Talk:Comparison of web browsers

order

I have edited the discussion page in order to restore the initial order. In order to preserve this order, please use three equal signs (===) to create sub-sections rather than sections. And please add your post at the bottom of a section.
Possible topics are Standards support issues, Features you want mentioned or Browsers you want added/removed. Grey (talk) 03:43, 7 January 2008 (UTC)

Features

IE voice control

Hi! I think internet explorer should have "partial" instead of "no", since, if running windows vista, you surely can control the browser with your voice!213.64.150.116 17:05, 16 April 2007 (UTC)

Addition of encoding support table?

One thing that is a good idea to have is a table on encoding support (Unicode and legacy encodings), which would probably go on a separate page due to size issues. This is useful because support on encodings is very different between browsers (for example, IE is the only browser with ISCII support) still need to test this -- Prince Kassad 18:06, 26 October 2007 (UTC)

If you can find any reliable sources, why not... --Grey 15:44, 27 October 2007 (UTC)

I strongly support this request. Maybe even extending it to a table with 'http' support indicating supported Content-Types (gzip, deflate, etc) and charset (UTF-8, Latin-1, etc). —Preceding unsigned comment added by 84.198.72.247 (talk) 11:48, 30 December 2008 (UTC)

AdBlock

It kinda bugs me that Firefox has "no" under "ad filtering" when everyone uses AdBlock. Should this and some other entries be changed to something like "via plugin"? 75.152.220.162 (talk) 09:42, 21 May 2009 (UTC)

It says so with a note [1] but it would be a good idea to prominently show them different. Platonides (talk) 21:05, 9 June 2009 (UTC)

Why does all the browsers that filters ads with a plugin have no with a cross, except firefox says yes with a cross, shouldn't it be a no as well? 72.218.100.143 (talk) 08:19, 8 January 2010 (UTC)

Epiphany (GNOME Web Browser) should definitely say "yes" for ad filtering under "accessibility features"! It's not the same as with Firefox, because here you certainly HAVE the filtering option "out of the box", installed with the browser itself, no need for downloading any add-ons. OK, you can remove epiphany-extensions package, but there is no GNOME-release coming without that, so it's practically included... Arpadapo (talk) 09:02, 24 February 2010 (UTC)

System Requirements

Hi All. Thanks for the amazing page. I know it would be hard, but easier than some of the other tables - a table of system requirements, hard disk, OP, RAM size especially would easily allow a reader to determine if he could download and run it. In some cases it's hard to find this data, for example I've just been looking at the Opera site and I can't find any of this information. Considering how much RAM Firefox has been known to use in the past (horribly large amounts) and that not everyone has a new computer, this information would be extremely helpful. Just my two cents worth. Thank you. Blucat David. —Preceding unsigned comment added by 198.142.19.27 (talk) 20:05, 29 July 2009 (UTC)

Browsing History

I notice that there is no information about the topic of browsing history. On this page there is an oblique reference to "full text browsing history search", but do all these browsers really keep a basic type of browsing history (i.e. links visited)? This is a current topic, as some browsers have recently implemented modes that do not keep browsing history (e.g. Chrome's Incognito Window and IE's InPrivate Browsing). Is "browsing history" a feature of most browsers, is it notable enough to mention in the Web browser page, and if so, should a comparison of each of the browser history functions be included in a table here? Twocs (talk) 01:49, 31 July 2009 (UTC)

Why no mention of WebSurfer browser?

I'm guessing it was mainly Twenty-Something's who wrote this article?
I remember in the early-to-mid 90's seeing ad banners that said, "WebSurfer 5.0 The BEST browser". I laughed every time I saw it and said, "yeah right". I never clicked on it, never downloaded it. I already had Netscape Navigator 3.0, which was The King back then.
But this WebSurfer browser should be mentioned here. It was indeed one of the also-rans who fought to get noticed in the shadow of Netscape Navigator.

Anyone else remember this? You can still see the ads in The Internet Wayback Machine web site. I was in my early twenties back then and was a Netscape Navigator Fanboy, but today's Firefox Fanboys put me to shame, with their insanely blind romantic love for a piece of software. I'm waiting for California to legalize Fanboy-Firefox marriages, that's how out of control it has gotten. Anyway...WebSurfer 5.0...the best damn browser on the planet. Dropmeoff (talk) 03:15, 17 August 2009 (UTC)

Other missing browsers

There are a number of other browsers missing, it's as if RISC OS never existed. I don't have all the dates so I can't add this stuff myself... however you ought to be able to find information on: ArcWeb (Stuart Brodie, 1996), Webite (Doggysoft, 1995), Webster (Andrew Pullan - written in BASIC!, 1996), Fresco (ANT, 1994), Browse (Acorn, 1996?), Fresco/Voyager (modified version for ArgoNet), Oregano (Oregan Networks, 2000), Oregano 2, NetSurf (yes, it started on RISC OS), Bon Echo (port, Peter Naulls I think), plus the expected port of Lynx.
Some of this stuff is available at ftp://ftp.demon.co.uk/pub/archimedes/www/
Fresco has its own Wiki page as it was deployed in a number of ARM-based (RISC OS internal) STBs.
I believe that Browse was one of the first browsers to fully support PNGs. It is written up well at http://www.vias.org/pngguide/chapter02_02_05.html and as an update, full sources to Browse are available as a part of RISC OS Open.
HeyRick1973 (talk) 19:52, 9 January 2010 (UTC)

Additionally Epiphany is not included here. --Khatores (talk) 18:23, 18 July 2011 (UTC)

Call for academic reviews of the web browsers....

