Talk:User agent

From Wikipedia, the free encyclopedia
Jump to: navigation, search
WikiProject Computing (Rated C-class, Mid-importance)
WikiProject icon This article is within the scope of WikiProject Computing, a collaborative effort to improve the coverage of computers, computing, and information technology on Wikipedia. If you would like to participate, please visit the project page, where you can join the discussion and see a list of open tasks.
C-Class article C  This article has been rated as C-Class on the project's quality scale.
 Mid  This article has been rated as Mid-importance on the project's importance scale.
 

Bad English[edit]

The sentence "Spam bots and Web scrapers often use fake user agents. For example, the Android browser identifies itself as Safari in order to aid compatibility[4]." implies that the Android browser is a 'Spam bot' or 'Web scraper.' The phrase should be changed or moved to make it more clear to the reader, and to eliminate any unintended implications. 71.206.27.210 (talk) 22:42, 13 August 2011 (UTC)

Safari implementation[edit]

Small point:

"...Safari implemented systems whereby the user could select a false User-Agent string to send"

hmmm..

open Terminal defaults write com.apple.Safari IncludeDebugMenu 1 then select from the debug menu.

It's not an option available by default or one Apple actively puts forward to consumers, it's a 'debugging' option, so although what you say is correct it is misleadingly weighted in this context.


An example of a Macintosh browser that offers/invites the user to change the User-Agent string through regular (non-debug) preferences would be iCab, and which as far as I can recall from testing on my own website logs doesn't append iCab to the end of the string either.

Opera 9 Preview 2 - I don't believe the user agent string has changed, it still reads Opera 9, etc. Someone want to confirm and add this? Preview 2 was released on Feb 7, 2006

IE7[edit]

In case anyone is wondering, I got the IE7.0b UA from this MSDN blog Aidan 08:14, Apr 28, 2005 (UTC)


Revision format in Gecko browsers[edit]

Can someone confirm the "rv - 1.7.8" format in Gecko browsers? I haven't seen these in the wild, only "rv:1.7.8"

Sorry, that's a mistake, in a search/replace I did, when reformatting the lists. I intendended the "—" to appear *before* the agent string, and did not intend to change the agent string. I beleive it is fixed. Sorry, I'll be more careful in the future. --rob 05:17, 18 September 2005 (UTC)

Stats[edit]

Most statistics software are able to recognize Opera and count it as such. use of Firefox or IE is not exaggerated because of UA spoofing.

Yes, Opera's spoofed user agent strings are always easily detectable as coming from Opera. However, Firefox extensions such as User Agent Switcher are able to change Firefox's user agent string so it can't be distinguished from the legitimate IE user agent string. I think it would be safe to say IE usage is probably overestimated, and usage of all other browsers may be underestimated. -- Schapel 03:10, 30 November 2005 (UTC)

There is a new phenomena that I've run into - the number of Linux OS reported is going down while the unknown is going up rapidly. I don't know what it means - is it ubuntu?

I also read about a IP fingerprinting system that detected which OS it was based on other details.

Moozilla userAgent is appearing in webserver logs[edit]

What's the deal with this ? I've looked up Moozilla ( http://www.moo.ca/moozilla ) and it just says it's a Moo client (whatever that is). The stats for the page there show only 45,000 hits but somehow it managed to log 900,000 hits to our website (11% total browser share). I've seen similar large % stats on other sites.

There's something fishy going on with Moozilla and I don't think it's just that obscure browser.

Anyone have any ideas why Moozilla is showing up in such large volume?

Spoofing in Firefox[edit]

I would debate that Firefox supports useragent spoofing. I don't think that manually editing config files is normal configuration by users. This greatly contrasts browsers such as Opera, which have menu items for user-agent strings. --BarkerJr 21:34, 16 May 2006 (UTC)

You don't even need to edit a config file. Just go to about:config and add a value for general.useragent.override. The way for normal users is to install an extension that adds this value automatically. The only reason Firefox extensions can change the user agent string is because Firefox has "implemented a system whereby the user could select a false User-Agent string to send," exactly as the article states. -- Schapel 22:33, 16 May 2006 (UTC)

Opera User-agent Masking[edit]

