GMail, we need to talk. We need to talk about how substring searches do not work, 11 years after your launch. We need to talk about how you've slowly engineered out everything that was good and unique about GMail. We need to talk about how you "flattened" the style, removed colors from threaded messages, and just decided to use shades of grey everywhere. We need to talk about how difficult it is to tell one message from a forwarded message from a divider line from attachments from the next message in a thread. We need to talk about how you've decided to replace the window manager by insisting that my messages be a window inside of a browser window, but ones that adhere to no UI standards. We need to talk about how you've hidden the date of a message, and hidden the To: address, and the Subject, so that what used to be a simple tab action now requires multiple mouse-clicks.
Imagine this: you are Google. Raking in billions. PhD level scientists are hired to mop floors, and serve coffee to everyone. You hire top level designers, pay them $250k/year and tell them to improve Gmail.
And they do...Gmail gets better. You make more billions, hire more designers, pay them more and ... tell them to work on Gmail.
You now have 10 designers, "designing" GMail. Come end of year performance review all those highly paid, world class professionals will want to say "I did <x>, <y>, <z> and I deserve a raise". Nobody will want to say "Yeah, GMail is actually great so I left it alone and didin't muck with it, can I have my raise?". So, they go to town and re-design it every 6 months. Review comes and well they all can claim they have successfully re-designed, flattened, styles etc, added custom iframe, windows, etc.
This happens with software teams, that is why so much software is re-written, people don't get rewarded for maintaining as much as for creating. So they have to create, and when things get crowded, that means tearing down and then re-create.
> Nobody will want to say "Yeah, GMail is actually great so I left it alone and didin't muck with it, can I have my raise?"
Does this incessant desire for change/to be seen to be doing something have a name?
It came up in a political commentary I was reading (about how politicians have to be seen to do something, even if the status quo is better), and I've noticed it in the "curve of website disillusionment" for clients ("Our website is 3 years old ergo a new shiny one will be much better. No, we didn't do any research but it's obvious 'cause it's old")
I've referred to it in conversations as movement vs progress: things that progress you forward toward a goal, or simply movement which could be in circles, backwards, side to side, hand waving etc.
Interesting point on managers. Yeah I've seen that happen. They are highly paid, engineering team is smart, small and self reliant, So management doesn't have much to do. But that doesn't mean they won't do much. So they throw themselves full on into "process", "communication", "meetings", new agile workflow, reorganizing the team, implement new metrics etc, etc.
A lot of it makes sense from the point of view of end of year reviews. They have to say they did "something" to justify pay raises and bonuses. The sad part is that a lot of times that busy works makes everything worse.
MS Office is much more usable now than it was in 2000. Consider a new user, with no experience in office software at all. They're presented with either two rows of incomprehensible buttons, no logical order, and additional features hidden in the menu bar, compared to having features (reasonably) logically organised behind tabs and grouped into sections and frequently with labels.
Sure, for power users, some things take two or three clicks which required one previously and I understand that's frustrating, but Office's old interface was a complete mess which got by on inertia - and there are still keyboard shortcuts and the customisable hotbar.
I hate this trend. Do people design soldering irons or cars to cater for newbies who didn't spend even 5 minutes trying to learn how it works? Programs like MS Office or Google Maps are tools. By dumbing them down, companies are making them less and less useful. And apparently nobody cares about power users (i.e. people who will actually use your tools) because there's more money in throwaway users.
Google, could you please make Google Maps Pro, with consistent interface and advanced functions available all the time? I'll be a paying customer.
MS office post 2007 is an utter mess for novices and power users alike. I say that as a "power user" who used office apps 8 hours a day pre-ribbon and post-ribbon and has to help people who don't.
The ribbons manage to unify the drawbacks of both the button bar and the menu bar that you mentioned:
- it is composed of "rows of incomprehensible buttons" with no logical order (Where is the most-frequently-sought insert-column button in Excel? Hint: not in the "insert" ribbon tab, that would be too easy. Where is the function to create collapsible rows/columns? In the "Data" tab...)
- it has additional features hidden so well that nobody can find them without googling - Excel's most powerful features like creating named cell styles are now only reachable by clicking through several layers of unintuitive unlabeled mini-buttons tacked on as an afterthought. Yes, I have the muscle memory for remembering half a dozen triple-chord keyboard shortcuts like Alt-D-F-F, but nobody who didn't use Office 2002 will be able to use the new version at anything approaching full speed - that is exactly the opposite of UI discoverability.
>Consider a new user, with no experience in office software at all.
But is that even a reasonable share of the market any more? I'm hardly a spring chicken and I was exposed to Office in elementary school. I would imagine that the proportion of people who've never used an office suite (and can't use Google to search for answers for how to do things) is fairly small at this point.
In my experience, adding the Ribbon was a huge step backwards in terms of usability, not because the Ribbon is inherently bad, but because it was a very big, very abrupt change from the File | Edit | View ... menu system that preceded it. I was working on an IT helpdesk at the time the Ribbon rolled out (with Office 2007), and the backlash from users who couldn't apply the same set of steps that had always worked for them to do things like adjust paragraph spacing was immense.
I feel like Microsoft was blinded by its focus groups. Perhaps the focus groups included people who weren't familiar with Office, or with computers at all, and that skewed the results enough to make it seem like the Ribbon was a good idea. But in practice, the cost of adopting the UI is not just the cost of learning the new UI, but also the cost of forgetting the UI that preceded it. The Ribbon might be great if you're starting from scratch with a new UI, but the existing Office 2000 UI was so entrenched, the cost of forgetting was huge.
The problem is, new users are "new" for a few weeks or months at most. After that, they join the massive groups of people that have to spend years using tools.
New users should be given tutorials, online help or other kinds of training that successfully bring them up to speed. Those kinds of materials are then cleanly separated, and they get out of the way forever once the user is acquainted with the system.
Uh, Windows 2000 was good on supported hardware but had really poor support for typical consumer PCs. XP was a security nightmare, although XP SP2 did improve things. Hardly a "peak". Newer versions of Windows, with malware protection enabled by default, do a much better job of protecting customers.
The big win for me is that somewhere between Windows 7 and 10 (not sure when), Microsoft added a feature so that the OS wouldn't bluescreen if the display driver crashed -- it would fall back to its "basic", Microsoft-written driver, and try to restart the actual driver. This is just one example but imho Windows actually became more reliable over the years.
> The big win for me is that somewhere between Windows 7 and 10 (not sure when), Microsoft added a feature so that the OS wouldn't bluescreen if the display driver crashed -- it would fall back to its "basic", Microsoft-written driver, and try to restart the actual driver. This is just one example but imho Windows actually become more reliable over the years.
My brother actually managed to play through a good portion of Bioshock on Windows Vista with a broken graphics card that faulted every couple minutes. The screen would just flicker black for a fraction of a second as Windows restarted the driver. I was rather impressed.
No doubt that Windows 7 or 10 are much more stable. They could have achieved that without making massive changes to the UI. My point was more about the UI capabilities. the general capabilities visible to a user haven't changed much between Windows 2000 and Windows 10 but they have made massive UI changes at least 4 times requiring the user to re-learn how to do the same thing.
I agree. I think to a certain extent, this curve has always been the case. However, in the old days, you as a customer had an easy choice to keep your particular piece of software at the top part: Just don't use the newer version. Of course now, thanks to the "cloud everywhere" mentality and the agressive push for auto-updates that seems less and less possible.
