Nothing much surprises me about NHS IT after I worked for the Health and Social Care Information Centre (HSCIC) which was the name before they were re-branded (£££) as NHS Digital. My experience was they outsourced the system development / support to a large Indian company, then realised that due to security and data sensitivity that non of the non resident Indian employees would be able to gain any access to the live system. The systems was hacked up legacy code that had cost a lot of money to develop, and even more to support.
9 posts • joined 6 Jun 2017
The outfit where the NHS England Digital boss is headed? Turns out their code is 'not technically suitable' for the £6.4m NHS App
Time/Cost/Quality - pick two. Computer Says NO!
None of this surprises me, the result of outsourcing to the cheapest provider.
I'm a developer who worked at BA in 2014. Worked on their Java based orders data store, which was used internally to reduce costs as they did not want to pay the money to query the central Amadeus booking system.
Of course, Tata was brought in later to reduce cost (and quality) further.
Oh well - I hate BA because of the way their management sacked everyone - always fly with other airlines.
British Airways' latest Total Inability To Support Upwardness of Planes* caused by Amadeus system outage
From what I recall; BA take the Amadeus information for its flights, process and then keep an internal copy. Mostly this is (was) to reduce costs as a query to Amadeus was deemed as too expensive, and so they use their own (nasty) Orders Data Processing (ODS) system. Worked on it for about six months, until again, to reduce costs (and quality) they sacked and replaced most of the BA UK workers.
I'll expect this added to the delay as all the BA internal systems would have had to wait until ODS had processed the information after the Amadeus outage.
Without doubt BA had the worst management I've experienced in twenty five years working in IT and the military.
I worked (very hard) on this project [iApply/iDecide/iSearch] for one of the supplier companies.
Problems was always the Home Office, with their stupid ideas. For example someone decided that since the computers were running 24hrs each day, we should have a shift system in order to get more work done. Issue was a common one in software development, they did not know what they wanted and had even less of an idea how to achieve it. They also split the work between various large supplier companies, creating political problems as they were always looking for someone to blame!
I did hear that they binned two of the systems, which we had working, and went back to the CID system. We suspected it was more about unions and the HO staff wanting to keep the old system as this would mean a dependancy upon their legal skills. Keep themselves in well paid work.
Expect they will also kick off another versions of the good old PBS (Points Based System), (I also worked on that one) if they have not already!
Gladly I got out of government IT, and not looking to return!
Yeah, they [BT] always outbid their competitors, only to outsource the work to Indian companies who produced cheap but low quality software. Of course, it didn't take long for the Indian companies to then approach the customers and argree to cut out the expensive middle-man which BT became. Seen this during my time on a few of their projects.
This new "strategic thinking" will be more of the same, hope they finally wake up realiase that they need to add value and get back to implementing quality solutions, like they did in the days of Syntegra.
Re: Single point of failure?
Nothing that a bit of transaction handling and idempotent receivers would not solve. Again I have to say that I've worked on large systems that had persistent service buses. Sounds like they were not wanting to spend the extra money on faster servers, so took the option of using non-persistent messaging - sounds like BA all over.
In my brief time at BA, I saw SOAP messages using JMS internally, and externally we were connecting the JMS queues to MQ endpoints. We also had some Jini, old nasty technology from 1998. Never did see any service buses, maybe they existed downstream.
I worked briefly for British Airways on their Order Data Store as a contractor a few years back. A heavily hacked version of a SABRE product which was overly complex [imho]. Yes, SOA was the fashion, with everyone using SOAP messages but many of the different systems did not quite fit together. Noticed a serious lack of understanding of system resilience and many single points of failure. But they culture was to "save" money.
Staff were not valued, less so the contractors. Suspect the senior management thought they could just replace the British workforce with cheaper Indian staff, and save more money. A lot of knowledge would have been lost. Maybe that was (and is) the real problem, but you will ever hear that from BA.