Search posts, tags, users, and pages
I recently tried a few WYSIWYG editors. They are utterly unnecessary, they tend to teach you to think about web design in a completely wrong way, and you should steer clear of them. I get the appeal of such software, but it's not how web stuff works at all.
In my learning, I've been realizing more that, in some way, coding is about making hundreds of decisions about how to solve your problem. In this case, it's your content structures, layouts, and how you want your dynamic elements to work. Your tool needs to be expressive enough to describe those decisions, and I don't see how you could handle all of that without coding in some language. It is just like how you can never drag and drop your way out of writing an essay. You have to think, decide what to say, write it, edit it, and edit some more. In the same way, I don't think you can drag and drop your way out of writing proper HTML/CSS/JS/etc. You have to understand the problem at hand first, know the code, and then, you have to make all the right decisions to code them correctly.
It's interesting to me that, as far as I checked, there have been numerous attempts at the same kind of WYWIWYG idea. None of them seems to stick, and I don't see why this one is different from the rest of them. It may be a sign of people trying in vain to find a technological solution for the utterly wrong approach.
Got to agree, wysiwyg type editors are limiting, generally produce poor code and I steer clear. Wordpress i may consider for a simple site on a budget, Wix is just awful. But Wappler is not a simple drag and drop wysiwyg product, it is a fully data driven, extensive programming environment but instead of manually calling framework elements ( vue.js, boostrap etc) the process is done in a visual manner. I remember in the 80s when the first spreadsheet appeared on the market (visicalc). The programming world shouted hoax! It cant possibly be done. Now the world would be lost without Excel. Perhaps you should take a look, you may rethink your position (As I did)
I already have before I took a hiatus for several years. I was a subscriber of DMXZone and also purchased DataAssist extension from WebAssist at some point in the past. I also tried Pinegrow and Webflow quite extensively.
Tooling at this high level of abstraction suffers significantly from the rule of leaky abstractions. It works until it doesn’t, and then, when it doesn’t work, nobody understands why until you spend days going back and forth in a support forum or a support ticket system. You get stuck long enough and either settle for a workaround/compromise (often by writing custom codes, ironically) because of the limitation imposed by the software/platform/ecosystem or scrap everything and write code for real outside the inherent boundary imposed by those tools. If you’ve had this happen once or twice, you’ll be reluctant to try again.
To me, this type of tool is like adding another layer of learning and troubleshooting. With traditional coding, I will still struggle to learn and figure things out, but it is the material itself and whatever I learn through that process is applicable beyond a particular tool. With a GUI tool like Wappler, it is that process plus learning how underlining core tech translates to how to use the tool and also learning all the little quirks and workarounds for limitations and glitches. I can't help feeling like I'm doing double-work by merely trying to use this type of tool. I get the appeal of it, but I found them to be an illusion of simplicity.
Furthermore, I don't think your Excel example parallels the Wappler example. Spreadsheet isn't about how well Excel generates a .csv file. It's about the act of analyzing and manipulating numbers itself, so what goes on behind the scene has little effect on the result (data analysis) besides that a bloated or buggy codebase can deteriorate usability. The same is mostly true in the Wappler example, but the underlining code in web development has more implications as an end product itself especially for the frontend of things. The way you write HTML/CSS has a direct impact on things like accessibility, SEO, and a few other factors, and both users and developers at a much higher level of abstraction are involved in it beyond merely using a tool. I believe these are two different paradigms, and I think it is why GUI for some things became widely adopted while it never became a norm in some other stuff like front-end development.
Besides, it's not like Wappler is unique in trying to build a "visual web coding" tool. Macaw, for example, claimed to be "the beginning of the end of HTML handcoding" when they first came out, and even Jeffrey Zeldman published an hour-long podcast episode interviewing the developer and hyping it as some kind of web design revolution for graphic designers. The hype lasted less than two years before the developer abandoned the ship. Dreamweaver has always had server behavior and data binding feature that inspired DMXZone to build their enhanced equivalent back in the early 2000's, but most competent developers at the time cursed at Dreamweaver, NOT for making things a little more approachable, but rather for generating horrific codes advocating accepting such codes as if they were acceptable in the professional environment. Dreamweaver is still around, but I find it ironic that Adobe has made it more coder-centric in the past two CC editions. They even market Dreamweaver more in that way now.
I suspect that Wappler can fill some particular niche user case scenarios. But, even then, I also suspect many users will eventually find it necessary to get their hands dirty and code manually even just to get it to work right. That was definitely the case with WebAssist's DataAssist extension. It was a horrific usability mess, I was always in their support forum looking for answers, and the developer constantly had us copy and paste custom codes to work around a wide variety of issues we encountered as a user of that extension. Not surprisingly, I just don't see any signs of Wappler being different in their support forum. In fact, I peeked in their forum just to see and found it was pretty much the same when it comes to the kind of bug reports they are discussing there. In that case, why not just do it right from the beginning, work directly with the code and become more efficient at it without a crutch that is Wappler? At least that's what I concluded after trying a few of these tools myself. I would much rather come up with a way to make my coding more effective (not just fast and easy... speed is not a true measure of productivity) while working with it directly than putting a thick abstraction layer that only gives me an illusion of efficiency but sucks up all my time trying to make it work in reality.
