Top Web Design Mistakes of 1999

• Any time you don't have to set things up yourself is a win.

• The built-in features, including markdown, flairs, moderation tools, RSS, and comments, are great. The more so with RES (the reddit enhancement suite browser extension(s)).

• I'd really prefer a better management of images. I use imgur for image hosting, and with RES these _tend_ to appear with content, but users not using RES get a sort of bastardized version. Since I tend to use images for conveying more information (charts, graphs, and similar data visualization) this matters.

• Post archival. I'm reaching the point where some of my earlier posts are going to get archived ("frozen") and will no longer be editable or commentable (6 months after creation). Since I'm building a reference trove and want to be able to revise content as I develop ideas (my Questions post in particular: http://redd.it/1v052g), this is a bit of a drag. I've got a secondary blog that I may rev up as this becomes an issue. I can still access the source for those to port them elsewhere, but it's still a bit of a pain.

• Post length. Turns out that for text-only subs it's 40k chars, not 10k, which was kind of nice to realize. That's around 26 pages rather than 6 per post (typewritten, 250 words/page), which should be enough even for me.

• CSS. I've clearly restyled the site a bit, and I really like being able to do that. About 98% satisfied with what I've accomplished, and a few recent fixes (such as being able to create hierarchical header and list numbering in the Wiki) are really nice.

• Post flair. It's not as complete as tags would be, though the discipline is somewhat useful. I wish it were easier to maintain, though all told it's OK.

• Draft mode: A negative. I really wish I could start to compose something and save it without publishing it. Blogging platforms offer this.

• RSS. I'm absolutely and unapologetically in love: http://redd.it/1sxfar

• Those short links aren't bad either.

• Engagement from elsewhere. After a few months I am routinely getting far more (and better) engagement than I was on G+ ("the Other Place"). Links to other subreddits (particularly popular ones) have grown engagement markedly.

• Better moderation and engagement stats. It's nice to see my traffic stats, though it would be even better to have breakouts for 1) most visited posts and 2) specific user engagement stats. I use RES's user tagging to identify useful and not-so-much people, color-tagging as well. But seeing who's really engaged and contributed would be nice.

• Ways to deal with deleted posts. I've had a couple of instances of people posting good information only to have the post later disappear. I'm aware that there are some sites which archive reddit content, but I've learned to reincorporate that content where and when possible.

• RES again: the RES post editor is fucking amazing. Side-by-side view of markdown and rendered text, in particular, full page width and height. I'll <F-11> the tab to full-screen and write without any distractions. Compare to the clusterfuck that's G+ and weep.

• File storage. I've got a few things I want to be able to post which aren't images, and sorting out how to do that is a bit of a stumper. I'm thinking

• Comments search. As noted in the FAQ, it'd be really nice to have that. Oh well. Also a wiki search -- the wiki in general is handy but not particularly full-featured. I'm curious as to why reddit went NIH on that instead of incorporating, say, MediaWiki. I may branch out my Wikiing efforts to a different platform eventually.

• Search syntax: What reddit lacks in completeness it gains in features. The ability to look for content by user, subreddit, and other elements is a huge win, and I use search a lot. The ability to key my own subreddit's search URL into my browser's reddit search keyword means I get my own styling applied to reddit site searches. Yay me!

• Community support. There's a ton of help to be had on moderation, reddit features, CSS styling, and the like. That's been great.

• Expertise. I've got a former LLNL researcher and a nuclear powerplant operator among my readers, which for my areas of interest is great.

• User voting. Sadly, abuse of up/down arrows seems to be growing on reddit. While I'm a believer in some user input on content quality rating, I'm increasingly convinced it needn't simply be open to anyone. Though it also lets me know when I'm touching on controversial topics.

• A projects board. Along with drafts, I've got a large number of writing ideas I'd like to tackle, and sorting out how to keep track of those (preferably integrated with the site) would be cool.

• Data liberation. I cap on Google a lot, but the ability to extract all of my content from the site (and having created an archive extraction tool: http://redd.it/21t7im) is actually pretty cool cheese. Reddit doesn't have a similar feature that I'm aware of, though I can iterate over posts.

I'm not sure that another tool (possibly Drupal, which everyone seems to rave about) might not address more of my needs, but on balance I'm still pretty happy with reddit.

My primary challenges now are building a relevant audience, actually creating content (I've got a half-dozen to a dozen half-written postings floating around), and the like.

In the past, revolutions in browser models didn't happen through evolution of existing browsers so much as their wholesale replacement.

