This is the mail archive of the cygwin-apps mailing list for the Cygwin project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]
Other format: [Raw text]

Re: Planned setup.ini changes for early 2018


Jon Turney writes:
> The purpose of my original email is to start that discussion.

Well, sorry then for not recognizing that, but from your other answers
in this thread I've got the impression src/ is a done deal that you
didn't want to discuss.

> If I've understood correctly, your objection is that this will break
> installing source packages from mirrors which choose to mirror
> selected subdirectories (e.g. x86_64/ and noarch/).

My personal objection is that it will add another two hours to mirroring
the Cygwin repo via a HTTP proxy (or I just give up on mirroring the
sources).  But yes, not breaking things for folks who try to be nice to
mirror operators is next on my list.  Then come the mirror operators
themselves.  I must admit I hadn't considered the Cygwin Time Machine,
this adds another twist that I haven't fully thought through yet.

> I guess I consider that counterbalanced by the ability to selectively
> not mirror src/.

I don't, due to the conspicUous naming of the source archive files that
effect is easily achieved by all mirroring methods that I have
personally used.  So it doesn't really add any value that I can see to
counterbalance the other ripple effects it will have.

Unfortunately I don't have a lot of time on my hands right now, but I
think (very preliminarily) that aside from keeping the directory
structure intact, we should also keep the old setup.ini as-is (not just
compatible) and instead move the "new" setup.ini one level up and have
it encompass both supported architectures using a more compact syntax.
It should be possible to generate the "old" x86{,_64}/setup.ini by
transforming the "new" setup.ini so you can hopefully drop some more
baggage in calm.  The main benefit of using a modified format would be
to not have the redundant information in it as the current format would
need to carry.  As long as we keep the old format around that redundancy
is effectively still with us, but we can isolate it more easily.


Regards,
Achim.
-- 
+<[Q+ Matrix-12 WAVE#46+305 Neuron microQkb Andromeda XTk Blofeld]>+

SD adaptations for Waldorf Q V3.00R3 and Q+ V3.54R2:
http://Synth.Stromeko.net/Downloads.html#WaldorfSDada


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]