Version 2008 looked good, was really nice to use, easily customizable
From 2010 on it turned into a gray, depressing mess with less features (toolbar buttons almost indistinguishable from each other, no macros, toolbars can't be customized by just dragging things) .
I wish this observation would be confined to software - it's basically the most annoying - defining - aspect of capitalism. It's on one side the engineers and designers trying to do something for their own sake - and on the other side it's the constant encouragement to change something about a product to lure customers into purchasing the new "improved" version of ... the jeans, the car, the shoes - you name it.
There is an accompanying ideology that regards a proliferation of choices as an emerging consequence of greater multiculturalism/individuality. This is a very convenient analysis for market apologists, and more importantly, it puts off a kind of anxiety-filled absolute reckoning of reason and utility in our decision making. If we were more rational we would likely have to confront an essential emptiness...or maybe that's just how it seems to those of us embedded in the prevailing values.
I've long used Latitude/Location History and it's just been getting worse as well. With the newest incarnation ("Timeline") they've made it prettier while breaking it in unfortunate and non-obvious ways.
It's like that person that your tough interview process is designed to weed out - looks good, but when you hire them and put them to work you discover that they can't code, just crib stuff from Stack Overflow and discussion boards.
With "Timeline," you dig in and find things that just don't work right like arrival and departure times (if you're going to show me at a location until 1:30PM why when I go in and drag the slider are there further data points bouncing around that location until 3PM?) and filtering of bad location recording (apparently I make several 1-minute trips a night 1 mile over to a corporate office park nearby with a cell tower).
The previous version was bad (or at least irksome and clearly designed and tested only by people on 4K monitors or with very odd mouse habits for the in-day timeline), but this one is worse because it SEEMS good until you actually look at the data.
What? If you left-click in a place, a small card appears with an arrow that you can click to get directions. If you right-click, you can get directions from or to the place. And if you search, you also get a very visible round button with an arrow and the label "Directions".
How is it hard?
EDIT: Sorry if this came out as aggressive. I'm just dumbfounded, that's all.
I'm not a UX expert and don't have the vocabulary to describe what's wrong with it. What I know is that I used to enjoy using it. Not just for directions, but I'd use it for fun to explore new areas. And now I dislike and avoid using it altogether. Similar to Gmail. It's just feels like a mess of UI elements that show and hide and move dynamically, don't have fixed spots on the page, don't work consistently, etc. Like it was excessively focus-grouped or committee-designed.
I know how to get directions in the new interface. I just don't like doing it. Is it more clicks? More elements taking up space I don't care about? Hijacking the native right-click menu? Aggressive resizing/panning when I didn't ask for it, to accommodate what it thinks the new context should be? I don't really know.
Here's a few more: randomly zooming out to full-earth view, changing the area of my search after I've already zoomed in, crashing all the time on mobile, randomly giving me results halfway across the earth, randomly changing my destination when I zoom in, randomly zooming out to full-state view.
These are the obvious ones. There are more subtle UX fumbles too. If I click my house while I'm in my house and click directions, I probably don't want directions to my house. I probably want directions from my house.
Edit: Oh! That ridiculous thing it does when my phone vibrates. Someone decided they could detect frustration by using my accelerometer. Now any time I get a text message and am using maps, maps gets insecure and asks me to submit feedback.
I'm 100% with you. I don't know what's going on exactly, but five years ago it was an example I'd point to of a beautiful, instinctively easy-to-use interface, and today it seems like it never does what I want on the first try. Doesn't matter if it's the web interface or on my Android phone, though I think those are screwed up in different ways.
It has become gnome of map applications. It used to open so fast 10 years ago it was surprisingly amazing, now it takes more time to load and crashes my browser randomly. I dread opening google maps now thinking if I will have to reopen all my tabs, and I literally use it only on chrome now.
I feel uncomfortable using maps these days too. Something about the dynamic moving and bouncing of elements when I'm just moving my mouse makes me insecure. I can't easily figure out where to find things.
On mobile: Do I press, or do I double tap, or do I long press, or is it in a menu, or is it just not available?
I know that tapping will drop a pin, and I don't want to drop a pin, (because there's already a pin from my search and that's in the right place) so I don't want to tap. But tapping is actually what I need to do to get rid of the interface and just leave me with the map. But only after I've done whatever it is I need to do to make the GPS dot active. I'm on a bus. The GPS dot shows me where I am. The pin shows me where my destination is. Maps also shows every bus stop (I complain, but this is one of the things that makes me feel like I'm on StarTrek) but you only get the bus stops if you zoom right in. So, I spend some time zooming in and out. The map images keep loading and unloading. I kind of wish Maps would realise that I need the bus stops. I know there probably isn't a setting for that (because FUCK YOU USER WE KNOW WHAT YOU WANT) so I just keep pinch zooming and unzooming, and hoping I don't miss-pinch and un-drop the pin.
Maps is great if you use it often enough to learn it. I use it just often enough to not remember it next time I go to use it.
(All maps are amazing. I remember using Autoroute from 5 1/4" floppies and a Hercules graphics adapter and that was amazing, so GMaps on that slab of glass does feel like magic. I moan too much.)
On my phone, if I type in the name of a store, it gives me the most popular store of that name in the United States instead of the nearest one, because I have wifi locaiton turned off. That's the only reason. If I turn it back on (I don't want it on), the results are sorted as expected.
This is clearly an intentional anti-pattern, and it's been there for over two years. I have to actually type, for example: "Fry's electronics, city, state" (remember, this is on a TOUCH KEYBOARD) to get where I want to go.
It shouldn't really be in debate that the UX of maps has gone way downhill. There's simply no comparison to the older version and I've never met anyone who wasn't disappointed by the changes.
>On my phone, if I type in the name of a store, it gives me the most popular store of that name in the United States instead of the nearest one, because I have wifi locaiton turned off. That's the only reason. If I turn it back on (I don't want it on), the results are sorted as expected.
I'm assuming you do have GPS location turned on, and that you're allowing location access on that page and waiting for a GPS lock before searching, correct? (And if not, how do you expect it to know where you are?)
Maps used to be much better at searching within the area I'd already focused on. Now whenever I search it seems to either search near my current location or zoom out and search my whole city. It makes it very hard to e.g. look for a bank branch on my route home from work.
This 100x. I don't get it: if I've zoomed in the map or panned, and then make a search -- why isn't it obvious that I want the store closest to the area I'm looking at, and not one 30 miles away?
They know what cell tower you're connected to, don't they? The original iPhone didn't have a GPS receiver, so this was the only way Maps could infer your location.
I also liked the one period of time I used to type in my US zip code for Movie Showtimes and somehow be routed to theatres and showtimes in a town outside Naples, Italy. This was on a desktop, not a phone.
Not the biggest deal in the world but it seems every useful thing Google comes up with seems to degrade in quality over time and increase in complexity.
My theory is that this has something to do with the average Googler's tenure lasting 3 years. With such a high turnover, the overwhelming majority of employees are probably much happier achieving the supposed glory of making something new rather than understanding and maintaining the old system.