• http://scholar.google.com/scholar?as_q=internet+explorer+google+chrome&num=10&btnG=Search+Scholar&as_epq=&as_oq=&as_eq=&as_occt=title&as_sauthors=&as_publication=&as_ylo=&as_yhi=&as_sdt=1.&as_sdtp=on&as_sdts=5&hl=en

--124.78.210.114 (talk) 13:20, 11 February 2010 (UTC)

• http://scholar.google.com/scholar?hl=en&q=internet+explorer+google+chrome&btnG=Search&as_sdt=2000&as_ylo=&as_vis=0 —Preceding unsigned comment added by 124.78.210.114 (talk) 13:29, 11 February 2010 (UTC)

The references cited do not seem to be from academic journals very much--124.78.210.114 (talk) 13:32, 11 February 2010 (UTC)

ahahaha,Google? is wiki sponsored by them?

Table notes converted into group references

I notice that the {{refun}} had recently been converted into {{ref}} which caused the numbers of references and table notes to overlap. Since I figured that might be a tad confusing and {{refun}} is up for deletion I converted the tables to use the group list feature of references. This means that table notes should automatically be linked up properly. Hopefully this will make it easier to keep the table foot notes in sync. I left most of the bodies of the notes defined below the tables except for a few cases where nested references forced me to move them to avoid citation warnings. The rest could be moved up if desired, I just wanted my change to be easily diff-able. If you have any questions about how this works, see Help:Footnotes or Wikipedia:Footnotes or ask. Note the occasional use of {{#tag:ref|...|group=note|name=...}} syntax to handle the special cases of normal references nested inside grouped references. PaleAqua (talk) 10:29, 10 March 2010 (UTC)

Thanks— I should have done it that way in the first place. ---— Gadget850 (Ed) ^talk 13:50, 10 March 2010 (UTC)

No problem. Was a fun little exercise. PaleAqua (talk) 01:08, 11 March 2010 (UTC)

Fun. It has taken hours to work that little template. It has been used in so many odd and wonderful ways. Tables full of fracking dagger symbols that link to different notes. One last article to do... ---— Gadget850 (Ed) ^talk 04:39, 11 March 2010 (UTC)

Acid3: Not passed / 100/100

Opera (for desktops) has been listed on Acid3 as not passed with an otherwise perfect score. It has rendered the test correctly for the last few versions (at least with me). The rendition is normally smooth and fast here. Is there something important that fails Opera, or is it just an old forgotten update? 158.37.73.13 (talk) 00:31, 20 March 2010 (UTC)

As far as I know, it still doesn't pass the performance aspect. Does the report say that no subtests take more than 33 ms? Are you running the test on a laptop? If both answers are YES, then perhaps it passes. On the other hand, we can't add that information to the article because it's original research. -- Schapel (talk) 00:55, 20 March 2010 (UTC)

I am running it on a laptop (HP dv9575eo/Win7), and test 26 and 69 usually come out less than perfect. But it generally does the test better than Chrome here, which is said to pass it perfectly. Here's a screen-shot of the shortest of around thirty tests ran February 17 (only six ran for IE): ImageShack. I know it's not statistically significant, but Chrome has not consequently ran the test better on this machine. 158.37.73.13 (talk) 01:43, 20 March 2010 (UTC)

From the screenshots shown, Chrome does better on the performance aspect. It ran test 26 in 43 ms, compared to over 100 ms for Opera. It is not the total time (speed of animation) that is the issue, but time for each subtest (smoothness of animation). -- Schapel (talk) 01:55, 20 March 2010 (UTC)

Though, the total time used calls for a better test than the fact that one single test is a bit slower or faster, does it not? Chrome might be fast on that specific test (but not faster than 33 ms as stated as perfect earlier), but the total time says that other tests added up to 1150 ms slower than Opera (in the quickest of all the tests). The 62 ms difference in a single stress test is small compared to the total. Why does Chrome pass when it never scored less than 33 ms on test 26 in any of my tests? Is a single lucky pass on any one system all that is needed for a browser to pass it? I would think it had to be consistent. Would be nice to get some different views or clarification on this - I'm no expert in the area.

For reference: the above tests were done on a fresh OS with unmodified browsers. 158.37.73.13 (talk) 02:41, 20 March 2010 (UTC)

But it's not up to you or me. The creator of Acid3 said it is the smoothness that matters -- each test must complete within 33 ms or less on the reference hardware. Did you run the test on the reference hardware -- a top-of-the-line Mac laptop? -- Schapel (talk) 03:35, 20 March 2010 (UTC)

A "top-of-the-line Mac laptop" is not a specific system. Acid3 was created in 2007, and 3 years is a huge amount of time in technology. I've stated my system earlier in this thread, which was a fairly top-of-the-line laptop back then., and Chrome does not pass for performance. Opera even has different versions running for Windows and Mac. The test should not only be valid for Macs when used in the context of this article. Should that not be stated to begin with, otherwise?

