What is the cost of Sybase Sql Server Anywhere to deploy per seat?

I finally managed to find a price on the Sybase site.

Can you buy licences in blocks of 5 for $72.  Seems like the price is too high... unless of course software sells for a very high price point which is not the situation in this case.

Any thought?

SQL Anywhere is a full-blown no-compromises relational database management system.  It scales from a cell phone to a 16 processor server and runs on a host of different operating systems and platforms.  I am kind of surprised you find $72 per seat to be high.

I am not sure what kind of software you are selling that would be crushed by adding $72 to the cost but you will have to sell an awful lot of it to make any money.  Unless you are going to sell thousands of units, a software product priced at $100 will never pay you a reasonable sum for your time.

If this is a serious product and your volumes are high enough to make it worthwhile, you can talk with Sybase about an OEM license to embed it in your software.  If you are buying 100 licenses at a time for a small, embedded application, you might be able to get each license for substantially less.

Your other choice is to go with an open source database.  Everybody and his brother uses MySQL which is, in spite of SUN buying it, not nearly as feature rich and robust as SQL Anywhere.  PostgreSQL / EnterpriseDB is much more mature and comes closer to having a comparable feature set and is the way I go when I need to use an open source database.  Personally, I would jump al over SQL Anywhere, even at $72 per deployed license.

So what, if you don't mind us asking, is this software product and how does the database fit into it?

Regards,
Bill

I know Sybase do a quality product, although I have no experience of Sql Anywhere.

I figure that $72 is okay for a sales force, an executive, something that plugs into a corporate infrastructure, or a piece of work being sold for > $400 a piece.

This is is a situation of large set of observations with relationships arround how the observation set came about.  Perhaps if I can partition observations from the rest.  But now I have bespoke work for the observations, and need to chose another database for the relationships (ie. how the observations came about + what they relate to).

Maybe there is an open source project that can help me apply simple relational operators < > <= >=

Is SQL Anywhere easy to set up on a users PC?

I just discovered http://www.firebirdsql.org/index.php?op=history.  I wonder if this is up to the task.

What is the performance of SQL Anywhere like?

Bill: thanks for mentioning the OEM possibility.  

SQL Anywhere is THE embedded database of choice.  It has over 70% of the mobile market in spite of Oracle and IBM investing great gobs of money (unsuccessfully) to make inroads into that market space.

On of the nice things about SQL Anywhere is that it does not need a lot of care and feeding.  Properly configured, it can run for years in an embedded application with no DBA support at all.  If you troll the EE archives, you will find there is no shortage of questions from folks who have very old SQL Anywhere databases (v6 and even v5) that have been running forever and now have a problem because the hardware has died.  Nobody knows anything about the systems/dbs because no one has had to touch them literally in years.

SQL Anywhere has a very small footprint, especially given its high level of functionality and rich feature set.  It is easy to install and maintain remotely.

I did a project for a large bank in which re-architected a trust management workstation product that talked with the vendor's own mainframe system.  I used SQL Anywhere on the trust manager laptops both to improve performance (response time in minutes was reduced to seconds) and to all for local caching of data for disconnected operation.  Installation and updates were performed in the field by the trust managers themselves and we never had a problem.

SQL Anywhere has always been known for excellent performance.  It benefits from having a small code base as well as the ASE database theorists around to draw on.

Again, you really did not say what the application is or why you feel it will be priced at around $100.  What are your data volumes and what form do your "observations" take?  Are these CLOBs, BLOBs, or simple relational records?  What platform(s) will this run on?

Regards,
Bill

Hi Bill,

thanks for taking an interest.  The observations themselves are simple (just an n element vector).  My prototype database in access is a small data set of 402,000 rows.  An upper bound to growth is probably about 100 M.  The average deployed size will probably be 15 M rows.

Nothing fancy in the way of relationships.  1:N

I was always impressed with Sybase for performance..

On the target audience, you are right, I really don't know enough about them at this stage.  I think that possibly there are enough people so that I can make a small but healthy income to supplement the day job.  I don't think there will be enough people to build a business out of.

It's a bit like a university project that has a practical application.  I like quality, and it will give me a chance to consolidate my IT experiences into one place.

Well done on your banking solution!  A company would view this as added productivity / possibly makes something business critical more robust.  The value to them is high.

Would a consumer also view this in the same way?

BTW: back to technology.  Do you think that stored procedures would be comparable to pulling the data out and doing C++ to process on a piecemeal basis?

I really like the unheard of sybase reliability you mention.  Wow.  Deploying something on par with ms access is a headache I can do without.

Platforms: I'm an ms/intel-o-phile, so x86.  Going cross platform would mean learning modern development languages (C# or Java).  Perhaps some other platforms.  

-Steven

As an architecture for database applications, I always stuff as much as I can (within reason) into stored procedures.  That is usually the highest performance route as the SQL is pre-parsed, pre-optimized, and pre-compiled.

Using stored procedures also centralizes the code, allows for much easier tuning and debugging, and generally simplifies your application code.  As a rule, I avoid putting SQL in application code at all.  It is a very rare exception that I break that rule.

As far as going cross-platform is concerned, you really don't need to learn another language; particularly not C# as it does not run cross-platform well.  My guys do a lot of API and some application development work in C++ using Visual Studio.  As an example, we do development through unit test on Windows then build it on Linux for further test and deployment.  It is quite possible to write C++ that targets both Windows and Linux without code changes for 99.9 percent of what you would normally do but you have to avoid the incompatibilities when you write your code.

Depending on your requirements, there are other languages like Python and Ruby/Ruby on Rails that greatly reduce the amount of code you have to write compared with C++ and Java but still provide more than adequate performance on todays hardware.

Best of luck to you with your project.

Regards,
Bill

Hi Bill,

I'd be interested in hearing more about writing in C++ for multiple platforms.  Are you doing Intel only?  Do you consider apple platforms?  Do you consider the same approach for mobile applications?

I have considered using adaptors for bespoke UI in different target platforms.  The write once run anywhere crowd think they've got the advantage when it comes to the new wave of technology deployment.

-Steven T.

As always, you ask a question and then you get to what you are really looking for.  I really like Sybase, but unfortunately they did not come to the table because the company I work for is new and doesn't have a credit history (BTW: they don't use credit + believe in being debt free.  So Sybase will probably never look favourably upon us.  Hey that's okay.  I don't hold grudges, I'll still use them in the corporate space when they are the better choice)

maybe this is not a bad thing, because frustration lead to looking at other offerings... some of which are more tailored to the exact need.

Many thanks to Bill for his high level advice on the next generation launguages + platform decisions.  Stay in touch Bill.