Le Tuesday 13 November 2001 à 01:01:07, Jean-Luc Giraud a écrit: > I don't understand why there should be a debate here. For me, this is > not an Open Source vs Closed Source issue: I am willing to use Dave's > applet now and trust that he will make the source available soon.
It depends how you define "Open Source". David said "The license currently restricts redistribution.". I really hope it will change and I think it will. I am just pushing a little bit :-) > To me, the real question is: does Dave think that a "release often, > release early" policy is appropriate for his applet. It is not because > the Linux kernel is developed according to such a policy that it is the > best one for the musclecard. This is exactly the point I think David is wrong on. - pcsc-lite-1.0.0B was released 23th of August 2001 - I send a patch to Dave 9th of September 2001 - Dave sent my back pcsc-lite-1.0.0G on 10th September 2001. This version was not public. - Dave annonced pcsc-lite-1.0.1 on 7th November 2001. So during 2 months I used a version 1.0.0G that was not public, I corrected a bug and other developers could not benefit from it (only few of them received 1.0.0G or newer versions). My questions are: - what can be my motivation to work on pcsc-lite if bug corrections I sent are released 2 months after? - what can be my motivation to work on pcsc-lite if potentially another developer found the bug and already corrected it and sent a patch to Dave. Dave sent a new version to this other developer but not to me. If i correct the bug and send a patch to Dave I will just lose my time (and Dave also) since the bug is already corrected in Dave's version. The model "release often, release early" works well because other developers see the code evolving and improving. They know that if they find a bug and send a patch the patch will be included and a new release will be done so the bug correction will not be done by someone else. With Dave's project management many developers use different versions of the code (which I imagined Dave wanted to avoid: multiple different vesions from the same code base). And more importantly I lost my motivation to find bugs and correct them. I would like to see features included on pcsc-lite but will not work on it since it may be a duplicate effort. What I could do is maintain an unofficial branch of pcsc-lite with patches sent by anyone to me. I release a new version each time I receive a patch so that people can work on the latest version and not a 2 months old one. And again this is exactly what Dave wants to avoid. > Instead of criticising Dave, why don't we thank him for letting us use > his applet, which he did not have to do at all. I am criticising Dave just to try to improve his responsiveness and make pcsc-lite a better program _faster_. I already asked him to create a "Unstable" section for http://www.linuxnet.com/middle.html but without success so far :-( You may notice I used the term "developer" and not "user". Users should use a stable release for their day-to-day work. > At least Dave has defined an Open standard which we can start using. > Be patient for the rest. Yes I am patient. But Dave treats me as a simple user that would complain to hell if he finds a bug. I just want to help him provides better programs faster. Dave has done a tremendous good job until now but I don't find his project management is optimal in term of development resource usage. Regards, PS: for those without PowerPoint, the presentation musclecard.ppt is available as html with .png bitmap images [1] and also as one .tgz [2] with all the .png and .html files. The Perl script used to generate the .html files is also provided, in source code :-) and under the GNU GPL licence :-). I sent the .tgz to Dave two days ago but no news of it since and not yet on linuxnet.com so I provide it on my web site. [1] http://ludovic.rousseau.free.fr/softwares/musclecard/musclecard/Diapositive1.html [2] http://ludovic.rousseau.free.fr/softwares/musclecard/musclecard.tgz [3] http://ludovic.rousseau.free.fr/softwares/musclecard/musclecard/convert.pl -- Ludovic Rousseau [EMAIL PROTECTED] -- Normaliser Unix c'est comme pasteuriser le Camembert, L.R. -- *************************************************************** Unix Smart Card Developers - M.U.S.C.L.E. (Movement for the Use of Smart Cards in a Linux Environment) http://www.linuxnet.com/ To unsubscribe send an email to [EMAIL PROTECTED] with unsubscribe sclinux ***************************************************************