I don't mean to directly question the validity of the Acid3 test - I believe it's a good test, but while Opera performed better in my testing, Chrome passes because of what can be seen as unreasonable test requirements. I am sorry for singling out Chrome versus Opera, but those are the browsers I can personally reference. 158.37.73.13 (talk) 04:17, 20 March 2010 (UTC)

If your was a top-of-the-line laptop a few years ago, it would stand to reason that a top-of-the-line laptop of today would be more than 30% faster and not more than three times faster. That would mean that Chrome would pass the performance aspect and Opera would not. Your tests seem to verify what our sources say. -- Schapel (talk) 13:41, 20 March 2010 (UTC)

@158.37.37.13 - Please feel free to question the validity of the test all you like, it was created by an individual, and not an organisation or collection of peers, so it is by its very definition biased. The performance aspect is widely regarded as being preposterous and the creator of the test himself has actually stated that for Acid4, one of the requirements be: "no performance stuff as part of the test". However, it is not within Wikipedia's purview to comment on the validity or perceived bias of what it documents (doing so would likely violate NPOV guidelines).

Also, the reference hardware is not a 3 year old top-of-the-line Macbook, it's defined as "whatever the top-of-the-line Apple laptop is at the time the test is run" - this means even figuring out what the reference hardware is is a completely unscientific moving target, to make matters even more unreasonable.

@Schapel I'm not sure how your interpolation of Opera's vs Chrome's results on a modern PC Laptop works, but given Opera performs the test more quickly than Chrome in the tests in the linked screenshot, surely it would also do so on a new laptop. Either way though, this is moot unless the tests are run on a bootcamp Windows install - otherwise, as far as the test is concerned, a stable build of Opera 10.50 does not exist. _ɹəəpıɔnı 02:03, 21 March 2010 (UTC)

I'm sure Opera performs the whole test more quickly than Chrome, but Chrome performs the test more smoothly than Opera. Each subtest must complete within 33 ms, and because Chrome takes at most 43 ms per subtest on a 3-year-old laptop it seems reasonable that it would take less than 33 ms on a new 2.66 GHz Intel Core i7. Perhaps we should make a composite image of Opera's rendering and the report given in the screenshot above and use it to demonstrate how you can tell whether a browser passes the performance aspect of the test in the Acid3 article. People don't seem to understand how to interpret the report. -- Schapel (talk) 12:59, 21 March 2010 (UTC)

For me both Browsers stop too long on 69 and the rest goes rather fine —Preceding unsigned comment added by 60.240.237.227 (talk) 20:42, 2 April 2010 (UTC)

You need to run the test on the reference platform to know if a browser passes the performance aspect. -- Schapel (talk) 18:16, 4 April 2010 (UTC)

What do you mean reference system. Also Opera clearly passes with the same results as WebKit based browsers, this section of the article is clearly biased towards WebKit as Opera clearly passes with the same results (Core i5 based system) as any WebKit browser, Opera also performs the test faster. I will be changing this to yes as most browsers seem to have the same problem and still receive a yes. Also I can list my exact specs if you care that much as this is a PC I built. (EDIT) I have also noticed that the Opera Version does appear less smooth this is in fact due to the fact that some tests are running faster than others as where webkit I believe may have some smooth enabled, but this is irrelevant as the performance is still valid smoothness is insignificant where as the test is really asking for the performance to be Under 33 MS. SandySoup (talk) 03:03, 23 April 2010 (UTC)

You need to run the Acid3 test on a top-of-the-line Apple laptop. Opera clearly does not pass the performance aspect of the test. It does not matter whether Opera performs the entire test faster. It must perform each subtest within 33 ms on the reference platform to pass the test. This section of the article is not biased in any way that I have observed. In any case, all you need to do to state that Opera passes is to find a reliable source that states that Opera fully passes Acid3. We cannot use the results of your test as that would be using original research. Go ahead and read this entire thread for all the gory details. -- Schapel (talk) 21:35, 23 April 2010 (UTC)

Where did you get that from because using the term top-of-the-line Apple clearly implies bias either on Wikipedia's part or Acid's part. I am more inclined to believe it is Wikipedia which is biased as I ahve never heard of a Reference System (FYI Yes I have searched it up and came up with nothing related to Top of the Line Mac). Second of all what defines top of the line is different and can be argued, (EG My friend believes his Core 2 Duo mac is top of the line). Finally it should be Windows with a minimum hardware or even better any OS as Windows is used by most people not Mac OS X based Computers. Please stop this bias in favor of WebKit. Using Mac OS X applies only to 5- 10 % of all computer users versus Every other OS which applies to 90 - 95 % of all PC Markets. Opera had it first and the Wikipedia page on Opera clearly states that an internal Dev Build of it had passed Acid. So please tell me where you got this magical reference system information from. (Edit) Also where does it mention that the Computer had to be running Apple based Software and we can assume since its performance requirments are theoretically (As far as I am Believing) for a top of the line mac that a computer who has lower specs is irrelevant or better as it shows that the web browser can run as fast to pass acid on a slower system. (Edit 2) Should the Acid Team have listed Top of The Line Mac as their reference system then the entire integrity of the test should be questioned as it shows that the test makers believe a top of the line Mac is better for Web Browser tests than a custom built Linux based computer or a Dell, HP computer and this section should be question and removed as that shows the entire test believes Apple based technologies are better than other companies. (Edit 3) When the developer of Acid 3 worked with the WebKit team to repair the problem he also compromised the entire test as he gave the webkit team an unfair advantage against every other team as they did not have the knowledge of how this made (even with source code) and that the Webkit Team knew what everything that was being changed and the exact changes made compared to every other team. Webkit also had the advantage of starting work before every other team thus the entire test has been compromised. SandySoup (talk) 02:14, 24 April 2010 (UTC)

