Tyne & Wear Metro Fleet Replacement: Awarded to Stadler

[hondafan1999]

That may be - I don't know the technology or the other requirements that need to be met before deployment. The point was to illustrate cost build up, and why calling changes "expensive" often misses the point.

Working in the IT industry, and dealing with estimates and pricing, I will also observe that a frequent error is that developers say "it will take X" and the other costs of the change (test, deployment, documentation where applicable) get overlooked.

Yeah, I don't know the underlay technology they're using either, however it's such a simple change to the presentation layer, I understand the other costs involved with everything, however if such a change is going to cost in excess of 1000s of pounds, nexus and the UK taxpayer are being absolutely rinsed, I've worked on presentation layers for billboard advertising, and this is a change it would take myself, or anyone in my team a few hours to change, test and document.

I was with Tyler when we were told by a Nexus employee that it was a heft expense, and the change in question was changing the font on the display. As a software developer myself, I'd sure hope that their software isn't such a mess that it would take 10 days of effort to implement and test a change to the font! Of course, whether such a change is necessary is another question, but just mentioning this as a bit of context for where "a hefty expense" is coming from. Changing the text on the bottom right of the screen I can imagine being slightly more difficult, depending on how it's implemented, but as someone else has mentioned, it's still unlikely to take as long as 10 days.

Madness, a font change a hefty expense?
Assuming the font in question is under a open license and doesn't actually cost to use under commercial terms, that's absolutely insane.

 

[hopperelec]

Yeah, I don't know the underlay technology they're using either, however it's such a simple change to the presentation layer, I understand the other costs involved with everything, however if such a change is going to cost in excess of 1000s of pounds, nexus and the UK taxpayer are being absolutely rinsed, I've worked on presentation layers for billboard advertising, and this is a change it would take myself, or anyone in my team a few hours to change, test and document.


Madness, a font change a hefty expense?
Assuming the font in question is under a open license and doesn't actually cost to use under commercial terms, that's absolutely insane.

There were two fonts in question. One was Calvert, which isn't an open font, but Nexus definitely already has the license to use. If I remember correctly, the other was an open font.

 

[hondafan1999]

There were two fonts in question. One was Calvert, which isn't an open font, but Nexus definitely already has the license to use. If I remember correctly, the other was an open font.

Whilst we don't know the figures of what the driver meant by 'hefty'
It could mean anything really, one number to you can be very different to me. Having said all that surely this is something that needs some sort of inquiry, at the end of the day it's us taxpayers money that is being spent here, changing the font and changing the presentation layer of a PIS is not hard and doesn't need to be so beurocratical

 

[northeast10]

Looks like units 02/03/15/20 have had the updated software so far.

 

[Paul_10]

Looks like units 02/03/15/20 have had the updated software so far.

So what's the update? Door pre-select disabled?

 

[DanNCL]

The update has changed it so that door pre-select only arms the doors on the side that the passenger presses the button, it will no longer arm the doors on the opposite side.

For a repeat of the Simonside incident to occur, in addition to the driver releasing the doors on the wrong side, a passenger would also have to pre-select the doors on the wrong side. The chances of that happening are incredibly slim.

 

[Peter Sarf]

Yeah, I don't know the underlay technology they're using either, however it's such a simple change to the presentation layer, I understand the other costs involved with everything, however if such a change is going to cost in excess of 1000s of pounds, nexus and the UK taxpayer are being absolutely rinsed, I've worked on presentation layers for billboard advertising, and this is a change it would take myself, or anyone in my team a few hours to change, test and document.


Madness, a font change a hefty expense?
Assuming the font in question is under a open license and doesn't actually cost to use under commercial terms, that's absolutely insane.

Depends how the font was coded. It might need several edits all through the code.
Also changing a font inevitably runs the risk of upsetting the layout if it makes a line of words longer.

Testing its always worth doing to see if something totally unexpected happens.

Testing - should they test it in combination with the rest of the systems running on the same hardware ?.

