Some of the benefits of SIF are listed here:
Some of the benefits of SIF are listed here:
Mike, thanks for the acknowledgement! If we have concrete walls around SIMS that is just as it needs to be â€“ we donâ€™t want any accusations that pupil data isnâ€™t safe! SIMS was started 21 years ago but the current design bears little relationship to what we have today! You can get the data that you want to do â€śin year variationsâ€ť by using BOs. I donâ€™t think SIF or a report generator could get to the data you need.
Olcinm, I can see why you think that we are sending out mixed messages by using a modified SIF in our partnership exchange. We are using a bastardised version of SIF as a data transport mechanism for a limited range of data between schools that share pupils. If the ZIS annual pupil charges remain as we have seen then we may indeed move away from SIF. Our implementation would not get a SIF approved tick. The IMS standard doesnâ€™t include assessment or attendance data so it isnâ€™t suitable.
We have at least 100 partners (who make profits) that link to SIMS via the run-time report generation or using Business Objects. This gives them full access to the data in SIMS and in the BO case they can write data back as well. SIF would not replace most of these partners needs because they want to get to the data that SIMS stores. Take for example document exchange, SIF doesnâ€™t support it and even if it did broadcasting documents around the network would not be very efficient.
Interoperability is not easy. If SIF is funded it might just work but it will require massive changes in all MIS to make it happen. Then the question will be whether to take things to the lowest common denominator or not.
In my view small players will not be able to afford to produce SIF compliant systems.
Hi again Phil
Thanks that's really helpful.
If SIF only provides for large scale interoperability with major vendor applications and does not provide a cost effective route for innovative small developers then I think it's the wrong route. We need an interoperability solution that allows for innovation.
If IMS cannot provide interoperability of assessment and attendance data then it's not suitable either. I thought the IMS specification Capita presented to Becta in January 2004 did provide for some assessment transfer. So IMS can do assessment transfer IF vendors agree and develop the data model and write software to the model. Same as the SIF issue.
Are you sure the SIF work that's being done isn't because Capita see SIF transport standards as a useful technology and that you have to provide for other MIS systems and those system developers want to write SIF agents not IMS connectors?
Moodle (probably the most common VLE in UK schools and colleges) as people probably know has the beginnings of an IMS interface. This in core this is file based though the MINTED extension project has it working with a web service. There is a also a SIF agent though this is "old" and I think costs because its only given "free" with a ZIS bought via CPSI in the USA.
If you're just after LAN based Interoperability LDAP may be your route especially if you decide to provision your LAN directory from SIMS automagically with your own schema. Moodle does LDAP for authentication and user/course creation quite nicely I believe.
Where are other VLE developments in this regard? I guess those linked with MIS providers RM Kaleidos and Serco Facility have tight integration there but what connectors are they using?
As said before I thought UniServity and NetMedia were going down the SIF route for their data model anyway even if they don't use the SIF ZIS transport method. I also thought Pearson were SIF implementors and that RM were using a third party to help with SIF. I also thought Fronter was going down the IMS route and that RM had done significant IMS work for GLOW.
Phil Neal who knows better than me says RM, UniServity, Fronter, Viglen (ITS Learning) are working on IMS.
Confused or what?
Come on Becta. Make a decision and get Gordon to fund it!
And please don't forget that we also have RBCs working towards being Regional IDPs ... we have to start thinking about migration of data over the NEN based on identity provision too (just thought I would throw that in since LDAP has been mentioned).
Code to export from SIMS below, you will need to open in vis basic and reference the Microsoft DTS data pumps (3 modules) then compile, this does year3 to year3.csv
Change paths and tags accordingly, works with SQL2005 only, change db name from SIMS2005 to SIMS for older (unsupported) versions of sims.net
I bet this upsets some people eh?
CODE IS HERE: http://brains.primaryt.co.uk/simsexport.txt
on a side note from the code I just posted, we are moving over to shib this year so we will stop working on our sims.net data extraction tools.... unless we are not happy with the single sign on services supplied by the LEAS or regional consortiums..
