Ultimaker S5 | impressions & opinions

[Marco_TvM 34]

11 hours ago, asb said:

So without these requirements which I can only guess, it uses an APIPA address; in my case it was 169.254.85.135. APIPA addresses are special IP addresses like 127.0.0.1 (locahost), and they are volatile (= they change).

Correct. If there is no DHCP server on the network the printer is connected to, or the DHCP server is MAC restricted, it will use the 169.x range.

 

11 hours ago, asb said:

Again, this is why a static IP address is crucial. If you try to work productively with APIPA addresses, everything gets totally borked after a couple of days. This is well-known behaviour since the UM3 and not a new bug in the S5, it's just how APIPA works. So the next challenge will be to resolve the static networking issue.

I disagree with the statement that a static IP address is crucial. Or are you enjoying to have to manually install an IP address on every device like a pc, laptop, mac, tablet, mobile phone, any wired/wireless device able to connect to a network?

I'm not - having many devices, I rather have a well-configured router/switch that provides a DHCP service which can be configured to give out 'fixed' IP addresses to specified MAC addresses.

 

11 hours ago, asb said:

After a couple of retries I got an error message in the printer: "An unspecified error has occured. Restart the printer or go to ultimaker.com/ER27". The explanation there says that an unspecified error has occured and I should update the firmware. Not very helpful.

 

11 hours ago, asb said:

Addendum: On the USB stick I found a a large number of (276) logfiles.They are labeled like so: "cap_Ultimaker-005f24_1970-01-01_00.09.10_N1_X46.5Y33.5_sensor_reset.log" (size: 0 Bytes) or "cap_Ultimaker-005f24_1970-01-01_00.09.10_N1_X46.5Y33.5Z7.79745.log" (size: 26 Bytes; content: "7.79745 // ok N202 P14 *B941", all dated January 1st, 1980.

The error you received seems like a fault that could not be recovered from. Did you by chance look at the leds of the cores and remember the color? I think this error was caused by some stalling communication between opinicus and marlin, leading marlin to believe the printer stopped working and halting operations (turning off heaters and stuff).

 

All those files on the USB stick are helpful in analyzing and backtracking the problem that has popped up - they are copied from the system to USB (if possible) so it is a sort of snapshot to give insights in what was going on with the printer during the time the problem occurred. If it is possible, can you send them to us using wetransfer.com so we can have a look?

[IRobertI 520]

3 hours ago, Marco_TvM said:

I disagree with the statement that a static IP address is crucial. Or are you enjoying to have to manually install an IP address on every device like a pc, laptop, mac, tablet, mobile phone, any wired/wireless device able to connect to a network?

I'm not - having many devices, I rather have a well-configured router/switch that provides a DHCP service which can be configured to give out 'fixed' IP addresses to specified MAC addresses.

Disagree all you want, the problem is that some companies NEED this functionality, it is just that simple. It's a question that has been popping up regularly since the UM3 was introduced.

[Dim3nsioneer 557]

53 minutes ago, IRobertI said:

Disagree all you want, the problem is that some companies NEED this functionality, it is just that simple. It's a question that has been popping up regularly since the UM3 was introduced.

 

I can confirm that larger companies do indeed have that request.

[tinkergnome 927]

49 minutes ago, IRobertI said:

the problem is that some companies NEED this functionality, it is just that simple. It's a question that has been popping up regularly since the UM3 was introduced.

 

Yes - of course no one configures static IP-addresses for every mobile device nowadays, but for servers and printers it is quite common, isn't it?
 

...and: it's already possible (and a quick task), but it is not documented and you have to activate the developer mode (at least temporary).
So: please document it and/or make it more accessible for customers and resellers (the same goes for the MAC-address).

[nallath 1,120]

I think the thing here is perceived and actual need. It doesn't really matter what a company actually needs, it matters what they think they need. So even though you might be able to solve the static IP thing in another way (which may or may not be better), it does not mean that they don't need the feature (because they expect it to be there and convincing them otherwise is harder than building the thing in the first place). 

[Marco_TvM 34]

1 hour ago, tinkergnome said:

Yes - of course no one configures static IP-addresses for every mobile device nowadays, but for servers and printers it is quite common, isn't it?
 

This statement already contradicts itself.

