has Xojo Inc. ever officially guaranteed (in 2020) a 10-year support for Api1 and Web1? or, today, does Xojo Inc. officially guarantee at least 5 years support for (…) how can you be comfortable with the fear that tomorrow your (…) application in production may break?
This ladies and gentlemen is exactly the reason not to touch anything business relevant with Xojo (or to be fair, with similiar scams or tools).
I cant say if it is the framework, the compiler or IDE. What I can say is that, even with paid subscription you are in that constant fear when working “on the bleeding edge” of technology. And I can affirm, the life-cylces in bigger companies, esp. in the producing industry and engineering branches are total far away from everything in the tech industry with already obsolete products on release and factually non-existent support because the whole tool-chain cannot be rolled back easily.
Slack and Discord both use Electron. You may not be a fan of how it does things, but it’s impossible to ignore. I’m pretty sure 1Password is Electron now too.
The question that really should be on the table is “why didn’t they pick Xojo?”
I actually bought that up with Xojo and they declined to answer. I also mentioned that perhaps they should contact 1 Password and ask if they’d be willing to share WHY those chose Electron and not Xojo. My guess is that my suggestion was ignored.
If I ran Xojo, my goal would be to get these companies to use Xojo, sure it would probably cost a lot in engineering terms, but the favorable exposure they would earn from it would be priceless. Which is why I believe that Xojo is in a maintenance mode, investing the bare minimum to keep it ticking over. Allowing customers with problems to walk, seems cost effective, but over the long run if you don’t have a growth mindset, your business isn’t going to grow.
Personally, I think the main mistake that was made was to make Web 2 only work with API 2, first because API 2 wasn’t completely hashed out at that point, but also because it made the transition that much harder. Had we been able to tell users that they’d be able to make a few layout tweaks, and then introduced API 2 later, things would have gone a lot smoother.
API 2.0 made Web 2.0 worse. It’s funny how the CEO still refuses to accept that API 2.0 was and still is a problem.
Even if he did, at his point I think it would be a BIG issue if they did go back, so i don’t think it could happen, even if Geoff realized API 2 was a big mistake.
Well realistically, it’s the lack of mobile support. 1Password absolutely needed iOS and Android, which wasn’t available at the time, and even today isn’t really available. Requiring the use of multiple projects is a huge weakness. Combined with the fact that Xojo’s options for code sharing between projects are absolutely pitiful, you might as well be using different tools between desktop and mobile. When it comes to one project that works everywhere, there are very few toolsets. There are plenty that come close, but not one have I seen that made me say “yep, this is it.” And it’s so frustrating because Xojo is so close, other warts aside. If they “just” spent the time to allow multiple products in the same project… well my app would be available on mobile in a few months. I’ve spent effort on it, but with each product requiring a separate project and no practical option to share my core modules, it’s unrealistic.
Xojo’s roadmap has some stuff I’m looking forward to, but not one of them are what I’d call great. I believe if Xojo wants to grow, the language needs to continue to grow and they really need to embrace their cross-platform strengths. As I said in another TOF thread, I want to see promises, closures, and async/await. I want to add null coalescing and traits to that list. What was our last language feature? If()?
FWIW I did come to finally realize HOW to do this - but now I dont work there
It would be a decently large undertaking - but DEFINITELY doable
So much would be affectedAutocomplete, as it is/was, would need such an enormous overhaul it might as well be rewritten to do lookups like the compiler does
And nearly everything hinges one compat flags
But - it should be doable
I’d even started doing some work in that vein but have no idea if those branches still exist or not with all the changes theyve made
