2021-06-22

[Technology] Issues I have with the web

Following my last post, I'll highlight some issues I have:

  • infinite scroll and chunking
  • see once and disappear
  • approximate, imprecise linking (e.g. Facebook comment highlights)
  • inefficient web design (facebook's emphasis on big images in posts)
  • transient content (considerations for privacy, vs. value of permanence)
  • opaque and obscure web structure
  • sharing information via screenshot
  • copyright holder abuse of fandom (Sailor Moon Says)
  • cyclic reinvention; e.g. messaging + voice/video
  • search-first over organization, and low discoverability
  • poor support for attribution (Pinterest is the devil)
  • dynamic articles
  • infinite privacy leaks and lacklustre permission control
  • challenges managing large volumes of information and making a profit
  • the stupid disorganization of ecommerce

infinite scroll and chunking

The example that annoyed me most recently was opening the comment section of a Facebook post in my mobile web browser.  There were several hundred comments.  I did indeed want to load The Whole Comment Section.  Instead, they just allowed me to load like 20 comments at a time.  This really wastes my time and gives me a very fragile context.  It doesn't take much to lose your context on mobile devices; e.g. a message comes in, so you switch to that messenger to reply, and you have like a 50/50 chance of your app/web browser's state being preserved (yay) so you can resume where you left off, or of having it reset (aggressive efforts to conserve memory, I guess), and losing your context (potentially "irretrievably"). 

This is part of what annoys me about infinite scroll.  If I'm interrupted, and my app/page has its state reset, I've totally lost the context in which I was, and my new context is likely different (with new content appearing and old viewed content disappearing).  

Some sites are better at this than others.  reddit breaks things up by pages and encodes references to anchor where you are in a collection of posts.  Twitter is awful; once I've passed a tweet, there's a good chance I'll never find it again.

see once and disappear

I noted Twitter above, but I use Facebook more and this is a huge pain point for me.  Once I've viewed something, if I lose context, it might be gone forever.  A common problem is a friend will share a link when I'm viewing on mobile; I open the link and read a decent article and by the time I'm done, I don't recall who shared it.  When I switch back to Facebook, it has now reloaded the app/browser tab!  I have no idea who shared it now!  My workaround is to pre-emptively like posts, so that at least I can go to the buried Activity History in settings (my interface is in German, so I'm not sure what it's called in English), and see what posts I last liked.  Ideally, there'd be some section that showed you which posts you had read chronologically, so you can always go back a bit.  

approximate, imprecise linking (e.g. Facebook comment highlights)

This baffles me slightly.  Facebook's feed will often highlight a post and in particular a comment made by a friend on the post.  In theory, I should be able to click on the comment they've surfaced, and it should take me to the comment section for that post opened to that comment.  Perhaps that now works on the app, though I no longer use the app.  It used to not, and it still does not on the mobile website, which is very frustrating.  

Rather than taking me to the indicated comment, it just takes me to the top of the comment section.  For a busy post with hundreds of comments, combined with Facebook's inability to actually load all the comments for me in a direct manner (thanks, chunking), means that it can be nearly impossible for me to actually interact with my friend's activity on the post.

inefficient web design (facebook's emphasis on big images in posts)

Presumably this somehow helps keep users engaged?  I find it disrespectful to my time and just an unpleasant experience.  I would appreciate a much more compact interface, so I can easily browse recent activity, choose what I want to interact with, and then drill down into that.  Instead, on my phone, each post in my feed now takes up almost my full screen.  If someone has shared a link, they always find some image to use a preview that I don't want.  On my laptop, I can often see 2 posts.  Also, I'm unsure, but it feels like Facebook might be more likely to promote lengthier posts or posts with images/links in my feed, than more concise posts, which is very unfortunate, as I often care more about actual updates/questions by friends.

