End of traditional IT departments, more white collar job losses on the way

[spyguy]

4 minutes ago, Futuroid said:

Yeah, but hardly anyone actually writes software anymore in a corporate setting.

They basically configure someone else's (SAP, Oracle, Salesforce, etc).

Wooh there.

Those are pre-rolled, corporate tools. They need a lot of setting up.

Now most compnaies rely on a lot of custom built software.

Not because their people inist on re-inventing the wheel. Most because most business is software.

If you could but the software off the shelf, then so could another company. Where the competitive advantage?

 

[Futuroid]

1 minute ago, spyguy said:

Wooh there.

Those are pre-rolled, corporate tools. They need a lot of setting up.

Now most compnaies rely on a lot of custom built software.

Not because their people inist on re-inventing the wheel. Most because most business is software.

If you could but the software off the shelf, then so could another company. Where the competitive advantage?

 

Only the real hard core are building custom software from the ground up. Even fairly specialist applications like commodities traders are moving to packaged software. 

The competitive advantage is in the art of joining it all up. Or applying existing solutions to a new application - for instance the use of barcodes and GPS tracking tech in the logistics sector. 

I build internet facing systems - if I see anyone with a C compiler I have them handcuffed and escorted out of the building. The potential for security flaws goes up by a huge factor the closer to the "metal" you get. 

[eek]

10 minutes ago, Futuroid said:

Yeah, but hardly anyone actually writes software anymore in a corporate setting.

They basically configure someone else's (SAP, Oracle, Salesforce, etc).

There is a lot of money in configuring those software packages - probably far more than writing custom code as so much is already built in.

The fun bits are when you have people who don't know the system asking stuff.

Can you add a new field (stuff that used to take days in old websites) - 30 seconds later there you go.

Can you change the logic here (stuff that would take say 1/2 day) - yep that's still half a days work. 

 

Oh and if anyone knows microsoft products and wants the next bandwagon - Dynamics 365 and powerapps.

[spyguy]

2 minutes ago, Futuroid said:

Only the real hard core are building custom software from the ground up. Even fairly specialist applications like commodities traders are moving to packaged software. 

The competitive advantage is in the art of joining it all up. Or applying existing solutions to a new application - for instance the use of barcodes and GPS tracking tech in the logistics sector. 

I build internet facing systems - if I see anyone with a C compiler I have them handcuffed and escorted out of the building. The potential for security flaws goes up by a huge factor the closer to the "metal" you get. 

Not digging. but what do you sue instead of C for those sort of systems.

Ill give you my answer after Ive heard yours.

 

[Futuroid]

3 minutes ago, spyguy said:

Not digging. but what do you sue instead of C for those sort of systems.

Ill give you my answer after Ive heard yours.

 

It depends. For bigger things it will Ruby on Rails running against PostgreSQL.

Some smaller projects it is a LAMP stack with a PHP framework and Nginx in front of Apache.

I find the big issue with website performance is rarely the server (unless the load is extreme). More often than not it's the structure of the page delivered to the browser and the size of images, Javascript, etc. Using CDNs / Edge networks, compressing Javascript, inlining things, etc will get the end user more benefit than saving a few CPU cycles on a server on the other side of the world.

Also, hardware power is so cheap now vs brain power. 

[spyguy]

8 minutes ago, Futuroid said:

It depends. For bigger things it will Ruby on Rails running against PostgreSQL.

Some smaller projects it is a LAMP stack with a PHP framework and Nginx in front of Apache.

I find the big issue with website performance is rarely the server (unless the load is extreme). More often than not it's the structure of the page delivered to the browser and the size of images, Javascript, etc. Using CDNs / Edge networks, compressing Javascript, inlining things, etc will get the end user more benefit than saving a few CPU cycles on a server on the other side of the world.

Also, hardware power is so cheap now vs brain power. 

Ruby? really?

PHP is terrible. Junk.

The above are all front -end stuff.

These are not robust, redundant systems that scale across multiple CPUs.

Erlang/OTP for all performance systems.

If you want web framework stuff then Elixir.

[Scunnered]

14 hours ago, dgul said:

It's fine.  They can all be employed keeping the HPC website up.

Maybe the HPC website will end up being the driver of the entire UK economy.  How ironic.

[Futuroid]

7 minutes ago, spyguy said:

Ruby? really?

PHP is terrible. Junk.

The above are all front -end stuff.

These are not robust, redundant systems that scale across multiple CPUs.

Erlang/OTP for all performance systems.

If you want web framework stuff then Elixir.

PHP is like Javascript. It can be horrible, it can be elegant. But it is performant and extremely well supported/understood. 

Facebook runs on PHP (OK, they translate the PHP into C++ then compile it now).

Between 50-70% of the cost of software is maintanence. Ruby is easy to maintain and today is an easy sell to clients. Java was in the past.

[Funn3r]

20 minutes ago, spyguy said:

 

PHP is terrible. Junk.

