Comments on: Are third party dependencies evil? http://www.ytechie.com/2008/09/are-third-party-dependencies-evil.html Productive software development using ASP.NET, C#, Adobe Flex, and other technologies and tools. Mon, 26 Jul 2010 02:01:44 +0000 http://wordpress.org/?v=2.9.2 hourly 1 By: Will http://www.ytechie.com/2008/09/are-third-party-dependencies-evil.html/comment-page-1#comment-672 Will Fri, 12 Sep 2008 13:35:02 +0000 http://www.ytechie.com/2008/09/are-third-party-dependencies-evil.html#comment-672 Huh? Rhino follows KISS? There's at least three different ways to do the same thing in the latest Rhino release. MoQ is a much KISSier framework. Rhino suffers from code bloat held over from previous versions. Huh? Rhino follows KISS? There’s at least three different ways to do the same thing in the latest Rhino release.

MoQ is a much KISSier framework. Rhino suffers from code bloat held over from previous versions.

]]> By: Jason http://www.ytechie.com/2008/09/are-third-party-dependencies-evil.html/comment-page-1#comment-671 Jason Thu, 11 Sep 2008 02:27:22 +0000 http://www.ytechie.com/2008/09/are-third-party-dependencies-evil.html#comment-671 Odi, I think I'm in agreement with everything you mentioned. Your DataGrid is the first I would check out if I needed anything fancy. In fact, I had tried it out a little bit until I found out Microsoft had came out with one. In my case, I honestly only needed something as simple as possible. I saw yours a few years ago, and just based on the fact that you seemed to have one long before the competition starting thinking about writing one. I have no doubt that your grid will stay around and see great success! Odi, I think I’m in agreement with everything you mentioned.

Your DataGrid is the first I would check out if I needed anything fancy. In fact, I had tried it out a little bit until I found out Microsoft had came out with one. In my case, I honestly only needed something as simple as possible.

I saw yours a few years ago, and just based on the fact that you seemed to have one long before the competition starting thinking about writing one.

I have no doubt that your grid will stay around and see great success!

]]> By: Odi http://www.ytechie.com/2008/09/are-third-party-dependencies-evil.html/comment-page-1#comment-670 Odi Wed, 10 Sep 2008 21:58:02 +0000 http://www.ytechie.com/2008/09/are-third-party-dependencies-evil.html#comment-670 Awesome that you responded with a blog post, thanks. I agree with your conclusion paragraph, but have seen that developers often do not realize they have outgrown the built-in toolbox control, and end up spending a lot of time trying to work around the limitations. As far as reliability is concerned, sure, the stock textbox is ultra-reliable. But what about all the extensions you have coded around it to get it to do what you need it to? That often lowers the reliability. For example, when you suddenly realize that the WPF textbox doesn't have "masked textbox" functionality, and spend a few hours implementing and testing it, is it still ultra-reliable? You may have been better off opting for a 3rd-party control which has been better tested and been around the block. As for building your own DataGrid, I can't imagine that would make sense, especially in WPF where you should really spend a lot of time designing your architecture or you can get burned. In Jan. 2007 we had logged about 11K person-hours to design and develop v1.0 of our WPF datagrid, and am quickly estimating nearly 30K hours at this point. That's a lot of dollars put into this one WPF control, built entirely in North America by experienced developers. Sure, you wouldn't build all the features we did in your custom grid, but it still is a considerable undertaking, and there's risk in building software, you aren't guaranteed to do a better job, finish on time, or even achieve decent results. But like you said, if you can find a control that's "in widespread use, mature, and under current development" then strong consideration should be given versus building it yourself. How do 7 major releases and 12 service releases for a control less than 2 years old sound? And thanks to the free version, a huge number of users. Awesome that you responded with a blog post, thanks. I agree with your conclusion paragraph, but have seen that developers often do not realize they have outgrown the built-in toolbox control, and end up spending a lot of time trying to work around the limitations.

As far as reliability is concerned, sure, the stock textbox is ultra-reliable. But what about all the extensions you have coded around it to get it to do what you need it to? That often lowers the reliability. For example, when you suddenly realize that the WPF textbox doesn’t have “masked textbox” functionality, and spend a few hours implementing and testing it, is it still ultra-reliable? You may have been better off opting for a 3rd-party control which has been better tested and been around the block.

As for building your own DataGrid, I can’t imagine that would make sense, especially in WPF where you should really spend a lot of time designing your architecture or you can get burned. In Jan. 2007 we had logged about 11K person-hours to design and develop v1.0 of our WPF datagrid, and am quickly estimating nearly 30K hours at this point. That’s a lot of dollars put into this one WPF control, built entirely in North America by experienced developers.

Sure, you wouldn’t build all the features we did in your custom grid, but it still is a considerable undertaking, and there’s risk in building software, you aren’t guaranteed to do a better job, finish on time, or even achieve decent results. But like you said, if you can find a control that’s “in widespread use, mature, and under current development” then strong consideration should be given versus building it yourself. How do 7 major releases and 12 service releases for a control less than 2 years old sound? And thanks to the free version, a huge number of users.

]]>