Drupal Views 2.0 UI disappointment: Cleverness versus Usability

We encourage users to post events happening in the community to the community events group on https://www.drupal.org.
dreftymac@drupal.org's picture

There is something really nice about the new Views 2.0 GUI, and something really disappointing.

The really nice aspect is this: it appeals to the developer in me, because it's really clever.

The really disappointing aspect is this: it appeals to the developer in me, because it's really clever.

....

The developer in me is the part that feels motivated to contribute to open-source projects, out of a sense of idealism and a sense of independence and even a sense of pride in accomplishment.

But the user in me does not like cleverness. It's the user who ultimately has to use the stuff that developers make to get his or her work done. The more clever a user interface is, the more the user has to think about the UI. The user does not want to think about the UI. The user does not even want to think about Drupal. It's just a means of getting work done so he can go home sooner when the work is done.

It's hard to quantify what counts as cleverness but one rule of thumb is pretty clear: if you have built something that looks like nothing else of its kind, and it was fun to build, you may have just fallen into the trap of "cleverness at the expense of usability".

No offense meant, but that's exactly how the user in me feels about the new Views 2.0 GUI. It misses out on some pretty basic common practices of GUI usability. That's very frustrating.

....

Anyone who develops UI might want to take a look at these links. It's from Microsoft, but that's not the point, you can freely despise some of their practices while still gleaning knowledge from them, in the rare case where they actually dispense it for free.

http://blogs.msdn.com/jensenh/archive/2008/03/12/the-story-of-the-ribbon...
http://msstudios.vo.llnwd.net/o21/mix08/08_MP4s/UX09.mp4

Anyway, not to be totally negative ... there's obviously a lot of good stuff in there. Hat's off to all who contributed to all the significant improvements.

Comments

Sorry, 'clever' is a crutch

merlinofchaos's picture

Sorry, 'clever' is a crutch to describe something that's different from what you're not used to. I don't actually agree with that description.

Giving me links to Microsoft UI design is useless. This is not Microsoft Windows, this is the web. We're extremely limited and this is special purpose. I think the user has to think about the UI the first few times they use it, that's absolutely true, but it's not because of the UI, it's because they have to think about what they're doing. A lot of people mistake the fact that they have to think about their data as having to think about the UI. That's entirely the wrong thing. Unless you want the UI to do all that for you, in which case let's just get rid of the UI and hope that a click sets everything up the right way.

I'm sorry you don't really like the design of the UI. I personally find it wonderful and easy to use and it goes a long way toward accomplishing our goals. Is it perfect? No. Unfortunately, I can't find anything in your criticism that will help improve it. Instead, I find a very abstract "This isn't perfect, this isn't [blah]" but, well. Tough. Configuring Views is a difficult task. It's a hundred times easier in Views 2 than it was in Views 1. In that sense, I've really met my UI goals.

We weren't trying to be clever. We were trying to do something that is intrinsically hard with a web based UI system. This would be a picnic if I had a real suite of OS windowing system tools to make things easier. It's not.

Criticism without direction is pretty useless. I'm sorry you don't like it, but there's nothing I can do about that.

dreftymac@drupal.org History

skor's picture

dreftymac@drupal.org

History

Member for
2 hours 38 min

Harsh criticism from somebody who joined less than 3 hours ago.

To be fair his d.o account

merlinofchaos's picture

To be fair his d.o account is almost a year old: http://drupal.org/user/211297

It misses out on some pretty

merlinofchaos's picture

It misses out on some pretty basic common practices of GUI usability. That's very frustrating.

What's frustrating is that I'm rereading this trying to pick something useful out of the criticism, and this is the most useful sentence I have. Sadly, I don't know what pretty basic common practices of GUI usability you're talking about. I'd be tickled if you could define them.

My gut instinct tells me that most of what you could list here are things I've thought about and rejected for various reasons, but it's hard to say. Since I can't read your mind.

:Harsh criticism from

dreftymac@drupal.org's picture

:Harsh criticism from somebody who joined less than 3 hours ago.

@skor: Critique, but sincere and out of respect for the project and the people who've contributed. Many times people contribute in ways that are not readily apparent from online stats alone.

:Sorry, 'clever' is a crutch to describe something that's different ...
:Giving me links to Microsoft UI design is useless ...
:This is not Microsoft Windows, this is the web ...
:We were trying to do something that is intrinsically hard ...

@merlinofchaos: With full respect, these responses all seem to indicate that you did not even glance the links I provided, and you have no intention of doing so. If so, fair enough, that is your right.

I do realize, however, that this will be viewed by more than just you and me, so some other people may benefit from the principles enumerated there, it's quite good stuff.

