FMPro layout mode lockup has been persistent since 20, and remained in 21
The other items are less prevalent, but are widely reported.
There is also the issue that after a deployment, a privilege set is commonly screwed up, and has to be rebuilt.
Also not a bug, per se, but the order of execution of user accounts in Linux has been reported to be alphabetical, not listed order (not tested so hearsay).
There has been a couple more serious bugs in 21 on Linux and Mac, that have caused some of the major FM shops to revert clients to 20 (although same bug reported on 20.3.1 as well), but I've not experienced those issues. The one on large scale data loss is a biggie. Claris has not yet been able to replicate AFAIK
Thanks @Kirk - crashes and other misbehaviors I can live with, but data corruption issues seem scary. I think I'll stay with my current versions for now and hope FM gets those critical issues fixed (or comes up with settings that prevent it from happening).
I am loathe to upgrade (client applications) to current versions of FMS, FMP, and OS until it's really truly fully stable and reliable. For personal tinkering and testing/learning it's fine, but I find that "selling" clients on the concept of FM licensing (and associated infrastructure) is a big enough challenge without dealing with quirks and "breakdowns". Most of our clients are patient and understanding (to a point), but failures or interruptions in business operations erode confidence in the platform, and potentially extending on to us.
Pushing for a fixed schedule of version upgrades (as well as with Mac OS upgrades) is absurd IMHO. My users care more about stability and reliability, not new shiny things. That said, security updates are of course appreciated.
Always give the client the cost comparison of using SaleForce and Quickbase - the other 2 similar appdev environemnts - FileMaker is way cheaper for far more deployment flexibility.
Of course, @Kirk . My clients don't consider SF or Quickbase. Different use-case. Typically the request for comparisons is with custom web-apps. We can do both, but prefer FM when appropriate. The "selling" of FM has more to do with updates, upgrades, OS compatibility, etc. Plus some of our clients are smaller and balk at the first spin-up costs when they think they don't need 5+ licenses. (It almost always changes to more licenses after they interact with a well-designed solution.) It's all fine; sometimes we provide a web app and as they request more features they admit going with FM would be better.
Most of my clients come from a bad experience with an browser-accessed solution via web server, so the concept of FM (and similar) is new to them.
IDK - SF's dev environment and Quickbase are both direct competitors to FileMaker. Similar tools, more brand recognition, fewer deployment options and significantly more costly. I've run into both, but mostly in larger client opportunities. Smaller orgs don't have the visibility into what tech exists.
And in any case, I never sell "FileMaker" per se; it is always cost/benefit and the underlying platform, in most cases, is completely irrelevant.
My larger clients don't care about platform (at all), but they do care about interruptions of business operations. Smaller clients are definitely more cost-sensitive and tend to underestimate the benefits of a well-designed solution.
I don't work with clients who have shopped different platforms; my environment is one where clients ask me to solve problems for them specific to my areas of expertise, both in data management solutions and certain business specifics. Plus, most of my clients are modest in size; generally fewer than 150 active users, with a couple as small as four (4) users (for now).