Truer words...
“All too often, technologists solve problems by introducing additional layers of technology abstractions and disregarding simpler solutions, such as outreach and engagement,” he wrote.
The Google-sponsored Chromium project has cleaned up its act, and the result is a marked decline in queries to DNS root servers. As The Register reported in August 2020, Chromium-based browsers generate a lot of DNS traffic as they try to determine if input into their omnibox is a domain name or a search query. Verisign …
Ah, but it is all about revenue generation, by integrating the search and address bar you will automatically direct a lot of people to where you want them to go (based on money) rather than where they want to go.
Most will simply click on the top result on the page without looking if the expected page does not appear.
So google merge their search and address fields into an "omnibox" and it makes billions of unnecessary DNS requests. Now they've calmed it down a bit, that's an example of how great they are at engaging with "the community"?
Shouldn't they have done the community engagement thing before merging the fields?
"I like the merged box. If nothing else it's simpler to explain to the mundanes."
Maybe I'm missing something, but "type address in the box marked 'address', and search in the box marked 'search'" doesn't sound all that complicated, and isn't something I've ever had trouble explaining to anyone. It is, after all, how things worked for quite a while, and somehow millions of ordinary people were perfectly capable of using the internet. Make no mistake, this was not a change made to simplify things for users, it was solely intended to drive more traffic to Google's search, and hence advertising, engine.
He has a point though: Oversimplifying stuff, so the Great Unwashed only do what they are supposed to do (consume), is the trend. So an unique entry box where the subjects enter all their input and then let Google decide what to do with it sounds just right: "Don't you worry your pretty little heads, uncle Google knows best what you (should) want".
Well, maybe there are people who would benefit from a combined oven and toilet, and let the device figure out what to do with what has been put into it. Like Chrome, it could start by assuming that it's food and cook whatever was deposited in it Then if it gets an adverse reaction, it knows that that was the wrong operation and it can flush it down the sewer instead.
Most people however don't mind having to choose whether to use a toilet or an oven, and get it wrong about as often as they do in choosing whether to click on an address box or a search box before typing.
"Well, maybe there are people who would benefit from a combined oven and toilet, ..."
Those are *exactly* the same people who would benefit from the omnibox.
I like to check the box in settings for separate search field, but some browsers don't make it easy. The thing I find especially maddening is when I type https:// in an omnibox and still get a search!
Maybe I'm missing something, but "type address in the box marked 'address', and search in the box marked 'search'" doesn't sound all that complicated, and isn't something I've ever had trouble explaining to anyone.
"So you definitely typed it in the address box?"
"What kind of moron do you take me for?! I know the different between the address box and the search box by now, and it's still not working."
"Send me a screenshot."
*looks at screenshot*
"You've typed it in the search box."
Users have a goal: they want to access a resource. DNS addresses, like IP addresses, have become behind the scene gubbins that they don't really care about.
"DNS addresses, like IP addresses, have become behind the scene gubbins that they don't really care about."
But, um, WHO is driving that dumbing-down of the users of our systems? It ain't those of us who are running and maintaining those systems. It's those who stand to profit from having even the most ignorant hick still able to consume their goods and services. Just because they can, doesn't mean they should.
"One of my users has been doing just that since 1996 (except it was excite)..."
It's very, very common, particularly among domestic and home micro business users. I show them, they say "Ohh, right" but next time I see them they're still typing the URL into Google and clicking the link. Eventually, I just give up and let them get on with it.
It certainly doesn't sound complicated, but I don't think Chrome has ever labelled the address bar. I find that people completely ignore the address bar and, having finally managed to open Chrome, type the internal hostname into the search box on the start page.
If they get completely stuck, I'll offer to help with Quick Assist. Generally that goes along the lines of:
"Click the start button"
"What's that?"
"The Windows icon at the bottom left"
"Where?"
"The button at very bottom left of the screen, it used to say Start, but now it's just a Windows logo"
"Ok, it says Apps and Features, Power Options..."
"No, left click"
"On what?"
"The start button"
"I can't see it, that list is in the way"
"Just click anywhere to close the menu"
"OK, it's opened Device Manager"
"OK, now left click on the start button and type Q"
"Where do I type it?"
