bb.donnay-software.com

There is a possibility of combining many databases, even in the simplest case, with the same structure, one file? I need to work with dozens and perhaps hundreds of thousands of identical databases (database - part), related by one-to-one, as a single database. However, I hope that each of these databases - the parts can be up to 2 GB. I need it because I can not create a database with a number of fields, a little more than 1,500, and an alternative to such a structure is to use many of the same database, related by one-to-one. Create more than 1,500 databases in the same folder is also not desirable. And 2 GB - it's too little

Eugene Lutsenko wrote:I need it because I can not create a database with a number of fields, a little more than 1,500, and an alternative to such a structure is to use many of the same database, related by one-to-one.

you shoud overthink your Data Design ... it is not common to have so many fields even using SQL Server.
you can store big Array into binary Memo but you need a lot of Handle ( Ask Alaska for a Fix )

what about "compress" you Data ?

p.s. you can handle maximal 65000 Workareas in a Workspace

Actually, I'm very much looking forward to the ADS, but I guess you just have to use dozens of databases on the field in 1500. But logically it complicates the software implementation. And what is the maximum size of the memo field? I tried to book the size of the arrays in memory - does not work, takes off for lack of memory. It may not be any use their own standard databases, work with these databases as files? Perhaps AK can operate with other standards databases not DBF, that do not have such severe constraints on the number of fields and database file? And that is, in this respect, suitable in Alaska Fix?

I will make your standard database without any restrictions, and the basic functions of working with him and I will use it.

Maybe you should consider using arrays for your data and saving the arrays to array files.

You can use DC_ASave() and DC_ARestore().

if it were possible to reserve a huge array dimension to work with them, written in the form of files, read their files, and so on, it would be the most convenient solution, even better than the database. But I tried to back up such an array of columns on 30000 100000 lines and it unfortunately does not work on memory.

In general, it is strange that the arrays are created entirely in memory without swapping (without the use of disk space as RAM)