|
| quote: | Originally posted by Yoepus
Monopolies are very good when you are dealing with standards. Its great our government has a monopoly over money, because if there were 12 different currencies we could trade in, it would be a headache for everyone and more costly, as companies would encure more costs in printing, time calculating, and IT spending... etc.
|
There are more than 12 different currencies - and you *can* use them for trading as long as your trading partner agrees with your choice of currency.
| quote: | Originally posted by Yoepus
Now I'm not advocating a monopoly in the software environment, I am just really happy at the current situation, M$ has a defacto monopoly, yet has the pressure to keep its software competitive. So we basically enjoy the best standards.
What I dread is a world with 25 desktop operating system of equal market share, or a world with only 1 desktop operating system. But so long as it is as it is, 1 desktop operating system with hundreds of small competition. I think it is for the best.
|
The problems you foresee do not exist as long as every company would used standard formats for data, and had their programs interpret data in the way it is supposed to be according to standards.
M$ is notoriously bad at this point. In spite of being heavily represented in standard comitees like w3, they consistenly fail to adhere to the standards.
Furthermore, for years organizations such as ISO has provided standards, which have worked quite well. There is no need for a monopoly in order to have standards.
| quote: | Originally posted by Yoepus
Like it or not Windows is a standard, and if I'm a program developer I know I will be able to program in a windows environment and I will have an audience to sell it to. If there were 25 environments, I would spend perhaps most my time porting the software to the 25 different envrionments, wasting my time on porting instead of the features of software. So you either lose money, lose time, or lose features if such a world would exist with 25 desktop envrionments. Not to mention a complete loss of standards.
|
Well, your statement is highly dependent on the type of software project you intend to take on. If your are constructing a piece of firewall software, or software for a highly specialized task, you would be more concerned with having a platform which fulfils your requirements rather than having the software run on the majority of PCs. And when evaluating a platform on something which is not software back-catalogue or size of user-base, Windows does tend to lose out.
| quote: | Originally posted by Yoepus
As for installation, I don't know but for me "configuring" then "making" and then "making install" every program I want to add to linux (otherwise the precompiled binaries give me problems 90% of the tiem) takes MUCH longer than double clicking on a .exe in windows.
|
Well, for me, I like that my programs are placed - in their entirety - in a directory, and can be removed by a simple rm-command. On Windows, you have libraries and configuration files spread over your entire directory tree, and its [IRONY]ingenious[/IRONY] referencing system will ensure that removing the program through the control panel leaves ample room for left over files, which eventually starts cluttering the system.
| quote: | Originally posted by Yoepus
Ya, but you need to open the package handler, etc.... Windows, you just double click, its quick and easy. All the installation part is on the onious of the programmer, not the user.
|
Well on Windows, you need to find a place to download the installation file, go to the folder where you downloaded to, and then double click it. With something like synaptec you just type the name of the package or a file included in it, and it is downloaded and installed in one go.
| quote: | Originally posted by Yoepus
Sure windows has no sensical services, so does linux.. I remember going into the kernel build and turning on things that weren't available and turning off those things I completely didn't need. But it was much quicker for me to go into Windows services, and click all those off I didn't need during startup.
|
The difference is that when you turn off things in Windows, it means "do not use this thing", but a series of checks is still performed during operation and the memory requirements for the kernel are unaffected. When you compile a new kernel for linux, the parts you do not want are not included in the running code, and you hence gain more efficient execution.
With regards to Windows being user-friendly: It is true that if you ask the average joe to do something with a Windows PC and then ask him to perform the same tasks with linux, he will probably have a harder time using the linux distribution. But I claim that is mostly to do with habbit, as the newer distributions like RedHat and Mandrake have similar configuration tools and applications as Windows.
Only if you start messing with low level stuff do you need to be an expert. But you would also need to be an expert if you should do low level stuff (e.g. repartitioning or changing file system) on Windows.
|