On 2012.04.21 10:57, Vincent Pelletier wrote: > Le samedi 21 avril 2012 10:50:29, Jose Pablo a écrit : >> It seen you guys are taking the project by force > > My feeling, exactly.
Which is not surprising, as, in case you haven't followed discussions on the libusb mailing list over the past few years, you should realize that this is a hostile fork (but then again, does there exist a non-hostile fork?), prompted by by the fact that the sole libusb maintainer for the past 18 months, Peter, has continuously refused to pay any attention to the advice of many of the core list members, let alone requests from libusb users, and very openly ignored the many people who had an issue with his vision. Among many fairly damaging things, he appears to consider that not producing a release for 2 years, despite 250 accumulated commits, is OK as long as you say "sorry it took so long"... Therefore, please understand that this is not a "We wanted to add new APIs but they were rejected" kind of fork but an "Unless we take action now, libusb is going to die" kind of fork. > To libusbx team: > > Please understand that you are technically abusing sonames. > > Of course, you can argue you are only adding functions, not changing existing > ones hence not falling in a case where soname change is required by the > book... Then what should happen to libusb now that/when it lacks functions > existing in another "libusb-1.0" ? Move to another soname ? > > I was originally quite exited to get more API to wrap in python-libusb1. Then you shouldn't consider libusbx as another version of libusb, with each one offering different features and with the ability to coexist. Instead you should consider for what it is: a means to abandon the unhealthy libusb project if, as we think everyone should by now, you have concerns about its future. > Now > all I can feel is sympathy for Peter: he was here first, he got to choose a > soname and now has enough "market[1] share" to be able to reasonably claim it > as his. And yet he has repeatedly displayed complete disdain for libusb users, such as yourself, by not releasing any new version or only doing so as a response to a fork. If you can't ready a release in 18 months, despite plenty of features and bugfixes, or, as has been his norm, delay the processing of important patches for months, then whatever "market share" you have can only be considered as up for grabs. > Because there were not enough releases in libusb, you now effectively > prevent any further release or force him to follow your API additions... The fork wasn't prompted by a disagreement over adding a new API. The API additions are only a by-product of having our own fork and, for lack of a better analogy, being freed from Peter's tyranny. > Is there any difference with taking over libusb project ? Hopefully not. At least as far as I'm concerned, I would say the sooner libusb dies and libusbx takes over, the better it will be for users. Unfortunately, this very much requires users realizing that libusbx is indeed the better alternative, which means understanding the reasons for the fork and why we had very little choice but to try to take over the libusb project in our own fork. Regards, /Pete ------------------------------------------------------------------------------ For Developers, A Lot Can Happen In A Second. Boundary is the first to Know...and Tell You. Monitor Your Applications in Ultra-Fine Resolution. Try it FREE! http://p.sf.net/sfu/Boundary-d2dvs2 _______________________________________________ libusbx-devel mailing list libusbx-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/libusbx-devel