"Nowhere, just press Q on the keyboard after you've left clicked on the start button"
"Oh, yes I see"
"Now click Quick Assist"
"Now it says Run as administator, Open file location...."
Every time, guaranteed.
I like the merged box
And I absolutely really REALLY [insert string of expletives here] hate it.
I'm really peed off with typing a valid URL into the box, only to find it's done a search (and needless to say, not found the internal resource I'm after). I'm also peed off with typing a search that the browser decides looks like a URL (yes, they do exist) and then tells me that the URL can't be found.
If you think teaching users to use a search box for search and an address box for addresses is hard - try teaching them the intricacies of whether they should type https or http (yes there are still legitimate sites not using the "s"), and yes they do have to put the "://" exactly like that ...
I have this sudden urge to hack bash so that if you enter a URL it starts a browser instance with it, and for any failed path search it starts a browser instance with the command line as the search query.
Address bar + search box + bash.
I mean, I wouldn't use it. But I'd enjoy inflicting it on my enemies.
Of course it's entirely likely someone has already done this. (I mean, people create readline plugins for filename completion that do network searches, so you know there's someone to add any daft thing to command-line processing.) I'd look but it would just depress me.
Not bash, but some (many?) terminals do recognise hyperlinks (and local file paths) and you can click / interact with them.
If you want to open a URL directly from the command line, you're probably after kde-open. It will do its best to recognise the URL and how to deal with it (e.g., it launches Dolphin for a file:// URL pointing to a directory, Kwrite for one pointing to a text file, Kaidan or similar for a xmpp:// URL and so on).
WHAT? You mean design the thing and look for possible inefficiencies and cockups before building it? It'll never catch on.
Simpler to just chuck together a prototype, lob that into beta, fix the serious user gripes and chuck hardware at any performance issues.
Costs so much less and makes the development process more......erm......agile. Yes. That's a good word, agile. Makes it sound like a method rather than just a massive fuckup looking for a place to happen.
This post has been deleted by its author
The modern software world.
Rapid software release policies where we in perpetual beta and constant updating mode.
I started using office365 two weeks ago, was previously using the now obsolete office 2010.
Within 30 minutes I discovered 6 bugs (enough that I started listing them to prep a bug report), 3 of them were moderately serious.
1 - Renaming custom ribbon items didnt work.
2 - Creating rules randomly crashed outlook.
3 - Accessing addin options crashed outlook.
I then noticed was an option to change update channel from current to a two monthly one, I switched to it and the bugs I listed above all fixed. The default channel seems akin to Windows 10 where the users are all beta testing.
If I'm searching for a term, say "widget", it's probably trying to find widget.com, widget.org, widget.co.uk, widget.net and a whole bunch of other domains. Yes, those queries will go to my "local" DNS server (which could be a corporate one or my ISP). Unless they've cached that information already, they need to look to the root servers to find the answer, hence it hits the root servers eventually. They should cache that for future requests by the next person searching for those domains reducing future traffic, though.
"
They should cache that for future requests by the next person searching for those domains reducing future traffic, though.
"
Not sure that caching non-existent domains would be a good idea. There are going to be so many different search terms that the cache will get very big very fast. And probably still not reduce traffic by a great deal, because with a few exceptions, not many people will be entering the same search terms within the cache entry expiry period.
Caching non-existent domains is a very very good idea. So much so it's been part of the DNS protcool for nearly 25 years. RFC2308 was published in 1998.
DNS caches might get big very fast, but not very big. They usually stabilise in a few hours and stop growing after that because the rate of cache entries tends to match the rate of cache expirations. Even the biggest DNS caches at a huge ISP are unlikely to fill more than 1-2 GB. What does that much RAM cost these days, $50?
It's also a very good thing that the zillions of DNS lookups for fuckup.whatever in your network hit the cache of your local resolver instead of going to the root.
Bingo! Is the send every fucking letter to google one at a time in packets of 1k...its the ultimate amplification attack, so they can send you back matches for each keystroke, lookup up the resultant cacaugphony (sp?) every time. This plus their 'generate random string and query it to determine if dns is being hijacked'
The only reason they fixed it is because, today, even Google employees are working from home and no longer have 1G-100G sub-millisecond connections to their servers. I would love to have been able to listen in on the bitch sessions for every time someone said "but it worked just fine on friday in the office, not its not responding at all"... serves you right, bastards.
