# $Id$ PyKota - Print Quotas for CUPS and LPRng (c) 2003, 2004, 2005, 2006 Jerome Alet This program is free software; you can redistribute it and/or modify it under the terms of the GNU General Public License as published by the Free Software Foundation; either version 2 of the License, or (at your option) any later version. This program is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for more details. You should have received a copy of the GNU General Public License along with this program; if not, write to the Free Software Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. ============================================================ PyKota NEWS : - 1.24HotFix2 (2006-05-30) : - Introduced the 'askconfirmation' directive to interact with the end user through any script (pknotify is recommended) and see if he really wants to print when knowing the cost of the job. - Now uses pkipplib if present to directly retrieve the job information from the CUPS server over IPP. - Added a lock to ensure that a single cupspykota backend is processing datas for a particular print queue at any given time. - Now exports more environment variables for the subprocesses. - Fixed some problems with DateTime objects. - Added support for PySNMP v4.x in addition to 3.4.x. - Make the MySQL schema work with old releases of MySQL. - Fixed several character encoding problems. - Ensures that the denied banners counter is properly initialized with relationnal databases. - 1.24 (2006-04-08) : - Improved documentation and packaging stuff. - 1.24beta (2006-03-26) : - Finalized the pknotify command. - Updated the messages catalogs for translations. - 1.24alpha18 (2006-03-15) : - Now uses MIME encapsulation for email messages to avoid some problems with 8 bits messages being rejected. - 1.24alpha17 (2006-03-06) : - Fixed a major problem in the PostgreSQL backend. - pknotify can now check a remote user's name and password. - 1.24alpha16 (2006-03-02) : - Introduced the pknotify command line tool to dialog with remote PyKotIcon applications. - 1.24alpha15 (2006-03-02) : - optimized pkbcodes, pkprinters and pkusers using edpykota as the model. - 'edpykota --list' and 'pkusers --list' are now authorized to mere mortal users, but the list is restricted to informations about the current user. - Extended the functionnality of the 'onbackenderror' directive to allow for configurable retries. - 1.24alpha14 : - Finished optimizations in edpykota and fixed some minor problems with LDAP. - 1.24alpha13 : - Introduced the pkusers command line tool to manage users and users groups. This command doesn't deal with print quota entries, which continue to be managed through the edpykota command line tool. - Major rewrite and incompatible changes to command line options in edpykota. This command now only deals with print quota entries. Printers, users (or users groups) now have to be created first using pkprinters or pkusers. - 1.24alpha12 : - Introduced the 'duplicatesdelay' directive to specify after which duration two duplicate jobs are not considered duplicate jobs anymore. - Introduced the 'balancezero' directive to specify any value as the balance limit instead of 0.0. This can allow admins to give a few free credits to users, for example. - 1.24alpha11 : - Huge speed improvements when modifying billing codes and printers. - Both pkbcodes and pkprinters now display the percent done for all actions. - The pkinvoice invoice generator command line utility is now included. - 1.24alpha10 : - Huge speed improvements when deleting users, printers, and billing codes. - 1.24alpha9 : - Preliminary support for MySQL, thanks to Matt Hyclak. - 1.24alpha8 : - The 'preaccounter' configuration directive was introduced to control which parser to use for precomputation of the job's size. When unset, the internal parser is used just like it was previously. See conf/pykota.conf.sample for more details. - Now subprocesses launched through the overwrite_jobticket directive can also output 'CANCEL', to represent the user's own choice to cancel the print job. - 1.24alpha7 : - pkbanner now handles new print limiting modes correctly. - The cupspykota backend wrapper now accepts to be interrupted with a kill -INT. This causes an immediate and clean abort of the current print job. - Matt Hyclak contributed the possibility to store the printing history in a MySQL database. See the contributed/ directory for details. - Added 2006 to the copyright's years. - 1.24alpha6 : - Fixed internal SNMP and PJL handling code for very fast printers : PyKota didn't see them going from 'idle' to 'printing' to 'idle' again. - Now waitprinter.sh accepts SNMP version number and community name as arguments, thanks to Brandon Holbrook. - Now command line errors and Ctrl+C repectively cause PyKota commands to exit with status -2 and -3. - 1.24alpha5 : - All command line tools now accept -A | --arguments filename.args as a way to bypass the shell's maximal command line size : just put each argument on a line by itself in a file, and pass the file's name as the -A | --arguments command line option's argument. For now this won't be documented in commands' helps. - dumpykota can now dump the complete database in XML format. - 1.24alpha4 : - Added support for the SQLite3 backend. See pykota/initscripts/sqlite/README.sqlite - 1.24alpha3 : - Fix for mailandpopup.sh : smbstatus can't be run setuid, so we regain priviledges before running an external mailto directive. - 1.24alpha2 : - Introduce the "onbackenderror" directive in pykota.conf, to control if the user must be charged or not whenever a real CUPS backend fails. The default value is 'nocharge'. Set it to 'charge' if you think your users can cause a CUPS backend to fail. - Now mark original backends related problems as such in the printing history. Modified printquota.cgi and pykota.css to make them be more visible. - 1.24alpha1 : - Fixed a problem which caused backend errors to be ignored. - The sample LDIF file now contains the two necessary user entries to have an LDAP PyKota Admin and an LDAP PyKota ReadOnly user. The sample configuration files were updated to contain the default passwords used in the LDIF file. - dumpykota is now more strict when checking arguments. - 1.23 : - Many minor bug fixes and code cleaning. - Improved documentation. - Added Polish translation. - Improved French and Swedish translations. - 1.23beta : - Now the mailto directive is honored in all cases where the job is denied or a warning has to be issued, and the correct reason can be sent back to the user. - The PYKOTAREASON environment variable, if defined, contains the textual reason why the job was denied or a warning has to be issued. - 1.23alpha31 : - Added support for --maxjobsize, --passthrough and --nopassthrough in pkprinters. These modes are now also honored when printing. - Allows the admin to restrict the number of print queues which will be treated by pkturnkey by passing queues' names as additionnal arguments on pkturnkey's command line. - pkturnkey now supersedes pkhint with the help of the -c | --doconf command line option. - 1.23alpha30 : - Database changes : you MUST use upgrade-to-1.23.sql again if you use PostgreSQL, or copy the new schema and restart your LDAP server if you use LDAP. - The -C | --comment command line switch was added to edpykota to attach textual information to modifications done to a user's account balance. - The -l | --limitby command line switch to edpykota was modified to allow 'noquota', 'nochange', and 'noprint'. See the edpykota command's help or manual page for details. - The -i | --ingroups command