Question 3: What are the emerging tools, standards and infrastructures?

I have been trying hard to absorb all the valuable insights provided in this thread.

I think everyone agrees that linked data / interoperability is the way / pathway to better globalised information sharing.

I appreciate also the fact that several tools - Agricdrupal, AgriOceanDspace, CMS etc..., have been mentioned, but I am really questioning myself on whether we are being democratic enough and whether we are helping the community ...

So in addition to what Krishan and Ajit have already mentioned ...

I would have been more keen to learn the other way round and focus on the features that need to be present for any information system, however small it may be, to play its role fully.

For instance, I want to run a library management system and I want to share my contents internationally, in this case, other than document/media management aspects (circulation, loans, barcoding etc..) of the system, on what criteria should I be chosing a particular system? What should there be in the tummy of this software, for it to be able to communicate outside and share what we have?

Other than features, let us not be overwhelmed with technology and overlook the human factor and processes involved. As long as a partcular system is secure, stable and has the desired features / components, why not use the easiest and most aesthetically pleasing ones.Why make life difficult!

Bottom line:

1. Yes we have certain tools out there that can work wonders. But what's in a name!

2. Our priority here is sharing information. The way - linked data. For that purpose, it is not singled out tools that we need.

3. We need to understand as a community the core of these tools, so that we have the freedom to choose the one we feel at ease with. This is how we will all start using systems that can communicate among themselves. Otherwise, we will end up frightening a large fraction who won't be on the same boat.

4. It would be helpful if we (especially the IT people) could come up with a list of features / standards required for particular systems to effectively communicate with each other.

5. This will be our buy in - Where we tell the community - Listen... if we want our systems to communicate, we need to use for XYZ system that has ABC features ... and here are a LIST of tools that have them. Have your pick and let us start!

Apologies if I went out of context!

(Johannes - I could not help smiling at this: Real world problems are resolved only by handling these technical details.

Bro... am not sure if nature works this way...)

I was quite intrigued by your  post , San_Jay, and some on this forum (at least Valeria, Johannes) know why. This post addresses the question at what level CIARD should be involved with tools, their selection, support and development.
I can see a  number of levels:
- List the features that certain tools should have to share your information with the rest of the world, as CIARD is advocating. It should for example be harvestable with the OAI-PMH protocol or produce sitemap files for indexing in search engines. This to me seems to be what you are advocating. One of the iterations of the "tools piece" that we are struggling with did exactly that. We were not sure  whether this was helpful enough for all: it helped setting up a shopping list but gave no information about brands and addresses of shops.
- A next iteration under development now attempts to list tools that have a number of these desired features. Such a list should come with more relevant faetures ranging from hard technical issues (operating system) to more soft issues (skill level required to run it) This approach will only be succesfull of the CIARD commmunity as a whole is willing to share it's experience to maintain all this.
- A next level of involvement coud be assembling a package or toolkit that fulfills the neds of most of the community. A sort of filled shopping bag with what many of us need anyway. FAO has assembled Agridrupal to support a number of partners. FAO contributes it as an option to the CIARD community but other CIARD partners follow different lines. Do we expect this level of support from CIARD?

I think it comes back to the question of standards. We can only achieve all the things we want if we follow certain standards.  As Johannes said standards are not set but accepted. Standards will not be accepted if there are no tools that handle them. For example: XML and XSLT would not have been accepted if Michael Kay would not have developed the Saxon processor, to get people started. Do we need such seminal technoogies for the standards that we are tring to get accepted?

I suggest we use question 4 of this discussion to discuss how CIARD should proceed with all this.

Hugo ... Intriguing? I wish I also knew why?

Thanks for commenting on the post. You got the picture right and the points (levels ...) brought forward are pertinent. The issue of standards is a priority.

We shall definitely catch up on Q4, but till then, I would also appreciate some reflection on below:

I believe it is not only about standards and features but also of 'freedom to choose'. This is very important for the community to join the wagon. If not, the process will either be limited or fail.

If an organisation has choice, then it becomes easier to SELECT tools that can INTEGRATE existing institutional infrastructure and / or cohabit with existing tools. It hardly bears financial costs and is more easily accepted by management (Let's not forget that we need the green light from our bosses before overhauling anything).

Again, if there is choice, it is most likely that the easiest system adapted to current needs will be selected - less staff training will be required and adaptation time will be less.

And so forth...

The community is made of thousands of institutions scattered around the globe. Most of them are under staffed and under budgetary constraints and CIARD's initiatives might definitely not be their priority of the day. Should they be willing to share information, which tool will they readily opt for or at least try: One selected out of many that meets and suits their current needs and situation or will it be singled out ones!

I think what Hugo is saying here is very interesting and gives a good answer to Sanjay's concerns, which have been often expressed also by other colleagues when talking about giving advice also on which tools to use for better information sharing.

In particular, when Hugo says that the list of "features" _may_ not be enough ("it helped settingup a shopping list but gave no information about brands and addresses of shops"), I think he is raising a point that is very important when you think of the decision making process: if you give a list of features to a technical person, this person may be able to evaluate several tools and identify those that have the necessary features, but managers who don't have technical staff or time to conduct this evaluation may just want to know which tools have the necessary features ("brands and addresses of shops") and they might look for a source of such information that has some consensus and "authority", and maybe CIARD at this point could represent that consensus, but this will happen only if the community sharing their experiences with tools is large enough and if there is some agreement on which "features" are most relevant for which needs. 

And one personal and questionable (and it will be questioned I'm sure) opinion on being completely tool-agnostic: the fact is that there _are_ differences between tools, and being too neutral can also be damaging. There are tools that have proven more suitable to information management and integration of standards than others, and more sustainable.