So, thanks, but no thanks.
Jason Knight , do you have anything to add or correct me on this? I thought maybe you could explain the Excel example a lot better than I did, for example.
I remember in the 80s when the first spreadsheet appeared on the market (visicalc). The programming world shouted hoax! It cant possibly be done.
Funny, I don't remember anyone -- especially programmers -- shouting "hoax" in 1979... though I remember Visicalc fondly on my Trash-80 model 1 and later the model III. I think I even had it on the Coco. Linear processing an array based on input of interpreted fields? Not exactly programming gymnastics even on the hardware of the time. PARTICULARLY given that each field was little more than a line of interpreted code. It's not like interpreters were some radical breakthrough given most microcomputers had 8 to 16k of them in ROM out of the box that did far, FAR more than Visicalc ever did.
I do remember Byte, 80 Micro, and Creative Computing gushing over how wonderful it was pretty much from day one! So sorry, I kind of have to call bull on that claim of programmers screaming "hoax". I know, I was there.
So yeah, not a great example.
The only people I could even see making such a claim were the big iron guys, but even at the end of the '70's the microcomputer folks had given up on taking any of those dinosaurs seriously. We knew the days for DEC, WANG, WYSE, and IBM were numbered even if they themselves refused to see it, entirely because they refused to take anything microcomputer based seriously. They honestly didn't think the little 8 bit home computers were capable of anything which is why they ridiculed every major software package for them. See their attitude towards CP/M.
There's a reason that by the early 1990's Unix and posixisms were all but dead in the 'real world', only vague vestiges of them lingering in fringe OS or deeply buried under the hood until Linus came along and revitalized it with fresh blood, fresh ideas, and less corporate elitism.
Though by '83 Visicalc had already seen its day come and go. The doubling of the price mated to failing to keep up with the overnight boom of competitors on features -- and in particularly the introduction of Lotus 1-2-3 in '83 was the nail in its coffin, which is why VISI inc had what? A five year run before going bits-up face-down? Introduced '79, for all intents and purposes dead as a doornail by '84.
The spreadsheet example is also flawed as there were not an endless string of miserable failures prior to visicalc. When it comes to WYSIWYGS and other 'visual' development tools there are so many dead, defunct, and dying ones the web is more littered with their remains than a WWII eastern front battlefield. It is simply put not a history of success. It is a history of preying on the hope, dreams, and ignorance of beginners or those who are never exposed to the better way of doing things common to each generation.
Again why I'm often so ... volatile on this subject is that for all intents and purposes, it's just watching people make the same mistakes I've seen made a half dozen times over twenty years. Mistakes I learned better than ages ago.
It's the same dance, just a different tune... and unfortunately the entire album songs are being pulled from is a "best of 1979 Eurovision" collection. Sie ritten um die Wette mit dem Steppenwind, tausend Mann... Hu! Hah! Hu! Hah!
Amaya, Frontpage, Netscape Composer, Dreamweaver, KomPozer, Google Web Designer, Expression, Wix, Webflow, Maqetta -- each a bigger failure than the one before it on doing any practical serious work if you care in the slightest about accessibility, usability, graceful degradation, or just plain doing the job properly. WORSE they are by their very nature riddled with bad advice, bad practices, and an utterly and completely backwards approach to creating a website. This is why they have and continue to be derided by anyone who has even the slightest clue of what HTML is, what it is for, or how to use it properly.
Said backwards approach being the mere notion of thinking that you can start out with what the site looks like. The same backwards approach that allows artists under the delusion they are "designers" con people daily.
As I've parroted countless times the past decade and a half, HTML is for more than what it happens to look like for a perfectly sighted user sitting at a screen -- as such starting out with screen layout as your focus is nothing more than beginning at the wrong end of the process.
Hence why visual editors universally output non-semantic inaccessible broken HTML, that are bloated with presentational markup. I've never seen one that didn't.
Though if you were to show me something made with one of these tools that wasn't a train wreck of -- again those two words -- ignorance and incompetence, I'd change my tune in a heartbeat. Show me the money... I didn't see anything built with Wrappler that I would even consider allowing a client of mine to deploy, as for many of my clients (in banking, public utilities, health care, government agencies) it would be a very short period of time before it landed them with fines, civil lawsuits, and court dates thanks to contractual obligations involving the WCAG, country specific rules, and laws like the US ADA and UK EQA. In terms of usability, accessibility, semantics, or any other important factor in web development -- regardless of how pretty any of the results might look -- everything I could find in my brief walk-around kicking the tires was a complete failure on those fronts.
