I have an autohotkey script that does the same sort of mapping in windows. It doesn't work in all dialogs but it works in most apps (including vim). I haven't really tried any keyboard hackery in OSX
With the rise of Javascript front-end frameworks (Ember, Angular, etc.), is there really a serious place for large opinionated server-side frameworks any more?
Is Rails destined to be a framework for writing APIs to feed front-end frameworks? And if so, is that enough?
You are quite famous for being loudly dismissive of Rails critics.
But do you ever get the urge to learn from your experience (and mistakes) and build a new framework that's different from Rails?
In other words, if you could burn Rails to the ground and start over without the need to maintain any sort of backwards compatibility, what would you do differently?
That's true from the point of view of an app developer. However it is different from the point of view of a device maker.
The issue is that lots and lots of very popular apps use the proprietary Google APIs. You can't make a device that runs all of these very popular apps without licensing the proprietary Google APIs from Google.
And Google won't license you their APIs unless you follow certain rules that are ostensibly about preventing fragmentation of the Android ecosystem, but are in reality about preventing non-Android devices from being able to run Android apps.
This is why the Amazon Fire and Blackberry 10 platforms could only offer limited Android support. They weren't making vanilla Anrdroid devices, and so Google wouldn't license them their APIs. And therefore they could not run the many very popular apps that rely on the Google APIs.
As a long-time Perl fan, I found Ruby to be an absolute delight. It's as expressive as Perl in some ways and more expressive in others.
I started my career as a Perl hacker, and then moved on to PHP (bleah), before making the transition to Ruby on Rails.
I got my first Ruby job without actually having any Ruby work on my resume. I had the general experience, the business experience, the right attitude, and a desire to learn.
I think most enlightened employers will be more interested in the right fit for the company (personality, general ability) than in finding someone with a precise set of skills on their resume.
X-Subpoena-Report: 9.9 hits, 6 required;
* 1.1 -- BODY: Refers to you as 'the first party'
* 4.5 -- BODY: Contains many words like 'whereas' and 'aforementioned'
* 2.0 -- BODY: Text is in ALL CAPS and poorly formatted
* 1.1 -- BODY: Includes a link to a reputable law firm
* 1.2 -- OTHER: Looking through your files, and it appears you've been BAD
Maybe it would help you think of a novice-oriented desktop environment as a special application that provides a specific (limited) interface to certain system resources (files and printers mostly), and acts as a launcher for productivity apps that also respect this sandbox view.
Think of this type of application in a corporate environment, rather than just in a home user environment
Any good Windows network administrator will attempt to lock down as much of the desktop interface (Control Panel, Run command, etc.) as is possible so that computer users can get on with typing that letter or entering those numbers, without the desktop getting in the way (and without the user mucking up the computer).
Now why shouldn't this desktop environment be running on a Linux computer? Or do you think that a Windows desktop is per se more easily and thoroughly secured than a Linux desktop?
Certainly, I remember having a shell account on my ISP's Sun box and running Pine for my email back in the days when I knew nothing about Unix. Try thinking of the modern corporate desktop in terms of that shell account. Sure, the actual hardware has been moved from the server room to the desk, but it should be possible to retain that model of use.
Look at it this way: The modern desktop user in today's corporate environment is analagous to the single user in the multi-user systems of old.
I'm not sure I understand the argument against drive letters, either from a technical point of view, or from a user interface point of view.
I'm really more interested in drive letters as a UI feature: as a shortcut to a logical "slice" of data; a miniature file system, if you will. H: for data, (C:/data) K: for downloads, etc. (C:/data/downloads) L: for config data (C:/data/config)
This allows quick access to important data areas, and splits out those areas well (e.g. for backup purposes).
Sure, it may appear to be a poor man's ln -s, or useradd -d, but it actually provides some functionality that these *nix tricks don't allow:
Drive letters are valid from any working directory (unlike symlinks).
Each drive letter maintains its own working directory (unlike pseudo-user home dirs).
Not that I wouldn't love to have symlinks under Windows of course. But I really wish that *nix had something as flexible as Windows drive letters. G: is so much nicer than/mnt/cdrom, especially when you can type: H:\projects\misc>cd g:\archive\980909 H:\projects\misc>copy g:.
There is aparently an adventure shell which I've never used, but I've always imagined that it would allow you to pick up and drop files like objects, and to manipulate them using spacial metaphors:
$ get group You take a copy of the group file
$ up You shimmy up the narrow crawl into the nook in the root. (/)
$ xyzzy The world shimmers for a moment and is gone. No wait! It's back again. You are in your home directory, a well house for a large spring.
$ inventory You are carrying: a red hat (being worn) a group file a strange configuration file
$ drop file Which do you mean: the group file, or the strange configuration file?
$ group Dropped.
$ examine configuration file You put the configuration file into the strange emacs machine, which gurgles happily.
You are in a maze of twisty control sequences, all alike...
Well, by default, most windows apps install to something like:
C:\Program Files\[Really long advertisement for Vendor]\[Really long advertisement for program]
That's a filing system that's optimized for finding programs by vendor. By looking at the Programs directory on a random computer, I can quickly answer questions like "Are there any applications by Globby Softworks Inc. on this computer?".
