Sorry… while on one hand not paying the subscription denies Xojo a small amount of revenue that could go towards “fixing” what I percieve to be “wrong”… but on the other hand, I am not one to buy products that do not meet the current day requirements. And I am sure I am not the only one.
Perhaps if I weren’t on a limited income (unlike a few years ago)… and my disposal income warranted it, but that is not the case
.[quote=458316:@Robert Livingston]Constructive criticisms and suggestions are fine.[/quote]
and alas, based on events of the past few days, and what I hear occurred during the Beta, these fell on somewhat deaf ears.
[quote=458282:@Robert Livingston]
Progress in the development of programming languages is difficult and always accompanied by the gnashing of teeth. The jump from Python 2 to Python 3 dragged out much longer than was hoped. But you have got to do something.
.[/quote]
Its a while ago since I did a Python upgrade, but I cannot recall it consisted of renaming large sections of code for no other reason that the creators fancied a name change… you are not comparing like with like.
At least with Swift (which has undergone similar, but much much smaller syntax changes)… Xcode would point them all out (yeah ok, Xojo does that too)… BUT… Xcode would suggest (95% correctly) what the change should be, and a single mouse-click updated each line… (yeah, Xojo does NOT do that)
[quote=458316:@Robert Livingston]Dave,
I do think that you should continue to pay your annual subscription.
Constructive criticisms and suggestions are fine.[/quote]
Why exactly should someone continue to pay for a product that no longer meets their needs? Why would a developer undertake a very significant about of work to upgrade a project to the new environment when it brings no material benefits? Now you may be willing to put the effort in in the expectation the good things will come somewhere down the line, but Im not. Im going to wait and see what is delivered in the next 18 months or so before I commit to that effort.
This is not the point I was trying to convey. It is fine, in my opinion, to leave the Federalist party to join the Socialist party. Who could object to that?
But to spend your time trying to tell the Federalist party what to do and what their platform should be all the while never voting for them and supporting the Socialist party might not sit so well with actual Federalists.
Anyway, I will back off after this note. S, in particular, has made many contributions to Xojo on this forum that were valuable to me. So for me, he has earned a lot of slack despite my disagreement with his current stance.
I personally am delighted by much of what I see in r2. But I was happy when Apple went from Classic Mac to OS10, when they moved away from Objective-C toward Swift, when USB-C was adapted, when Flash was abandoned, when 4D introduced more object-oriented programming. I was disappointed when the Xojo. frameworks were not pursued. I like subscription-based software.
As Cadillac suggested, the worst things in my programming life have occurred when some complex application or programming environment, in which I have invested lots of intellectual capital, dies. All the users of such applications, whether they like it or not, are tied together. So personally I encourage people to support the applications that I use/like.
To support the background work that needs to be done.
When my inhouse users ask me what I am doing I tell them “go away or I will kill You”… ok, funny… but really, sometimes changes are not visible as we all know.
Xojo team should have a roadmap with a vision what they want Xojo to become.
I agree with Robert… even looking for another backend, not supporting Xojo anymore is not a good solution; if I am going to will have changed by opinion next year…
And for me, as I don’t use Xojo’s GUI I don’t suffer from event renaming.
As a backend, glueing to MBS is simply done and straightforward.
At the moment I am learning Perl 6 (no, Raku!)… My bet is it will go up on tiobe index.
So, why not have n backends? JS → Perl 6 (websocket) → Xojo (database access)
And ui site Elm → JS
Umm - isnt that what members are supposed to do? Or have I completely misunderstood American politics and the big meetings are just garden parties?
And that Tim Dietrich stopped using Xojo (after he left Filemaker for Xojo) is a lot worse for Xojo than Dave (and me) not renewing and tells its own story.
@Robert Livingston Like your style man! API2.0 offers many upgrades that improve my code significantly and the IDE improvements seem to be ignored.
What I like:
DateTime is a marvelous improvement on Xojo.Core.Date which was an even better improvement on Global.Date (see I’ve been around a while :))
SelectSQL & ExecuteSQl remove the need for BindTypes - wow does that make safe db access easier.
Split has been deprecated and literal.split isn’t available, so now my strings are constants and are now localizable - yes a little (very) bit of work, but I’m being encouraged to become a multilingual developer
The IDE seems snappier and I like the new debug icon that points to the debug line much more obvious to us nearly blind old fellas.
I’m also finding muscle memory to replace Dim with Var is surprisingly easy .text to .value a little less so, but it will come.
What I don’t like:
This is a hard question, I’ve spent > 250 hours in this beta and it has become normal to me. Most of my personal library of work-around’s have been superseded, my entire db accessor class is now redundant, but that’s actually awesome - I’m not having to maintain it any more.
Sorry, spend some time, unlearn & relearn there are too many serious improvements here to take a break for me.
[quote=458355:@Wayne Golding]
SelectSQL & ExecuteSQl remove the need for BindTypes - wow does that make safe db access easier.[/quote]
They also impose a significant speed penalty especially in db’s where reusing a prepared statement can actually speed things up
They recreate the prepared statement each time - and you have to resort to the old methods to make it so you can do
[quote=458352:@Markus Winter]
And that Tim Dietrich stopped using Xojo (after he left Filemaker for Xojo) is a lot worse for Xojo than Dave (and me) not renewing and tells its own story.[/quote]
Yikes, yes! Wasn’t he the one who started an online Xojo club chat and all of that? The fact he’s left Xojo (at least for now) should leave them a bit worried, no doubt.
He prototyped a chatroom, but it wasn’t a permanent installation. If you’re looking for the club chat, the discord server is active daily and is a safe place to talk about Xojo. https://discord.gg/nqY9Hm
I agree to some extent with @Robert Livingston regarding supporting the company because I basically love the product; I’ve supported them since 2006 or maybe earlier. It’s just tremendously disheartening to me that glaring things remain broken in the IDE since RealStudio was abandoned (6 years now), while resources are focused on shiny new gimickry and change for the sake of change. It makes me wonder if the developers actually use the product - how can they not notice this stuff? Fields in the Inspector immediately de-select on mouse-up, necessitating a second click. Syntax help doesn’t work right. The project navigator never remembers its expansion status. Vast swaths of wasted screen real estate. Little things that make it a chore to use instead of a joy. It’s been requested by others in the past and ignored - fix the stuff that’s broken before adding new stuff. Can’t someone just be assigned the task of making a list of broken IDE features and fixing them? How hard would that be?
@Paul Rodman: and now you want to do re-development from scratch in an app that surely has >100kloc? How long do you think this would take? A couple of years?
My guess would be that you use outdated declares which should be relatively easy to fix. Large parts of the warnings can be fixed by doing search and replace.
@Julia Truchsess: I haven’t a clue why the Xojo persons don’t get the poor usability of the IDE. I use the syntax help every day. Is this so terrible to understand that I want to see the help after typing the first parameter?
Anyways. 2019r2 is there. The new API is waaayyyyy better than the awful namespace one. I hate unlearning something. But the new API isn’t too bad. Except for the silly event renaming and AddRow.
Version 5.1 of my app is done in 2019r1, version 5.2 is done in 2019r2. Going backwards and forwards in versions is just a recipe for disaster. You can’t find the small problems if you switch IDEs.
4484 warnings in the main project remaining. Sigh…
