Want to read Slashdot from your mobile device? Point it at m.slashdot.org and keep reading!

 


Forgot your password?
Close
typodupeerror
×
Data Storage Programming IT Technology

Specialized, Open Source Databases? 34

Posted by Cliff from the preventing-the-reinvention-of-the-wheel dept.
PyTHON71 asks: "I've been asked the head of the Wichita State University Anthropology department to help fix his growing biological anthropology collection database. It's currently in Access (insert boo's and hisses), but he is willing to have it developed in a reliable open source format. Now, there are a lot of university departments out there that need to have specialized work done on a small budget. They can't rely on just any developer, because the developer has to know biological anthropology (in this case) as well as MySQL, etc. And since it's not in the Hacker Code to duplicate work that's already been done, I was wondering what specialized database projects are out there & available for general use (not the data, just the structure)."
This discussion has been archived. No new comments can be posted.

Specialized, Open Source Databases? More

Specialized, Open Source Databases?

Comments Filter:

  • Why? (Score:2, Insightful)

    by LordNimon ( 85072 )
    Why does the developer have to know biological anthropology? Can't the developer just work with the scientists to understand the data formats, and create the database for them, and then teach them how to use it?
    • I might offer a suggestion... Because a developer who already knows the field will need less training
      Also the relationships between the data will be more obvious to a developer already familiar with the background... And usually never happen to show up during training... Nobody ever thinks about them until problem X occurs, this client already seems to know that, and wants a solution where developer A writes something, and BANG they don't have to see him again... instead of the ongoing relationship a solu

    • Re:Why? (Score:3, Informative)

      by AndyElf ( 23331 )
      I guess his question is mis-phrased: he wants a ready application/database for that very purpose that is not proprietary.

  • What about Access? (Score:4, Insightful)

    by GTRacer ( 234395 ) <gtracer308NO@SPAMyahoo.com> on Monday April 07, 2003 @01:08PM (#5679781) Homepage Journal
    While I don't have a great love of the "Evil Empire" that is MS, what's so bad about Access? It's certainly not perfect, but tell me what Windows-based alternatives exist that will:

    * Let me create front-end forms populated and controlled by code and query

    * Support relational tables, indexes, complex JOIN queries, aggregate functions, etc.

    * Design complex reports with page/section formatting and behind-the-scenes code controls

    * Can access ODBC data sources

    If you can show me the way, I'll take the first step! I'd love to cut some of the distribution costs here!

    GTRacer
    - Needs help with gcc + PS2...

    • Re:What about Access? (Score:3, Interesting)

      by SLiK812 ( 518195 )
      The problem is Access's data storage is in a flat file format. Which means it'll be slower than dookie after it gets to be a certain size and complexity. Especially when sharing the database with many users. I'd suggest SQL Server, with a vb frontend or even an Access frontend with the SQL Server tables linked into Access.

      Unfortunately the question was pertaining to open source, probably because the owner of the data, doesn't want to drop a chunk of change on a major database. Yes Access comes with Win

      • Re:What about Access? (Score:3, Interesting)

        by metacosm ( 45796 )
        I am not sure about the data storage of access. It was my understanding it was NOT a flat file -- but it is very limitted.

        I recommend you use the "upsize" wizard (which will automatically relink to your existing forms) and goto MS SQL 2000.

        This does a few things for you. #1. Huge number of people familar with Access. (Better chance of finding a "biological anthropology" major with experience [I should note, I find this requirement a bit silly, have a guy from the CS department set it up, and you use it]
        • I recommend you use the "upsize" wizard (which will automatically relink to your existing forms) and goto MS SQL 2000.

          I agree wholeheartedly, but I don't have any money to spend and I'm trying to make the most of what I have. I may be able to convince I/S to give me SQL Server space, and if so, then I'm all set.

          But they may tell me to take a flying leap, and then I'll be forced to roll my own solution. My boss supports some small expenditure for this effort, but I'd rather spend the money on a true pro

        • I agree. If the jump to MySql or Postgres is too great, then Access should work out fine.

