Oliver Wong wrote:
> "Roy Schestowitz" <newsgroups@xxxxxxxxxxxxxxx> wrote in message
> > How I saved over $4000 on software?
> > ,----[ Quote ]
> > | Anyways, while setting up my new laptop, I had an
> > | intriguing idea. Why not consider installing
> > | all/most non-commercial software to cater to all
> > | my daily computing needs?
> > | Total Savings (in USD) $4235.95
> > `----
> > http://www.nilkanth.com/archives/2006/10/24/how-i-saved-over-4000-on-software/
> That said, this blog entry is interesting in that if you don't know what
> alternatives exists for a given piece of non-beer-free software, you might
> be able to look it up in the table provided.
> Unfortunately, the comparisons are not always apples-to-apples.
> Thunderbird is not the equivalent of Outlook,
> for example (TB has no calendar),
This is typical "Microsoft Think" - you assume that Outlook is "better"
because it bundles e-mail, calender, and other functionality into a
In Unix/Linux, context switching is about 10-20 times faster than it is
for Windows XP, and about 100-200 times faster than Windows NT 4.0.
Which means that there is very little advantage in putting all of the
functions into a single application that runs as a single process.
Linux can easily run the same functionality, and do each core function
more effectively, by putting each core function in a separate process.
This also makes maintenance and support easier.
Netscape Communicator/Mozilla attempted to put all of the applications
into a single process, and it was slow, difficult to manage, and hard
to support. They split that core functionality into FireFox,
Thunderbird, and plug-ins. The components were more successful than
It's more important that the applications adhere to standards such as
protocols and file formats, than it is for applications to be bundled
into a single monolithic solution.
> nor is Paint.NET the equivalent of Adobe Photoshop.
Granted. And there are some commercial graphics applications available
for *nix that would put adobe to shame.
Yes, you can compare a free OSS application like gimp or paint.net to
$600 photoshop and say the free one isn't quite as good.
Which then begs the question
"If I wanted to spend $600 for an application, what could I get"?
The point is that if you use the Linux applications for the things that
are not "critical" or strategic, but image processing is critical, you
now have $4,000 that can be directed to those strategic projects which
will help you generate the most money from the least effort, in the
area where it counts the most.
Look at the server market. Need a database, you can use MySQL or DB2,
but if you use MySQL on some of the "peripheral" databases, for less
critical tasks, you can focus more resources into a really "beefy" DB2
machine, and have the MySQL engine offload some of the work, like
lookup tables and code to description tables. With proper blending and
Component oriented Model/View/Control design, the result is better and
you get more "bang for the buck"
You can use Apache, Jakarta, Struts, and JBOSS, or you can use
WebSphere. If you are using canned OSS solutions, you might go the OSS
framework. On the other hand, if you need to make sure that you can
maintain compliance with regulatory standards such as HIPPA, you might
find WebSphere more than pays for itself in that specific form of
> A site I occasionally check out to find open source alternatives is
> - Oliver