My inability to use Facebook and some similar sites efficiently is one of the main reasons why I don't use them much anymore.  It just feels overwhelming and frustrating to use.

transient content (considerations for privacy, vs. value of permanence)

This is really killing me recently.  I think Instagram's app now defaults new posts to be stories.  I hate stories.  Stories disappear after 24 hours.  I hate this because it disrespects me and my time.  It bullies and pressures me into using the app at least twice a day if I'm interested in actually keeping up with my friends' activity.  

Some of the possible benefits to Instagram are having to store less information (but I think they might actually store them beyond 24 hours, visible to just the original poster?), possible increase in privacy (less easy to be haunted by your old activity if it's mostly ephemeral), possibly encourage people to interact with the app more (besides bullying people into using it at least once a day to not miss content, perhaps the limited time that your content is up encourages people to then add new content often, and encourage them to share fleeting thoughts and content that they never would consider giving a permanent home to?).  

A more controversial view on it relates to activism over transient social media (so including Snapchat and Facebook Stories and Instagram Stories).  For some reason, when issues get attention like Wet'suwet'en, BLM, Palestine, etc., people love to post important content ... in a format that will just disappear in 24 hours?  I sometimes wonder if a platform encouraging transience in group discourse like that is a strategy to limit politicization of the platform.  E.g. a number of friends of mine quit using Facebook because the dominant content grew too depressing for them, and they moved to Instagram/Snapchat/Tiktok.

A decade ago, I feel like it was common to say "nothing truly is gone once it's online".  But now I think it's the opposite.  Things rarely persist online.  Even the general web.  5 years ago, I feel like I used to be able to search certain topics like ICON computers used in Ontario in the 1990s, and I could find hundreds of references.  10 years ago I could google my old elementary school (since demolished) and still find photos of it online.  Now I actually can't!   Platforms like GeoCities have gone away.  Google Cache is just not available for a lot of sites anymore.  Efforts like the Wayback Machine are woefully incomplete - almost every time I turn to it, I am disappointed.  Sigh.

Somewhat disturbingly is the intentionality behind part of it.  In Europe, some people advance a "right to be forgotten" which I find disturbing as it gets used by some bad actors in society to cover up their misdeeds and escape accountability.  I am actually in favour of obliging content creators and hosts to correct inaccurate information (e.g. if I was falsely accused of a heinous crime and was cleared, I'd like to oblige hosts to correct that; but I wouldn't want to actually entirely erase the whole affair; it would be worthwhile that the original false accusations were still a matter of record). 

opaque and obscure web structure

This kills me, and it's a somewhat niche problem.  As a computer programmer, I have opportunities to improve my user experience by making my own modifications to the sites I use.  I can write custom style sheets, GreaseMonkey scripts, bookmarklets and extensions to adapt a website's presentation and functionality to better suit my needs.

However, for various reasons, websites make it increasingly hard to work with their websites in interesting ways.  Few websites have what I'd consider a stable "API" or even a human-friendly unstable one.  In particular, CSS classes are all unsemantic, obfuscated nonsense on the user's end, making it a challenge to properly select elements to do interesting manipulations to a page, like converting all dates to a sane Y-M-D format.  

As an example: Since I fell in love with my favourite T.V. show ever, She-Ra and the Princesses of Power (Netflix 2018), I've started collecting a lot of media related to it.  It's really important to me when collecting that, or memes, or just whatever, that I be able to accurately attribute authorship, the source (e.g. platform), the date-time it was created/found, etc.  Consequently, if I'm going to save a piece of fan art or fan fiction, I like to have a bookmarklet to pop up the key info (e.g. creator username, date-time posted, platform) so I can include it in the saved filename.  However, each website makes it a pain to identify the elements holding those.  