Shibboleth is only going to get you trusted federated authentication not account creation and what about the rest of data interoperability like assessment, attendance etc. How will you or your LA fling that about? The thread says data extraction so I guess we're not talking about data going back into the MIS but don't we really want data unification with several authoritative sources etc.
What experiences does anyone have of either BizTalk or Identity LifeCycle Manager (ILM) from Microsoft? Are these the ways in which you will get the accounts or are you so small and uncomplicated in Bradford that you can have a big Active Directory forest from school LAN up to LA? Some LAs have huge numbers of schools linked by varying connections and with a myriad of MIS systems and LAN directory systems. Secondary schools are less likely to want to lose directory control on their LAN and the structure of say a AD with 500+ schools carrying the attributes necessary for Shibboleth authorisation into lots of different services could be really difficult to manage.
What an LA has to do is find a solution that works with several different systems. Its fine for school IT staff to get solutions working for them but kids move in and out of schools and colleges, staff can work in more than one school, SEN provision has special demands on data, parents can have kids at more than one school. Identities need to work inside several LANS and on the Internet and need to be moved between lots of partners. Yes this is scary and there are all the scare stories about centralised government IT data secuiryt issues. Who owns the data? The MIS provider, the DCSF, the school, the Network Manager, the teacher. Anyone asking the parents and children?
If we can help kids, parents and teachers from constantly filling in on-line and paper forms but instead data checking and updating on-line then great. But of course we have to get the security right. The responsibility for the data being verified shifts and the validation becomes simpler (less often). The issue are always going to be around money and people. And the biggie is who gets the blame when something goes wrong.... Who in school is currently responsible for data security? Are they really in danger of losing their job if the data gets in the wrong hands or is lost because it's not been backed up properly. Is this a shared responsibility, do we really have this all written down, explained, discussed and signed up to?
If the LA is going to take on a more centralised role for data (identity, content and other data) then the balance of responsibility shifts and heads will role if it goers wrong. Wouldn't many of rather it was someone else's problem? As long as the identity doesn't stray into the financial and stays in the educational then we may be OK but whatif the system starts to include school dinner money or trip or uniform payments? Would you want the educational identity connected to the financial one? This is the way we are heading and I for one would not want to be the bod in a school looking after that lot.
Last edited by olcinm; 17th February 2008 at 03:43 PM.
I joined this debate a bit late and only skimmed through the last page.
If we are to go with the title of this tread, "Extracting data from an MIS", one needs to thinks of 'open sytems' and the UK government own main trust in this area - namely e-GIF.
Some of you with other industry experience may know about e-GIF (electronic Government interoepratibility framework). e-GIF explicitly defines 'open systems' demands open access for data freely and standards for exchange of data between products, suppliers and systems. This goes back 1999/2000.
http://en.wikipedia.org/wiki/E-GIF and http://www.govtalk.gov.uk/schemasstandards/egif.asp
Yet, sadly DCSF and BECTA appear to be totally insular implementing very basic of e-GIF. Does it not make you wonder why? Lack of true 'free-market'? By all means put forward your theories.
e-GIF is backed up with full legistlation and hence there is basis for DCSF /Becta to demand from all suppliers MIS and other to comply to e-GIF's open systems policy or not allow to be allowed to access business/money from public funded programmes!
It may that schools are not making enough noise but this threead had 6,249 views - so it must be interest to a lot of schools/EduGEEK fans.
The purpose behind SIF is for DCSF suck data from schools rather than anything else. It far from providing anything like an 'open systems' that schools should have and I think this thread is discussing! Correct?
No school MIS supplier could survive in the market if it didn't allow 'open system' policy. UK importa so many US ideas/practises - sadly this wasn't one.
Last edited by Tiger; 17th February 2008 at 06:32 PM. Reason: Corrected number of views to 6,249 & e-GIF links
http://www.clip.local.gov.uk/lgv/aio/41579 To quote it
A snippet from the document linked above.
Q; What about the e-GIF? I thought that was the UK standard.
