Cloud Foundry: Design and Architecture
Derek Collison discusses the goals, the design premises and patterns employed in creating the architecture of Cloud Foundry, VMware’s open source PaaS, unveiling internal architectural details.
The content has been bookmarked!
There was an error bookmarking this content! Please retry.
Posted by Jonathan Allen on Oct 04, 2011
As of version 11, macros will no longer be available in Visual Studio. This marks the first version in nearly a decade to not allow for crating ad-hock extensions to the IDE. Fully compiled extensions can still be created and new project templates are included to make that process much easier.
According to Microsoft’s usage tracking data, macros in Visual Studio are currently being used by less than 1% of all developers. This alone isn’t enough to cause Microsoft to drop the feature, the macro feature in Visual Studio has also a disproportionately high maintenance cost. Unlike most other features, the macro support has to be updated and exhaustively tested for each new version of Visual Studio. In theory one should be able to record and replay any feature via the macro IDE, which places a huge maintenance burden on Microsoft.
According to Matt Kaufman, the macro IDE hasn’t been updated for several versions. Firing it up, one will quickly notice that it looks like an old version of Visual Studio. More troubling is the fact that it still only supports Visual Basic. One cannot use C# or any of the newer .NET languages to create macros.
As mentioned before, the support for extensibility now includes several new project templates. The current list, which is offered under both VB and C# follows:
In addition to the language specific templates there is a generic extensibility folder containing:
Visual Studio vNext: ALM features for Agile Planning, Team Collaboration
Using ALM to Drive Business/IT Alignment
How ALM Supports Business Processes
In recent years, for me it's a surprising trend that developers don't customize IDEs. There are multiple strange consequences.
We know that average users don't customize settings either. Yet a lot of developers tell them "there's a setting for that, if you wanted it, why didn't you set it at the first place?"
Also, for dynamic languages, the IDE support in recent years was terrible. Once another php architect came to me and told me, I can't be an architect like him, as I use vim, and a real architect should use an IDE. I asked him: what features do you use of that IDE? Do you build custom modules? No, we use what came with it. Do you use refactoring features? No, it doesn't work in PHP in our IDE. Then what on Earth are you talking about? I'm full of vim refacroring macros, custom editor shortcuts and I regularly kill of antipatterns in the code using automated refactoring. Still he wasn't convincible...
(When I'm doing java, I prefer IntelliJ as it rocks, especially in refactoring which I love there; I don't have too many realistic options when I'm doing c#, of course...)
Also, developers tend not to build editors for their own domain-specific languages, or some even can't configure a custom deployment system, which surprises me especially in xml and cloud settings (the ability to build a visual editor to our yaml-based form and entitity descriptors, and one-click deployment w/ jira handling were the sole reason I started to standardize on netbeans, despite having terrible php/js support, and questionable java support)
So, today's developers are just users: they don't think into it that the tools they use are simply code as well, and they don't try to recognize their own use cases and improve their own situation, which is really sad and provokes further questions on the role and future of our profession as well.
There are explorations on bringing javascript-based editors to javascript developers, like Cloud 9, Orion, or the former bespin... We'll see if the problem lies in the familiarity of language or environment, or there are more serious causes to consider...
This story completely lacks context.
Besides being woefully outdated, Microsoft has hired Erich Gamma of Eclipse fame to work on various sprints to improve Visual Studio. Macros don't save nearly as much time as features like Eclipse's Mylyn project, and other timesaving features that better integrate a developer's workflow. If a developer wants "macros" for Visual Studio, then they have tons of third party tools that do the job better, such as CodeRush. Incidentally, CodeRush just announced it is ready for Visual Studio 2011.
It is not like you can't do keybindings in Visual Studio, either. There are even Emacs Emulation Mode plug-ins for VS 2010. It would be funny if people, voting with their feet, used Emacs Emulation Mode, or a VIM plug-in, or CodeRush instead of the Microsoft-supported product. The phrase "too clever by half" comes to mind for the macro IDE.
I understand that much of what people used the Macros IDE for can be done as an extension, I just think it's unfortunate that MS is forcing everyone to now write such extensions rather than use what already works in prior versions of Visual Studio. We make regular use of a handful of advanced macros and now have to consider the cost of converting these . . . if we "upgrade" to VS-11.
But the Macros IDE aside, can anyone tell me where I can now go to simply record and playback keystrokes in VS-11? Since this (legacy) feature uses the Macros IDE transparently in the background, it too is now gone. I have used this feature most every day in Microsoft compiler products dating back 20+ years . . . and now they've decided to take it away?!?
Really -- what is the replacement for this? If MS can supply us with a record/playback keystrokes extension, I'm fine with that; I just don't know if that's even possible.
Derek Collison discusses the goals, the design premises and patterns employed in creating the architecture of Cloud Foundry, VMware’s open source PaaS, unveiling internal architectural details.
Andrew Watson talks about the work of the OMG, where CORBA is alive and well (hint: in your car), UML and UML Profiles vs. custom Modeling languages, DDS and other middleware, and much more.
Sohil Shah discusses creating iPhone and Android enterprise mobile applications based on cloud services using the open source platform OpenMobster.
Paul Sanford presents the transformations supported by data throughout its life cycle, and how that can be better done with Splunk, an engine for monitoring and analyzing machine-generated data.
A common “best practice” for unit tests is to only write a one assertion in each test. I intend to question this advice by showing that multiple assertions per test are both necessary and beneficial.
John Rauser presents the architectural and technological evolution of Amazon retail websites starting with 1994 and ending with adopting Amazon Web Services.
Michael Stal discusses system architecture quality, how to avoid architectural erosion, how to deal with refactoring, and design principles for architecture evolution.
Every developer has had to integrate with another system, API or component. Tis article provides strategies to handle the change and for he separating system boundaries.
3 comments
Watch Thread Reply