The Case Against GUIs, Revisited 720
snydeq writes "Deep End's Paul Venezia advocates the importance of the command line, in light of the increasing use of GUIs in today's technologies, as well as the increasing perception among admins that proponents of the CLI are dragging computing back to the 'dark ages of the C:\ prompt."
This, perhaps... (Score:5, Insightful)
... speaks more of the admins who assume that "CLI" == "C:\ prompt".
Or the idiots who think "CLI" == "the GUI in front of me is therefore made unusable". The people at "GUI Industries" can't make a link or shortcut to the appropriate script?
Why would you trust an admin who can't, as TFA indicates, edit a text file?
Dead on (Score:4, Insightful)
This is dead on. Human beings invented symbolic language because it's simply more expressive than pointing and grunting. CLIs are superior to GUIs for the same reason.
I don't get it. (Score:4, Insightful)
I haven't used windows since '99.
looking around my desktop right now, while posting to slashdot, I have chromium running, and 7 xterms. Two of'em are running irssi, the others are just nice little windows to do various bits of work in.
I live and breathe in a CLI environment. I can't really remember doing much useful in GUI's except lookup information (for which it's suited perfectly well).
But why on earth would you do configuration in a GUI? Why would you ever program in a GUI, instead of vim or emacs?
I just don't get it.
Re:Nobody needs a GUI or CLI (Score:4, Insightful)
Some UI may go this way eventually, but I imagine most written/typed communication is still incredibly valuable.
I imagine it won't be long before we communicate with computers very much the same way we do with people.
For casual simple tasks, that means mostly voice (which computers suck at today) and a bit of gestures (which computers are OK at with a mouse).
For anything complex, though, communication between humans is typically written - and I expect it'll continue to be so for computers for as long as people interact with them -- not because it's great for the computer, but because it's the best humans come to a high-bandwidth precise recordable communication channel.
Re:Dead on (Score:5, Insightful)
Human beings invented symbolic language because it's simply more expressive than pointing and grunting. CLIs are superior to GUIs for the same reason.
CLIs are superior to GUIs in the same way romance novels are superior to sex. Sometimes taking the time to carefully describe things in clear, unambiguous detail is important, but sometimes pointing and grunting is both more effective and and more satisfying.
And when's the last time you edited photos, video, or audio with a CLI?
Dark ages of the C:\ prompt (Score:5, Insightful)
Anyone who thinks that a command line prompt starts with a 'C:\' has no idea what they're talking about.
CLI != DOS (Score:5, Insightful)
While the C:\ prompt would really be the Dark Ages all over again, command lines done right can be significantly faster to use for experienced admins. Not to mention they can also be used in diverse interface environments like serial connections in almost exactly the same way that they are used anywhere else. It's significantly easier to script from the command line as well.
Just about every shell available on a UNIX-like OS is at least ten times better than DOS ever was. Lumping in bash with DOS is like taking a BMW and saying it is equivalent to a Yugo because they both have four wheels and an internal combustion engine.
There's certainly a place for a GUI, and well designed GUI apps are lifesavers in big environments where seeing a graphical representation of 1000 servers as icons and being able to click and drag to select and execute group commands is a much easier way to work with that many servers. Having 1000 lines of text zoom by is going to be hard to take in at a glance. I love the GUIs that I use for certain tasks like visualization and management of complex environments.
Still, anyone who considers the command line to be the Dark Ages either doesn't know how to type, doesn't understand how to use a CLI, or is simply trying to sell you a pretty GUI app.
Re:CLI is no longer essential (Score:5, Insightful)
Uh, using a GUI doesn't preclude you from editing text. Windows is developed and compiled in Visual Studio, which is a GUI-based IDE.
Re:I don't get it. (Score:4, Insightful)
easy of use. I get far more from VS then either emacs or Vi provide. And yes I have used them extensively.
A good GUI make configuration a snap. Should it be the only way? of course not.
It's funny, people complain about GUI speed, but then never use the shortcuts. They say CLI is better, but then use all kinds of shortcuts.
... and a far higher payoff function (Score:2, Insightful)
The difference between a GUI and a CLI is that a CLI rewards understanding with capability.
GUIs are appropriate to non-expert systems where functionality must be patently clear. Think of your smartphone / PDA, or a consumption-oriented computing device (iPhone, iPad, PowerBook).
If you're administering or creating content, you'll eventually want the power of a real scripting tool.
The Memory Connection (Score:4, Insightful)
Command lines discriminate against those with poor memories. GUIs make it possible for people who can't remember detailed shit to be productive without having to constantly refer to some other resource(s).
I learned English well enough that I rarely require a dictionary, but I was never so lucky with programming languages and other syntaxes. I love my GUI... when it's implemented correctly. Paul Simon wanted his Kodachrome, and I want my GUI.
Why not both? (Score:5, Insightful)
Why does this always have to degenerate into a Campbell's Chunky Soup "Fork or Spoon" debate? Why not just use the most appropriate interface for the task at hand?