Note it is possible that a very unlucky software change can corrupt other software loaded or crash the whole machine.
What if the other software is safety critical ?.

I was charged out for a lot of money just to work out why a mainframe was crashing and fix the poor software that caused it.
It could have just corrupted other software loaded but luckily seemed to crash the whole machine.
Imagine if that machine had been safety critical.

Whilst we don't know the figures of what the driver meant by 'hefty'
It could mean anything really, one number to you can be very different to me. Having said all that surely this is something that needs some sort of inquiry, at the end of the day it's us taxpayers money that is being spent here, changing the font and changing the presentation layer of a PIS is not hard and doesn't need to be so beurocratical

Yes "Hefty" is not an actual figure.
But it won't be cheap and it certainly will not be free.
If the customer cannot get their requirements correct then they should pay if they change their mind.
It is why we were always reminded to get anything required in writing and then get it agreed.

 

[inais20]

The update has changed it so that door pre-select only arms the doors on the side that the passenger presses the button, it will no longer arm the doors on the opposite side.

This seems much more logical to me and is probably how it should’ve been from delivery. Why on earth would you ever need both sides armed after selecting just one side? Very strange default!

 

[hopperelec]

This seems much more logical to me and is probably how it should’ve been from delivery. Why on earth would you ever need both sides armed after selecting just one side? Very strange default!

There are two cases where I can think of this being useful:
- For people who don't know which side it will be on, particularly for Airport and St James
- When it's busy and someone realizes they're on the wrong side when approaching their stop and want to make sure they have a chance to get past other people before the doors lock
But, of course, safety is more important than these minor cases!

 

[Paul_10]

There are two cases where I can think of this being useful:
- For people who don't know which side it will be on, particularly for Airport and St James
- When it's busy and someone realizes they're on the wrong side when approaching their stop and want to make sure they have a chance to get past other people before the doors lock
But, of course, safety is more important than these minor cases!

For me it just feels like a bit of a gimmick to show how much we can do with technology these days.

Pushing a button once the train stops I'm sure 99% of the time has worked for people so it does feel a little unnecessary in my view.

 

[pokemonsuper9]

For me it just feels like a bit of a gimmick to show how much we can do with technology these days.

Pushing a button once the train stops I'm sure 99% of the time has worked for people so it does feel a little unnecessary in my view.

But it's nice to be able to hit a button and know it will open at the first opportunity rather than having to keep your finger on/just above the button waiting for it to be unlocked.

 

[hopperelec]

For me it just feels like a bit of a gimmick to show how much we can do with technology these days.

Pushing a button once the train stops I'm sure 99% of the time has worked for people so it does feel a little unnecessary in my view.

The main reason the pre-select is so necessary on these new trains is because most people expect to be able to open the doors as soon as the train stops. However, once the train stops, it takes a few seconds for the sliding step to extend, and the doors aren't allowed to open until the sliding step has extended. I've seen plenty of people get frustrated at why the door isn't opening when they press it and, on multiple occasions, have seen them start walking away to try another door, only for the door they technically "pre-selected" to finally open. But it's also useful for people who might have a lot of luggage and things they need to collect together before they leave the train- they can press the button, then pick up all their things without needing to worry about needing to press it with everything in their hands.

 

[Paul_10]

The main reason the pre-select is so necessary on these new trains is because most people expect to be able to open the doors as soon as the train stops. However, once the train stops, it takes a few seconds for the sliding step to extend, and the doors aren't allowed to open until the sliding step has extended. I've seen plenty of people get frustrated at why the door isn't opening when they press it and, on multiple occasions, have seen them start walking away to try another door, only for the door they technically "pre-selected" to finally open. But it's also useful for people who might have a lot of luggage and things they need to collect together before they leave the train- they can press the button, then pick up all their things without needing to worry about needing to press it with everything in their hands.

Sounds a bit like the doors on the prototypes which I remembered back in the 90s would take ages(relatively speaking) to open that sometimes people pushed them to open. They did get much quicker in the end and maybe some modifications are needed for the doors on the 555s if indeed people are worried the doors are taking too long to open

I will hopefully finally be able to sample the 555s when I'm up there next month, will certainly be different for sure.