The Web itself supplanted a whole set of proprietary document formats, both static and hypertext, of which Tim O'Reilly has written and spoken about this in talking about the origins of O'Reilly & Associates. His Open Sources essay covers this tangentially, though I think there are better examples I'm not finding at present: http://oreilly.com/catalog/opensources/book/tim.html

In this regard, it's interesting to point out that the software industry's first attempts to improve on the web interface for active content--technologies like browser-side Java applets and Microsoft ActiveX controls--failed because they were aimed at professional programmers and could not easily be copied and implemented by the amateurs who were building the Web. Vendors viewed the Web in software terms, and didn't understand that the Web was changing not only what applications were being built but what tools their builders needed.

The upshot: HTML wasn't pay-to-play, you weren't locked within a proprietary "ecosystem". It also wasn't subject to the format limitations of postscript and PDF. It's even been backwards-bolted onto existing information infrastructures such as manpages and GNU Info documents via tools such as Debian's dwww:

https://packages.debian.org/wheezy/dwww

And so proprietary doc formats gave way to HTML. Lynx begat Mosaic, which added graphics, then Netscape, which extended graphics and a number of other capabilities, until it fell over under its own bloated mass of bugs and technical debt and birthed Mozilla.

Tabbed browsing came by way of ... um, whatever it was that graced HN a week or so back, in 1997/98, though didn't reach high visibility until around 2000 by way of Opera, Skipstone, and Galeon, from which it was rapidly adopted by Mozilla, which itself grew emphasizing Xul (remember that) until it collapsed and was replaced by a user's one-off, Firefox. Google got into the game with Chrome which seems headed in the apps engine direction (and from which I suspect it will similarly collapse as a Web browser).

For the lightweight document reading and management interface, I suspect we'll see emergence from some of the existing eBook reader / bibliography / online simplification tools. I suspect that sites will start publishing in a simplified format (ePub, Markdown, LaTeX, DocBook, possibly even strict HTML5), which can be rendered to client preferences and in a device- and format-independent manner. Our problem isn't a want of semantic markups, it's running QC over the ones we've got and ensuring people are using them properly. Absent some sort of publishing authority (think along the lines of Debian's package management system) I don't know how this could be enforced. Perhaps adding smarts at the webserver level: "Hey, that document you just loaded, those tags you use, I do not think they mean what you think they mean." A validation step integrated as part of the publishing process.

Or maybe a nudge from search engines (though Google's own craptacular markup in its G+ pages makes me suspect this is wishful thinking). A well-formed, semantically-correct page gets a higher SERP ranking than one that's a clusterfuck. And hey, the SEOs will love me because this gives them yet another contracting/consulting service opportunity. Sigh.

Candidates to grow into this niche on the client side include Calibre, Zotero, Moon+Reader (and other ePub readers), and perhaps some of the generalized reader tools such as Evince and Okular (the GNOME and KDE readers). I'd like to explore the Internet Archive's BookReader as a _local_ client (it's presently oriented at Web content) -- ironically it's actually got vastly better presentation performance than most local readers I've seen. I expect dedicated PDF readers to go the way of the dodo. Pandoc is another tool to keep you eye on. I've played with that recently, with some script glue, to convert an ePub to LaTeX source and then PDF output (still somewhat preferred for local viewing, ePubs still annoy me on Linux desktop). Nothing makes me happy at present, but I'm seeing some glimmers of light.

The apps platform space: some sort of sandbox in which remote apps can play but only in a limited sense. Effectively creating a mobile device environment within the desktop (on mobile, the apps already have their space). Key elements of this will be standardizing on widget sets, state preservation (apps should be immune to having the rug pulled out from under them), and user-controlled privacy controls (lacking in present app and browser models). A data liberation standard would also be very useful, so long as I'm compiling my ideal wishlist. I've got less experience here than elsewhere, and I'm probably missing a lot.

Seeing apps which are functionally oriented rather than site oriented would be a huge plus. The "download our app" annoyance must die yesterday.

On the commercial space, we're already seeing proprietary dedicated ecommerce platforms -- that's what the mobile shopping apps from Apple (iTunes), Google (Play Store), and Amazon (Amazon Marketplace) are, effectively. Each is, of course, a silo. The next step is going to be for someone to come up with a generalized shopping interface which can access any of the back-ends. Google's doing some interesting work here with its Google Shopping Express, and the really interesting element of this would be, IMO, to integrate a few different areas of shopping:

• Information gathering. Essentially, reviews of both merchants and products. A cross between Yelp, Google Reviews, Wikipedia, and Amazon's reviews.

• Merchant info. Often I simply want to go to a physical location and get standard information (hours, available merchandise, service availability for, say, a barbarshop or restaurant or auto shop).

• Stock info. Is X product in stock? This is actually where I think the most interesting aspects may happen, as it could see a tie-in of merchant systems with the rest of the supply chain. As with what I've discussed recently on disintermediating of search, providing local merchants the access to global supply-chain and inventory management which is presently only available to big box stores could be a huge disruptor. It's not an easy problem to solve, but it could be immensely profitable. There's actually a name for this, "Federated Retail": http://instagov.com/wiki/Federated_retail

• Ordering. Requesting a specific item.