I know full well that developing for the Web is difficult. It's difficult to do even a mediocre job, which is why I salute you and your hard work. Nevertheless, good design principles apply everywhere. Good design principles are not contingent upon the toolset. No matter how fancy or limited the tools are, there's always a way to improve how those tools are used.

:I personally find it wonderful and easy to use ...

I should hope so. If even you had difficulty with it, that would be a very bad sign indeed. From a usability standpoint, however, that doesn't provide much validation. That's not criticism against you, that's just a basic principle of ordinary usability review -- the most useful tests are those done with people who are not already intimately familiar with the software component being tested.

:I'm sorry you don't really like the design of the UI.

Please re-read what I said. There's a part of me that likes it a lot. Unfortunately, that's only half the picture.

:This would be a picnic if I had a real suite of OS windowing system tools ...

No, it wouldn't. That's the point. Building excellent user interfaces (from the user standpoint) does not just magically happen based on the sophistication of the tools. It takes a lot of deliberate effort to get things right based on example, study, and willingness to accept feedback.

I'm sure we won't see eye to eye on everything. I'm glad you took the time to respond. Thanks again for all your efforts, and congrats on the areas where you've found success.

@merlinofchaos: With full

merlinofchaos's picture

@merlinofchaos: With full respect, these responses all seem to indicate that you did not even glance the links I provided, and you have no intention of doing so. If so, fair enough, that is your right.

I did glance and them, and that's all I did. I'm allergic to web video. Actually I'm basically offended when the answer is "Watch this video". I hate screencasts and I hate watching video taped presentations even more. You want to talk usability problems? The whole web video phenom that's going on is a usability problem right there.

Sorry. If you've got concrete things to say, say them, but 90 minutes of video is not an investment I make for you OR the guy offering me a 3 day cruise, either.

Nevertheless, good design principles apply everywhere. Good design principles are not contingent upon the toolset.

I'm starting to find this offensive. If you can't name what good design principles I violated, please don't reply. Because now you've got TWO posts full of generalities and backhanded accusations.

No, it wouldn't. That's the point. Building excellent user interfaces (from the user standpoint) does not just magically happen based on the sophistication of the tools.

Excuse me? I do find this offensive. Please pretend, for a moment, that I'm a professional programmer that's done this for 20 years. I know a thing or two about UI, and I don't need Microsoft (who is responsible for some good UI but also some of the worst UI ever invented) to tell me what good design principles are.

I should hope so. If even you had difficulty with it, that would be a very bad sign indeed. From a usability standpoint, however, that doesn't provide much validation.

And your generalities provide no validation either. A group of people spent months designing this UI, and going through a process to do it.

Please re-read what I said. There's a part of me that likes it a lot. Unfortunately, that's only half the picture.

Your'e right. What you said is actually much worse, because it's an insult.

Your entire message is filled with that "Developers are stupid, developers don't understand UI, they don't understand how other people think," mentality that I just despise. People say that, then other people who don't know either side repeat it, and pretty soon we've got this users vs developers mentality (trust me, it's there) and the developers are evil people who are too stupid to do it right. It's a bunch of poppycock.

Give me something concrete to consider or shut the hell up.

:Give me something concrete

dreftymac@drupal.org's picture

:Give me something concrete to consider or shut the hell up.

@merlinofchaos: I will be happy to give concrete examples. You might want to reconsider your obvious offense, however. As I said previously, I realize that you and I are not the only people who may see this. Other people will have a chance to benefit even if you won't. If you feel your experience and skill justify your utter disregard and contempt for sincere and respectful feedback, you are more than welcome to feel that way.

This may offend you even more, but really it's more than just your personal feelings at issue here. Either you are willing to accept honest feedback, or you're not. If you're not then it won't matter if the examples are concrete or abstract anyway.

Trading resumes or decades of experience really is quite irrelevant. I've worked with plenty of top-notch programmers who simply are not that great at UI design, and vice-versa. It's not a matter of "good" versus "evil". I have decades of industry experience and still learn something new every day. Sometimes from users who don't know or care a thing about programming.

If you really are strongly offended by different perspectives and people who are enthusiastic about sharing ideas and honest viewpoints (as well as web video and however many other common trends that are constantly changing the state of the web) all I can wish you is the best of luck in blocking it all out.

so where are they?

catch's picture

@merlinofchaos: I will be happy to give concrete examples.

I don't see a single concrete example yet on this thread. have you posted any issues to the Views issue queue at all? How about an annotated screenshot showing where you think things went wrong?