The Acid3 test itself specifies that the reference platform is a top-of-the-line Apple laptop. There is no bias. -- Schapel (talk) 11:45, 24 April 2010 (UTC)

Yes, but it does not specify the operating system. Also due to the fact that those are reference (as in hardware which is a maximum for a browser to pass for it to be valid) anything that is slower than the reference hardware can be considered valid as well. Another thing I noticed was that your position changed from any top of the line Mac to a laptop specifically. SandySoup (talk) 01:40, 26 April 2010 (UTC)

Well, I sincerely apologize if I misspoke once. How dare I do that! In any case, Opera doesn't come close to passing on an older laptop, and it doesn't pass on a newer laptop, so I'm not sure what your point is. In any case, what we need is a reliable source that states that Opera fully passes Acid3. As you are not presenting one, I see no case for stating that Opera passes Acid3. I'm still not seeing any bias. -- Schapel (talk) 02:48, 26 April 2010 (UTC)

←← Opera 10.50 for Mac is now a stable release. It's faster than Safari 5.0, which is faster than Safari 4.0 which passed the performance test, so I don't think it's reasonable to continue the exclusion of Opera at this stage. Thoughts? _ɹəəpıɔnı 22:23, 19 June 2010 (UTC)

First, we cannot make an edit based on a conclusion -- that is original research. We need a source to cited instead. Second, when some tests say that Opera is faster than Safari, that doesn't mean it's uniformly faster at everything -- it is almost certainly faster at some things and slower at others. We cannot conclude whether Opera completes subtest 26 of Acid3 faster than Safari based on a completely different benchmark. In any case, when Opera does pass Acid3, Opera Software ASA will surely crow about it. -- Schapel (talk) 22:27, 19 June 2010 (UTC)

I vote for removing any references in regard to pass/not pass. This has become more a religious topic than a factual topic. A browser only passes if they claim they do, and as soon as they claim they do, there is a source to cite and that's taken as a fact? And only because a certain browser never claims a pass, they never do? That's all ridiculous.--DeTru711 (talk) 01:02, 16 June 2011 (UTC)

Whether a browser passes or not is fact. Whether we can say so in Wikipedia depends on whether there's a reliable source to cite. Any reliable source will do. Those are Wikipedia's rules, and if you don't like them you can debate them in the appropriate places. As far as your implication that there's a discrepancy between what Wikipedia states about which browser pass Acid3 and which browsers actually do, I've never seen any evidence that there is any such discrepancy. Do you have evidence that a browser Wikipedia claims passes doesn't pass, or that a browser Wikipedia claims doesn't pass passes? -- Schapel (talk) 17:19, 16 June 2011 (UTC)

Wireless Application Protocol Bitmap Format

I would add the Wireless Application Protocol Bitmap Format (short wbmp), but I'm not sure if I should add this row to the Mobile Web technology support- or to the Image format support-table. what do you think? mabdul 17:45, 30 March 2010 (UTC)

removed references

Two references are removed about opera support. They aren't outdated or unneeded: I do think everything should become more referenced! Why remove these as facts? mabdul 04:33, 7 April 2010 (UTC)

I considered them leftovers from the initial announcements. First off, the ‘references’ in question were directly linked instead of ref'd. If we were/are to put them back, they should be marked up as references.

Bittorrent ref: http://my.opera.com/desktopteam/blog/show.dml/154823

IPv6 ref: http://my.opera.com/desktopteam/blog/show.dml/271098

Secondly, these are not really references. If we want to reference it, we should google for a third party article. The main reason I bothered to remove them, however, is that Opera is the only* one ‘referenced’, making it look like "yes we support it BUT..." I agree references are good, but in this case they're hardly necessary, as they're not disputed features. Anyway, that's my argument. It was probably a bad idea to say ‘removed references’ in my edit note, as that's not what I consider them to be. *The only other field in that table that's referenced is OmniWeb's Gopher support, but feel free to add proper refs if you feel they're valuable. WOFall (talk) 17:23, 7 April 2010 (UTC)

Bad Table Rendering

I've noticed that at least a couple of the tables on the right of the screen overlap the table of web browsers by year on the left of the screen. It looks to me like someone designed the page in IE and ironically enough failed to do a comparison of how it looked in any other browser. I've tried Firefox in Win & Linux, along with both Opera and a Chrome clone in Windows and the page only looks decent in IE. Maybe someone who knows how to work these Wikipedia tags can fix it because I'm not sure how to. 132.170.67.1 (talk) 18:23, 3 May 2010 (UTC)

This could probably be rectified by setting a simple right align on this template. However, I will not do it, as this template is used on quite a few other pages, and I don't want to get yelled at ;) WOFall (talk) 22:47, 4 May 2010 (UTC)

Language Mix Up

In the Internationalisation section, the last language supported by Google Chrome is listed as "Vietnamese (cy)". "cy" is the abbreviation for Welsh, so this needs to be changed to one of these here:

• "...Vietnamese (vi)"
• "...Welsh (cy)"
• "...Vietnamese (vi), Welsh (cy)"

N4m3 (talk) 16:32, 15 May 2010 (UTC)

Check version number before editing vulnerabilities

In the vulnerabilities table both Chrome and Safari seem to have been edited to reflect problems that were found in their 4.x versions after the 5.x versions were released. I removed the errors and just wanted to remind prospective editors to make sure to check against the latest version number before editing. 132.170.50.236 (talk) 15:05, 11 June 2010 (UTC)

Per-site security configuration - Chrome

The note on partial support for chrome is "Starting with version 4, Google Chrome can disable Cookies, Images, Javascript, Plugins, Popups, and Geolocation individually." but I fail to see how this is "partial", even from the note it sounds like chrome has supported per-site configuration fully since version 4 (which is even unsupported for being out of date now). Zamadatix (talk) 00:17, 12 August 2010 (UTC)

BMP Image format support

A BMP column should be added to the Image format support table. —Preceding unsigned comment added by 144.136.83.19 (talk) 13:22, 22 September 2010 (UTC)

BMP's not useful in webcontent since it's not compressed, and unlikely to be widely used. If it's not widely used, it's not topical TEDickey (talk) 13:24, 25 September 2010 (UTC)

But it's one of the handful image formats supported by all major browsers. 144.132.204.79 (talk) 05:20, 28 September 2010 (UTC)

ehm, its handful since every application/operating system support that file format. as you can see here that every major browser engine supports that. but as tedickey said: nobody should use this format and there are (png,gif,jpeg,etc) better formats. mabdul 10:03, 28 September 2010 (UTC)

This is about browser support, not format desirability. 144.132.204.79 (talk) 12:24, 28 September 2010 (UTC).

But we cannot list every feature supported by browsers. We need to exclude little used features to make the article tractable to read and edit. Let's not try to turn Wikipedia into a collection of trivial facts. Let's keep it an encyclopedia, which is a collection of articles that contain the most important information about the topics they cover. -- Schapel (talk) 17:13, 28 September 2010 (UTC)

If eight image formats not supported by all major browsers are included, surely one of the four that is shouldn't be left out. I for one came to this article to find out the extent of BMP support. Wikipedia should be, well, encyclopaedic. 144.132.204.79 (talk) 23:09, 28 September 2010 (UTC).

Yes, Wikipedia should be encyclopedic, meaning broad in scope or content. Encyclopedic does not mean including every bit of information no matter how trivial. Wikipedia should leave out most information on most topics. BMP is not a commonly used format for web pages. JPEG, GIF, and PNG are the main ones, with APNG, JPEG2000, TIFF, MNG, and SVG occasionally used. I'm not even sure why XBM is listed in the table. -- Schapel (talk) 01:04, 29 September 2010 (UTC)

TEDickey: Actually, the BMP format is compressed using RLE.

Schapel: I use BMP files every day. They're far more common on the web than JPEG2000, TIFF, MNG, or even SVG files. It's definitely worth including in the entry. Ideally, a browser should be able to read BMP files. Not being able to read them would be a drawback and it would be notable for someone not seeing a picture they were supposed to see, right?--Best Dog Ever (talk) 02:43, 29 September 2010 (UTC)

It's unlikely that you can provide a WP:RS for the remark. However, some examples of sites using BMP's would be constructive TEDickey (talk) 08:08, 29 September 2010 (UTC)

Here are several thousand sites using BMP images: [2].--Best Dog Ever (talk) 08:19, 29 September 2010 (UTC)

That's 12,000 hits. SVG gives 4 million. We're looking for something that supports your remarks TEDickey (talk) 08:37, 29 September 2010 (UTC)

MNG files get 327 hits: [3]. So, that means you're applying a double standard to MNG and BMP. In any case, there are, at the least, thousands of sites using BMP images. And those are just the pages Google knows about. That makes the format notable in this context to me.--Best Dog Ever (talk) 09:07, 29 September 2010 (UTC)

Agreeing with you that MNG also is nonnotable trivia, like BMP TEDickey (talk) 09:26, 29 September 2010 (UTC)

MNG is little used because there's little browser support. The same goes for JPEG2000 and TIFF. BMP is not widely used despite being supported by every browser for many years. It's not really a double standard. -- Schapel (talk) 11:09, 29 September 2010 (UTC)

I think that will change as more people switch from dial-up to broadband connections and from broadband to fiber-optic connections. It will also become more common as hard drives increase in capacity. In the future, image compression will be obsolete. Movies will also be streamed to our computers uncompressed. I have come across BMP images on the web by chance -- yesterday before I came across this conversation, in fact. So, I would be interested to find out which browsers support it and which don't.--Best Dog Ever (talk) 20:32, 29 September 2010 (UTC)

BE Bold! mabdul 21:09, 29 September 2010 (UTC)

Where is the information being added to the article about BMP support coming from? -- Schapel (talk) 19:01, 1 October 2010 (UTC)

Should we assume it is original research because there are no citations for the information? -- Schapel (talk) 12:44, 4 October 2010 (UTC)

Couldn't we nearly rate everything as OR since we don't have any reference at the moment? Maybe I / we can start a userfied version of this list and add the references to every cell. mabdul 16:28, 21 October 2010 (UTC)

Modern Browser Only Condensed view option

Many people coming to the page are only interested in comparing current browsers to decide which to use. However, because all of the comparison tables are multiple screens long and compare every browser that has existed, they find the page to be mostly useless. We need a way to show only current information for those who want it, in order to increase utility. A method for allowing the viewer to sort information and suppress what they don't need would also be good.