          Currently our Access 2000 database weighs in at 740 MB's with 4-5 users. And it's running fine. However, I'm planning to migrate to Postgres very soon.

          Proper choice of Primary/Foreign Keys, Indexing, and NORMALIZATION of the tables is *critical* to size and performance.

          3rd Normal form will do wonders for size and data integrity. Indexing/keys will do wonders for speed.

          If the developer is not properly versed with
      • I'm not sure where you get this from. Access is fully relational. The file format is just a packed collection of files stored as a single file. Similar to an oracle database which uses a single tablespace for all tables and stores the entire tablespace in a single datafile.

      • Perhaps this is (-1, Offtopic), but why suggest SQL Server with an Access front end, when you can suggest an Open Source/Free server, such as PostgreSQL, or MySQL, and an Access front end?

        I've never tried it myself, but I've heard that PostgreSQL [postgresql.org] can be an ODBC datasource on Windows machines.

        If you've already got Access, and you already have a viable data structure and forms/reports that work with it, you should be able to move the data to Postgres, change the tables in Access to ODBC links, and pretend l

    • Re:What about Access? (Score:2, Interesting)

      by Anonymous Coward
      Most of the time, whenever anyone posts about database, I like to point out Adaptive Server Anywhere (from Sybase). It's ....

      - inexpensive

      - probably the easiest DB to administer

      - fully SQL compliant, supports ESQL/C, ODBC, OLEDB, JDBC

      - available in linux, windows, mac, aix, etc. versions

      - available for free download (as an evaluation copy) off the sybase website

      - advanced query optimizer (blows most open source engines out of the water .... really)

      - fully ANSI transaction-oriented; stable in the eve
    • All you need is available with powerbuilder. Quick, easy reports, connect to most databases, use standard or 'enhanced' SQL inline, and you can compile it to native code for Windows or Unix. Easy language; main thing is learning what not to do in powerbuilder. kinda quirky. Maybe the latest is better.
      Not a plug by any means, but Access is nasty. My experience with it is when they need to port apps to a more robust, less buggy environment.

  • Sure (Score:1)

    by jsse ( 254124 )
    http://bioinformatics.org/ [bioinformatics.org]

    End of thread

  • database? (Score:3, Informative)

    by Kevin Stevens ( 227724 ) <kevstevNO@SPAMgmail.com> on Monday April 07, 2003 @01:20PM (#5679852)
    Why dont you explain what you mean by 'database' first. A database can mean different things. I am not sure if you want a complete database application, front end and all, or if you just want a database design- A UML model or E-R diagram, from which SQL DDL can be created. (or perhaps by open source you meant SQL code to Create a database). Your requirements are kind of blurry to me. Doing something like this should not really be all that difficult though. If you have a developer look at the kinds of data you are trying to store, and the relationships between those data (in depth knowledge is not required of what the data actually means, though I guess it helps) a developer should be able to come up with a Data model and front end relatively easy.

  • Do you really have to have this answered? (Score:3, Insightful)

    by dacarr ( 562277 ) on Monday April 07, 2003 @01:21PM (#5679862) Homepage Journal
    Asking for this beast is like asking for a version of M$ Excel (or, to be fair, Gnumeric) that is specifically designed toward balancing a company's books.

    A database is a database. There is no "specialized" database for any purpose, with the possible exception of design for platform (IE, DB/2 for AS400) - you build the structures in the DB program and go. It's what makes apps and engines like MySQL really cool.

    If you want a DB that is geared for bio-anthropology, fire up MySQL or PostGreSQL or insert your favorite engine here, build the structure for the DB you need to run, put a pretty face on it for the users, and kick back as it does the rest.