As to other people benefiting from your feedback, I haven't seen any feedback yet, just a lot of generalities and truisms. If you're feedback was 'sincere and respectful' it would amount to more than "here's some videos from Microsoft, watch them, then guess where I think you went wrong". Lots and lots of people have given concrete feedback on the Views 2 UI, and others were involved designing it other than merlinofchaos, spending a bit of time to actually research some of the previous discussions might have helped you enjoy a more productive conversation here, If you've done all that research (and reading further you say that you have), then there's no evidence of that here. I'd like to see some answers to Bill Fitzgerald's questions.

I like some of what I saw

mfb's picture

I like some of what I saw when quickly scanning thru the video. I have to say if you want everyone to watch a 90-minute video about MS Office, couldn't you also make some concrete UI suggestion for Views (in descriptive or screenshot format).

Sure, screenshots are coming

dreftymac@drupal.org's picture

Sure, screenshots are coming ... I should emphasize that the video was just for an overview of general principles, clearly it is not Drupal-specific and clearly there is a LOT there. The main point is it's available for anyone to see. There's a lot to be gained from reviewing ideas and methods of the company that has arguably done the most work in the area of Usability.

Speaking of more "Web-relevant" stuff, however, there is still a lot of freely available information for people who care a lot about usability and excellent Web-GUI design.

http://developer.yahoo.com/ypatterns/
http://en.wikipedia.org/wiki/Don%27t_Make_Me_Think
http://www.usabilityinstitute.com/resources/userInYourFace/userInYourFac...

Anyway, these are all just principles. People always want specifics (if only to have something concrete to disagree with) so that's coming also.

Way off base here

bonobo's picture

And I say this from the perspective of someone who has written end user tutorials for Views and Views2 --

The Views2 UI is incredibly easy to use. Views is powerful; leveraging that power requires time to get familiar with the options.

Given a choice between 90 mins watching a usability video or 90 mins learning how to navigate the Views2 UI, I'd recommend the latter.

Also, if you have such strong feelings, where were you when http://groups.drupal.org/node/6288 and http://groups.drupal.org/node/8429 were occurring?

Or, look at different approaches, such as http://drupal.org/project/simpleviews or http://drupal.org/project/advanced_help (getting into more documentation embedded in/alongside the UI) -- there are avenues for constructive involvement here, and you will lend credibility to your criticisms by including specifics and creating solutions.

But, as one example, the UI for configuring arguments in Views2 is phenomenal. In Views2, you can configure very complex views without the need for any argument handling code. This was not possible in Views 1, which placed arguments beyond the reach of non-developers.

I have no illusion that I will convince you, but I hope at least that you will take the time to ground your opinion in specifics before tossing out very general criticisms of high quality work that has been undertaken in public for the last year.

Cheers,

Bill


FunnyMonkey
Tools for Teachers

:I have no illusion that I

dreftymac@drupal.org's picture

:I have no illusion that I will convince you ...

Convince me of what? Who said I disagreed with you?

You seem to be making some very bold assumptions:
1) that I am not familiar with the work that has gone on here;
2) that my "newness" in these proceedings somehow represents a detriment (you ever hear the benefit of "fresh eyes"?); and
3) that I don't recognize that excellent work and improvement has been done here (I do recognize that and I've acknowledged it since my very first post in this thread);

I am familiar with all the content you linked and I wouldn't have bothered chiming in here if I hadn't done homework and gained some background first. You can't be disappointed in something you don't care about in the first place.

There is a very basic point here. People who would be really impressed by Views 2.0 (and by extension, Drupal itself) will not be as avid as you or I about taking the time to watch a screencast about how to use the GUI ... sure, they may do it, but this very thread proves that not everyone is a fan of web video, even if they are given direct links. This translates into less people using Drupal, and fewer people uprading to vesrion 6, and fewer people bothering to join this group and contribute.

Bold assumptions? On the contrary.

bonobo's picture

Bold assumptions? On the contrary. I strive to be meek when I assume.

All kidding aside, fresh eyes are essential in a design/build process. However, so is an understanding of context, and, up until your response to my comment, you did not mention/reference, at all, that you had read through the past thoughts on the Views2 UI -- and these threads go back over a year, and represent countless person-hours of thought and development.

RE: "that I don't recognize that excellent work and improvement has been done here" -- fwiw, that was not apparent from your posts here. I re-read them several times before responding, and that still has not come through. I definitely get that your intentions are good; however, your approach to giving this feedback could be more effective by grounding that feedback in specifics.

And it does get back to specifics. If you find the existing UI too "clever," what are your plans for making improvements? How does simpleviews fall short? Why won't advanced help achieve these goals? What should the UI for configuring filters look like? Should the whole thing be more wizard-based? I'm not necessarily saying that any of the above suggestions are good, but they are at least grounded in specifics.


