I have read a popular Itlaian blog this morning which theorized a serious problem for the popular Linux distribution Ubuntu, that would be the incapacity to comply with the growing shift towards the web of more and more applications which would shortly make local computing obsolete in favor of "Thin clients" devolving all the dirty job to super powerful remote servers.
... O_o' ...
The topic is seriously interesting, although I really fail to see how this would be a problem of a specific operating system, even more a specific distribution of it. I am a proud Ubuntu user and I really don't see the problem. If anything, from a user point of view, it seems to me that the web 2.0 integration is one of the fields where linux does shine! But let me put aside the polemics, this is not the point of my post.
The fact is that my inner man 1.0 completely disagrees with this vision in general for a variety of different reasons:
1. The mainframe-terminal architecture reminds me of '80 sci-fi movies à là Tron, but was overcome by the revolution of the personal computer, since then we have seen more and more powerful personal computers (alas, my handheld is orders of magnitude better than the computer that sent 3 men on the moon).
2. As pointed out by someone in the comments to the blog I have read, the network computer never broke out because a lot of tasks are just better accomplished on a local basis. The network shift is partly for tasks that are better accomplished over the web (for mobility, accessibility, sharing or whatever other advantage), and partly as a sort of "exploration". Being an entirely new field it is useful to test the web version of whatever comes to the developer's mind. Wether most of these services will still be there in five years I really don't know... perhaps many will have disappeared and others will be born that we cannot even think of today.
3. Personal computing demand grows faster than the web based services: five years ago I had a laptop with 5Gb of storage and it seemed a lot, now both my personal music collection and my photo folder, each exceed by far that limit (let us not speak of videos). Today an on line storage services offering 5Gb of space is considered just right, an it is, because it replies to very different needs than local storage (again mobility, social use, remote backup). The same discourse applies to more "processing dependent" uses, I am all in favour of Google Docs, which I use quite frequently, but I still need OpenOffice and I cannot see this situation changing too soon, same for The GIMP and any game I can think of (hehe, untill they will not come out with the falsh version of Urban Terror ^_^').
4. What about security? All hail the network, but the best way to sleep comfortable at night is always be to keep our little scary secrets offline!
5. A last tought from what I can remember of economics at the university: there is a growing cost for adding each unit of "processing power" to a machine, I guess it's called marginal cost. So If I can build a machine with power suitable for one user, a machine with enough power for two users would not cost the double but something more... and so on. Scale economies and technical progress can overturn this rule for a while, creating an exception, but I think that in the long run there is a limit to what even a 2.0 man needs and want, and the general rule will apply again.
That's it, and it is not a complete list... of course I might be wrong, but if I should bet my wage on thin clients with minimal local OSs or small powerhouses with the power of current supercomputers I'd surely go for the latter (after all I have recently read of some guys who built a TERAFLOP super computer just stacking some Playstation 3 and running Linux!).
Slashdot It!
Nessun commento:
Posta un commento