TSB meltdown...

[vetran]

Just because someone is more expensive doesn't necessarily mean they're better. The reason why Henry Ford was so successful was because his cars were cheaper than everyone else's.

Obviously you need to outsource to a company who can do the job, and no large company simply gives an outsourcing contract to the lowest bidder, cost is simply one factor.

Offshoring in India began in the 1960's if it was a failure companies would have long since given up on it. Most major Software companies have development centres in India including Google.

Actually quality was a big part of Ford's success. The fact his cars were affordable AND reliable drove success. The reason his products were cheap were driven by efficiencies not poor service.

As we can see here TSB has outsourced to a company that can't do the job and they failed to manage it correctly.
Combining that with it being core functionality is a real issue. Its entirely reasonable to conclude cost was a significant factor.


Amazon outsource their delivery but keep it so well managed because it is a core competency.

IT is about all public banks do now. There are few branches and all the products are identical.

http://www.evancarmichael.com/librar...s-Factors.html

[BlasterBates]

There are tariffs on goods, but not on services

There should have been both and very high - directly depending on things like taxes in any given country, social security offered - the lower those are - the higher tariffs should be, now that's fair and sensible

That would be almost impossible to achieve without some multi-lateral deal. No-one cares about protecting highly paid IT professionals.

Tariffs effectively push up costs for everyone else, which is fair price to pay to protect workers in industries and regions where they couldn't easily get jobs.

[BlasterBates]

Actually quality was a big part of Ford's success. The fact his cars were affordable AND reliable drove success.

As we can see here TSB has outsourced to a company that can't do the job and they failed to manage it correctly.
Combining that with it being core functionality is a real issue. Its entirely reasonable to conclude cost was a significant factor.


Amazon outsource their delivery but keep it so well managed because it is a core competency.

IT is about all public banks do now. There are few branches and all the products are identical.

How He Moved The World: Ford's Success Factors by Henry Ford

No-one disputes all that, but what this isn't is proof that outsourcing doesn't work, it's simply a failed project.

I remember hearing a statistic 20 years ago that 70% of projects failed. Recently a bridge collapsed in the US, it was simply because it was a sh*te project.

[AtW]

That would be almost impossible to achieve without some multi-lateral deal. No-one cares about protecting highly paid IT professionals.

Tariffs effectively push up costs for everyone else, which is fair price to pay to protect workers in industries and regions where they couldn't easily get jobs.

Prices would not have gone up if it was done from the beginning - there would also be incentive to build quality stuff that lasts, not cheap mass produced tulip that later pollutes the planet

WTO will fall apart and future will be run by a few major trade blocks - EU, North America, a few pacific countries - crazy time to Brexit

[Mordac]

No-one disputes all that, but what this isn't is proof that outsourcing doesn't work, it's simply a failed project.

This. I've worked on many projects that would have been judged failures (usually due to very late delivery) because someone, usually the client, didn't understand what they were asking for, or didn't understand the full implications.
Outsourcing companies seem to have a habit of saying "yes, we can do that" and then finding out a year down the line that they really can't, because the scope has expanded while no-one was watching.

[AtW]

So clients can't wcope and outsourcing can't mitigate it effectively: it follows outsourcing is a failed strategy...

[vetran]

So clients can't wcope and outsourcing can't mitigate it effectively: it follows outsourcing is a failed strategy...

nope outsourcing based on lowest price and no maintenance of quality is a failed strategy.

Its a failed project because the people running the project did not know what they are doing. The ones implementing it had little investment in the outcome.

[BR14]

nope outsourcing based on lowest price and no maintenance of quality is a failed strategy.

Its a failed project because the people running the project did not know what they are doing. The ones implementing it had little investment in the outcome.

Wot? - like, contractors??

[fool]

This isn’t (just) the PM, it’s those who probably bought into the agile/DevOps/cloud ideas without truly understanding the idea of quality to the the left.

The Devs who love the idea of CI/CD without the microservice architecture to go with it (you can only do fix forward if you’ve changed a tiny part of the code), and think that discipline to test properly stops them from deploying quickly are probably more responsible.

And the management who are too weak to control the Devs.

Feckwits, like I said.

This doesn't really make sense.

CI/CD and releasing small changesets is just a way to lower your risk. The idea is pretty simple. Having the automation gives you the ability to actually test your changes, and have a historical record of what you've done, proven against manual interference so you can go forward in confidence that what you've tested for is actually passing.

This does not require you to have a microservice architecture. But it does require you to have a reasonable pipeline combined with a reasonble test suite and do a whole bunch of the hard stuff that you should have been doing already.

Sure, there's a problem in the industry with devs who just want to fire and forget. But it has more to do with a lack of desire to retain and pay for experts and instead a lot to do with the abbundance of non-technical or junior people being the sheppards of highly technical system, navigating through worlds of buzzword bingo, pretending that this time they can successfully plaster over complexity with bling, despite it never working in the past.

Maybe though, if TSBs head of infrastructure (and presumbly their head of dev) actually had a background in running large infrastructures and and system, maybe they wouldn't have failed at doing the job (or at least on babysitting the consultancy bods that failed to do it).

Either way though, being down so long suggests nobody really understands how the system works (and also has the power to change it). We can dress is up in whatever type of failure we want (we usually just blame it on something that challenges our niche), but it still probably usually comes down to treating people like replacable cogs, leading to a loss of institutional knowledge, leading to an IT disaster.

[darmstadt]

Seeing as so many people on here are experts and know exactly what to do in this situation, then this is for them: https://www.linkedin.com/jobs/view/633114556/