Authorship attribution in general is a failure of the modern web that makes me very sad, given the power we have to do better.  Pinterest and reddit are good examples of failures there.

sharing information via screenshot

This kills me as well.  The dominant way people seem to share textual information around me now is by screenshot.  I can see some appeal to that; it appears literally exactly as the person first encountered, possibly exactly as the author had hoped it would appear.  However, it also makes it nearly impossible to work with.  You can't easily translate it (though machine translation tools actually are working fairly well with image text; but even on Android with the amazing Google Lens, it's still extra steps and more friction and a new, unnecessary risk for error).  You can't easily cut/copy/paste.  You can't easily search it.   It wastes space and bandwidth (an image of text is often many more bytes than just the text itself).  

It's doubly bad because this intersects with transient sharing.  So many Instagram stories of interest: if I want to preserve what I'm seeing, I have to screenshot it!  I can't copy/paste it anymore.  I can't save a permalink to the story (since it will just disappear shortly).  I end up with dozens/hundreds/thousands of poorly-organized screenshots of interesting/important stuff in a folder on my phone.  It's hard to follow-up on, and it's infuriating.  

It feels as time goes on and the Internet "evolves", information is becoming less accessible and more opaque.  I guess there are some benefits to that, but that depresses me.

copyright holder abuse of fandom (Sailor Moon Says)

An on-going issue with the web is the abuse by copyright holders of fair usage by fans.  What kills me about this is when a copyright holder attacks an activity that directly benefits the copyright holder.

I'm quite enjoy the Sailor Moon anime, despite the variety of issues it carries.  I think it's very reasonable for the copyright holders to go after streaming sites that host entire episodes online, when the copyright holder has ensured there is a "reasonable" means to access the content legally.  But in this example, there has been a very popular YouTube channel called Sailor Moon Says that has hosted hundreds of amazing, amusing, entertaining short clips from the series.  This has been an amazing resource for fans to revel in the show that they love and commiserate over so many great moments.  However, last week YouTube removed the channel due to alleged copyright violations. 

Part of my complaint is how easily unfair complaints make it through a platform's reporting process to unfairly penalize fan activity (I wish YouTube had the capacity to give better review and had a greater interest in defending users), but the other part of my complaint is the readiness with which companies attack their customers and fans.  Like, please stop undermining your own communities.

cyclic reinvention; e.g. messaging + voice/video

This makes me feel dead inside.  It's another failed promise of technology and its evolution.  I'm going to pick on messaging + voice/video chat.  NetMeeting, MSN Messenger, Windows Live Messenger, Skype, Teams.  Google Talk, Gmail Chat, Google Hangouts, Google Duo, Google Allo, Google Chat + Meet.  iMessage, Facebook Messenger, WhatsApp, Instagram Direct/Chat, Telegram, Snapchat, Signal, etc.  I'm fine with multiple approaches competing against each other, especially from different vendors.  But what feels stupid is the repeated reinvention of the same tools, and re-living the growing pains each time.  

I'm sure there's value in it.  Older approaches had limitations and newer approaches hope to avoid those (while discovering new ones).  It's just exhausting, because each time a new approach starts and matures, users have to relive the the associated growing pains.  And we end up with a ecosystem of dozens of "solutions" that don't interoperate, forcing us to split ourselves and our attention across multiple platforms, complicating the management of our lives and communications.

I am so grateful for e-mail.  A standard format that different providers all (mostly) interoperate well with.  Shame on those who seek to replace it.  Especially with proprietary, closed solutions.

I really loved communications technology, back when the big question was XMPP/Jabber vs. SIP.  I had so much hope and optimism when most of the leading chat networks were switching to XMPP, including facebook Messenger, Google Talk, maybe AIM?  What killed that?  I remember a lot of fuss over adding voice/video on top.  Maybe that was the problem?  Jingle.  SIP.  Eventually people were trying to add their own proprietary extensions.  Next they just gave up and went with their own private protocols.  Was it just depressing business decisions, "why interoperate when we can try to lock people into our platform?!"

Perhaps the open solutions at the time were just ultimately inadequate to enable the rich experiences communications providers strived for.  The fact that friends keep pressing me to install creepy apps that overreach for access on my phone, though, will continue depressing me. 