That's real useful.
If you're interested in keeping track of what's installed on a computer, either because it's your own, or because it's one of several that you manage, it is useful to design your filing system around being able to quickly determine which types of programs are installed, and why. For example, installing web and email apps under "C:\Program Files\Internet" will help you quickly answer questions like "Which email client is on this system?".
Also, the default install often means that newer versions overwrite older versions. If you're a web designer, you probably want to preserve your old versions of Netscape for testing purposes, so you have to pay attention during its install and choose sensible install locations. And you have to be able to remember those locations later, when you edit your four identical Netscape icons in the start menu.
I suppose the irony is that Windows installs default to giving the end user this degree of control over file locations, while most of it is wasted on Windows, considering how frequently you have to reinstall.
That, and the fact that this is one question that most *nix apps don't ask you (sometimes you really have to dig); which is odd, considering the longevity of a typical *nix system.
Even in KDE, which was had good keyboard support, there were apps that wanted to do things differently. For example, ALT-F pulls down the File menu in the KDE text editor, but brings up the Find dialog in Netscape...
On a Mac you can't access the standard menus with the keyboard. (ALT, on a Windows box). This sometimes traps you in a dark place and you have to reach for that mouse...
If you really want to find fragmentation on Windows, you don't have to look far:
It's in DLL conflicts and VB runtime version mismatches...and the fact that IE 3 and IE 4 can't coexist in the same Windows installation...
It's in the fact that certain of the MS products require a particular version of IE...
It's in the reality that Outlook and Outlook express are two entirely different beasts, sharing only a name; and that (when IE4 was originally released) you couldn't send email messages between them...
It's in the fact that I have copies of MSIE 3 that swear blind that they are actually copies of MSIE 4 (this when queried via Javascript under oath).
The question is: is fragmentation a technical problem or a marketing problem?
A more important question is: how does fragmentation actually harm a platform? There seems to be a general fear of fragmentation about in the Linux community. But fragmentation doesn't seemed to have harmed Microsoft any; If anything, the confusion just encourages their customers to go out and buy the very latest of everything "just to be on the safe side". In the realm of Consumer Linuxism, this would translate into the latest Redhat or the latest Caldera, wouldn't it?
Slashdot Mirror
I use a text file in the todo.txt format (http://todotxt.org).
I keep the file in Dropbox, and use SimpleTask to edit it on Android.
Good point!
(Replying to undo moderation typo)
I have an autohotkey script that does the same sort of mapping in windows. It doesn't work in all dialogs but it works in most apps (including vim). I haven't really tried any keyboard hackery in OSX
I feel your pain. I bypass the whole issue by mapping CAPS+IJKL to be cursor keys at the xwindows level. Then they work in (almost) any program.
With the rise of Javascript front-end frameworks (Ember, Angular, etc.), is there really a serious place for large opinionated server-side frameworks any more?
Is Rails destined to be a framework for writing APIs to feed front-end frameworks? And if so, is that enough?
You are quite famous for being loudly dismissive of Rails critics.
But do you ever get the urge to learn from your experience (and mistakes) and build a new framework that's different from Rails?
In other words, if you could burn Rails to the ground and start over without the need to maintain any sort of backwards compatibility, what would you do differently?
That's true from the point of view of an app developer. However it is different from the point of view of a device maker.
The issue is that lots and lots of very popular apps use the proprietary Google APIs. You can't make a device that runs all of these very popular apps without licensing the proprietary Google APIs from Google.
And Google won't license you their APIs unless you follow certain rules that are ostensibly about preventing fragmentation of the Android ecosystem, but are in reality about preventing non-Android devices from being able to run Android apps.
This is why the Amazon Fire and Blackberry 10 platforms could only offer limited Android support. They weren't making vanilla Anrdroid devices, and so Google wouldn't license them their APIs. And therefore they could not run the many very popular apps that rely on the Google APIs.
As a long-time Perl fan, I found Ruby to be an absolute delight. It's as expressive as Perl in some ways and more expressive in others.
I started my career as a Perl hacker, and then moved on to PHP (bleah), before making the transition to Ruby on Rails.
I got my first Ruby job without actually having any Ruby work on my resume. I had the general experience, the business experience, the right attitude, and a desire to learn.
I think most enlightened employers will be more interested in the right fit for the company (personality, general ability) than in finding someone with a precise set of skills on their resume.
If you start with Vim you'll get a great text editor with a focus on actual text editing. You can then add plugins to add IDE-like features.
My favourite plugins:
* fugitive
* dbext
* syntastic
* YouCompleteMe
* Command-T
* Unite
* multiple-cursors
* gundo
* sparkup
* easy-align
Replying to this to undo a moderation mistake. Great post!
Except that after World War III, all the C64s lying dormant in all our closets will be the only computers to survive.
So I'd have to say that in the long term, they are a viable development platform.
After all, there's got to be something left for the cockroaches to play with...