I'm not sure where would be best to put it, but the article should mention somewhere that Opera 9 can 'mask' itself as IE/Mozilla which doesn't have any mention of Opera in the user-agent string, as opposed to "others – e.g. Opera – duplicate the User-Agent string but add the genuine browser name to the end." from the article.

eg when Opera 'identifies' itself as IE, it has the user agent

Mozilla/4.0 (compatible; MSIE 6.0; X11; Linux i686; en) Opera 9.02

but when Opera 'masks' itself as IE, it has the user agent

Mozilla/4.0 (compatible; MSIE 6.0; X11; Linux i686; en)

BigBrownChunx 23:52, 1 November 2006 (UTC)

long list fire fox and opera[edit]

I fail to see the point in having such a long list for firefox and opera. But can't be bothered deciding which ones to delete. —The preceding unsigned comment was added by 81.232.51.243 (talk) 22:43, 14 January 2007 (UTC).

I removed many Firefox and Opera UA strings, mostly where there were several UA strings for the same version, and old beta and release candidate versions. I tried to keep variety in hardware and operating systems. Feel free to put one or two back if I did remove some important variety. -- Schapel 04:14, 16 January 2007 (UTC)

Lists[edit]

Something needs to be done about these lists. It's unencyclopedic to have "non-list articles that are overrun with laundry lists. As a rule of thumb, if more than about 30% of a non-list article consists of a laundry list, it may be a problem." In this article, it's a problem. (Narkstraws 22:49, 22 February 2007 (UTC))

OK, would a separate article with just a list of user agent strings be appropriate? I was extremely disappointed to come here today and find that great resource (the most excellent list of strings) GONE. It was the best resource out there. --angrykeyboarder (a/k/a:Scott) 00:55, 19 May 2007 (UTC)
They are incredibly useful for testing web links, though. Perhaps much of the list could be moved into Wikipedia namespace. –EdC 23:11, 23 March 2007 (UTC)
I think the huge lists should not be included in this article, especially under a section labeled "Example user-agent strings." IMO, this section would be better off showing just a few select strings to get the general idea of a user agent. However, I must agree that this list is extremely useful, and I'd hate to see it disappear. Rktur 01:33, 5 April 2007 (UTC)
I can understand why the list was removed fom this article (as previously discussed), but that list was incredibly useful. Does it still exist anyplace in Wikipedia?
This great list has now been deleted and replaced with an uglier much less useful list. As far as i can tell there is no suitable alternative on the internet. Great... - meconomy 5/1/2007
There's an regularly updated list at user-agents.org, so I put a link to that in the article. That's the best list I've been able to find; the last update was two days ago, and there's an organized way to update it. I also reorganized the external links into three sections, and took out some of the less relevant links. We still have too many links to "click here to see your user-agent" sites, though. --John Nagle 02:22, 19 May 2007 (UTC)
After searching that site, I found far fewer Firefox User-Agents than had been listed here (thankfully, those us us in the know, can check history - at least that's still around).
I did not see how any visitor could update the list on that site either. --angrykeyboarder (a/k/a:Scott) 11:51, 17 June 2007 (UTC)
If you need the list so badly, why don't you copy it from the history: [1] and then put it up on the web somewhere. Also I didn't remove the list, I just think it doesn't belong. (Narkstraws 19:32, 25 June 2007 (UTC))
Yes, perhaps on a blog or a wiki or something... —Ryan (talk) 10:59, 10 February 2008 (UTC)
I'd like to see a small list, say Opera, FireFox, IE and Safari. 203.129.33.32 (talk) 01:34, 27 August 2009 (UTC)

This article becomes less useful with each edit[edit]

I just don't get it. --angrykeyboarder (a/k/a:Scott) 11:51, 17 June 2007 (UTC)

arbitrary string[edit]

What happens if you just set your user agent to be a message like "Hello World"? —The preceding unsigned comment was added by 207.189.98.44 (talk) 23:07, 2 March 2007 (UTC).

