Installation Issues and Known Faults |
|
| This page provides information about installation problems and known faults. If you are a user of TRIST and can not find a resolution here then please email support@trist-online.co.uk to see if we can a way around the problem. Please remember that TRIST is a free to use tool and no guarantees are made about its ability, stability or security. | |
| Missing terms in the index | In the first realeases of TRIST, a fault in exception handling meant that
some inserted keywords returned an exception to indexing tool causing it to
skip the insert of some terms. This stems from an unpredictable error returned
from the Access database when an SQL Insert statement was used. It only occurs
in the creation of the Token table and not any other inserts. The exception
is thrown even though an insert has taken place. To get around this problem:
|
| Slow opening when Indexed documents is bigger | |
This is because the system preloads the token table into memory when loading. The choice to do this was to provide better response to the Token View option. If your index contains a larger volume of files, in testing we reach 2000 documents with 54 000 keywords, then the indexing tool will be noticeably slower at loading. |
|
| Find New Files slow with a large index | |
You should expect the find new files to get slow as the index increases. The tool must compare new file names with a much larger number of documents. If your index is 500 documents, the tool must do up to five hundred comparisons for each new document it finds. The tool uses the most efficent comparison techniques available to it but patience is the key to working with big indices |
|
| Maximum Capacity of TRIST | |
The theoretical limit of TRIST is only the size constraints of the database. The Access platform would allow over 4 billion documents to be referenced. In practice the tool will take approximately 6 hours to index 1000 documents, so time is more a governing factor. If you have to index more then 500 files for web-use then TRIST would become too slow. It would still function but slowly. A comercial solution might be more suitable. If your only concerned with personal documents then the limit is higher. For really large numbers of files, overnight indexing is preferable. The TRIST tool will happily batch process the whole list of pending documents given enough time. Bear in mind that the tool might require a second pass to fill in gaps caused by Microsoft Access exceptions. |
|
| Tool Appears Unresponsive | |
The TRIST indexing tool produces output to the screen only if run from classes. The download is only available as a Jar file. As a result, the output disapears into computer limbo! Future updates aim to provide better communication with the computer. To see if the tool is operating, use System Monitor (pre windows 2000) or Task Monitor to measure CPU use. |
|
| Stopping the Index Mid Way | |
The indexing process can not be stopped manually once instigated. You have to force the program to stop using the CTR+ALT+DEL Options. The indexing process is designed to be a batch process and so user-intervention acts a drain on resources and computing time. Only begin the indexing process if you have the time to let it finish. The time taken will depend on the length of the Pending List. |
|