A GUI can be shit for some things, and (unless you live and breathe CLI) a CLI can be too complex and unwieldy for other things.
Re:This, perhaps... (Score:4, Insightful)
GUI = makes it easy to do one off tasks, because the interface can be made intuitive.
CLI = makes it easy to do repetitive tasks, because they can be easily scripted.
Even repetitive image manipulation is best achieved with scripted command line tools. Don't believe for a second someone had inserts watermarks into photos!
Re:Raises hand (Score:4, Insightful)
Re:I don't get it. (Score:4, Insightful)
But why on earth would you do configuration in a GUI? Why would you ever program in a GUI, instead of vim or emacs?
Because it's way, way easier. Don't get me wrong, I love my VIM, I use sed at least weekly, I've built 5,000-line programs in awk (awk, not gawk, and god help me not by choice), but here's why I use a GUI to program (when it's available):
Do I miss stuff from VIM when I'm working in Visual Studio? Sure. But if you know the tools equally, and I know VIM like I know the palm of my hand, Visual Studio is just plain faster. But, your entire premise is invalid--a GUI or CLI is not a binary choice for a subject as large as "programming". Individual activities are better suited to one or the other, and a savvy user will use the most effective tool for the task at hand. A GUI can host a CLI or simulate it entirely, a CLI can't do either, therefore GUIs are objectively better unless the resources required for the UI is a concern.
Re:This, perhaps... (Score:4, Insightful)
I rely heavily on scripting for one reason: scripts can be captured in version management repositories. GUI interactions can't.
Aside from that, my database creation scripts are huge. I'd hate to have to open each script in a GUI tool to run it instead of being able to run them all as a batch (as intended) through a bash script (which in turn invokes the compent bash scripts that actually invoke the database CLI.)
Re:This, perhaps... (Score:5, Insightful)
Why would you trust an admin who can't, as TFA indicates, edit a text file?
Or write a safe regex:
$ echo '123a123b123c123' | sed 's/123.123.123./321.321.321./g'
321.321.321.123
Gotta love the way he hand-waved the "half hour on a Wednesday" to do the whole transition, too. As someone who's done it, it is never, ever, ever that simple. Have fun finding bugs like the above, fixing them, finding out that your fix invalidates your entire solution, approaching it from a different angle, looking up an obscure syntax you can't remember even though you use sed/awk/grep every day, finding out you've got a slightly older version or yours wasn't compiled with --enable-double-buttgrep ... sure, the "30 minutes" sounds great when you're trying to make a point, but it also sounds ridiculous to anyone who's actually done it twice.
I love me some CLI, and I'd lose my mind if I didn't have it available, but promoting either one to the exclusion of the other is asanine.
Re:This, perhaps... (Score:5, Insightful)
Re:This, perhaps... (Score:2, Insightful)
I think you are looking at it in the wrong light.
The GUI would certainly take longer to develop. No question. Thing thing it is tends to save a lot of time on the end user side of things. That's where that 10x extra development time is paid back many fold more.
I've developed in open source text editors and in Visual Studio. I'd pay for VS in a heartbeat because of the time it saves in coding.
Spend 10x more time developing the good GUI, charge the same as the CLI only product, you'll find you might generate 1000x more sales because your product is much easier to use than the CLI only competition.
Re:First post (Score:5, Insightful)
By your own admission you've devoted half a lifetime or more to developing computer skills. Should everybody have to do that? Are people who don't devote half a lifetime specifically to computing skills "stupid' and "fearful"?
This is just a form of elitism. I'm not even going to call it intellectual elitism, as preferring CLIs is no more a demonstration of superiority or intellect than preferring GUIs is a demonstration of inferiority or stupidity.
Re:First post (Score:5, Insightful)
The trouble with CLIs though is that the functions aren't always named whatever you would expect so you *have* to look up the function name and formatting.
If a command line could be written as:
"Take this image, resize it by 50% and increase the contrast 10%" then people would use CLIs all the time.
Instead it goes:
ImageProcessor ... /help ... ... /help ... .. /help ...
ImageProcessor -scale 50p...
Image Processor -scale 50p -filter (contrast,1.1)
etc...
You have to keep looking through a huge documentation system. GUIs at least present you with the documentation *AS* the interface.
You don't need to know what the "Contrast" filter is exactly called, you just look through the list by default and choose the one that closest matches what you want.
If we had smarter more adaptable CLIs people would be far more familiar instead you find out you missed a comma somewhere and it takes you 5 minutes to track that down. Or you try to figure out if adjusting contrast is even a function at all.
Re:First post (Score:5, Insightful)
Yes, and yes, if their career is administering computers (or computer networks, which is what the article is about).
Re:This, perhaps... (Score:5, Insightful)
GUI = makes it easy to do one off tasks, because the interface can be made intuitive.
CLI = makes it easy to do repetitive tasks, because they can be easily scripted.
That task split between GUI and CLI is exactly what I think went wrong when GUIs were first designed.
