On Sun, 2005-11-27 at 06:13 +0530, A. Mani wrote: > On Sunday 27 November 2005 01:22, Rajarshi Guha wrote: > > On Sat, 2005-11-26 at 22:33 +0530, A. Mani wrote: > > > On Saturday 26 November 2005 04:56, Rajarshi Guha wrote: > > > > On Sat, 2005-11-26 at 04:02 +0530, A. Mani wrote: > > > > > > <snip> > > > > > > > How does the CLI lead to a more cost-effective solution for the Third > > World? Can you elaborate on this? > > > Easy we can consider the $50 PC. The graphics demand makes the cost of the PC > higher...that is a direct part. This way we can help some sections at least.
True - but I assume that even a $50 PC would have and onboard graphics card. Actually my desktop has a 16MB NVidia graphics card and it handles a GUI fine. As for the rest of the system, even 128MB would be enough to run stuff like XFce But I seriously doubt people are going to use a $50 PC if they are forced to use a CLI. In fact, if I were to place a $50 PC with just a CLI in front of somebody whose native language is not english, what happens then? Does this approach require a whole translation of the commands themselves? 'echo' might have a Bengali translation, but if that translation has to be made at the syntactical level, how are they supposed to learn Unix from other sources (which I assume are based in English) If you then say, that no, keep the standard Unix commands, then you are essentially saying that a user must simply learn by rote a set of commands which are sometimes cryptic to English speakers, so I expect they would just be random letters to a non-Enlgish speaker For the non-expert, a CLI is definitely the wrong way to go. > > Does that mean you would ask programmers to give up their IDE's? > No, ... but programmers generally forget about most IDE features which > involve the mouse very soon. Those who use Emacs will not look at IDEs. Agreed - but that approach sounds a lot like dogma. If a GUI tool does improve workflow it would be unprofessional to not use such a tool, if available. Actually in bioinformatics/cheminformatics, GUI based tools such as Pipeline Pilot and Taverna) allow very sophisticated systems to be built. They're built by dragging icons and connecting them - clearly a simplified interface. However the result is that they are able to perform quite involved tasks. The fact that each of these environments can be extended (via writing programs and modules) is a nice thing. But the important thing is, a huge amount can be done without having to learn how to program (though I have to admit, a bioinformatics or cheminformatics person who can't program is at a serious disadvantage) > I use Kile...but without the mouse. > I am more of a GUI user, only when I started looking in did I understand how > insufficient these GUIs are. Thats my whole point! You wanted to go and understand what goes on. Obviously a GUI hides a lot of stuff. So it is in your interest to use the CLI. But for a person who just needs to write an expense report, the CLI is superflous. It seems this distinction has been made clearly, and I don't see any disagreement with this. So what are we arguing about then? I think everyone would agree that an interface to a computer is optimized for its function - GUI's are fine for occasional word processing. GUI's are fine for spreadsheets. A CLI is required for sysadmin tasks. A CLI is useful in a stats program if the GUI is not enough. And so on. ------------------------------------------------------------------- Rajarshi Guha <[EMAIL PROTECTED]> <http://jijo.cjb.net> GPG Fingerprint: 0CCA 8EE2 2EEB 25E2 AB04 06F7 1BB9 E634 9B87 56EE ------------------------------------------------------------------- Eureka! -- Archimedes -- To unsubscribe, send mail to [EMAIL PROTECTED] with the body "unsubscribe ilug-cal" and an empty subject line. FAQ: http://www.ilug-cal.org/node.php?id=3
