I am beta-testing Windows XP. The betas all ship with smart-tags disabled by default, if this is because Microsoft's tag repository is nearly empty, or because it's good practise, is anyone's guess.
When turned on, the smart tags differ from regular hyperlinks, in that they're underlined in a different manner, and when you hover above them, you get an icon to click, as opposed to simply following a link.
This seems to allow the user to either click the link and end up where s/he expects to end up, *or* follow the smart tag-link for Microsoft-ified "information".
Currently, only Microsoft's own webpage seems extensively smart-tagged, on/., the only thing popping up is the word Microsoft.
As a former employee of Ericsson Data (subdivision of L M Ericsson AB), which most of you may know as a cellphone-manufacturer, I've made an observation on policies in large, multinational companies;
- There will always be small, independent groups within said company that will go their own route, when it comes to using and publishing open-source, choice of OS, etc.
- These groups will as a rule of thumb get penalized by the overhanging organization.
- If the gains from using the "deviant" methods or software are large enough, the group will pay the fines and continue.
When I worked at Ericsson, we had what was called ESOE, Ericsson Standard Office Environment, which blasted our computers with an image of what they thought our system should look like. Solution: Make sure the network cable was unplugged between 1am and 3am.
We used both open-source software, contributing our changes to the appropriate repository, freenixes (Linux, *BSD) and we even bought our own workstations, DECPC Dual PII's, which blew the standardized, corporate deal HPs away.
In the end, our productivity and quality of development clearly outweighed the fines we had to pay the overhanging corporate police, so there was nothing anyone could do. One of two things usually happens to such groups; either the mother organization values profit more than conformity, and casually continues to penalize it, but nothing else. Or, it disbands it, re-assigning people to other departments.
My point here is [vague], but there's something to be said about pure commercialism. If it's high enough in the corporate structure, it sometimes benefits the "low" individuals, in letting them be at their most productive and choosing their own tools.
(My department got disbanded approximately 2 years after it was formed, leading to 90% of us leaving Ericsson.)
I use Solaris extensively. I like it very much, probably because I know more about it than any other Unix, but I also like Linux, because it's easier to just blow onto a box when I need to create a workstation or test something trivial.
I dare say that I am moderately to very experienced in compiling freenix (Linux, *BSD) soures to work with Solaris, both GNU and of other origin. A few times, I have submitted patches of what I have done, not necessarily "correcting" an "error", but most often dealing with Solaris idiosyncrasies that can be "fixed", but typically might not be (think DNS round-robin and the result when using INADDR_ANY).
However, an overwhelming amount of times, the people maintaining the sources have just shrugged and said "well, post it on some Solaris messageboard, we don't run Solaris", or given some other response in the same vein. I'm not saying everyone's guilty of this, but a lot of developers think that if it can be fixed on the system, it should be, not in the software. This is unfeasible when speaking about server installations, that cater to a multitude of other applications. I will not break, or even force reconfiguration of, iPlanet, for instance, just to be able to run a small util.
Slashdot Mirror
I am beta-testing Windows XP. The betas all ship with smart-tags disabled by default, if this is because Microsoft's tag repository is nearly empty, or because it's good practise, is anyone's guess.
/., the only thing popping up is the word Microsoft.
When turned on, the smart tags differ from regular hyperlinks, in that they're underlined in a different manner, and when you hover above them, you get an icon to click, as opposed to simply following a link.
This seems to allow the user to either click the link and end up where s/he expects to end up, *or* follow the smart tag-link for Microsoft-ified "information".
Currently, only Microsoft's own webpage seems extensively smart-tagged, on
As a former employee of Ericsson Data (subdivision of L M Ericsson AB), which most of you may know as a cellphone-manufacturer, I've made an observation on policies in large, multinational companies;
- There will always be small, independent groups within said company that will go their own route, when it comes to using and publishing open-source, choice of OS, etc.
- These groups will as a rule of thumb get penalized by the overhanging organization.
- If the gains from using the "deviant" methods or software are large enough, the group will pay the fines and continue.
When I worked at Ericsson, we had what was called ESOE, Ericsson Standard Office Environment, which blasted our computers with an image of what they thought our system should look like. Solution: Make sure the network cable was unplugged between 1am and 3am.
We used both open-source software, contributing our changes to the appropriate repository, freenixes (Linux, *BSD) and we even bought our own workstations, DECPC Dual PII's, which blew the standardized, corporate deal HPs away.
In the end, our productivity and quality of development clearly outweighed the fines we had to pay the overhanging corporate police, so there was nothing anyone could do. One of two things usually happens to such groups; either the mother organization values profit more than conformity, and casually continues to penalize it, but nothing else. Or, it disbands it, re-assigning people to other departments.
My point here is [vague], but there's something to be said about pure commercialism. If it's high enough in the corporate structure, it sometimes benefits the "low" individuals, in letting them be at their most productive and choosing their own tools.
(My department got disbanded approximately 2 years after it was formed, leading to 90% of us leaving Ericsson.)
I use Solaris extensively. I like it very much, probably because I know more about it than any other Unix, but I also like Linux, because it's easier to just blow onto a box when I need to create a workstation or test something trivial.
I dare say that I am moderately to very experienced in compiling freenix (Linux, *BSD) soures to work with Solaris, both GNU and of other origin. A few times, I have submitted patches of what I have done, not necessarily "correcting" an "error", but most often dealing with Solaris idiosyncrasies that can be "fixed", but typically might not be (think DNS round-robin and the result when using INADDR_ANY).
However, an overwhelming amount of times, the people maintaining the sources have just shrugged and said "well, post it on some Solaris messageboard, we don't run Solaris", or given some other response in the same vein. I'm not saying everyone's guilty of this, but a lot of developers think that if it can be fixed on the system, it should be, not in the software. This is unfeasible when speaking about server installations, that cater to a multitude of other applications. I will not break, or even force reconfiguration of, iPlanet, for instance, just to be able to run a small util.