Which release of Cura is STABLE without additional errors?

[3D4Me 1]

I gave up on trying to get a Cura update because every time you fix something, yet more problems are introduced. Stop releasing something with fixes that just gives more issues to deal with. Just release a version that has previously noted errors fixed and work on your upgrades at a different time.

I have settled on 2.1 because it does at least work. I paid £2600 for this wonderful printer but the software development is a joke and really lets it down.

I see the latest version is filled with errors as well, which is why I refuse to install it. But if there is a stable version out there, I'd like to know about it.

You should have this:

Version X - Beta

Version X - Stable (with somebody or team just working on fixes for that build and overwriting the Version X build with each fix)

Version Y - Beta (Version X Stable but with added features/improvements)

At least that way we could stick with a stable build if we choose to so that our workflow isn't interrupted with problems we hadn't noticed until ten hours into a print, or whatever.

Seriously, people paying thousands of pounds deserve better than this. Great printer - software made by amateurs, it seems.

Get your act together... PLEASE!

[Dim3nsioneer 558]

While I find words like "amateurs" a bit harsh I like the idea of detaching bugfixes from new features. So a logical next step would be Cura 2.7.1.

[burtoogle 516]

I think what google does having devel, beta and stable versions of chrome is good because then those people who want to live on the bleeding edge can and by the time the stable release is updated, there's a good chance that most of the worst bugs have been spotted.

As a contributer to Cura, I think one of the most difficult aspects is getting any new code or even bug fixes sufficiently well tested before it gets exposed to the masses. A real difficulty is that slicers are expected to cope with a very wide range of input due to the many different models/setting/etc. that users throw at it. Personally speaking I simply cannot test the stuff I produce against a wide enough range of inputs to be confident that it will work correctly for everyone. That is why it would be very good if development builds where made available on a regular basis because then the developers would get some feedback before it gets to the major release stage which is what happens now. So, for example, 2.7 has some fresh bugs (not mine, honest) that weren't spotted before the release. They probably would have been spotted if they had been incubated for a month or two in devel and beta releases (that assumes that brave people would actually try those early releases).

The bottom line is really that people have to test the fuck out of any beta release to catch regressions and other problems. If the testing doesn't happen, the bugs will get through.

Well, that's my take on it.

[Dim3nsioneer 558]

These are a few very good remarks! Having more than "just" one Beta release before release would certainly do something good to the code.

On the other hand it would probably lead to a longer interval for stable releases. It would be interesting to read some more thoughts about this here.

[darkdvd 988]

Maybe it would be good to give back information on how to trace the bugs that are found on CURA (via github)

For my part, I do not see which bug is blocking or so important that it prevents tu use CURA (I'm talking about version 2.7)

[3D4Me 1]

While I find words like "amateurs" a bit harsh I like the idea of detaching bugfixes from new features. So a logical next step would be Cura 2.7.1.

No, they are not amateurs - far from it. But it certainly 'seems' that way given that every release has additional bugs. Very frustrating.

[DaHai8 74]

I've been a developer for 30+ years, and rarely its the programmers who are pushing out new releases - especially at Cura's feverish pace. Perhaps upper management needs to made aware the damage being done to Ultimaker's reputation having so many releases back-to-back with continuous bugs.

Its just not that feverish with other slicers:

Slic3r - 2+ years

Simplify3D - ~1 year between releases

Crap(ft)ware - 2+ years...

So, my 2c: Push back to management to slow down. 1-2 stable release per year. One team working on bug fixes for each release, and the other team working on enhancements for the next release while incorporating the bug fixes of the previous one.

There is no way any Q&A team can sufficiently test Cura against all the current 3D Printers on the market. That's where the community comes in and bug-fixes for stable builds over time. There is just not enough time to do all that when new versions are pushed out every 2-3 months - or quicker.

[darkdvd 988]

At the risk of repeating myself: I do not see in CURA 2.7 bugs sufficiently important to make this version unusable.

Do not forget that CURA introduces in every version in addition to the corrections a significant number of new functions.

I hope, for example, that I do not have to wait a year before I can use a function to move the position of the Infill...

In computing, there are only programs that do not evolve that do not have (or have no more) bugs.

[ahoeben 2,012]

> I have settled on 2.1 because it does at least work.

As someone who has worked on Cura 2.1, I would advice you to use at least Cura 2.3.1. Unfortunately you don't mention what issues you have with the later versions, so I can't offer advice on how to get newer versions to work.

[nallath 1,124]

There is no QA team on earth that can test Cura with all configurations. If I look at how much resources it takes to "just" test the printers that we have now (2 full time on Cura and a number of System testers per release), it's just not going to work.

I very much doubt that there is a bug that has lived in Cura since 2.1. The bugs that do end up in a final release are not so big that they "ruin" prints (eg; Anything that damages printers or guarantees that a print fails are labled as a "blocker" with us, which means that we won't release if it's still in there).

Most of the bugs that are found in the beta are fixed. We don't release a new version, as most bug fixes in the beta tend to be low impact. If we release more beta versions, we noticed that each version gets tested less (People losing focus?). So we need to "sugar coat" our minor bug fixes with new features.

As for helping you; You're giving very little info on what your issues actually are. Could you write down a number of issues that you have or is it just that because some people don't like a new version that you just don't want to take the risk? If the second option is the case, I'm going to have a bit of bad news; There are always going to be people that don't like change, so any version is going to get annoyed people (who can be quite vocal about it).

As someone who worked on pretty much every version of Cura, you're probably better of with upgrading to 2.3.1. 2.1 was okay, but it had way more issues than 2.3.1

As far as I've been able to see; 2.7 is pretty darn stable. Granted, it still has a long way to go (Speed improvements being high on my personal list), but i definitely would not call it amateurish or unstable.

All that being said; I think it's smart to consider having LTS (Long time support) versions.

[burtoogle 516]

All that being said; I think it's smart to consider having LTS (Long time support) versions.

I think that would be very good. With, say, 2 years between LTS releases and only major bug fixes triggering sub-releases, users would benefit from the stability while those who like to live more dangerously could use the non-LTS releases. Even they don't need to be more frequent than 2 per year (IMHO).

[nallath 1,124]

It will have a big impact on our overall productivity though; It will require more testing and more development. So it's not a decision to be taken lightly.

[darkdvd 988]

Cura works very well (2.7) and I see no reason to use a previous version, especially with an Ultimaker machine.

I NEVER had print failure due to a slicing error in CURA: never.

When I had "problems" with CURA, I always found a way to get around the difficulty, changing settings in CURA or changing the object to print.

For me, 3D printing is not plug and play. If the object is bad, the printing will be bad.

We are talking here about printing a solid from a three-dimensional structure drawn on the screen of a computer, not to reheat a coffee in a microwave oven!

Ignoring a technique often leads to errors of judgment: if I have not understood the function, I will always use it in the wrong way and the mistakes will always be the same ...and in the end it will be the fault of the software.

(Excuse my bad "google english")

[3D4Me 1]

2.6 was the last version I tried and I think the introduced error then was a wrong nozzle size. Prior to that it was something with only part of the moddle appearing.

If 2.7 is relatively stable and has't reintroduced a massively long print time in some of the previous incarnations, I'll give that a try.

I agree that feedback from users is essential for improvements to happen but it is a serious flaw to just bundle those fixes in with another update that in itself has other issues.

I use the Ultimaker 2+ Extended, by the way.