This is a general comment applicable to all comparison articles. —Preceding unsigned comment added by 136.142.206.28 (talk) 22:35, 15 October 2010 (UTC)

I think the page is just way too long, with lots of original research. We should weed out the browsers that have never had a significant fraction of usage share and even pare down the number of columns in some of the table. Then we can work on citations for the data that is left. -- Schapel (talk) 00:26, 16 October 2010 (UTC)

There's a fine line between original research and common knowledge. This constitutes common knowledge since anyone can acquire the information with minimal efforts. However, collecting the amount of information shown here would be impossible without a team of people working on it. Wikipedia is the perfect community to maintain this page, and it is extremely useful, functioning qutie similarly to Consumer Reports (without the opinions). Bob the Wikipedian ^{(talk • contribs)} 04:16, 16 October 2010 (UTC)

As I said above: Let us create a userfied version of this page and start working on adding references for all cells! mabdul 16:34, 21 October 2010 (UTC)

How about putting scores on every chart, then the last chart tells the score on every chart. :) Coleet (talk) 11:52, 21 October 2010 (UTC)

really: I can't imagine that this feature helps? Is the reader then smarter? mabdul 16:34, 21 October 2010 (UTC)

Charts for Beta Versions

Hmm... Wondering if we can add details for the beta versions; just need to include the date we have tested it, the version and what features are still missing. I'm Using Internet Explorer 9 beta and can help! :D and there are some people (including me) who wants to experience beta versions, and would like to check if it could harm or slow down my computer based on others experience. Well, Good luck! Coleet (talk) 11:47, 21 October 2010 (UTC)

Betas are not finished version. these trable are complex enough. If the betas remove again a ferature the confusion is pre-programmed. mabdul 16:32, 21 October 2010 (UTC)

Disputed

I've disputed the claim that Chrome and Safari pass Acid 3 having repeatedly asked for a reliable source for this claim, including a user page discussion and have been meet with a mixture of diversions on the subject. With no reliable source forthcoming after multiple requests it brings to doubt if there are any, so, I have marked the section as 'factual accuracy is disputed.'. Regards, SunCreator ^(talk) 01:48, 2 November 2010 (UTC)

Considering I could not find discussion of the topic on your talk page, nor on the google chrome talk page, I can't vouch for any discussions you've had elsewhere (I can't find them). However, as on the Google Chrome page there is a link to the official chrome releases blog here which outlines Acid 3 compliance. Sylvanelite (talk) 10:46, 4 November 2010 (UTC)

Thanks for your reply. The discussion has been both in the edit summary and on the reverting editors talk page. The source you supplied is the best yet suggested, unfortunately the Chrome blog is not a reliable source but a self published blog, same issues of not being a reliable source applies to the Apple blog for the Safari browser. Regards, SunCreator ^(talk) 14:44, 29 November 2010 (UTC)

Self published information by corporations is generally considered a reliable source whether it is in blog style or web page style. What matters, whether its a corporation or individual, is are they an expert on the subject. I think google can be considered an expert on whether their browser passes Acid 3 tests. Doubly so, since google was the main developer of Acid 3 tests. :)   Daniel.Cardenas (talk) 16:43, 29 November 2010 (UTC)

Self-serving promotional claims are rarely considered an acceptable reliable source but I'll let others comment. Please raise the issue on WP:RSN. Regards, SunCreator ^(talk) 17:05, 29 November 2010 (UTC)

Do you have an example of corporate published info not considered a reliable source? Anywhere you look corporate info is used as a reliable source throughout wikipedia. Daniel.Cardenas (talk) 05:34, 30 November 2010 (UTC)

Are we looking at the same Wikipedia? By definition such sources are not independent, and they are (or should be) treated with great care. Drmies (talk) 05:35, 30 November 2010 (UTC)

The information is from their official website, and as no company is not going to flat-out lie about something that easy to check, it should be regarded as reliable and true until proven otherwise by a reliable source.-Zyrath (talk) 04:31, 11 December 2010 (UTC)

Oh, I'm happy someone noticed it too. I can confirm that on my computer, Chrome gets 100/100 but fails the test (#26 and #69, like Opera). Does the result change if you use another computer? Shouldn't the information be changed? -Zaba 17:26, 13 February 2011 —Preceding unsigned comment added by 80.11.199.18 (talk)

First, you're doing original research which is not allowed in Wikipedia; you need to have a reference to a reliable source. Second, I suspect that you're not running the Acid3 test on a reference computer as you are supposed to for the performance aspect of the test. Third, we have confirmed that Chrome's official blog is indeed a reliable source. So, no, the information should not be changed. -- Schapel (talk) 16:43, 13 February 2011 (UTC)

Too long

This article is far, far, too long. Something needs to be done about this. Now.-Zyrath (talk) 04:35, 11 December 2010 (UTC)

any examples? What do you think? I do think everything is fine... mabdul 09:40, 11 December 2010 (UTC)