I really wish I just had one app with an interface that suited me.

search-first over organization, and low discoverability

I'm going to pick on Windows and modern messengers for this one.  Back in my day, it felt easy for me to browse and discover my operating system to understand what it was capable of, in part because of an imposed organization.  I could go to the Start Menu, and programs were pre-sorted into categories!  It was amazing.  Now on operating systems or my phone, if I pop open the list of available apps, it's a disorganized grid, often sorted alphabetically or (for better or worse) recency (where I lose positional context as things keep shifting around as I use them).  Windows settings and configuration experience feels like a mess; there are often multiple views into the same settings, and the different views have different capabilities.  Control Panel/Settings feels like a war between the old paradigm (kept around to keep older users happy and for eternal backwards compatibility) and a desire to have a "fresh new approach" which centres search.  

My issue with search-first is that you need to already know something exists or might exist to know to search for it!  If you're not aware that something is possible or may exist, you can only find it via search by dumb luck.  When things are categorized, and potentially with a hierarchy, you now have a space in which you can browse and explore and discover what's possible.  I'm pretty happy with my settings on, say, Fedora Linux, and how they're organized in GNOME's control centre.  I'm a bit less happy with the settings on Android and Firefox - in part because of that earlier problem of cyclic reinvention.  Re-designing your settings and re-arranging where they're found is incredibly painful when you're forcing users to relearn how to use their devices every 2 years.  Firefox's settings in particular are also becoming very space inefficient, which is making them exhausting to navigate.  

The other area where I'm really hating search is in messaging apps.  When they have it.  Back in my day, a lot of chat histories/logs were well-organized by date.  In some messaging programs, I was able to easily navigate by year, month, day.  I could also do an actual, proper search against the text.  Facebook Messenger, despite having hundreds of millions of users, is abysmal at this.  My two ways of accessing old messages are: infinite scroll (which is horrible, as incoming messages will kick me back down to the bottom, and which suffers from that small chunking problem from before), plain incomplete search results, and no way of selecting by an actual date/range.  

I actually will use the Facebook Download tool to export my messaging history to my computer, and then just use grep to locate what I need.  This can take some more time, but is reliable and successful.  I am still stunned that I can search for keywords that are definitively present in a conversation's history, and Facebook just simply doesn't find them. 

poor support for attribution (Pinterest is the devil)

I am so depressed by this.  Earlier I cited my keen enthusiasm for the new She-Ra.  I'm a member of several online communities for it.  One of the most frustrating things is that people will post fan art they like to Facebook groups or their Instagrams, without actually attributing the artist.  Sometimes it also creates an implication that the poster is the actual creator, when they're not.  Thanks to Google Image Search, I can do a reverse search and often track down the original and identify the artist to give them credit in a comment.  A lot of users are reluctant to put in effort to identify and credit an artist, despite appreciating their work, which feels pretty disrespectful to me.  

In the header, I specifically pick on pinterest.  This is in part because when I do a reverse Google Image Search on an image, I invariably find 90% pinterest hits of people who have all saved the same image to their pinterest, and 99% of them have added no context at all.  This is much worse than on other platforms that have common attribution issues like Facebook and Instagram.  For She-Ra, reddit's SPOP community is actually fairly decent at crediting the artist (wow, given how horrible the rest of reddit is at attribution).  Tumblr and Twitter actually come out on top, in part because of how their platforms work, with reblogging and retweeting and quote retweeting actually linking directly to the original post (yay).  

Besides having pinterest burn to the ground, something I'd love to see is better, more durable embedding of authorship metadata.  I know some creators actually don't really want identifying information of theirs to be associated with their works.  I guess I wish all image/content creation tools more prominently emphasised addition of metadata to images/content (e.g. like how metadata is an important part of music files), and that image/content viewers more prominently displayed it (and not just hiding it behind some properties panel).  I also wish the image transformation tools were more biased towards to preserving metadata (e.g. when cropping/resizing an image and exporting to JPEG, I'd love it if the original authorship metadata carried over by default).  I also wish that when saving media from websites, web browsers would embed the originating URL into the image metadata (!), possibly along with a timestamp (!).  