  • The requirements seem kind of vague, but if you don't want to go all the way, you could always just export the data to PostgreSQL or mySQL, and use the the existing access interface for a frontend, connecting over ODBC to the Open Source alternative - at least until you've transitioned it to something else.
  • Here's what I understand:
    you already have everything up and running on an access database
    you want to use a different database platform

    doesn't sound like you need to worry about designing anything, its already there.... especially if you want to go from access to MySQL (or any other of the same family), just build the database structure to match the old one, export to .csv from word, and import to your chosen database... interfacing should be similar, it depends on how you're accessing it (if web based, and

  • SQLITE (Score:2, Insightful)

    by ddriver ( 613483 )
    It supports almost all sql92, is blazingly fast, it has an ODBC interface and native api's for c and c++, is has a python db api 2.0 compliant interface. It has a really cool and versatile command line tool. It is imbedded though, so if you want to you it as a server you will want to write the implementation. I don't think that it would be too hard though, depending on what you wanted to do with it. I have just started a payables application withit and it looks like it will work out quite well.

    Oh, and it i

    • Re:SQLITE (Score:3, Funny)

      by ddriver ( 613483 )
      try www.sqlite.org

      Here is the licensing terms from source: /*
      ** 2001 September 15
      **
      ** The author disclaims copyright to this source code. In place of
      ** a legal notice, here is a blessing:
      **
      ** May you do good and not evil.
      ** May you find forgiveness for yourself and forgive others.
      ** May you share freely, never taking more than you give.
      **

      Dude I so dig that!
  • A generalized forum such as "Ask Slashdot" probably isn't the best place to ask this type of questions. The problem is that the domain is too specific. What you probably want to do is talk with other biological anthropologists and see if anyone has a schema that you can adapt to your needs.

    That being said, one thing that you might want to do is search Google for XML schemas. I do a lot of (relational) database design and I've noticed that there are XML schemas for just about everything. (Although I gen
    • A generalized forum such as "Ask Slashdot" probably isn't the best place to ask this type of questions.

      You'd think that...but amid all the noise and chatter of people pimping their own pet databases, they got your actually useful comment.

  • You will also have to ask who will continue to support the app/db combination in the future. With any dedicated database server, who will keep it patched and secure, and backed up in the future? I see the following options:

    1. Stay with Access:
    Security is based on file sharing, and up to the Network Administrators. However, the problems with Access, you are probably well aware of; a biggie is that the db can lock and require reboot of the server (which can really piss off your Network Administrator if the

    • 2. Migrate to MySQL (or other free DB) back-end with Access front-end:
      You will have to figure out who maintains the server setup. Access does not like to migrate to non MS databases, and you may have to totally redo the db structure and Access front-end.

      Actually, it's very easy to migrate Access to MySQL. Install the MySQL ODBC driver, then import the database and data into MySQL. You can then use Access as the frontend to the MySQL ODBC data source.

      • There are a few caveats depending on how a particular Access db is set up. He may have to re-structure some tables and re-write portions of the Access front-end to accomplish this. You're right that some people may not experience any problems. See this [mysql.com] for a bit more information.

    • 3. Migrate to MySQL (or other free DB) and rewrite front-end:
      Again, you will need to figure out server maintenance, as well as finding an open-source front end. I'd be positively giddy to find/learn about an Access/Oracle Forms-like open-source tool.

      I know that PostgreSQL has a semi-decent front end program called pgaccess (I know that it's in Linux, but I'm not sure about Windows), which you can use to create forms/reports for. And the best thing about it is that all of the form data is stored in the dat

  • Matt Dillon [backplane.com] (of FreeBSD fame, not the actor :-) has a new database startup company called Backplane [backplane.com]. They are creating a high-performance, distributed database which should be massively scalable (using many small servers ala Google's web farm). It's open source, too.

  • As I understand it, the *NIX equivalent of Access is Berkely DB [sleepycat.com]. Small, lightweight, stores all its information in a single file. They've got 4 different versions depending on your needs and support pretty much any OS.
  • Something that's important to look at is who will support this application down the road. Funding at universities can be dicey, and this thing may have to be supported by inhouse flunkies. What can they reasonably support? This doesn't necessarily mean that it needs to be in access because the inhouse flunkies only know access. If you could write a C++ front end that was stable and reliable, with an easy-to-maintain connection to a backend that they can support, that might be a better solution. That sa

Slashdot Top Deals

Recent investments will yield a slight profit.

Close