The problem is Docker. The named volumes initialize in a different way, 
Mayan relies on that behavior for writing the initial setting files. I see 
the merit in changing that but it requires changing the way Mayan itself 
works. For the moment we are trying to keep the major functionality intact 
and focusing on usability, tests, minor features and bugs. 

We are discussing the possibility of another fork. Mayan EDMS NG will 
remain 100% compatible with Mayan EDMS. The new fork will deviate and will 
be the place where we test all the ideas and changes we would like to see 
in Mayan on the long run. Stuff like changing the initialization and make 
it work with standard Docker volumes. Things that work well and make the 
product better will be ported from the new fork back to Mayan EDMS NG and 
eventually to Mayan EDMS. We want to call this something different to make 
it clear that it is not completely the same as Mayan. Any idea for the name 
is welcome.

On Thursday, March 1, 2018 at 9:50:11 AM UTC-4, RW Shore wrote:
>
> Good. So far the existing docker image is working fine for me, though I 
> need to figure out how to install the EXIF processing extensions. One 
> request I would make, though, and that's to get rid of the named volume, 
> for the following reasons.
>
> I run mayan-edms in a docker swarm, with glusterfs-based shared disks 
> across the three hosts involved. This means that if/as the service migrates 
> from host to host, the service sees the same data files, independent of the 
> host it's running on. Having a named volume in the mayan-edms image means I 
> need to add a couple of steps to the build process:
>
> 1. I need to install a special volume driver, local-persist, so I can 
> re-direct the external definition of the volume out of 
> /var/lib/docker/volumes.
> 2. I need to define the volume on each of the hosts in the swarm (a volume 
> definition is always local).
>
> By making the external instantiation of /var/lib/mayan a normal mount 
> rather than a volume mount, I can more easily add hosts to the swarm 
> without having to remember to define the mayan_data volume, and I can use 
> bind mounts in the service definition rather than volume mounts, avoiding 
> the need for the local-persist driver.
>
>
> On Wed, Feb 28, 2018 at 7:36 PM, <lonevi...@gmail.com <javascript:>> 
> wrote:
>
>> Yes we do :) But are trying to push some improvements before releasing 
>> it. Eric is working on modernizing the existing Docker image. He is also 
>> trying to merge it with the ARM Docker image to provide a single multi 
>> architecture image that would work the same on a PC and on a Raspberry Pi. 
>> We are also experimenting with Alpine Linux to reduce the image size. Once 
>> finished we will announce it on the list.
>>
>> On Wednesday, February 28, 2018 at 2:07:59 PM UTC-4, RW Shore wrote:
>>>
>>> re version 2.8: do you plan to have a Docker container as well? (hint 
>>> hint :-)
>>>
>>> On Tue, Feb 27, 2018 at 5:40 AM, <lonevi...@gmail.com> wrote:
>>>
>>>> Hi all,
>>>>
>>>> We love Mayan EDMS and have come to depend on it for our daily work. 
>>>> Like many of you we are worried about the effects hurricane Maria has had 
>>>> on the development of the project. So we decided to do something about it.
>>>>
>>>> We gained good knowledge about the internals of Mayan EDMS from 
>>>> customizing 
>>>> it for our own use. Looking at the code for the next version we could see 
>>>> that much work had already been done. We were confident we could finish 
>>>> the 
>>>> remainder tasks to be able to take the code to a point where a version 2.8 
>>>> would be possible.
>>>>
>>>> We contacted Roberto with our idea. While his communication options 
>>>> are limited, he nonetheless provided us with some great guidelines. We 
>>>> don't pretend our fork to replace Mayan EDMS so we are naming it Mayan 
>>>> EDMS 
>>>> NG (Next generation). To make things easier we continued the version 
>>>> numbering. 
>>>>
>>>> So, we are proud to release our work as Mayan EDMS NG version 2.8.
>>>>
>>>> Our forks is just a continuation of the work already started by Roberto 
>>>> for version 2.8. Therefore it is 100% compatible with any existing data. 
>>>>
>>>> Our work focused on the following areas:
>>>>
>>>> * Finishing the API test refactor. The purpose of this refactor is to 
>>>> test each API function for permission and access failure and success. The 
>>>> API tests we also updated to conform with the new API Test class interface 
>>>> which reduces a lot of boilerplate code. There are now a minimal of two 
>>>> tests for each API function.
>>>>
>>>> * Some minor permission changes were made when obviously needed such as 
>>>> the Workflow Create permission not having any effect.
>>>>
>>>> * Roberto had already backported the notifications feature to the 
>>>> "next" branch and our work just focused on giving the final polish to this 
>>>> feature by removing duplication.
>>>>
>>>> * Some user interface fixes and updated we added. These were inspired 
>>>> by discussion on the mailing list and from looking at the Open Paperless 
>>>> fork by Tina Zhou.
>>>>
>>>> * Introduction of the MERC (Mayan EDMS Request for Comment) as a means 
>>>> of documenting the internals of the project, proposing features, changes 
>>>> to 
>>>> the code as well as to the related process of the project. This type of 
>>>> documentation mechanism has been very successful for other things like the 
>>>> Internet (RFCs), Python (PEPs) and Django (DEPs) and we are sure it will 
>>>> also help Mayan EDMS a lot.
>>>>
>>>> The repository for this fork is 
>>>> https://gitlab.com/Michael.Price/mayan-edms. From there you can 
>>>> examine the changes that we made to get the code to release level. Our 
>>>> work 
>>>> is available from PyPI as mayan-edms-ng.
>>>>
>>>> We hope you enjoy it, thank you.
>>>>
>>>> Michael Price, internals and process
>>>> Eric Riggs, user interface
>>>>
>>>> -- 
>>>>
>>>> --- 
>>>> You received this message because you are subscribed to the Google 
>>>> Groups "Mayan EDMS" group.
>>>> To unsubscribe from this group and stop receiving emails from it, send 
>>>> an email to mayan-edms+...@googlegroups.com.
>>>> For more options, visit https://groups.google.com/d/optout.
>>>>
>>>
>>> -- 
>>
>> --- 
>> You received this message because you are subscribed to the Google Groups 
>> "Mayan EDMS" group.
>> To unsubscribe from this group and stop receiving emails from it, send an 
>> email to mayan-edms+...@googlegroups.com <javascript:>.
>> For more options, visit https://groups.google.com/d/optout.
>>
>
>

-- 

--- 
You received this message because you are subscribed to the Google Groups 
"Mayan EDMS" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to mayan-edms+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

Reply via email to