FunnyMonkey
Tools for Teachers

:How does [the existing UI]

dreftymac@drupal.org's picture

:How does [the existing UI] fall short?

Really it falls short only if you consider that (at least from what I've seen) it scares people. That much seems to be readily acknowledged by most folks. At very least, there's an apparent open admission that there is a learning curve attached to the UI itself that is above and beyond the learning curve attached to the basic underlying principles of data manipulation that would be present regardless of what the UI looked like.

I believe there is some unnecessary complexity that can be dealt with easily.

For example, take a look at screen000.png. The first thing users see is pretty much an admission that they probably won't be able to figure out what's going on. Even if this were software for an air-traffic control system, or surgical robotics, that can't be a good sign. Complexity should not be a code-word for incomprehensibility.

Another example, its great to have context sensitive help, but not so great to have something that is not intelligible without absolute reliance on every single line of the help, along with a screencast or two in support of the help before you can get a "feel" for how the UI even works. Imagine how a user feels from looking at screen001 for the first time. If they stick in, its because they are already a Drupal fan and they know it will be worth it.

Another example, there are buttons and functions that are closely related, but they are strewn disparately about in some places. Add display and Remove display are a good example, see screen002. If you already know what they do and what they're there for, its no problem. The complexity of the system, however, is no reason to keep these from being organized together in a clean and logical way.

Another example, screen003 the "gear" idea seems like a nice "javascripty goodness" way to selectively obscure and display details when the user wants to fiddle with a specific setting, but again, why introduce a totally new metaphor that does not exist in any mainstream web application? Has anyone seen this kind of thing before they got used to the Views UI? Selectively obscuring screens is not a new invention, not for the web or any other design context. Why invent a new way of doing that? Even if it is a well-known thing that I've just never seen before, I've definitely never seen it in Drupal before.

Now for the part where you get to disagree with me in a concrete way ...

First off, it would be great if there were a single visual framework that held all of the interaction that the user needed to make with the UI. See concrete001. Sure, you may hate the layout being used here, but the point is there are many many applications that use this kind of structural metaphor. No wheels to re-invent.

It doesnt take much thought to understand the layout. Moreover, it gives the user a visual cue of how the entire module is organized in terms of features.

Secondly, there's definitely got to be a way to group items selectively such that its never necessary to invoke the "gears" metaphor. If content is too verbose to fit into one screen, it can be separated out into an individual item within a pre-existing hierarchy, or it can be ordered as a consecutive series of links (kind of like the way "themes" does it when you toggle between the available themes that apply to a node).

What's worse, there are clearly places where the "gears" metaphor could have been used, but it wasn't. That means its not only novel, but inconsistent.

Finally, the items that include a "plus" icon (for adding items) and "arrow" icons (for reordering) are similarly disjointed and separated from any visual cue of what they actually do, at least until you click on them or spend a few iterations of watching screencasts. All the items that require the "add item" and "arrange items" metaphor could similarly do this all in once place. Sure, its the web and its hard to do, but it's not impossible. (see concrete003 for an example).

Thats about it for now, there are other items of course, but that should be enough to convey the basic point. I know this is not the first time some of these points have been made btw.

Multiple buttons strewn

merlinofchaos's picture

Multiple buttons strewn separately on the screen

You don't put buttons next to each other just because they're buttons. You try to put buttons near to what they DO or operate on so that there is some visual indication. 'Remove' is kind of a weird one. IMO it has to be somwhere inside the display tab because it's operates on the display so it can't be in the drawers to the left.

troubling indication that this won't be self-explanatory

Yes, welcome to the wonderful world of query building. If you can make query building self explanatory, I'll happily give you the keys and go retire. I personally don't think it's possible. There's a lot of data, and new users come to Views without any clear idea of what it does or what they want to do with it.

Windows pretty much sucks too when you don't know what it does or what to do with it too. So does MacOS.

Another example, screen003 the "gear" idea seems like a nice "javascripty goodness" way to selectively obscure and display details when the user wants to fiddle with a specific setting, but again, why introduce a totally new metaphor that does not exist in any mainstream web application?

Yes, you're right. I set out to obscure things. Absolutely.

In fact, the gear mechanism exists because of the order that choices must be made in. As much as I really didn't want to put the gears in, you have to make some choices before other choices can be presented, and the gear mechanism exists to effectively do multi-step forms within the system. I know, it's much easier to think I did it because I was really excited over my cool new toy, but no, sorry. I wasn't cumming on my keyboard over an opportunity to do cool javascript stuff there.

