Re:How do I go about learning Emac?
on
The Birth of vi
·
· Score: 1
But I am glad that I was open minded enough to learn a more elegant weapon for more civilized times.
Save your condescension for when you can actually make a point. Almost everyone these days knows how to touch type and few people feel the need to learn the unintuitive double-keypress commands for movement. I can use the keys I described far faster than the modifier+X keys, simply because they are single keypresses and I can home my right hand on them with no effort at all without looking at the keyboard, especially on the Thinkpad keyboard I use most of the time. Likewise, if you think a modern GUI-based text editor requires a mouse to switch between buffers, you've been living in the past for a long time. Many modern editors have even caught up with Emacs on text-completion-based buffer switching, but unlike Emacs, they are actually able to also display all the open files and loads of other useful hierarchy information in a sidepane.
Telling a modern computer user to distinguish between text editors and electronic typewriters is like telling a cell phone user to distinguish between a rotary dial and a keypad. Anachronistic in the extreme.
There is no reason why these two tasks should be logically separate from each other.
Re:How do I go about learning Emac?
on
The Birth of vi
·
· Score: 1
Actually, they are. Home, End, PgUp, PgDn, Ctrl-up/dn/left/right, Alt-up/dn/left/right, Ctrl-F. Those are the keys that work properly in most modern GUI text editors.
You know how I learned these, and many others? By looking at the main menus and the shortcuts noted in them. This is why I despise both Emacs and vi. They provide no facility whatsoever for UI-based, online help and exploring of options. They have no concept of UI. Thanks, but I'd rather use a less featureful GUI editor (with no mouse necessary) and write some macros than internalize all those key combinations in the blind.
And don't get me started on Emacs' "menus" or default indentation mode or "GUI" or any number of other quirks.
No, there is no standard. If there were one, you would be hearing a lot about it, and people would routinely be up in arms over what it should provide and how. (Freedesktop is making a lot of progress on this front, though.) As it is, if you want to write a GUI application, you must link to one of many different UI toolkits and associated libraries, and life will be a lot easier for you if you program your app according to the conventions of that desktop environment.
If you've used KDE and haven't noticed that all applications have a lot of recurring components and common structure, you haven't looked hard enough.
Clearly they don't rely on KDE, so they must be written to some generic desktop standard. So, why the mcnaming?
They do rely on KDE, namely kdelibs, which relies on qtlibs. Have you ever used a KDE desktop? All applications use a massive amount of environment-wide services like the prefs engines, dialogs, helper apps, kioslaves, text editing libraries like Kate/Kwrite, services, etc. No other desktop environment can touch the degree of integration and most importantly, consistency afforded by this code reuse.
This ease of development and consistency is what makes KDE so powerful, and why I'm a huge fan of the KDE development model. The people who design kdelibs really understand what it takes to provide a tightly integrated desktop environment.
You are defending a crutch of an application, created for a grievously misguided library whose main motivation appears to have been that "if Windows is doing it, it must be good enough for us", to expose options which are not user-configurable because of an inability to organize the applications' GUIs. Then you go on to complain that an environment that avoids all these problems is a support nightmare. I have news for you: Microsoft Windows is even easier to administer if you want your users to have a minimum of options in completely disjointed applications with idiotic facilities for adjusting functionality that didn't make it to the UI.
GNOME releases regularly and gets new features out into general use quickly.
Yeah, and KDE devs hold all their features back for themselves. WTF does this have to do with anything? The Gnome and KDE point releases occur at about the same frequency.
Stick with Gnome and Firefox if you need the comfort afforded by cutting functionality from programs. "Less is more" has brought Gnome and associated applications to a point where I cannot use them without disgust. KDE always seems to have lived by the principle of "more options, better organized" when preferences are concerned, and what it needs to do is expose even more, and organize better. There are applications where prefs dialogs are organized poorly. Does that mean those options need to be hidden? Hell no. Just reorganized.
