Appliance update: texisScripts/applianceman-18.0.2, texis-7.06.14893...

Post Reply
User avatar
Kai
Site Admin
Posts: 1271
Joined: Tue Apr 25, 2000 1:27 pm

Appliance update: texisScripts/applianceman-18.0.2, texis-7.06.14893...

Post by Kai »

Several new appliance updates are now available: texisScripts-18.0.2, applianceman-18.0.2, and texis-7.06.14893...

texisScripts-18.0.2 contains the following fixes:

* Fixed Javascript Memory not allowing 'mb' suffix

* Fixed some canonical URLs not replacing non-canonical URLs if the canonical URL is encountered first

* Improved canonical URL parsing: was truncating at spaces, not HTML decoding

* Avoid erroneous Ignoring "Outside ... (Stay Under)" rejection for Base URL message when Ignore Case and Stay Under are both set

* Ignore Case Y with Required Prefix could cause some upper-case URLs (e.g. acceptable offsite links) to be rejected, regardless of prefix

* Extra Domains with REX-sensitive characters might have caused problems

* Exclusion Prefix with uppercase protocol or host was not rejecting URLs

* Any Exclusion Prefix, Allow Extensions or Exclude Extensions setting value was permanently changed to lowercase if Ignore Case set, even if unset later

* A successful Generate Browse URLs task no longer shows its exit reason as Error: unknown ExitReason [ok]

* Browse URLs by Folder could create separate folders for variant upper/lowercase protocol/host

* List/Edit URLs, List/Edit Enums links to metasearch target profiles now give proper links for localhost target profiles

texis-7.06.14893... contains the following fixes:

* Using a PAC (proxy auto-config) URL and threads (parallel fetch) of 2 or more could potentially cause an ABEND, if the PAC URL times out or otherwise fails

* Fetching a page with <link>, <app>, <applet>, <meta refresh>, Gopher or offsite/rejected frames may have resulted in a memory leak

* Fix for JavaScript ABEND for certain instances of running out of scriptmem

* Certain complex searches, such as a keyword search with a site: query, that require post-processing now check the Allow Post-Processing setting, and thus may issue a `Query ... with current WHERE clauses would require post-processing' error if that setting is N (the default). Enable Allow Post-Processing if the error occurs, to allow the full queries (at a potential performance cost). Previously, some such post-processing queries were performed in full regardless of the setting.

* A fetch of an NTLM (Windows authentication) protected page may have cause a small (~30-byte) memory leak per fetch

* A page with large complex JavaScript was causing an ABEND sometimes; now fails the JavaScript

* A page with transfer or content encodings may erroneously get a Max page size exceeded (truncated) error if the original entity size is just under (by a random amount, typically no more than a few hundred bytes) the actual max page size

* A fetch of a page with chunked transfer encoding and Keep-Alive may timeout, take a long time, or receive connection reset sometimes, if the remote data is returned in large read() blocks; more likely if gzip content encoding also used

* Fix for ABEND in some JavaScript code

* Fixed full permutation thesauruses sometimes missing terms

* A fetch of an NTLM (Windows auth) protected URL could ABEND under certain circumstances, if NTLMv1 enabled in Authentication Schemes

These updates are available via the System -> System Setup -> Update Software menu.
Post Reply