Developers of Sympa

[Soji Ikeda <address@concealed>]

Marc, I never have told you not to use mailing list for discussion.

Someday I told (on the chat) that, people may use the way they prefer, either 
list or GitHub issue tracker. I don’t want to force using gh. However, you 
have repeatedly complained about interface of gh, and at last proposed 
limiting use of gh.

It’s in vain.  Instead of behaving such, why don’t you use mailing list? For 
example: Why don’t you write about wiki system you made and ask folks the 
comments? (Are we, “core developers”, less important than audiences of fosdem 
on this issue?)

Therefore I ask you two things:

  - Please feel free to make discussion, exchange information and so on, 
using the way you prefer.
    Anyone have not denied your effort.

  - However, please consider learning the other ways the other people prefer
    For example, make a few hours in each week to check github issues and PRs 
to know:
      - what is up
      - how people put the development forward
      - and how they make decision among them

After experiences for a while (months), if you still feel change is needed, 
please propose.

Btw, you asked me if I disagree to all of your proposal: Answer is no, but I 
feel you seems not understanding reality of development ongoing (e.g. removal 
of CGI support —- have you run Sympa with  CGI more?)

Regards,
— Soji


2018/01/08 17:03、Marc Chantreux <address@concealed>のメール:

> hello Soji,
> 
>> I can’t agree to all
>>  of your post below.
> 
> do you mean you disagree everything ?
> 
>> Especially, I definitely disagree to your argument that “core
>> developers” should not use function of GitHub.
> 
> please notice that none of us said that.
> 
>> You should tell us the basis of your argument to persuade us.
> 
> sympa-developpers was the way to exchange those kind of exchanges
> way before gh and worked well. we agreed that we should test the gh
> issues to replace the tracker, not the mailing list. so please persuade
> us to adopt a change in the workflow you did without debating it with
> the community.
> 
> if you are open to the debate, i'll enjoy reading you but as you asked
> me, i can give you my opinion on it:
> 
> * i don't follow every issues but i carefully read developpers
>  (as i did on other projects)
> * fixing bugs and talking about architecture isn't the same
>  at all so i think it's nice to open the list archive to
>  know about the direction of sympa (not doing some wishfull digging
>  into tickets)
> * as racke said on my very last gh comment: we should be careful
>  to restrict information. in your last issue, you just wanted
>  to introduce behavior changes in a patchlevel version. Worse:
>  you removed the CGI level in a patchlevel version. As system
>  administrator as a patchlevel should be applied confidently so
>  if i was one of them, i would be very upset. This should be debated
>  but GH issues aren't the good tool to have a thread that is
>  potentially long and large.
> 
> you did an amazing job by taking care of the entrant issues, tagging
> them and releasing new versions of sympa and i think every members of
> the community is grateful on it (i really am) but David took this
> community bound by listening to everyone, not taking abitrary decisions
> without even discuss them.
> 
> We have to walk on his steps of widsom and that's what the hackathon
> was: i wanted something to happens that respect this idea the way David
> did the things so i had a list of suggestions, all debated, some
> rejected because i care what people think (especially using such an old
> version of perl when most of us are running 5.24 and more). I even
> polled sympa-users for those kind of decisions. You removed CGI support
> without even pinging developpers.
> 
> Now our arrival (David and I) at Renater is done, David and I will be
> involved in the sympa code and i will use this list to make RFC as
> long as there is no clear majority to use gh to talk about design.
> 
> Regards
> Marc
>