# old school installation

**URL:** <https://forum.ansible.com/t/old-school-installation/25471>\
**Category:** AWX Project\
**Created:** [September 13, 2017, 9:00am UTC](https://forum.ansible.com/t/old-school-installation/25471 "2017-09-13T09:00:36Z")\
**Posts on this page:** 6\
**Page:** 1

<div class="post-metadata">

**Author:** ![swiedernix](https://sea1.discourse-cdn.com/ansible/user_avatar/forum.ansible.com/swiedernix/32/9806_2.png) [@swiedernix](https://forum.ansible.com/u/swiedernix)\
**Post date:** [September 13, 2017, 9:00am UTC](https://forum.ansible.com/t/old-school-installation/25471/1 "2017-09-13T09:00:36Z")

</div>

nice, finally tower is open source.

but I’d like to do an old school installation, any chance to see this within the next weeks?

bye

---

<div class="post-metadata">

**Author:** ![matburt](https://sea1.discourse-cdn.com/ansible/user_avatar/forum.ansible.com/matburt/32/3612_2.png) [@matburt](https://forum.ansible.com/u/matburt)\
**Post date:** [September 13, 2017, 11:41am UTC](https://forum.ansible.com/t/old-school-installation/25471/2 "2017-09-13T11:41:18Z")

</div>

We have no current plans to implement a different install target in the open source AWX repository.

---

<div class="post-metadata">

**Author:** ![Paul\_Archer](https://avatars.discourse-cdn.com/v4/letter/p/839c29/32.png) [@Paul\_Archer](https://forum.ansible.com/u/Paul_Archer)\
**Post date:** [September 16, 2017, 5:10am UTC](https://forum.ansible.com/t/old-school-installation/25471/3 "2017-09-16T05:10:20Z")

</div>

Are you opposed to, as Stefan said, and “old school” installation, or just not planning on it yourself? Would you consider a PR if someone implemented it?

---

<div class="post-metadata">

**Author:** ![matburt](https://sea1.discourse-cdn.com/ansible/user_avatar/forum.ansible.com/matburt/32/3612_2.png) [@matburt](https://forum.ansible.com/u/matburt)\
**Post date:** [September 16, 2017, 3:14pm UTC](https://forum.ansible.com/t/old-school-installation/25471/4 "2017-09-16T15:14:55Z")

</div>

In a way I am a bit opposed to it, even from a PR. Having an alternate bare-metal/VM based install opens us up to maintaining and supporting that. On the Tower side we already have experience with that and it’s nearly a full time job as it is. I’m not opposed to someone else owning it, but that person absolutely has to be the one to own it.

---

<div class="post-metadata">

**Author:** ![Paul\_Archer](https://avatars.discourse-cdn.com/v4/letter/p/839c29/32.png) [@Paul\_Archer](https://forum.ansible.com/u/Paul_Archer)\
**Post date:** [September 18, 2017, 5:38pm UTC](https://forum.ansible.com/t/old-school-installation/25471/5 "2017-09-18T17:38:45Z")

</div>

Matt, is the code that does that build still proprietary? If it is, as you say, nearly a full time job to support this, wouldn’t it benefit you to release that code to allow the community to help with it?  
And isn’t it more work to deal with three build processes (docker, kubernetes, bare metal), than one or two? I don’t quite understand why, when Tower is built one way, you would take the time to use docker to build it a different way.

Thanks,

Paul

---

<div class="post-metadata">

**Author:** ![Greg\_DeKoenigsberg](https://avatars.discourse-cdn.com/v4/letter/g/c68b51/32.png) [@Greg\_DeKoenigsberg](https://forum.ansible.com/u/Greg_DeKoenigsberg)\
**Post date:** [September 18, 2017, 5:55pm UTC](https://forum.ansible.com/t/old-school-installation/25471/6 "2017-09-18T17:55:27Z")

</div>

> Matt, is the code that does that build still proprietary? If it is, as you  
> say, nearly a full time job to support this, wouldn't it benefit you to  
> release that code to allow the community to help with it?  
> And isn't it more work to deal with three build processes (docker,  
> kubernetes, bare metal), than one or two? I don't quite understand why, when  
> Tower is built one way, you would take the time to use docker to build it a  
> different way.

The docker build is much simpler by design. It's a great way to  
constrain weird variables, and for small deployments, it makes sense  
to constrain as much as possible to limit weird failure cases.

Ensuring supportability of complex deployments is an enterprise  
feature, and for good reason. That's something that's generally needed  
by large organizations, and it's something they can and should pay  
for. That's our business model.

As Matt said before, others are free (and encouraged!) to build and  
support their own deployment mechanisms.

--g