My craziest desire connects to my problem above with the use of screenshots to share information.  I kind of wish that a screenshot process was intelligent enough to identify the elements on screen that had metadata, that the screenshot would aggregate that metadata and associate it to their relevant region of the screenshot (!).  I see risks to privacy with this, but as an example: if you took a screenshot of several overlapping windows open, one with a photo, one with a piece of fan art, one with a Word document, one with a PDF, and one with a song playing), I'd want the screenshot to have multiple metadata records for each of those items it captured.  I want to be able to look at a screenshot and hover the piece of partially fan art captured in it, and be able to tell who its author was, and the datetime the fan art was created (as recorded in the original image).  Similarly with the author and title of the Word document, etc.  There's a definite problem of people inadvertently capturing information they don't want to share, though, so this isn't that simple of a solution.  

I'd also really like it if some of the larger web companies actually supported content registries, where creators could register content they had produced as theirs, similar to how Content ID works in YouTube.  As an artist, I'd like to be able to upload copies of my work registered to me.  Then, a user could encounter some of my work (perhaps a screenshot of it, or a re-compressed version, that isn't bit-identical), and the registry would ideally work in such a way that it could identify close matches.  This somewhat bypasses the data-heavy approach I dream of with metadata, putting the onus on, say, the web browser or image viewer to maybe generate a representative digest of the media I'm curious about and match it in a registry.  

Regarding reddit in particular, there's a large incentive somehow for people to farm karma with others content.  People will knowingly repost content that was never there's, and sometimes wilfully conflate themselves with the original artist.  I'd really enjoy it if reddit posts actually had a feature to add a link back to a source and an author.  Also, to deny a poster the benefit of karma accrued on a post if it is not their content.  E.g. if they're reposting someone else's original post from 2 years ago, I would like upvotes on the new post to go to the original user's profile if still active.

dynamic articles

This isn't actually a great way to describe them.  However, I'll try to explain.  When you click on an article at a site, and reach the bottom, and suddenly you're scrolling to a new article, endlessly.  Or you're in an article, and a lot of sections unrelated to article proper dynamically interweave themselves into it, similar to advertisements but actually just content from the same site.  E.g. you go to read a story about a new video game and after two paragraphs down there's a video ... not for that video game, but from a feature story on some other game or even a film.  This happens with general news sites a lot, too. I guess it's an attempt to grab the reader to keep them pulled in.  They're reading one story, and you want to keep their attention so you keep highlighting other stories too.  

I hate this.  It's so disruptive and aggravating.  I avoid entire websites due to these practises.  I used to really love Kotaku.  Besides decent coverage on actual news, they also spent a lot of time adding value to their reporting by considering human and social impact atop the facts.  Now it is unreadable.  Pages load painfully.  The dynamic content (videos from other featured stories, partially loading comment sections, other article pages connecting when I scroll past the end) including the abusive volume of advertisements, creates jerky pages where content keeps shifting as I scroll.  Perhaps their greatest sin is splitting up lists like this

  • list item header 1
  • big advertisement
  • list item content 1
  • list item header 2
  • big advertisement
  • list item content 2
  • ...

