Michael, [EMAIL PROTECTED] said: >This discussion is going to come up repeatedly since this mail list is >not archived and I don't see CTM making the change any time soon. I don't >think it is a simple thing to change search engines woven into a program.
If the codebase is ObjectOriented in structure, which I suspect it is at least partially, then it's not a lot harder than replacing calls to the search engine interface object and having another kind of index build process. As Applescript support is so well developed in Powermail it is at least theoretically viable to build a separate search engine, if you'd have the time and the motivation. Not that I actually suggest that as a solution. It's much better if PowerMail solves this problem as elegant as they removed the search bugs in 4.1.x. I certainly have belief that they can, if they want.While that is not without cost, it is within reach of an interested developer group. I guess I'm questioning CTMs dedication to the problem issues with PowerMail. Letting them know that you want improvements in the searching area, can only help everybody. If you have another problems you'd like to see solved first, may I suggest you put that forward instead, first to the list and then to CTM? >Again, you make an example of something and provide no information on >where these show up in the RANKINGS. Did you search for "bon" and the >"bön" listings showed up first? If so, then it's wrong. If not, then it >is just a function of an engine that locates all the possibilities and >ranks the best ones first in a list. This would only be the case if it would ALSO list "bin", "ban", "ben", "bun", "bån" and "bän" further down the list, because in most languages "ö" is a totally different wovel than "o". And as I said before, if it isn't in a language, the behavior should be influenceable by the user. I haven't found that this response pattern is what PM is producing. Feel free to prove me wrong here. >Bob Parks and I were clear >that our experiences with search engines allows us to make use of >PowerMail's search engine better than many others. Yes, but your implicit message was that the interface couldn't be improved to give better expectations. I disagree. >Even if all of this was explained in a manual as you seem to wish, I >doubt it would change much. The majority of people don't read manuals, >nor do they understand the subtle differences among search engines. Agreed, it's the interface that have to explain why it gives the results it gives. There may be a multitude of ways of doing that. I could suggest a few approaches, but not before CTM have expressed any interest in exploring a solution to the problem. I suppose their biggest problem is that they'd have to acknowledge there was a problem to begin with.