 

PHP is OK for amateurs on piddly little sites that nobody cares about. facebook.com was built using it for example

[Mikhail Liebenstein]

<duplicate>

Edited October 14, 2016 by Mikhail Liebenstein
<duplicate>

[Mikhail Liebenstein]

<duplicate>

Edited October 14, 2016 by Mikhail Liebenstein
<duplicate>

[Mikhail Liebenstein]

<duplicate>

Edited October 14, 2016 by Mikhail Liebenstein
Duplicate

[Mikhail Liebenstein]

3 hours ago, Greg Bowman said:

 

You will see Office 365 stickers in  all of those terribly depressing high st repair PC shops normally incorporating a fix your mobile phone counters as well. These people aren't industrial scale commercial technologists like us they are break fix monkeys hanging on and are quite happy to earn 2% for  selling office 365 without a a a partner agreement. Microsoft has completely volte faced on Office 365 which it tried to sell direct and now sells through the channel for a healthy 20-25% margin which is ok for a light touch recurring revenue product similar to tele comms. Although above about 50 users your better off with  a premise based system for cost and flexibility.

 

 

Yep agree with you on this. I work for a fairly large firm, and we have a global presence, and so IT thought Office 365 made sense, especially as we do have partnerships with MS.

But so far I have not been impressed.  Office 365 started out OK, but it seems to be grinding to a halt - it isn't my device as I have tested different devices with differnet OSes etc and different networks.  Something somewhere in O365 doesn't scale very week.

On the flip side, as decent alternative to crappy Yahoo, I am happy to fork out for Office 365 for home use.

 

 

[FreeFall]

1 hour ago, spyguy said:

Wooh there.

Those are pre-rolled, corporate tools. They need a lot of setting up.

Now most compnaies rely on a lot of custom built software.

Not because their people inist on re-inventing the wheel. Most because most business is software.

If you could but the software off the shelf, then so could another company. Where the competitive advantage?

 

I can't talk for most companies, but I work in IT for a global multi-billion enterprise, and I can safely state that we develop very little bespoke software - in the area I work it's only considered as a last resort.  COTS is the primary route, commodity is great if you can use it.  The core principle is that there's very little competitive advantage to be had from a specific software - and where it exists it's temporal, generally pretty fleeting, and significantly outweighed by other factors like business process, strategy, analytical insight, or just sheer luck.  I suspect that unless you're a tech company, tech has very little to do with competitive advantage - it's too easily replicated / bettered / disrupted.

[Mikhail Liebenstein]

1 hour ago, eek said:

There is a lot of money in configuring those software packages - probably far more than writing custom code as so much is already built in.

The fun bits are when you have people who don't know the system asking stuff.

Can you add a new field (stuff that used to take days in old websites) - 30 seconds later there you go.

Can you change the logic here (stuff that would take say 1/2 day) - yep that's still half a days work. 

 

Oh and if anyone knows microsoft products and wants the next bandwagon - Dynamics 365 and powerapps.

 

Ah you need to act more like the old EDS, Logica and DEC days.

Sharp intake of breath, suck teeth, "that'll be £50K for the change request, another £200k for implementation and testing."

I did once here an old EDS story about the RAF and some MoD project manager requesting that they change the name field on an Old VMS System because they had an airman who have 25 middle names.   Apparently there was a sharp intake of breath and a huge quote was produced.

[spyguy]

31 minutes ago, FreeFall said:

I can't talk for most companies, but I work in IT for a global multi-billion enterprise, and I can safely state that we develop very little bespoke software - in the area I work it's only considered as a last resort.  COTS is the primary route, commodity is great if you can use it.  The core principle is that there's very little competitive advantage to be had from a specific software - and where it exists it's temporal, generally pretty fleeting, and significantly outweighed by other factors like business process, strategy, analytical insight, or just sheer luck.  I suspect that unless you're a tech company, tech has very little to do with competitive advantage - it's too easily replicated / bettered / disrupted.

I guess it depends on the type of business.

Any org where digitised data forms a core of what the company does - be it logisitcs, trading, banks - tends to go for for custom s/w products.

On the  subject of banks - banks are just software ledgers with some people and branches at the edge.

[JoeDavola]

3 hours ago, Futuroid said:

Yeah, but hardly anyone actually writes software anymore in a corporate setting.

They basically configure someone else's (SAP, Oracle, Salesforce, etc).

I work with one of these big 'configurable' products for a large organization. It was sold to them with the idea that they'd be able to get rid of their developers over time.

That hasn't been the case. It's still a great thing to have in-house developers who can build good working relationships with business users and deliver bits of software that do exactly what they need. I'm pretty much betting my job that this will continue to be the case, despite the consultants trying to push products that can be 'configured to do anything', but in actual fact just cost a fortune in license and consultancy fees.

But I think it's important to have developers who can actually work with people, who have good analytical consultancy-type skills themselves, as opposed to the idea of a developer as a 'code monkey'.

Edited October 14, 2016 by JoeDavola

[wsn03]

"The Cloud" .... what a load of pretentious sh1t...I'd like to punch the inventor of that term squarely in the face. As someone rightly said on here its just your stuff on someone elses server (and now open to hackers). It's such ******, I've been hearing about the new and clever ways to package and bundle IT blah blah blah my whole career, its still the same sh1t everywhere - access to information. The technology changes, but the concept doesn't, and the same people are still in the industry - their job titles change but little else.

