Crossrail - Construction updates and progress towards opening (now expected 24 May 2022)

[coppercapped]

I've been told that, back in the day, a group of guys were challenged to write an operating system over a weekend. What they wrote was the core of UNIX. Not surprising it's still got bugs.

SNIPPED

I know this is going off topic, but it's only one post!

I don't know who told you that about the origins of UNIX, but they are misinformed.

UNIX has evolved dramatically since its first appearance in 1969 and it is now available in several different flavours and is used world wide. Bugs in the kernel of the OS have long since not been a thing.

Originally running on DEC PDP/11 machines it now works on a multiplicity of types of hardware. However, things move on and for the last 20 years Linux has become the most widely used Unix-like operating system, with other variants of Unix (apart from macOS[1]) having only a negligible market share.

[1] Since my old company Apple adopted Unix as its core operating system in 1997 all Apple machines have used it. It's also the core of iOS, watchOS and so on. In total the installed base of Apple products is some 1.5 billion items - all of those built since about 2000 using Unix. It's rock solid.

 

[InOban]

I am well aware that Unix in one flavour or another runs almost every computer which isn't running Windows. I've been told that even Hotmail runs on Unix.

 

[mrmartin]

Moving back on topic, I think it's becoming clear from reading that performance dashboard posted AND the fact there is still no concrete date, things aren't going well in terms of reliability at all.

Look at today's performance - it's a a complete mess against timetable performance (assuming they are actually trying to hit that currently?)

Realtime Trains | Departures from Farringdon Crossrail around 1213 on 17/03/2022

Train information at Farringdon Crossrail around 1213 on 17/03/2022. From Realtime Trains, an independent source of train running info for Great Britain.

www.realtimetrains.co.uk

In fact looking back this week there isn't really a day where you could say it was fully reliable. At least an hour or so of problems causing delays. I don't think you can launch the service like that, it would basically be on minor/severe delays every day in a row.

I assume there isn't many more software updates they can do before June now? Obviously one over easter, not sure if there would be time to do another one after?

 

[JonathanH]

Look at today's performance - it's a a complete mess against timetable performance (assuming they are actually trying to hit that currently?)

As has been pointed out before, none of us know what disruption scenario was being tested today. They are running to a timetable, yes, but with simulated disruption so it would be expected that it should not run to time.

The recent video said that they were getting close to the completion of this stage of testing. There would then be a lull and then they would operate to the planned timetable before opening. If there is disruption during that later stage there is a problem but at the moment 'delays' on the trial operations should be expected.

 

[J-2739]

Is there no testing taking place tomorrow on the core?

 

[JonathanH]

Is there no testing taking place tomorrow on the core?

Trial operations doesn't appear to be the only testing that needs to be done.

 

[Horizon22]

I guarantee there will still be "bugs" even after the full launch of service, some of which won't be fixed in the lifetime of the 345s.

Indeed. Many trains - and technology generally - have odd quirks and faults. The general difference is that they have a known solution & work around until they are dealt with (if they ever are) which is seen as proportionate and satisfactory.

345s have a lot lot fewer ETCS incidents than they once did and obviously this becomes more critical on a two track rail with a relatively intense service. It can already cause carnage and then delays for an hour or so when a 345 (or HeX 387) is stuck trying to reboot in the Heathrow tunnels.

 

[Nicholas Lewis]

Back a few decades ago (last century) and software was released when it was expected to be stable and bug free. After thorough testing. Showing my age.

Nowadays it is seen as acceptable to roll out new software or a new version of old software and expect bugs to be found while it is in use. This scenario is forced on customers by software manufacturers phasing out support for older versions.

That approach has not yet become acceptable with safety critical systems and I assume (hope) it never will. So Crossrail takes time.

But the argument goes that there is always a law of diminishing returns and that at some point the risk from the new software has to be taken as the risk is outweighed by the advantages. Again heavily skewed for safety critical systems. I am in the camp that says software should be written properly and tested properly. It is lazy to do otherwise and I have been confronted with repairing utter rubbish that suggests the programmer cannot write or amend software.

Umm it takes as long as it takes isn't a robust way of running a project and it seems to me software has now been allowed to get so complicated its almost impossible to test it.

Need to get back to the level of simple, but very effective, electromechanical electronic control that was deployed on the Victoria Line in the mid 1960's.

