Its notable that after nearly 20 years as a REALbasic/Xojo consultant Bob has moved into a new role at BlackBag Technologies. Probably not a bad move for him but still a significant change in the community.
Personally I have a pretty good idea what transpired to lead to this change. And while I might be disappointed things culminated in Bob taking a role there I also understand the motivation.
I have no idea if his new role will have him using Xojo or not.
But, the community has lost one of the better examples of building a consultancy using Xojo. Thats disappointing.
I wish him nothing but the best and all the success in the world in his new role.
Often in code you might see checks for NIL written as
if foo = nil then
and sometimes you might see
if foo is nil
Whats the difference between the two ?
In Xojo the IS operator compares the references. It doesn’t compare the contents of the objects in any way. It answers the question “do these two references refer to the exact same instance” – in some languages this would be comparing whether two pointers point to the same location.
= will do much the same IF you have NOT implemented operator_compare for your specific classes. However if you HAVE implemented operator_compare then you can get VERY different results as your implementation of operator_compare will be called, even if one of the items is nil. In the following code
if foo = nil then
operator_compare for the instance foo would be called with the rhs parameter as NIL.
You can try this for yourself very quickly.
Create a new desktop project. In that new project add a Class named myClass. Add a method to myClass, Public operator_compare(rhs as myClass) as Integer
Simply put a break statement in the code
Now in Window1.open put
dim foo as new myClass
if foo = nil then // you will hit your break statement in the class
end if
if foo is nil then // you will NOT hit the break statement in the class
end if
The advice here is to be REALLY careful about when you use IS or =, or other comparison operators like >= <= <> etc, as they could mean VERY different things. And without reviewing code you can’t be sure whether a class does or does not have operator_compare implemented.
Use IS when you mean “are these the referring to the same instance”
Use the others when you _may_ mean something else.
I was doing some reading on the weekend and came across some really good thoughts on how to make decisions in code – but not by using conventional if or select branching mechanisms or others that masquerade as flow control statements under other guises.
There’s a whole movement for “if-less programming”.
And it fits very well with using OOP and the capabilities of polymorphism in ways that make your code clearer and easier to maintain and extend.
Let’s say you have a class that represents vehicles. Your program permits you to have many types of vehicles with various capabilities. And somewhere in that class that represents a vehicle you might have code like :
Function calculateMaximumWeightCapacity() as Double
if ( vehicleType = "Transport" ) then
// we have to account for the fact this has 18 wheels and 5 axles
elseif ( vehicleType = "Truck" ) then
// we have to account for what kind of light duty truck this is
elseif ( vehicleType = "Car" ) then
// we have to account for what kind of car this is as a mini cant
// carry as much as a impala
end if
End Function
Yes this is contrived but you get the point
IF you have classes that display this kind of code then you probably should consider using polymorphism in a way it can make you life easier.
In your application, if you have this kind of code, then EVERY time you want to add a new kind of vehicle you have code to adjust. You’ll need to update every possible place this kind of selection goes on. And if you miss one you have bugs.
If However we used polymorphism to help us we could be sure that any time we added a new vehicle type we definitely covered all the required functionality AND that we didnt have to adjust a lot of code in many places.(1)
Instead if we had
Class Vehicle
Private Sub Constructor()
// this is so we make this class "abstract" and
// you cannot create new vehicles - you MUST create
// instances of subclasses only
End Sub
Function calculateMaximumWeightCapacity() as Double
// since we have NO way to FORCE a subclass to override this
// we put a break in here so we can figure out which subclass
// did NOT override this
BREAK
End Function
End Class
Class Transport
Inherits Vehicle
Function calculateMaximumWeightCapacity() as Double
// we have to account for the fact this has 18 wheels and 5 axles
return 30000 // just for fun !
End Function
End Class
Class Car
Inherits Vehicle
End Class
Class Mini
Inherits Car
Function calculateMaximumWeightCapacity() as Double
// we have to account for what kind of car this is as a mini cant
// carry as much as a impala
return 500
End Function
End Class
Class Impala
Inherits Car
Function calculateMaximumWeightCapacity() as Double
// we have to account for what kind of car this is as a mini cant
// carry as much as a impala
return 1000
End Function
End Class
.. and so on ...
And this gets rid of the need to use an IF to test what kind of vehicle it is to determine how to compute the weight. The specific subclass just does it knowing that its a transport, a truck, a mini or an impala.
Function calculateMaximumWeightCapacity() as Double
return self.calculateMaximumWeightCapacity()
End Function
Note that this code is much simpler to follow.
And when you need to add a new kind of vehicle you add a new class type, recompile and off your program goes because you havent had to hunt around and find all the places where you need to check vehicle types and maybe introduce new bugs.
Polymorphism is definitely your friend here.
footnotes
(1) there are some things that Xojo does NOT provide that would make life simpler. A means of forcing a subclass to override a specific method defined in a super class does NOT exist. IF we use a base class we can only catch that a subclass has not overridden a method. Some languages have a means for a base class to define a method that subclasses MUST implement. Xojo does not.