And journalists are people paid to write stories, most of them simply repeat whatever crap they are told by whoever they interview - the reason why for instance people think house prices can only go up. What a phoney nonsense of a world we live in.

[BristolBuyer]

The customers of the IT consultancy I work for are still stampeding towards AWS, so that's what I'm learning about at present. On the other hand, if the customers modifiy their bespoke applications to use the AWS-specific APIs they will be quite tied-in to AWS. I suspect that moving to a different cloud provider at a later date could be expensive.

As a customer-facing application developer/support person, I'm sometimes surprised at how much my employer's in-house hosting services cost, and how long it takes to get the in-house Windows/Unix/DBA administrators to fix things when they break, so maybe the cloud will help there, but only time will tell.

Edited October 14, 2016 by BristolBuyer

[Fairyland]

One of our meetings got cancelled and we talking about this article with global peers. It seems similar news are running globally. An Indian colleague pointed us to a slowing down IT sector - Infy, TCS nos show IT industry is dying slow death   

Their government has started cutting down rates but official inflation figures do not reflect real life, he said.  "Welcome to our world" I thought. 

 

 

Edited October 14, 2016 by Fairyland

[CentrinoDuo]

1 hour ago, wsn03 said:

"The Cloud" .... what a load of pretentious sh1t...I'd like to punch the inventor of that term squarely in the face. As someone rightly said on here its just your stuff on someone elses server (and now open to hackers). It's such ******, I've been hearing about the new and clever ways to package and bundle IT blah blah blah my whole career, its still the same sh1t everywhere - access to information. The technology changes, but the concept doesn't, and the same people are still in the industry - their job titles change but little else.

And journalists are people paid to write stories, most of them simply repeat whatever crap they are told by whoever they interview - the reason why for instance people think house prices can only go up. What a phoney nonsense of a world we live in.

Yea what a crock sh*t that Cloud is. No one would be foolish enough to use it. It's so crap that it's on track to generate some $10bn for Amazon this year. 

[Futuroid]

2 hours ago, JoeDavola said:

I work with one of these big 'configurable' products for a large organization. It was sold to them with the idea that they'd be able to get rid of their developers over time.

That hasn't been the case. It's still a great thing to have in-house developers who can build good working relationships with business users and deliver bits of software that do exactly what they need. I'm pretty much betting my job that this will continue to be the case, despite the consultants trying to push products that can be 'configured to do anything', but in actual fact just cost a fortune in license and consultancy fees.

But I think it's important to have developers who can actually work with people, who have good analytical consultancy-type skills themselves, as opposed to the idea of a developer as a 'code monkey'.

TBH some of them (SAP in particular) are so complex that configuring them IS development.

You can also write your own code to customise things, I seem to remember SAP has "customer exits" - essentially hooks you can use to call your own code at specific points*.

It makes complete sense to start with a platform that supports the basic functions mosts businesses will need (invoicing, stock control, payroll, etc) and then build your own customisations on top it. Nobody sits down and thinks about writing their own Word Processor these days, they simply choose from the available options.

The key is being able to get the software to mirror your own processes as much as possible and not as I saw in a large SAP R/3 rollout, having to re-engineer your business around the software!

* Knowing the rate for a good SAP bod that advice has cost you £30.

[Chunketh]

The great thing about working in IT is that you don't get to rest on your laurels. Advance or get off shored.

I don't see the threat of AI taking over my job function, I keep learning. Being off shored is the primary threat and the primary target of automated systems (in this context).

[Chunketh]

2 hours ago, wsn03 said:

"The Cloud" .... what a load of pretentious sh1t...I'd like to punch the inventor of that term squarely in the face. As someone rightly said on here its just your stuff on someone elses server (and now open to hackers). It's such ******, I've been hearing about the new and clever ways to package and bundle IT blah blah blah my whole career, its still the same sh1t everywhere - access to information. The technology changes, but the concept doesn't, and the same people are still in the industry - their job titles change but little else.

And journalists are people paid to write stories, most of them simply repeat whatever crap they are told by whoever they interview - the reason why for instance people think house prices can only go up. What a phoney nonsense of a world we live in.

I thought that for a while but the landscape has changed. As another reply has stated, Amazon are doing it right and making a killing from it.

Why would you want to go to the trouble of setting up a DC when you can rent the CPU cycles you need at a fixed cost?

[Mikhail Liebenstein]

2 hours ago, BristolBuyer said:

The customers of the IT consultancy I work for are still stampeding towards AWS, so that's what I'm learning about at present. On the other hand, if the customers modifiy their bespoke applications to use the AWS-specific APIs they will be quite tied-in to AWS. I suspect that moving to a different cloud provider at a later date could be expensive.

As a customer-facing application developer/support person, I'm sometimes surprised at how much my employer's in-house hosting services cost, and how long it takes to get the in-house Windows/Unix/DBA administrators to fix things when they break, so maybe the cloud will help there, but only time will tell.

Do be careful with AWS, I know several firms that have ended up with surprise bills. One for £200k