line switch to repykota allows PyKota Administrators to only lists members of the specified group or groups (separated with commas). - The -I | --increase command line switch to edpykota allows to increase or decrease page limits. - Added the precomputed job's size and price to the history for controlling purpose. - 1.23alpha29 : - Fixed a small problem with billing codes. - Exports some environment variables earlier, to render the script launched by the overwrite_jobticket directive more powerful. - Now autopykota also creates print quota entries for the user on the printers groups the current printer is a member of. - 1.23alpha28 : - Added users groups management to pkturnkey. - Fixed umask in setup script. - 1.23alpha27 : - Replaced the --dryrun command line option by the --force command line option in pkturnkey, with exactly the opposite effect (for security reasons). - 1.23alpha26 : - Allows parametrized pjl and snmp hardware accounting methods, to be able to respectively set the TCP port and the SNMP community. - 1.23alpha25 : - Added a --dryrun command line switch to pkturnkey. - Improved pkturnkey's general usability. - Added a manpage for pkturnkey. - 1.23alpha24 : - Allows different databases to be used depending on the user's rights. This allows PyKota admins to directly use a master LDAP server for read+write, while letting normal users using the replicas for readonly access. NB : using different storage backends is also possible, but you're on your own there. - 1.23alpha23 : - Fixed a problem with banner accounting. - Introduction of the new 'pkturnkey' command line tool. - 1.23alpha22 : - The cupspykota backend was almost entirely rewritten from scratch. It's now much more readable, maintainable, etc... - LPRng support is OFFICIALLY dropped for now. LPRng users are advised to download earlier releases or purchase the PyKota v1.22HotFix1 Official release. LPRng support MIGHT be re-added later. - 1.23alpha21 : - Fix for IPP requests which contain more than a single job-originating-host-name attribute. - 1.23alpha20 : - Added the overwrite_jobticket directive. See pykota.conf.sample for how to use it. - 1.23alpha19 : - Fixed an LDAP filtering problem when managing billing codes. - Added the unknown_billingcode directive. See pykota.conf.sample for how to use it. - 1.23alpha18 : - Added the dump_bfd.php contribution. - Several fixes in the internal PJL accounter. - Now autopykota doesn't create unneeded print quota entries anymore. - Increased the timeout in the PJL hardware accounting code to accomodate some printers (HP5000 at least). - 1.23alpha17 : - pkpgcounter is now distributed separately only. This means that people must now download and install pkpgcounter v1.56 or higher prior to being able to run PyKota. - Removed all the autoconf/automake stuff which was outdated and confusing. - 1.23alpha16 : - Extracted the GPL blurb from all commands' help, and use a static version only once. This saves space big time, and eases future translations. - 1.23alpha15 : - pkbcodes now works with the PostgreSQL backend. NB : PLEASE RUN upgrade-to-1.23.sql AGAIN !!! - 1.23alpha14 : - Logs a warning when the user/group doesn't exist when running edpykota. - 1.23alpha13 : - Minor fix for user groups because of latest database changes. - 1.23alpha12 : - Big database changes, use the migration scripts ! - 1.23alpha11 : - Small fix to the PCL3/4/5 parser for some drivers. - 1.23alpha10 : - Fixed a bad exception name in cupspykota when parsing IPP messages. - Improved the IPP messages module : it can now generate IPP messages in addition to parsing them. - 1.23alpha9 : - The denyduplicates directive now accepts a command line as an argument beside yes/no. The command is launched and must print either ALLOW or DENY to its standard output. Then the job gets allowed or denied by PyKota. This will allow user interaction in the near future. - Introduced the new 'striptitle' directive to remove ugly prefixes from print job's titles whenever a new print job is submitted. - Fixed internal SNMP accounter for printers which only have a life time page counter, and not a volatile page counter. - 1.23alpha8 : - The data dumper can now summarize payments and history with the help of the --sum command line option (or checkbox in the CGI). - The message used when the job's filename, title and options are hidden because of the privacy directive is now translated at retrieval time, instead of at saving time. - Minor fix in the LDAP backend code. - 1.23alpha7 : - Improvements to the internal PJL and SNMP accounter : now should support more printers and waste less time. - 1.23alpha6 : - The data dumper's start= and end= date filtering options now accept a positive or negative offset to be added to the special keywords 'yesterday', 'today', 'now', 'tomorrow'. e.g. : dumpykota --data history start=yesterday-5 end=today - 1.23alpha5 : - The data dumper's start= and end= date filtering options now accept the special keywords 'yesterday', 'today', 'now' and 'tomorrow'. - 1.23alpha4 : - The data dumper now accept start= and end= to specify the starting and ending dates to dump the history. - 1.23alpha3 : - The destination link for the logo used in CGI scripts is configurable. - Data dumper improved : it's faster and now datas are correctly escaped. - The URL to the logo displayed in the CGI scripts is configurable. - 1.23alpha2 : - Fixed edpykota --prototype with balance. - Major rewrite of the IPP message parsing code to definitely fix the job-originating-hostname IPP attribute problem. - 1.23alpha1 : - Sets current username to the user CUPS runs as instead of "root" when printing test pages from CUPS' web interface (because the username is empty in this case). - Fix for strange values in job-originating-hostname - 1.22 : - Big improvements to the documentation. - 1.22beta : - Improved stability in pykotme and pykotme.cgi. - Fixed a permission problem in pykotme and pykosd. - Fixed an inconsitency between SQL and LDAP backends' behavior in edpykota. - Improved stability and allowed tracebacks in CGI scripts. - Improved pykotme.cgi - 1.22alpha8 : - Fixed history retrieval for nonexisting printers or users. - Clicking on the billing code in printquota.cgi's job history now shows all jobs matching this billing code. - Fixed the dump in CUPS' page_log format which didn't dump the oldest entry. - The data dumper now accepts billingcode= and hostname= as filters when dumping the job history. - The job billing code if present is now saved into the database, and exported to subprocesses as the PYKOTAJOBBILLING environment variable. - Ctrl+C now makes PyKota commands exit without a traceback. - Fixed an i18n problem with Python 2.1 - 1.22alpha7 : - Fixed some problems with the retrieval of the history from LDAP. - 1.22alpha6 : - Early errors now produce meaningful error messages (finally !) - 1.22alpha5 : - SNMP and PJL improvements with regard to PowerSave / StandBy. - 1.22alpha4 : - Fixed charset problem under MacOSX. - Allows pkbanner to accept more info to display on its command line. - 1.22alpha3 : - Fixes some case related problems with some LDAP backends. - Fixes some problem for printers which directly go from "idle" to "warmup" and never to "printing" before returning to "idle". - Some permissions settings may have prevented cupspykota from parsing cupsd.conf, this is now fixed. - 1.22alpha2 : - Introduction of the web enabled print quote generator. - Fixed a potential race condition when several backend wrappers were present in CUPS' backend directory. - Fixed dumpykota's output in CUPS' page_log format to