Dial 'M' for Microsoft's new programming language As if you haven't got enough languages already, Microsoft is about to give you one more: M, part of its Oslo development and service-oriented strategy. The company is today due to announce M, for building textual domain-specific languages (DSLs) and software models with XAML. Microsoft will also announce Quadrant, for building …
Some but probably only about 30% - and I'm a programmer; albeit one still maintaining 10 year-old Windows applications written in a 12 year-old language.
I do sometimes wonder whether anyone actually uses all this technology? It's not like MS sells any of it as it's all included in an MSDN subscription. To me, it's just more crap I have to wade through when looking for the Visual Studio disc in my MSDN folder. I've worked in IT for nigh on two decades, know dozens of programmers personally and have met many others but I don't know anyone who uses any of the crap mentioned in this piece. I suppose they must be out there somewhere. Probably in India where all the large-scale projects seem to be based nowadays.
In the real world I'm beginning to suspect, like me, most people are stuck maintaining and enhancing ancient systems (our oldest - and most important - was originally written in 82) that still do the job and that nobody dare replace because no one person can quite remember how it's all bolted together or what the knock-on consequences would be.
"It's a fresh take on an existing model and code re-use concept."
Doesn't sound like it. Domain-specific languages have been around since forever. Some of them even made it mainstream, like yacc/lex. These days, DSLs tend to be written *in* (not just compile down to) a general purpose language. There's no clear boundary between a DSL and a decent domain-specific library. As they say: Language design is library design, and vice versa.
Using XML as the underlying syntax is a dreadful idea because raw XML is about as readable as Reverse Polish or Backus-Naur. (Yes, it can be done, but no, it's not pleasant.) Therefore, you need specialised tools to convert the XML into something human-readable, so you suffer vendor lock-in before you've got as far as "Hello world". MS seem to speak of this aspect as a feature. (Yes, it will be cross platform, but you have to buy the IDE from us. Buwahaha!)
"I do sometimes wonder whether anyone actually uses all this technology?"
Who cares? With new languages there will be new books, new training courses, new certifications, new blogs, new magazines and new opportunities for me to scream with undergrad students who want to take a CS course to "learn the new hot cool language". Little twerps.
I've seen languages come and go, and my money is still on C, Java, and .NET. Everyone is always trying to make software development easier (or "democratize", whatever the hell that means) and all that ever happens is that you get more monkeys how think they can write code. Software developers are more productive then they were a long time ago, but the advances come from libraries and programming languages (garbage collection, string classes, automated memory management, ever expanding libraries, etc), not trendy "functional" languages, or BS marketing buzzwords.
They seem to be mainly for the large-scale, new projects. Used to work as the Network Admin for a small web development company who specalised in .Net. Did a lot of banking system for some very big high street banks, as well as one of the worlds biggest food companies and various telecoms providers.
Can't really tell you what UML is - but it was used a lot in my previous place. Visual Studio Team System was a big deployment that the coders, projects manager and testers liked quite a lot. Was around 2006 we started doing work with the new features on SQL 2005 too.
All about the money really. The tools are normally pretty good, and have a lot of worth. However the resource to learn about them is quite high. Therefore it's only projects where the big money is that'll pay for devs to learn about them - as the overall aim of the tools is to reduce the cost of developing. More out, less in etc.
But yeah - I imagine it's probably used even more in the big, big projects normally based in India.
Microsoft has only one purpose, to generate revenue. They have to keep producing stuff, which they can then use to justify the subscription fees they charge to the large corporates. And the corporate directors can feel comfort in the fact they have an 'innovator' holding their hand. Of course it's included 'free' in MSDN, of course it's useless for nearly everyone. But they can sell the idea to the non-techies that hold the cheque books, job done.
They don't produce software for the greater good. Or for technical excellence. Imagine if they had produced C. Then they sat back and said, "well, it's got everything it needs to have." They be out of business. So they would produce 'C, in egg shell blue', and then 'go faster stripes for C', 'upside down C' and probably 'Cloud C'.
Do you need any of those variations? Probably not. But if they sell the idea of 'doing C in a Cloud' to a muppet with some money, then they have business.
It's pointless to attack this in a technical sense as another new language. it's not about a technical language, it's creating something new to sell. Whether that's in a financial selling sense, or in the sense of 'look, we're not a stagnant company we produced this...'
"Therefore, you need specialised tools to convert the XML into something human-readable, so you suffer vendor lock-in before you've got as far as "Hello world". MS seem to speak of this aspect as a feature. (Yes, it will be cross platform, but you have to buy the IDE from us. Buwahaha!)"
And you're not struck by how familiar this sounds - w.r.t. Microsoft?
I expect the idea is to raise the level of abstraction so that, yes you have to use tools in order to make it readable but in so doing, you get more of the donkey work done for you for free (or rather for the cost of the tools). Compare with high level languages requiring the use of a compiler but giving you more productivity/power of expression/etc than assembly languages.
But because it's MS we are talking about, it's likely to be a world of pain that fails to deliver much more than a lot of work for a different type of programmer. HoHum
Perhaps Microsoft is creating a new language to hide the fact that they've failed miserably doing anything reliable with old languages?
Or is it that they don't own enough of the I.T. world and want a language whose very symbols they have trademark and patent claims to, so that no matter what anyone writes, they have the final say on whether it gets to see the light of day?
Personally, I don't care anymore. If it comes from Microsoft, then it's sole purpose is to lock people into Microsoft products. I don't play that game.
Ahh yes Microsoft "cross platform", so it'll run on Win2K (maybe), WinXP *and* Windwos Vista. *rolls eyes*
A "cross-paltform" product isn't cross-platform if it's tied to platform-specific products such as Visual-Studio and SQL Server, and especially if the remainder required to do ANYTHING on another platform is vaporware.
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2024
