Re: [HEADS UP] Replacing %post/%postun -p /sbin/ldconfig

2018-02-25 Thread Orcan Ogetbil
2018-02-14 15:59 GMT-05:00 Igor Gnatenko : > > Your options: > > * Speak up and tell package names I should not touch because … (you should > complete this sentence). > * Fix up packages and tell package names I should not touch because you did > that already. > * Tell package names you want to rem

Re: [HEADS UP] Replacing %post/%postun -p /sbin/ldconfig

2018-02-17 Thread Antonio Trande
On 14/02/2018 21:59, Igor Gnatenko wrote: > As guidelines changed[0] and now require maintainers who package libraries in > default library path (/usr/lib, /usr/lib64) to use %ldconfig_scriptlets in > case > they want spec file to be compatible with all Fedora/EPEL versions or drop > them > entir

Re: Re: [HEADS UP] Replacing %post/%postun -p /sbin/ldconfig

2018-02-17 Thread nicolas . mailhot
De: "Fabio Valentini" > Additionally, there are lots of packages that look like their > maintainer hasn't touched them in years (for example, the only git > commits are from mass rebuilds), Actually it's hard to find the motivation to touch a package just for a little syntactic cleanup. Mass c

Re: [HEADS UP] Replacing %post/%postun -p /sbin/ldconfig

2018-02-17 Thread Fabio Valentini
On Sat, Feb 17, 2018 at 11:02 AM, Remi Collet wrote: > Le 17/02/2018 à 10:05, Igor Gnatenko a écrit : >> On Sat, 2018-02-17 at 07:08 +0100, Remi Collet wrote: >>> Le 16/02/2018 à 15:18, Mark Wielaard a écrit : >> I had to tweak it a little though so the spec could still be build older RH

Re: [HEADS UP] Replacing %post/%postun -p /sbin/ldconfig

2018-02-17 Thread Roberto Ragusa
On 02/16/2018 10:35 PM, Jason L Tibbitts III wrote: >> "RR" == Roberto Ragusa writes: > RR> Was that a valid consideration? Has something changed on that front? > > It was, and packages will now fail to build (via brp-ldconfig) if they > don't package those symlinks. Though in practice pack

Re: [HEADS UP] Replacing %post/%postun -p /sbin/ldconfig

2018-02-17 Thread Remi Collet
Le 17/02/2018 à 10:05, Igor Gnatenko a écrit : > On Sat, 2018-02-17 at 07:08 +0100, Remi Collet wrote: >> Le 16/02/2018 à 15:18, Mark Wielaard a écrit : > >>> I had to tweak it a little though so the spec could still be build >>> older RHEL or Fedora (I reuse the spec to build on RHEL and with SCL

Re: [HEADS UP] Replacing %post/%postun -p /sbin/ldconfig

2018-02-17 Thread Igor Gnatenko
-BEGIN PGP SIGNED MESSAGE- Hash: SHA256 On Fri, 2018-02-16 at 15:18 +0100, Mark Wielaard wrote: > Hi Igor, > > On Wed, 2018-02-14 at 21:59 +0100, Igor Gnatenko wrote: > > Your options: > > > > * Speak up and tell package names I should not touch because … (you should > > complete this se

Re: [HEADS UP] Replacing %post/%postun -p /sbin/ldconfig

2018-02-17 Thread Igor Gnatenko
-BEGIN PGP SIGNED MESSAGE- Hash: SHA256 On Sat, 2018-02-17 at 07:08 +0100, Remi Collet wrote: > Le 16/02/2018 à 15:18, Mark Wielaard a écrit : > > > I had to tweak it a little though so the spec could still be build > > older RHEL or Fedora (I reuse the spec to build on RHEL and with SCL)

Re: [HEADS UP] Replacing %post/%postun -p /sbin/ldconfig

2018-02-16 Thread Remi Collet
Le 16/02/2018 à 15:18, Mark Wielaard a écrit : > I had to tweak it a little though so the spec could still be build > older RHEL or Fedora (I reuse the spec to build on RHEL and with SCL). > Maybe something like the following is better for people who have a spec > file they might reuse on systems

Re: [HEADS UP] Replacing %post/%postun -p /sbin/ldconfig

2018-02-16 Thread Jason L Tibbitts III
> "RR" == Roberto Ragusa writes: RR> When I proposed this kind of optimization in some mailing list RR> (maybe this one?!), I was answered that my method was not entirely RR> safe because there could have been problems for some rpm scripts RR> calling libraries that had been just upgraded (e.