better match CUPS format. - 1.22alpha1 : - Import into Subversion. - 1.21 : - Improved documentation. - Improved dependencies checker. - Updated some translation files. - 1.21beta2 : - Fixed an LPRng specific bug. - Several improvements on stability issues with manually modified LDAP trees. - Documentation improvements. - 1.21beta : - Prepared translation files for next release. - 1.21alpha27 : - Allows software() (no arguments) to the software accounting directive. This saves a lot of CPU since the job is not parsed again : the precomputed size is used directly in this case. - epykota's --prototype option now works fine with accounts limited by balance. - 1.21alpha26 : - Fixed the handling of the 'maxdenybanners' directive. - Introduced the 'denyduplicates' directive. - Fixed some database related glitches. - 1.21alpha25 : - The job's size is now always precomputed, even when 'enforcement' is not set to 'strict'. - Introduction of the new 'trustjobsize' configuration directive to workaround some problematic printers. See the sample configuration file for details. - 1.21alpha24 : - Big database structure changes. An upgrade script is now included for PostgreSQL. LDAP users may just want to copy the new schema in place. - edpykota now recognizes the -o | --overcharge command line option to set an overcharging or undercharging factor on each print job's cost for a particular user. - repykota's output format has changed an is not limited to 80 characters wide anymore. People who used to parse repykota's output in their own tools are strongly advised to parse dumpykota's output instead. - Introduction of the new 'maxdenybanners' configuration directive to allow up to N (the directive's value) deny banners to be printed even if the print account is forbidden to print. - Better detection of the number of copies in the PostScript parser. - 1.21alpha23 : - Traditional Chinese translation added. - Several fixes and performance improvements with the LDAP backend. - 1.21alpha22 : - edpykota's --delete command line option doesn't consider '*' to be the default argument anymore :-) - 1.21alpha21 : - dumpykota.cgi can now be run as a normal user as opposed to a PyKota administrator. No read access to pykotadmin.conf is necessary anymore. This CGI scripts now honors the REMOTE_USER environment variable just like printquota.cgi does, see the file cgi-bin/README for details. The command line tool dumpykota must still be run by a PyKota administrator. - Fix exception's name error due to a change in recent versions of PygreSQL. - 1.21alpha20 : - Big bug fix wrt the datelimit field which was never reset. - 1.21alpha19 : - Introduction of the new dumpykota.cgi CGI script, which allows third party tools like phpPrintAnalyzer to be interfaced in real time with PyKota. To use it you may have to change your web server's configuration, please see pykota/cgi-bin/README - 1.21alpha18 : - Now dumpykota can dump the job history into CUPS' page_log format (even when the printing system is LPRng). This allows people to use tools like PrintAnalyzer or phpPrintAnalyzer (two completely different beasts) with PyKota. - 1.21alpha17 : - Fixed field type's name in XML dumps. - Now drops priviledges to user 'pykota' when launched as root (directly or as a CUPS backend), if such an user exists. - Fixed PCL3/4/5 parser for some drivers' output. - 1.21alpha16 : - fixed several recently introduced bugs in LDAP backend code, thanks to Matt Hyclak. - 1.21alpha15 : - Normalized field names between LDAP and PostgreSQL backends in dumpykota's output. - dumpykota filtering capabilities are now supported within the LDAP backend too. - 1.21alpha14 : - dumpykota now supports extended filtering possibilities with the PostgreSQL backend (LDAP support will come soon). - 1.21alpha13 : - Added a --debug command line option to pkpgcounter, so that people may see what really happens, and what will be available in the future :-) - 1.21alpha12 : - Fixed a bug with non standard pkbanner command lines. - Now logs as errors the differences between precomputed and computed job sizes. **CUPS only for now** - 1.21alpha11 : - Now user and group print quota entries can be stored directly below the user or group with the LDAP backend, using them as branches. Thanks to Stefan Wold for the patch. See the sample configuration file for details. - TLS is now supported with the LDAP backend. Thanks to Stefan Wold for the patch. See the sample configuration file for details. - edpkota now accepts the -U | --used value command line argument to preset the page counters to an initial value. Thanks to Wilson Roberto Afonso and Matt Hyclak for the patch. - 1.21alpha10 : - Now automatically reconnects up to three times if the LDAP server returned an error, possibly due to a timeout. - 1.21alpha9 : - Now searches the configuration files in 'pykota' system user's home directory. The 'pykota' system user and group have to be created by the system administrator, and given sufficient permissions to PyKota's configuration files. - 1.21alpha8 : - Added the new reject_unknown directive to reject the creation of user or groups which are unknown to the system. - 1.21alpha7 : - Introduction of the pkmail email gateway. - 1.21alpha6 : - Fix for LPRng filenames. - Added Turkish translation thanks to Baris Cicek. - 1.21alpha5 : - Improved ESC/P2 miniparser for Stylus Photo printers. - Implemented the --savetoner command line option to pkbanner. - Improved pkbanner's help and messages. - 1.21alpha4 : - Fixed job's data file detection code for LPRng. - Fixed a problem with PJL and PowerSave mode, thanks to Wilson Roberto Afonso. - 1.21alpha3 : - PyKota can now generate its own banners, and either account for them or not, depending on newly introduced directives in pykota.conf Thanks to Matt Hyclak for the patch. - 1.21alpha2 : - Installation script is now non interactive again, as it should have stayed from the beginning. Dependencies checking can be done by launching the checkdeps.py command. - 1.21alpha1 : - added an IPP messages miniparser. Extraction of the client hostname is now considered to be fiable, at least until the CUPS developpers change something... - 1.20 : - Version number changed. - 1.20beta : - Norwegian (Bokmål) preliminary translation added. - German translation was finalized. - Greek translation was finalized. - For urgent legal reasons, a "privacy" directive has been added to pykota.conf, to hide jobs' title, filename and options in the history. - Minor changes in the debian directory. - 1.20alpha27 : - Fixed many problems with translation files. In particular the Brazilian Portuguese translation file was incorrectly installed into the french "breton" locale directory. - Finished French translation. - 1.20alpha26 : - LPRng support code should now better detect network printers' hostnames or IP addresses. - Fix for recently introduced bug in LPRng support. - Fix for incorrect locale setting when starting. - 1.20alpha25 : - pkprinters allows --list to normal users, but require user to be a PyKota administrator when another command line option is used. - pkhint doesn't use absolute path to search for helper commands anymore. - Spanish translation ready for 1.20. - 1.20alpha24 : - Now warnpykota only warns users limited by balance once. - Now warnpykota only warns users who have ever printed, to not cause confusion to users who have just opened their account. - Prepared help messages for translations. - 1.20alpha23 : - Add workaround for locale problem. - 1.20alpha22 : - Hopefully final fix for data encoding charset. - edpykota now refuse to work if the user is not