Xojo 'Average User'

When I see Geoff posting what the ‘average Xojo user’ does I hear 'I don’t use this feature so why would anyone else?"

SMH, that seems to be the average response to something technical these days on TOF.


Geoff has totally lost control. He refuses to listen to his clients, he continues to “improve” the product by changing syntax without adding any value, he insists this is to create an environement in which “newbies” can thrive (when in fact it imposes knowledge that is difficult to transition to other languages)


Do you have a link?

A link to what? To Geoff making such comments? Just read TOF, that is about all he spouts these days

Yes, I don’t read all of TOF, so if someone mentioned a posting of Geoff, a link would be nice.
And I did search for it and look for recent threads, but can’t find anything.

1 Like

I think you have to learn to read ‘average Xojo user’ as as ‘I’
When a post has

the average user doesnt understand …
read that as
I dont understand


Things just make more sense that way and are much less hand wavy

I find thats often true esp if its geoff posting

I think they’ve deleted the entire thread as I can’t find it now.

The usual…

Same answer as the other clasic, “WE know Better”.

But the tool being in a very ugly path the last decade+ says other thing. Lot of “good intentions” went very wrong, lot of time WASTED pursuing a shorsighted vision.

1 Like

I do recall a thread in which subclassing was mentioned–it might have been an issue with API 2 conversion. Xojo’s reply was that the majority of their users don’t use subclasses. This was in a Pro/Testers channel. I couldn’t help but infer from this that the very people who could even read the thread weren’t of most concern to the company. I’ve personally been a satisfied customer overall, but this was disconcerting.

The thread had to do with the uint comparisons and how the compiler does some string things. I think Thom started it and at one point it had well over a hundred responses. Then it was locked. And I think deleted since I can’t find it now.

Was it Something funky about UInt32 comparisons?

Oh that thread is gold, Geoff talking about maintaining legacy issues just because, then he took the decision to break everyones code who uses TextOutputSteam with alternative encodings in 2019r2 and didn’t even bother putting it in the patch notes and he took almost 2 years to confirm it happened! Brilliant. feedback://showreport?report_id=59351

As for average user, I guess they are people that use about 2% of the features of xojo, so would be pretty much those that slightly alter the examples to do the bit of office automation that they need. Yes, I agree with you Bob, its used as an excuse to be an average niche product.

In that thread when I said I used unsigned integers … Geoff said I was was not the typical user… So i asked him why, as I am a “Citizen Developer”, he thought I was not a typical user… and his is reply boiled down to saying I was!

I remember scratching my head about that one, but I think the thread got locked before I could follow up on that.


1 Like

I’ve not reread the whole thing recently, but I think Geoff was implying that your ability surpasses that of an “average user”, ergo you aren’t one? Surely he didn’t mean that because your project uses a big number that you’re suddenly a different class of user?

And even then, when every one was telling him that the problem was the silent break of code, and I told that at least he should make a note in the docs telling people of the behavior CHANGE to avoid people loosing their minds on how somthing was working before…

And he closed the FC with a note about the problem being saaving with NO encoding and that was what made it to the docs. WTF?

As for a recen example… If i dont use radio butons horizontally no one must use them:

Maybe… but I’m not all that talented or knowledgeable … just persistent and have been using the product for 20 years…

Over the years I have seen a good number of “citizen developers” who can run circles around me…

Citizen developer does not mean someone who is clueless… those are the ones that are both new to XOJO and new to coding, as well as not earning a living coding.

So does Geoff think that the average Xojo user is someone completely new to coding and have used Xojo for less than a couple of years? I sure HOPE that is not the case!

I really think that thread was about Geoff skirting the real issue which is they don’t have compiler expert on staff these days like they used to. Granted there is less need for one since they went to LLVM… but not having one likely leaves gaps in the types of things they can efficently address (or chose to address) these days.


Well, I hope they don’t target their average user.

For me the target is more the people with Pro licenses, but only something like 10% of Xojo users may pay for a Pro license and subsequently pay for add-ons.

The average Xojo user may just have a small license, doesn’t use plugin, may never do remote debugging, builds only for one platform or even use Feedback app.

I know, I was in disbelief as well, I was very close to posting about it on TOF but I didn’t bother as it would be more use talking to the wall.

Geoff totally missed the point in that ticket, it wasn’t about no encoding, it was about write and writeline changing the encoding of the passed string even when someone has taken the time to force the encoding into something they want. So now all previous users will have undocumented errors (all previous examples are also wrong) and any users going forward who actually change the encoding will have to jump through an additional hoop of setting the encoding again when writing and changing it depending on the type of the encoded string which is neither simple or intuitive which I assume is the requisit for the “average user”. All to make things “simple”.

Almost seems simpler to create your own “ProperTextOutStream” class and have it just use a binary stream internally to write data out properly rather than mess with a busted text output stream

And yet, in the docs for Control Sets, we find this:

Really??? The “average user” aka noob, is going to jump right into creating controls on-the-fly instead of dragging them into a window for a simple, intuitive WYSIWYG RAD?

On one hand we have Geoff, advocating for noob-friendliness, and on the other hand we have the dev team, who always come across as computer scientists.