Three fundamental questions

[whugemann]

As I told you, I intend to produce some basic tutorial videos for Magick.NET beginners in the near future. When arguing why one should start using Magick.NET (if you are not a C# expert), three questions arise

1. What are the advantages of C# in regard to other programming interfaces for ImageMagick?
2. For what type of projects you should switch from OS script languages to C#?
3. Why should you rather use Magick.NET instead of calling the ImageMagick command line tools in a separate process?

Especially for small projects, the latter might be the more straightforward approach for a start: If you are familiar with the command line tools, you don't have to find out how their Magick.NET equivalents work. I have asked some large KI models (especially ChatGPT) to write a Windows GUI solution that rectifies a photo by providing five lengths between four points picked by the user, and all solutions started magick.exe in a separate process. At least ChatGPT was however able to use Magic.NET instead, when asked for it. The KI code results were not perfect, but could actually be compiled to a working application.

In regard to the first point, I personally chose C# because I had to switch from Pascal to C at some point of my (small) programming career and have always missed Pascal's strict type concept. Although C# code runs on other OS, C# is probably mostly chosen by Windows programmers (?).

Answering the second point is easy for Windows programmers: Batch files are too limited and VBScript will be abandoned in the near future. Thus you have to switch to one of the various IM shells available, if you need more advanced user interactions, calculations, data structures or if you want to use the Windows API. This can already be the case in rather small console projects, as I am willing to proof. Any other reasons?

As these questions are of more general nature, I don't need a prompt reply, but I would appreciate if you could shortly comment on them sometime in the future.

[Insire]

C# is probably mostly chosen by Windows programmers (?).

I'd argue that C# and .NET are popular, because

• they are backed by MS
• C# is a modern feature rich language
• they run basically anywhere
• .NET is a rich ecosystem of libraries and tools (such as Magick.NET)

2. For what type of projects you should switch from OS script languages to C#?

Anything you like. But probably anything used longer than 5mins. Especially OSS and enterprise things

3. Why should you rather use Magick.NET instead of calling the ImageMagick command line tools in a separate process?

Performance. Launching a new process for each image processing task, is slower than launching one process and doing everything in there

[whugemann]

With exeption of 'backed by MS', the mentioned points are met by several other languages/programming environments, such as Python or Java.

Regarding point 2: Of course you can do anything you like in C#, but how would you argue when trying to convince a newbie to use exactly this programming environment?

In regard to point 3: Although there is probably a better performance when you process thousands of images, I personally have never encountered real performance problems with VBScript calling IM's command line tools in 20 years of use, even when processing thousands of photographs. And when dealing with dozens of photographs, I can't perceive any difference in performance between my new C# solutions and the old VBScripts.

I don't doubt that there is a performance gain, but it might have to be weighed against all the overhead that comes along with using a compiler language.

So basically, I am especially searching for arguments that address the novice user of Magick.NET and C# (like me). Or maybe I'm in the wrong place over here ;-).

I have considered myself to be pretty familiar with IM's command line tools and scripting, but I find it quite hard to switch to Magick.NET/C#.

[dlemstra]

I will pick up the conversation here because Insire their answers are similar to what I would say so I will add my comments here.

In response to question 1 and 2. I think it's probably decided by the place where you work and the programming languages that they are using. I have used many languages and for some specific situations one language would be better than the other but for a lot of situations it doesn't matter. So it's not easy to give a good reason why one should use C# instead of another language.

For question number 3 the advantage is that you can call the methods directly instead of going through the command line parser. And when you are using this library in an application that keeps running (for example a web application) the initialization of ImageMagick will happen only once. But if you have an application that runs and exit there is probably not a big performance difference.

I really wonder why it's hard to switch to this library. One of the things that I have done is to keep the method names close to the command line arguments. What can I (or maybe we) do to make this easier for new users of this library?

[whugemann]

Why it's hard to switch to this library?
I am probably not the best sample 'switcher', having worked with IM's command line tools for 20 years and then switched to the C# interface: If people know something by heart and then have to switch to something quite different, they always complain ;-).

The command line tools are still my point of reference: Before I implement anything in BAT, VBS or C#, I have always tested the functionality of the operations needed with the command line tools. This would also be my advice to novice users of Magick.NET: Test everything with the command line tools before putting it into code.

What can we do to make this easier for new users of this library?
I think I would help if one (you ;-) would explain some general concepts, for example, why the methods are so heavily overloaded (to reflect the flexibility of the command line tools in regard to their parameters). I guess that key values (that can be -set) translate to properties of the image. OK, every command line option (like -resize) translates to a method of the same, capitalized name – that's the easiest part of it …

Some explanations may be needed how you implemented methods like -distort, where you can have different distortion functions: Using '-distort perspective', you really have to see some code in order to figure out, what parameters are expected.
Some comments on which parameters are int, unit or double might also help. Why is it MagickGeometry(int x, int y, uint width, uint height) ? (Because x and y can be negative in some operations.) If you draw on an image, coordinates are however real (i.e. double) values.

If a newbie is not embedded in a working environment where other, more experienced people help him to get started, he needs some simple examples that get him started (which I will try to provide with some YouTube videos). Picking everything up from sample code works on the long run, but it's really tedious.

In regard to Fred's scripts which you translated to C#/Magick.NET:
That really helps, but you obviously assume that everyone using Magick.NET is very knowlegeable in regard to C#. I would prefere if there were some out of the box examples, where you can first watch what they do and then try to grasp how they do it. Fred's well-documented scripts and your translation of them were one reason that I could be convinced to go for C# instead of other shells.