About Us › Forums › Future of OSS in South Africa › Future Linux Desktop OS for SA Gov › Implementation suggestions
- This topic has 5 replies, 2 voices, and was last updated 3 years, 4 months ago by Danie van der Merwe.
12th Dec 2019 at 12:00 pm #2463
A topic to discuss ideas for implementation and adoption in SA Gov. We can also look here at what went wrong with the previous drive so as not to repeat the same mistakes.
- This topic was modified 3 years, 5 months ago by Danie van der Merwe.
12th Dec 2019 at 12:01 pm #2464
From Llewellyn Gush: I would like to propose that as part of this initiative that pressure be brought to bear on education and industry… Education .. as the current curriculum is appalling. Why is it necessary to specify the MS document suite for everything submitted to university? ?
Industry where they intersect with Gov . The DiData’s etc all develop solutions using proprietary software. It is very hard to find a solution that is OSS based
10th Jan 2020 at 6:02 am #2493Prince SebapuParticipant
I still believe in this motto “make it work, they will all want to use it”. I suggest after we agreed which community distro we are going with let’s say Ubuntu 18.04 LTS, then the following happens to ensure we are faster innovating a custom distro for public sector:
1. We need to create a google shared spreadsheet where we will all populate with current commercial software we know our departments are utilising, this will help us start thinking and considering replacing them later on with OSS alternatives, or be prepared to keep their underlying proprietary OSs (partial migration to OSS). We may need the spread sheet to have column for “OSS Alternative”
2. Everyone of us download that common agreed Ubuntu version and install it on a VM or straight on a physical machine.
3. Install agreed office suite and start testing >> share the results.
4. Start experimenting with installation of BAS, LOGIS, etc (common government systems). By the way I guess we may ask the developers to recompile these systems on Linux so we can have a Linux executable to
5. Start installing the OSS alternatives of other existing proprietary applications as loaded on the spreadsheet >> share the test results
6. Have presentations in our meetings on progress made.
7. Agree on at least a BETA that we can start presenting at various forums.
8. Present to the right people at SITA and Treasury with indication of the cost savings we will make for government.
9. Present to committees like the 4IR committee, GovTech and various other forums where we can be able to make noise and be heard.
- This reply was modified 3 years, 4 months ago by Prince Sebapu.
11th Jan 2020 at 10:11 am #2498
Agreed, we essentially need to take stock of all proprietary software running and suggest alternatives for testing that can be presented.
We must just agree on the aims: do we do server side first, desktop OS, or end user apps with a view to readying the OS for change. Windows 7 support coming to and end means Microsoft has gov in a never ending spiral of payment and change.
But yes this list will provide a checklist of what must be tackled. I’m just thinking also about how broad we tackle this: A product across all Depts, or do a specific Dept first and migrate them for all products.
Obsidian Systems has done at least one survey across Gov and may also have some good inputs regarding change.
11th Jan 2020 at 11:19 am #2500Prince SebapuParticipant
I suggest we start with the desktop side, for the following reasons:
1. We can still allow users to remote to those Windows based systems we cannot get rid off immediately, e.g. Pastel and AccPac, it can be installed on central Windows server and allow VNC to various users using Linux desktop.
2. It may not be easy to immediately migrate their server side as they have already invested in many expensive systems which need careful migration strategy, but desktop systems as per 1. above
3. We can realise immediate huge savings on dekstop licenses, Treasury may be happy because desktop licensing was identified as one of their huge costs per annum.
11th Jan 2020 at 12:12 pm #2502
Yes desktop is more visible and actually less impact.
Some server side changes may not be difficult eg.OpenText ECM currently running on Windows servers can run on Linux, so could be swapped out. But yes if an enterprise Linux (paid) is opted for there are extra costs.
We also need to give guidance to Depts for new installations and provide options. So many default to expensive options that vendors suggest, without being aware.
Making me realise how much work there probably is to be done. We really need a proper gov standing committee on OSS and to re-establish a proper FOSS project like was mandated in SITA years ago.
- You must be logged in to reply to this topic.