And tools aren't just "technology", they are the instrument through which you implement the standards and in the end you make your information accessible.

In a recent seminar, someone said in a presentation that "standards are like toothbrushes, everyone wants their own". And that's a really tricky situation. Most of us cannot afford building our own standards; and even if we ever do, we cannot deploy them in a sufficiently large scale to become accepted. Yet, we often find existing standards (if any) lacking or inadequate for our own purposes.

Since I work for a small research institute, we have elected to borrow somebody else's toothbrush. We agreed on some minimum requirements (XML, OAI-PMH) and then went out and looked at what is available out there that meets our needs. We were fortunate to find something that did (in our case EML - Ecological Metadata Language) but this came at a cost.

The protocol is general-purpose enough and the software is forgiving enough that you can deposit any of our data in it. But this means we have to give up powerful search and indexing capabilities and sophisticated ways to manipulate the data. This could only be done using robust semantic technology which we do not have. It is like buying pants that are a couple of sizes too big but missing the belt that will keep them up.

It is virtually impossible to translate knowledge from one ontology to another without running into problems. But we are too small to do such work. Fortunately a consortium has recently emerged (DataONE) to specifically look into this for the ecology space. Amongst what they do, they are now working on making their technology (the EML-based system we use in our institute) interoperable with other platforms.

I think this probably goes some way in delineating what can and cannot be done for research organisations and for consortia: like us, a bottom-up approach for the former based on immediate needs. For the latter, a top-down strategy focussed on interoperability (semantic and technological). 

As I have been following the forum it appears the emphasis remains sharing research information within the research community across the globe. Perhaps that is where most of the sharing is needed. But there is a question that I have been thinking of the last couple days, and that is to what end. My guess is that:

1. occassionally the sharing of information will lead to a completely innovative idea, that can then be researched in detail, etc.

2. More commonly it is simple a sharing of similar information that reinforces or fine tune already estabished innovations. As such if gives the scientists confidence in what they are doing. But since agriculture is always a very local science how critical this is may be a worthy question to address.

3. It also tend to be commodity specific or even sub-commodity specific looking at fertility, weed control, pest management separately and treating the research commodity as if it were the only commodity involved.

This is all well and good and is done in an ideal research environment with few if any limitations on resources needed to conduct the experiments. As mentioned before this gives excellent results as to the physical environment potential, but does not measure the drag on this potential that result from the limited resources the end user may have when attempting to apply the research to their specific fields, and integrate it into the rest of their farm enterprises, both crop and animal as appropritate.

My concern along with that of several other members is how the research information gets to the farmer end user, particularly the smallholders with the very limited resources to implement it and in the overall economic and administrative environment found in most developing countries. That is virtually no tax base to fund agriculture support services.

The need here is for more general information on all the activities needed to optimally producer a crop. This is normally done as strictly an educational effort for detailed recommendations and assuming that farmers can readily accept the research result once they understand it. If not it is the failure on their part to learn and they need to be repeately taught. But what happens to the research result when the basic crop establishment, by a hungry exhausted farmer who can only work 3 or 4 hours a day, takes 8 weeks instead of the anticipated 2 weeks? Will the rest of the detailed message still be valid, or will the plant population, weeding, fertilizer response be severely compromised? How does this drag get fed back into the research program so adjustment can be made? I wish the forum could spend a couple days address that issue.

Also, most smallholder are involved several crop and animal enterprise the management of which has to be integrated to obtain the maximum yield of all farm enterprises and priortize their crop program, usually giving priority to subsistence crops over cash crops. Again that can compromise the best research effort. But are there any return links so the researchers can better appreciate the constraints under which the farmers are operating and not simply blame them for failure to learn and appreciate the research results. The alternative is to keep pushing out solid research results that are beyond the means of smallholder farmers to utilize.

This still leaves the question of the means used to convey research information to the farmers. With most extension officers concnetrate on collaborating with development project as a means of getting some supplemental income, most of this is through the donor assisted NGO. They then try to operate this though farmer organizations and cooperatives, but this leaves the question is this best method, or dispite the social ideal is the accompaning business model just to cumbersome and inconveninet that the farmers are better off taking their business elsewhere. I fear the latter is the case and an issue that needs to be address in some forum, if not this one.

Thank you.

 Very good points related issues mentioned from the post of Nabeel Abu-Shriha. I have some limited experiences on participating the projects serving the practical end users (farmers and enterprenuers) with ICTs. there are very good project designed well but failed to generate a sustainable effect after the some local service center invested. it seems that the end users wanna some "fast food" that easily understood to act, and can bring them benefit in a see-able future, and better free of charge. other wise, they'd like to ask the real person.  

the PC, mobile, even CATV have been utilized as network terminals already, and there are certain orgs input region-oriented info, but is that enough? can we make make more valuable customized service and regarding them as knowledge farmer/agri-entreprenuer, as what we did to users in business world?

and finally, I do feel there are much more resources for research community comparing to those for those end users, is it really hard to meet their requirement on information based on the existed tools, standards and infrastructures? 

 I agree with what is written. LOD is the way to go. Content management systyems will play a leading role in mashing up different contents to get analyses of interest. There might be new tools and approaches that are in use and hence the more justification of our meeting. However , application of these tools and approaches will require capacity development especially in the South. 

• The models of social networks such as Face Book, Linked In, Twitter will determine issues of interoperability in a big way. User generated information is here to stay
• The concepts of semantic web design will also feature in a great way.
• Cloud computing is clearly the way to go.
• Translation will be a key issue to be addressed – countries like Africa where the languages are quite diverse will continue to pose a challenge. Thus the support of initiatives such as google translate are key
• Mobile technologies and platforms cannot be ignored