The humane way of interleaving advertisements would be to be between list items, not breaking up an item between its header and content.  Presumably they feel like having ads humanely placed would just make it easy for a reader to subconsciously ignore the ads, and just look at the space in between.  But now they've annoyed me so much and made trying to read it so hard, that I just avoid their site and don't look at any of their ads at all.  The "ads breaking up content" is doubly frustrating when the content is actually images, so if you basically have an image caption (the header) separate from the image itself, by another, unrelated image (the ad), and the cognitive dissonance in trying to associate the header/caption with which image is supposed to be the content gives me a headache.

infinite privacy leaks and lacklustre permission control

My mind is blown by the state of privacy, security, and application distribution.  Privacy abuses by app developers are rampant.  Systems like Android provide APIs that help reveal the type of access that an app intends on having, which is kind of nice.  You can review it.  But if you don't like it, there's little you can do about it.  Why does this GPS-powered compass need access to my network connection and microphone?!  There is some limited ability to control these, but it's laughably limited.  At one point I remember having the option to revoke some permission only after I would install an app - so the app once installed would happily have several/dozens of seconds from installation to when I could "successfully" revoke permissions, to siphon my precious private data off to their servers.  Huh?

Really, the model I want is a) for an app to access any information or context outside of itself, to go through a controlled API, and b) to be able to - at the time of installation - choose whether I want to allow it real access to that service/information, a limited/virtual access, or no access.  E.g. an app to backup texts requests access to external storage and network connections.  I may want to say yes to the former and no the latter.  I'd rather have an app potentially crash from unexpectedly getting no/fake access to a network, then to have to implicitly trust it.

I'm sometimes stunned that desktop/laptop computer OSes went so long with such poor control over resource access.  It blows my mind how trivial it has always been for a simple "benign" utility to get installed, and instantly pop open a network connection to receive remote commands or exfiltrate sensitive data.  I rely on a network of trust on Fedora Linux, where I have to hope that the volunteer community of packagers and open source projects maintainers are sufficiently trustworthy, helped in part by the open source nature of their work.  There are still regular failures (e.g. an open source application, Node.js dependency, or browser extension gets "bought" by a bad actor who then adds malicious code :\), but it's still somewhat better than the closed source "hip new camera filter app" on the app store gaining 100,000 downloads with ridiculously unsavoury permission requirements.  

I'm grateful that Flatpak and Firejail are trying to do a better job of isolating programs from resources to control risk, but there's still a long way to go.

challenges managing large volumes of information and making a profit

I understand that a lot of what I hate about modern web design is driven by the challenges of managing large volumes of information that people mostly just want to consume briefly and move on.  But these choices still make using a lot of services nearly miserable and present a continuingly failed promise.

I also understand that service providers have real costs, and have real aspirations to grow and evolve, and to do so in our capitalist society, they need to generate a meaningful profit.  Which generally means advertising.  I'm not fully opposed to being advertised too, nor am I opposed to personalized advertising.  In fact, I am very grateful when I see advertisements that actually pertain to my interests, rather than generic ones that are sometimes downright offensive.  (The web should know well enough I'm a vegan by now, so please stop advertising meaty dishes at restaurants to me.)

I do wish advertising networks also just gave me more direct control.  I would really love to be able to actually register some of my interests with them, and some of my disinterests.  Also, advertising fails in a big way with tunnel vision.  I do a single search for a record player, and suddenly I see ads for that same record player all week everywhere I go.  What they're getting wrong: if I already bought that record player, I do not want to see more ads for it.  Advertising something I've already bought to me is wasted advertising opportunity.  If I have not already bought it, then there are two likely reasons: I didn't like the price, or I didn't like that specific product though I wanted something in its class.  I'd really appreciate it if advertisements would focus on trying to actually match me with a better deal (I'm looking at you, 2TB SSDs), or a better version of that product (it took me a long time of searching myself to find the air conditioner I wanted, with advertisers missing out by just showing me ones I'd already looked at and passed on already).  

the stupid disorganization of e-commerce

How does anyone ever manage to successfully buy anything from Amazon?  For such a massively successful company, I remain perplexed by how unusable it is almost every time I go to use it.  Some common issue:

  • I go to search for an item on amazon.com and find it but they do not ship to Canada
  • I search for it on amazon.ca but they do not have it (it's only on amazon.com)
  • I find it on amazon.ca, but it's at a Much Worse Price than amazon.com. 
  • My specific, explicit search terms pull irrelevant results (HOW?!)
    • e.g. I search for a "2TB SSD" and the results are muddled with irrelevant results for 2TB HDDs (where the product page doesn't reference SSD at all) or 1TB SSDs (where the product page doesn't reference 2TB at all).  I'm guessing some sellers are somehow stuffing references in somehow to appear inappropriately in more searches?  Or Amazon is trying to do some "clever" approximate matching, assuming I don't actually know what I'm looking for?  
  • I search for something specific with necessarily ambiguous words (e.g. "bike rack", hoping for a rack for bikes to carry saddle bags, and instead get racks for cars to carry a bike), and have no way to filter what cluster of "bike rack" meanings I mean.  
  • sometimes-available, sometimes-mutually-exclusive filters?  This one makes my head almost explode.  I may be on a site like BestBuy, and I want to look at computers, and they provide me with a variety of filters on the left-hand side, for processor speed, RAM amount, SSD capacity, etc.  I set one of them, and then some of the other filter options disappear (?!).  
  • poorly-structured and inconsistent product information: I'll pick on BestBuy again.  One of the main things they sell are computers.  There are many key specs of a computer that people are interested in when searching and comparing.  And they manage to encode them ambiguously or incorrectly into their product records.  E.g. some HDDs will have their speed in RPM listed, and others won't!  The level of detail on two monitors will differ, as well as the presentation/format of that data, making it hard to effectively compare multiple monitors' on contrast ratio, without having to search each of the product's pages on their manufacturer's websites.  (Don't get me started on the uselessness of Home Depot's website when it comes to shopping for shower heads...)
  • Oh, and sizing on clothing O_O  And the stupid disproportionate amount of packaging (e.g. too lightly packed fragile stuff; ridiculously overpacked tiny stuff), oh and and and...

Ultimately, I accept that there must be something wrong with me here, as many people seem to successfully do a lot of shopping on-line, and e-commerce companies make billions.  I just feel so depressed because it's easy to visualize what an actually-easy and frictionless experience should be, and see the gap between that and what we have. What we have is in some ways more flexible and agile; there's less explicit, specific support coded and maintained for each different type of product, yay?  I am a big fan of actually paying to do the work needed to get an optimal outcome, though.  I want utopia, damn it.

thank you

If you actually read this rant, wow.  Thanks.  Sorry for the typos.  Please leave a comment just so I know you exist.

I miss the predominance of blogs, once upon a time, outside the social media silos.  I'm old now, I guess.

[Technology] Complaints

I sometimes feel like most of my thoughts on technology in the past several years are too negative.  I mostly focus on deficiencies and frustrations.  I end up not wanting to share those, as they don't often feel constructive.  I think that's because it's easy to express them as a value judgement. 

I dislike criticising a technological choice while implying some sort of inferiority in the developers that made that choice.  People make bad choices, and so do I, but also many choices that appear "bad" are actually very reasonable given the context, and it's hard to know what that was.  So I don't like squawking about it.

That said, I think it might be worthwhile to sometimes express my issues with technology, as those are often the points I'm interested in trying to improve.  So I'll try to do some posts about that in the future.

2021-03-13

[Technology] "Add to Home Screen" broken by PWA's "start_url"

Problem

This has been driving me mad.  At some point, I was increasingly unable to add direct links to web pages on my home screen on my phone.  E.g. to specific Twitter accounts, or to Instagram's Chat page.  I would go through the usual steps of visiting the specific page, going to the menu and choosing "Zum Startbildschirm zufügen" (aka. "Add to Home Screen"), but the shortcut would only go to the main page for the site instead, twitter.com or instagram.com, making it somewhat useless.

Today I tried to investigate the problem:

https://android.stackexchange.com/questions/225002/chrome-and-firefox-cannot-add-specific-url-to-home-screen

The problem is that these sites are acting as Progressive Web Application (PWA).  In their Manifest file, they can specify a value "start_url", which then overrides the actual URL I'm trying when adding to my home screen.  

