Link style sheets v. inline styling

Hi,

Here's what i feel.. it might be a bit sarcastic..

These big company get big names because they do stuff that normal designers and programmers like us wont do.

They will cram up all their code in one single page and make it almost impossible to decipher what they really plan to do.
We designers will try to keep it simplified and give one whole package, and make it so simple that even your client can change things.
Whereas they must be doing to so that the client comes back to them to get changes done , so that they can charge on an hourly basis for the work done...

And i think many of the sites you might have seen that use tables and different inline styles might be Joomla! or some other CMS based, as they have 3rd party components that dont care abt proper coding.

mihir

I believe the larger companies get to a point where they need changes made, they've had 300 designers/developers work on their site, no one knows what the other did so they have to hack the code up to get it to do what they want to do, ala inline styles.

They also have enough money so that they dont care if their site is bloated or takes forever to load, because people will always come back.

Exactly my point???

mihir

You both sound right. I've kinda figured it had to do with job security. It's still strange. I mean, with all the money big companies spend on talent, standards, etc., you might think they would stress more efficient and portable approaches.

A friend feels another factor might simply be that prototyping is often done inline, and deadlines just prevent more elegant refinements later on.

I guess I am relieved. But not completely yet. Even YUI, ExtJs and other pretty sophisticated web 2.0 libraries will use a LOT of html source and inline styling, even when they also have a lot of CSS.

Finally, another friend thinks they may be trying to just set up their source so it can work in primitive browsers, phones, etc.. I have my doubts about that, because phones need almost an entirely differnt structure anymore, much of the time.

Hi,

I wont agree with ur first friend. Why... doing an inline style is far too much difficult than doing it in css. it doesnt simplify the matter, it makes it much much worse.
Secondly, many designers and developers sometimes seldomly care abt stuff they release , cause they have built it for one of their clients and their job is done.
Again ur second frnd is wrong cause almost all the times you would define a seperate sheet for mobiles as you will want to avoid using images and graphics, and only stick to text and borders. check ALA for more info on this.

mihir

I'm always confused as to why people are confused :?

Standards, Semantic web, CSS-P these are all matters that are still regarded with a little bit of disdain by the web development world, perhaps contary to popular belief? scripters/programmers (those people along with graphic designers) who drive web development very really care much for markup or CSS as long as their script outputs and the graphic designer can get his pictures up there somehow all is well and good.

Good JS libraries will respect and allow good coding I'm working a lot with jQuery and I don't accept or allow any bad code to slip through as I have responsibility for clean valid code and it isn't a problem with a decent library, it's down to the coder!

It is most certainly not being done to aid older browsers or 'phones' though

Rendering for mobile devices is becoming rapidly more sophisticated and requires and accepts as much thought and care as rendering for larger browsers, they also don't necessarily require a separate structure.

Ok, selated to this, is that I've seen even commercial components such as menus, nav trees, and other widgets where every icon will be an IMG tag with swapping handled by ponderous JS handlers in the source. All mouseover JS functions will be embedded in the source, as well, rather than simply adding all such event handlers to the DOM tree during onload (or jquery.ready), from a linked js script. Now sure, that requires more programming than some designers want to do, but surely there's enough examples out there for it to be somewhat routine at this point.

And what effect is Flex going to have on all this? I see more and more sites, large and small, going to Flash/Flex/Air solutions, and more and more clever widgets being designed for data controls. And with SilverLight coming up behind it, is the HTML/XML/CSS paradigm going to be sink into history? I have never done any Flash, so I don't really know just how versatile it is. but since action script is mostly JS anyway, it seems it could not be too limited. I know many of the criticisms have been addressed, and worked around, but it just seems that any proprietary solution that costs that much is going to have trouble becoming ubiquitous. I assume it's CSS under the hood anyway, for styling, but at some point, if the components are architected well enough, an every easier styling logic could emerge.

Ok, I am talking completely through my hat here. Anyone have some insight on where Flash/Silverlight will fit, say 5 years out? Will they be just one more component to be used where required, or will it take over the whole shootin match and be the environment for constructing an entire web-app?

And you thought I'd be talking about point to point over ethernet.

First, many/most of the large interweb companies had web presences prior to general css support, ca. 2001/02. There is still a large inventory of legacy table based pages that act as prototypes for current pages. CSS gets worked in as inlined style rules, analogous to the html attributes they were using.

Second, so many sites are/have been developed in Dreamweaver. DW is responsible for an unbelievable volume of silly browser sniffer, image preloader and image switch javascripting. None of it can be considered best practice, eg., onLoad/-MouseOver/-MouseOut/-Click inline event handlers [improper capitalization theirs]. Further, older versions inlined the style attribute—again, as a substitute for the font tags and attributes, etc., and newer versions create non semantic id and class tokens, moving the rulesets to the style element.