This illustrates an incredibly fine point. In the "Default: Style" screenshot you have, that's not possible, because the bottom half of that form relies on choices made in the top of that form. It needs to be two step to function.

Really it falls short only

merlinofchaos's picture

Really it falls short only if you consider that (at least from what I've seen) it scares people. That much seems to be readily acknowledged by most folks. At very least, there's an apparent open admission that there is a learning curve attached to the UI itself that is above and beyond the learning curve attached to the basic underlying principles of data manipulation that would be present regardless of what the UI looked like.

I strenuously disagree that the learning curve attached to the UI is above and beyond the learning curve attached to the basic underlying principles. In fact, there is plenty of evidence to show that people who really understand what Views does learn the UI in a VERY short period of time.

It's scary for people who don't understand what Views does and are intimidated by the fact that there are hundreds of options.

The learning curve is primarily in understanding things like "What is a display", "What is the default display" and overrides. I don't ever recommend screencasts, though a lot of other people like them. The getting started has a couple of nice tutorial bits to walk people through setting up a view, but you'll note that those tutorials are designed to explore data concepts as much as they are UI concepts.

Finally, the items that

merlinofchaos's picture

Finally, the items that include a "plus" icon (for adding items) and "arrow" icons (for reordering) are similarly disjointed and separated from any visual cue of what they actually do, at least until you click on them or spend a few iterations of watching screencasts. All the items that require the "add item" and "arrange items" metaphor could similarly do this all in once place. Sure, its the web and its hard to do, but it's not impossible. (see concrete003 for an example).

Again, these are very nearby what they do. There is a visual box, and inside this box is a list of items. You can add items to the list or rearrange the list. There's some argument that perhaps the add should be after the list not before the list, but it's consistent with Panels, which also contains the add icon in the upper right hand corner of the box where you're adding an item.

The + icon is fairly universally recognized as 'add', and they are near the things they manipulate. The tough one is rearrange, but it was difficult to come up with a metaphor that just said "re-order-this-list" to the user without taking up a whole ton of space. You can't just put drag & drop right on the summary page because a degradability requirement meant that no form data could actually be on the summary (risk of data loss too high) so rearrange had to be a click through. You can't have 'remove', 'add' and 'rearrange' all at the same place. Remove operates on the currently selected item, while rearrange operates on all existing items and 'add' operates on items that don't yet exist.

Another item

dreftymac@drupal.org's picture

Another item

  • "Changes to this view have not been saved yet"
    This is indicated with an asterisk next to
    Manage View : Save Changes
    (optionally also next to any unsaved individual
    settings)

Also attempting to re-attach images that didnt make it.

Reasons I'm not fond of this

merlinofchaos's picture

Reasons I'm not fond of this screenshot:

This setup makes it pretty hard to see what's in a view. You have to drill down through extra clicks to see anything, and you can't see enough stuff together. This means you have to rely on human memory to get anything big picture.

You don't really save the user anything by hiding so much stuff, since a user is going to have to click through stuff to see anything.

You want to talk about violating UI principles? 'Save changes' absolutely should be a button. What is 'Exit'? That's a nice concept if this is a windowed application.

You have the arrange button there but it's not near what you're arranging. Right there it's near one of the the things you're arranging. I assume that's intended to show a kind of hierarchical drill down, where first I click sorting and get a list of everything I have sorted on in the View, then I click on one to get its settings.

The first time I responded

merlinofchaos's picture

The first time I responded to this I think I had missed the idea that you had all the forms for the criteria together. I abandoned this because each criteria defines their own forms; they are separate and unrelated, and trying to merge them all together introduces difficulties that I was specifically avoiding. That's why each item in the list is edited independently.

stop... count to 10.

markfoodyburton's picture

Very disappointing thread,
Views 2 is awesome, and the UI needs to provide access to that power, while remaining easy to use. I believe whats there is as good an approach as any other, but I will confess to being a little scared by it. Except by removing access to some of the power of views 2, it seems to me inevitable that the interface will be scary.

I have some concrete suggestions.
I do not believe the approach that has been taken is the wrong one, far from it, I believe its a very good approach, so my suggestions are 'deltas' from it.
Equally, I am not convinced that any of these suggestions are necessarily the right ones, so I do not think they should be submitted as feature requests until they have been discussed.
Finally, None of these suggestions 'scratches' me enough to want to fix them for my own sake - I believe what is there is perfectly workable. However, the rules of this game, as far as I am concerned, is that it would be perfectly reasonable to expect me to contribute to fixing these things, if enough people said they wanted them fixed.