Maybe another reason for the change is that, back in the Marissa Mayer days, hiring standards used to be much more brutal. Google was still reaping the rewards of bucking the trends of web portals jammed with ads, but eventually lost its confidence in nerd-driven products and instead began to imitate Apple, Facebook, and inexplicably, Microsoft...
My personal one is that they just can't stop tweaking. It's not like there aren't some fundamental improvements with a lot of their services but everything needs to get touched and then it ends up like that car Homer tries to design in that episode of The Simpsons.
I know that you're supposed to constantly iterate and try and improve performance, design, etc in software but is there anything in the general literature / case studies of 'improving' too much and degrading the core product(s)?
Not to mention that it's always funny to search for an issue on Google, find the links to the forum, see all the people that chime in with a similar issue, the few souls that try to help, then, as you scroll, the further desperate comments of "They don't care about us." Shouts in the wind.
Are you using the actual native app or just the mobile website? If you have wifi off, it should still have access to your GPS to get your location, unless that is either off or blocked/unavailable.
If the website then how would it know your location if you don't allow it?
There are several better failbacks than "most popular store in the US." Starting with "Store closest to the centroid of past N addresses that I have clicked."
Its actually worse. Only this Saturday I searched for a ice cream shop and it directed me 2.5 miles away. So now I am riding the prescribed route when I see a different branch of the same shop in 0.6 miles. Baffled I looked again at the search results and for some reason this nearest branch was listed at number 5 and I had to scroll down to find it. This when I had GPS and 3G on.
Google has very good geo-IP location detection, almost creepily good in some cases. Scroll down to the bottom of the search results on a desktop browser sometime and see for yourself.
Yes, there's a round "Directions" icon. This time. Frequently, searches for a given address don't bring up that icon or any other UI affordance that indicates how I might get directions to the destination.
However, Maps does reliably fill half the screen with photos, random chunks of white space, and other irrelevant visual junk as seen in the above screenshot.
I had absolutely no idea that Maps had a right-click menu, so thanks for that tip (seriously). Right-clicking on an AJAX-based Web page is obviously not a standard or expected UI concept... but given what Google has done with Maps over the last year or so, literally anything helps.
> "Yes, there's a round "Directions" icon. This time. Frequently, searches for a given address don't bring up that icon or any other UI affordance that indicates how I might get directions to the destination."
This is sadly a really common UX failure in a lot of different apps.
One thing that many people still don't seem to realize is that it's more important to be consistent than being smart. I like to call these "hide and seek" UI elements - they display conditionally in an effort to be smart (e.g., no directions are available to this place, so we hide the directions button). But in practice it just confuses users.
One thing I keep encouraging, to varying levels of success, is that UIs should always be consistent. If you are unable to provide a feature, disable that element with a helpful tip (e.g., graying out the Directions button and say "Directions not available from this address" or some such).
My favorite example of this is deleting a draft in gmail. Go make a draft and try to delete it. There's no button to delete it. Unless you click the checkmark! Then the button appears where previously there was whitespace. Ugh.
You have to find a real address before it can give you directions. Otherwise on the default maps page, you can switch directly to directions mode to input both start and end addresses.
My beef with GMaps is its performance. The UI is for the most part fine (though there some awful blunders that were later fixed).
Loading GMaps takes forever now and brings my entire browser to a freezing stop (Chrome OSX) until it's done doing whatever it's doing. This is a modern desktop (i7, 16GB RAM, discrete graphics). Even after it's done "loading" if you try to interact with the map right away it does so at a crawl - inputs take multiple seconds to register and panning/zooming runs at about 2 frames a second.
It takes about 10-15 seconds after load for the site to be fully functional and performant. That's insane.
Ditto the Satellite view. Yeah, having 3D buildings is pretty cool and all, but the performance is atrocious on even fast, modern machines. If that's the tradeoff I'll take the 2D satellite imagery back please.
Overall interacting with GMaps is itself a chore, and it almost entirely comes down to bad performance. This is especially frustrating on mobile if you put the performance of GMaps on iOS next to competitors like Apple Maps - the rendering performance is insanely low even on the most up to date hardware like an iPhone 6s.
Beyond performance problems, one thing Google didn't fix until recently was that it was literally impossible to see information on an address/place and nearby transit at the same time. They changed Transit from a persistent overlay that can be toggled to a special view mode itself, which would disable itself the moment you clicked on any point of interest on the map. It makes me wonder if anyone on the Maps team lives in a major city.
On a puny Macbook Air with 4gb ram and Chrome with 30 tabs open I'm getting perfect framerates and instantaneous UI... perhaps this is something local to your machine?
It's hard to imagine that Google would make a UI that slow given their extensive optimizations.
It's true! Google Maps is really fast on some computers and impossibly slow on others. The only fix is "Lite Mode" as mentioned above, and unfortunately there's no link for that on the main page.
They aren't testing their cool WebGL version widely enough, so it fails badly for many users. (Edit: or wait, is it just a 2D <canvas>? Either way, it should be fast in Safari and it's not.)
You used to be able to just click the directions tab and fill in the from and to in the textually-labelled boxes on the left. Now you have to provide it the incantation in the order it wants; in particular you must put the destination in first and then go back and put the start point after. I don't even know if you can do "via" any more, never mind how. And it seems incapable of remembering what mode of transport I want any more.
You can do via, but not for public transport, and indeed it magically hides the add destination button if you're not in one of the modes it supports (including if you're in the auto-mode and it's suggested a public transport route).
I can kind of see their reasoning, since public transport is often time dependent, but often I don't care about specifics, I just want to get an idea what links exist, and how long I'm going to spend in transport visiting a bunch of places.
You don't have to put the destination first: in the search box, there's a directions button, just press it and it'll show both boxes!
I'm with you on the "it forgets the transportation mode" complaint, by the way; even Google Now seems incapable of using it, despite being able to parrot back the setting in the config panel.
The magical mystery UI is the worst part, actually. I (evidently) have no idea what the capabilities are, because the input boxes appear and disappear arbitrarily when you do things.
Maybe GMaps is just polarizing software that seems awesome to some people and awful to others? Unless you're a Googler and/or on the Maps team? :)
One thing that is absolutely brain-dead is that (in Chrome) I can't type in maps.google.com and tab-search for my address. I can do this with Amazon.com or Netflix, but not an actual Google property? Weird.
Another thing that is just bafflingly absent is that GMaps doesn't know where my home and work are. Not just as locations, but as concepts. (Snide part of my brain wonders if the Maps team just lives at work.)
Another thing is that to make GMaps work with my Mac, I have to have a Chrome Extension to stop the map from freaking out. And then simple things like double-clicking to zoom in seems to have a non-predictable behavior where sometimes it zooms in and sometimes it tries to put a marker on the map.
Not everything in Gmaps is horrible. There are many nice features. But they've allowed the critical path through the software to become obscured and eroded.
Also I'm not sure but it would seem you have some kind of plugin that might be interfering as GMaps is very smooth on Mac Chrome and Safari.
And you might be single-clicking on the maps which brings up a gray marker to show you more detail of the nearest address to where you clicked. It's easier to just use the scroll wheel to zoom in/out.
And you can just put the address in the search bar and the first result on the results page should be the address detail plus links for directions.
I frequently, in Chrome on OSX, get Google Maps in such a state that attempting to interact with the map at all will cause it to "freak out" and randomly zoom or pan somewhere unexpected.