The latest compilers, with the default settings, produce code that's bloated for the tiny machine you describe. I suspect GCC 4 with "-Os" would do better, as long as you were going Gentoo anyway. Also, what happened most likely was that your video setup was suboptimal and had minimal or no acceleration. In general, I find it very hard to believe that a system with the later 2.6 kernels, GCC4, and properly configured video and multimedia packages would have trouble beating an old 2.4 system, especially if it's Gentoo, not saddled with a bunch of services starting by default.
But the fanatics, who buy/rent/go to a movie or two every month, or even every week, are maybe 10% of the population, and they amount to 80% or more of the profits.
What?!
Most people who don't use filesharing extensively buy, rent, or go to watch movies more often than once a month. Your analysis in general, and that datum in particular, is way off base.
Have 2 different groups: Development - that works on a backlog of tasks and incrementally improves various pieces of the product; and Release - that takes versions of the development pieces, matches them with marketing requirements, makes a cohesive product and releases when the marketing requirements are met.
Yeah man. It's called release engineering and all good enterprise (and not-so-enterprise) software projects have it. There's a bunch of groups working on sub-projects, and then there's releng which is responsible for rolling all that into releases.
- data should be stored on the server (centralised backup, provision for web mail when you need it, ability to have an administrator control it, access from multiple hosts)
It's called IMAP.
- server-side spam filtering which can also take easily feedback from the client on what proved to not be spam, or what was and was missed.
Yeah. SpamAssassin and friends.
- server-side addressbook
LDAP.
- should deal only with plain text - non plain text should be flattened to plain text. It would be nice to automatically bounce office files with a message to tell the person to send stuff as PDF or plain text.
News flash: email HAS to be all plain text for servers to understand it. Attachments are still encoded, of course. Writing a filter for mailfilter/procmail/whatever that bounces based on attachment type takes all of 30 seconds. I'm not sure if it can be done in Thunderbird, but I'm sure there's a plugin out there for it.
- effective searching - very responsive client for reading mail - very responsive client for writing mail
Already done. Have you even USED Thunderbird lately?
- effective communication between client and server that doesn't require the user to wait
What hardcoded paths? What has to be scanned? The profile is a bunch of files formatted in a standard way with nothing preventing you from relocating it. FFS, I copied my whole profile from Windows to Linux, started Thunderbird, and there it was, using all the data I previously had. If the devs can't make a basic function for importing other profiles from their own program, well, that's just ineptitude on their part.
On a totally unrelated note, the junk mail scanning in Thunderbird is horribly broken, and worse, I can find no way to turn it off altogether...
Do robots dream of electric sheep?, do bears really shit in the woods?, would a Slashdot editor recognize a proper English sentence if it hit him in the face? The answers will probably come when hell freezes over(pretty soon).
... well, I guess I did say that, so I concede that AMD still has some performance parity left in the 4-core market, if discounting all factors other than the number of cores.
AMD still rocks - without K8, we would not have Intel offering anywhere near the performance for money that we have today. And all my CPUs at home are AMDs, since they were bought before Conroe came out.
E5320 is a 1.86 GHz chip. 8218 is a 2.8 GHz chip - top of the line, an overclocked Opteron that AMD added after Intel started making good server CPUs again. An apples-to-apples comparison would be 8218 vs. X5355 (2.66 GHz and a faster FSB). That's 43% faster clock speed and some effect from a faster FSB. This will still be behind the 4-socket system by a bit, but I never said quad-cores will scream past dual-cores, I said they offer a hell of a lot more value for my workloads, particularly in rack density and heat.
Look, you can be paranoid about your privacy all you want, but if you cannot appreciate the versatility and freedom a car allows - not just in the United States, but especially here - you're either a sorry hermit or are being intellectually weak. I won't even start listing the number of ways having constant access to a car has improved my life.
I think the capital expenditures on public transportation are too low in the US, and I would gladly use public transportation instead of a car to commute were it available in those places where I can't bike to my workplace, but that doesn't prevent me from appreciating the tremendous freedom a car allows me.
they want the "freedom" of being able to decide where they go (even though everything related to cars is heavily regulated.)