Re: [HEADS UP] Replacing %post/%postun -p /sbin/ldconfig

2018-02-16 Thread Mark Wielaard
Hi Igor, On Wed, 2018-02-14 at 21:59 +0100, Igor Gnatenko wrote: > Your options: > > * Speak up and tell package names I should not touch because … (you should > complete this sentence). > * Fix up packages and tell package names I should not touch because you did > that already. > * Tell package

Re: [HEADS UP] Replacing %post/%postun -p /sbin/ldconfig

2018-02-16 Thread Roberto Ragusa
On 02/14/2018 10:14 PM, Matthew Miller wrote: > I'd be super-interested in benchmarks comparing before and after > install times. I guess since the plan is to do this _after_ the mass > rebuild, we'll need to wait until after the *next* rebuild to see how > much impact this has. Many years ago, I

Re: [HEADS UP] Replacing %post/%postun -p /sbin/ldconfig

2018-02-15 Thread Jonathan Wakely
On 14/02/18 21:59 +0100, Igor Gnatenko wrote: * Fix up packages and tell package names I should not touch because you did that already. boostdenisarnaud jwakely Done in https://src.fedoraproject.org/rpms/boost/c/4c456d525c8779b5ea8ef8b2031ad4eab6b66c61?branch=master mysql+

Re: [HEADS UP] Replacing %post/%postun -p /sbin/ldconfig

2018-02-15 Thread Pierre-Yves Chibon
On Thu, Feb 15, 2018 at 08:12:27PM +, Tomasz Kłoczko wrote: >It took me personally more than few years to come to conclusion that rpm >packages scriptlets idea is wrong. That exact sentence makes me wonder, did you watch the videos from Will at flock or DevConf? Because what you're tal

Re: [HEADS UP] Replacing %post/%postun -p /sbin/ldconfig

2018-02-15 Thread Tomasz Kłoczko
On 15 February 2018 at 15:19, David Shea wrote: [..] > Can we maybe step back and give other developers the benefit of the doubt > instead of immediately attacking an attempt to provide information? This is > really unnecessarily hostile. > What I wrote is not about hostility or attacking anyone

Re: [HEADS UP] Replacing %post/%postun -p /sbin/ldconfig

2018-02-15 Thread Miroslav Suchý
Dne 14.2.2018 v 21:59 Igor Gnatenko napsal(a): > msuchy abrt satyr * You should not touch because I done the change in upstream, it will be propagated during next release. Miroslav signature.asc Description: OpenPGP digital signature ___ devel m

Re: [HEADS UP] Replacing %post/%postun -p /sbin/ldconfig

2018-02-15 Thread David Shea
> Determinism level is about level of *variations* of the results on > repeating the same operations starting from exactly the same initial state. > Executing ldconfig after each package libraries installation/upgrade or > executing the same ldconfig only one time after install/upgrade libraries >

Re: [HEADS UP] Replacing %post/%postun -p /sbin/ldconfig

2018-02-15 Thread Robert-André Mauchin
On mercredi 14 février 2018 21:59:27 CET Igor Gnatenko wrote: > * Tell package names you want to remove ldconfig scriptlets entirely instead > of replacing them with %ldconfig_scriptlets and get fix **for free**. I wish you remove the scriplets entirely for Rawhide only on my packages: ec

Re: [HEADS UP] Replacing %post/%postun -p /sbin/ldconfig

2018-02-15 Thread Tomasz Kłoczko
On 14 February 2018 at 23:26, Will Woods wrote: [..] > I don't think this single change will make a huge difference within > the existing ecosystem, but I think it's an important step in a larger > shift toward make package installation & image composition a) > introspectable and b) deterministic

Re: [HEADS UP] Replacing %post/%postun -p /sbin/ldconfig

2018-02-14 Thread Vít Ondruch
Dne 14.2.2018 v 21:59 Igor Gnatenko napsal(a): > > * Speak up and tell package names I should not touch because … (you should > complete this sentence). > * Fix up packages and tell package names I should not touch because > you did > that already. > * Tell package names you want to remove ldconf

Re: [HEADS UP] Replacing %post/%postun -p /sbin/ldconfig

2018-02-14 Thread Will Woods
On Wed, Feb 14, 2018 at 4:14 PM, Matthew Miller wrote: > I'd be super-interested in benchmarks comparing before and after > install times. I guess since the plan is to do this _after_ the mass > rebuild, we'll need to wait until after the *next* rebuild to see how > much impact this has. I don't

Re: [HEADS UP] Replacing %post/%postun -p /sbin/ldconfig

2018-02-14 Thread Matthew Miller
On Wed, Feb 14, 2018 at 09:59:27PM +0100, Igor Gnatenko wrote: > The whole purpose of this is to make installation of packages FASTER and > obviously to comply with guidelines. Most of packages would be possible to > automate, however some would not and you would need to deal with it youself. I'd