a PyKota admin. - Fixes a recently introduced bug with users groups and PostgreSQL. - 1.20alpha21 : - Now dumpykota refuses to launch if the current user is not a PyKota administrator (a PyKota administrator is an user allowed to read the pykotadmin.conf file). - Now any PyKota administrator can use repykota and warnpykota with additionnal arguments, not only the root user. - Improved documentation a bit. - 1.20alpha20 : - Reduced delay in internal SNMP and PJL_over_TCP handlers from 2 to 1 second. - dumpykota can now output XML. - 1.20alpha19 : - Improved user groups and user groups quotas dumps in dumpykota. - Finished implementation of LDAP support in dumpykota. - The md5 sum of the job's datas is now available as an hexadecimal digest in the PYKOTAMD5SUM environment variable. - The description, filename, title, and options fields are now stored as UTF-8 strings in both LDAP and PostgreSQL. Proper charset is used during extraction from the database. - 1.20alpha18 : - Experimental autopykota command to initialize user account and quota entries on first print while not resetting the balance value if not needed. - Added the new "winbind_separator" directive to strip out the Samba/Winbind domain name while printing if needed (Off by default). - 1.20alpha17 : - Added --hardreset command line option to edpykota. - Fixed problem with accented chars in title, filename or options and LDAP backend. - 1.20alpha16 : - Internal handling of socket based PJL queries over port tcp/9100 is now available with the magic 'pjl' value in the hardware accounter. - 1.20alpha15 : - LDAP pykotaAccountBalance objects were always created attached to the pykotaUser objects. This is now fixed and the value of balancebase in pykota.conf is now correctly used. - 1.20alpha14 : - Belgian translations added (copy of the french ones). - External hardware accounter now loops until a result can be read correctly. - pkhint now suggest the internal snmp handling instead of the external one. - 1.20alpha13 : - Hardware accounter now accepts the special 'snmp' value, meaning ALL is done from main code without launching any subprocess. You need the http://pysnmp.sf.net Python module for this to work (tested with v3.4.2) - Increased delay between two loops iterations to 2 seconds in waitprinter.sh to diminish CPU load. - 1.20alpha12 : - waitprinter.sh rewritten from scratch to wait for printer status stabilization. - 1.20alpha11 : - Generic data dumper 'dumpykota' introduced. Works only with PostgreSQL backend for now. - 1.20alpha10 : - Added fix for incorrect job's size computation when hardware accounting fails. - 1.20alpha9 : - Small fixes wrt subprocesses. - 1.20alpha8 : - Added full support of PCL3 : HP Deskjet printers are now supported. - 1.20alpha7 : - First draft of PCL3GUI (Deskjet 640C for example) analyzer. Needs more work, but no documentation is available. - 1.20alpha6 : - Improved mailandpopup.sh script. - Small fixes for LPRng. - Now exports PYKOTAPRINTERHOSTNAME for network printers' hostname or IP address. Set to 'localhost' when not meaningful. - Fixed issue with how mod_auth_ldap Apache modules fills the REMOTE_USER environment variable. - Fixed a bug in LDAP user deletion code. - 1.20alpha5 : - Introduces the new configuration directive 'onaccountererror' to specify what must be done whenever the software or hardware accounter fails. Allowed values are 'continue' and 'stop' (default). - Minor fix in software accounter for a recently introduced problem. - 1.20alpha4 : - Added support for more ghostscript devices in ESC/P2 analyzer : stcolor, st800, escp, escpc are now supported. - 1.20alpha3 : - Small improvement in PostScript parser to avoid being fooled by a change in number of copies : The user can't ask for 200 copies of page 1 and 1 copy of page 2 to fool PyKota anymore : this would have been accounted for as 1 copy only, but now the maximum value seen is used. - Relax checks for PCL5 header to accomodate some strange printer drivers. - Now hardware accounting method works like software accounting method wrt the subprocess' output. - Hardware and Software accounting stop the print queue when the accounting subprocess didn't work. This will be configurable in the near future. - 1.20alpha2 : - Now software accounting method reads accounter's answer until a line with a single integer is obtained or until all of its output has been read. This should fix the problem some people have when strange locale related messages are sent when running subprocesses (Solaris only ?). - First version of ESC/P2 analyzer. Seems to work ok at least with 'epson', 'epsonc', 'eps9mid' and 'esp9high' ghostscript devices. Needs more work to correctly analyze other devices. - 1.20alpha1 : - Fixed problem with number of copies in PCLXL parser. - Fixed problem with number of copies in PCL5 parser, when number of copies was set outside of and before first page. - Added workaround for number of copies directly set in PostScript code. - Added german translation. - Fixed problem on first print for user group quota when "enforcement: strict" is used. - Fixed interpolation problem in french translation. - 1.19 : - Updated existing translations. - Added final Thai and Greek translations. - Minor last minute bug fixes. - 1.19beta : - LPRng support for both software and hardware accounting seems to be OK. - Translation files fixed because of fuzzy matching problem. - Fixed path to Swedish translation: 'sv' instead of 'se'. - Added first draft of Thai translation. - Fixed some minor bugs with Pychecker. - 1.19alpha35 : - Generic PDL analyzer now supports Binary PostScript documents in addition to DSC compliant ones. - 1.19alpha34 : - LPRng support : software accounting seems to work fine, but hardware accounting doesn't work yet. - 1.19alpha33 : - pykosd now accepts command line arguments. Enjoy ! NB : due to a bug in xosd, you need xosd version 2.2.8 or higher if you don't use the default C locale. - 1.19alpha32 : - Early (not working) version of the LPRng support rewrite. - Saner default configuration file. - 1.19alpha31 : - Fixed a double Python2.1 incompatibility with newer versions, which prevented the PCL5 and the PCLXL parsers to work with this older version of Python. - 1.19alpha30 : - Introduction of preliminary version of the pykosd command - Usernames, Groupnames and Printernames can now contain ANY character, expected '/' and '@'. - Integrated most of the Debian work from Sergio González González - 1.19alpha29 : - PCLXL parser should now correctly handle number of copies set for each page. - Added testsuite for generic PDL analyzer. - Typo fixed in PCLXL parser. - Correctly dispatches error messages depending on the situation : tools which can be run interactively will ALWAYS output their logs to stderr, excepted for their DEBUG messages, for which the "logger:" directive as set in pykota.conf will be honored. Tools which can't be run interactively will ALWAYS honor the "logger:" directive as set in pykota.conf - 1.19alpha28 : - The description field for printers can now be set, and is displayed in reports. - Fixed a small problem in SQL backend code. - Improved pkhint - First working PCL6 (aka PCLXL) parser ! Doesn't handle copies yet. - 1.19alpha27 : - Additionnal 10% speed increase in PCL5 parser for GhostScript generated files (untested yet with other PCL files). - Reintroduced the use of mmap in PCL5 parser : this speedups PCL5 parsing by almost 4 times. This 4x with the 3x below give us in fact a twelvefold increase in performance for PCL5 parsing ! - The Python accelerator Psyco is now used if available in the generic