I find it entirely unusable on a Macbook touchpad, and even worse with a magic mouse. Sometimes I forget and go to the site, only to have it freak out on me part way through trying to do whatever I'm doing, while lagging like crazy of course, because at some point Google stopped giving even half a damn about performance, company wide.
Google Music kicks on my fan after a while. The only other thing I run that does that is Kerbal Space Program. One simulates space flight with nice (ok, decent) 3d graphics, the other plays f*cking music.
> One thing that is absolutely brain-dead is that (in Chrome) I can't type in maps.google.com and tab-search for my address. I can do this with Amazon.com or Netflix, but not an actual Google property? Weird.
Works for me, and that's definitely one I've never explicitly set. According to my settings, the tab search just redirects to https://maps.google.com/maps?q=%s
I have the same complaint. I can't explain how it has changed, but it used to be easy. Now it is difficult and error prone. I guess I just never paid attention to it when it was easy.
I can't run through it right now, but I guess there are simply too many steps to go from address to navigation. It used to be, as best I recall, input address -> click navigate. The error part comes in from making me visually choose from a list of addresses after I have already put in a specific address (search is not needed). The difficult part comes in to play when I am trying to get somewhere and the app is making me navigate through several interactions.
But you don't need to search. Open maps.google.com, then type the address and click on the arrow; it'll open the two directions input screen with the destination already filled in. Now you can just choose the origin (e.g. click on "my location") and it'll show you the route.
This is the problem. You can't argue a user into accepting that their experience is better. You have a whole swath of people saying maps has gotten significantly more cumbersome over the last few years. Telling the user they are wrong and giving them coping strategies to handle the UX awkwardness is just insulting.
I'm not even a Google employee, let alone a Gmail developer, so I don't see how is it insulting. I'm simply explaining why I don't see the problem being reported.
I also have no problem with gmail currently. I suspect only the people with complaints are being vocal here.
It's gotten leaps and bounds more useful than it was 4-5 years ago. Transit directions are much easier to use. One of the only things I can complain about is that on mobile, if I've conducted a transit direction search then the results I get are for the given time. If I miss a bus or don't leave until a bit later and want to re-run the search again at the current time (some buses may have stopped running if I get delayed leavng party for example), I have to actually input the time on my phone; there's no way to re-run the search with the current time.
It's mind-blowing to me that they wouldn't make it easier to have a transit search use the current time, and instead continue to use a route from anywhere from 15 minutes-3 hours ago. Why would I even be performing the search again if I wanted the same routes.
I have the same problems with modern Maps. Personally, I always type something in maps, weird search cards I don't want at all popup. I click the map which is only a small stripe above the cards to try to get to a map with my placemarkers, then Google Maps forgets my search and hides all the placemarkers. Rinse and repeat several times and eventually Maps gives me my darn map with Placemarkers that I wanted in the first place. I find directions then a whole separate pain in the ass compared to how they used to be as well. Sure, the thing looks better than it used to, but it works terrible.
Try Here maps. This is the old Nokia maps now owned by Android. I use this as my go-to for driving. It's much better interface and route selection is as good as Maps. Also I love the GPS speedo and speed warning. I tend to use Maps for location based searches like restaurants for the better reviews.
HERE is pretty great. I've been amazed how basic things like a list of recent searches (something that can be easily stored locally on your device) aren't even available on Google Maps if you turn their cloud location stalking off. Whereas HERE provides a pretty functionally useful app without me even having an account with them.
It might be tied to some of the search history settings? It's tied to one of the cloud privacy settings. If you turn it off, it doesn't even provide recent searches on the device itself.
I want that feature, but that shouldn't entitle Google to have that data on their servers.
Mostly it's indiscoverable. The v2 overhaul shifted everything toward a semantico-geographical google search, except it didn't improve the ergonomics, and even made the software 5x slower. The latest upgrade fixed some stuff but it still feels like a waste compared to the v1. btw, you something is wrong when you go to bing maps and you feel happy because "it" .. works.
It's pretty well hidden, but you can still compose messages in a new window. For new messages, hold down Shift when clicking on the Compose button, or just hit Shift-C (but only if you have vim-style keyboard shortcuts turned on[1]). For replies, you have to click "pop out reply" first and then hold Shift while clicking the (horribly labeled) middle button in the resulting reply pane / thing[2].
...and the fact that I had to write this comment proves your point.
Ahh, lovely.
What I really like is the strip along the top.
I want to use Drive. I click on ... wait for it ... "Drive".
Genius ... (not being sarcastic either)
I've very recently moved away from GMail for various reasons, but part of the transition was using their web panel to tidy up some things.
Wow, I agree with you that it's visually worsened. I'm shocked at how much harder it is to visually parse information in there. It's not harder to use overall per se (from what I remember) but the number of visual indicators defining what's clickable or a message in a thread or a sliding panels hiding tags or pop-up reply panels that are just harder to work out is very surprising. I actually stumbled on an interesting piece today discussing this exact point, I forget where I saw it[0].
I admit to using desktop/clients for virtually all email stuff so going back into the web panel was a surprise to say the least.
It is amazing that Microsoft is spanking Google in web interfaces of all things. Google does great stuff, but their disregard for the end user experience just ruins it.
Office 365 for email is dramatically better than GMail. Google Drive is much better, but I'm sure they will find a way to design that up too.
I honestly believe there's a revolving door of Product Managers at Google that come in, decide to leave their "mark" on a product by making a bunch of arbitrary changes, then move on to destroy other products.
Hard to complain when they're free, but... I do my best.
The example presented are the kinds of emails I actively avoid. Perhaps the author is right, and Gmail is not working with them on purpose. If that is the case, it may be because many users are like me, and they tend to not want fancy-looking marketing emails, and just want basic rich-text features.
The Schema initiative - I didn't know it was called that until now - is in the other direction, and I like it. Emails that have valuable information for me, such as airline flights, are presented not in the way those that send it want it to be shown, but in a way that my mail client thinks is best. Frankly, I want sameness in my emails.
Marketing emails are not the only ones that developers want to look not broken.
To pick an instance at random: Confirmation emails for purchases made, especially if they have tickets as an attachments. If this email looks broken, customer thinks it's phishing or something, and might end up trashing it, just because styling is not supported consistently.
And I take issue with your argument of "marketing emails are bad." You may not like them, and that's fine, but there is a large part of the internet ecosystem that do like marketing emails- such as my mother who likes learning about deals/sales/coupons at her favourite outlets.
Such a minimalist design may work for some products/brands. But just as many companies, especially companies that put a lot of focus on brand management, may want nice images and responsive design. It's not an unreasonable request.
Even if it's not as effective for click-through/customer re-engagement/etc. you should let the customer decide on what kind of emails they want. Data will prove if nice images and ads are less effective than simple HTML. Gmail shouldn't be forcing that decision, it should be empowering the greatest number of people to do the most amount of things.
I disagree. If you don't want your email to look the way Gmail displays it, use a different client. I don't want a million configurations for how I could possibly have my mail displayed. Show me a good option. If I don't like it, I'll use a different option. That being said, one thing I really miss in Gmail compared to other options is being able to do things like sort by the sender and that sort of thing. Instead I have to search for that sender which often gives me many false positives.
Responsive design? Have you ever tried implementing an HTML email that will work across different clients? I'd be thrilled if 2005-era floats worked. As it stands, <table>s inside <table>s is state of the art.