On the list of views, I think the 'default' should be to somehow order the list so that disabled views fall to the bottom...
It's currently alphabetical, I'd suggest to have an 'enabled' and a 'disabled' section.

When editing a view, there is an update button on the panel that opens at the bottom to edit 'details' (of whatever your editing). At least to me, it wasn't obvious that I should also hit 'save' once I've finished updating all that I wanted to update. - I'm not sure how to best address this, but maybe calling the button 'refresh' rather than 'update' may just about do the job!

I also wondered if it would be possible, when you click on something to edit the details, if the browser window could be moved down to the edit box, because when I first used the UI, I didn't realise it was there (and I was using a small-ish window, so I didn't really notice it)...

Ok - so those are my concrete suggestions - I hope others will make concrete suggestions (and maybe much better ones than these).

Cheers

Mark.

Concrete points!

merlinofchaos's picture

On the list of views, I think the 'default' should be to somehow order the list so that disabled views fall to the bottom...
It's currently alphabetical, I'd suggest to have an 'enabled' and a 'disabled' section.

This has been requested 3 or 4 times, and I agree with it. There's a fair bit of work that could still be done on the list screen and this one can definitely happen. I also want AJAX updating and $_SESSION storage of the changes so that contexts don't get lost, but they weren't quite important enough to push back a release.

When editing a view, there is an update button on the panel that opens at the bottom to edit 'details' (of whatever your editing). At least to me, it wasn't obvious that I should also hit 'save' once I've finished updating all that I wanted to update. - I'm not sure how to best address this, but maybe calling the button 'refresh' rather than 'update' may just about do the job!

Hmm. This is a tough one. Generally, web-based interfaces always require you to submit changes...except where they don't, of course, and the rise in auto responsive interfaces has definitely confused this issue for a lot of people. My intention was that the 'Changed view' message would be enough to highlight that 'Save' is necessary, but your experience here has also been echoed by others. I haven't come up with good ideas for how to improve this yet, but this is a place where ideas are welcome.

I also wondered if it would be possible, when you click on something to edit the details, if the browser window could be moved down to the edit box, because when I first used the UI, I didn't realise it was there (and I was using a small-ish window, so I didn't really notice it)...

Yes, this is a problem. Some browsers are a bit finicky about moving the view. At the time, I actually didn't know how to do this in javascript. I tried to ask for help on that one and none of the js experts had anything useful for me. I think I've since figured out how to do it, though, so I may give it another try. I'm not sure how much I like it, though. Changing the viewport can be very disruptive, and because of different window sizes, what works for one user will be completely unusable for another.

Thanks!

Concrete example of the

dreftymac@drupal.org's picture

Concrete example of the "sort" settings without the need for a separate "add new" and "rearrange" floating widget all by itself.

Usability and learning curve

Alice Heaton's picture

When I first used Views 2, my first feeling was "ugh, this is not like what I'm used to with the old views !". At first I didn't the see the point ; the old views interface worked well.

But now that I've learned to use it, nothing would make me go back. Yes, there was a learning curve. So what ? It doesn't mean it's not usable. There is a difference between low learning curve and usability ; assuming that to have "usability" you must have little or no learning curve is treating the user like an idiot.

I don't think Drupal is a CMS for users who do not want to learn it - I think there is quite a learning curve to getting what you want out of Drupal. Not because it's too complex, but because it doesn't impose any choices on you. If you don't want to make the effort to learn, then you're better off opening a Blog account at Wordpress or somewhere similar.

Different "world

dreftymac@drupal.org's picture

Different "world views"

:Yes, welcome to the wonderful world of query building. If you can make query building self explanatory,
:I'll happily give you the keys and go retire. I personally don't think it's possible. There's a lot of data,
:and new users come to Views without any clear idea of what it does or what they want to do with it.

@merlinofchaos: I think this nicely summarizes how you and I probably see things very differently. Having both
built and documented (web-based) query-building interfaces (admittedly not for Drupal, but relational databases
aren't exactly a new invention, so how different can it be) this is definitely a dramatic difference in perspective and
opinion.

(e.g.,
I personally don't think it's possible
users ... [don't have] ... any clear idea of what it does or what they want
)

From a standpoint of principle, I think these are simply very unproductive premises, but I can totally see how
you derive your findings if these are your basic underlying assumptions. I can also totally see why you would
be satisfied with the status quo, and offended by honest disagreement.

To the extent I agree that users
may not be completely aware of what's possible (with any system), there are two issues, one is familiarity
with Views (the interface and terminology that has evolved in Drupal) and the other is familiarity with
Data Manipulation (fundamental issues of how to arrange data in a tabular format in general).