X-Subpoena-Checker-Version: SubpoenaAssassin 0.15
X-Subpoena-Report: 9.9 hits, 6 required;
* 1.1 -- BODY: Refers to you as 'the first party'
* 4.5 -- BODY: Contains many words like 'whereas' and 'aforementioned'
* 2.0 -- BODY: Text is in ALL CAPS and poorly formatted
* 1.1 -- BODY: Includes a link to a reputable law firm
* 1.2 -- OTHER: Looking through your files, and it appears you've been BAD
Maybe it would help you think of a novice-oriented desktop environment as a special application that provides a specific (limited) interface to certain system resources (files and printers mostly), and acts as a launcher for productivity apps that also respect this sandbox view.
Think of this type of application in a corporate environment, rather than just in a home user environment
Any good Windows network administrator will attempt to lock down as much of the desktop interface (Control Panel, Run command, etc.) as is possible so that computer users can get on with typing that letter or entering those numbers, without the desktop getting in the way (and without the user mucking up the computer).
Now why shouldn't this desktop environment be running on a Linux computer? Or do you think that a Windows desktop is per se more easily and thoroughly secured than a Linux desktop?
Certainly, I remember having a shell account on my ISP's Sun box and running Pine for my email back in the days when I knew nothing about Unix. Try thinking of the modern corporate desktop in terms of that shell account. Sure, the actual hardware has been moved from the server room to the desk, but it should be possible to retain that model of use.
Look at it this way: The modern desktop user in today's corporate environment is analagous to the single user in the multi-user systems of old.
I'm not sure I understand the argument against drive letters, either from a technical point of view, or from a user interface point of view.
I'm really more interested in drive letters as a UI feature: as a shortcut to a logical "slice" of data; a miniature file system, if you will.
H: for data, (C:/data)
K: for downloads, etc. (C:/data/downloads)
L: for config data (C:/data/config)
This allows quick access to important data areas, and splits out those areas well (e.g. for backup purposes).
Sure, it may appear to be a poor man's ln -s, or useradd -d, but it actually provides some functionality that these *nix tricks don't allow:
Not that I wouldn't love to have symlinks under Windows of course. But I really wish that *nix had something as flexible as Windows drive letters. G: is so much nicer than /mnt/cdrom, especially when you can type:
H:\projects\misc>cd g:\archive\980909
H:\projects\misc>copy g:.
There is aparently an adventure shell which I've never used, but I've always imagined that it would allow you to pick up and drop files like objects, and to manipulate them using spacial metaphors:
$ get group
You take a copy of the group file
$ up
You shimmy up the narrow crawl into the nook in the root. (/)
$ xyzzy
The world shimmers for a moment and is gone. No wait! It's back again. You are in your home directory, a well house for a large spring.
$ inventory
You are carrying:
a red hat (being worn)
a group file
a strange configuration file
$ drop file
Which do you mean: the group file, or the strange configuration file?
$ group
Dropped.
$ examine configuration file
You put the configuration file into the strange emacs machine, which gurgles happily.
You are in a maze of twisty control sequences, all alike...
I'm quite Floccinaucinihilipilificationary myself.
C:\Program Files\[Really long advertisement for Vendor]\[Really long advertisement for program]
That's a filing system that's optimized for finding programs by vendor. By looking at the Programs directory on a random computer, I can quickly answer questions like "Are there any applications by Globby Softworks Inc. on this computer?".
That's real useful.
If you're interested in keeping track of what's installed on a computer, either because it's your own, or because it's one of several that you manage, it is useful to design your filing system around being able to quickly determine which types of programs are installed, and why. For example, installing web and email apps under "C:\Program Files\Internet" will help you quickly answer questions like "Which email client is on this system?".
Also, the default install often means that newer versions overwrite older versions. If you're a web designer, you probably want to preserve your old versions of Netscape for testing purposes, so you have to pay attention during its install and choose sensible install locations. And you have to be able to remember those locations later, when you edit your four identical Netscape icons in the start menu.
I suppose the irony is that Windows installs default to giving the end user this degree of control over file locations, while most of it is wasted on Windows, considering how frequently you have to reinstall.
That, and the fact that this is one question that most *nix apps don't ask you (sometimes you really have to dig); which is odd, considering the longevity of a typical *nix system.
Which WM is most keyboard friendly?
I've only briefly tried one or two.
Even in KDE, which was had good keyboard support, there were apps that wanted to do things differently. For example, ALT-F pulls down the File menu in the KDE text editor, but brings up the Find dialog in Netscape...
On a Mac you can't access the standard menus with the keyboard. (ALT, on a Windows box). This sometimes traps you in a dark place and you have to reach for that mouse...
The question is: is fragmentation a technical problem or a marketing problem?
A more important question is: how does fragmentation actually harm a platform? There seems to be a general fear of fragmentation about in the Linux community. But fragmentation doesn't seemed to have harmed Microsoft any; If anything, the confusion just encourages their customers to go out and buy the very latest of everything "just to be on the safe side". In the realm of Consumer Linuxism, this would translate into the latest Redhat or the latest Caldera, wouldn't it?
Personally I want everybody to do it my way.