 

[mr_jrt]

UNIX has evolved dramatically since its first appearance in 1969 and it is now available in several different flavours and is used world wide. Bugs in the kernel of the OS have long since not been a thing.

Minor point of order - every piece of complex software that is still being changed is riddled with bugs. Most very inconsequential enough they'll never be found, some quite the opposite...

Case in point: https://nickgregory.me/linux/security/2022/03/12/cve-2022-25636/

The Discovery and Exploitation of CVE-2022-25636​

12 Mar 2022
A few weeks ago, I found and reported CVE-2022-25636 - a heap out of bounds write in the Linux kernel. The bug is exploitable to achieve kernel code execution (via ROP), giving full local privilege escalation, container escape, whatever you want.

It's not just Linux - even the Apple kernel you mentioned has severe bugs discovered from time to time, i.e: https://www.cvedetails.com/cve/CVE-2006-4392/

The Mach kernel, as used in operating systems including (1) Mac OS X 10.4 through 10.4.7 and (2) OpenStep before 4.2, allows local users to gain privileges via a parent process that forces an exception in a setuid child and uses Mach exception ports to modify the child's thread context and task address space in a way that causes the child to call a parent-controlled function.
Publish Date : 2006-10-03 Last Update Date : 2018-10-17

I'm a software developer, so this is very much in my area of expertise. Complex software engineering is very, very, very hard, which is why the mantra is always make things as simple as you can, because it's primarily the interactions between components and systems that introduce spiralling complexity.

 

[Roast Veg]

Need to get back to the level of simple, but very effective, electromechanical electronic control that was deployed on the Victoria Line in the mid 1960's.

The day to day expenditure of maintaining this is higher than that of any software, and only grows with time. It's just a bad business decision.

 

[Nicholas Lewis]

The day to day expenditure of maintaining this is higher than that of any software, and only grows with time. It's just a bad business decision.

Maybe when it got old and the technology had been abandoned, i don't like to use obsolete as there is this temptation to believe everything new has to be better than what preceded it, but that simple technology delivered 10's millions journeys safely and reliably. Im not a luddite and realise there have been huge benefits but what im contending is technolgy in the second decade of the 21st century has many examples of it just getting too complicated in part I believe due to too much inter connectivity between systems. So in future we should just question if we want or need all this functionality that computers and software can provide. Reality is hardware is relatively cheap but software engineers aren't and certainly testing it all isn't.

 

[AM9]

Maybe when it got old and the technology had been abandoned, i don't like to use obsolete as there is this temptation to believe everything new has to be better than what preceded it, but that simple technology delivered 10's millions journeys safely and reliably. Im not a luddite and realise there have been huge benefits but what im contending is technolgy in the second decade of the 21st century has many examples of it just getting too complicated in part I believe due to too much inter connectivity between systems. So in future we should just question if we want or need all this functionality that computers and software can provide. Reality is hardware is relatively cheap but software engineers aren't and certainly testing it all isn't.

Ironically this post just underpins an argument that I've posted on RUK more than once that the railway (and especially the UK railway) has allowed a plethora of random technical standards with indifferent interfaces to prevail, particularly since privatisation. This is where buyers flock to the cheapest bid where manufacturers offer their pet systems ignoring integration, that not insignificant element of cost of ownership, that excalates as incompatible systems and interfaces clash. The buyers have got what they paid for. Unfortunately, 'its buy cheap, struggle through with it' , - there's usually no 'buy twice' option.

 

[Roast Veg]

Ironically this post just underpins an argument that I've posted on RailUK Forums more than once that the railway (and especially the UK railway) has allowed a plethora of random technical standards with indifferent interfaces to prevail, particularly since privatisation. This is where buyers flock to the cheapest bid where manufacturers offer their pet systems ignoring integration, that not insignificant element of cost of ownership, that excalates as incompatible systems and interfaces clash. The buyers have got what they paid for. Unfortunately, 'its buy cheap, struggle through with it' , - there's usually no 'buy twice' option.

Software can be patched into compliance, if and when a decent standard is finally paid for, which is desperately needed.

Unfortunately you can't patch hardware, which underscores the point that mechanical/electromechanical solutions of old are not suitable.

 

[AM9]

Software can be patched into compliance, if and when a decent standard is finally paid for, which is desperately needed.

Unfortunately you can't patch hardware, which underscores the point that mechanical/electromechanical solutions of old are not suitable.