Like (almost) any of the comparison-of topics on Wikipedia, a large fraction of the content isn't really interesting to most readers. However, pruning it seems hard. Splitting it into manageable-sized topics is probably the way to go. TEDickey (talk) 13:07, 11 December 2010 (UTC)

What do you ave in mind? I think we could remove (although I don't think that this would be a good idea) the web (/Internet) standards parts (except the protocols) because these parts are already handled in the comparison of web browser engines-series and the reader has only to translate the engines with the corresponding browser. Would this be manageable and satisfy solution? mabdul 14:06, 11 December 2010 (UTC)

Maybe a good thing to start with is to move off (separate article?) all the browsers whose development has been discontinued.--DeTru711 (talk) 00:57, 16 June 2011 (UTC)

I think this article can be split into two separate articles: "Comparison of Discontinued Web Browsers" and "Comparison of Modern Web Browsers". – ShiViMan (talk) 20:16, 18 September 2011 (UTC)

I support this initiative.1exec1 (talk) 09:14, 21 September 2011 (UTC)

Release Table-template

I have no idea where this comment goes, but the very first table (Web browsers by year) should be corrected to reflect the advent of Chrome 8 in 2010 (right now it says Chrome 4 through Chrome 7). Thanks! - JobiWan144 — Preceding unsigned comment added by JobiWan144 (talk • contribs) 08:32, 13 December 2010 (UTC)

I moved you're comment at the bottom of the talk page. next time use the preview option! I willadd the new chromes. mabdul 09:41, 13 December 2010 (UTC)

Why so many old versions of browsers in Vulnerabilities?

The Vulnerabilities table is huge now that so many older versions of browsers have been added. I can understand IE 6, IE 7, and IE 8 in the table because they are all supported and in common use. I can understand having Firefox 3 and Firefox 4 in the table after Firefox 4 comes out while both versions are supported and popular. Perhaps sometimes we can have two versions of the other browsers, again just after a new version comes out and both old and new versions are popular. But the last three versions of Chrome? The past three versions of Opera? The past four versions of Safari? Why? Many are no longer supported, so they are not getting vulnerabilities fixed. Many of those browser versions are barely used any more, so information about the vulnerabilities in them isn't very useful. If someone wants to look up that old information, they can always go to Secunia and SecurityFocus and look it up. -- Schapel (talk) 12:33, 17 December 2010 (UTC)

Regarding the size of the table, I don't believe that it is either huge or too large. It is still smaller than most of the other tables on this page and the actual browser data just fills a single screen on my laptop. However, I do believe it is larger than necessary due to the formating of the "{{start" template(?). The table is mostly blank/whitespace because of it. If there was a way to shorten the output so instead of "December 10, 2010;" it produced "Dec 10, 2010;" then we could either move the number of open advisories onto the date line or bring the "8 days ago" information up to the date line saving quite a bit of vertical room (~30%).

I included the additional browsers/versions because they are being used (more than you would expect anyways) based on metrics obtained from Netmarketshare and W3Counter (the same metrics utilized in Usage share of web browsers). For example, Chrome versions 6 and 7 are in greater common use than version 8, so I included them (would you prefer that version 8 be removed instead?). FF 3.5 and 3.0 are both in the top 10 list on the W3Counter, and according to the Netmarketshare, FF2 is nearly as popular as Chrome v6. Safari 4 and 5 were included because they're also in the top 10 list at W3Counter. I included Safari 2 and 3 because people using older versions of OS X are not able to upgrade to the latest version of Safari (similar to WinXP/IE6); However, in hindsight, I do agree that both of these versions should be removed. Opera version 9 and 10 are still in greater use than version 11. Also, I removed Seamonkey because it doesn't appear in any of the metrics. Currently, there are 17 browsers/versions in the table, removing Safari 2 and 3 would bring that down to 15. Personally, I feel that removing Chrome 6 and 7, for example, is robbing people of potentially valuable information, and hopefully, it will encourage more people to upgrade (and many clearly aren't based on the metrics). -- Lhignight (talk) 13:41, 18 December 2010 (UTC)

I think the latest version of browsers should always be included in the list, so no, I would not argue that Chrome 8 be removed. The statistics you list are from November, so of course Chrome 8 was not used much then, as it was released in December. This is also why Chrome 7 is popular in the stats you cite -- Chrome 7 was the latest version in November. I think the first thing to realize is that the data you're using for your argument is outdated. I can understand having Chrome 7 on the list while it's supported and popular, in addition to Chrome 8. Similarly, Opera 11 was recently released, but I can understand having Opera 10 on the list while it's still supported and popular. I can understand having Firefox 3.5 on the list because it's supported and popular. I can understand having Safari 4 on the list while it's supported and popular.

As for Chrome 6, Firefox 2, Opera 9, and Safari 2 and 3, those are quite old, not used much, and I don't think supported any more. After browsers are not supported, it seems that security vulnerabilities are not reported against them any more. I don't think even Firefox 3.0's vulnerabilities are updated any more; it's inconceivable to me that no security vulnerabilities were discovered in it since Mozilla stopped supporting it about a year ago. The data for these old, unsupported versions are not reliable for this reason. I would recommend removing these older versions form the table to make the data more reliable and reduce the size so it can be read more easily. Nearly all the tables in this article are huge and nearly unreadable. I think this should be fixed by making the unreadable tables more readable, not making one of the few still readable tables also unreadable. -- Schapel (talk) 14:37, 18 December 2010 (UTC)

funny, since when IE is insecure? safari wont even start if you have mandatory DEP/ASLR on, same with chrome though i dont have the courage to touch it nowadays — Preceding unsigned comment added by 85.130.6.148 (talk) 00:19, 23 July 2011 (UTC)

SecurityFocus information is incorrect?

