You could draw a frontend easily. You could glue together fairly high level components trivially. You could connect window/component/widget clicks and actions to code trivially. Want a thing that sits in the systray, scans a directory for new files, reads lines out of them and spews them out somewhere over the network? Easy. Make a test frontend that calls a bunch of scripts and dumps the results in a grid? Done.
It had a bad rep because in a lot of ways it felt like a toy, and because a lot of bad software was written in it by people that didn't really know what they were doing. OTOH that also made it a really accessible first step for some folks.
(For those that don't want to click the link: (turing-unaware) race car drivers are programmers that know a single high-level technology very well, but are unaware of the things that the technology is based on. Somehow like Formula 1-drivers.)
https://sites.google.com/site/steveyegge2/age-racecar-driver
http://web.archive.org/web/20080325033938/http://www.stifflo...
By the way, Richard Stallman in response to this email posted me a short lecture on how in English you only use "You" capitalized to refer to deities (I spelled it this way in the email because this is the custom in Poland when referring to the recipient in letters) but ignored the questions and further attempts to get those out of him.
To the question "What do you think makes some programmers 10 or 100 times more productive than others?"
Guido van Rossum answers "Genetic differet brain structure." (sic)
Tim Bray answers "The surprising variability of the human mind."
I think they're basically the same answer, but Tim Bray's is the very American politically correct "hooray to everybody!" version. Given that the title of the blog post refers to "great programmers", Guido's answer could be read as "my brain is super awesome and yours is not". Which, I bet, is not what he meant.
I hate - really hate - working with it, but I have to acknowledge that PHP has made a lot of things possible that otherwise might not have been because it made it much easier for people to start doing web stuff.
- Excel
- Visual Basic
- Lotus Notes
- Microsoft Access
- Foxpro
For the company who is modernising their IT, perhaps "moving to the cloud", they have to replace a lot of the functionality that was built on these platforms (especially if the vendor doesn't support them any more, or wants to charge an unviable amount to do so).
What is the modern web equivalent to the "drag and drop programming" paradigm for simple form applications? There are the scaffolding frameworks for proper programming languages (Rails, etc), but they just simplify the process, rather than allowing non-programmers people to design their forms, and then having a little bit of workflow which either they, or their technical colleagues, add on?
Or these things an antique from a time before multiple front-ends were necessary, and when they were safely run on a trusted network alone, and there will never be a place for their return?
Delphi and C++ Builder, the simplicity of VB for laying out forms, with the performance of native code. These still exist by the way.
I think what Linus wanted to highlight was that VB was a great environment to quickly get results in the real world, something which I think he highly regards. It might not be the most elegant language to program in but it had an incredible IDE, tons of prebuilt components and looked exactly as it was developed after it was deployed.
> I have a soft spot for Andrew Tanenbaum’s "Operating Systems: Design and Implementation".
I too liked VB6/VBA. Actually a really productive language and development environment. Joel Spolsky describes his part in designing the language: http://www.joelonsoftware.com/items/2006/06/16.html
(And I don't think Tanenbaum/Linus have much problems with each others, anyway.)
On the subject: The problem with tools like VB is that they are hard to grow up in; something which both support the first steps ("baby language") and serious development.
Today Linus might argue that the scripting languages (Ruby, Perl, PHP, etc) fill this niche in the FOSS world. But with the web as GUI.
Linus disagreed on the microkernel stuff of Minix and decided to build its on. So, he surely likes the Tannenbaum book on Minix, otherwise he would not have used it as a forking point.
CPAN comes closest for me or the old NeXTSTEP object catalog, but VB was amazing for it.
I picked up a book (forgot the name) to learn programming as a hobby. This book had a project based approach to teach. Step by step it helped me write a tic-tac-toe program using Visual Basics. It was easy, it was fun, it was hands on. It helped me quickly relate to basic programming constructs like conditional statements, loops and branching.
It turned out not only a wise choice for someone who was experimenting with programming, but also a natural choice because, as soon as i looked at the books content and illustrations, i knew what i would be able to get out of this exercise and it seemed both interesting and achievable.
We used it for doing quick user interfaces and "shell" applications for servers. It was well suited for this; you could use much of the Windows API surface, and VB could host DLLs for doing heavy lifting.
Once you got over the "Ewww, I'm using VB" sensation and turned off the obvious brain damage (ON ERROR NEXT, for instance) it was fine. There have been many absolutely brain-dead and horrible languages in the world, but VB is not automatically one of them.
Doesn't mean I'd use it today though.
I wish I didn't know this.
This was an unnecessary comment and having read the original interview I don't really agree with it either.
Don't bring back that "Guido Van Rossum, for example, comes across as kind of a jerk" again!
Sad Comment :/