Simply said: you want to install a DHCP service for every device out there, except it should not give out IP addresses being used by servers and printers... Hence multiple places to manage IP addresses. You have to record every static IP address; to make sure a) it cannot be given out by the DHCP server (using pools of addresses) and  b) it will not be manually assigned again and c) if a server is replaced, it will get the same ip number again...

 

But sure, I guess IT companies love making money doing things much harder than they really need to.

[Marco_TvM 34]

3 hours ago, IRobertI said:

Disagree all you want, the problem is that some companies NEED this functionality, it is just that simple. It's a question that has been popping up regularly since the UM3 was introduced.

I disagreed with the statement that having a static IP is crucial. It is not.

As @nallath mentioned: people WANT it - but with current technology there is no need for it.

Speaking from experience having being part of network engineering and having to fix all kinds of weird static ip issues, mostly due to bad bookkeeping leading to double or even triple use of the same ip address - which are a hassle to find and fix.

[korneel 176]

7 minutes ago, Marco_TvM said:

This statement already contradicts itself.

Simply said: you want to install a DHCP service for every device out there, except it should not give out IP addresses being used by servers and printers... Hence multiple places to manage IP addresses. You have to record every static IP address; to make sure a) it cannot be given out by the DHCP server (using pools of addresses) and  b) it will not be manually assigned again and c) if a server is replaced, it will get the same ip number again...

 

But sure, I guess IT companies love making money doing things much harder than they really need to.

nonsense. really, this is not your field of expertise. this reply actually annoyed me.

let me break it down as simple as i can because your reply really does not show any understanding for the average company.
so any decent company is splitting up their networks into networking segments. you've got your DMZ, you've got your wireless, your guest access, usually layered server access and things like that. using firewalls, communication between those segments is controlled and it will help you protect yourself  against things like broadcast storms, unauthorized client access etc. 

this is then usually managed by using separate DHCP scopes (not services) from a centralized and authorized DHCP service with usually 1 or 2 backups. 

anyway, some segments have from a security point of view a MAC filter so no unauthorized devices will enter, and some companies (like universities) simply run B class public IP ranges where they need tight control over the usage of IP addresses. they can't just run a dhcp scope.

so there are millions of super valid reasons for companies to require static IP addresses and use ACL's based on MAC addresses to control access to the network. 

yes, they actually do record every IP address (which are usually in ranges where there is no DHCP scope configured so that point is moot).  and yes when a service is replaced it usually get's the same IP address.

no, IT companies don't do this to make money, your view is based on sub 100 computers. your view works great for a home network, and does not work on the enterprise market. happy to have a discussion about this, but this is essentially my daytime job. 

the way it currently works (you only get a mac address if you have a valid DHCP scope configured for this segment) is a mega weird strategy. i can live with the non-static IP addresses, but this makes me weary of the networking stack on this device if basic functionality does not exist. 

please @Marco_TvM take these comments seriously. you are dismissing them in a way that shows you have done zero investigation in this area.

 

[korneel 176]

18 minutes ago, Marco_TvM said:

I disagreed with the statement that having a static IP is crucial. It is not.

As @nallath mentioned: people WANT it - but with current technology there is no need for it.

Speaking from experience having being part of network engineering and having to fix all kinds of weird static ip issues, mostly due to bad bookkeeping leading to double or even triple use of the same ip address - which are a hassle to find and fix.

and that's where you're wrong kiddo. 

the people that get to decide what is crucial and not are the people that give you money. those are your target audience. if enough come back stating they need this, it just became crucial irregardless of your opinion. 

 

and if we're throwing in the engineering credentials; CCNA all the way to qualifying for the CCIE exam, you're looking at a properly educated networking engineer. 

 

please listen to the people who give you money. they don't care about your opinion what is needed or not. they care about buying a product that meets their expectations.

 

edit:

still puzzled with the there is no need statement.

using your new knowledge i described above, why is it not needed? there are simply customers that are not allowed to have DHCP scopes for certain segments by their security department.. 

 

Edited July 23, 2018 by korneel

[Marco_TvM 34]

@korneel Not going to quote all the points, but...

