ssamayoa wrote:
Quote
I thought, in 1999-2001, that linux will be in at least 30 or 40% of the
desktops of the world.
But now, falling 2005, there is no sign of linux on the desktop.
I think that Windows will be the {*word*109} OS on desktop at least for the 5
to 10 years in the future. As matter of fact I'm advising to my customers
to stay with Windows, I'm lefting linux only for server user.
i think this is due to a lack of business applications
i mean, for office purposes OpenOffice+firebird+thunderbird is perfect
but many other daily used applications are missing
for corporate companies : ERP clients, Lotus Notes clients...
and for small companies : accounting, budget, appointments ... softwares
and to complicate this each country have is own official required
documents formats(tax dept), VAT in US is a little bit complicated...
also, there is weight of habitude...but things are changing slowly
Quote
Why bother with cross platform if M$ will be in the desktop for the next
lustrum or decade?
I'm begining to play and develop with the new VS 2005.
I left my technician and anti-M$ hat in the closet, I want to make money
with my apps.
it's difficult to be objective
but at least from the technical point of view, i've 3 remaks regarding
this NEW intermediate layer that MS created and is hardly marketing
called 'NET Frameword'
1?technical innovation
on one side CPU powerfullness is always improving
and on another side a new intermediate layer is introduced.
claming no, no it's not pcode, it's compiled.
look a little bit CIL specs, it's an unexisting cpu, with an unexisting
instruction set, opcods, data structures...
all this have to be compiled ok but then ... no cpu can uses this nativeli !
it's the same as making a AMD64 compiler in 1990, resulting code
should have to interpreted again to be usable.
so all these compiled stuffs have then to be translated
by an engine that will traduce this in native code.
this only have consequencemake to make YOUR code extremely {*word*60} to the OS
and perhaps in 2010 they will release a new MS CIL CPU understanding
natively wich is for the moment nothing else than another pcode :)
first there was only VB now they plan to put the complete OS in PCODE !
very good
2?needs
earning money is coding faster complex things and bug free (nearly free :) )
this could be achieved with higher level objects, more business
frameworks with big rock solid objects libraries...
if they not exist i must make my own and if i have to they must be
portable...
instead of this, what new thing do you find in VS :
-TLabel, TCheckBox, Timage, TTreeview...
ho, it's seems something 'déj?vue'
just reinventing a 10 years old wheel !
very good
3?remember what happened to MFC ->deprecated
jet engine, ado remote dataset ->deprecated
cautiousness, it seems that lifetime ms technologies can be short and
can change like the wind...
the risk is to drop the time you spent to learn it, drop your code and
learn again.
