It's http://fox.wikis.com/wc.dll?Wiki~MatXtab~VFP

I made few preliminary comments. I will need to look at this program closely and test it better, so I'm going to do it tonight at home...

>Hi Nadya,
>
>I've started a new topic on the wikis on MatXtab. We can discuss how to improve this better in that environment.
>
>Thanks, Val
>
>>>Hi Nadya,
>>>
>>>I don't have your test data so can't comment. <s>
>>
>>I sent to you an e-mail with the structure of test data, which failed the program. Anyway, I'll send you the test data.
>>
>>>FYI - I will start a MatxTab wiki for a new version of this program. I've already check'd it with Steven B.
>>>
>>>- Val
>>
>>Give me a link, when it would be ready, ok? Thanks a lot in advance.
>>
>>>>>Hi Nadya,
>>>>>
>>>>>Matxtab works very well for millions of rows. It is a single pass solution. This does not mean that alternatives exist. I will look at Alexanders code as well.
>>>>>
>>>>>With Matxtab, you don't need a progrees indicator - just use the built in from VFP.
>>>>
>>>>Ok. Do you know, why it failed on my simple test data? I probably look closer and send it to home.
>>>>
>>>>Thanks again.
>>>>
>>>>>
>>>>>>>Nadya
>>>>>>>
>>>>>>>It seems you don't understand what I wrote. I never suggested you want to add thermometers to the SQL. MatXTab does its cross tab with one SQL command. If this other program does that, the SQL thermometer would preclude the need for some custom thermometer. If it does not use SQL to do its cross tab generation (as you seem to mean when you say its "scanning cells cursor"), then adding the thermometer would slow it down. I've used Matxtab with millions of records on much slower hardware and I just took advantage of the SQL thermometer.
>>>>>>
>>>>>>Hi Mike,
>>>>>>
>>>>>>Could you please send it to my home address (nosonov@msn.com), so I can take a closer look in it? As I explained, I haven't really looked into the logic, because I got another program working, and it was very quick. I ran a quick test of MaxTab on my data and got an error, so I haven't tried to solve it.
>>>>>>>
>>>>>>>>>But if the program is ilke MatXtab it basically does the cross tabbing with an SQL and the SQL thermometer would be available. Adding a custom one would slow down the process.
>>>>>>>>
>>>>>>>>In the FastXTab program final result is created by scanning Cells cursor. So, theoretically, if the original file had MM records, thermometer would help. As I said, it should be optional, with default as NO. I'm not suggesting to add thermometer to SQLs.
>>>>>>>>
>>>>>>>>>
>>>>>>>>>>>>5) Add thermometer (this would be an option)
>>>>>>>>>>>
>>>>>>>>>>>Nadya,
>>>>>>>>>>>I'm not sure that it's really needed, especially on modern computers. What are you thinking about it?
>>>>>>>>>>>
>>>>>>>>>>>Alexander
>>>>>>>>>>
>>>>>>>>>>In my examples your program took less than half of a second, however, what if the original file has millions of records? If so, I guess, final scan of Cells cursor may need a thermometer... Basically, I think, this portion of code could be taken from VFPXTab program (this program already have a thermometer). By default it would be turned off, but there would be a logical flag, which allows to add it, if needed.