I don't make the decisions on what is going to be done - I only gave my opinion based on my experiences with big and small networks, for big and small companies at the time I was involved in this - that included commercial and non-commercial sectors including University networks etc. So don't make assumptions on sizes I have worked with, please, even tho it is already 15-20 years ago

21 minutes ago, korneel said:

the way it currently works (you only get a mac address if you have a valid DHCP scope configured for this segment) is a mega weird strategy. i can live with the non-static IP addresses, but this makes me weary of the networking stack on this device if basic functionality does not exist. 

 

At the time of implementation of the UM3, the use for static IP was moot (at least for those making the decisions). The stack, connman, should be very well able to handle it (being used on Android devices too afaik) but interface wise (hmi) nothing is defined. 

 

Also, the impression of how it currently works is wrong.  The MAC address is always there (for both the cable as wifi connection) but we lack in the display to show this information (when the network is not active). I think that is an oversight from our part, as we just should show both MAC addresses.

Unless I misunderstood your statement I quoted...

 

[kmanstudios 1,120]

1 minute ago, Marco_TvM said:

but we lack in the display to show this information (when the network is not active). I think that is an oversight from our part, as we just should show both MAC addresses.

Sounds like an easy fix, eh?

[Marco_TvM 34]

1 minute ago, kmanstudios said:

Sounds like an easy fix, eh?

Not as easy as it sounds unfortunate.

[kmanstudios 1,120]

Ahhhh, but one that should be looked at rectifying quickly for large, security customers that you guys are trying reach, eh? :)

[Marco_TvM 34]

2 minutes ago, kmanstudios said:

Ahhhh, but one that should be looked at rectifying quickly for large, security customers that you guys are trying reach, eh? ?

As said before - I'm not the one making the decisions...

[korneel 176]

13 minutes ago, Marco_TvM said:

@korneel Not going to quote all the points, but...

I don't make the decisions on what is going to be done - I only gave my opinion based on my experiences with big and small networks, for big and small companies at the time I was involved in this - that included commercial and non-commercial sectors including University networks etc. So don't make assumptions on sizes I have worked with, please, even tho it is already 15-20 years ago

 

At the time of implementation of the UM3, the use for static IP was moot (at least for those making the decisions). The stack, connman, should be very well able to handle it (being used on Android devices too afaik) but interface wise (hmi) nothing is defined. 

 

Also, the impression of how it currently works is wrong.  The MAC address is always there (for both the cable as wifi connection) but we lack in the display to show this information (when the network is not active). I think that is an oversight from our part, as we just should show both MAC addresses.

Unless I misunderstood your statement I quoted...

 

well, based on your reply i am making assumptions. the universities in NL that I worked with over the past 10 years, all required static IP addresses and the MAC address to be registered against the ACL. amongst those are the 5 large ones in NL. we tried deploying IP Phones that could not do static IP addresses and that meant a whole design change request to go through the board. if we could have just had static ip's in those phones that would not have been needed..  

i was referring to your earlier statement about how no-one would ever need static ip addreses and how DHCP servers would fix everything and then you make a statement about how the only reason for it would be for IT companies to make money. no-where did it use the sentence "in my opinion" and it is very clearly said by someone who is part of team Ultimaker. that's perception. 

no-one in the networking field would ever make that statement in my opinion that manually assigning IP addresses is not needed.. but hey ho what do I know.. 

 

my apologies for being unclear, I fully understand that both the WIFi and the Network adapter have their own MAC address, what i mean twas that for me as a user, i don't get that address, or as you describe it, get to see the address, unless a cable is connected AND a connection is made. that is mega weird. 

sounds like a bug that can be fixed.

 

let's not get into this DHCP fixes everything discussion, is plainly not true and i don't feel like typing 3 pages to proof it's not true.

[WesleyE 14]

Both properly displaying the MAC addresses of both interfaces and the requests for static IP's are on our backlog. Not displaying the MAC addresses has indeed been an oversight that we are looking to fix quickly, but I'm not sure it can be worked in before our next release, so no promises.

 

With most users so far we have been able to find a workaround for the static IP's, so this has not been a top priority for us before. However, we are getting a lot of requests (not only on these forums) for supporting more network configurations and are actively looking to build support for those configurations. It is on our backlog, but I can not promise you guys a timeframe for now.