It's absolutely ridiculous to imply that a car doesn't offer a massive increase in personal freedom. If you disagree, you probably aren't using your freedom much.
So the moment the health visitor told you that, you should have told them to take a walk and reported them to their supervisors, and told them that if that's their policy, they need to drop it. If this bullshit (government officials in a mandatory test telling you not to educate your kid too much) makes your blood boil, as it does mine, then stand up to it and do your part to make it stop.
I want AMD to succeed as much as the next guy - I definitely appreciate the fact that they spent the last 20 years working hard to gain superiority over a much larger competitor. But I think Hector Ruiz (the AMD CEO) must feel like Yamamoto, the Japanese WWII admiral who allegedly said, "I fear all we have done is to awaken a sleeping giant and fill him with a terrible resolve".
(sorry, didn't read the whole comment before replying)
On most workloads I'm concerned with, Intel's new dual-FSB, quad-channel chipset (5000P) provides enough of a performance parity against AMD's on-die controllers to make the overall performance very competitive against AMD's 4-sockets. Most public benchmarks I've seen on these configurations are pretty ridiculous, using non-parallelizable or I/O-bound consumer software. Granted, my workloads are not terribly memory intensive, so if you're looking for a database server that has to juggle the whole database in RAM continuously, AMD might still be the killer. On the other hand, even when I think about desktop workstations, I'd rather not have a 4-socket monster on my desk and take the slight performance hit using Intel's quad-cores. And of course Intel's 2x quad core is very substantially cheaper than AMD's 4x dual core.
No. At 8 cores, you'll be well advised to buy 2 sockets with 2 quad-core Clovertowns. Which will still wipe the floor against all but the 2.6+ GHz Opterons on all but the most memory-intensive workloads, and have about twice or more the performance/watt of any Opteron. AMD is quite screwed everywhere except on 8-socket servers.
Slashdot Mirror
But I am glad that I was open minded enough to learn a more elegant weapon for more civilized times.
Save your condescension for when you can actually make a point. Almost everyone these days knows how to touch type and few people feel the need to learn the unintuitive double-keypress commands for movement. I can use the keys I described far faster than the modifier+X keys, simply because they are single keypresses and I can home my right hand on them with no effort at all without looking at the keyboard, especially on the Thinkpad keyboard I use most of the time. Likewise, if you think a modern GUI-based text editor requires a mouse to switch between buffers, you've been living in the past for a long time. Many modern editors have even caught up with Emacs on text-completion-based buffer switching, but unlike Emacs, they are actually able to also display all the open files and loads of other useful hierarchy information in a sidepane.
Telling a modern computer user to distinguish between text editors and electronic typewriters is like telling a cell phone user to distinguish between a rotary dial and a keypad. Anachronistic in the extreme.
There is no reason why these two tasks should be logically separate from each other.
Actually, they are. Home, End, PgUp, PgDn, Ctrl-up/dn/left/right, Alt-up/dn/left/right, Ctrl-F. Those are the keys that work properly in most modern GUI text editors.
You know how I learned these, and many others? By looking at the main menus and the shortcuts noted in them. This is why I despise both Emacs and vi. They provide no facility whatsoever for UI-based, online help and exploring of options. They have no concept of UI. Thanks, but I'd rather use a less featureful GUI editor (with no mouse necessary) and write some macros than internalize all those key combinations in the blind.
And don't get me started on Emacs' "menus" or default indentation mode or "GUI" or any number of other quirks.
No, there is no standard. If there were one, you would be hearing a lot about it, and people would routinely be up in arms over what it should provide and how. (Freedesktop is making a lot of progress on this front, though.) As it is, if you want to write a GUI application, you must link to one of many different UI toolkits and associated libraries, and life will be a lot easier for you if you program your app according to the conventions of that desktop environment.
If you've used KDE and haven't noticed that all applications have a lot of recurring components and common structure, you haven't looked hard enough.
Clearly they don't rely on KDE, so they must be written to some generic desktop standard. So, why the mcnaming?
