xsharp.eu

There's some text re "Free 2-User Server" version - did you understand this? I.e., could that mean, that after the trial period the install continues to work, accepting two users at max?

As far as I was able to understand, the fre 2-user server is completely identical to the normal server, but is limited to 2 users. And it is in the basic package, so one could try it out before buying the full server.

The ApolloDB server seems a very good value for its price.... use a server without installation, and multi user too...

Wolfgang

wriedmann wrote:I have done some tests with the ApolloDB engine, and I like the idea of an SQL layer over DBFs very, very much.
[...]
But I had to rethink about this:
[...]
- third the SQL layer seems to have some shortcomings. One of the most important IMHO is that you cannot use a "select * from" when the table contains a memo field, and you cannot use memofields directly in a select.
[...]

Hi Wolfgang,

When i look at the Apollo homepage it seems that the "Appolo dbf utility" uses datas created with the dbcre8 tool. So i created with that tool a 20000 record test.dbf / test.fpt ( memo field "notes" included ) and started a query with Foxpro 2.0 Dos. Yes, already in the early 1990th FP 2.0 supported "Select" on DBFs .

SELECT * FROM test WHERE test.notes like "%11%"

the Foxpro result shows 1451 hits

if Appolo really fails with memos, i´m wondering what the guys did the past 30 years ?

regards
Karl-Heinz

Hi Karl-Heinz,

you are right... At least a "select *" should work in this case.

But even Advantage does not permit memo contents in index expressions. Therefore I had to add extra fields (duplicated) in some tables, like an index on the first 100 bytes of the article description.

When the DBFCDX RDD is ready, I hope the development team creates both a SQL Engine over the DBFCDX and s standalone server - both as commercial products (or as subscriptions) to earn some money for the project. I'll be happy to buy them as they would introduce a clear plus for my applications and my development work.
These would make my applications easy to distribute, and I like the combination of SQL and direct access.

Wolfgang

Wolfgang,

wriedmann wrote:When the DBFCDX RDD is ready, I hope the development team creates both a SQL Engine over the DBFCDX and s standalone server - both as commercial products (or as subscriptions) to earn some money for the project. I'll be happy to buy them as they would introduce a clear plus for my applications and my development work.
These would make my applications easy to distribute, and I like the combination of SQL and direct access.

We can't make any promises about that. We will most certainly have a look at this.
However there are also other things on our list of features that are "nice to have".
But you should realise that we are adding DBF support to X# because that is needed for backward compatibility. But I personally would not use DBF anymore for new standalone code.
In that case something like SQLite is a much better choice I think.
For multi user code SQLite may not be the best choice. In that case any of the other SQL solutions should do. PostgresSQL if you want a free solution or otherwise MySQL, MsSQL or Oracle. This all depends on your market of course.

Robert

Hi Robert,

the most installations of my applications are either single PC or small networks with 2 to 3 PCs. And since our basic software is cheap (we are starting from 800 Euro for a software than can do only invoices, complete with form personaization), the installation should be done with copy only.
For single user installations SQLite works very well, it does not need anything that cannot fit in a zip file.
For larger installations with a dedicated server a PostgreSQL server is the best, but the gap is between them.

Wolfgang

Wolfgang,
i wonder, why you are shy to install non-dedicated PG in the small environments? It's not big, the install is dead easy,(even i managed to succeed <g>), the background imprint on performance is imho neglectible. If needed, copy a customized postgresql.conf, restart PG and forget it...
Re SqLite, as a user i suffer from it, as my banking app uses this - i have to admit, never had data loss. But performance is rather underwhelming with db size under 5MB.
Re Apollo: i'm not too keen to introduce another dependency to a closed source product where you never will know when it closes shop or get's bought out like Ads.

I don't want to be smart, just curious

Karl

Hi Karl,

I need a copy-only install, and an installation that every technician is able to move over to a new PC. With DBFs, they copy over the folder and all is done.

I don't know if such a thing is possible also with PostgreSQL.

For a larger web project, we are using PostgreSQL instead of MySQL together with PHP, and the two guys that work on it are very happy with it. But the entire software is running on a dedicated Linux server managed by myself.

For single user, SQLite is really great, I'm using it on my Door Configurator project, and there the entire configuration is stored in a SQLIte database that is changed on every program update, and every order is a single MySQL database that can easily be transferred back and forth.

Wolfgang

Hi Karl,

I have checked now better, and it seems PostgreSQL does not need any installation, and you can do all from your program itself without the need of a special installation program.

The steps to make PostgreSQL run seems to be very straightforward:

https://www.petrikainulainen.net/programming/tips-and-tricks/installing-postgresql-9-1-to-windows-7-from-the-binary-zip-distribution/

Wolfgang

Robert,

Would be great to have 2 solutions; a data processing very speedy batch oriented solution, and a GUI interface related solution.

For Advantage users, the basic API in ACE.DLL and DBFAXS library, would be a fantastic start.

We built our own super fast datasets (using Anthony Smith's RAMDBF as inspiration). The datasets cache/track changes, and generate/execute SQL code when ready to write the changes. Really good for large amounts of data processing.

Really like what we are seeing with xSharp.

Paul Bartlett