[korneel 176]

1 minute ago, WesleyE said:

Both properly displaying the MAC addresses of both interfaces and the requests for static IP's are on our backlog. Not displaying the MAC addresses has indeed been an oversight that we are looking to fix quickly, but I'm not sure it can be worked in before our next release, so no promises.

 

With most users so far we have been able to find a workaround for the static IP's, so this has not been a top priority for us before. However, we are getting a lot of requests (not only on these forums) for supporting more network configurations and are actively looking to build support for those configurations. It is on our backlog, but I can not promise you guys a timeframe for now.

thank you for this open and workable answer ?

 

[asb 7]

For a couple of days I have been talking back and forth with my vendor about developer mode and factory warranty. The vendor (ELV) forwards everything to the German distributor (iGo3D), so getting answers takes some time.

 

I am afraid we have a serious dealbreaker here. As far as I understand the rather cryptic response, you will loose factory warranty and guarantee by enabling developer mode.

 

Also, the distributor refuses to provide documentation or support for developer mode.

 

This is a Catch 22. Without developer mode it is impossible to set up a static IP address, and also, without developer mode it is impossible to calibrate X-, Y- and Z-axis to gain better print accuracy for technical objects with precise measurements.

 

Honestly , I don't like to gamble with an 6,533 Euro value, without even knowing that the issues actually can be resolved with developer mode.

 

I am quoting the original statements for your information.

 

Distributor (iGo3D) statement, as of July 27th, 2018:

Quote

Da der Entwicklermodus frei zugänglich und nicht durch ein Passwort gesichert ist, kann der Kunde diesen gerne nutzen. Der Kunde sollte aber schon sicher sein was er da tut, denn treten durch Veränderungen hier Fehler auf, dann kann die Garantie u.U. abgelehnt werden wenn diese durch die falsche Nutzung des Entwickler Modus entstanden sind. Eine Dokumentation für den Endkunden gibt es hier leider nicht.

 

Since you can not know exactly what you are doing when there is no documentation, this means pretty clearly that whatever happens after enabling developer mode once, factory warranty and guarantee can be refused without further discussion.

 

Thereon I asked to confirm my understanding of this statement and to give a clear yes/no answer.

 

Vendor (ELV) statement, as of July 31st, 2018:

Quote

[…] nach Rücksprache mit unserem Llieferrant müssen wir Ihnen leider mitteilen, dass dieser seine bereits getätigte Aussage nicht weiter konkretisieren kann. Wir als Ihr Lieferant raten Ihnen im Zweifelsfalls von einer Nutzung des Entwicklermodus ab, da die Garantie gemäß des Herstellers ggf. in Frage gestellt wird. Sofern Anforderungen an dieses Gerät nicht erfüllt werden, senden Sie das Gerät im Rahmen des Rückgaberechts an uns zurück.

 

Paraphrased this says that the distributor refuses to make a unambiguous statement. The vendor recommends not to use the developer mode because of a possible loss of factory warranty and suggests to return the printer.

 

To state this clearly: I do not want to return the printer. But I need to set up a static IP address, and I need to calibrate it.

I do not care how this is accomplished, but I am not willing to risk warranty for it.

 

Question to Team Ultimaker: Is this handling really the way you want it?

[korneel 176]

well. My german might not be best, but I think your translation is a bit more harsh then what they said.. 

how I read it is that if you enable developer mode, you're free to do that. you should however realize that developer mode is the wild west, they will not provide you with in depth documentation and if you change stuff that damages the printer (let's say you crank up the amperage so now your drivers break or you change the parameters of the printer so the stepper motor rams the print head into the sides until it breaks) they can refuse warranty on that. I see no harm in that statement.. 

 

furthermore.. if you download the full system image (search the forum here) and you do damage something like that, flash the system image back and no one is any wiser that you ever enabled dev mode.. so i don't see the point of the statement to be honest.. 

i think it's more of a "we have to tell you this or our lawyers get angry" kind of statement then any real "yeah, you will loose all warranty as soon as you enable dev mode". 

 

just be prepared that if you enable dev mode, do weird things that make the software crash, and you have to take the printer in to reinstall the software if you were enable to do it yourself, they might charge you for it.. and i see no problem for that.

then again, I don't work for Ultimaker ?

[SandervG 1,521]

Hi @asb , thank you for your post. 