Third, developers who work in that environment have simply let the technology pass them by, and since what they do "works" for them, they see little value in bumping into the learning curve. These are what have been called the new amateurs. The true professional is continually upgrading their skill set, and these new amateurs are stagnating.

cheers,

gary

Kudos, Gary,

Much of that rings true. I've always wondered about that. The casual observer thinks "Macromedia/Dreamweaver. Big player. Must be motivated to develop the state of the art." In fact, most of the time, such leaders just sit on their leads and make what they can. I haven't used it since DW 2004 MX. I suspect not much has changed.

I am really trying to encourage the Aptana people to take on the DW market, but with a snippets orientation that can deploy good code in a versatile, flexible, open source IDE. It's so powerful now, that I think another year or so, and a Wizzy module will come forth and DW will be in real trouble.

If you've never seen Aptana, it's a great fork of Eclipse, with a terrific community, and an amazingly powerful and stable multi-language editor. Check it out here:
aptana.org

Sadly Flex is likely to become popular due in part to it's extreme ease of use I've had to look into it recently and it's very seductive as it requires little knowledge to turn out interfaces that have lots of bells and whistles and uses a kiddie style style approach to coding with lot's of selecting features/functions from a click and paste list, although it does require a modicum of source code editing but that's straightforward enough.

I'm afraid that a realistic view of the way things are going is firmly towards a modular code building structure that requires only a very basic understanding of what we hold dear.

Along with what I said earlier about many developers simply not prepared to lend any credence to Standards and all it stands for we do not really have a happy picture ahead of us, the large proprietary apps and UI/JS libraries simply have too firm a hold on things and seem way too seductive for people to resist along with the fact that they allow people to become developers :? without really having to learn too much about the craft.

Not too sure that Aptana isn't part of this same sort of world really it may be an IDE but I'm never too sure that IDE doesn't just mean "you need to understand less , as we'll do the tricky stuff for you"

That reflects my gut instinct too. Of course, the question is, could the flex tools mature to the point where all or most of what we do now is simply done more easily? I mean, after all, the point is the application and ease of use and appearance of the UI. If another technology gets there with a few trade offs along the way, that to me, is acceptable. I worry more about the start up cost and market hegemony of another proprietary technology. What's the start up cost for flash/flex? $500+?

I don't mind progress, if it really is progress. It seems the Flex snowball is picking up steam very fast.

What I can't really get my arms around yet, is how fast is this happening? Each day, I am seeing more and more adobe gadgets handling everything from video viewers, to data controls. Always with some pretty sexy visuals and behaviors. If it's really as easy as making a flash movie, that's nothing to sneeze at. I guess it depends on how versatile it will be, and whether third party components can easily be used.

As for Aptana, it's by no means "that sort" of IDE. It's hardly for the novice on any level. It's an enterprise level fork of the famous open source project called "Eclipse," widely used and supported by IBM, Sun, Oracle, and other big players. It's a tool suite that programmers can use to program in all or most of popular languages and libraries with a common editing platform. It has built in FTP/SFTP, Subversion, Mercurial, internal versioning and a ton of other tools that previously required separate apps that rarely worked well together.

Below is the elevator description, which doesn't come close to explaining it. Checkout the demo, or just install it. In Linux, just unpack and run from any folder (or there's a Windows installer). If you've battled a lot of editors that always came up short, I think you'll be impressed. It still lacks quite a few button palettes and other dreamweaver style IDE features, but it's designed to accommodate them when they get around to it (or someone wants to write a plug-in). In the meantime, it's already about the best open source, free editing suite anywhere. It's probably also going to become a preferred flash IDE, as it's really the first one to bring everything together. The idea is to have one editor, no matter what languages or tools you're working with.

The Aptana IDE is a free, open-source, cross-platform, JavaScript-focused editor and development environment for building Ajax applications. It features code assist on JavaScript, HTML, and CSS languages, FTP/SFTP support and a JavaScript debugger to t

http://www.aptana.org/

I don't see anything there that can't be done in Emacs[1] plus the usual development stuff, except for the gui. I'd feel a lot better about the product if their home page exhibited best practice for javascript, and some valid xhtml.

As to wysiwyg/IDE editing, wysiwyg is a waste of bytes. There is no best, only some not as bad—but I don't recall which that could be. The IDEs (Builder types) usually derive their power from drag and drop widgets. My experience has been that they're great for rapid prototyping, but the real work still needs to be done and could have been done faster and more efficiently from scratch. The widgets are overly complex, as are their APIs, because they must work in the general case.

For html and css, which holds most of the interest here, wysiwyg editors are worthless, with no redeeming social value. I think that was one Supreme Court Justice's definition of pornography.

An Integrated Development Environment may be helpful to anyone using an archaic OS[2]. On Linux, it is trivial to glue together the tools you need ad hoc. Further, any programmer will have his own library of favorite functions, or a library such as jquery, PEAR or CPAN at hand to copy/paste/adapt or use as is. It is my sense that these libraries far exceed anything similar in a single app, no matter how many plugins are in place to bog things down.