PDL parser : it speedups PCL5 parsing by almost three times automagically. - 1.19alpha26 : - Fixed PCL5 parser according to the sources of rastertohp. - Fixed number of copies handling in PCL5 parser : the number of copies could vary from page to page. - Added more debug information to PCLXL parser. - 1.19alpha25 : - Added new "description" field to printers in both PostgreSQL and LDAP. No code to handle this yet. - Reversed the fix for power save mode in printers : no paper or tray open would have caused people to be undercharged. - Preliminary version of MySQL initialization script. - 1.19alpha24 : - Native PDF parsing code now works perfectly with Python 2.1 and above, and is way faster than before. GhostScript is not used anymore to convert PDF to PS, because PDF is parsed directly. - 1.19alpha23 : - Fixed waitprinter.sh wrt PowerSave mode and some printers. Thanks to Ryan Suarez. - Smart PDL analyzer now recognizes PDF too : two PDF parsing methods are used : a native one which is very fast but requires Python 2.3 or above, and a slower one which uses Ghostscript. PyKota detects automatically which method to use depending on your version of Python. - All tracebacks now include PyKota's version number. - 1.19alpha22 : - Ensures that the file pointer for job's data is reset to the start of the job's datas after the job's size pre-computation phase. - Improved PostScript detection code for bad drivers which don't correctly switch languages between PJL and PS. - Even better debugging code. - 1.19alpha21 : - Additionnal check. - 1.19alpha20 : - More complete logging messages wrt accounters. - Now catches early failures occuring when the configuration files are parsed. - 1.19alpha19 : - Fixed over-verbose exits when displaying help or version. - Extended some PATH in the shell scripts - Deleting an user now also deletes his payments history. - Doesn't explicitely ignore SIGCHLD anymore. This seems to cause problems. - 1.19alpha18 : - History of payments is now stored in the database, and updated each time an user's balance is modified with edpykota --balance. This history represents each time the user was given some positive or negative credit, but doesn't reflect price paid for jobs, which appear in the jobs history instead. - Jobs history web reports now include the job's size in bytes. Also columns were reordered to render the report more easily readable. - 1.19alpha17 : - Mysterious bug (program exited unexpectedly without leaving any log message) which occured once in a while when cancelling jobs should be fixed now. - New 'crashrecipient' directive. Disabled by default. See sample configuration file to see how to activate it. It is suggested that you activate it. - New PYKOTAJOBSIZEBYTES environment variable exported, which contains the job's size in bytes. Empty jobs are now denied. This value is now stored in the job history (but not yet displayed from the web interface) - Preliminary work on payments storage : the PostgreSQL and LDAP schemas should be OK now, but no code is written yet to support this functionnality. - 1.19alpha16 : - the newuser and newgroup LDAP specific directives now accept a second parameter to attach(), "fail" or "warn". If no entry is found to attach PyKota attributes to, and "fail" was used, then action is aborted and program exits in error. If "warn" is used, a new entry is created but a warning message is logged. - The job's originating hostname is now stored in the database, and exported as PYKOTAJOBORIGINATINGHOSTNAME in the environment. - 1.19alpha15 : - The old pykota filter is now removed. This means that currently PyKota doesn't support LPRng anymore, but LPRng support will be re-added, improved, in the near future. - 1.19alpha14 : - PYKOTAPRECOMPUTEDJOBSIZE and PYKOTAPRECOMPUTEDJOBPRICE are now exported as environment variables. - Heavy CPU usage is now completely fixed it seems, by introducing some 0.01 second sleeping when nothing to do. - Pre-computation of the job's size is now only done when enforcement is "STRICT", this saves CPU for people who don't want to use this new feature. - New 'enforcement' directive which accepts either STRICT or LAXIST. See sample configuration file for details. - 1.19alpha13 : - Get back to an old version of the polling loop which worked fine, but non-optimally. Will be optimized as time permits. - 1.19alpha12 : - Job's size is now precomputed unconditionnally using the generic PDL analyzer. NB : This value is not yet used to pre-allow/deny the job. - 1.19alpha11 : - The pykota filter is now completely deprecated and can't be used anymore. LPRng users MUST download an older version of PyKota. CUPS users MUST use cupspykota. Better support for LPRng will come in the near future. - 1.19alpha10 : - pykotme now uses generic PDL Analyzer directly instead of spawning a new process. - Improvements to generic PDL Analyzer. - 25% Speedup in PCL parser. Getting more of the Python version will probably be difficult, since the completely table driven parser I wrote is slower than the original algorithm. Then, the original algorithm is still used for now. Probable that more speed could be achieved using mmap, at the expense of severe memory consumption. - configuration simplifications : the "accounter" directive's 'hardware' arguments now accepts what was in the "requester" directive as it's content. The "requester" directive is not supported anymore. See sample configuration file for details. - 1.19alpha9 : - pkpgcounter is now just a wrapper around the PDLAnalyzer class - pykotaOptions LDAP attribute now accepts non-ascii characters too. - First try at cupspykota main loop rewrite - 1.19alpha8 : - 'querying' accounting method is now called 'hardware' - 'external' accounting method is now called 'software' - 'stupid' accounting method doesn't exist anymore - 'snmp' requester doesn't exist anymore - code simplifications in external requester - 1.19alpha7 : - Database schema change (both LDAP and PostgreSQL) to allow storage of the job-originating-hostname attribute. An SQL upgrade script is present. The code is not yet capable of setting/retrieving this attribute. Support for this functionnality will be added in the future. - 1.19alpha6 : - problems connecting to SMTP servers are now catched. - pykotme now uses pkpgcounter to compute the size of the job. This modification adds PCL5 and soon PCLXL (PCL6) support. You can now get quotes for PCL in addition to PostScript. - pkpgcounter now includes first try at PCLXL parser. It is currently disabled until it's finished. - 1.19alpha5 : - cupspykota polling loop fix for MacOSX (at least) - logging subsystem now logs PID too, to ease diagnosing problems under heavy load. - pkpgcounter enhancements. - pkprinters --skipexisting now tells pkprinters to not modify existing printers when --add is used. pkprinters logs a warning message whenever --add is used and a printer already exists. - LDAP schema now allows accented chars in pykotaFileName and pykotaTitle attributes. - 1.19alpha4 : - a bad copy&paste prevented pkpgcounter to work at all, this is now fixed. - improved PCL support in pkpgcounter, with the use of HP PCL/PJL Reference Set PCL5 Printer Language Technical Quick Reference Guide - pkpgcounter now uses mmap to speed things up : duration is halved with the sample 87 pages PCL document I use for testing. - 1.19alpha3 : - pkprinters now accepts the --list command line argument to list informations about printers. - 1.19alpha2 : - pkpgcounter is now included as a smart external accounter which can handle both PostScript and PCL. - 1.19alpha1 : - PYKOTALIMITBY environment