Forum OpenACS Q&A: APM and FTS

Collapse
1: APM and FTS
Posted by G. Armour Van Horn on 03/07/02 10:20 AM
At one point, trying to open /acs-admin/apm/ logged lots of errors (three lines, repeating every 30 seconds) indicating that FTSEngine wasn't found, and wouldn't display the page. This is on RedHat 7.2, PG 7.2, and yesterday's nightly tarball: 
[07/Mar/2002:01:03:22][1218.2051][-sched-] Notice: Running scheduled proc search_indexer...
[07/Mar/2002:01:03:22][1218.2051][-sched-] Warning: APM: FtsEngineDriver does not exist
[07/Mar/2002:01:03:22][1218.2051][-sched-] Notice: Done running scheduled proc search_indexer.

Using the command from https://openacs.org/doc/openacs-4/nextsteps.html#install-next-openfts to install FTS 

[nsadmin@vox Search-OpenFTS-tcl-0.2]$ ./configure --with-aolserver-src=/usr/local/src/aolserver/aolserver
things ran smoothly until 
checking looking for postgresql include files... configure: error: couldn't find postgresql include files
I also tried 
./configure --with-aolserver-src=/usr/local/src/aolserver/aolserver --with-pgsql=/usr/local/pgsql
with no change. Anything else I should try? I searched the archives and found a number of messages that seem to address problems with installing OpenFTS, but none quite this early in the process. Has anyone been through this with something like my setup? I'd love to get it figured out to hand to Vinod, the current docs only address Debian.

Van

Collapse
2: Response to APM and FTS (response to 1)
Posted by Neophytos Demetriou on 03/07/02 10:33 AM
Van Horn, there are no problems installing OpenFTS (the openacs package). The message you get is just a warning in the aolserver logs that says there is no FTS engine installed (not necessarily OpenFTS). Remember the search package supports multiple search engines through the use of service contracts (FTSEngineDriver).

Now, w.r.t the problem you get when you try to install OpenFTS (not the OpenACS package but the real thing) is due to the fact that the installation scripts are based on PostgreSQL 7.1. Postgresql 7.2 changed the place where include files are stored (serve side includes). Therefore, the error you get. As you understand it was difficult to foresee such a change last Summer when OpenFTS was first released. The new (upcoming) OpenFTS version is based and requires 7.2.

Collapse
3: Response to APM and FTS (response to 1)
Posted by G. Armour Van Horn on 03/07/02 10:56 AM
Glad to see you back to answer my questions, thank you.

Are you suggesting that I don't need to install the OpenFTS program to use the OpenFTS ACS package? The nextsteps.html of the docs suggest that it is a prerequisite. Do you know how long it is likely to be before the program is updated for PG 7.2? I certainly can live without the function for a while.

Van

Collapse
4: Response to APM and FTS (response to 1)
Posted by Neophytos Demetriou on 03/07/02 11:12 AM
<i>Glad to see you back to answer my questions, thank you.</i>
<p>
No problem. I'm still around. I'm just taking some time off to sort things out (too much work these days).

<p>You still need to install OpenFTS (just that the installation instructions and scripts are PG-7.1.3-oriented). The OpenACS package is only an interface (a driver) to the actual search engine.  If you want to use OpenFTS, you should either settle with PG 7.1.3 or try to modify Makefile.global under the OpenFTS src directory. I did that once but I can't find the files anymore, probably deleted, so I'm not sure if you have to change anything else.
<p>
On a similar track, I think that Dan has checked in the TCL port of the new OpenFTS version (note: I'm talking about the search engine, not the openacs package -- you can get it from http://openfts.sourceforge.net). Next few days, I'm gonna prepare the documentation for the new OpenFTS version but it's only after this release that the new version is going to be available for OpenACS (I'm going to write about this in another thread soon, I hope).

Collapse
5: Response to APM and FTS (response to 1)
Posted by Dan Wickstrom on 03/07/02 02:02 PM
It's checked in but not tested.  Hopefully I'll get to it this weekend.
Collapse
6: Response to APM and FTS (response to 1)
Posted by Don Baccus on 03/07/02 03:09 PM
It would be great if you could.

The above confuses me just a tad bit:

Will we have installation instructions (either through our docs or OpenFTS docs) that make it clear how to install OpenFTS under either PG 7.1 or PG 7.2?  Will we have clear pointers to the old and new OpenFTS packages?

I'm sure we can get Roberto/Vinod to add some stuff to our install documents.

Without clear documents that are highlighted somewhere we'll be getting endless questions from users, over and over again ...

Collapse
7: Response to APM and FTS (response to 1)
Posted by Dan Wickstrom on 03/08/02 02:18 AM
Openfts is in what is likely to be one of many stages. The openfts developers are experimenting with various features which should eventually result in a full-featured search engine within postgresql.

The latest incarnation has more natural interface which is easier to understand. In the current system, search queries look something like the following: 

    SELECT
          txt.tid,
      FROM
          txt
      WHERE
          (txt.fts_index @ '{14054652}')

and the new system that I'm working on porting uses a natural language approach that supports boolean operators and it looks like the following: 

select * from foo where titleidx @@ '(the|this)&!we';

This is quite an improvement over the previous approach, and I can't wait to see how openfts evolves.

In any event, Neophytos has stated that he will do the documentation for this new version, and I'm sure that he will include references on how to build it for both 7.1 and 7.2.

Collapse
8: Response to APM and FTS (response to 1)
Posted by Don Baccus on 03/08/02 02:37 AM
Yeah, I've been following the Russian's work and the new-style searches will be really nice, because it makes it easy for any client to implement "advanced" boolean searches.

I just want to make sure that we'll have good docs for installation on both versions of PG and it sounds like you guys have it covered.