I don't mean to be all that negative about IDEs. I tried both MSFT's and Borland's C++ IDEs, and settled on Borland's C++ Builder as the better. I also used Turbo Pascal which was the first IDE. C++Builder, Delphi and JBuilder were on my machines until only recently. They were removed because for one, I no longer wrote anything requiring a gui (except for Java which I have stopped using altogether), and two, I had found Emacs, which lets me do everything in any language and keep my hands on the keyboard for better efficiency.

cheers,

gary

[1] HTMLTidy has native support for Emacs, and Emacs in nXML mode has on the fly xhtml validation. CVS has native support for Emacs, and Emacs has native support for CVS and two other versioning systems. Emacs supports syntax hi-liting and formatting for more languages than I knew existed. FTP, SFTP, SSH, SCP, directory editing, mail, news, text browsing and multiple shells are all run from within Emacs. Emacs can run over a network, in either direction.

[2] You may assume I refer to MSFT's offerings. Windows is kinda wysiwyg. What you see is easy to do, but what you don't see is damned near impossible to do.

Not sure I want to go here...

Gary, with all respect, you're a emacs guy. If most of the world wanted a command-based editor with a primitive gui wrapped around it, that would be fine. But I think you're not even in the 1 percentile, these days. I work with some of the most VI/emacs centric coders around, and they'd never argue that emacs is a step up. It's just what they're used to, and even they admit it would probably make sense to break the habit someday. Many of them love Aptana, and are slowly waving goodbye to their emacs and VIs. For millions of others, it's like asking them to do FTP in terminal mode, or add print tags to their Multimate documents. It's just not going to work for the next generation that couldn't care less about the past. Hence, the Aptana approach is simply an open source editor/ide framework for lots of convenience-oriented tools to get added to a rational environment. It's not for you, clearly, and that's perfectly fine. But for the vast marketplace of typical programmers, it has huge value. And it's sure a better approach than the proprietary Dreamweaver environment.

Here's an amusing, but informative post (and screen shots) about Emacs that sort of captures my sentiments about it, but in more irreverent prose than I might have used

Emacs Sucks

notepad2 > * tbh.

You're right, I'm in a declining minority. Even worse, I'm not some Unix fossil (though I'm of a certain age), but rather an ex Windows fanboy, going back to 3.1 ('89 or 90?). Other than starting on a TRS-80, it's been GUIs all the way. Oh wow! Mouse pointers all the way!

I mentioned already having used the very first IDE (Turbo Pascal). The move up to Borland's Delphi, C++Builder and JBuilder was fantastic. Then I met Linux. What a shock to the system (mine, not the computer's). All of a sudden I was working from the console and needed an editor. Talk about a vertical learning curve … oh, wait. I only needed to learn what I used, not the rest. OK, that's easy. Weaning myself from the mouse wasn't as simple. We all have bad habits that are hard to break. But, I did, and learned how much more efficient and productive it was to keep my hands on the keyboard.

So, I moved from a GUI/IDE world to a text interface because, 1) it's actually easier to use, and 2) it's more powerful.

Now, just a comment on your link:

Ach, that is so wrong on so many levels. No file selection dialog? It's "at the bottom where it says 'Find file' to select a file".

I don't know about you, but I usually know which file I'm going to open. Type in a few characters, and hit [tab] for auto completion. If you really need to browse for the name, hit [tab] twice and get a list. From there, type a few characters and hit [tab]. Bam! You're there. If you just must use the mouse, put the mouse cursor over the item you want, and middle click. How simple is that? In fact, considering auto-completion, it's easier to use than the usual GUI file browser.

This guy apparently expects the same user interface on every app. That he's not willing to learn the Emacs interface does not reflect badly on Emacs, but rather on himself. Emacs is considerably more powerful than any other editor, and has literally hundreds of key-bindings. For any particular mode, the user should at least learn the common bindings he uses. This guy didn't even learn how to use C-x C-f to open a file. Sh*t!

I will grant you the apparent advantage of a common interface among various apps toward ease of entry, but I will not grant any advantage in productivity or efficiency.

But, that's the direction people are moving—toward simplicity, even at the expense of power, efficiency and versatility. :shrug:

cheers,

gary

Um, wouldn't a steaming snowball just melt itself?

Maybe. I guess that's what I'm wondering about. But some snowballs also suckup everything under them and just get bigger. I keep an informal watch on how many times I see those flash widgets, maps, charts, sliders, faders, and other dohickies, and they just keep swelling in numbers. If the snowball is melting, the runoff is invisible to me. Perhaps it's a fad that will fade. I have a hunch it won't.

But I am not sure it's a bad thing yet.I don't have enough experience to know. It seems that CSS is still the styling engine of Flex, and assuming the HTML is structured well, I assume the worst case is if the flexibility is crippled by some kind of fixed component architecture that forces styling only one way. As much fun as it is to guess in a dark room, I suppose it's coming time for me to run it and turn on some lights.

Gary, I think it's best we drop the GUI or Not-Gui thread. It might start a land war in Asia

I used Turbo Pascal at college. Loved it. We then moved onto Visual Basic. Hated it.