Which is why -- and this is just my opinion -- anyone thinking they can build a serious website with that is deluding themselves; either out of ignorance or just plain willful spite rooted in nothing more than "waah, I don't wanna type!"
... and to be frank, if the only reason people want this type of thing is "waah, I don't wanna type" then they probably in the wrong business, or in reality don't actually want a website. They want an artist to draw them a poster to hang on the wall.
... and as always I don't mean ignorance as an insult here -- it just means you don't know. Ignorant can be fixed. It would just be nice to catch people and educate them before the results from these types of systems whips around and bites them in the backside. Something that is usually not so much a matter of 'if', but 'when'.
See all the people who used all those other WYSIWYGS I mentioned the moment something goes wrong. Or the moment a user with accessibility needs wants to use their site. Or the moment those failings result in not ranking well in search. Or the moment they try to do something the WYSIWYG wasn't designed to do out-of-box. Or the moment when using ten to fifteen times the bandwidth starts making them butt heads with hosting costs. Or the moment when the site is failing to generate revenue despite 'looking great'. Etc, etc, etc... Legitimate problems that happen to the sites built with these types of tools all the time, for all the reasons the detractors state.
... and mind you, I would LOVE to see someone pull it off! To make such a tool that wasn't such a disaster. But I don't see how it is possible, I certainly don't see it happening with trash like bootstrap or other front-end "frameworks" in the equation since they too are equally if not more flawed in both concept and execution. I can't conceive of a method of doing so, and am not convinced such a concept is consistent with the reality of what the web is, is for, and who it is for.
Who websites are supposed to be for being everyone regardless of infirmity or limitations. The blind, the partially sighted, quadriplegics using mouth-sticks or puffers, people on bandwidth limited connections in places that might not see upgrades for decades since the old copper is still being paid off.
That might make me an unrealistic dreamer, a naive idealist... so be it. It'd rather aspire to a greatness that is inclusive for users and visitors, than to settle for a mediocrity that has the potential -- as I'm always saying -- to tell large swaths of potential visitors to websites to sod off.
Even if that inclusiveness for visitors comes at the cost of being exclusionary as to who can build websites for businesses. Toys like Wrappler being cute for blogs for grandma and not much else. If you accept it for what it is, it's fine. Just don't go trying to make it something it is incapable of being, a proper tool for business... and don't come crying when things go pear-shaped because of the incorrect choice of tool.
There's a joke my buddy Dan used to tell all the time on Sitepoint and DigitalPoint back when said sites were actually relevant; but I'm trying to be on my best behavior.
Ironically I was there at he same time, trash 80 bought in 1979 (still have it as a reminder) and read the same magazines, the joys of soldering upgrades to the solder tracks designed for the expansion interface as it was the cheapest way of upgrading, piggybacking a 1k static ram chip to give the 8th bit for display and having lower case letters. Funny how people can read the same article in the same magazine and see different words. Took me three months to disassemble the o/s Rom so I could integrate Z80 machine code extensions to extend it's capability. Now i prefer tools that just do the job, minimum of fuss, maximum productivity.
Now I prefer tools that just do the job, minimum of fuss, maximum productivity.
That's the part I don't understand. It wasn't like that at all when I tried various extensions from DMXZone and WebAssist, especially the one involved in connecting to a database. I gave them a try for about two years before finally giving up on them. It was a lot of fuss trying to get them to do things, and I found it more productive to learn to code and working directly with code later on.
And, of course, the responses to that were always similar to one of the following,
"I have not experienced any of those problems you are describing. It works great for me. Therefore, it must be you."
"You should send a support request to the developer and stop complaining."
"You've got to know how to use your tool properly."
"Have you uninstall and reinstall the extension? Have you reinstalled the OS? The bug is probably on Dreamweaver, so you should contact Adobe instead."
I particularly loved this one:
"You should help the developer make the software better. You should be more patient in troubleshooting."
It felt more like being a beta tester than being a paying user using a tool to do something useful. And it gets tiring when the vast majority of the first response in a support forum is often about how I was the odd one having "strange problems" that were supposedly so rare among other users that it must be just me. It's like being in a religious cult in some way.
And that's why I'm learning to code now. I feel far more liberated than ever compared to the time I was stuck relying on those types of tools. I can confidently say that I spend much less time troubleshooting my tools (it's just a code editor and a browser) and spend more time building something with the code. I can code in any environment (I can do stuff in my code editor, JSFiddle, CodePen, whatever... it just doesn't matter anymore) and apply the same coding skills. I just code stuff wherever I need to. To me, that is the kind of productivity and flexibility I would like to have more and more.