On Wed, Jul 27, 2016 at 11:27 AM, Tomas Strachota <[email protected]> wrote:
> +1 for option 4. From what I saw it's common approach in other projects. > It's simple enough for start and we can re-evaluate later if we find it > problematic. > due to the fact that number 3 is not that simple either, I'm starting to lean towards 4 as well, as it will allow us to unblock yet will not break any packaging / testing etc, we could still decide its a blocker for 1.13 release regardless. Ohad > > T. > > On 07/26/2016 05:25 PM, Tomer Brisker wrote: > >> Hi, >> >> I meant to reply to this much sooner but it somehow slipped between the >> cracks. >> First, a big thank-you to Daniel for all the effort put into getting >> this done. >> I think either 3 or 4 make the most sense, but at this point anything >> that will let us move forward is fine with me. >> I still haven't managed to understand why we insist on building a ton of >> RPMs just so we can provide them to Koji. >> All the node modules are only needed for one step in building the rpm >> and are never used for anything else or provided to users. >> I would understand the need to separately package them if they were >> needed for running Foreman, but this is not the case. >> We are literally wasting weeks of time and I have no idea how many hours >> of engineering work for something that can be achieved in a simple tar >> ball or a single package containing everything. >> Lets finish this, please, one way or another, so we can concentrate on >> building a better Foreman for our community. >> >> Tomer >> >> >> On Fri, Jul 22, 2016 at 10:26 PM, Ohad Levy <[email protected] >> <mailto:[email protected]>> wrote: >> >> On Jul 22, 2016 9:01 PM, "Daniel Lobato Garcia" <[email protected] >> <mailto:[email protected]>> wrote: >> > >> > On 07/22, Daniel Lobato Garcia wrote: >> > > On 07/21, Daniel Lobato Garcia wrote: >> > > >> > > > 3. Put node_modules in a new package (foreman-npm-assets?) that >> > > > foreman-assets %install section puts in the right place? >> > > > >> > > > - Upsides: >> > > > - relatively easy to do >> > > > - jenkins can automate it >> > > > >> > > > - Downsides: >> > > > - less control or information for users about what's >> installed with >> > > > foreman, although we could specify the dependencies under >> > > > "Provides: bundled-npm(xxxx)" >> > > > >> > > > - Success with POC: haven't tried >> > > >> > > Update - I tried to make a POC for this quickly >> > >https://github.com/theforeman/foreman-packaging/pull/1240 >> > > >> > > I think it could work if there's a way of telling RPM to not care >> about >> > > the requirements -i.e: it notices there are multiple >> npm(tap-stream) >> > > versions and complains about that... >> > > >> > > To be fair I didn't expect RPM to read the contents of >> > > %{nodejs_sitelib} and figure out all the dependencies! >> > >> > Update 2: >> > >> > It's working! -http://i.imgur.com/c8CPRqq.gif >> > >> :) >> > Brief summary: >> > >> > The way foreman-node_modules works is by dumping node_modules in >> > /opt/foreman/node_modules, then copying that sources to the Foreman >> root >> > to ensure webpack:compile can run. Aside from that the Foreman >> package >> > itself now has a BuildDependency on node itself (& libuv) to run >> > webpack. >> > >> > That way we ensure no other program finds stuff it shouldn't on >> %{nodejs_sitelib} >> > and we don't conflict with the nodejs-XXX packages provided by the >> > linux distribution. >> > >> > It's not building in Jenkins, but it's just because I have not put >> the >> > sources for foreman-node_modules anywhere yet. You should be able to >> > build it with mockchain locally. >> > >> > I think this solution would be a good compromise considering the >> > difficulties 2. (bundled-npms) had and also it's much less >> intrusive - >> > other packages need not to care about our npm deps at all. >> > >> >> It sounds great but I'm really looking for Dominic feedback here. I >> hope it's enough to unblock merging webpack and can be improved if >> needed. >> >> Thanks for your efforts! >> Ohad >> > > >> > > -- >> > > Daniel Lobato Garcia >> > > >> > > @dLobatog >> > >blog.daniellobato.me <http://blog.daniellobato.me> >> > >daniellobato.me <http://daniellobato.me> >> > > >> > > GPG: >> http://keys.gnupg.net/pks/lookup?op=get&search=0x7A92D6DD38D6DE30 >> > > Keybase:https://keybase.io/elobato >> > >> > >> > >> > -- >> > Daniel Lobato Garcia >> > >> > @dLobatog >> >blog.daniellobato.me <http://blog.daniellobato.me> >> >daniellobato.me <http://daniellobato.me> >> > >> > GPG: >> http://keys.gnupg.net/pks/lookup?op=get&search=0x7A92D6DD38D6DE30 >> > Keybase:https://keybase.io/elobato >> > >> > -- >> > You received this message because you are subscribed to the Google >> Groups "foreman-dev" group. >> > To unsubscribe from this group and stop receiving emails from it, >> send an email [email protected] >> <mailto:foreman-dev%[email protected]>. >> > For more options, visithttps://groups.google.com/d/optout. >> >> -- >> You received this message because you are subscribed to the Google >> Groups "foreman-dev" group. >> To unsubscribe from this group and stop receiving emails from it, >> send an email to [email protected] >> <mailto:[email protected]>. >> For more options, visit https://groups.google.com/d/optout. >> >> >> >> >> -- >> Have a nice day, >> Tomer Brisker >> Red Hat Engineering >> >> -- >> You received this message because you are subscribed to the Google >> Groups "foreman-dev" group. >> To unsubscribe from this group and stop receiving emails from it, send >> an email to [email protected] >> <mailto:[email protected]>. >> For more options, visit https://groups.google.com/d/optout. >> > > -- > You received this message because you are subscribed to the Google Groups > "foreman-dev" group. > To unsubscribe from this group and stop receiving emails from it, send an > email to [email protected]. > For more options, visit https://groups.google.com/d/optout. > -- You received this message because you are subscribed to the Google Groups "foreman-dev" group. To unsubscribe from this group and stop receiving emails from it, send an email to [email protected]. For more options, visit https://groups.google.com/d/optout.
