Does Platform Matter?
I've been thinking about web development platforms a lot lately, and wondering how organizations without access to a lot of technical resources distinguish between platforms when faced with proposals that might span quite a range.
In my "Build vs. Buy" post last month, I talked about how mature some kinds of web platforms are (such as CMS, CRM, and others) and how one should exhaust all "buying" options before considering building. But if you choose to build, you will invariably be faced with questions like:
In my "Build vs. Buy" post last month, I talked about how mature some kinds of web platforms are (such as CMS, CRM, and others) and how one should exhaust all "buying" options before considering building. But if you choose to build, you will invariably be faced with questions like:
- If a vendor is suggesting custom development, does it matter whether they are using frameworks such as Ruby on Rails or CakePHP, or not?
- Does it matter whether the application is written to run on UNIX (most often LAMP) or Windows?
- Does it matter whether they will be writing their code in PHP, Python, Ruby, Java, or .Net?
- Maturity: how mature is the framework/language? Has it been around a while? Are there lots of users and developers? Examples of the other end of maturity are the proprietary language/platform ColdFusion, which is ebbing in use and popularity, and what used to be the web programming language of choice, Perl, which has now been relegated to the trash heap (for some very good reasons.)
- Support: will you be able to find support for your project beyond just the vendor who will implement it? Are there user communities, and other vendors and consultants who use it? How big is that community? How many others support it? This might be an issue when selecting a vendor who a platform in a language (like Ruby or Python) that is not as widely used as a language like PHP.
- How does this jive with other systems that you have? (For instance, if you are a completely Windows shop, and you are implementing a web project internally, it might make a lot of sense to choose a Windows platform based on .Net.)
- What kind of scalability/robustness do you need, and will the platform/framework/language handle it with ease?
- For custom development: these days, the availability of frameworks such as RoR and Cake (there are many others) means that development time can be significantly shortened. If a vendor is suggesting custom development without using a framework like this, it might be worth asking why (they might have built their own libraries over time, etc.)
Permalink  
![]()
6 Comments:
Laura - spot on. Even just a few years ago at NPower - we spent a lot of time considering/delving/probing into the platform question before starting a project. In the last couple of years, though -I think it has become much more difficult to paint yourself in a corner. Even though NPower has settled on Plone and Salesforce - I know that I could use different tools and still help nonprofits with their web and database needs. Support matters more, I think.
I tell my customers to consider when they last bought or leased a copy machine. I haven't seen much innovation or clear winners in that area in the last handful of years -but some vendors are better than others, some have slightly better hardware (but just slightly) and most customers are looking for a great service contract and a great price. All of the copiers do a fine job copying!
Thanks, Patrick! As much as I'd like to take credit for this great post, though, this one's actually by Michelle...
I suppose that's what happens when a feed reader doesn't show the whole post!
-Patrick
This post has been removed by a blog administrator.
This post has been removed by a blog administrator.
This post has been removed by a blog administrator.
Post a Comment
<< Home