It's true, when you enable dev mode you basically 'open the hood', which offers you some handy tools but these tools / values could potentially damage your printer. Like in Korneel's example. But just enabling it, shouldn't result in a direct loss of warranty. Caution is advised though.

 

I'll see if I can collect some instructions on how to set up a static IP via dev mode, so you know where to look and you should have no risk of damaging anything ?

 

Hope this helps, and our apologies for the inconvenience. 

[SandervG 1,521]

Update; I've sent an update to @asb and offered to help him set up a static IP. I will do so in 'private', because it involves entering dev mode which has its own set of risks and we prefer if no one does it. But if necessary, guided. 

[kmanstudios 1,120]

6 minutes ago, SandervG said:

Update; I've sent an update to @asb and offered to help him set up a static IP. I will do so in 'private', because it involves entering dev mode which has its own set of risks and we prefer if no one does it. But if necessary, guided. 

?

[asb 7]

If we use the analogy that enabling developer mode is 'opening the hood', it would be the same like popping the hood on my car. I can do that without loosing factory warranty, and in the glove compartment of my car I have a detailed manual where stuff like spark plugs is and how I can change them. I don't know about other brands, by mine is maintenance-fiendly. So if the S5 would be engineered like a car in this analogy, we wouldn't have this discussion.

 

Having that said, I confirm that @SandervG send me partial instructions how to set up a static IP address, which is really cool. So thank you! It's basically the same what is (or was) on the forum somewhere last year when we had the same issue with the UM3. However, the instructions are for the Ethernet interface, that's the one I can not activate on the S5. So I need to read about using ConnMan how to configure the Wifi interface first, which is a bit more complicated.

 

Also, going onto the S5 requires SSH, which requires routed networking. So far I'm getting those pesky APIPA addresses which give me a "no route to host" from our subnet. So I need to figure out how to access the S5, then figure out how to use ConnMan, and then I can take a look how to resolve the actual problem - calibrating the printer so it outputs mechanical parts that are usable for assembling.

[Bigbrit 8]

Hi Everyone,

 

ive had my ultimaker S5 for a couple of weeks now and I have some observations.

 

its a great printer, I love it and it works really well but

 

1) the instructions in the box were very limited.  The setup wasn’t as straight forward as the instructions on the touch screen lead you to believe.  For instance loading the filament is glossed over. The nozzle types are confusing, why do I put the BB in the second slot and not the AA?

 

2) intermittent connection.  Despite there being a strong wifi in the location of the printer I will occasionally have the printer disconnect.

 

3) no way to remotely monitor the printer, can look at the camera if you are on the network - pretty useless for me.  I would love a way to look at this (when the printer is connected!) whilst I am not at home.

 

4) the camera is in an odd position, the secondary filament primes a big blob in front of it by default. I would prefer it to be angled down more

 

its a beautiful machine but I just think its a shame it has some short comings at this price point!  

 

 

[kmanstudios 1,120]

56 minutes ago, Bigbrit said:

The nozzle types are confusing, why do I put the BB in the second slot and not the AA?

This is an assumption that you will be using a support filament. You most definitely can put any other core in there for two color prints. That is why you got the  extra AA core....not just a replacement.

 

57 minutes ago, Bigbrit said:

Despite there being a strong wifi in the location of the printer I will occasionally have the printer disconnect.

From Cura Connect? This has been mentioned in other areas and is a known issue. One thing to try is to reboot your wifi setup. I had a situation where Cura Connect would not see either the S5 or the two UM3Es. Reboot solved that one. But, the S5 has been noted.

 

59 minutes ago, Bigbrit said:

no way to remotely monitor the printer, can look at the camera if you are on the network - pretty useless for me.  I would love a way to look at this (when the printer is connected!) whilst I am not at home.

There are ways to do this with a VPN. I am not the one to ask though.....

 

59 minutes ago, Bigbrit said:

its a beautiful machine but I just think its a shame it has some short comings at this price point!

I find this to not be an issue simply because it is such a solid printer in my experience. There are other issues thought that do raise that comment up to high validity. But, I just do not worry about the issues you mention. The other ones I do worry about. However, this does not invalidate your opinion....just a varied opinion on that statement is all.

 

The prints I am pulling are amazing in consistency and I like to make big things.....of course, that means a great deal more in expense when it comes to filament..... ?