It's pretty infuriating and undermines my control over my device and experience, and impairs me from accomplishing my goals.

The site lists some workarounds.  One of them is to use 3rd party apps to create the short-cut.  One of my qualms with 3rd party apps on Linux is that so many request excess permissions beyond what they need for their purpose, and despite having some limited control over that, it's rarely enough to feel safe. :\

Solution I'm Using

javascript:{
  if (meta=document.querySelector('meta[name="mobile-web-app-capable"]')) {
    meta.parentNode.removeChild(meta);
  }
  if (mani=document.querySelector('link[rel="manifest"]')) {
    mani.parentNode.removeChild(mani);
  }
  void(0);
}

The user that suggested this, "user185038",  notes that Chrome seems to strip "javascript:" away if you just copy and paste that in, so they paste something slightly different (javascriptx:) for that protocol bit, and then just edit to be what's intended.

And it worked!

Dieses Blog durchsuchen

Labels

#Technology #GNOME gnome gxml fedora bugs linux vala google #General firefox security gsoc GUADEC android bug xml fedora 18 javascript libxml2 programming web blogger encryption fedora 17 gdom git emacs libgdata memory mozilla open source serialisation upgrade web development API Spain containers design evolution fedora 16 fedora 20 fedora 22 fedup file systems friends future glib gnome shell internet luks music performance phone photos php podman preupgrade tablet testing typescript yum #Microblog Network Manager adb apache art automation bash brno catastrophe css data loss debian debugging deja-dup disaster docker emusic errors ext4 facebook fedora 19 gee gir gitlab gitorious gmail gobject google talk google+ gtk html libxml mail microsoft mtp mysql namespaces nautilus nextcloud owncloud picasaweb pitivi ptp python raspberry pi resizing rpm school selinux signal sms speech dispatcher systemd technology texting time management uoguelph usability video web design youtube #Tech Air Canada C Electron Element Empathy Europe GError GNOME 3 GNOME Files Go Google Play Music Grimes IRC Mac OS X Mario Kart Memento Nintendo Nintendo Switch PEAP Selenium Splatoon UI VPN Xiki accessibility advertising ai albums anaconda anonymity apple ask asus eee top automake autonomous automobiles b43 backup battery berlin bit rot broadcom browsers browsing canada canadian english cars chrome clarity comments communication compiler complaints computer computers configuration console constructive criticism cron cropping customisation dataloss dconf debug symbols design patterns desktop summit development discoverability distribution diy dnf documentation drm duplicity e-mail efficiency email english environment estate experimenting ext3 fedora 11 festival file formats firejail flac flatpak forgottotagit freedom friendship fuse galaxy nexus galton gay rights gdb german germany gimp gio gjs gnome software gnome-control-center google assistant google calendar google chrome google hangouts google reader gqe graphviz growth gtest gtg gvfs gvfs metadata hard drive hard drives hardware help hp humour ide identity instagram installation instant messaging integration intel interactivity introspection jabber java java 13 jobs kernel keyboard language language servers languages law learning lenovo letsencrypt libreoffice librpm life livecd liveusb login lsp macbook maintainership mariadb mario matrix memory leaks messaging mounting mouse netflix new zealand node nodelist numix obama oci ogg oggenc oh the humanity open open standards openoffice optimisation org-mode organisation package management packagekit paint shedding parallelism pdo perl pipelight privacy productivity progress progressive web apps pumpkin pwa pyright quality recursion redhat refactoring repairs report rhythmbox rust sandboxes scheduling screenshots self-navigating car shell sleep smartphones software software engineering speed sql ssd synergy tabs test tests themes thesis tracker travel triumf turtles tv tweak twist typing university update usb user experience valadoc video editing volunteering vpnc waf warm wayland weather web apps website wifi wiki wireless wishes work xinput xmpp xorg xpath
Powered by Blogger.