I can't take the liberty to publish the letter itself (to the forum admins) but I don't think the reasoning should be kept secret and not discussed.
Basically he states that lack of support is the reason for departure.
In detail I understand that the comparison with systemd is of a really messy system with tremendous support (some heavy duty multinational corporations financing and paying people to provide support through their involvement in distributions) and systemd alternatives that lack such support, but very complex for users to live without.
Eric and JM are two people who have contributed thousands of hours of expertise based work to provide what we have, as obarun, as 66, and as viable way to utilize s6 to its maximum, something that the s6 developer hasn't provided himself (a single entity with some small community of experts who help in providing ideas for improvement and development). So basically we are enjoying the fruit, free of commitments and cost, of the hard free open work of three people.
For a couple of years I have tried to help a bit with documentation and the wiki, and randomly when I think I know the answer to a problem around 66 I offer a suggestion.
Objectively, if one uses the installer and installs Obarun, one gets a fully functional system. Installing more software is just like arch, no difference in procedure. Adding and subtracting services, daemons, and modules, takes "a little bit" of reading of the material. One doesn't need to become an expert in 66, let alone s6 which doesn't need to be touched at all when using 66, but requires some reading, a bit of playing around, and although not as simple as runit, it is much closer to "common sense" than sysvinit or openrc. After what, 30 years, of sysvinit, there should be nobody in linux who doesn't understand what the difference between run level 4 and runlevel 5 is. What is it?
Obarun is around for 6 years, and 66 for 2? 66 is on version 0.6 which means 66 version 1.0 has a little bit more to go still. Any bug that has been discovered hasn't prohibited anyone from doing work on Obarun, it is usually a minor little detail of refinement. What is there to compare it with? Runit? Runit is the 4th most common init/service management system and has been around longer than s6, and hasn't seen a single letter of development in the past 3-3.5y? Yet is relied upon as a "good" alternative to systemd.
90% of the problems addressed in the forum about 66 are answered by text in the wiki (about 15' of careful reading) and software documentation (a couple of hours maybe worth of reading for those with a deeper interest). Questions are answered regardless the lack of any reading attempt by the user. They could have been dealt with classic Arch arrogancy: "read the manual, read the wiki".
This is not Ubuntu. Some canonical employees have 6 digit ($/Eu) salaries for the involvement in support and documentation provision. This is why they so fiercely defend the hands that feed them.
Can it be done otherwise? Yes!
Eric can retain his git account, publish his work on 66 alone, nothing else. Document it, just as s6 is documented (good luck if you don't have a graduate degree in CS) and you are on your own configuring and installing it in any distribution you wish.
Sinit was published for many years, it didn't need any further development, all 30 lines of it. It worked. It took someone to write some more lines and develop ssm to make it work in Arch as display. Support? Maybe if you are lucky you may get 2 lines of response a week after you sent an email. That is Spark-Linux. Arch with sinit and ssm. Easy! It does 1/10000 of the work s6 and 66 do. No forum, no discussion list, no publication or instructions on how to really make a functional system out of it. A bug took 6 months to be addressed and the response was "since it is not working anymore, don't use it". This was about using file checking on boot. Someone can say it is better, it is easier, it is not complex, ..... and like it. Some wood crafters/luthiers still prefer a hand powered drill to open holes on musical instruments.
Working on obarun, 66, repackaging arch's crap, even writing documentation or correcting it, doesn't pay for shoes, food, cigarettes, or ISP. It doesn't even pay for a server where all of this information is reliably available globally.
It is mind boggling, thousands of people were sad because Antergos, a "distribution", was abandoned. This was arch with colorful themes and icons. "But it had great support in various languages, specifically in Spanish and Portuguese".
I am not sad because Lizzias is leaving, I wish he could stay, I am sad for the reasoning of his departure.
Basically he states that lack of support is the reason for departure.
In detail I understand that the comparison with systemd is of a really messy system with tremendous support (some heavy duty multinational corporations financing and paying people to provide support through their involvement in distributions) and systemd alternatives that lack such support, but very complex for users to live without.
Eric and JM are two people who have contributed thousands of hours of expertise based work to provide what we have, as obarun, as 66, and as viable way to utilize s6 to its maximum, something that the s6 developer hasn't provided himself (a single entity with some small community of experts who help in providing ideas for improvement and development). So basically we are enjoying the fruit, free of commitments and cost, of the hard free open work of three people.
For a couple of years I have tried to help a bit with documentation and the wiki, and randomly when I think I know the answer to a problem around 66 I offer a suggestion.
Objectively, if one uses the installer and installs Obarun, one gets a fully functional system. Installing more software is just like arch, no difference in procedure. Adding and subtracting services, daemons, and modules, takes "a little bit" of reading of the material. One doesn't need to become an expert in 66, let alone s6 which doesn't need to be touched at all when using 66, but requires some reading, a bit of playing around, and although not as simple as runit, it is much closer to "common sense" than sysvinit or openrc. After what, 30 years, of sysvinit, there should be nobody in linux who doesn't understand what the difference between run level 4 and runlevel 5 is. What is it?
Obarun is around for 6 years, and 66 for 2? 66 is on version 0.6 which means 66 version 1.0 has a little bit more to go still. Any bug that has been discovered hasn't prohibited anyone from doing work on Obarun, it is usually a minor little detail of refinement. What is there to compare it with? Runit? Runit is the 4th most common init/service management system and has been around longer than s6, and hasn't seen a single letter of development in the past 3-3.5y? Yet is relied upon as a "good" alternative to systemd.
90% of the problems addressed in the forum about 66 are answered by text in the wiki (about 15' of careful reading) and software documentation (a couple of hours maybe worth of reading for those with a deeper interest). Questions are answered regardless the lack of any reading attempt by the user. They could have been dealt with classic Arch arrogancy: "read the manual, read the wiki".
This is not Ubuntu. Some canonical employees have 6 digit ($/Eu) salaries for the involvement in support and documentation provision. This is why they so fiercely defend the hands that feed them.
Can it be done otherwise? Yes!
Eric can retain his git account, publish his work on 66 alone, nothing else. Document it, just as s6 is documented (good luck if you don't have a graduate degree in CS) and you are on your own configuring and installing it in any distribution you wish.
Sinit was published for many years, it didn't need any further development, all 30 lines of it. It worked. It took someone to write some more lines and develop ssm to make it work in Arch as display. Support? Maybe if you are lucky you may get 2 lines of response a week after you sent an email. That is Spark-Linux. Arch with sinit and ssm. Easy! It does 1/10000 of the work s6 and 66 do. No forum, no discussion list, no publication or instructions on how to really make a functional system out of it. A bug took 6 months to be addressed and the response was "since it is not working anymore, don't use it". This was about using file checking on boot. Someone can say it is better, it is easier, it is not complex, ..... and like it. Some wood crafters/luthiers still prefer a hand powered drill to open holes on musical instruments.
Working on obarun, 66, repackaging arch's crap, even writing documentation or correcting it, doesn't pay for shoes, food, cigarettes, or ISP. It doesn't even pay for a server where all of this information is reliably available globally.
It is mind boggling, thousands of people were sad because Antergos, a "distribution", was abandoned. This was arch with colorful themes and icons. "But it had great support in various languages, specifically in Spanish and Portuguese".
I am not sad because Lizzias is leaving, I wish he could stay, I am sad for the reasoning of his departure.