Its possible to use an interface to force implementors of an interface to implement a method BUT this MAY be the wrong approach for the specific problem.
I’ve now had use of a Pro+ license for nearly a year.
When this license type was introduced Xojo told us that the Pro+ included
Top Priority Support is just as defined on that page, we’ll bump your support questions to the top of the queue and we’ll review your Feedback cases first. And of course you can set up a call with us anytime too. Fast Fixes is as defined on that page as well. Because of the difference between bugs and what may be involved in fixing one, we can’t promise that we will fix “X” bugs for you. But we can promise to help you track down an issue in your code, get a bug you’ve discovered fixed more quickly or help you find a workaround. And yes, we did previously offer a meeting with Geoff but people did not seem to want or use this so we dropped it. Geoff is always interested in hearing from customers, if you want to reach out to him on the forum you can anytime. Or email hello@xojo.com to set up a call.
https://forum.xojo.com/conversation/post/429116
So far my experience is that this is a nice statement but the reality is different. As I write this I still have several cases awaiting review – some for several months
I’ve actually had to ask about getting these reviewed more than once yet there they still sit.
What Top priority support meant wasn’t clear to me what that meant so I sent an email asking
On the page for Pro+, which I now am a holder of (designated), it says Top Priority Support. We understand that some Xojo users need support issues resolved ASAP. With Xojo Pro Plus, you’ll get support questions answered first and bug reports will be verified first. Is that for all reports I submit while I hold the license ? Or just ones I specifically bring to your attention ? Thanks
https://forum.xojo.com/conversation/post/429116
And the reply was
Hi Norman, Pro Plus users cases are automatically flagged in Feedback, so you don’t have to do anything. A case is also flagged if you add your name to it, so you don’t have to be the creator. To be clear, it does not mean we will fix it right away, but it will receive a higher priority. If you encounter anything show-stopping please reach out and bring that to our attention so we can look into it.
I pointed out that I had several cases that had been sitting and I also wrote back saying
I’m not expecting that it will be fixed right away
I just have a number of cases that are still in the Needs Review state and have been for a couple of weeks at least
And got a reply saying
Ok, we’ll take a look.
That was in August and most of these cases have remained in the Needs Review state since then.
In Nov I again asked about the reports still in the Needs Review status and was informed that they would endeavour to get caught up. To date they are not. Some are marked reproducible, some reviewed and some still needs review.
The other thing mentioned in the post is “Fast Fixes” but it was unclear on how many are included with a Pro+ license. So I posed that question and was told
There’s no set number that is listed but 2 per year is what I would work off of.
I haven’t asked to use any of the ones included with the license I’ve been assigned but not having a number stated as part of the license is bothersome to me. Is it more for some customers and fewer for others? Does it depend on the nature of the Fast Fix itself? And what are the criteria for what is and isn’t a Fast Fix?
I certainly can’t answer that. Xojo also wasnt cleear on what might qualify as a Fast fix.
The last item in the list is being able to schedule a call with Geoff. As of today I have an outstanding request to schedule a call with Geoff. So far that has not been scheduled nor have I had any indication that it will be. The email reply I got back made no mention of scheduling a call and focused entirely on other issues unrelated to my request for a call with Geoff.
In my opinion I’m not seeing that holding a Pro+ license actually makes any difference in how fast the reports I file get reviewed or fixed, or that it has any influence on any of the other aspects it claims to have influence over. (1) As of this writing ~60% of my cases are “Open” – either needs review (29 of 396), reproducible (88 of 396), reviewed (92 of 396), or verified (15 of 396).
If I were making a purchasing decision today I would opt for a Pro license. I might get 2 of them if I really needed twice as many installs (Pro+ includes 2x installs). I would save the extra $600. Xojo could alter the Pro+ to permit 3x activations and remove this as a consideration for opting for multiple Pro licenses instead.
Things don’t actually seem to move along any faster because its a Pro+ user reporting it or not.
YMMV. Anyone else have a Pro+/Enterprise license and have anything different to report ?
(1) Note this could be simply because I am who I am and Xojo does seem to treat me differently than other customers. I cannot say why they do this. It just seems to occur and others have mentioned they notice this difference to me.
Some companies like to pose certain programming questions as part of their interview process.
I’ve run into a variety of these. They’ve been SQL related questions, code writing, and algorithm descriptions. Some were fairly straight forward. Some less so.
One of the best I heard was a series of “How would you write this code if you had to write it …” with varying restrictions put on things. And the code was pretty simple. Write code to count all the upper and lower case letters in a string of ASCII characters.
First write the code however you wanted. Loops, if then else, select case.
Depending on which branching style you used the first time then the follow on was to write it using the other. If you first used a massive if then elseif statement then switch to using a select case.
And of course take into account that Xojo IS not case sensitive with string comparisons.
And finally to not use a branching type at all. This requires some out of the box thinking. Literally “count all the upper and lower case ASCII letters without the use of IF, SELECT or any other branching control.” All you have to use is a loop (while/wend, for/next whatever you wanted)
This first two bits are fairly straight forward.