It does not necessarily follow that unfamiliarity with the former == lack of understanding in the latter.

This point about user mastery was nicely addressed in the video link I shared from the beginning btw, as
well as several decades of best-practice and evolution in user interaction design.

Nevertheless, it's good to know your basic underlying assumptions, it helps to give context to the end result.

Simple views?

robertDouglass's picture

dreftymac@drupal.org - what do you think of the simple views interface? http://drupal.org/project/simpleviews

I think it's very good ...

dreftymac@drupal.org's picture

I think it's very good ... it will very likely help introduce people to the power of Views who might otherwise not be inclined to investigate further.

I know

Pasqualle's picture

you are just jealous. views UI is the best, please do not waste merlin's time with useless complaints. he needs to work on the new panels UI :)

I agree the UI is excellent.

Flying Drupalist's picture

I agree the UI is excellent. It didn't take me long at all to figure out what to do with it. The only thing that I missed was the whole display thing. I would say, perhaps add seperate links to add feed/block/page, rather than a drop down menu for 3 options.

UI is good

dwees's picture

The UI is good, and like most programs, it takes a while to get used to it, but once you realize 1 or 2 important things, it makes creation of Views way, way easier.

What will be helpful is when there are a few more sites and tutorials out there that explain the basics of the UI and what the various vocabulary words mean. The advanced help stuff is fine, and I'm glad the link to the advanced help is enabled by default.

After creating 4-5 views, I feel totally comfortable with the new UI and will never go back to the old system where I had to clone the same view 5 times in order to present it slightly differently in different locations.

Dave

It's a learning curve...

bb_matt's picture

Anyone who makes a statement "The user does not even want to think about Drupal." should probably not use Views.
Views pretty much demands you get to grips with the concepts of Drupal as a platform.

It's almost as if the thread starter is advocating that you "shouldn't" have to learn key concepts and that they should just come naturally.

I'm sorry, but whilst good UI is an important starting point, the concepts behind the UI are more important still and become intuitive once you start to understand those concepts.

I'm really struggling to understand exactly where dreftymac is coming from here. Comparing his example UI to the actual UI leaves me scratching my head, as it's clear the work done on Views 2 is simply streets ahead of this simple diagram in terms of UI.

Having used Views 1 extensively - being thrown into the deep end on 3 different commercial Drupal builds - from what I've seen and meddled with in Views 2, it's fantastic - a big improvement over views 1.

As someone who is as comfortable using a console as a GUI, I do get rather tired of the weight of emphasis some people put on a UI.
There's no excuse for being too lazy to learn key concepts. With something as complex as Views, you have to understand the key concepts before getting stuck in. The UI of any complex application is designed around core functionality, not the other way around.

I'm extremely comfortable using Views 1, having made it do some ridiculously complex things (a multi-category bookshop, with several layers of "special offer" functionality, for instance) - I think I've used pretty much every option of Views 1 to the point where I was beginning to require additional functionality. I'm looking forward to using Views 2 on an upcoming project is a major upgrade of a site that has 3 years of data on it, from Drupal 4 to 6 (yep, go figure)

In short, I strongly disagree with everything dreftymac has outlined.
I've used views extensively - creating this site, my second Drupal build, was exceptionally painful:

http://www.swedenborg.org.uk/

I must disagree

alicia's picture

Drupal has many very powerful capabilities, but truth be told, to configure custom sites using it takes a geek with a lot of time. Especially if you want to build a site that is visually appealing and organizationally accessible to the user audience. And, for reasons known primarily to the geeks building drupal, the drupal developer community has decided to have 3 major versions being built concurrently (5,x, 6.x & 7.x), all with different capabilities, API & hook methods, user interface changes, core module changes and various contributed module compatibilities/incompatibilities. Changing the complex Views module UI currently widely in use in 5.X as the current developers have done, with no apparent rhyme or reason makes it virtually impossible to upgrade from 5.x to 6.x and beyond unless the local drupal geek has the time (lots and lots of it) and patience (even more) to try to learn the new model, and make the old stuff fit in the new model.

If Darwin were watching, I'm not sure he would agree that this is the best evolutionary model for ongoing survival of the species.

Hello, Alicia, There are

bonobo's picture

Hello, Alicia,

There are some errors in your initial statement. For some background, read these posts:

On backwards compatibility: http://drupal.org/node/65922

For full documentation on the existing core apis: http://api.drupal.org/

On version info, and what it means: http://drupal.org/handbook/version-info

RE Views2 being rewritten without rhyme or reason: sorry, but this is just silly. Do a side by side comparison of the functionality, and what Views2 offers relative to Views 1 is pretty amazing. Additionally, the development was completely transparent, and is pretty clearly documented in this group. See http://groups.drupal.org/node/8429 and http://groups.drupal.org/node/6288 for two threads that were pretty central, but there are several other threads in this group that are relevant.