If you do that you will get blocked by sites that use automated software to block bad web robots. Many blogs may also block you because you will look like a spam bot.(4.231.129.11 (talk) 00:36, 23 March 2010 (UTC))

Web servers won't know which browser you are actually using and may not send you the version of the site that you would otherwise see. Stats for that site will also be off. For example, it appears to cause Google Maps not to show you satellite photos. Jason McHuff 21:45, 13 March 2007 (UTC)
A cute user agent bug: if the site has a Coyote Point load balancer, and you send a USER-AGENT string ending in "m" but with no earlier "m" in the string, and the USER-AGENT field is the last field in the HTTP header, it drops the packet. Even fails on Coyote Point's own site. The thing parses HTTP headers with regular expressions, and one of the built-in regular expressions seems to be broken. I'll bet someone wrote "\m" where they wanted "\n". Took me two days and a Python program that tried a huge range of USER-AGENT strings to diagnose that problem. --John Nagle 21:01, 25 June 2007 (UTC)
I think I just found the different-pages-for-different-user agents that takes the cake--if you go to a page on http://www.ngm.com/ with a blank user agent, you'll get sent to Network Solutions' site. I was really thinking that the ngm.com domain registration had lapsed. Oh, and another example would be that Java will not work right in Mozilla Firefox without the correct UA, an issue that has been around for YEARS (SINCE 2001!) Jason McHuff 08:26, 15 September 2007 (UTC)

bad JavaScript[edit]

This article contains the sentence:

"However, out-dated JavaScript, which effectively locks out browsers other than Explorer or Navigator, is still in use - especially on smaller, non-corporate, websites."

I totally agree with the first part of this sentence, regarding JavaScript, but it ought to be sourced. However, the part about "especially on smaller, non-corporate, websites" is totally POV, and I intend to remove it (based on my experience that it is actually larger websites that malfunction on non-Windows-based computers). 68.49.208.76 03:04, 7 September 2007 (UTC)

It is NPOV, the W3C has declaird Useragent sniffing to be a bad javascript practice. Developers should instead use Browser capabilities testing. I think if a standards organization like W3C says it's a bad practice, it's a bad practice. However, I do conceed that while the content was not NPOV, the phrasing could use a little correction before being reentered into the article. --Robert Wm "Ruedii" (talk) 08:32, 18 April 2013 (UTC)

Google Chrome[edit]

Chrome is taking it to a new level: Mozilla/5.0 (Windows; U; Windows NT 5.1; en-US) AppleWebKit/525.13 (KHTML, like Gecko) Chrome/0.2.149.27 Safari/525.13 —Preceding unsigned comment added by SmilingBoy (talkcontribs) 17:22, 3 September 2008 (UTC)

Edit on 13:51, 20 October 2008[edit]

The export restrictions were "effectively eliminated" in 1996 whereas the first version of Opera was released first in 1997. Sources are http://en.wikipedia.org/wiki/Export_of_cryptography#PC_era and http://www.opera.com/company/about/milestones/ —Preceding unsigned comment added by 78.27.90.194 (talk) 13:55, 20 October 2008 (UTC)

Clarity[edit]

This article isn't very clear about what a user agent is... -- Eraserhead1 <talk> 19:35, 19 January 2010 (UTC)

Can the removal of my clearer introduction text please be justified properly? A user agent and a user agent string are generally used interchangeably... -- Eraserhead1 <talk> 21:56, 19 January 2010 (UTC)
That is clearly sloppy usage, a UA is foremost the application or software component, the article now properly delineates the identification as well. Kbrose (talk) 21:59, 19 January 2010 (UTC)
Fair enough. I've moved the redirect User agent string to point at the User agent identification section. -- Eraserhead1 <talk> 22:03, 19 January 2010 (UTC)

It seems that the majority of this article is about User agent identification in computer protocols. Only the lead section talks about a UA being a software program. The rest (identification, string format, spoofing, sniffing, and encryption sections) are all subtopics of identification, and in these sections the term user agent is actually used to mean the “User-Agent” header field, rather than the software program itself. Perhaps it would be better to turn the article “inside-out” or inverted so the article name is identification, with just one section defining what a user agent program is? Vadmium (talk) 05:12, 2 August 2011 (UTC).

Crypto Strength Token[edit]

This will be removed in Firefox in version 4.0 (see bug 572668).

