sqlite server windows

if any one have this working with Xojo


i love this!!

Well SQLite is definitely not my engine of choice, but if that’s what you’re looking for, cubeSQL is very hard to beat.

my system need to be use be able to used by one person (standalone) or a few users (server) using the same data file. With CubeSQL all i need to do is just change the indicator standalone or server. my program has already cater for both.

thom, what database do u used instead??

Can it be use by one person without setting up the server and then move to multiuser by setting the database server and putting the data file into a folder in the database server???

Not really. I use PostgreSQL for my anything where possible and I need data integrity. PG’s strength, in my opinion, is how well it handles your data.

I’ve brought up this example before, but I’ll type it again:

Consider a VARCHAR(10) column, which you are trying to insert a 12-character string. What do you expect to happen?
SQLite: Will accept and store all 12 characters.
MySQL: Will accept but only store the first 10 characters. No errors or warnings.
PostgreSQL: Will error.

SQLite is a simple engine, requiring you to do all your data type checking in code. This sounds fine in theory, but in practice, we’re only human. Mistakes will happen. In this example, MySQL is perhaps the worst though.

PG also has MUCH simpler prepared statements. With a simple extension method, I can fire off a parameterized SQLSelect with a single line.

Don’t get me wrong, I use SQLite for small-scale stuff like in-app support databases. But for mission-critical data, it simply isn’t enough.

how much difference is there between SQL Statement for PostgresSQL and SQLite?? can we simply move view on SQLite to PostgresSQL??

the CubeSQL is good but very expensive.

i looking for Embedded Server

not when they have their sales on. Also if it is less than 4 users , then it is free.

I successfully migrated a large SQLite database to PostgreSQL. But it took a lot of work and required a Xojo console app to “massage” the data so it was acceptable.

Taking my earlier example, since you’ll get an error in PG if you try to move that 12 character string into a 10 character VARCHAR, you’ll get an error. So you need to decide what the right thing to do is. Should the column become a 12 character VARCHAR, should you truncate, or should you drop the row entirely? Only you can answer that question.

Mostly PG’s syntax is nearly identical to SQLite. A few notable differences:

  1. There’s no autoincrement primary key. Instead, you should define your primary key as “SERIAL PRIMARY KEY”. This will create a sequence for you, set the column to INTEGER, and set the default value to nextval(sequence). It’s a minor distinction, with powerful abilities.
  2. Limit 0,1 syntax doesn’t work. You need to use OFFSET X LIMIT Y instead.
  3. Aggregate functions are a total pain, because they are strict. This is hard to explain, but it basically means you cannot use a GROUP BY unless you select the column, for example. After you get through the initial cursing and read the docs, it’s obvious that the other engines are actually wrong… but convenient.
  4. Prepared statements are easier because your variables are numbered ($1, $2, $3, etc) and do not require typing in Xojo. It’s perfectly acceptable to use the same numbered variable multiple times in a statement, so you only need to bind it once. Because it does not require typing like SQLite does, you can use a ParamArray of variants to automatically bind the values for you, and turn it all into a single line. Dramatically more convenient.

There’s a whole ton of differences, but those are the most notable I’ve found when it comes to actual syntax and code.

[quote=61147:@Alexis Colon Lugo]the CubeSQL is good but very expensive.

i looking for Embedded Server[/quote]
Sorry for hijacking the thread.

Considering the amount of work Marco has put into cubeSQL, I’d be surprised if SQLitening can be anywhere near as good. Making a multi-user database server is WAY more work than one would expect initially.

This is a “get what you pay for” scenario. Do you want something good or cheap? I don’t think you can have both in this case.

to me is no a problem to pay $999.00 for one time.

but pay them every year to distribute is very expensive.

Thom, so according to u, it need a lot of work migrating and just a one way migration.

[quote=61152:@Alexis Colon Lugo]to me is no a problem to pay $999.00 for one time.

but pay them every year to distribute is very expensive.[/quote]

why do u need to pay $999 for up to 3 users when u can get it for free??

I didn’t think you could deploy to your customers for free with CubeSql. I thought you had to purchase a license to sell commercial software utilizing CubeSql regardless of the amount of connections.

I asked Marco a while back if it was acceptable for my users to download/register their own free copy of CubeSQL to use with my software.

He said it was.

CubeSql is a great choice for a SQLite database server… plus Marco has always been very helpful when I’ve had any questions. I’ve had a couple of schools running some of my software on CubeSql server for about 1.5 years now and they love it.

i always grab hold of unlimited connection when they have the sales on. i just install for my client with my information or if the client insist, ask marco to change it to the client’s name if the database server is throw up as part of the application or something we tell the client to go to the website to buy the appropriate one.

[quote=61157:@Johnny Harris]I asked Marco a while back if it was acceptable for my users to download/register their own free copy of CubeSQL to use with my software.

He said it was.[/quote]

That is a possible solution but I suppose it may not be viable for some developers who want to deploy to many-many users of which I’m sure some would not want to be bothered with installing additional software. That being said all things being equal (pricing especially) I would use CubeSQL if I actually had an app to sell to the masses.