HTML makes receipts more readable at-a-glance. You can emphasise more important information, format things into proper table columns, show an image of the item, etc.
Because even a 'simple' email need to carry significant amounts of information, all with varying levels of importance.
A newsletter will multiple items will need headings (for people to scan through if they're short on time), a paragraph for people who need more details. Images, if the articles require them, also help.
Corporate email - not just advertisements, but also invoices, notifications, as such, almost always require some sort of branding. Customers expect them - a plain text email might be regarded as phishing or fake. And why force people to open up an attachment when HTML exists and can mark up those elements perfectly well?
Email also need to carry information of less importance - unsubscribe links, confidentiality notices, signatures, most of which are required by law or lawyers, but which can be made smaller or greyed out so they can be ignored when reading.
You can't do any of the above with plain text. What you need is markup, which guess what, HTML is.
I'm sure that you don't have a problem if confirmations are unstyled. But I'm also sure that a lot of users differ in that regard and would expect at least some styling on that non-marketing, but still system-generated emails. Examples include things like confirmations, notifications (that you've opted in to), or just richer and more complex conversations. I mean: to end users, the email experience is part of the brand, and if the brand emits terribad emails, or unstyled emails, it erodes confidence.
I'll grant that there's a lot of yucky and annoying html in some marketing emails. But there is still a use-case for html-styled emails, and just having plain text with "click here to see images of the food you're getting next week" is pretty darn 1994.
The article claims that Schema has not caught on, but in fact you have to be approved [1] to be able to have Gmail pick them up in emails you send and it's not really an open standard like it should be.
Moz has a great article which covers it pretty well. Basically you have to send 100 emails per day for some length of time.
Since you can't just add them to your email and expect them to work, I think that is why it hasn't caught on. I'd really love if it did though.
I completely agree. I am totally unsympathetic about it being difficult to send consistent looking advertisements and emails that resemble web pages. If anyone from the gmail team happens to be reading, please let my comment be a contrary data point to this TC article and keep doing what you're doing!
I'll agree there. One of my favorite things about using Thunderbird as my mail client is that I can force simple HTML view on all incoming mail. It ignores CSS and other garbage, while keeping basic things like bold, italic, and links in tact.
> Many users are like me, and they tend to not want fancy-looking marketing emails, and just want basic rich-text features.
Oh god, yes! I, sometimes, get really offended when people are asking my email for sending the warranty on something I bought in real life. And then, extremely upset when they start sending their colorful marketing prospectus through this address. This inbox (or any other one) is carrying very valuable informations for me to remain organized and accurate; Therefore, it must stay sorted and categorized. I do not have time, nor attention to evaluate your christmas sales on stuffed donkeys.
>If that is the case, it may be because many users are like me, and they tend to not want fancy-looking marketing emails, and just want basic rich-text features.
Of course it wouldn't be that Google is the world's largest advertising company and they are in direct competition with those advertisements that they break.
Maybe I am just really, really old at this point, but sending emails in plain text solves all these problems. I don't think I've ever received an HTML email that couldn't have been adequately represented in plain text and maybe an attached image if necessary.
I'm sole responsible for acquisition on a startup and I solely (for now) use direct cold email (it is B2B, email sent to public adresses, so it is not spam).
I only use plain text emails. I don't even use email tracking so that hidden pixel doesn't trigger any annoying warning on the email client that ruins the experience.
Sending unsolicited email to public addresses in order to sell something is spam. Calling it 'cold email' to make it sound like a legitimate process does not change the fact you are spamming.
That's not true, by either the letter of the law or the spirit of it. I own a business, and I post my email publicly because I want people to contact me. Some of those people will be selling me things--I want that! It's true that I end up with a lot of spam as a result, and that sucks, but it's better than never receiving unsolicited business email.
Sending unsolicited email to consumers is, indeed, spam. B2B is different.
EDIT: CANSPAM (US) is a bit ambiguous on this point. (Helpful blog post: http://customer.io/blog/Send-email-without-unsubscribe-link-...) CASL, its Canadian equivalent (I'm Canadian) is very clear on this point, though--unsolicited B2B to a publicly posted email address is a-ok, as long as it's not automated.
you may be right regarding definitions of spam, but as a user I treat relevant-and-targeted(-yet-unsolicited) emails very differently than I treat automated junk email.
I don't actually sell, as these business don't pay us anything. Rather they list they business on our website to their (and ours) value. Is it still spam? (honest question)
Edit: They are a business and have a "contact us" email on their website. Should I really not email them just because im not a customer, but a possible partner?
This argument never holds up. Being a doctor or an engineer is almost always thought of as a good path. But what if everyone was a doctor or engineer? There'd be no one left to do basic tasks in life and the positions available for doctors and engineers would be so in demand that these people would work for very little.
Your point can be rephrased as, "What if everyone ignored the realities of the job market?" There's no danger that everyone will do that, so there's no need to use ethical or other philosophical arguments to discourage it.
On the other hand, spamming or otherwise misusing email to deliver your unsolicited "message" at everyone else's expense is a classic instance of the Tragedy of the Commons. There is no incentive for any individual actor to do the right thing. Kant's categorical imperative is one of several reasonable ways to think about it.
If everybody visited one of this business website, the server wouldnt support. Yet, it is ethical that anyone visit. And if only people that really have an interest on visiting it do it. No problems.
I think the same for the email. If,everybody who can make a legitimate business proposal email them, no problem. And I apply the same ethics when I had my own company with a public email on my website (now im an employee and emails dont get to me)
In some countries, business addresses aren't considered PII (private information). I, as a business, regularly get emails because of the Whois, as well as I receive letters which I dutifully send back to sender, but the law doesn't identify the latter as spam. Is it the case that emails to businesses aren't spam?
No. Not by the spirit on not by the letter of the definition. Spam is bulk mail, or otherwise annoying a whole group of people. Sending a targeted e-mail to a single person (even if unsolicited, even if commercial) is not spam.
Yeah. Sadly, plain text email seems to be endangered species. But perhaps there is still hope, markdown and pals have made some inroads as all around useful document formats.
I know this is an extreme but I like this extreme.
When I say plain text, I mean plain-text. I mean not even markdown. Just plain text. The client can choose to make links clickable if they like.
Any formatted document should come as an attachment (I'd prefer PDF). I guess it should be an attachment I expect or I probably wouldn't open it. In any case, I don't like emails with tracking. It feels forced. If you really care, send delivery and read receipt requests but I doubt we do.
Markdown should stay where it belongs. It should convert into HTML in the server, not on my client.
For the first few years that I used gmail, I was set up to use plain text with the terminal theme. Why? It was the only theme that I could find that had a fixed width font, and I read enough emails with cut and pasted code that I wanted a fixed width font.
They took that option away from me in 2011. No, I don't care about a picture. All I wanted was fixed width font, and plain text. Yes, I could apply style sheets to read plain text, and they would eventually break them for me. I'm not bothering with an uphill battle to get it back. I'm just moving on.
Oh, and about rich text? If I happen to be cutting and pasting something that was bold over here, I usually don't want it bold in my email! In fact I've wanted that approximately 0 times in my life. And it is a very good approximation.
I'm with you on this one, though I didn't realise we were extreme!