This is also apparently going to be removed in IE 9, if it hasn't been in IE 8 - maybe it should be noted somewhere. —Preceding unsigned comment added by RyanJones (talkcontribs) 22:55, 2 July 2010 (UTC)

Most of this article is really about the "UserAgent" in HTTP[edit]

Most of this article talks about the piece of software that acts as a client in a HTTP transaction, and how it is identified by the UserAgent header in a HTTP request.

In general, a "User Agent" is a software agent that acts on the behalf of a user, and (usually) with the user present (since otherwise, if there is no user, it would just be an Agent.)

However, in the case of a web browser talking to a web server, the client is the user agent, and the term has blurred in usage, even though many other protocol standards use the terms closer to their original meaning. (E.g., a MUA is a Mail User Agent).

Masinter (talk) 06:52, 16 December 2011 (UTC)

Some xx.wp references[edit]

Martijn Koster, no article, now of historical, see ALIWEB, EI*Net?. PS: what's Bad English? — Preceding unsigned comment added by 177.2.173.189 (talk) 18:59, 22 September 2012 (UTC)

User-Agent Sniffing is considered non-standard according to W3C[edit]

User-Agent Sniffing is considered non-compliant according to w3c standards. Should we not mention this, be it in a NPOV way. --Robert Wm "Ruedii" (talk) 08:33, 18 April 2013 (UTC)

I cannot find this in the W3C standards. Are you referring to this page from the W3C Web Education Community Group Wiki? --Guy Macon (talk) 18:54, 28 November 2013 (UTC)

Security implications[edit]

I have long thought that whilst a website may need to know what browser is in use for the sake of page layout issues, there is little justification for disclosing what OS the browser is running on. Doing so increases the risk of a compromised website being able to take control of the visitor's computer by way of a known vulnerability. The end of support for Windows XP has highlighted this issue, though it's one which has existed for many years. Basically, the Internet is a hostile environment, and the less you tell a potential crook about you, either by UA or scripting, the better. --Anteaus (talk) 08:19, 10 April 2014 (UTC)

User Agent Switcher is your friend. --Guy Macon (talk) 19:44, 10 April 2014 (UTC)
Yes there are several ways to stop the OS being reported. However, the more pertinent question is whether it should be reported without the user's permission, after all it could be seen as an unwarranted disclosure of private data. --Anteaus (talk) 21:23, 10 April 2014 (UTC)

External links modified[edit]

Hello fellow Wikipedians,

I have just added archive links to one external link on User agent. Please take a moment to review my edit. If necessary, add {{cbignore}} after the link to keep me from modifying it. Alternatively, you can add {{nobots|deny=InternetArchiveBot}} to keep me off the page altogether. I made the following changes:

When you have finished reviewing my changes, please set the checked parameter below to true or failed to let others know (documentation at {{Sourcecheck}}).

You may set the |checked=, on this template, to true or failed to let other editors know you reviewed the change. If you find any errors, please use the tools below to fix them or call an editor by setting |needhelp= to your help request.

  • If you have discovered URLs which were erroneously considered dead by the bot, you can report them with this tool.
  • If you found an error with any archives or the URLs themselves, you can fix them with this tool.

If you are unable to use these tools, you may set |needhelp=<your help request> on this template to request help from an experienced user. Please include details about your problem, to help other editors.

Cheers.—cyberbot IITalk to my owner:Online 02:49, 19 March 2016 (UTC)

Anotherbigal (talk) 12:16, 10 June 2016 (UTC)== Make it clearer for non technical user ==

I presume to someone with technical knowledge this article make a lot of sense. The discussion on this 'Talk' page seems to support that. However, what the main page fails to do, at least to me it fails to do, is explain to the non technical 'average' viewer exactly what a 'User Agent' is used for, why someone would want to use one and in what circumstances the use of a 'User Agent' would be beneficial and why. Just a clear, everyday paragraph for the average computer user. — Preceding unsigned comment added by Anotherbigal (talkcontribs) 12:15, 10 June 2016 (UTC)

Good point. I just edited the lead paragraph too try to make it clearer. Please let me know if that helped. --Guy Macon (talk) 13:35, 10 June 2016 (UTC)