Reminder : Libgourou is an open source ADEPT protocol implementation (ePub DRM management from Adobe) that helps download ACSM files on Linux system (and remove DRM).

Just two little fixes after a user reported an issue downloading his ebook from Kobo. Unfortunately it doesn't fix his issue as it's more a problem with new cloudflare firewall setup by Kobo.

Changes :

• Bugfix : Operator URL only written when the full certificate is retrieved
• Raise an error for HTTP request with status != 200
• Set up cookie jar file for each session
• Fix for compilation in Android (strptime() format)

You can find source code and binaries in my forge

Reminder : Libgourou is a free ADEPT protocol implementation (ePub DRM management from Adobe) that helps download ACSM files on Linux system (and remove DRM).

No revolution for version 0.8.3 but only a bug raised by J.M. and a little feature coming from this bug :

• Bugfix : bad ID used for loaned files
• Server is now notified (if desired) when downloading file & loan return. Can be disabled with --no-notify option

You can find source code and binaries in my forge

Reminder : Libgourou is a free ADEPT protocol implementation (ePub DRM management from Adobe) that helps download ACSM files on Linux system (and remove DRM).

Libgourou v0.8.2 is now out ! Few changes since v0.8.1 :

• libgen.h missing
• Makefile updates (GCC 13 compilation, PREFIX and DESTDIR variable management)
• Bugfix : hexadecimal strings were not decrypted in PDF (can be used for table of content)

What's interesting is that all errors/fix has been reported by libgourou users. I specially want to thanks Berwyn H. for his kind donation !

You can find source code and binaries in my forge

Reminder : Libgourou is a free ADEPT protocol implementation (ePub DRM management from Adobe) that helps download ACSM files on Linux system (and remove DRM).

Libgourou v0.8.1 is out now. It's a minor release if we only look at the ChangeLog, but a very intersting one for a lot of aspects. First, the ChangeLog :

• DRM with encryption key size 192 bits now fully supported by adept_remove
• New -D option to specify .adept directory
• Check for user ID before trying to decrypt a file
• Default ADEPT directory is now /home/USER/.config/adept and can be specified by $ADEPT_DIR environment variable
• -f option can be omitted and is replaced by last argument (ie: we can do ./acsmdownloader EPUB)
• Add man pages
• Add install/unintall and install_headers/uninstall_headers Makefile targets
• PugiXML is no more statically linked, we use the system level shared library

Except one bug fix which add a major feature and new default path for .adept, all work has been done in Makefile part which is not code related. And this is interesting because it enlight the fact that libgourou is becoming a popular software (535 downloads within 4 months for v0.8), in "concurrence" with Knock (command line) and Calibre's plugin DeACSM (GUI). I get feedback from 5 different peoples that creates a port to MacOS, create packages for Gentoo and OpenSuse (ArchLinux has an outdated package) + a docker version. The reference implementation (utils) is becoming the only implementation and now has to be more user friendly. This is unexpected as, initially, the project has been designed as a library (so the name "lib" gourou) that should have been integrated by embedded devices, and a reference implementation to be used for Linux users. But I didn't expected that so many people were interested by the command line interface. If the project is still growing like that, I have no doubt that people will ask for new nice features, so wait and see !

You can find binary packages (compiled + AppImage) here.

Almost 3 years since the last news about IWLA. It does not really reflect the continuous development & maintenance of this wonderful tool. But don't worry, version 0.6 is now out ! The main change is the complete move from Python 2 to Python 3, but we may also mention :

• Users requests are no more saved (except if keep_requests is set) which allow to save a LOT of space
• A favicon is available
• Fresh synchronization with AWSTATS data
• Users need to do at least one hit per viewed page to not be marked as a robot
• Feed detector has been enhanced
• Track users plugin has been replaced by filter users which allows to define complex filters
• Users can be enlighted in all visits page
• IP can be anonymized (for public statistics)

The full ChangeLog is available here

While working on it, I realized how we can easily extend it. It's a real pleasure comparing to so big one PERL file code of AWSTATS, plus having it modular allows to implement our own rules which makes statistics really more precise. The only issue compared to AWSTATS is that IWLA is only focused on web statistics, but it has been design for it, not for everything related to log parsing !

New : A demo instance (for indefero.soutade.fr) is available here

I also decided to give up the old style branching model with master and dev. Using git and its lightweight branches, it's better to have a model with tags for stable releases and features branch for development. Code is not often updated and it makes no sense to have a master branch updated every 3 years with only one merge commit while dev is living.

I recently had look on concurrence, especially with Matomo and I was really afraid to see how users are tracked ! Everything is managed from pages viewed to cursor moves, user system information retrieval, time spent... All of this generate extra traffic and requires to execute Javascript code to obtain a lot of information about users's environment. But it's not the worst tool as it doesn't use commercial tracking (like Google Analytics) and keep data on webmaster's server and it's certified RGPD compliant. Commercial trackers are really a nightmare for consumer's privacy. Using it, webmaster can obtain really good statistics, but everything is stored on (abroad) commercial servers to create your profile ! Your profile is then sold or used to display you personalized advertising. Unfortunately, almost all websites are using them. In opposite, IWLA requires no cookies, no Javascript, no awful banner. It only parse and analyze log requests from webserver and generate a static HTML report which is the only right way to do !