What we should have done was created a GUI which has exact formal one-for-one-correspondence with the underlying CLI, so that for any given task, you can use the interface of your choice that works for you - or create new graphical interfaces for the special custom jobs you end up doing multiple times.
And it wouldn't have been that hard to do. GUIs involve message passing, but we chose not to expose those messages as scriptable 'commands' at the same level as a CLI. We hid them and required C++ and other system/application languages to send those messages in full, then exposed a tiny fraction of them to various "scripting languages". But you can't write all applications in scripting languages, so it's a lossy conversion filter.
The result is, we've got at least three levels to our systems: a reasonably transparent 1960s-batch-system derived command/process/file based layer, an opaque 1980s Smalltalk-derived object/component layer, and finally combined applications/GUI/preference files launched by the 1960s layer which orchestrate the 1980s layer to do stuff. We can't use the 1960s layer (CLI) to do everything the 1980s layer (GUI plus services/components/DLLs) can do, because the 1960s layer has fundamental connection concepts like piping and "everything is a file which is an I/O stream" which simply don't exist in the 1980s "everything is an object" layer. It's impedance mismatch at the OS level.
The big problem with an impedance mismatch is that you can't solve it by adding more lossy-conversion layers. It can only be really "solved" by creating a new abstraction which has the features of both, but that seems impossible now that we have a huge weight of legacy code. I'd been hoping that Linux in the 2000s would give us some hope of breaking away from the Windows cruft and coming up with a simple unified desktop framework. But GNOME and KDE seem to have just fragmented the situation even more.
Wish I could see a plausible way out of this mess.
Re:Raises hand (Score:5, Insightful)
If a human is required, it's editing.
That's your opinion. I think the point the original poster is making (certainly the point I'd make) is that our interfaces should be agnostic as to whether a human or a decision-making algorithm is driving them.
Because otherwise, how are we going to take control of our workflow and delegate the automatable parts of it to automation, if the interfaces we use stubbornly insist that 'no, a human must be in the loop to do that!'
This attitude, which sadly is rife among GUI designers, is keeping us stuck in the dark ages. It fundamentally shouldn't matter if a human or a program is doin any job on your desktop. End of story. It's not the interface's job to decide. If a script can take a look at a JPEG, apply a beauty algorithm and decide to fudge the colour contrast and recrop - and if a human can look at that afterwards and decide that it's good and keep it - well, then that's editing, isn't it? But it's done by a partnership of human and machine, as it should be.
Don't force us to decide between human and machine for every job, and especially, as an application designer, don't impose your conception of what tasks should be done by which. As a user, closer to the coalface, I might well have a better idea. If your software gets in the way of my automation, it's wrong.
Re:This, perhaps... (Score:3, Insightful)
and once again destroyed the ability for users to do their own interface design and to program their own workflow.
There's a reason for that: 99% of users don't want to do their own interface design, or program their own workflow.
User interface design, and workflow design, are difficult. Users expect the software developers to have already done that work for them.
Re:First post (Score:5, Insightful)
It is often worth doing cost-benefit analysis and determining if the time spent reading the documentation is worth the time that will be saved by using CLI.
If someone asks me, for example, to add a user to active directory, it doesnt really make sense to research how to get LDIFDE or whatever to do an import properly. If, however, they hand me a CSV file of 100000 contacts and asks me to import them to active directory, then you better believe Im breaking out cygwin and awk documentation, and touching up on the active directory CLI tools. I may spend a few hours learning how to work them, but I will save hundreds of hours in time.
This isnt a "heres the right answer" scenario. You have to use your brain and your experience to choose the right tool for your job. That, as much as anything else, is the mark of a good systems administrator, not knowing every CLI in existence. A knowledgable SysAdmin can be inefficient if he needlessly ignores some tools (GUI) that are perfectly suitable to a task just so that he can use the same tool for every task.
Re:This, perhaps... (Score:3, Insightful)
That task split between GUI and CLI is exactly what I think went wrong when GUIs were first designed.
What we should have done was created a GUI which has exact formal one-for-one-correspondence with the underlying CLI, [...]
And that GUI would be unusable for most of the time. Making a good GUI is not about just slapping together a number of widgets, it is about usability. For example, take curl. It has 139 or something options. A GUI would have to present all of them and if the GUI is to be usable it shouldn't be too hard getting to any of them, that is, it should not be hidden away somewhere deep down. Either you get a very cluttered GUI with every option visible or a GUI that is more suited for a certain task but that is not as good with general tasks.
I think that the split is exactly what is needed to make the right tool at the right time. There is no silver bullet. For some tasks GUI is superior, for other CLI rules.
As an analogy, think of communication. Sometimes we prefer written (contracts for example). At other times we prefer verbal (talking to your therapist). You cannot say that there is ONE right way to communicate and all forms of communication should adhere to the same strict doctrine.
Having said that, it should be obvious that having skill to use BOTH CLI and GUI gives you flexibility and power. The problem today is that too many people are afraid of using a CLI and try to find GUI versions, complaining when they aren't powerful enough. The problem is most likely because of the task they wish to do.