[...] and no longer have 1G-100G sub-millisecond connections to their servers [..]"
A customer made postcode look-ups a central server function - with client iterative requests doing the licensed database accesses. Worked ok on the lan in head office. Helpline staff on a 2mbps connection complained that they could go and make a cup of tea while their PC found the caller's address.
It's not that simple. If my local DNS doesn't know the answer, it doesn't just go straight to root. It goes to another level (often the ISP DNS). If that doesn't know the answer, it will look elsewhere, eventually getting to root. Any organisation that goes straight to root is being very rude.
When you do a DNS lookup on the local DNS server, and the query is for a domain that the local server does not have cached (which will usually be the case when the query is in fact a search term and not a domain at all), then the local DNS server has to forward the query to the root server. So you are impacting on the traffic to both the local DNS server and the root server.
If the user starts an entry with http(|s):// or even www. or ends an entry with .blah, it gets treated as a URL and triggers a DNS search.
If not, the query goes to the search engine of choice.
One might reasonably expect that Google actually have a list of URLs that could be tested for before troubling the DNS servers, no? But that would be using Google's hardware rather than the commons, and perhaps they don't like that.
> If the user starts an entry with http(|s):// or even www. or ends an entry with .blah, it gets treated as a URL and triggers a DNS search.
>
> If not, the query goes to the search engine of choice.
But then the user has to enter a scheme and/or www, and it won't work for other subdomains.
I've a simpler solution: have one box for addresses, and another for searches.
You do know that FF comes with the freaking awesomebar enabled and no search box?
You have to go to options/customize to enable the search box. Then you have to go to about:config and disable a bunch of crap in order to get rid of the monster.
Now tell me how many ordinary people even bother to customize their browser. Yeap, that's what I thought.
Any ideas what you need to disable in about:config to switch it all off in Firefox 85? I accidentally "upgraded" from 84 and it all reappeared. Despite disabling everything again, I still have an extra line which appears below the address bar offering to send the string I'm entering to a search engine with one simple click. No, allowing me to choose which search engine it goes to isn't sufficient, I want to switch it off. And the address bar gets slightly bigger when I type into it for no reason that I can fathom.
But I did manage to re-enable the search box (which, until you type into it, still has "Search" written in grey to let you know what it does) at least.
Firefox can be hard to like sometimes.
Thanks for taking the time to post this link, and I'll check it out later when I'm at home to see if it helps. I do have to mention that the link points to a two-year-old page, so won't determine what changed between version 84 and 85 though.
Again, thanks for the help.
>I've a simpler solution: have one box for addresses, and another for searches.
Whilst that would mean no DNS lookup on query searches, it doesn't solve the partial address lookup problem as the browser searches for autocomplete suggestions.
It would seem that until you have a 'full' URL ie. user has hit return, either none of the autocomplete lookups should be going beyond the local DNS server's cache or they get treated as a query search with the search engine just returning a set of URL's. That way 'www.b' would most probably result (in the UK) in only the URL www.bbc.co.uk being offered.
If the user starts an entry with http(|s):// or even www. or ends an entry with .blah, it gets treated as a URL and triggers a DNS search.
It also checks for a trailing /
Browsing to 192.168.1.n gets got a Google Search.
Browsing to 192.168.1.n/ will actually allow you to access local resources.
You can also go and type the whole http:// bit as well for the same effect, but a trailing slash is quicker (there's even one on the numpad!).
EDIT: Just checked and actually, the trailing slash is no longer necessary. Looks like some bright spark thought it'd be a good idea for Chromium to check whether the entered string looked like an IP address before doing a search. And not before time...
They also used to have real menus, but not sandwiches.
And they used to offer to do helpful things like clear your download history every time you close the browser, WITHOUT having to nuke your entire browsing history at the same time.
It's like things used to be separate in the bad old days, but streamlining and collapsing have made things soooo much better now. We have such a bright future ahead of us.
Considering the UK and other countries use the DNS lookups to block access to illegal sites then you could run the risk that a innocent search term you start to type might actually send a DNS query for an illegal website.
Think typing a search for 'dog sex in a Labrador pup' would at some point send a DNS look up to your ISP for dogsex domain which could actually exist and be on the banned list