So, what is the optimal time to wait before updating to 21?

[JST]

I generally do not like to be an early adopter, I have been scorched too many times doing that. And we have some issues with 2019 to suggest moving slowly, since you will never again open some designs if you have to go back.

So, how long are you going to wait before switching over, assuming you are NOT jumping right in?

There must be an optimal time in which anything that is a real problem will likely show up once a new program version goes live. (yes, even though there was a beta)

 

[Hunter]

I think it typically depends on two things:
1) How good this release was tested by the Alibre developers.
2) How many users adopt it early on and find bugs that were not tested for, or cases that weren't anticipated, or user specific hardware setups that trips Alibre.

If everyone waits, then the missed bugs will remain in the software for a longer time (and there will be bugs, no process in which humans are involved is error free).

Jump in, use it, and help us all?

 

[HaroldL]

For me it doesn't matter to wait or not, I updated to v21 at its "first release" and have now installed the latest release. Can't say that I've seen any big issues with the new features. The new constraints seem to be more robust in that can manipulate an assembly with linkages and it moves a lot smoother than the old constraints allowed.

Everyone will have to judge for themselves since you won't know if there will be issues until you do update. The optimal time for any issue to show up is when it does and you are least expecting it.

 

[dwc]

I update immediately, but do an archive of my designs first so I can go back if I have to.
Don

 

[DavidJ]

Generally NOT in the middle of an important or urgent project.

If collaborating with others, discuss and agree when to all update.

Make an archive copy of anything critical before updating (so you can go back if needed).

If you have access to a 'spare' computer, perhaps install there first and gain some confidence in a no-pressure environment.

If one of the new features will make your next design much easier, or if there is a fix for something that has been causing you grief, then there is an incentive to move sooner.


v21 doesn't involve any major changes of back-end technologies, just some updates of those. Hopefully that means less chance of unexpected surprises.

 

[Max]

I think it typically depends on two things:
1) How good this release was tested by the Alibre developers.
2) How many users adopt it early on and find bugs that were not tested for, or cases that weren't anticipated, or user specific hardware setups that trips Alibre.

I would add perhaps the most important one, which is what's in the release. Last few releases have had fundamental overhauls to major systems (graphics pipeline, ACIS/DCM/etc, for example) which tend to disperse random bugs thoroughly and it can be harder to find them all. This release is filled with relatively atomic features, as David points out above, in that they don't tend to touch many different systems, and so it is more likely it will be more stable.

 

[astroverted]

Certainly not today, as I am finding I cannot create even a simple 2D sketch (not tied to a part instance) with constraints applied without seeing either bizarre behavior (constraining a point at the intersection of two existing lines deletes one of the lines) or else the "inconsistent sketch" error message and refusal to apply the constraint when no conflict should exist. Grumble....

 

[JST]

Naturally not in the middle of something important or "rush". Although there may be no such "clear time"!

Certainly not today, as I am finding I cannot create even a simple 2D sketch (not tied to a part instance) with constraints applied without seeing either bizarre behavior (constraining a point at the intersection of two existing lines deletes one of the lines) or else the "inconsistent sketch" error message and refusal to apply the constraint when no conflict should exist. Grumble....

Ah HA! Will wait a bit......

 

[astroverted]

A further update: after multiple attempts, I was able to create my sketches. To get there, I had to avoid dimensioning or locking down entities prior to selecting them as intersection constraints to locate additional sketch construction points, whereas in 2019 I don't recall having to do that when making similar sketches. The disappearing line trick did not recur, however one attempt was plagued by line entities failing to appear after I had terminated them; in those instances closing the sketch made the lines I had drawn visible. Closing Alibre and reopening seemed to get me out of that particular rabbit hole. Weird.