 

[FusionZenFlame]

Does anyone here have any idea on when the rest of the 555 schedules will be back and running? I would've thought it was going to be today but it has proven otherwise.

Any helpful information would be handy thanks!

 

[AndrewP]

There were two fonts in question. One was Calvert, which isn't an open font, but Nexus definitely already has the license to use. If I remember correctly, the other was an open font.

For reference some similar typefaces to Calvert which was originally designed for a French town but dismissed as too English (so why employ an iconic British designer?) are Rockwell for a standard MS Windows typeface and Kameron which is a Google font but the question is - would these be close enough?

 

[norbitonflyer]

Does anyone here have any idea on when the rest of the 555 schedules will be back and running? I would've thought it was going to be today but it has proven otherwise.

Any helpful information would be handy thanks!

Is it still just two? Hoping to go up tomorrow to try them out (and do Ashington)

 

[FusionZenFlame]

Is it still just two? Hoping to go up tomorrow to try them out (and do Ashington)

Sadly today it was. Im hoping either tomorrow or Wednesday it becomes three or four at once.

 

[typefish]

For reference some similar typefaces to Calvert which was originally designed for a French town but dismissed as too English (so why employ an iconic British designer?) are Rockwell for a standard MS Windows typeface and Kameron which is a Google font but the question is - would these be close enough?

I don't quite understand why there would be an issue with using Calvert, it is most certainly available in TTF format amongst others - and from what I've figured out from watching the passenger information system reboot (thanks to YouTube) the PIS definitely is running what appears to be a modern version of Linux.

I guess we'll need to wait and see what happens next I guess

 

[hopperelec]

I don't quite understand why there would be an issue with using Calvert, it is most certainly available in TTF format amongst others - and from what I've figured out from watching the passenger information system reboot (thanks to YouTube) the PIS definitely is running what appears to be a modern version of Linux.

I guess we'll need to wait and see what happens next I guess

The same person who uploaded the video in the screenshot, also mentioned in a separate group that they saw it again, this time with the distro version. Debian Buster (10) which was released in Sep 2022. So yes, definitely a modern version of Linux.

 

[pazanate]

555006 operating passenger service T121 towards St. James today

 

[hacman]

The same person who uploaded the video in the screenshot, also mentioned in a separate group that they saw it again, this time with the distro version. Debian Buster (10) which was released in Sep 2022. So yes, definitely a modern version of Linux.

Buster is now out of support, and has been since last June (unless they've switched to the unofficial ELTS train).

To be honest, Debian is a strange choice for this as it typically has shorter lifecycles than other distributions, and given what Stadler is doing, it's a little surprising they're not just rolling their own distro.

 

[hopperelec]

Buster is now out of support, and has been since last June (unless they've switched to the unofficial ELTS train).

To be honest, Debian is a strange choice for this as it typically has shorter lifecycles than other distributions, and given what Stadler is doing, it's a little surprising they're not just rolling their own distro.

It's out of support, but my point was just that it's certainly modern enough to support rendering custom fonts lol. Also, I don't see what benefit rolling their own distro would have?

 

[hacman]

It's out of support, but my point was just that it's certainly modern enough to support rendering custom fonts lol. Also, I don't see what benefit rolling their own distro would have?

Modern enough to support custom fonts basically covers anything from the last 30 years.

Rolling their own distro has quite a few benefits - mostly surrounding tight control of package versioning, ensuring that the footprint is kept minimal (for security and also management), and it also ensures they control the support timeline and lifecycle. There are a few other more technical benefits too.

I used to design and maintain PIS and other railway systems, and we rolled our own Linux distro, or where we had to use Windows, we built a heavily customised image of the Windows Embedded variant; ultimately it ensured we always had 100% control of the platform.