variable is now exported too. - Inclusion of Michele Baldessari's work on autotools. - 1.18 : - Improved documentation - 1.18beta2 : - Fix for raw jobs with cupspykota - 1.18beta : - Small fix for new waitprinter.sh script - Better tracebacks on error conditions - Improved translations - Improved documentation - 1.18alpha15 : - waitprinter.sh replaced by a new and possibly faster and more powerful one. - Names can now begin with a digit. - Stability improvement for quotas set on printers groups. - 1.18alpha14 : - Catches more (all ?) problems with SMTP servers instead of breaking. - Italian translation added. - 1.18alpha13 : - Pre and Post hooks are now available in the pykota filter as well. - The pykota filter doesn't check last user's quota anymore when delayed hardware accounting is used : this will be done anyway the next time the last user will print again. This saves some database queries which are unneeded. - 1.18alpha12 : - Pre and Post hooks to external commands with many many environment variables available are useable from the cupspykota backend. See sample configuration file for details and help. - 1.18alpha11 : - Hopefully final fix wrt groups (users and printers) - 1.18alpha10 : - Fixed nested printer groups accounting. - Fixed user groups + printer groups problem. - 1.18alpha9 : - Fix for LDAP problem when job price was 0. - Fix for empty user groups with LDAP. - Preliminary version of the snmpprinterstatus command added. - 1.18alpha8 : - Don't retrieve data when not needed. This avoids many database queries. - 1.18alpha7 : - New experimental ldapcache directive to improve LDAP caching. - 1.18alpha6 : - New configuration directive added : maildomain to specify the domain to send mail to when user email address is not set in PyKota database. - More complete email message headers. - 1.18alpha5 : - Debian packaging specific files, contributed by Peter Hawkins are now included in the project's tree. - Preliminary work on future low-level LDAP caching to speed up modifications (at least). - 1.18alpha4 : - Applied some of the modifications suggested by Peter Hawkins, for better Debian integration. Most notably cupspykota, pykota, and the shell scripts are now installed into /usr/share/pykota instead of into /usr/bin BEWARE : double check that your symbolic link to cupspykota is correct, and that your pykota.conf file use correct path for the shell scripts like waitprinter.sh and al. - pkhint command added, to help in configuring correct accounting method for PyKota. - 1.18alpha3 : - Fixed a rare bug in cupspykota backend which gave incorrect "CUPS backend died abnormally" messages. - 1.18alpha2 : - Added the pkprinters command line tool to manage printers exclusively. - 1.18alpha1 : - Preliminary work on Relationnal Database Independance. - Sample report made from OpenOffice.org added as PDF and SXC. - 1.17 : - Finalized release. - 1.17alpha2 : - Fix stupid software accounting bug in the CUPS backend. - 1.17alpha1 : - User/Group deletion code in edpykota rewritten. - 1.16 : - Preliminary documentation on how to use OpenOffice.org to create personnalized reports for PyKota + PostgreSQL. - Spanish installation guide for CUPS+PyKota+PostgreSQL added to the documentation. Doesn't yet document v1.16 though. - Portuguese Portuguese translation replaces Brazilian Portuguese Translation. Brazilian portuguese translation is still available under the 'br' locale. - 1.16beta3 : - Swedish translation added. - Small fix in documentation. - 1.16beta2 : - Small fix for job cancelling code while job already printing in CUPS backend. - Big fix wrt printing policies for unknown users. - 1.16beta1 : - Manpages upgraded. - 1.16alpha26 : - Job price is now computed like this : SUM((NbPages * PricePerPage) + PricePerJob) For current printer and all the printers groups it is a member of, if any. This may be difficult to grasp, but offers unprecedented flexibility. - 1.16alpha25 : - Email address can be set at user creation time, using the following syntax : $ edpykota --add username/email@yourdomain.com - 1.16alpha24 : - CGI script now can show the printing history. - Bug wrt number of copies should be fixed. - Major code refactoring => It is now possible to automatically add printers on first print. - 1.16alpha23 : - Fixed potential accuracy problem if a user printed on several printers at the very same time. - 1.16alpha22 : - edpykota now allows adding printers to printer groups - 1.16alpha21 : - Allows the dot in user and printer names. - Printer groups are now cached too. - 1.16alpha20 : - Fixed several small bugs with the help of PyChecker. - 1.16alpha19 : - Fixed a bug in PostgreSQL backend's code. - 1.16alpha18 : - Initial code for printer groups support. Printer groups have to be added with external tools like psql or gq for example. Printer groups seems to work fine with users. Untested with user groups for now. - Several minor visual problems fixed in reports. - Fixed a bug in LDAP date handling. - Fix for LDAP problem with some combination of OpenLDAP + Python-LDAP + some particular settings. - mailandpopup.sh shell script added. See sample configuration file for details. - Several environment variables are now exported when executing external commands (accounters, requesters, mailers) : PYKOTAUSERNAME => user who is printing PYKOTAPRINTERNAME => printer on which the job is being sent PYKOTATITLE => Job's title PYKOTAFILENAME => Job's filename PYKOTACOPIES => number of copies asked PYKOTAOPTIONS => print command options (-o with CUPS) - 1.16alpha17 : - Added portugues translation. - Major code refactoring in accounters. - 1.16alpha16 : - Code refactoring, and addition of a real HTML reporter, to have nice looking and useful web based reports (in the future). - Added utolower configuration option to convert all usernames to lowercase during printing. - 1.15alpha15 : - Now includes the papwaitprinter.sh shell script to wait for AppleTalk enabled printers. - 1.16alpha14 : - Now printing on a printer not defined in the Quota Storage, results in the job being cancelled or stopped, depending on the printing system. - 1.16alpha13 : - Small bugfix wrt Python v2.1 - If username is an empty string, user root is assumed. - 'postgresql' is not recognized anymore as a storage backend, you have to use 'pgstorage' instead. 'postgresql' used to be supported for compatibility with very old PyKota versions, this is not the case anymore. - 1.16alpha12 : - Job price added to job history, to keep accounting correct if a printer price per page or per job is modified and the history is not reset : the history would have given a false amount of money charged before the printer's prices were modified, this addition solves the problem. - 1.16alpha11 : - Some preliminary modifications to the database structure for PostgreSQL and OpenLDAP. A script to upgrade a PostgreSQL database is now provided. LDAP attributes now support eq, pres and sub indexes. - Job history now contains additional informations : number of copies, filename, title and print options. - 1.16alpha10 : - Some modifications done to external policy handling. See sample configuration file for details. - Spanish translation added. - 1.16alpha9 : - mailto can be set to external(...) to warn users above quota with any command of your choice (e.g. linpopup). Several parameters are available, see sample configuration file for details. - 1.16alpha8 : - Now cupspykota transmits SIGTERMS to the original backend when the user deletes a job being processed. - The documentation is now licensed under the terms of the GNU GPL too. - 1.16alpha7 : - The CUPS