Tetra WebBBS Documentation
Tetra-WebBBS 5.30 (2008-2)
(1) locking routines for databases revised
The old way of database locking was cumbersome.
Since this was originally invented almost 10 years back operating systems have improved as well. The new way of locking now relies on the underlying operating system. Esp. on busy forums or shared hosting accounts or while searchengines are indexing the forums the new locking is supposed to drastically decrease serverload.
(2) optimized interaction with archives
The new $archive_URL variable allows for HTTP redirects to the archive in case a message is no longer on the board.
This can be used even on multiple cascading archives to redirect requests further backwards subsequently.
This is a major improvement for all those having their boards indexed on searchengines. Users and searchbots will be redirected to the new location of the message. That annoying *Message no longer on the board bla bla bla* message will no longer bother visitors.
(3) optimized handling of uploaded images
The old way was troublesome and sometimes even unpredictable, for example if archives got involved.
The whole stuff about handling uploaded images has been refactored. Scripts now are autosensing for images related to messages or profiles.
The new $live_subdir variable (to be used for archives only) will allow for autosensing and proper deletion of images even when messages have been archived.
(4) protecting access to profiles and restriction of setting preferences
As you know access to profileeditform is based on the username stored in the cookie.
This new version prevents users from getting usernames stored in their cookie for which somebody else has already registered a profile. To store a registered username in the cookie users will have to provide the appropriate password - on posting form as well as on set preferences form.
Login to editing the profile will only be granted if either the cookie holds the appropriate name/password pair or users fill a little login form.
(5) forum access for searchengines and other bots
The scripts are now reconfiguring on the fly to a default set of configs reasonable for searchengines.
Example: Searchengines won't see the posting form or the corresponding link - so they can't index.
On top the new $SearchFriendlyProfiles allows boardadmins to either allow or disallow indexing the profiles.
(6) decreased load on indexing by searchengines
Unfortunately the old scripts prompted searchengines with a 10-year message index. - Everybody knows what that means on forums more than 50-100 K messages.
This behaviour has changed. Now searchengines see the index the same way users do.
That allows boardadmins to trim their forums and archives to optimized load - according to what their server/account allows for.
(7) usage optimisation for archives
The new $browsable_archive variable allows for easily enable archives for browsing.
Basically setting this variable to *1* or, even better, *2* is a shortcut for a set of configs trimming the board for load and speed optimisation.
Boardadmins are still free to chose whatever $DefaultTime appropriate to *their* archives. This allows for further load optimizing.
Here is the place to say thanks to Bill. He sponsored the teaser feature.
Be aware enabling teasers *does* put some load on your server. Use carefully, consider showing teasers only for *new* messages.
More on this to be found in the docs as soon as possible (hopefully).
(9) SMTP and authentication
It's possible to send notification mails using SMTP servers requiring authentication.
The appropriate data for the mailbox used is located in a separate file *smtp_settings.pl*.
(10) all parsing routines refactored
The scripts are now using standard parsing routines provided by various Perl modules available on every server instead of a home-brewed parsing code. Errors such as the old Opera bug or the relative new *WebKit* bug are supposed to be gone.