I think we're in agreement here. Of course, even considering going back to the simplicity of an electromechanical system a la the mid '60s would be bizarre, what next, - mechanical intelocking on track and signals?

 

[mmh]

Software can be patched into compliance, if and when a decent standard is finally paid for, which is desperately needed.

Unfortunately you can't patch hardware, which underscores the point that mechanical/electromechanical solutions of old are not suitable.

Vauxhall Nova drivers, class 153, 769 (etc) owners, and the directors of Vivarail wouldn't be as certain that you can't patch hardware...

 

[YSTrains]

Only the Class 345s are cleared for the tunnel?

 

[Nicholas Lewis]

Software can be patched into compliance, if and when a decent standard is finally paid for, which is desperately needed.

Unfortunately you can't patch hardware, which underscores the point that mechanical/electromechanical solutions of old are not suitable.

1960's Victoria Line never needed patching though because they spent the time perfecting the technology in parallel with the construction works so it was working from day 1 not needing over four years of testing (345 test trains were first running through crossrail in 2017)

 

[dosxuk]

1960's Victoria Line never needed patching though because they spent the time perfecting the technology

Indeed - and also why it's all still there today, having never needed to be changed at all. Perfection at it's best!

 

[Roast Veg]

Vauxhall Nova drivers, class 153, 769 (etc) owners, and the directors of Vivarail wouldn't be as certain that you can't patch hardware...

I take your point, though I strongly suspect you may not believe it yourself. The critical difference is that this would fall under "aftermarket" mods rather than a manufacturer update.

 

[JonathanH]

Only the Class 345s are cleared for the tunnel?

Yes, as far as passengers are concerned. There is also some engineering access.

 

[Watershed]

Yes, as far as passengers are concerned. There is also some engineering access.

Although there are restrictions on the use of diesel engineering trains.

 

[43066]

I reckon when it does eventually open it will be low key with a press release after it has open. Maybe with an 'official' opening with Prince Charles cutting a ribbon a couple days later.

It would be nice if the queen was able to open it. Sadly that’s a lot less likely now than if it had opened on time!

 

[345 050]

Is there no testing taking place tomorrow on the core?

Getting this thread back on topic, it does look like all services through the core are cancelled from yesterday onwards. A bit of a break before Shadow running? New update being rolled out?

 

[JonathanH]

Getting this thread back on topic, it does look like all services through the core are cancelled from yesterday onwards. A bit of a break before Shadow running? New update being rolled out?

Back on Tuesday by the looks of things.

 

[345 050]

Back on Tuesday by the looks of things.

Ah cheers, I only looked as far as Monday... Oops

 

[Malcolmffc]

Indeed - and also why it's all still there today, having never needed to be changed at all. Perfection at it's best!

Umm, except it isn’t. The original 1960s signalling needed to be ripped out and replaced with modern signalling when the 2009 tube stock were introduced. The old signalling couldn’t support 32tph

== Doublepost prevention - post automatically merged: 19 Mar 2022 ==

1960's Victoria Line never needed patching though because they spent the time perfecting the technology in parallel with the construction works so it was working from day 1 not needing over four years of testing (345 test trains were first running through crossrail in 2017)

A more accurate explanation is that there was basically zero investment in the tube after the Victoria Line was built until the late 1990s

 

[mr_jrt]

Wasn't there the '83 stock on the Jubilee, the Central Line's ATO in '86, and the DLR in the late 80's?

 

[snowball]

A more accurate explanation is that there was basically zero investment in the tube after the Victoria Line was built until the late 1990s

Surely there was the first phase of the Fleet/Jubilee line?

And the rebuild of Angel station?

 

[317 forever]

If yesterday was the last big passenger station exercise, They must have all gone ok? Does anyone know if they need to redo any?

There is a train software update over Easter and then maybe 2 weeks full running?

I'm going the end of May myself, In time for the Queens events.

That would be a very good date, safety permitting.

It would be embarrassing having the Elizabeth Line open late for Queen Elizabeth's Jubilee celebrations. I may well ride it on June 4th if it is indeed open by then.

 

[CyrusWuff]

It would be embarrassing having the Elizabeth Line open late for Queen Elizabeth's Jubilee celebrations. I may well ride it on June 4th if it is indeed open by then.

Slightly less embarrassing than the Jubilee Line opening two years after the Silver Jubilee though! (Ignoring, for a moment, that CR should have opened in 2018.)