was identical between two producers? Yes, we have a ways to go, but we've come along way since the early '80's in terms of computerized management systems.

[Mufferaw]

Perhaps the solution is to stay away from ODBC and only use open standards like csv, dbf, etc. Most of these can be attached to the Microsoft JET Engine and hence used in Windows programs like the current Settler software, etc. I note that too often even I refer to ODBC as a standard which it is not. ODBC is proprietary technology by Microsoft. Standards are set by organizations like American National Standards Institute (ANSI). As per the level of analysis required by farmers, I think the success of the PigCHAMP program at University of Minnesota is a prime example of the level of sophistication farmers are willing to support. It is not the analysis itself that the farmers on this program require, but rather the decision support that comes out of it to improve their operations from a herd health and profitability perspective. You will also note if you have ever worked with PigCHAMP data that it is relatively quite open and hence can be imported fairly easily into the university's Oracle database for subsequent analysis with SAS, S-Plus, or whatever other tool they may choose. I agree we have come a ways in the past decade or so, but much yet is to be done, and the last thing we need to do is throw future hurdles in our paths. You are correct that no two farmers use the same chart of accounts, but that in and of itself is one more reason for keeping data in open formats so that it can be more easily determined how the farmer did chart his/her accounts at a future date. It impresses me that perhaps a more professional approach is required to advising farmers on management information systems. A good analogy is that of a patient going to the doctor. The patient only expects one thing; to get better with the doctors help. The patient may have no knowledge of anatomy, epidemiology, biochemistry, etc., but that does not excuse the doctor from using the current knowledge from these fields to serve the patient no matter how many years it has been since s/he was trained. Indeed, if the doctor does not, and the patient suffers, the doctor is liable.

[Mufferaw]

It may be that Settler liked the gooy interface that Access has. Personnally I prefer FoxPro front end and an Imprise (Borland) backend. This provides flexibility and openness.

[Mufferaw]

I have worked with both FoxPro and Access. I don't see much difference at the front end, but still prefer our Java front ends (lower investment risk to the client in the longterm). As per the backend, anything but Access almost. Current project I am working on we are using Postgres on Solaris. Very nice for a piece of free software. It is actually the origin of Informix Illustra.