5 ms·

Two examples of bad user experience caused by HTTP prefix being hidden in Chrome

64 points by augustl 16y ago

konad 16y ago
concur, hiding the http is annoying to me
bruceboughton 16y ago
These are poor counter-examples.
- augustl 16y ago
  If you're referring to my examples, I'd love to hear some arguments :)
- al_james 16y ago
  I honestly think the first example is a feature not a problem (its a good thing!) and the second is an edge case that will never happen to the vast majority of web users. Whereas hiding the http/https prefix simplifies the web browsing experience for the vast majority of web users. Worth keeping!
citricsquid 16y ago
I repeatedly posted about this being terrible on reddit, I was repeatedly told to stop bitching. It's a pain, surely this is something that you don't do, the user highlights "foo.com" and is instead given "http://foo.com http://foo.com. It's as bad as the sites that use Javascript to inject "read more @ http://bar.com.. http://bar.com... It wouldn't be so bad if it was optional, but don't expect that of Chrome, oh no. It also leads to an inconsistent experience, I'll quote myself: "If the URL you're using is http://website.com http://website.com and you copy website.com, you get http://website.com http://website.com, but if the URL you're using is https://website.com https://website.com and you copy website.com, you get website.com. It not only messes with copying by changing what you copied into what it thinks you want, it's inconsistent with that guessing."
- augustl 16y ago
  Let's hope my issue helps. I wrote it as to-the-point as I could, with no bitching and hopefully well reasoned arguments. They're obviously tired of people claiming suckage without explaining why they think it sucks, and probably tired of even hearing about the issue in general. They've merged LOTS of dupes into http://code.google.com/p/chromium/issues/detail?id=41467 http://code.google.com/p/chromium/issues/detail?id=41467
- Devilboy 16y ago
  It's as bad as the sites that use Javascript to inject read more @ bar.com... No it isn't don't be ridiculous
  - citricsquid 16y ago
    uh, yes it is. It causes the same level of annoyance and has the same solution. Load up Chrome and go to http://hackerne.ws http://hackerne.ws, copy "hackerne.ws" from the address bar (all that displays) and you clipboard contains "http://hackerne.ws http://hackerne.ws, to only get "hackerne.ws" in your clipboard you must paste somewhere else and remove the http:// http:// Load up dailymail.co.uk and select an article, highlight a portion and copy it. Now your clipboard contains your selection + "Read more at...", now you must paste it somewhere else and remove "Read more at..." if you want your selection. Both cases can be argued it's "for the user": The Chrome example, "oh well obviously the user wants to have http:// http:// there!" and in the dailymail example "Well obviously if they're sharing that part of the article a link to the full article would be good!". Both examples cause the same problem and have the solution, therefore they are comparable.
    - St-Clock 16y ago
      I think he was being ironic (look at how it was quoted: @bar.com and not @http://bar.com http://bar.com)
    - sausagefeet 16y ago
      > Load up dailymail.co.uk There was your first mistake
paol 16y ago
I find Chrome full of little annoyances like this. I tried to move to it because of the huge speed advantage over firefox but gave up pretty quickly. I understand - and generally approve - the drive towards simplicity, but things should be made as simple as possible, and no simpler. (With apologies to Einstein.) Chrome crosses that line in several ways which ends up causing more confusion than it solves.
- redthrowaway 16y ago
