What will come after the great CEO

22

A little over 6 years now

Created: 2020-01-19 07:53:24 UTC

23

And many (most?) of our predictions on the state of Xojo have come to pass in those 6 years and maybe even accelerated a bit. It’s become more expensive to use and less relevant to the developer community despite adding more targets. It’s clear that ‘quantity of targets’ is not equivalent to ‘quality of targets’.

6 Likes
24

With the advent of agentic coding, tools like Xojo and Filemaker will become entirely irrelevant. Agents working in more powerful IDEs, with more powerful languages and APIs solve many of the citizen developers problems. Tokens are a better investment than some random xyz license.

1 Like
25

Absolutely

  1. change event & property names
    Predicted - will make all old code irrelevant and useless to learn from
    AI’s proved that one and continue to every day

  2. fail to fix bugs
    Predicted - people will abandon the product for poor quality (coupled with # 1)
    While we have no sales #’s the severe slow down in activity on the forums surely indicates that is the case
    Also the lack of requests for consultants points to that also being the case (none in close to 6 months)

Any others ?

2 Likes
26

It can’t be more irrelevant. It can’t be more outdated. Compare it with IntelliJ based IDE’s. Raider, CLion, Android Studio. Only as examples. Since the mac world changed there is no real need for an outlaw anymore. That was realbasic. But you can build on mac for all targets today. You have all languages and tools on macOS like you have for Windows or Linux. This times are ended up and so XoJo has a real problem. Remembering when we wrote the first TCP stack for mac before near four decades. Mac was totally isolated. Windows started to have networking (Novell network was the first big one). Mac had nothing like that. In this environment the Idea for realbasic was born. At the end of the 90th real basic became commercial. I forgot, before it was named CrossBasic. But there is no big relevance today compared with the times when VisualBasic was released for Windows and was the first simple development environment. CrossBasic was this environment for mac. It was discussed. it had many users. The situation today is totally different.

27

Looking on the market: XoJo is entirely irrelevant. It is a seldom used tool compared to C#, Java, C++ and objective C or also Python. Looking on the amount of people using this tool it is dramatical. There are more successful companies in that environment with many targets. Like for example codenameone.

@npalardy the biggest problem is: they have changed and changed back. What the heck is this change of msgbox to messagebox? I have no clue. It is to make it simpler for beginners. They really believe that beginners are to stupid to learn: msgbox(). That’s XoJo. And with the renaming all learning of AI is blocked between API1 und API2. So the AI can’t learn. Cause it can’t separate between old and new API. That’s something they did not get. Instead of developing the mobile target to the end.

28

Your history is a bit off
Mac had AppleTalk and decent sized networks predated Windows & Netware
I worked on one in 1991 that spanned Canada and was connected round the world by 1994

The rest of your post is quite accurate
Beginners are, were a big target initially, but those people have largely moved to something else like Zero Code and Low Code tools
In that environment Xojo might as well be C# or C++ for a beginner

1 Like
29

Yes but mac-Windows networking was a b ig problem. There was in the beginning only rumbapcs as networking system for mac with novell networks

30

Yeah but there was no AppleTalk for Windows so they couldn’t join some pretty huge networks at the time :slight_smile:

Eventually they all settled on common protocols and by the mid 90’s where I worked had a global mixed network

31

That was in the later 90th. In the late 80th there was no connection possible. When mac started to be standard in printshops they had huge problems with networking. Today: it is one world, one big network.

32

Your history is incorrect (see AppleTalk - Wikipedia )
AppleTalk networks existed as early as 1985
Apple laser printers used it
AppleTalk was later reimplemented on top of TCP/IP and wasnt finally discontinued until 2009 in macOS 10.6

I used this a LOT and when AppleTalk over ethernet came about the company I was at used it
That switch was before I left in 1994 or so

1 Like
33

Yes but no networking between PC and Mac. That was a big problem.

34

Worst:

In 1990, I booted an Apple computer (computer cist was < $1.000, from Les Ulis, near Paris) on an OS located on a server at Marseille (1,000Km away) using AppleShare via internet and a standard phone line.

The apple is far better than the windows…

And remember from where Windows comes… (from Apple Computer Inc. “You give me a license for macOS and I create Word and Excezl for your Macintosh”)

1 Like
35

Precise the year. Apple computers weith Ethernet card exist from a long time. Read my previous entry (Ethernet on Macintosh existed for many years before 1990.)

1 Like
36

Not to forget: I spoke about networking between macOS and Windows / Dos. That was a problem. Today it is not anymore. Not more not less. It is no question that apple was able to do networking. But this two worlds existed without connection.

37

Not if you were a Mac only shop :slight_smile:
like this company was

1 Like
38

I’m not so sure. I get the sentiment. Perhaps yesterday’s/today’s programming languages will be abstracted away by prompts and .md files (which could be really cool, if more verbose).

Then will come the optimization and hand-tuning, because there will be times when it’s just faster to dive in and find and fix something.

I have this sneaking suspicion that RAD will stay because there is going to be at least some portion of the population that wants/needs to manipulate pixels without typing “move that box 50 px right…no, wait, 40…no, wait, make it a squircle with 20 px radii, and a deep red…no, a deeper red…how about blue”

1 Like
39

I miss working for Mac-only shops.

2 Likes
40

I never had that kind of luck. Always Unix, Dos, Windows and Mac around. Mac only would be nice cause that would not be any problem. But when your customer is a manufacturer of milling machines he needs CAD (Unix, Catia), DTP (MAC the good and wonderful Aldus Pagemaker) and Dos (for his billing software) and Windows for I dunno what ever. This environment was to run into one network. Unix and Dos was possible. But not Mac and Dos or Mac and Unix. That was a really hard time. On the other side: yes, Mac was okay for itself alone like the others also.

@mikey yes, this could be possible and I guess: it will be so. People like this “pixel perfect” design until they realize that on an other device … and restart. I have some of this kind of people around. For me it is not from interest. But … that mean nothing

41

I get it. It can be difficult, and often not worth the effort. We were doing this in the late 80’s and early 90’s with NetWare (SPX/IPX) at CWRU, and it was more than DOS, multiple flavors of *nix, and Mac. We also had two or three DEC OS’s (VMS and…something), a Cray or fifty, IBM, and other assorted OS’s. Was it trivial? No. Did it work? Yep. Did we have too many hypernerd students compiling Mosaic one minute and trying to diagnose why some prefessor was offline AGAIN the next? Yep.

At least on ios, we have not had any issues with apps being write once, deploy on everything in the same tier (i.e. every phone or every ipad). When we build for ipads, we are deploying on minis, airs, regulars, and pros, across several generations (we tend to retire ipads after five years). We just retired our ipod touches, but before we did, we didn’t redo anything between the touches and the iphones, even though the displays were very different. Scaling just seemed to work for most of the screens.

I’m surprised how little android there is in corporate clients in the USA. IT directors seem to very heavily favor ios, and, what is it, 70% of the phones are iphones? I get that it’s much harder to carry the layout paradigm across platforms.

Obviously when you need real reactive design, for web browsers and the like, or because you don’t want to have one layout for phones, one for tablets, one for web, and, if you’re using android, one for every manufacturer, you don’t lay out in absolutes. It’s still far easier to select several objects on a screen with your mouse and apply some property than it is to read through a .yaml or .xml or json file, hunting for the line that has what you want that has the property you need, for all the objects.

1 Like