They do rely on KDE, namely kdelibs, which relies on qtlibs. Have you ever used a KDE desktop? All applications use a massive amount of environment-wide services like the prefs engines, dialogs, helper apps, kioslaves, text editing libraries like Kate/Kwrite, services, etc. No other desktop environment can touch the degree of integration and most importantly, consistency afforded by this code reuse.
This ease of development and consistency is what makes KDE so powerful, and why I'm a huge fan of the KDE development model. The people who design kdelibs really understand what it takes to provide a tightly integrated desktop environment.
Do you even realize how moronic you sound?
You are defending a crutch of an application, created for a grievously misguided library whose main motivation appears to have been that "if Windows is doing it, it must be good enough for us", to expose options which are not user-configurable because of an inability to organize the applications' GUIs. Then you go on to complain that an environment that avoids all these problems is a support nightmare. I have news for you: Microsoft Windows is even easier to administer if you want your users to have a minimum of options in completely disjointed applications with idiotic facilities for adjusting functionality that didn't make it to the UI.
GNOME releases regularly and gets new features out into general use quickly.
Yeah, and KDE devs hold all their features back for themselves. WTF does this have to do with anything? The Gnome and KDE point releases occur at about the same frequency.
Stick with Gnome and Firefox if you need the comfort afforded by cutting functionality from programs. "Less is more" has brought Gnome and associated applications to a point where I cannot use them without disgust. KDE always seems to have lived by the principle of "more options, better organized" when preferences are concerned, and what it needs to do is expose even more, and organize better. There are applications where prefs dialogs are organized poorly. Does that mean those options need to be hidden? Hell no. Just reorganized.
WORD.
Thanks for stating eloquently the contempt many of us hold Microsoft in for bullying the software market into mediocrity.
The latest compilers, with the default settings, produce code that's bloated for the tiny machine you describe. I suspect GCC 4 with "-Os" would do better, as long as you were going Gentoo anyway. Also, what happened most likely was that your video setup was suboptimal and had minimal or no acceleration. In general, I find it very hard to believe that a system with the later 2.6 kernels, GCC4, and properly configured video and multimedia packages would have trouble beating an old 2.4 system, especially if it's Gentoo, not saddled with a bunch of services starting by default.
But the fanatics, who buy/rent/go to a movie or two every month, or even every week, are maybe 10% of the population, and they amount to 80% or more of the profits.
What?!
Most people who don't use filesharing extensively buy, rent, or go to watch movies more often than once a month. Your analysis in general, and that datum in particular, is way off base.
Have 2 different groups: Development - that works on a backlog of tasks and incrementally improves various pieces of the product; and Release - that takes versions of the development pieces, matches them with marketing requirements, makes a cohesive product and releases when the marketing requirements are met.
Yeah man. It's called release engineering and all good enterprise (and not-so-enterprise) software projects have it. There's a bunch of groups working on sub-projects, and then there's releng which is responsible for rolling all that into releases.
WTF? This has already been all done.
- data should be stored on the server (centralised backup, provision for web mail when you need it, ability to have an administrator control it, access from multiple hosts)
It's called IMAP.
- server-side spam filtering which can also take easily feedback from the client on what proved to not be spam, or what was and was missed.
Yeah. SpamAssassin and friends.
- server-side addressbook
LDAP.
- should deal only with plain text - non plain text should be flattened to plain text. It would be nice to automatically bounce office files with a message to tell the person to send stuff as PDF or plain text.
News flash: email HAS to be all plain text for servers to understand it. Attachments are still encoded, of course. Writing a filter for mailfilter/procmail/whatever that bounces based on attachment type takes all of 30 seconds. I'm not sure if it can be done in Thunderbird, but I'm sure there's a plugin out there for it.
- effective searching
- very responsive client for reading mail
- very responsive client for writing mail
Already done. Have you even USED Thunderbird lately?
- effective communication between client and server that doesn't require the user to wait
Um. Yeah. It's called broadband.
What?!??