Plain text cuts all the cruft and I can reasonably expect emails I receive to largely look similar so I get good at extracting relevant information.
Every client I have, desktop and mobile, displays and sends in plain text only and I'm pleasantly surprised how infrequently this breaks anything completely and how often it exposes tracking links etc.
While rich/styled text would be useful in many situations, I'd rather strive for simplicity. With Unicode, plain text is already complicated as it is and the diversity in email client software means that plain-text is probably our best bet. As far as I know, most office emails are pretty much like text messages anyways. Also, it is possible to leave stylistic/visual cues in a plain-text email such as Important (: Not really sure if that is a good thing or a bad thing to write like this...
If we must track users, perhaps we can just give them a link they can visit?
What do you think about using attachments for documents where formatting is essential?
I was not aware of that. Sorry. I just meant that if we want to send a document that should look exactly as we send it, perhaps we should consider sending a PDF document or something.
Something I found interesting in the Wikipedia MIME article is:
If I so much as write க in Tamil, I rely on MIME. (Sorry if this is obvious to readers. I am not an email expert.) For others like me, the Wikipedia article on email spells it out:
> Originally an ASCII text-only communications medium, Internet email was extended by Multipurpose Internet Mail Extensions (MIME) to carry text in other character sets and multi-media content attachments. International email, with internationalized email addresses using UTF-8, has been standardized, but not yet widely adopted.
Outlook formats plain text email with a square monospaced system font, making the rest of the world look retarded. The favour is reciporcal, because, as a gmail user, Outlook-sent emails often have this delightful pastel cyan rounded font, making them look like kindergarten artifacts.
If we're just displaying tables rather than interacting with them, that problem is solved.
+----------------------------------+---------+------------------------+
| Col1 | Col2 | Col3 |
+----------------------------------+---------+------------------------+
| Value 1 | Value 2 | 123 |
| Separate | cols | with a tab or 4 spaces |
| This is a row with only one cell | | |
+----------------------------------+---------+------------------------+
You cannot guarantee squat. Which is pretty much the gist of TFA, within a single vendor's various client interfaces to its own unified email system.
However, as I put on my True Scotsman hat and kilt, any true email client will present monospaced fonts, therefor those which fail to do so aren't real email tools.
In the old days instant messenger didn't work if the other party was offline, it didn't usually support more than 1 recipient, and it was just stuck in proprietary programs as it is these days. SMS is too short.
…except Gmail doesn't support plain-text email well either. They've been converting plaintext quotes to literal '>' characters for at least 3 years and counting:
GMail, we need to talk. Thank you for making it complicated for people to send me "rich" text emails. Thank you for making it harder for embedded tracking images to function correctly. Perhaps the moving target that are your rendering changes will lead to a simpler, more featureless, email landscape like we once had. I really appreciate the time and money designers are forced to spend on annoying me; if i have to deal with the full frontal assault on my eyes, they should at least have to work for it. Please keep up the good work and thanks for encouraging people to focus on the iOS market instead (those people will buy anything AMIRITE?).
> Gmail and its plethora of rendering quirks is a big reason why developers avoid working on email. Developers like well-defined and documented environments, and email is anything but. Although many email clients suffer from rendering issues, getting an email to display nicely in the various Gmail desktop and mobile clients gives developers the most angst.
I've become convinced that gmail's goal is not to improve email, but to extinguish and replace it. This article talks about one minor aspect of that. In addition:
* breaking established addressing conventions. Gmail (last time I used it) made it difficult to distinguish between To: and Cc:, both in received emails and when sending.
* messed-up quoting, and other broken conventions.
* the one that's been discussed to death here recently: it's on the way to impossible to send email and expect it not to be silently dropped if you're not on gmail.
* the email authenticity measures linked to the last point also screw up mailing lists and forwards. Most of the DMARC notifications I get from gmail now include a bunch of FAILs for people that forward from another domain to gmail. Soon, they'll simply stop receiving most mail.
Email is walking dead, and gmail is largely to blame.
Back in pre-history, they did a similar number on their usenet interface, again after having offered a decent product for long enough to lure people away from running local clients.
They killed usenet off good by providing a reasonable interface that encouraged people to link to it .. then destabilising the URLs. I've seen wiki links to google groups that have been changed 3 or 4 times before people simply gave up trying to re-find the original target.
> While the plain-text emails we created had some formatting capabilities (hyperlinks, tracking pixels, etc.), they were as close to plain-text as we could while maintaining the ability to track opens and clicks for testing purposes.
That sounds like "HTML email with minimal cruft" rather than "plain text" to me.
Oh god, HTML email. Gmail and Outlook literally make coding these things an absolute hell due to their dreadful lack of supports for any form of standard, and the author is damn well spot on about it.
Indeed, the only good thing I can say about Gmail is that they didn't try to use the Microsoft Word engine to render HTML like Outlook did. The way HTML email is set up and the table filled soup like code that's needed to use them well is about half the reason most developers fall back to things like Mailchimp templates rather than trying to make anything half decent.
It's also why it's a real shame that the old email standards project kind of got abandoned, because absolutely zero progress has been made on this front for years upon years (outside of Apple's smartphone email apps).
And yes, while plain text email does 'work', good luck convincing a company (or your boss) to go with something basic rather than the equivalent of a glossy print brochure.
If you haven't already, you are better leaving Gmail
1. Have your own domain for emails. This is important. Gmail (or any service provider) can block you/remove your account for no reason and with no recourse. Your email is very important if you are doing serious work and have financial sites and data linked to that email address.
2. The very reason that I used Gmail was speed and storage. The very reason that I left Gmail was speed. Storage is no longer relevant now. Speed become a problem in 2012-2014. It seems better now, but there seems to be "rush hours".
3. The Gmail web app is no longer good. I like a very simple interface because I write very long email. Why make the "compose" show up as a pop-up? I write emails and not chat messages.
4. The SPAM filter is good until it catches a couple business emails. This is when I decided that no SPAM filter is better.
5. If you don't pay for the product, you are the product. I left Gmail after using it for 8 years. I was fine for 12 years but as the business grow I thought about that line. Suddenly, I had strange thoughts about Google using all my data.
You have a really interesting argument against the oft-repeated "If you don't pay, you are the product." I think you are right, that there is great power in anonymized data, but ultimately motives matter. From what I saw as a former scientist, the goodwill of patients donating their tissues formed the foundation of great cancer (and neuroscience, and HIV, and...) research. IBut I think there is a social contract for scientists (and open-source programmers, and nonprofits, and...) that gives them credibility. Maybe it's our human nature to trust them, and for them to trust themselves?
But 1)corporations are obliged to care about profit, and 2) anonymized data can be de-anonymized with (to me) astonishing ease. In Data and Goliath Bruce Schneier cites a dozen cases of de-anonymized data, from health records to Netflix history to personal identification.[0] It is possible that I'm just more paranoid than the average GMail user, but this scared the bones out of me.
I'm an ex Google Apps user and now a happy FastMail user. I left Gmail for many reasons, some technical (poor IMAP support, unhappy with the available clients) and some ideological (I don't want to vote with my wallet for monocultures). However ...
> Have your own domain for emails
Gmail allows you to own your domain. You just subscribe for Google Apps, at $5 / user / month, which is reasonable.
> If you don't pay for the product, you are the product
Well, with Google Apps you very much pay for the product. But I don't agree with such thoughts not even for the free Gmail. Yeah, it's a cute line, but has nothing to do with reality, the reality being that products being sold have to be produced, the essence of capitalism being to lower the costs and increase the scale, turning sold products into commodities and users, by being people, are anything but commodities, being in fact entities that in comparison with hardware or whatever, can always get pissed off or switch to a competitor. Such lines also assume a very simplistic view of the producer / consumer relationship. Cash changing hands is nice, but in our world an ecosystem is much more complex than that, much like how in nature there's a very complex and life creating relationship between predators, cattle and grass, each link in that food-chain being essential for the survival of everything else. And the genius of Google has been to recognize that the welfare of users is paramount to their success, though unfortunately they seem to forget that ever since Google+ and Vic Gundotra.
More realistic would be to say that a zero price tag is in fact subsidized with something else, as there's always an implicit cost. The cost of the free Gmail is lock-in. And just like in nature where monocultures are unhealthy for everybody, free Gmail is hurting competition and worse, they are dropping the ball on standards. So Android is my favorite mobile OS, but can you believe that Android doesn't have a modern, well designed and standard email client? I mean, you're good as long as you're a Gmail user. Exchange also works. But for plain IMAP? Say goodbye to push email and to basic functionality, like archiving by swiping. They should also be ashamed that Android doesn't support CalDAV or CardDAV out of the box, though granted you can buy some add-ons.
Excellent points. I really like your reinterpretation as "a zero price tag is...subsidized with something else...The cost is lock-in." It is definitely more accurate with respect to Google Inc. Though to my ears it sounds like a more generalized version of "if you are not paying for it, you are the product." But don't we almost always think of Google when quoting this line anyway?
Android has bad IMAP support -- is this a limitation of Android, or of IMAP? (are there push notifications for iOS?) IMAP has been a thorn in my side for ever, but I assumed that was the fault of an old outdated protocol?
What I _do_ like, is that when I get an email with things like flight details that my client can add it to the calendar.
if I have any complaint with gmail it's that they're too successful, running your own mail server that's standards compliant and runs all the DMARC/SPF/DKIM + PTR is fine, but until you have a reputation with gmail 70% of people aren't seeing email from you (and I'm not talking marketing emails).
which is self-fulfilling, because my emails go into the spam bin, which means when the spam bins get churned into the anti-spam learning engine it has, it sees that "(S)He's been marked spam before".
So it somehow centralises email, which is at it's core decentralised. :\
Gmail, we need to talk about how you make our unpopular product unpopular even with people who want to like it. We also need to talk about how to let more junk mail into people's inboxes and maybe you could just skip the whole thing about sorting important mail out from Sears flyers?
techcrunch, we need to talk about your article titles.
Addressing the content of the article:
- "Gmail Makes Developers Shun Email", I can't think of any developer I know that shuns email, and they all pretty much use gmail.
- "How Gmail Breaks Email", to be frank... not even in the slightest. Sure, it sounds pretty annoying having to jump through all those hoops for a fancypants email.
- "Google should take the lead in bringing the web mail stakeholders from Yahoo, Microsoft and AOL together toward the support of a common set of CSS and HTML."
Who makes a demand like that of a company? Hey, Company X, gather every industry leader please and force them to all standardize to one spec. Am I the only one that would prefer companies to not be in the driver seat for these kinds of decisions? Isn't this what led to the IE6 debacle?
I really don't get these articles. Today I've read HN front page articles about horrible gmail is (my favorite email service for the past 10 years) and "Amazon kind of sucks". I must be out of touch.
It's an opinion piece from a guy who... apparently... wants to put image carousels into emails and is upset that GMail doesn't let him, and yet dares to intrude upon his beautiful iOS.
Hopefully no one exposes him to the ocean of old Outlook clients with crufty Word-based HTML renderers out there; he might die of a heart attack. Mentioning Pine/Alpine would be cruel as well.
Author here. Haha yes I write on experimental techniques like carousels mainly to get developers curious into the kinds of stuff they can do with just CSS.
I'm very aware of Outlook's backward HTML support, but at least its a single client. With Gmail, each of their clients has weird quirks that make it frustrating to code for.
As developers sometimes creating HTML emails are necessary, such as creating system generated reports for business users. Right now, the process is fraught with issues - including figuring out Outlook, so fixing this does not just benefit email marketers alone.
Media queries in email would allow a number of information leaks about how the email is being viewed, which I don't especially like. Allowing external images is already pretty awful.
Images are much less problematic for Gmail users since they started proxying them[1]. Now the sender can't tell if you opened the email by checking if the image was downloaded.
I was aware of the proxying, but didn't realize that gmail always prefetched the image. This implies that even sending each recipient a unique image URL will not work?
Author here. Gmail does not prefetch the image. Appending a unique image URL will still allow the sender to know when you've opened the email. The proxy eliminates your IP as well as user-agent so the sender can't tell where you are and what you're using to read the email.
Huh, I was going to argue that the User-Agent string would probably leak the same data as media queries, but if it's a generic proxy fetching them, then I guess not.
Yeah, I guess it could reveal which interface you're using or something, but it is probably static and doesn't reveal anything except maybe that it's Gmail's auto-fetcher.
Nobody seems to have mentioned that the author works for FreshInbox - an email marketing company.
Sorry Justin, but sending web-pages by email is exactly why email sucks. No messaging apps allow HTML+CSS markup to be sent, and there's a reason for this - it's not readable. Instead send a URI/Link and the recipient can view the HTML using a tool that's designed for that - the browser! Web-pages aren't messaging. That's what the WWW is for, not email. Apple is totally at fault for introducing HTML support in emails, they didn't see the spam coming. Email is a messaging protocol, not a news reader or web browser. NOBODY (except for advertisers) wants SHOPPING CARTS in their emails. What non-sense.
Allowing style to be included in a web based email client is just asking for trouble. It's not just hard. It's close to impossible thanks to CSS global nature. You would have to literally parse out the css and rewrite it as well as all the html content to automatically uniquely namespace it so it doesn't collide with the rest of your page. And meanwhile you've just created a new attack vector for your email service with security implications you may not want to take on.
All of that just so that you can pretend an inherently text medium is actually a way to deliver web pages directly to users. This is a lot of work for questionable benefit to GMail's actual users as opposed to Email Marketers.
I'm the author of the article. All the other web mail clients from Yahoo! Mail, Outlook.com, Comcast and so forth support the style tag. If it was a big security attack vector, all those clients wouldn't be doing that.
The other webmails rewrite ids and classes and prepend unique identifiers to prevent them from clashing with their own namespace (AOL is the only one that puts it in an iframe).
Interestingly Gmail webmail actually supports the style tag but strips out id and class selectors in them leaving element selectors - so they're actually doing some heavy duty parsing already.
As for it being an inherently text medium. I would say it depends on the recipient. Emails do not need to be web pages, but simple CSS styling can go a long way to ease readability in emails - especially in a long curated newsletter like Product Hunt.
There's this thing called "iframes" with an attribute called "sandbox". And to prevent tracking or other cruft with images or external content, well, writing a CSS filter (url:... values) and filtering out src="" attributes in the HTML is not THAT hard.
There is no valid reason anymore for not iframe'ing your HTML emails any more. Browsers are consistent enough in rendering that you can achieve trivial cross-platform pixel perfection (font rendering aside).
