On Wednesday 05 August 2009, Pranesh Prakash wrote: > On Tue, Aug 4, 2009 at 20:02, jtd<[email protected]> wrote: > > Only an idiot will commit a folly of proposing and approving multiple > > standards when there exists an opportunity of avoiding it. > > That assumes that there are costs involved in multiple standards, > which is what I'm seeking to question (not necessarily to refute).
We do not assume. This is an observable fact in EVERY case of multiple standards. > My > question is mainly about concurrent usage of multiple open standards > when the costs incurred are questionable (i.e., it is highly unlikely > that there will come a situation where a program like OpenOffice.org > will stop supporting .txt files). The Costs involved are hughe. 1) Costs of application development requiring support of multiple sandards 2) conversion costs (power, storage, cpu cycles, documentation, training etc) of data stored in multiple standard. 3) The above costs will recur continuously for the lifetime of the data. > Sometimes it is more convenient to > use .txt while at other times .odt is more convenient (at least at a > personal level, as I consider things like file size / support for > formatting / ease of opening for others, etc.). What would be the > arguments against such concurrent use? Processing a .txt file is a fundamental requirement of any text processor. It is like saying a computer-software system may support binary, hex and decimal numbers. Therefore it is supporting multiple standards. And the comparison between .txt - a file without any formating info and odt (or some other format) - a file system with huge amount of formatting info is specious - an apple to steel ball comparison, both being more or less round. You are comparing the most basic means of human computer interaction, with > Thanks for the figures on the railway gauge question. Now contrast > that situation with roads of different widths. The difference between > the two is that while the different sizes in the former (railway > lines) lead to incompatibility (for the trains), the different sizes > in the latter does not. It does and once again a very poor comparison. There are standards (in many parts of the world) set for road widths and types followed more in the breach than anything else. And who bears the cost? Everyone. While varying road widths dont prevent you from going from a to b (a basic requirement to be called a road), they impose massive costs on everybody except the contractor who built the road. Most of us in India are well aware of the problem. There are any number of roads where a car can enter but larger vehicles cant. There are innumerable national and state highways whose widths vary drastically, resulting in traffic jams and accidents and all the associated costs associated with the inefficiency of not following standards and not documenting in an open and transparent manner the deviations. Finally in the case of roads there are the occasional forces of nature that will dictate deviation from standards. With software, i am yet to come across deviations being forced by forces of nature. > So as long as the vehicle (the software) is > capable of driving on (using) the different standards, are they really > a problem? You arguments for multiple standards are very tepid. Even in the case of open standards, there will be a hughe cost on a day to day basis while catering to multiple standards, particularly when scaled up to several billion transactions a day. -- Rgds JTD _______________________________________________ network mailing list [email protected] http://lists.fosscom.in/listinfo.cgi/network-fosscom.in