A; In short, the DfES Common Basic Data Set of school data objects is e-GIF compliant and the SIF UK data model will be based on the CBDS. Furthermore the SIF specification is based on the same internationally accepted XML standards as used by e-GIF
http://220.127.116.11/gray_knowlton/a...ic-change.aspx ) I guess Capita will also have to embrace this. For example supporting more document formats and office systems than simply Microsoft Office.
Interestingly it looks like SIF and SCORM may be getting together too as SIFA and ADL have announced a partnership.
Capita has no concerns about losing market share through interoperability standards such as SIF. We have over 100 third parties that link to SIMS and compete with us already. If anything the cost of SIF to suppliers will cut out competitors!
The issues that we have with SIF are:
- the UK data model is not yet fully defined and will require considerable compromise, certainly on our part. We are working with SIFA UK to help refine the standards.
- the agent development costs and more importantly the infrastructure costs will have to be financed
- there are quite a number of processes that we think that SIF will not deal with such as primary to secondary transfer. (This is because of dual ownership of data items which is not permitted in SIF)
- the use of SIF needs to be mandated
- in the UK schools own the data not the LA. This raises data protection issues as a SIF agent has to respond to an annoyomous request for data. We would like a legal opinion on this point
Last edited by PhilNeal; 25th February 2008 at 07:53 AM.
These are important points that I'm sure Capita and others must have raised with SIFA UK. Totally agree about the mandate but I'm not totally clear on the financing needed. If the DCSF and Becta mandate SIF for Learning Platforms, MIS systems and census returns presumably by absorbing the data model into e-GIF and aligning it with CBDS then won't suppliers have to finance developments or get out the market?
Ownership of data as data controllers is very important but lets not for a second imagine that schools already conform. The reality of what already happens with data taken out of SIMS is far from kosher. I'm glad Capita are looking into the legalities of this. But isn't this in reality a bit like closing the door after the horse has bolted? There would need to be legal agreements and a framework for sharing data. Isn't this what MIAP are doing?
If the data can be exported easily into other MIS systems doesn't it make it easier for a school to change MIS suppliers and doesn't it mean that functionality in the MIS may become functionality in the Learning Platform instead? Doesn't this make Capita a bit nervous? Why do Capita muddy the water with things like curriculum and lesson planner. Surely that's a Learning Platform function?
Anyway lets see what happens this year shall we. Anyone want to open a book on a DCSF announcement about SIF in 2008?
Last edited by olcinm; 10th March 2008 at 03:40 PM.
The infrastructure costs arise from the LA or RBC having to set up and more importantly maintain the ZIF servers. The model being proposed is that the servers will be maintained at the LA or RBC and that will require staffing that does not exist at present.
We estimate that to modify SIMS to deal with SIF will cost >Ł1m and at least one other major supplier has come to a similar conclusion. Whilst we can spread our costs over our user base and it will amount to a little extra for every school our competitors are not able to do the same.
We have indeed raised our concerns over SIF with DCSF and Becta. We all want a solution to these issues but we must choose a model that will deliver the benefits.
Phil going back to your previous post - I find I cannot agree at all with your sentiments on SIF!
You complain that the SIF UK data model is not yet fully defined - but it wouldn't be - its work in progress. How can we introduce a new standard without any discussion or work?
You say that the agent development costs and infrastructure costs must be financed - of course they must be - but it looks like you've been asking your plumber for an estimate.
You think that primary to secondary transfer will not be dealt with by SIF, but surely SIF is ideally placed to do this kind of transfer. I would think that the process would be less fraught with difficulties using SIF than it is now.
I agree with olcinm about whether the DCSF /BECTA will mandate SIF early this year - where can I place my bets that it will happen?
How can there be anonymous requests for data in SIF? The parties to the SIF arrangement agree in advance to exchange data and it is not open to any Tom Dick or Harry to turn up and request data. There is nothing in principle why SIF cannot ensure auhtenticated and secure exchange of data!
Just to recount Capita has supported the proof of concept in Birmingham and between Northern Ireland and Durham University. We have given up a lot of time to working on the data model with Becta. Our concerns are from very direct experience. If we exchange data between SIMS systems SIF will lose lots of data - its a problem!
There are currently 1 users browsing this thread. (0 members and 1 guests)