It's email. SMTP. It's not instant messaging and it sure isn't a browser.
Point blank, it's easy to see that this is by design and not by quirk or fault. If anything Google are looking to maintain usability by keeping as close to RFC spec as possible.
One of the biggest problems facing email today is SPAM, and general deliverability. Once you veer far off spec it gets a lot more difficult to manage these two issues.
For example, look how nice Greylisting worked in theory, until you realize that most IT or hosting companies don't have a clue how to configure an MTA within RFC specs and do silly things like retrying delivery every 5 mins.
When your dealing with complex systems, it's better to keep each part of that system well oiled and within a predictable spec. And then Microsoft adds HTML support to Outlook so that marketing execs can throw 15 sales related images and some wording about saving trees...
Often, solution providers should have given an alternative when asked to do things that break from standards. As we often say in this field, "Just because you can, doesn't mean that you should."
>look how nice Greylisting worked in theory, until you realize that most IT or hosting companies don't have a clue how to configure an MTA within RFC specs and do silly things like retrying delivery every 5 mins.
that's what needs to happen for greylisting to work.
Thanks, but you misunderstand my point. Actually you pretty much prove my point that most MTAs are run by those who can't grasp that there are RFC standards out there.
An MTA should never resend after only 5 minutes. It should wait at least 25 minutes before even the first retry.
I've seen many set up to retry every 60 seconds, giving up after 5 tries. One IT tech even told me it was so his boss doesn't have to wait 3 days to get a non-delivery report! Lol.
I'm not going to add another "just use plain text" comment. I really want to send images (in the text, not just attached), and simple rich text features such as bold text can really help readability. But that's not a huge responsive webpage crammed into an email -- it's just rich text. It's html entirely without style, neither inline nor in a header. That should be the standard email format apart from plaintext. My client can style all emails like I want them in terms of fonts and so on. Consistently.
If the sender wants me to see a website, just send me the damn link?
Gmail- thanks for the free email service for the past years. I don't mind the pretty minimal advertising you include for a free service. I could run my own mail server, but yours is fine, and let's face it - far better than the email that provided by my ISP. (You would think ISP's would be more motivated to provide a better service with the single feature that is most likely to stop me switching provider.)
Oh- what's that? Do your competitors find it hard to advertise on your platform? That's so sad. I really value the service that email marketers provide.
Thanks again Google,
Your search is still ok - but it is your email account that is good enough to keep me as a customer. An Bing is awful. Except for the arial photos - they are better than the shitty ones on google maps.
Being the only coder at my startup, you have to do everything. And the other day, I had to spend the whole day troubleshooting rendering differences between Outlook and GMail. It reminds me of the old days, back around the time of IE6, when it was all-but-impossible to get FF and IE to render the same HTML - except worse, by at least an order of magnitude. It's astounding to me that the same companies who've worked their asses off to make sure that their browsers all render HTML and CSS exactly the same, haven't given the slightest thought to making sure that their email clients can get along. This is a huge time-suck for developers. Glad to see that they're getting the public shaming they deserve.
Author here. Glad to hear this from another developer. I think once developers are tasked to code an email, perhaps something as simple as a notification or a report, they'll quickly realize how truly foobar the email situation is.
Short emails may be fine with plain text, but once you need to present any sort of structure, HTML becomes necessary and there's no reason why email needs to be this difficult.
Why has Gmail made it so hard to delete emails? I use various imap connections to gmail, and they've all replaced the delete icon with an archive button. Only the web app still has an easy to click delete button.
As someone who reads all of my email in a terminal: I don't see the problem here. I think gmail has a lot of issues, but making it harder for you to send me spam is not one of them.
You say that like a terminal is a step back. I see it as the opposite: I started with email in web-based systems. Terminals improved my email experience 100-fold
I'm saying that your usage pattern has little in common with the 99.99% of gmail users, hence isn't super-relevant as a data point to be used in the design of gmail.
Schema has the same problems as Microformats where adoption only takes off if the market demands it, or it gains enough momentum that people can't imagine webapps without them.
There are countless other standards this article overlooks which have been implemented very well in recent years in Gmail, so why does it lambast the lack of Schema?
Yes! I fundamentally believe Matias Duarte and his "Design department" at Google is one of the worst things to ever happen to the company. As Material Design has taken over, product by product, performance and usability has taken a backseat to branding and animation.
Author here. The majority of non-personal emails contain some form of styling for readability. And the styling is done using CSS. The only issue with Gmail is that they force you to inline them which makes coding email very tedious.
I like that it forces the use of templates that are rigorously tested and will seldom change.
Because when I mark the email as spam... it will very likely bin all other emails produced by the same template.
I only slightly jest, but the best emails are the ones with the least possible styling. These are also the most deliverable emails, the ones most likely to make it through to an inbox.
I sure would love it if the gmail client supported more modern web stuff. The history of web development has always been a history of the Lowest Common Denominator: I'm held back from doing XYZ because this browser / email client doesn't support it and that represents a big fraction of my users.
It's not just the gmail client, of course; nearly all rich mail renderers are astonishingly baroque and broken (take a look at which engines Outlook uses to render html mail - sometimes it's the msword renderer).
Of course, we could all just agree to send plain text emails with a link to the web page-ified rich version of the email, but that's not satisfying. I work in the borderlands of rich email - I'm not sending advertising stuff, I'm just sending notifications to users, but those notifications look a lot better when they've got some good markup.
It's not just Google Mail that doesn't support proper HTML emails.
The Android mail client doesn't, the iOS mailclient doesn't, Outlook actually feature-retarded between 2010 and 2013, most webmailers break on anything that's not tables with inline styles, Lotus Notes Webmail is actually fairly good but unfortunately depending on the platform the full Notes client uses either its horrid internal engine or something provided by the OS that may actually be IE7 (shit, you design a beautiful newsletter for Mac Notes, and it falls apart on Windows; also weird weird suboptions hidden deep inside Notes make it balk on anything but color/background-color/height/width)...
The only mail client that can render full-featured HTML5 newsletters is Thunderbird.
GMail/Google, why are you tying Google Fi to my GMail account? I got it for my wife, who now has to have my GMail account setup on her phone for the service to work.
You say you don't have family plans yet, but it sounds like you will. Do you really expect me to be okay with having my GMail account setup on my children's phones?
Why use Gmail when IMAP will sync all your devices? Android has a good IMAP client. On desktops, use Thunderbird. What value does GMail add? Besides ads, of course. And Google building a profile of you.
the only thing from Google I really like, at this moment, is Golang. Products I used to love are either gone (greater) or ruined. It's amazing to witness a long, gradual process of unavoidable increment of entropy/mediocracy. so long, and thanks for all the fish!
I've been wondering about this and after reading that link, I think I'm going to switch from the Gmail iOS app to using the included Mail app. We'll see how it goes, but I do want my email as the developer intended.
As someone who just made the switch in the opposite direction (I've been using Mail for multiple emails accounts for a few years) I'd strongly suggest sticking with Gmail's iOS app. My email experience on mobile has been much more enjoyable, Mail was glitchy and difficult to use/organize for me.
I use the Inbox by Gmail app on iOS and I really love it. I never got into the regular gmail app. Inbox is far better than the standard Mail client for gmail IMO.
And apparently something about HTML emails too.