Alicia -- I don't know if you were feeling frustrated and venting, or if your post is an accurate representation of what you believe, but in either case some research or questions about the Drupal community in general and Views development in particular will be more helpful in getting any questions you have answered.

Also, from your profile page, you have linked to the Berkeley Public Library. Berkeley has a good user group; if you are located in the East Bay, you should also join the Berkeley group here on g.d.o.

Cheers,

Bill


FunnyMonkey
Click. Connect. Learn.
Using Drupal in Education

Drupal has many very

merlinofchaos's picture

Drupal has many very powerful capabilities, but truth be told, to configure custom sites using it takes a geek with a lot of time.

Yes, the Drupal community is aware that the complexity of the software is an issue. I'm pretty sure no one has ever suggested that this was easy.

Especially if you want to build a site that is visually appealing and organizationally accessible to the user audience.

Or fast. Or without effort.

And, for reasons known primarily to the geeks building drupal, the drupal developer community has decided to have 3 major versions being built concurrently (5,x, 6.x & 7.x),

Incorrect. 5.x is not being actively developed. 6.x is not being actively developed for core, but is frozen; updates are bug fixes and security updates only. 7.x is being actively developed but it will be quite some time before you will even see a 7.x.

all with different capabilities, API & hook methods, user interface changes, core module changes and various contributed module compatibilities/incompatibilities.

Yes, APIs change as the software grows. This is part of the way Drupal grows.

Drupal core is not responsible for contributed modules. That's why they are called that.

Changing the complex Views module UI currently widely in use in 5.X as the current developers have done, with no apparent rhyme or reason

What? no rhyme or reason? Ok, this is where I stop responding to you. If you think there is really no rhyme or reason to what I did, you either 1) are not paying attention, or 2) came here to be cause trouble. Either way, you are not welcome here. Please uninstall my software immediately. I spend a year of my time writing this stuff so you can suggest I am either malicious or capricious? Screw that.

After clicking through, I

merlinofchaos's picture

After clicking through, I see your account is not even two hours old. You created an account just to spread false accusations. You are reprehensible.

If Darwin were watching, I'm

merlinofchaos's picture

If Darwin were watching, I'm not sure he would agree that this is the best evolutionary model for ongoing survival of the species.

I lied. I have one more response. Darwin believed that the ability of the species to adapt to changing conditions was the most important aspect to evolution. Trying to tell the species that it should remain unchanged for your convenience sounds, to me, to be the opposite of what Darwin would suggest.

Instead, Darwin would think that the species that does the best job of changing to meet new conditions would survive. The species that would not change would be left behind. In nature, the only thing I can think of that has survived, unchanging, is the alligator, and it's only atop the food chain in the swamps.

If you want to stay in the swamps, I recommend you stick with Drupal 5.x and Views 1.

Complex site implies know-how.

Manuel Garcia's picture

If you want the power to create complexity, you can't expect the program to think for you. Views 2 gives you the freedom you need to create the complexity, without coding (which in itself is amazing), but it cannot know what kind of complexity you will create, nor should it do that - for if it did, it would be limiting your options, and that defeats the original purpose. Therefor, the complexity will need to be in the UI itself, and yes, you will need to know what you are doing, and use your head.

If however, you aren't creating complexity, then there is the Simple Views module, which will let you do 80% of what people use views for, without knowing how it all works, because it assumes what you will build, and thinks for you, limiting your freedom to create, but allowing you to stay comfortably not-thinking. I think it is a great addition to drupal by the way.

So now that I have cleared that out (I hope), there's always room for improvement on well, how intuitive the UI is. I think it is quite a challenge to fit all this complexity into a non-scrollling interface, and I sincerely admire the work that has gone there. I could suggest more intuitive icons perhaps, but then again, real-state is very limited. The main problem is that people don't like reading though, because that is all you need to get the hang of this UI, click around a bit, test it out, and I bet within an hour you will get the hang of it - if you know what views is of course. Also, reading the advanced help will get you going much faster, but then again, you do have to read and yes, it takes brain function to do that.

The only thing I would change is making the "Theming information" link stand up from the rest of the links, because I find people asking about how to theme views2 in IRC, when it is so much easier in this version than before. But hey, once they clicked it once, they never forget it... ;)

Views Developers

Group organizers

Group notifications

This group offers an RSS feed. Or subscribe to these personalized, sitewide feeds: