Short: Patch SpamFryer to ver 6:21 Author: AmigaPhil Uploader: AmigaPhil scarlet be (AmigaPhil) Type: comm/mail Version: 6.21 Requires: SpamFryer, ARexx Architecture: generic *** Re-upload: Minor change (fix) to the patch/installer *** script. (See http://tinyurl.com/nc8veze for more.) This is an UNOFFICIAL(*) patch script that will update an installed version of SpamFryer.rexx to version 6:21. Only the main SpamFryer.rexx script will be updated (and a new version of CheckLists.rexx will be added in the SpamFryer dir.). Documentation and other addon files will be left unchanged (newer version of SpamFryer.yam, SpamFryer.thor, and SpamFryer.smrx will be available separately). Your configuration settings and any custom icons will be preserved. Double-click on the Patch icon and follow the instructions of the installer script to update SpamFryer.rexx. (*) If anything goes wrong, DON'T BLAME Simon N Goodwin, this patch is not supported by the author of SpamFryer. You can send tomatoes to AmigaPhil ?? scarlet ? be :-) Included catalogs: french : Philippe "AmigaPhil" Duchenne - ATO France dutch : Johan Banis - ATO Nederland italian : Simone "Tuxedo" Monsignori german : Dirk Neubauer - ATO Germany Horst Schumann - ATO Germany (proof-reading) Bug fixes and improvements since version 6:18 : - The localization system has been upgraded to RexCat 2.2 - Added ASCII code ranging from 80x to 9Fx to the list of non-printable chars to remove in BROWSE mode (those can mess the listview window). - With the sorting method set to 2, when a rule ranked at the second place in a group (like "Subject:") got more hits than the rule in the first place (in that group), it was moved up by 2 positions (before the last rule in the previous group). Fixed: that rule is now taking the first place in the group as it should. - The character 'a0'x (non-breakable space) is now recognised and handled as a normal space ('20'x). - Removed the '|' (vertical bar) and '$' character in the fuzzy match translation (introduced in version 6:8 and 6:10). It wasn't really helping to catch more spam, but could lead to confusions when creating a 'Early-Ref:' type rule. Eg.: With fuzzyMatch set to 1, a rule like 'Early-Ref: P|()|R|N' would never have matched; unless written: 'Early-Ref: PL()LRLN'. - CheckLists.rexx: When testing a list for similar rules, the leading blank spaces in rules like 'From: " logiciel"' were ignored. Fixed. Added a command line option to save the output to a file. Moved the script from the 'Translation Tool' directory back to the main 'SpamFryer' directory. - Fixed a bug with the internal 'maxMailSize' rule. The 'bad' flag was re-initialised AFTER the test was made. - When sorting is activated (sortMethod set to 1 or 2), the keepList or loseList was not saved back to disk if the order of the rules remainded unchanged. To avoid loosing a 'Trigger Hits' incrementation, a list is now also saved when a rule scores a new hit without moving up in that list. - With the sorting method set to 2, the rules' Trigger Hits were scaled down only if the first rule ("From:" type rule) of the list had reached at least 250 hits. Now the Trigger Hits are scaled down as soon as ANY rule reach 250 hits. - CheckLists.rexx now also check for similar/redundant rules (Note that the script has more intensive tasks to do, so be patient :-) )