Hi Cort,

Here are some thoughts and replies, brainstorming as 'twere, giving feedback, don'tcha know ;), though very F'd. I'll snip what I don't answer.

Quote Originally Posted by Cort View Post
Yes, these are three important intermingled issues - deciding (a) what data we want to gather and then (b) how to collect it, store it and present it effectively and (c) how to analyze it.


So in an important sense those are first altho I think we should certainly proceed on all tracks as they we do have some ideas.

Quite so - if only to find out what is realistically possible.

Certainly if we could do it - it would be hugely helpful in terms of our ability to tweak it later as we surely would want to as new ideas present themselves and new findings come out.

Well... it seems to me quite doable for a considerable part using spreadsheets and html. That is, restated for my own site: Using e.g. Gnumeric and Frontpage I could easily display statistical information quite palatably and well-formatted.

One limit is that this is simply done on my HD with such data as are there but Gnumeric stuff - if only by simple greenscrab as gif, png or jpg: KISS - ought to be easily usable in principle (and there are other ways using html).

But unless one wants 'live internet gathering & displaying of data' that is in fact the best way: Simply writing articles for the web using a html-editor + a good spreadsheet.

However the form ends up in the end I hope there are people in the community who can educate themselves in the program and help to tweak it when necessary. I'm certainly going to try my best. I know my talents in that area are limited, however.

From my own experience: It helps if you can program. I learned to program when my ME was less severe, already knowing rather a lot of mathematical logic, from before my ME also. Learning to program from nought to a useful extent is difficult with ME and anyway not for more than 1 to 5% of healthy folks, so for my part disrecommended for persons with ME. (It requires a good and well-working memory.)

If a programming language is to be used, I propose FreeBasic, for the reasons given, and could give more - unless, of course, there be a plethora of C or C++-trained folks on the PR-forums . Also, programming for MS Windows without a good WinApi Resources Editor - the programmatical thingie that puts the windows and the editors and the lists and the graphs on screen, in schematic outline, WYSIWYG-editing wise, in a good one - is stooopid.

The other issue is the statistics end. I believe most computer languages have basic statistics functions we can easily use to get averages, proportions, etc. which we can use to visually display (via piecharts, etc.) interesting, if not statistically rigorous information.

Indeed, and for the major programming languages, all of which have open source implementations, there will be open source code available for such things. The problem is rather: (1) is there a good free compiler + rapid development environment for a reasonable language with (2) people on the forum who can or could program in it? (Could but alas cannot is also OK, for then one may have info.)

Gnumeric provides much more advanced statistics. I'm always wary about the difficulty of transferring programs to the web. How difficult would it be to hook up a function in gnumeric to datapoints in our relational database and then display it on the web? Is that simple, complicated but doable, problematic????

As above: No problem in principle to do a lot in a good spreadsheet and present the result in reasonable html.

statistics is certainly our weak point in this venture at this point - thanks for your insights :)

Yes, more fancy statistical analyses, including advice of what to (not) do or try, needs some professional advice, which I am seeking as I wrote before.

Absolutely we will wait until we have clarity with these issues before we jump forward. I think we're more at the beginning of the beginning than anything else. We're in the exploratory phase.


Best wishes,