• Payment. Note that ordering and payment are separate. I may well want to order online but pay in person.

• Fulfillment / shipping / delivery / pickup. The act of actually getting the thing you want. Note that while delivery to the home may be convenient, it often isn't, and one really useful function a merchant can provide is for warehousing of orders. I actually do special-order products frequently from local merchants (bookstores, hardware stores, clothing stores), especially where seeing, handling, getting fitted, or browsing products is a significant part of the experience. So long as the pick-up location is convenient to my daily habits, walking a few blocks really isn't a big deal. And it's often far more convenient than arranging to be at home at a specific time. At least until such time as homes and residences are constructed with package delivery as a consideration.

• Authentication if necessary. Note that often this isn't a requirement, and, frankly, authenticated deep behavior tracking is a massive turn-off for me on virtually any online shopping experience, and is among the key reasons I avoid online shopping where possible. I'm one of those reverse show-roomers: I'll often research a product on Amazon, then buy it locally.

One of the things I find most interesting about online shopping isn't its size or growth, but the lack of these. For the past several years, online retail has been underperforming growth estimates: http://redd.it/243in1

In 2010, TechCrunch cited Forrester Research predicting 8% of all retail sales by 2014.

From the US Small Business Administration we have "An Analysis of Internet Sales Taxation and the Small Seller Exemption[5] " giving surprisingly low numbers in November, 2013: "According to the U.S. Census Bureau, online sales accounted for approximately 5.3 percent of total retail sales in the second quarter of 2013. While online sales still represent a small share of total sales, they are expected to grow significantly in the future."

http://techcrunch.com/2010/03/08/forrester-forecast-online-r...

http://www.sba.gov/advocacy/7540/758295

For media, as noted, there are a number of existing tools which provide a superior playback experience to browser-embedded players. My main complaint with, e.g., VLC, is that it's unstable (tends to crash), loses playlists, and if playing, say, YouTube content, seems to establish a link to the present instance of the content URL which changes, so that if you're playing with interruptions, you'll lose the source and have to re-queue the original YouTube link to your playlist, which is a bit of a PITA.

• This requires JS enabled. Which I disable with ScriptSafe, so for a typical site, it's: enable JS, reload site, hit "read now" bookmarklet, pray that it actually opens.

• The page doesn't always render. In too many cases, the site simply fails to provide content to Readability. And since Readability's hidden their "feedback" button, I've got to hunt for that (or email them, which is better, since Readability doesn't provide feedback on feedback, even to the extent of providing an ack of your submission). After a few rough words with the crew I seem to be on better terms with them (and I really do love the tool and concept).

I realize this isn't the sort of thing that can easily be accomplished client-side, though sometimes I think it's still worth trying. People do the stupidest shit with HTML markup.

What problem, exactly, is this trying to solve?

It seems to be addressed at the popularity of apps. I'm generally not an advocate of site-specific applications. A function-specific app, perhaps. Say, an OpenStreetmap mobile app.

But if you're publishing content what you should focus on is semantic presentation, with a sane default styling, but allowing overrides by the user.

I'd prefer some sort of content syndication scheme / broadband tax, though that's clearly a political, not strictly technical, solution.

Phil Hunt's Broadband Tax proposal: http://redd.it/1vknhc

My Universal Media Payment Syndication idea: http://redd.it/1uotb3

Agreed - the common pattern I'm referring to is to have a button that the user clicks on to open a modal window.

This is ridiculously frustrating - it obscures the rest of the page, preventing me from navigating around that page or to other parts of the site without manually opening a new tab.

Occasionally those modal windows let me drag them around slowly with my mouse, but that's really only marginally better. I don't want every website to try and invent its own windowing manager, which is what many try to do.

And more often than not, this giant horizontal element refuses to scroll down properly and makes the whole page sluggish like hell. You gotta love 2014.

I cannot think of any myself, but I'm happy to be proved wrong.

The biggest issue with the modern web (IMO) is JS (but mostly JS frameworks). They're just so miserably slow on mobile. Slow to download. Laggy. Supplanting native controls with ones with a noticeable input delay.

It's gone so far as designers interpreting "mobile first" as "your default layout should be the mobile optimized one, and media-queries used to reformat for the desktop".

Why? It makes no difference to a modern smartphone. But it breaks IE8. For no reason. I'd like IE8 to go away as much as the next developer, but I also try not to deliberately break things if I don't have an actual justification. Especially when IE8 is a contracted requirement. It's just plum crazy.

Functional JS can be a great tool. I'm not suggesting it be completely banished. But JS abuse seems to be at an all time high. And to think the "conservative" developers were worried about abuse of AJAX back in the day...