Westlaw clone (Was: Re: Software to Block Ads)

From: Timothy Arnold-Moore <tja[_at_]mds.rmit.edu.au>
Date: Thu, 02 Apr 1998 11:05:28 +1000

Bob Stock <bstock[_at_]ucla.edu> writes:
>
> BTW, as a footnote, it's not clear to me that there would be absolutely
> no copying of the Westlaw software. In making the changes there might
> be some intermediate copying going on.

If you were simply changing Westlaw to Eastlaw, the only intermediate copy need be the RAM copy in which you substitue `Ea' for `We' and then write it out to disk again. Same goes for a change in phone number if the number has exactly the same number of digits. Its usually really easy to change strings in object code providing you don't change their length (unless the object code is encrypted but I doubt West's is).

I thought this group had reached a consensus that temporary RAM copies were not sufficiently permanent to constitue a reproduction. Its not a derivative work until you actually make changes so this imposes no extra complexity to the analysis.

> Finally, Ari makes it sound deceptively simple to use Westmate for some
> other service. It is very unlikely that just changing a few literals in
> the program would make it work with any other service unless all the
> software at the server end were virtually identical to what is located
> at West's servers.

I don't know a lot about the communications protocol used by West but I doubt its very complex at all.

You don't need `virtually identical' software on the server. You need software on the substitute server which speaks the same protocol which is very, very different. Ari did mention this.

Reverse engineering communications protocols is done all the time. (The only complication is if they do some sort of public key encryption of the data which then gets decrypted by the client. That would require the substitute service to have the private key used by West or for a substitute public key to be embedded in Westmate. I don't think they encrypt and if they do, given how long West has been in business, I doubt whether the key would be particularly difficult to crack ;-)

The search engine could be implemented completely differently as long as the query instructions could be translated into operations on the new search engine. The communications code will probably look pretty similar but that is largely determined by functionality and is therefore not in danger of infringing (at least not in the US).

> And a lot more would have to be changed even at the client end.

Sorry. I am struggling to think what would need to be changed in the client. Could you be more specific? The only factor I can think of is the West numbering system and the case headnotes.

Lets assume we make no further changes to the client. Eastlaw would then have to provide its own headnotes and classifications in order to fully utilize the power of the access software. Headnotes would not be a problem as they can be created independently It depends how hard coded into the software the classification system is as to the options here.

If they are not hard-wired (i.e. read in from a configuration file) then Eastlaw can provide a completely independent classification system and classifications so there is no danger of infringment.

If they are hard-wired either Eastlaw has to provide classifications using West's numbering system or the client needs further modifications. I personally see no reason why independent classification of the cases using West's classification scheme should be a problem but I'm not convinced every federal judge would agree with me. Any thoughts?

> I know this muddies up Ari's hypo, but, technically speaking, it is
> tough to make the whole thing work.

I'm not sure its as tough as you make out nor as easy as Ari suggests, but the potential rewards would be great. Therefore even significant tehnical hurdles may well be worth jumping.

I suspect the major disincentives are:

- fear of litigation from West (even if there is no infringement),
- cost of compiling such a comprehensive database,
- cost of starting up a service with so many access points

  (communications and hardware costs),
- relative ease of starting a Web-based access service (with less   risk of litigation from West, lower initial start-up costs and   none of the interface issues).

Technical hurdles are way down the list.

Tim Arnold-Moore, LL.B. (Melb)        | Multimedia Database Systems, RMIT  |
tja[_at_]mds.rmit.edu.au  B.Sc.(Hons Melb) | 723 Swanston St    -----------------
Tel: +61 3 9282 2487   Fax: ..2490    | Carlton 3053       |  simul iustus 
    http://www.mds.rmit.edu.au/People/Tja/tja.html         |   et peccator 

Received on Thu Apr 02 1998 - 01:05:40 GMT