backend is now fully functionnal (it seems). - From now on, the use of the pykota filter with CUPS is deprecated. The pykota filter still has to be used with LPRng though. - 1.16alpha6 : - Bug fixed : when edpykota was used to add a printer with no user or group name, a bad user/group name "*" was added to the database. - 1.16alpha5 : - New CUPS backend supports device enumeration ! This means that NO manual CUPS configuration has to be done. All can now be done from CUPS web interface. - 1.16alpha4 : - New CUPS backend added. - 1.16alpha3 : - Minimal FAQ document added. - 1.16alpha2 : - CGI script now partially supports browser's language preference - 1.16alpha1 : - Ensures that the LDAP connection is really closed at end. - More complete error messages in case of LDAP operation failure. - 1.15 : - Documentation improved. - 1.15beta : - Command line tools now accept multiple printer names or wildcards, separated with commas. - 1.15alpha9 : - External policy can be used to auto-add unknown users and any other interesting things you may want to do. WARNING : You can't auto-add printers for now. They have to be created manually before. Look at sample configuration file for examples. - 1.15alpha8 : - Job history can be disabled if needed. This feature may be useful for people who use the LDAP backend and don't want their LDAP tree to grow out of control. - 1.15alpha7 : - More work done on caching wrt groups. - 1.15alpha6 : - Performance bottleneck with LDAP solved. - 1.15alpha5 : - Caching mechanism for all database queries. - 1.15alpha4 : - Problem with Python 2.3 fixed. Needs some definitive modification. - 1.15alpha3 : - CGI script now tries to restrict view when the REMOTE_USER environment variable is set. This allows people to protect its use with .htaccess files. When unprotected, or if REMOTE_USER is 'root', then unrestricted access is granted. NB : REMOTE_USER is automatically set by the web server (e.g. Apache) when a page with restricted access is viewed. - 1.15alpha2 : - Better detection of LPRng. - Outputs a warning if printing system is unknown. - 1.15alpha1 : - Bug fix in SQL code. - 1.14 : - The PostgreSQL database connection is now always closed before exit. The LDAP database connection too, but it didn't suffer from the problem anyway, AFAICT. - 1.14beta2 : - An LDAP-specific configuration field was introduced, usermail. Use it to specify which LDAP attribute contains a valid email address for the user. See sample configuration file for details. - 1.14beta : - Two new LDAP-specific configuration fields were introduced, newuser and newgroup. Use them to specify if you want to add quota information to existing entries or elsewhere. See sample configuration file for details. - 1.14alpha9 : - Documentation improved. - pykotme now also prints the user's account balance value if the user is registered in the Quota Storage and his limiting factor is "balance". - 1.14alpha8 : - bin/waitprinter.sh is now included to wait for SNMP enabled printers to be idle again before asking for their internal page counter. See sample configuration file for details. - 1.14alpha7 : - edpykota now doesn't check if the user/group is above quota. The check is only done at print time. This prevents the sending of email messages when creating user quotas if quota is already reached at creation time. - 1.14alpha6 : - Configuration file split and moved to /etc/pykota/pykota.conf and /etc/pykota/pykotadmin.conf to prevent simple users to have Read/Write access to the Quota Database. Don't forget to : $ chmod 640 /etc/pykota/pykotadmin.conf - storageuser and storageuserpw configuration fields reintroduced for the same reason. - Code cleaning for future implementation of email addresses support in the PostgreSQL and LDAP backends. - 1.14alpha5 : - Big bug fixed when changing the prices for a printer - 1.14alpha4 : - When adding an user, the user itself and its account balance are now added in the same LDAP entry. Before they were added as two different LDAP entries. - 1.14alpha3 : - Email field added to PostgreSQL database. An upgrade script is included in initscripts/postgresql. - 1.14alpha2 : - Typo fixed when using the --prototype option to edpykota - 1.14alpha1 : - Poor man's treshold included : users with a low account balance (below 1.0 by default, configurable) are sent a warning message. - All warning messages to users are now configurable. NB : Warning messages to the administrator are not. - 1.13 : - Some small fixes wrt group handling. - The whole module code was passed through PyChecker and several small code inconsistencies were fixed, as well as a bug in the groups code. - 1.13alpha1 : - Bug fix wrt LDAP. - 1.12 : - More complete test added in edpykota. - 1.12alpha2 : - The previous bug fix was incomplete. This is corrected. - 1.12alpha1 : - Small bug fix. - 1.11 : - Now includes the pykotme command line tool, to produce printing quotes for users. - 1.10 : - Bug fixed wrt quota reports when print limiting is by account balance. - 1.09 : - Translations finalized. - Manpages updated. - Installation documentation improved wrt LDAP support. - 1.09beta2 : - The CGI script for quota reports works again, and is much more powerful. - 1.09beta : - edpykota finalized - 1.09alpha2 : - LDAP schema modified a bit. Please upgrade. - LDAP backend works ! - repykota and warnpykota allow the root user to pass users or groups names on the command line. Users and groups names may contain wildcards. - repykota and warnpykota only reports or warns the current user when launched by non-root users. - Minor bug fixes. - 1.09alpha1 : - More work on LDAP storage backend. Many options were added to /etc/pykota.conf to give some hints to the LDAP storage backend. - Severe database access optimizations were done by rewriting most of the Quota Storage backends drivers. - 1.08 : - Major bug fix wrt LPRng support for remote jobs. - Major bug fix wrt increase/decrease account balances. - 1.08alpha8 : - Code refactoring. - A single user/password pair is used to connect to the database backend. The storageuser configuration field, and its associated passwords storageuserpw, are not used anymore. - You can now set PyKota in debug mode, see sample configuration file for details. WARNING : only Quota Storage backend queries are logged in debug mode, for now. - 1.08alpha7 : - Very latest LDAP schema. - Code enhancement wrt easy pluggability of PyKota attributes and object classes into an existing LDAP directory. - 1.08alpha6 : - Minor bug corrections - More work on LDAP : new schema - 1.08alpha5 : - More good work on LDAP storage. repykota now works reasonably well with the sample LDAP directory. - 1.08alpha4 : - LDAP schema included. THIS IS JUST FOR PEOPLE TO DISCUSS ABOUT THIS. LDAP SUPPORT DOESN'T WORK YET, AND THE SCHEMA MAY CHANGE IN THE FUTURE. PLEASE TELL US WHAT YOU THINK ABOUT THIS SCHEMA ON THE MAILING LIST. - 1.08alpha3 : - External accounting methods were partly rewritten : - No more "broken pipe" should happen. - They now take care of the number of copies This may be unneeded though, if the postscript file already does this, because this would overcharge users (number of copies counted two times). NEEDS MORE TESTING. - The sample configuration file now contains an external accounting method example which should work with all DSC compliant Postscript files. - Some small bugs were fixed. - 1.08alpha2 : - Now works with net-snmp v5.0 and above. It already worked, but the sample configuration file didn't contain appropriate values... - 1.07 : Release of the Shame ! - The external accounting methods driver was left out of