Is the future of jrev in doubt? I have developed some java classes to access LH data directly from the files. I have also created a library of AREV migration tools. Is anybody interested?

Would anybody from RTI like to comment upon licence implications. I don?t want to upset Jim!

I am interested please contact me at data1448@pacbell.net

Thanks,

Tom

I'm interested in ANYTHING that can bring LH and most importantly

MV and AMV into the GUI world. Please respond to jim@firmsoftware.com

We are currently working with Larry Wilson on this. Thank you. JME

Vannessa,

At your convenience, could you please e-mail additional information.

Thank you very much,

rayc@symmetryinfo.com onmouseover=window.status=imagine … ;return(true)"

Symmetry Info

Ray Chan ~ Symmetry Info

Vannessa,

Very exicting news! I would like additional information too.

Thanks

Ralph Johler

Jefferson Center

ralphj@jcmh.org

Maybe you should ask who ISN'T interested!

I would deffinitely be interested in more information.

SBerger@coh.org

S

I would be interested.

steve@state-of-the-art.com.au

Please send more info

mrc@harbornet.com

Vannessa,

Cant see where there is much doubt .. although the primary developers of OI and JRev are no longer at the company I am sure that JA will bring in someone to complete the work.

dsig@teleport.com onmouseover=window.status=imagine … a company that supports their product from pride not extortion;return(true)"

David Tod Sigafoos ~ SigSolutions

dsig@teleport.com Cell:503-341-2983

Please send me additional info.

wilhelm@swissinfo.org

I'm looking for a way to retain the value of 10 years worth of VM and AVM data in a GUI way.ANTHING that can help me is of interest.

mikeb@primesight.co.uk

Thanks

please send me more information to

spiegel@cosmotron.cz

thanx

R.Spiegel

Has anyone been contacted by this person yet? Is this legitimate?

dbakke@srpcs.com

SRP Computer Solutions

We'll jump on the band-wagon as well . Could you please send more info to vsl@vernonsystems.com

Thanks, Matt

Are these Java classes able to stand on their own, or do they rely on calls to OpenEngine? Do they support data entry, file resizing, etc? If they do, and if you maintain the same linear hash structure, how are you doing this so it does not infringe on RTI's intellectual property rights? You said that you have AREV conversion routines. Do these routines convert windows and menus too?

Please send more info to DReese@concentric.net

I would be most interested in receiving information on you java conversion utility tools.

We have developed a very powerful Arev survey/test/polling system. We had hoped to convert it to OI but this proved to be to expensive from a number of points of view.

1. The conversion/development time and cost.

2. The deployment costs - runtimes etc

3. The lack of support for charts and the 'report' tools.

4. The availability of Arev/OI programmers.

5. etc etc

Your java utilities may be just the answer. I would appreciate receiving the information on your java utilities.

Mark

Hi Don and all you patient people,

My Java classes directly read the LH files. There is no requirement to have any RTI system installed, hence my concern about licencing.

I believe the original LH work was published by Witold Lutwin many years ago, and is in the public domain.

However if RTI have a legitimate claim to the file structure, and can stop people from accessing it without an RTI tool, then I will not be able to release my system.

I have not had any comment from Jim or anybody else at RTI on this issue. Does anybody know the extent of the copyright of a file format?

We assume that you mean Witold LITWIN. The query would be if you are convinced that this is the algorithm used - how are you so convinced? You can't decompile their object code according to their license and you can't reveal proprietary info gained during employment at RTI?

Either way this was 20 years ago - surely there must be some statute of limitations .

http://www.informatik.uni-trier.de/~ley/db/conf/vldb/Litwin80.html

The Sprezzatura Group

World Leaders in all things RevSoft

Hi Vannessa, I would would like to understand your solution better, If you have the opportunity, please give me a call at 978-247-7101, thanks Jim

http://www.informatik.uni-trier.de/~ley/db/conf/vldb/Larson80.html

and look where he works now?

http://www.research.microsoft.com/~palarson/

Vannessa,

One thing I've discovered is that things really get done when individuals like you want to contribute thier knowledge for the benefit of all without expecting anything in return (i.e. $$$$).

The bottom line is that there are thousands of small, medium, and very large applications around the world written using some version of Revelation (REV G, AREV 2.xx AREV 3.xx, etc.), and they deserve a migration path into the future.

I have a core system that can do the basics such as attaching LH volumes, opening data and dict tables, etc. There's menu/window/popup classes that directly read the AREV definitions.

Obviously the missing link is being able to use that small collection RBASIC code commonly referred to as "the application". Once you have this, most of the system functionality exists. I would imagine the only requirement is to not use any Revelation Technologies source code to fill in the gaps (without permission).

I have found an Open Source project that is currently working on a Pick-like BASIC to Java compiler. It works. All it needs is attention in the additional flavors needed (i.e. REV A-G, AREV).

Database access can go several ways. I was looking for a parallel AREV / Java operating environment, but it's not really necessary. The initial benefits would be to use the existing tools to build the windows, menus, popups, etc.

I don't mean to trivialize this, and I hope I'm not repeating another's thoughts completely, but all the pieces exist in one form or another. Let's give a few hundred thousand people a gift. My current high-level brain dump is below. Hope this helps.

Links

www.maverick-dbms.org - The Pick-like BASIC to Java compiler

                       and multi-value db environment.  This

                       is the best I've found.  Please share

                       any others you find.

sourceforge.net - Open Source management site.

* Core Classes - BASIC compiler, OCONV, READ, WRITE, LOCK, etc. AREV.EXE & SYSOBJ type functionality. Have some + utilize open source. Tool Classes - Like PAINT, DICT, EDIT. Many non-RevTech programs exist that can do these things. Compile them into Java and run initially, rewrite when time allows. Menu Processor - Have. Window Processor - Have. Popup Processor - Have. Report Writer - Using Third Party App. through JDBC. Could just compile your own RBASIC program and run it. SELECT Processor - Server should handle reduction. JLH Server didn't have it last I checked. Using temporary layer on client side. Application Classes - User developed code (dict symbolics too). Database Interface - Like RTP57, etc. JDBC Driver Rev LH Server - Have. Server=RevTech=$$. Limited functionality. UniData/UniVerse - Exists?, $$. MV DB Server - Your Multi-threaded server here please.

Curt Kastler

ckastler4@aol.com

-or-

curt@kastler.com (moving server, off now)

Curt,

Thanks for the information, especially the link to the Pick BASIC to Java complier. As you know it is very, very important. I was actually begining a project to do something along these lines myself but using the approach of an emulator similar to the LISP to Java system "Jess." Clearly the better approach is to try to contribute to this already ongoing development effort.

Tom Savell.

You need an intellectual property attorney to answer these questions, which am not. From what little I do know, you probably need to know what type of protections were used in the first place – was the original work copyrighted, patented, treated as a trade secret, etc? Did you ever have any access to any trade secrets and are those trade secrets used in your solution?

Even so, none of this would prevent you from creating a generic linear hash solution of your own, even if it was not 100% file compatible with RTI's Linear Hash structure. Users would just need to copy data from one linear hash structure to another, then use the new structure. The application should not care about the hashing algorithm in use, or the internal distribution of records in the frames.

You might want to check www.patents.ibm.com. I did a search on "linear hash" and got a number of hits, that lead to even more hits. Although I did not see any patents awarded to RTI, there are a number of active patents in this general field. Many provide specific functionality that probably would not apply to your project, but it is better to find out now instead of later.

View this thread on the forum...