 

[AndrewP]

I don't quite understand why there would be an issue with using Calvert, it is most certainly available in TTF format amongst others - and from what I've figured out from watching the passenger information system reboot (thanks to YouTube) the PIS definitely is running what appears to be a modern version of Linux.

I guess we'll need to wait and see what happens next I guess

I don't think there would be an issue but from a cost effective and design perspective that most people would not notice (some would - like er me!) there would be no need to dependent on Calvert (even if it would be best to use it)

 

[MetroTyler]

And if they didn't charge per change, you can bet that there would be a premium on the overall contract price to cover the costs of changes at will. That would be jacked up because they would assume, not unreasonably, that a customer will want to make more "free" changes than if the changes have to be paid for.

I've no idea what the price per change would be, but I'd also be careful about saying "it's a hefty expense" without thinking about the cost of making (including testing) that change. If their staff are paid an average of £50k (low in the IT industry), then a piece of work requiring 10 days of effort would cost over £2k simply on salary costs, before any other costs of employment or corporate overheads.

I'm saying its a hefty expense considering Nexus absolutely do not have the funding to really allow for any changes to such a system as an addition beyond the contract. Small things with the testing process and development process involved is simply expensive.

Buster is now out of support, and has been since last June (unless they've switched to the unofficial ELTS train).

To be honest, Debian is a strange choice for this as it typically has shorter lifecycles than other distributions, and given what Stadler is doing, it's a little surprising they're not just rolling their own distro.

Debian is a perfectly fine choice, its an LTS distribution that, although gets many releases, allows for a long life cycle and backported security updates for years to come. It'll be a long time till they need to upgrade the full distribution.

 

[35B]

I'm saying its a hefty expense considering Nexus absolutely do not have the funding to really allow for any changes to such a system as an addition beyond the contract. Small things with the testing process and development process involved is simply expensive.

Fair enough, but they then need to consider whether the changes really are important - and make their choices accordingly.

I’ve lost count of the number of times customer personnel have said “it’s vital” and then, faced with a cost, gone “well, actually, I didn’t really need it”.

I gave a very simple example of why things rack up charges, and how even a small piece of work actually costs a measurable amount of money. That cost is real, wherever it lies, and while I’m sympathetic to British local authorities for their tight budgets, that doesn’t mean I regard them as deserving charity.

 

[hacman]

Debian is a perfectly fine choice, its an LTS distribution that, although gets many releases, allows for a long life cycle and backported security updates for years to come. It'll be a long time till they need to upgrade the full distribution.

Debian Buster is out of support, it has exited LTS. It no longer gets backported security updates. You can get ELTS support for it, along with Jessie and Stretch, but it is not provided by the Debian project and is a commercial service from a 3rd party with a limited scope. Bullseye is the release that is currently in LTS, with Bookworm being the current stable. Upgrades between releases are not always straightforward, though in this case, you'd just re-image the system so that's less of an issue.

Debian is fine, but as I say, having done this sort of thing in the past, it surprises me that they don't just roll their own.

 

[norbitonflyer]

Sadly today it was. Im hoping either tomorrow or Wednesday it becomes three or four at once.

There seemed to be at least two on the coast/ South Shields route yesterday. I travelled on 006 working T121 (at Central station at 1445) , and shortly before that I saw another running through Byker coastwards at about 1430, possibly T123? Or was that a training/test run

 

[MetroTyler]

Debian Buster is out of support, it has exited LTS. It no longer gets backported security updates. You can get ELTS support for it, along with Jessie and Stretch, but it is not provided by the Debian project and is a commercial service from a 3rd party with a limited scope. Bullseye is the release that is currently in LTS, with Bookworm being the current stable. Upgrades between releases are not always straightforward, though in this case, you'd just re-image the system so that's less of an issue.

Debian is fine, but as I say, having done this sort of thing in the past, it surprises me that they don't just roll their own.

System should still be okay is the main thing. They'll have presumably built a lot on top of it however so we don't really know what Stadler have done with it, it might be a great choice for them.

Thanks for correction, I'd thought the backports phases had broader scope than they do.

 

[rg177]

555011 out on test today.