On 03/05/2015 02:44 PM, Zbigniew Jędrzejewski-Szmek wrote:
While I would love to see 100% migration, the benefits for leaf packages aren't that big. We have fairly good compatibility support, and only a small number of people are using each package.
Nobody can use those leaf packages unless the legacy sysv initscript has been renamed in them which is why FESCo could just as well drop those package this release cycle and where did you get the statistic that only small number was using the remaining unmigrated components?
And there's the problem of who should do the work. I don't think FESCo members can be responsible personally for the implementation, and so far nobody has stepped up.
You yourself experience first hand not so long ago how incompetence those individuals in FPC regarding the uid/guid where FPC only needed to follow their own approved guideline o_O
Something Ovasik would have done blindfolded, holding a beer in one hand and typing on the keyboard with the other a while riding a unicycle those 8 months back when this request originally crossed his path but no someone singular or plural decided in his, her's or theirs infinity wisdom to move that process and workflow in the hands of the FPC where something went from someone who knows his stuff and would have been done in a jiffy to 9 months to be ( semi ) completed!
You have already addended those FESCo meeting which among other things they did not regonize you as neither the primary maintainer of systemd in the distribution nor lead developer from upstream! ( and this is just what has been happening in the year 2015 )
I'm not volunteering to dive that deep into the other 95 packages.
I said I would complete the work I started under certain condition on that report ( I hate leaving things unfinished ) and as I mentioned before to complete this will require 300 man hours then there is another 1000 manhours work in an cleaning up process ( not talking about removal here ) so FESCo needs to decide on what they are going to do *before* someone starts doing this and other related work and on top of that the distribution or Red Hat needs to decide what they intent do regarding "factory reset" ( RHEL 8 or not ) . ( From the looks of that will take equal amount of work that replacing the init system has taken. )
So not taking a decision without the power to have it implemented is better than having it taken and ignored. If ajax manages to kill the 17 -sysvinit subpackages, that will be a good first step.
Is it good the majority of FESCo decided to revert their own ( fesco at that time ) ignant previous decision making? no since if they bothered to take the time to inform themselves about the topic at hand in the first place then they would not have to spend time their and others time in undoing previous own work. ( In this case if they would have installed those subcomponents that contain legacy sysv initscript to be used as an replacement or alternative to the components existing units and try to use those legacy sysv initscript in conjunction with existing units, through updates etc they would have kicked those packages to the curb then and there at that meeting instead of wasting cpu cycle and infrastructure space continuing carrying them )
JBG -- devel mailing list devel@xxxxxxxxxxxxxxxxxxxxxxx https://admin.fedoraproject.org/mailman/listinfo/devel Fedora Code of Conduct: http://fedoraproject.org/code-of-conduct