  I've found the simplicity of Chrome to be great. Mainly I'm bugged by how it just doesn't quite get flash right on OSX. Still, I've tried the latest FF and Opera releases, and I can't get over the extra toolbar real estate. Chrome just plain looks nicer, which, in combination with its other great features, solidifies its spot as my browser of choice. I just wish they'd fix the bugs.
- sid0 16y ago
  Agreed. Have you tried the latest nightlies of Firefox 4? They're much closer to what I imagined browsing would be like today, ten years ago, than Chrome is (Firefox has new features like Panorama, and is as hackable as ever. Chrome has... the http prefix hidden? Seriously?). Ever since I started using Panorama, my productivity has shot through the roof.
- sigzero 16y ago
  Chrome on the Mac is still lacking. In Safari if I right click and copy an image I can paste that directly in Mail. Using Chrome I get a broken image icon.
bambax 16y ago
Not directly related (but not totally unrelated), try to search for "PDF/A" in the "ominous bar" => error. What you're looking for: http://en.wikipedia.org/wiki/PDF/A http://en.wikipedia.org/wiki/PDF/A What you get: This webpage is not available. The webpage at http://pdf/a http://pdf/a might be temporarily down or it may have moved permanently to a new web address. The reason for this, is probably that anything with a slash in it has to be interpreted as a url, since there is no protocol left to make the difference between a web address and a search...
- redthrowaway 16y ago
  That seems like a pretty easy fix. Just ensure that you have a .something somewhere in the address, before the first slash.
  - drdaeman 16y ago
    Sorry, but at least http://localhost/ http://localhost/ disagrees. Also, remember about hosts file and NetBIOS names.
  - bambax 16y ago
    Maybe my post wasn't completely clear: I want to search for PDF/A, and actually the expected result is the Google result page for this search, ie: http://www.google.fr/search?q=pdf/a http://www.google.fr/search?q=pdf/a What I get instead is an error because Chrome tries to access the non-existent web page http://PDF/a http://PDF/a
    - redthrowaway 16y ago
      Sorry, no, it was my post that wasn't clear. That should be easy behaviour for Google to code for. The current implementation seems buggy; at least the fix wouldn't be difficult.
- ximeng 16y ago
  Add a question mark or use Ctrl+K for a search. Both the below work. Ctrl+K pdf/a Enter Ctrl+L ?pdf/a Enter
shrikant 16y ago
Why this sudden outrage? This has been the case ever since Google Chrome took the step of obscuring the HTTP protocol scheme way back in April 2010, and been mentioned way back then as well. See: http://code.google.com/p/chromium/issues/detail?id=41467#c19 http://code.google.com/p/chromium/issues/detail?id=41467#c19 for a developer's comment where he explains the 'feature'. This seems to be a case of reporting an existing 'feature' (however much you may dislike it) as a bug.
- augustl 16y ago
  Because from what I've seen, the Chrome devs appreciate concrete use cases. I've seen little of those, and many "plz fix it suxx!!!" posts, so I thought I'd contribute some, well, concrete use cases.
al_james 16y ago
If you have a website that respondes to HTTPS traffic and not HTTP traffic, and you are targeting everyday Internet users, you don't deserve traffic! Most people dont understand the difference between the cryptic labels HTTP and HTTPS. Why should they? Why force users to become experts in Internet protocols to use the Internet?
- augustl 16y ago
  I'm not targeting everyday internet users. As I mention in the ticket, it's our internal servers. Hiding the label makes sense, but the way it's solved in Google Chrome has some trade-offs that some (which according to the number of tickets on the issue is more than just a few) people may not be willing to live with. https://addons.mozilla.org/en-US/firefox/addon/4014/ https://addons.mozilla.org/en-US/firefox/addon/4014/ - sirn mentioned this Firefox addon in a comment, looks like a good solution. I'd also say the iOS-like animation I propose in the ticket makes more sense.
sirn 16y ago
I don't understand why Chrome don't implement something similar to LocationBar2[1]. It expands to include the protocol when URL bar is focused, and hides the protocol on unfocus. It also tries to make sure text width never changes when protocol is hidden by center-aligned the domain part, e.g. URL: http://news.ycombinator.com/item?id=1829087 When protocol is hidden, will be displayed as: URL: news.ycombinator.com /item?id=1829087 [1]: https://addons.mozilla.org/en-US/firefox/addon/4014/ https://addons.mozilla.org/en-US/firefox/addon/4014/
- citricsquid 16y ago
  The fact that they don't lends weight to the spdy conspiracy theory (can it be called that that?)
  - al_james 16y ago
    I think its more likely that they realise most people don't understand what the http/https bit is all about, so its much simpler if you ignore it altogether. I honestly think http/https should beside the point. Websites should transparently redirect to https if they need to (so the padlock appears), but there is no need to signal this to the user with some cryptic techno-gabble.
    - drivingmenuts 16y ago
      Shoulda, coulda, woulda ... The web largely of sites that could do things better, but don't. We have to work around that, because you and everyone else here knows in the deepest pits of our little black hearts that these sites won't be fixed.
      - al_james 16y ago
        I should imagine google, having access to a 'quite' large web index, would know if hiding the http/https would break the Internet for the majority of web users. I am guessing this actually only happens in a very few edge cases, and those edge cases should fix their sites, not hold back development for everybody else!
yock 16y ago
I want to know why your server is listening on port 80 only to kick out 404 pages. If you require SSL for all traffic, then you really only have two logical choices, either only listen on 443 or redirect from 80 to 443. Now, if you only use https for authentication, not redirecting to the SSL-enabled url is a far, far greater transgression against UX than hiding the protocol in the address bar. Simply put, this seems like a nitpick rarely exposed outside of corner cases and other UX mistakes.
- jancona 16y ago
  Saying that the solution is for server admins to do something misses the point. Even if you're right, as a user I have no way to get a random server's config changed. Therefore I want the option to configure my client so that I don't run into the issue.
  - abraham 16y ago
    If your client does not do what you want use a different client. Or you can run Chromium and add whatever features you want.
stfp 16y ago
I find it kind of silly to be hiding protocols in URLs, while, at the same time, sites are adding the ugly-ass #! shebang ajax crawlability thing[1], right in the middle of URLS - a scheme brought to you by ... google :p [1] see http://code.google.com/web/ajaxcrawling/docs/getting-started.html http://code.google.com/web/ajaxcrawling/docs/getting-started...
plaes 16y ago
Hehe.. check out the subtle tabs on the images posted in comment #36.
rwolf 16y ago
It's amazing how many people jumped into the comments section to say "I agree" or "I disagree." Whenever the Chrome devs get up this morning and check their tracker, I imagine they will be very annoyed.
- mkr-hn 16y ago
  I clicked the little star to register support, and now I'm getting every "I agree" post in my inbox. Very annoying.
nlawalker 16y ago
I still haven't seen a good reason for hiding it in the first place.

64 points by augustl 16y ago

konad 16y ago

concur, hiding the http is annoying to me

bruceboughton 16y ago

These are poor counter-examples.

augustl 16y ago
If you're referring to my examples, I'd love to hear some arguments :)
al_james 16y ago
I honestly think the first example is a feature not a problem (its a good thing!) and the second is an edge case that will never happen to the vast majority of web users. Whereas hiding the http/https prefix simplifies the web browsing experience for the vast majority of web users. Worth keeping!

citricsquid 16y ago

I repeatedly posted about this being terrible on reddit, I was repeatedly told to stop bitching. It's a pain, surely this is something that you don't do, the user highlights "foo.com" and is instead given "http://foo.com http://foo.com. It's as bad as the sites that use Javascript to inject "read more @ http://bar.com.. http://bar.com... It wouldn't be so bad if it was optional, but don't expect that of Chrome, oh no. It also leads to an inconsistent experience, I'll quote myself: "If the URL you're using is http://website.com http://website.com and you copy website.com, you get http://website.com http://website.com, but if the URL you're using is https://website.com https://website.com and you copy website.com, you get website.com. It not only messes with copying by changing what you copied into what it thinks you want, it's inconsistent with that guessing."

augustl 16y ago
Let's hope my issue helps. I wrote it as to-the-point as I could, with no bitching and hopefully well reasoned arguments. They're obviously tired of people claiming suckage without explaining why they think it sucks, and probably tired of even hearing about the issue in general. They've merged LOTS of dupes into http://code.google.com/p/chromium/issues/detail?id=41467 http://code.google.com/p/chromium/issues/detail?id=41467
Devilboy 16y ago
It's as bad as the sites that use Javascript to inject read more @ bar.com... No it isn't don't be ridiculous
- citricsquid 16y ago
  uh, yes it is. It causes the same level of annoyance and has the same solution. Load up Chrome and go to http://hackerne.ws http://hackerne.ws, copy "hackerne.ws" from the address bar (all that displays) and you clipboard contains "http://hackerne.ws http://hackerne.ws, to only get "hackerne.ws" in your clipboard you must paste somewhere else and remove the http:// http:// Load up dailymail.co.uk and select an article, highlight a portion and copy it. Now your clipboard contains your selection + "Read more at...", now you must paste it somewhere else and remove "Read more at..." if you want your selection. Both cases can be argued it's "for the user": The Chrome example, "oh well obviously the user wants to have http:// http:// there!" and in the dailymail example "Well obviously if they're sharing that part of the article a link to the full article would be good!". Both examples cause the same problem and have the solution, therefore they are comparable.
  - St-Clock 16y ago
    I think he was being ironic (look at how it was quoted: @bar.com and not @http://bar.com http://bar.com)
  - sausagefeet 16y ago
    > Load up dailymail.co.uk There was your first mistake

paol 16y ago

I find Chrome full of little annoyances like this. I tried to move to it because of the huge speed advantage over firefox but gave up pretty quickly. I understand - and generally approve - the drive towards simplicity, but things should be made as simple as possible, and no simpler. (With apologies to Einstein.) Chrome crosses that line in several ways which ends up causing more confusion than it solves.

redthrowaway 16y ago
I've found the simplicity of Chrome to be great. Mainly I'm bugged by how it just doesn't quite get flash right on OSX. Still, I've tried the latest FF and Opera releases, and I can't get over the extra toolbar real estate. Chrome just plain looks nicer, which, in combination with its other great features, solidifies its spot as my browser of choice. I just wish they'd fix the bugs.
sid0 16y ago
Agreed. Have you tried the latest nightlies of Firefox 4? They're much closer to what I imagined browsing would be like today, ten years ago, than Chrome is (Firefox has new features like Panorama, and is as hackable as ever. Chrome has... the http prefix hidden? Seriously?). Ever since I started using Panorama, my productivity has shot through the roof.
sigzero 16y ago
Chrome on the Mac is still lacking. In Safari if I right click and copy an image I can paste that directly in Mail. Using Chrome I get a broken image icon.

bambax 16y ago

Not directly related (but not totally unrelated), try to search for "PDF/A" in the "ominous bar" => error. What you're looking for: http://en.wikipedia.org/wiki/PDF/A http://en.wikipedia.org/wiki/PDF/A What you get: This webpage is not available. The webpage at http://pdf/a http://pdf/a might be temporarily down or it may have moved permanently to a new web address. The reason for this, is probably that anything with a slash in it has to be interpreted as a url, since there is no protocol left to make the difference between a web address and a search...

redthrowaway 16y ago
That seems like a pretty easy fix. Just ensure that you have a .something somewhere in the address, before the first slash.
- drdaeman 16y ago
  Sorry, but at least http://localhost/ http://localhost/ disagrees. Also, remember about hosts file and NetBIOS names.
- bambax 16y ago
  Maybe my post wasn't completely clear: I want to search for PDF/A, and actually the expected result is the Google result page for this search, ie: http://www.google.fr/search?q=pdf/a http://www.google.fr/search?q=pdf/a What I get instead is an error because Chrome tries to access the non-existent web page http://PDF/a http://PDF/a
  - redthrowaway 16y ago
    Sorry, no, it was my post that wasn't clear. That should be easy behaviour for Google to code for. The current implementation seems buggy; at least the fix wouldn't be difficult.
ximeng 16y ago
Add a question mark or use Ctrl+K for a search. Both the below work. Ctrl+K pdf/a Enter Ctrl+L ?pdf/a Enter

shrikant 16y ago

Why this sudden outrage? This has been the case ever since Google Chrome took the step of obscuring the HTTP protocol scheme way back in April 2010, and been mentioned way back then as well. See: http://code.google.com/p/chromium/issues/detail?id=41467#c19 http://code.google.com/p/chromium/issues/detail?id=41467#c19 for a developer's comment where he explains the 'feature'. This seems to be a case of reporting an existing 'feature' (however much you may dislike it) as a bug.

augustl 16y ago
Because from what I've seen, the Chrome devs appreciate concrete use cases. I've seen little of those, and many "plz fix it suxx!!!" posts, so I thought I'd contribute some, well, concrete use cases.

al_james 16y ago

If you have a website that respondes to HTTPS traffic and not HTTP traffic, and you are targeting everyday Internet users, you don't deserve traffic! Most people dont understand the difference between the cryptic labels HTTP and HTTPS. Why should they? Why force users to become experts in Internet protocols to use the Internet?

augustl 16y ago
I'm not targeting everyday internet users. As I mention in the ticket, it's our internal servers. Hiding the label makes sense, but the way it's solved in Google Chrome has some trade-offs that some (which according to the number of tickets on the issue is more than just a few) people may not be willing to live with. https://addons.mozilla.org/en-US/firefox/addon/4014/ https://addons.mozilla.org/en-US/firefox/addon/4014/ - sirn mentioned this Firefox addon in a comment, looks like a good solution. I'd also say the iOS-like animation I propose in the ticket makes more sense.

sirn 16y ago

I don't understand why Chrome don't implement something similar to LocationBar2[1]. It expands to include the protocol when URL bar is focused, and hides the protocol on unfocus. It also tries to make sure text width never changes when protocol is hidden by center-aligned the domain part, e.g. URL: http://news.ycombinator.com/item?id=1829087 When protocol is hidden, will be displayed as: URL: news.ycombinator.com /item?id=1829087 [1]: https://addons.mozilla.org/en-US/firefox/addon/4014/ https://addons.mozilla.org/en-US/firefox/addon/4014/

citricsquid 16y ago
The fact that they don't lends weight to the spdy conspiracy theory (can it be called that that?)
- al_james 16y ago
  I think its more likely that they realise most people don't understand what the http/https bit is all about, so its much simpler if you ignore it altogether. I honestly think http/https should beside the point. Websites should transparently redirect to https if they need to (so the padlock appears), but there is no need to signal this to the user with some cryptic techno-gabble.
  - drivingmenuts 16y ago
    Shoulda, coulda, woulda ... The web largely of sites that could do things better, but don't. We have to work around that, because you and everyone else here knows in the deepest pits of our little black hearts that these sites won't be fixed.
    - al_james 16y ago
      I should imagine google, having access to a 'quite' large web index, would know if hiding the http/https would break the Internet for the majority of web users. I am guessing this actually only happens in a very few edge cases, and those edge cases should fix their sites, not hold back development for everybody else!

yock 16y ago

I want to know why your server is listening on port 80 only to kick out 404 pages. If you require SSL for all traffic, then you really only have two logical choices, either only listen on 443 or redirect from 80 to 443. Now, if you only use https for authentication, not redirecting to the SSL-enabled url is a far, far greater transgression against UX than hiding the protocol in the address bar. Simply put, this seems like a nitpick rarely exposed outside of corner cases and other UX mistakes.

jancona 16y ago
Saying that the solution is for server admins to do something misses the point. Even if you're right, as a user I have no way to get a random server's config changed. Therefore I want the option to configure my client so that I don't run into the issue.
- abraham 16y ago
  If your client does not do what you want use a different client. Or you can run Chromium and add whatever features you want.

stfp 16y ago

I find it kind of silly to be hiding protocols in URLs, while, at the same time, sites are adding the ugly-ass #! shebang ajax crawlability thing[1], right in the middle of URLS - a scheme brought to you by ... google :p [1] see http://code.google.com/web/ajaxcrawling/docs/getting-started.html http://code.google.com/web/ajaxcrawling/docs/getting-started...

plaes 16y ago

Hehe.. check out the subtle tabs on the images posted in comment #36.

rwolf 16y ago

It's amazing how many people jumped into the comments section to say "I agree" or "I disagree." Whenever the Chrome devs get up this morning and check their tracker, I imagine they will be very annoyed.

mkr-hn 16y ago
I clicked the little star to register support, and now I'm getting every "I agree" post in my inbox. Very annoying.

nlawalker 16y ago

I still haven't seen a good reason for hiding it in the first place.