the CVS tree. Nobody could use it since it wasn't included ! - 1.06 : - Severe bug fixed : if you had a printer in power saving state which couldn't answer immediately, a bug was triggered. It was caused by a bad copy/paste which forgot to import a Python module at run time. - 1.05 : - External accounters are finally available ! This means that you can plug any page accounting method you like by setting the appropriate 'accounter' field in /etc/pykota.conf See the sample conf/pykota.conf.sample to learn how to do. NB : Both 'external' and 'stupid' accounting methods account a job size just before *this* job is sent to the printer. The original 'querying' method accounts a job size just before *the next* job is sent to the printer. - 1.05alpha3 : - A 'stupid' and unreliable accounting method was implemented to serve as an example on how to do this sort of things. This method only counts the 'showpage' statements in the input data. See sample configuration file for details. Pluggable accounting methods work, but I advise you TO NOT USE THIS ONE WHICH IS JUST AN EXAMPLE. It is not reliable enough to be used. Use the 'querying' accounting method instead. - 1.05alpha2 : - Pluggable accounting methods. - Better error handling. - 1.05alpha1 : - SECURITY file added to help improve PyKota's security. - Extracting the printer's internal page counter is now tried several times, waiting several seconds between two tries. This lets the time to warm up for some printers which don't answer when they are sleeping (my Apple LaserWriter 16/600 PS is in this case, maybe others too) - Small display bug fixed in repykota - 1.04 : - Default print policy for users/groups unknown from the print quota system is now DENY instead of ALLOW, since ALLOW can generate inaccurate results (incorrect job sizes charged to the wrong persons) - LPRng support works ! - A bug was introduced some time ago wrt printers' default policy for unknown users. It is now corrected. - repykota now displays prices per job and per page for each printer, if they are defined. - Minor display bug fixed in repykota. - Problem when running repykota with an empty database was fixed. - 1.03 : - Upgrade script included for earlier versions. Please look inside the initscripts subdirectory. THE DATABASE SCHEMA HAS CHANGED, PLEASE UPGRADE. - repykota now reports account balances too. - PyKota now has a nice logo, see the logos subdirectory. - Manual pages were finally updated. - Group quotas seem to work now ! - The new database schema allows to keep an history of all jobs as well as charge users per page and/or per job, and track users' account balance. - edpykota now accepts much more command line options to use the new functionnalities. See edpykota --help for details. - The installation script now allows to install the sample configuration file during first installation. - More group quota code works, but still not finished. - The CGI script displays a link to PyKota's website. - The job history is now kept, this will allow per-period reports in the future. - After having modified the quota for an user with edpykota, a quota check is done to eventually warn the user/admin about a quota which is too low to print. - A workaround is provided for HP Printers : their internal page counter is only saved to NVRAM in a 10 increment, so if you switch them off and then on, the reported page counter may be lower than the real number of pages printed. See http://web.mit.edu/source/third/lprng/doc/LPRng-HOWTO-15.html We unconditionnally set the last job's page count to abs(int((10 - abs(lastcounter(snmp) - lastcounter(storage)) / 2)) in this case. For a more accurate accounting, never switch your HP printers off. - A fix is provided for printers which only have a volatile page counter (reset to 0 every time you switch the printer on) This should allow PyKota to work reasonably fine with HP Laserjet 4L/5L/6L, not perfect, but better than nothing. See http://web.mit.edu/source/third/lprng/doc/LPRng-HOWTO-15.html For a more accurate accounting, never switch your HP printers off. This is untested, please report any problem. - A bug was fixed when edpykota --add was used with users who already had a quota on the specified printer. - A small display bug in repykota was introduced in preliminary 1.03 versions, and fixed later on. - Some minor bugs which happened in rare situations were fixed. - Support for AppleTalk printers was added, see sample configuration file for details. - Users and group printing can now be controlled (limited) either by print quota or by account balance. - 1.02 : - The installation script now checks for software availability and in case a software is missing asks the user if he wants to continue with the installation or abort it. - The configuration file pykota.conf is now expected to be found in /etc instead of in /etc/cups The installation script prompts the user to see if he wants to move an old configuration file to the new location if needed. - Improved documentation. - You can now set the tcp/ip port on which the Quota Storage Server is listening, see sample configuration file for details. - Better general error handling. - Upgrade script for pre 1.01 PostgreSQL database schema is now included as well. - 1.01 : - The configuration file now accepts an option to choose the recipient(s) of the email messages : - DevNull means no one will receive them. - User means only the user will receive them. - Admin means only the admin will receive them. - Both means the User and the Admin will receive them. - The configuration file now uses hard-coded default values when an option is not set. See sample configuration file for details. - Manual pages are included since 1.00, but I forgot to add this information to this file. - Redistribution terms for the official package have softened and are now fully GPL compatible : unrestricted modification is now allowed even for the version number. - 1.00 : - edpykota now accepts a --noquota option. This disable quota checking while still doing page accounting. This is really useful for people who don't want to limit their users but want to know how much pages they print. - Some untested scripts were added to retrieve the life time page counter of non-SNMP printers. - Every directory now has a specific README file. - 0.99 : - Under some circumstances while the user wasn't allowed to print, he didn't receive any email message. It is now fixed. - When an user wasn't allowed to print, the quota for the previous user wasn't updated. This is now fixed. - 0.98 : - Correctly handle the case where the printer is switched off. - Small bug wrt syslog fixed. - 0.97 : - edpykota accepts wildcards on its command line for users/groups too if the --add option is not set, e.g. : $ edpykota --printer lp --softlimit 50 --hardlimit 100 "jer*" - If no user name is passed at all, then a default wildcard of "*" which means apply the command on ALL users for this printer is used. - Small bug fixes. - 0.96 : - Options requester, policy, admin, adminmail and gracedelay can now be set either globally or per printer. The printer option has priority if both are defined. - More powerful configuration parser. - If all options are defined globally, there's no need to define a section for each printer in the configuration file anymore. Just define a [global] section and it's ok. - 0.95 : - External requesters for printers finally added. - Full internationalization (english and french are supported) - More complete quota usage report - CGI script to access to the quota usage report - Several bugs fixed. - 0.9 : - First public version