What hardcoded paths? What has to be scanned? The profile is a bunch of files formatted in a standard way with nothing preventing you from relocating it. FFS, I copied my whole profile from Windows to Linux, started Thunderbird, and there it was, using all the data I previously had. If the devs can't make a basic function for importing other profiles from their own program, well, that's just ineptitude on their part.
On a totally unrelated note, the junk mail scanning in Thunderbird is horribly broken, and worse, I can find no way to turn it off altogether...
Do you hear that sound, Mr. Cloricus? That is the sound of inevitability. That is the sound of a joke flying over your head.
Do robots dream of electric sheep?, do bears really shit in the woods?, would a Slashdot editor recognize a proper English sentence if it hit him in the face? The answers will probably come when hell freezes over(pretty soon).
Tovarish! Pishi pravilno!
Yeah, 20 blades with 40 5160s does good for some nice bit crunchin' :)
yeah, I meant 4-socket.
... well, I guess I did say that, so I concede that AMD still has some performance parity left in the 4-core market, if discounting all factors other than the number of cores.
AMD still rocks - without K8, we would not have Intel offering anywhere near the performance for money that we have today. And all my CPUs at home are AMDs, since they were bought before Conroe came out.
E5320 is a 1.86 GHz chip. 8218 is a 2.8 GHz chip - top of the line, an overclocked Opteron that AMD added after Intel started making good server CPUs again. An apples-to-apples comparison would be 8218 vs. X5355 (2.66 GHz and a faster FSB). That's 43% faster clock speed and some effect from a faster FSB. This will still be behind the 4-socket system by a bit, but I never said quad-cores will scream past dual-cores, I said they offer a hell of a lot more value for my workloads, particularly in rack density and heat.
Look, you can be paranoid about your privacy all you want, but if you cannot appreciate the versatility and freedom a car allows - not just in the United States, but especially here - you're either a sorry hermit or are being intellectually weak. I won't even start listing the number of ways having constant access to a car has improved my life.
I think the capital expenditures on public transportation are too low in the US, and I would gladly use public transportation instead of a car to commute were it available in those places where I can't bike to my workplace, but that doesn't prevent me from appreciating the tremendous freedom a car allows me.
they want the "freedom" of being able to decide where they go (even though everything related to cars is heavily regulated.)
It's absolutely ridiculous to imply that a car doesn't offer a massive increase in personal freedom. If you disagree, you probably aren't using your freedom much.
So the moment the health visitor told you that, you should have told them to take a walk and reported them to their supervisors, and told them that if that's their policy, they need to drop it. If this bullshit (government officials in a mandatory test telling you not to educate your kid too much) makes your blood boil, as it does mine, then stand up to it and do your part to make it stop.
I want AMD to succeed as much as the next guy - I definitely appreciate the fact that they spent the last 20 years working hard to gain superiority over a much larger competitor. But I think Hector Ruiz (the AMD CEO) must feel like Yamamoto, the Japanese WWII admiral who allegedly said, "I fear all we have done is to awaken a sleeping giant and fill him with a terrible resolve".
(sorry, didn't read the whole comment before replying)
On most workloads I'm concerned with, Intel's new dual-FSB, quad-channel chipset (5000P) provides enough of a performance parity against AMD's on-die controllers to make the overall performance very competitive against AMD's 4-sockets. Most public benchmarks I've seen on these configurations are pretty ridiculous, using non-parallelizable or I/O-bound consumer software. Granted, my workloads are not terribly memory intensive, so if you're looking for a database server that has to juggle the whole database in RAM continuously, AMD might still be the killer. On the other hand, even when I think about desktop workstations, I'd rather not have a 4-socket monster on my desk and take the slight performance hit using Intel's quad-cores. And of course Intel's 2x quad core is very substantially cheaper than AMD's 4x dual core.
No. At 8 cores, you'll be well advised to buy 2 sockets with 2 quad-core Clovertowns. Which will still wipe the floor against all but the 2.6+ GHz Opterons on all but the most memory-intensive workloads, and have about twice or more the performance/watt of any Opteron. AMD is quite screwed everywhere except on 8-socket servers.