IntraHealth International – NGO Interview series
The 3rd interview in the NGO series, was with Mark Hershberger (MH), Jeff Strope (JS) from IntraHealth International. I’m really learning a lot from these and someone asked me the point in them. Well, a lot of NGOs are using Open Source where they can and a lot are using it 85% of the time. I’m trying to ascertain why not 100%, find the pitfalls of where oss lets them down and see if we the community can help them.
How I’ve gone about this is asking NGOs to take part and overall the reception to this has been great as they want to tell everyone about how great it is, and also where they are let down by. In the end I’m going to use these interviews to go to other NGOs who don’t use oss and show them that they use it just as easily as their current choice without much effort and there is a community of friendly people to help them when they need it.
Czajkowski Can you tell us more about your organisation and how you use Ubuntu?
JS : IntraHealth International is a (mostly) US-government funded non-profit tasked with (mostly) promoting better health care systems in the developing world. Systems is used in a broader sense, not a technical one. We work with governments and regional leadership to ensure an adequate number of health care talent exists and that these individuals are well trained and equipped.
MH : We’re in the process of wrapping up the Capacity 2 funding cycle (yesterday was the End Of Project thing at the World Bank) and beginning on the country-specific AAs (Associate Awards — follow-ons for Capacity). There should be additional funding soon… but all that is kinda sketchy atm. But most of our work right now is around health care capacity in developing nations. Training and tracking providers. Our Mission statement is here . But one thing we do is the iHRIS Suite. Which I can go into more detail on. Jeff will be able to tell you how we use Ubuntu internally. But when we deploy our open-source iHRIS Suite, we tend to use and develop on ubuntu. iHRIS is an open source human resources for health application that Mark and his colleagues developed. It is now in use in many regions in the world.
Big problem in developing countries like Rwanda, Uganda, Tanzania, etc, where we’ve been working, is that health care providers (doctors, nurses) are out there, but no one really knows where they are or all the information is in paper files literally kept in containers. Not the easiest thing to look through when you have an Hep G outbreak. What iHRIS does is gives the Ministry of Health or Nursing Council a way to keep track of all that information by putting it in a digital format. (Actually, while Internet is spotty, there is electricity and thus, computers, at the health care centers. )
“no one wanted to touch them they were so dirty” so, of course, first you have the problem that it is paper and hard to get to, then it is dirty and no one wants to touch it. Just digitalizing the records would be a boon, but Uganda is now rolling out iHRIS to each of its districts and we’ll hopefully be working with them to integrate it with OpenMRS, running on Ubuntu where possible.
Czajkowski can you perhaps tell me some more about the challenges you come up against?
MH : Challenge: finding people in-country to support the software. We have one really really great guy in Uganda (and others in Kenya, etc) who is really interested in OSS, but he says most of his peers are MS focused. and we’re trying to work with Unis in-country to develop educational programs (see Intrahealth Open intrahealth.org/open) here was a really good story in the Linux Journal maybe 10 years back that talked about training people in-country to use accounting software he realized that the software would be no good if no one could maintain it or the computers… so he backed up and taught them dbl entry book-keeping.
But I think the biggest problem is finding people on the ground who are willing/able to even learn to maintain the software…. (as far as software use goes) and getting the right people trained on using these methods (not all of which means the software). We have workshops, but the first few of those have been attended mostly by managers and not the people who would actually be doing the work. Hopefully that changes.
JS: Trying to apply Western technologies into the developing world is probably the largest challenge, generally speaking. That and getting Western people to think in terms of the technologies that do exist in the developing world. We recently had meetings with a number of sister NGOs that have similar missions and financial tracking was the #1 issue for all of them.
Czajkowsk So what made you start this?
MH: Then USAID first started saying “we want to build capacity” and they were looking around for some HRIS software intrah started talking to consultancies to get a price for it the consultancy said something like $500k for the initial development + $100k/country/year after that which, when you’re talking to developing countries… is just ridiculous.So they had a developer at the time, Luke Duncan, and he and Dykki started working on the first version of the software think that was for Rwanda and Kenya saw it and re-implemented in Java .
