Welcome back ActiveX (Silverlight the Newest Silver Bullet)
I know I am in the minority here, but I am not at all excited to see the way things are shaping up. ActiveX is making a come back, and I am dreading it.
As an application developer I was one of the first out there on the web. I was building applications for Internet Explorer 1.0 on a MAC Performa using Claris Homepage, that I distributed with a disk because most people didn't have the internet. Then the boom came and off we went into the whirlwind of brochureware sites, crazy online applications, and the birth and death of the Fat Client and dll hell. Then came the browser plug-in movement, and the DHTML movement, during which we promised fat-client on the web and actually thought those technologies would be great for building complex applications found on a mainframe for massive enterprises. Uuhg. What were we thinking?
The last 6 gigs I have been on have had a web and a smart client (RIA using winforms for those that are offended by the old MS language 'Smart Client'). Every time we used what made sense to use. I have found myself watching a world of silver bullet pushers. Meaning, they still believe one is coming and intend to force one to exist. Their environment has been either all applications will be thin client, or all applications use this or that technology, or everyone will use this language. There is no thought going into the individual needs of each project.
This kills morale and productivity by forcing people to make a circle fit into a square all the time. With the next silver bullet, appropriately named Silverlight, I am dreading the day when I watch the silver bullet chasers put out the rule that it will be used on every project no matter what the needs or requirements are of a given project.
Don't get me wrong, I think it has it's place, just like AJAX, ASPX, FLASH, WPF Clients, and Windows Forms. I also like what I am seeing with Silverlight, especially the 2.0 stuff. But I have also been down the ActiveX path and I am having flashbacks of having to implement functionality you aren't allowed to use in a browser environment, and then spending weeks trying to create the work around because we MUST use the new silver bullet.
Getting the context right of a new technology is not easy and it takes some trial and error. Running a proper architectural business cycle promotes the use of proof-of-concepts (POC). With all these new technologies coming out there has never been a time when POC's were more important. But if unforgiving policy is setting the technology required on a project, the POC's don't hold much weight.
Will projects take the time to put Silverlight into the right context? I don't know. But for those of us who do have control over the environment on which we develop for, and there are no MACs or LINUX boxes there, we should be dismissing it, and recognize that it is a great technology for those that don't have the luxury of controlling their environments (those that should be using a browser, because they must).
There are some of us still around developing for the windows platform and I hope MS does not forget that. Smart Client is a dying MS term, I just hope the movement is not dying as well. We don't need AJAX, ASPX, or Silverlight on 80-90% of the applications we are using those technologies on (I am not talking about a worldwide scope, I am talking about the one currently around me), yet the lack of understanding and fear of fat client dll hell keep people using them inappropriately. There still are a lot of places that do not need to limit the user's experience by forcing that experience through a browser.
We keep trying to mold that perfect silver bullet, and the promise of its ability to kill our werewolves keep us buying into them.
At least the new version of ActiveX won't have us developing using the ATL (Active Template Library)…. The ATL experience equated to days of mental root canals….
As an application developer I was one of the first out there on the web. I was building applications for Internet Explorer 1.0 on a MAC Performa using Claris Homepage, that I distributed with a disk because most people didn't have the internet. Then the boom came and off we went into the whirlwind of brochureware sites, crazy online applications, and the birth and death of the Fat Client and dll hell. Then came the browser plug-in movement, and the DHTML movement, during which we promised fat-client on the web and actually thought those technologies would be great for building complex applications found on a mainframe for massive enterprises. Uuhg. What were we thinking?
The last 6 gigs I have been on have had a web and a smart client (RIA using winforms for those that are offended by the old MS language 'Smart Client'). Every time we used what made sense to use. I have found myself watching a world of silver bullet pushers. Meaning, they still believe one is coming and intend to force one to exist. Their environment has been either all applications will be thin client, or all applications use this or that technology, or everyone will use this language. There is no thought going into the individual needs of each project.
This kills morale and productivity by forcing people to make a circle fit into a square all the time. With the next silver bullet, appropriately named Silverlight, I am dreading the day when I watch the silver bullet chasers put out the rule that it will be used on every project no matter what the needs or requirements are of a given project.
Don't get me wrong, I think it has it's place, just like AJAX, ASPX, FLASH, WPF Clients, and Windows Forms. I also like what I am seeing with Silverlight, especially the 2.0 stuff. But I have also been down the ActiveX path and I am having flashbacks of having to implement functionality you aren't allowed to use in a browser environment, and then spending weeks trying to create the work around because we MUST use the new silver bullet.
Getting the context right of a new technology is not easy and it takes some trial and error. Running a proper architectural business cycle promotes the use of proof-of-concepts (POC). With all these new technologies coming out there has never been a time when POC's were more important. But if unforgiving policy is setting the technology required on a project, the POC's don't hold much weight.
Will projects take the time to put Silverlight into the right context? I don't know. But for those of us who do have control over the environment on which we develop for, and there are no MACs or LINUX boxes there, we should be dismissing it, and recognize that it is a great technology for those that don't have the luxury of controlling their environments (those that should be using a browser, because they must).
There are some of us still around developing for the windows platform and I hope MS does not forget that. Smart Client is a dying MS term, I just hope the movement is not dying as well. We don't need AJAX, ASPX, or Silverlight on 80-90% of the applications we are using those technologies on (I am not talking about a worldwide scope, I am talking about the one currently around me), yet the lack of understanding and fear of fat client dll hell keep people using them inappropriately. There still are a lot of places that do not need to limit the user's experience by forcing that experience through a browser.
We keep trying to mold that perfect silver bullet, and the promise of its ability to kill our werewolves keep us buying into them.
At least the new version of ActiveX won't have us developing using the ATL (Active Template Library)…. The ATL experience equated to days of mental root canals….
posted by tadanderson at 8:08 PM
0 Comments:
Post a Comment
<< Home