I’m usually against the idea of hacking quality CSS code to make it compatible with IE or older browsers that do not support modern standards, but in some cases it might be required. Of course what I’d really like to say is stop using crappy browsers on completely outdated hardware, but not everybody is a tech junkie. Not everybody has the money to buy the latest computer every year. Not everybody cares. The idea behind the web is that the end user should have to care about what browser or OS they are using. It should just work for them. Reality dictates that isn’t the case, but it’s a nice though and we are closer to that goal today than we have ever been.
I read over some of my statistics before writing this. I was surprised to see Windows 98 and IE 5.5. It was less that 0.1%, but still. The fact that somebody might actually still be using those is amazing. Mostly it was Firefox and XP. A few IE 6 and 7. A few Linux machines and Vista. You get the picture. Site statistics are a good way for a web designer to understand what they must support.
However, I’m probably not really a good web designer. I’m seriously thinking about not supporting anything prior to Windows XP and IE 7. I say those two because IE 7, whilst still a crappy browser, has far more support for modern standards and it’s only available for Windows XP and above.
My plan for browser support goes kind of like this:
Firefox 2.x+
Opera 9+
Safari 3+
IE 7+
Chrome
Konquerer
OS support (kind of theory on this one)
Windows XP+
Mac OSX+
Linux
*BSD
And of course, mobile support. Still studying up on this one. All I really have access to is my Crackberry Curse (I really do like my phone) running Opera Mobile 4.1, but I believe that mobile phone (device) support is important these days. Which means either building two sites or limiting the eye candy on the main sight. Namely flash. I like flash, but it’s just way to limiting still yet.
So that’s my thoughts. Probably going to change as I move along.
It’s late. I’m bored and I own a crackberry. ‘Nough said. :)
I’ve installed Kubuntu 8.10(beta) and Mandriva 2009 on my IBM ThinkPad T41. On the laptop they both work really well. Haven’t tested battery life or anything like that, but all the hardware seems to work and speed is good.
However, what impresses me today is how nice KDE 4.1 is. It’s actually brought some excitement back to Linux for me. I’ve been pretty much a gnome user for awhile now and was kind of hating KDE 4 before it’s release. I tried it out when it came out and liked it, but it wasn’t really functional for me. 4.1 though seems to be much better. My initial feelings here are that it’s a rock solid desktop manager and the skies the limit. It’s safe to say that I don’t really see myself going back to gnome anytime soon. I’d even venture to say that KDE might win some converts with 4.1. It’s just a really nice desktop manager.
As I grow in technical proficiency my one sided “hate Microsoft” mentality is fading away. That’s not to say that I’m running to Microsoft, but rather that I’m not simply ‘hating’ Microsoft because it’s Microsoft. I’m trying very hard to put bias behind me and view technology as an object with a function. However, the one fact remains. Microsoft is a company with a business practice of locking customers into it’s products. Not all of it’s design ideas are particularly geared to the advantage of the end user. That’s one reason why I still prefer other options. Vendor lock in sucks.
So what are my other options for an OS. Linux, Unix, BSD, Apple, Solaris, or some other OS variant. You know it appears, on face value, that I have a great number of viable options. However, that’s not really true. In many cases I’ve been force to buy Microsoft products simply because of that ol’ crappy vendor lock in situation.
For the most part I’m stuck with Windows as a server OS and a desktop OS. However, I’m constantly re-evaluating my options. So what are my needs? Well, I need a file server with user authentication and rights management. I need a capable print server that works with not only Windows based programs, but also some legacy DOS based programs. I need a DNS server. Those are my basic needs. Some optional wants are RDBS, HTTP/FTP server, DHCP, etc. Without really digging into it much all those needs and wants can be easily served by any of the major alternative OS’s. That’s a good thing.
******
It’s been a few days since I started this little article. Since then I’ve been playing around with CentOS 5. CentOS is a fine OS, but I still prefer debian based distro’s. However, learning a Red Hat based distro will probably serve me better in the real world than a debian based as most business running Linux are probably running either Red Hat or Suse. However, I digress. One of the reasons I playing around with Linux is because I want to learn about OpenLDAP, Samba, and Cups. So far I’ve really just been focusing on LDAP. With a background in eDir I’m not to far off understanding LDAP. Mainly because I use eDir like a LDAP server.
One of the interesting things about LDAP is the many schemas available for it. Even though I haven’t found any use for any schema beyond the core one included it is pretty cool that so many exist. With that said they maybe total crap. I don’t really now.
Anyway. I suppose the reason for learning LDAP is because I’m looking for an open source and standardized directory server for user and resource authentication. While AD and eDirectory are both fine one is tied to Windows and the other is controlled by Novell. So far I’ve just touched the surface of LDAP, but I’m already starting to understand how powerful a tool it is.
At this point I’m going to finish this up by saying that while my technical experience with LDAP is sorely lacking at this point I’m not giving up on it as a good method for authentication. While it won’t replace AD in a Windows environment because of Windows AD hooks it might serve as a backup method for it.
Software companies that create intentionally confusing and complex software licenses suck.
I was reviewing Microsoft Exchange licenses today and from what I could understand they are basically designed to screw the customer out of as much money as humanly possible. While I would love to say this is a Microsoft monopoly practice I see it with many software programs. The simple fact is that they create crazy licenses to generate as much revenue as possible. Simple schemes simply don’t create the same kind of cash flow.
Here’s how I would license software, but first I should mention that I support the open source model and free (as in beer) ideology.
First off. Server operating systems should be licensed via the server. No per core, per processor, per user, etc. crap. You have one server os and one server it goes on you need one simple license. Simple. It even works with virtual servers. Wow. How easy is that.
Secondly, desktop operating systems should be just as simple. One desktop os on one desktop computer. And no more multiple versions. One version. No need for a business edition, a home edition, an ultimate edition, a cheap edition, a third world edition, etc. One simple version for all desktops. (On a side note. An OS should be as resource un-intensive as the programmers can possible make it. The OS itself should require massive amounts of ram or the latest video card to work well.)
And now we get to applications. Application are a bit trickier in my mind. I believe you can license application software based on either physical machine or active user. However, I disagree that you should have multiple version of the same software that simply cut out features. If you are going to do that just let me pick the pieces I need.
I can see were you might want to license a program based the computer and I can see where you might license a program based on the number of active users. Personally I like the latter in a business sense because if you have 20 computers, but only 10 active users of a particular program then you save money by not buying 20 licenses. This methodology doesn’t bode well for application developers though and that’s what the end products is all about. Revenue.
At the end of the day though the only real choice for choice is free and open source software. Not because it’s quality or usability is better, but because the licensing is easier to deal with and much less confusing. Free isn’t a must, but why not. With software cost for most business going up getting a discount on the software is always good.