The third however is still one of my favourite “thinking out of the box” approaches.
Anyone care to take a guess at how to achieve that one ?
MacOS has been fully 64 bit since the release of 10.7 (Lion) in 2011.
Yet is has continued to support running 32 bit apps up until the release of 10.15 (Catalina) which removes support for 32 bit apps entirely.
There are ways to work around this like running an old version of macOS in a VM.
There are many apps that have not been updated for 64 bit. Things like AccountEdge (what used to be MYOB) wont be updated to be 64 bit. Old versions of MS Office need to be updated to newer ones as they too are 32 bit and will not operate on Catalina.
And anyone producing apps in Xojo should be building ONLY for 64 bit macOS. I’d encourage that even if you have to support old versions of macOS since all versions that Xojo supports are all 64 bit.
If you’re still building for 32 bit macOS you should update now and be building for 64 bit.
I’d encourage you to do the same for Linux as most modern distros are 64 bit.
The only target where you might need to still build a 32 bit and 64 bit version is Windows where you can get both 32 and 64 bit versions. While most machines sold today are 64 bit and installing a 32 bit OS on them doesn’t make much sense there may be legacy machines running a 32 bit version.
Xojo includes the ability for you to supplement the plist a macOS build creates.
You can’t overwrite the items the IDE will generate for you but you can add things like heys for App Transport security, Crash on Exceptions and even your own custom items if you want them.
The entire of creating and adding a plist is fairly simple.
Start a text editor. Notepad, textEdit, BBEdit or whatever one you happen to use (vi, emacs, nano, etc will all work just as well)
Make sure you create a “proper” plist that includes the normal header and footer. its just an xml file. The basic form is as follows :
<?xml version="1.0" encoding="UTF-8"?>
<!DOCTYPE plist SYSTEM "file://localhost/System/Library/DTDs/PropertyList.dtd">
<plist version="0.9">
<dict>
<!-- your additions go here -->
</dict>
</plist>
Then you do need to look up the specifics of whatever key it is you’re adding and how it should be written.
The really lazy way is to open Xcode and use its built in plist editor by selecting File > New File > macOS > Property List and then saving the file. But for folks on other OSes thats not going to work.
A lot of the keys like Crash On Exceptions are booleans which are written as
<key> THE KEY NAME HERE </key><true/>
or false instead of true if the item should be off. Others like Apple Event usage are strings as follows
<key>NSAppleEventsUsageDescription</key><string> YOUR DESCRIPTION OF WHY YOU NEED TO USE APPLE EVENTS HERE</string>
There are LOTS of possible keys that can be in a plist. Adding them at build time can make a tedious process a bit simpler.
Lots of times we might like to clean up our code by making sure we fix all the errors AND all the warnings as well.
And one of those you might find you get warned about is unused parameters. These might come about from implementing some interface for a nice consistent API but some implementors of the methods on the interface dont use all the parameters.
You can go around and read each warning and selectively insert
#pragma unused <whatever one the compiler warned you about>
Or you can actually be lazy and lie to the compiler and kust copy and paste all the method parameters and put a #pragma unused for each parameter. The compiler wont give you a warning or error about having lied to it even if you DID use the parameters after you said it was unused.
For example if you wrote a method like
Sub foo( bar as integer, baz as integer)
#pragma unused bar
#pragma unused baz
dim i as inteher = bar + baz
End Sub
the compiler will happily accept this and not warn about bar or baz as being flagged as unused
I’ve seen lots of discussions that talk about how certain bugs have been copied over from API 1.0 to API 2.0.
This is unlikely and mostly not correct. From what I can tell nothing was “copied” from one API to another.
Let me detail why I think this is the case with an analogy.
Suppose you had source code that you could compile into a dynamic library for macOS (a dylib), Windows (a DLL), or Linux (a .so). But in your Xojo application instead of having to write all the declares every time you wanted to call you dynamic library you created class, or module, that put a nice API on it that hid all those declares. A great example of this is something like macOSLib which hides a lot of the details of calling into macOS Cocoa API’s from you. Behind the nice Xojo API this presents are a lot of declares. But you dont have to deal with those. You just use the classes available.
Thats roughly analogous to API 1.0 – the classic API used from the days of REALbasic through to Xojo.
Now if you altered the Xojo code in all those methods and classes by renaming the properties and events in the Xojo API or added new ones, but did not alter the underlying declares, you would have a new, or modified, API. And if you brought all this code into existing projects you’d have to adapt all your existing code if you wanted to use this new API. But you would still retain all of the old bugs you had in the C++ code.
This is roughly whats happened with API 2.0
API 2.0 is a new API on top of much of the same framework that has existed for years. Hence there are old bugs still present along with new ones adding this new API has introduced.
Thats not to say the new API 2.0 work didnt alter any of the underlying code. There are some things like Folderitem which appears to have new more modern OS API’s in use on macOS (a very welcome thing) and a handful of other items.
But by and large API 2.0 sits on the same framework as API 1.0 did.
