Microsoft trips on Visual Studio Lightswitch Microsoft's Visual Studio LightSwitch, just released to beta, is a new edition of Visual Studio 2010 and will become the next step up from the free Express. In other words, it will be paid-for but cost less than the existing Professional version. Do not be fooled though: although this is a low-end tool it is one with high …
After a few slow years MS has shown why it took over every office in the land.
It's not developers who are important (although they are nice) it's people who think they are developers! All those middle managers defending their own little empire need their own apps and microsoft gives them that power.
I pity the developer who has to come in years down the line and turn this now 'mission critical' app into something that can be supported and extended but I guess it can't be as bad as excel and access.
"Some two years in the making, it aims to simplify the building of business database applications, particularly at the departmental or small business level, but to do so in a way that respects Microsoft's current ideas about best-practice software architecture."
Wouldn't it be possible to have a way of flagging articles like these in El Reg, or keeping them in a separate section, so that we all know which ones are pure unadulterated promotion - without any critical analysis - and save ourselves the effort of reading them, please? I thought this was The Register, not an in-flight magazine for gullible executives.
"Some two years in the making, it aims to simplify the building of business"
... the clue is in the word "aims". I'll add that I'm not sure any tool which generates 4500+ lines of code gloop can be described as creating a best-practice design; but it may still be better than tools like VB which positively encourage bad design.
I still think it is an interesting effort. One of the points not brought out in the article is that LightSwitch does model-driven development under the covers; it is even conceivable that some future iteration might target different platforms - HTML 5 for example - from the same model.
Whether it will survive long enough to get there is an open question though.
Tim
As a Microsoftie Dev I loaded it up to take a peek, for me I ran into the "wanting to customise things further" wall too quickly. I will persevere further just to make sure I know what it does, because even though it doesn't appeal directly to me, I know at least a dozen people that rely on Access/Excel that cannot afford a single full time dev that would benefit hugely from this approach. Now I just have to convince them to take a look.
I just know this list of comments though is going to be from Pro Devs exclaiming that it is rubbish because it is by Microsoft or it is too restrictive, when in point of fact it is not even targeted at them.
It looks like you met the classic problem with many RAD tools - they tend to make easy things easy and tricky things impossible.
I think the main objection from "Pro Devs" (of whom, I confess, I am one) isn't the provenance of tools like this, or the restrictions they impose. It's that they can land a company with an unreliable mission-critical application which has in fact cost a lot to develop and will cost a lot more to maintain, especially if the person who built it has moved on. Professional developers spend only a small proportion of their time making applications work. Most of their effort goes on "proving"* that they do what is intended and making sure that anybody with suitable skills can maintain and extend them.
* Logically, you can never really prove it, but you demonstrate to a high level of confidence.
Yet another "tool" to make simple DB driven applications - you'd have thought they'd have worked out it is not the building of database applications that is the problem - it is the design and logic required to make it work.
While things like MVC make GUIs easy - building the underlying relational db structures and processing logic still require knowledge, skills and experience that has yet to be replaced by some "smart application" (or tbh most large outsource providers but that is another story)
... in with Access. They target very different needs, which is why Access is part of Microsoft Office. And what are the real issues with something like FoxPro, apart from not being trendy due to the received wisdom of 'bloke said, down the internet' ... ? Well, there will never be a 64-bit version and the locking mechanism when using it's native DBF file format is increasingly flaky in the light of things like SMB2.1.
So to Lightswitch - yes it does fill a need, absolutely. Plenty of people just want to create solid, bespoke LOB applications without worrying about endless low-level faffing with ORM tools and the like, or getting involved with a framework like Strataframe.
Problems? Well, if they want to shift people off things like FoxPro they should provide an Entity Framework provider for it. Also, what about reporting? Why is an easy-to-use report designer always an afterthought ?
This is a good thing. There's a cottage industry for developers working on Excel/Access applications for large and small enterprises. Tools such as this can only create more work within this industry. There's no law that says a developer must always build to the latest methodology. Sometimes in business you need solution that just works and serves the immediate needs of the business.
As a "Pro Developer" I intend to use this to write the naff little bits of your app that just need to update that table over there every now and then. e.g., customer or address records, product lists, status codes. This will save hours of development time, or as is often the case, not having an app at all but relying on the IT department to write UPDATE statements.
Thank you Microsoft, this is a good product.
And I can think of many users in the business who write quite large Access applications, who could use this against our SQL server and have a app that had record locking, and works over a WAN/VPN. This would also mean I dont have to write DTS packages to get data into OLAP cubes or just the reports server.
As for "built in reporting" SQL Express 2008 comes with the Report Server/Builder. Job done.
Access is hopefully dead, long live LightSwitch?
Oh and the licensing is cheaper than buying Access for 20 users who dont actually design anything just use a couple of Access forms someone else wrote.
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2024