Czajkowski what made you chose Ubuntu/OSS ?
I think it was just because back in 2005 or so that was what Luke was using. That Canonical was behind it and pushing its LTS helped. so, time goes by… by the time they’re working on version 3, I come along and Carl L. comes along. Carl really pushed the software forward made a really complete framework. (I was mostly working on other OSS projects) the MoH in Uganda uses Suse on their servers (maybe even the iHRIS ones) because they bought it pre-installed, but mostly we’re talking about Ubuntu. (When I went to Rwanda, I took a USB HD with a mirror of the Ubuntu repo on it so that made it easier)
Czajkowski Do you use any proprietary software?
JS: We’re still a primarily Windows shop. We have 500-ish Windows workstations + an AD/Exchange environment.
MH: There is a small component inside iHRIS that isn’t strictly open source, but that is replacable — and I hope to replace it 😉
Czajkowski you say 500 are Microsoft Workstations, that’s a lot, how many are running Ubuntu so ?
JS: A dozen or so. Servers are probably 50/50 ms/linux.
MH: While we tried to get internal HR to use iHRIS, we’ve run into an HR problem. Hard to get them to want to eat their own dog food enough to pay for additional devs….
Czajkowski do you see that as an obstacle, your own people not using OSS , how do you expect others to?
JS: We do use OSS and I don’t think we “expect” others to use it either. We expect (or hope) people to use the most appropriate tool for the job.
MH: The concept of OSS is good, people get it, but its hard when they already have MS installed and are used to it. Hard when most people aren’t technical and don’t get all the “political” arguments for OSS. Remember 500 PCs … most of which are used by people with medical, not technical, backgrounds unless I’m wrong about the backgrounds, jstrope?)
JS: Not wrong at all. Few of the people here are technical.
MH: Other projects, like OpenMRS, have similar problems. OpenMRS depends on MS’s proprietary form software. Users at one site in Uganda said “we could run everything on Ubuntu and save lots on MS licenses if it weren’t for that requirement in OpenMRS”
Czajkowski what can Ubuntu do to help you?
JS: Laura — you hit the nail on the head there about training. Why would IntraHealth want to save $50,000 on licensing when it would cost us $1,000,000 in retraining to do so? That’s really why Ubuntu should focus less on desktop and more on web-based services. That’s how Microsoft is going to lose its grip. Because my MS users here might not be comfortable in a Gnome desktop, but they are (for the most part) very comfortable in a browser. It would be very, very nice to see better AD integration. AD stinks, but it’s not going anywhere. Focus on smoothing out the server and that will really help desktop adoption
MH: I would like to know what is needed to get some things into Ubuntu proper. what do I need to do to contribute things I’ve packaged? Debian Med is a micro-distro, for example, focused on Medical stuff that I’ve thought about and may still contribute my iHRIS packages to. Is there a possibility of something similar for Ubuntu? (would they be interested in the packages) JS is working on some Asterisk stuff… getting that easier to deploy on the server side would be good.
I see Canonical/Ubuntu very focused on the desktop and doing great things there, but the server side could use the same attention. Many things (e.g. AD integration, Apache single-sign-on) could be really, really simplified. And the AD stuff could really, really easily be automated and packaged w/o relying on any proprietary stuff. Make it simple to join a domain, etc. Make it possible, on the server side, to set up SMB and LDAP and Kerberos to work with or replace AD. I think this is possible now, but takes a lot of fiddling. make it fiddle-free. Did I mention I’d like to see a lot more love for the server side? I mean, it helps if the people we train in Uganda, Rwanda, Tanzania, etc don’t have to deal with so much… Debian has done a lot of good there, but Ubuntu could take it so much farther.
So I’d like to thank once again Mark and Jeff for taking part in the evening of interviews. I’m learning a lot during these meetings, and also being able to share with you and others in the community aspects in which I hope we can address, already I’ve passed on some of the server comments to people so lets hope we can help them in some way.