I see that someone has removed SecurityFocus information from the Vulnerabilities section with the justification that it is incorrect. I've never seen any claim before that the information was incorrect. I think the issue should be discussed before the information is removed. I also see there's a claim that the SecurityFocus information is hard to update. I would agree it's a bit of a pain, but if someone thinks it's too much of a pain to do, they can simply not update it. I have been updating the information and correcting others' mistakes for years. I only ask that if someone wants to update the information that they do it properly, and it they cannot do it properly they can let me or someone else who can do so update it. -- Schapel (talk) 15:40, 17 December 2010 (UTC)

I think what the editor who removed SecurityFocus information meant was that the links were wrong on the old versions of browsers he added to the table, and he didn't want to go to the trouble of fixing them. That, I think, is no reason to delete content from Wikipedia. I can fix the links for the browser versions that we reach a consensus on keeping. -- Schapel (talk) 23:36, 17 December 2010 (UTC)

Actually, I removed the SecurityFocus information because some of the information listed appeared to be incorrect AND I didn't want to have to update the new additions that I added to the table. =)
Let's start with the false positives that I noticed:

• IE6: If you click on the oldest unpatched link then on solution, you'll see that Microsoft has released a patch for this vulnerability.
• IE7: Again, a patch has been issued for the vulnerability.
  • In both cases, there appears to be conflicting information on the site;
    This isn't a complex issue like climate change -- either a patch fixed something it or it didn't -- how can you rely on the information otherwise?
• When I did a search for vulnerabilities in IE6 SP3 there were ~500 results total; I find it hard to believe that 473/500 are unpatched.
• There are supposedly 62 unpatched vulnerabilites in IE8, but only 26 in IE7? That seems very peculiar to me.

Regarding false negatives, both Safari and Opera have unpatched advisories according to Secunia, but there wasn't anything listed in the SecurityFocus column, which I attributed to difficulty culling the data. IMHO, the SecurityFocus data is inconsistent, far too difficult to cull, and frankly, unreliable. Personally, what I would like to have added is a column that summed the total number of unpatched days for all advisories for a given product, but since that data isn't readily available, I figured adding statistics on the total number of advisories, number of advisories in 2010 and percentage of advisories that were very critical seemed like a close second place. -- Lhignight (talk) 14:29, 18 December 2010 (UTC)

Java applet support

There are several ways of Java applets support in browsers: full-native (?), using JRE, using plugin (and JRE?), using JDK etc. The article has only the following comment for it: Java support is for built-in support by the browser without the plugin from SUN.. It's not clear for editors (that's why Firefox and IE randomly has Yes while Mozilla has No) and not really useful for readers: they probably need to see the ability of browser to run applets (via pre-installed software, official plugin, 3rd-party plugin etc) rather than core abilities for something not clearly defined (due to copyright issues?).

So I suggest to drop the old rule (about built-in support) for this feature and fill the column with different values for different cases: No (never), Plugin, JRE, JRE+Plugin etc. Or to set for most browsers Yes with footnotes. —Preceding unsigned comment added by 80.70.236.82 (talk) 10:34, 17 January 2011 (UTC)

Browsers have two ways of supporting Java -- built-in support and using the Java plugin. The possible options would be No, Built-in, or Built-in and Plugin. But all browsers that support plugins would support Java using the plugin, so it's not really making a comparison between different browsers, which is the point of the tables. I think that's why the options are No and Yes (builtin). We should really save footnotes for exceptional cases, not the common case. -- Schapel (talk) 14:46, 17 January 2011 (UTC)

Are there really two ways of supporting Java? Here is my little OR (i.e. copypasting official documentation):

Opera for Windows: For Windows, Opera uses the Java plug-in to run Java applets. If you do not have Java installed, Opera will automatically offer to take you to a download site when you encounter such content. To install Java, go to the Java website and follow the instructions. (c)

Opera for Linux: Sun Microsystems (now part of Oracle) created the Java programming language, but most Linux distributions will include a default Java implementation that was not created by Sun/Oracle. (c) The Java plug-in is part of the Java Runtime Environment (JRE), but Opera does not use the plug-in: it uses the JRE directly. To get Java working with Opera, you must install the JRE, and configure Opera with the correct Java path. (c)

Internet Explorer: Microsoft no longer ships Java with Windows, and in recent versions of Windows, Internet Explorer cannot support Java applets without a third-party plugin. (c) To run Java programs in Internet Explorer, however, you must install special Java software. (c)

Firefox/SeaMonkey for Windows/Linux: Before Firefox can run Java applets you must have the Java Runtime Environment (JRE) properly installed and enabled. Starting in Firefox 3.6, you also need the new Java plugin ... (c)

Firefox/SeaMonkey/Camino for Mac OS X: Java Runtime Environment that ships with OS X has major issues as a plugin for Gecko-based browsers. As such, mozilla.org has opted to include the Java Embedding Plugin by default in all of its projects. (c)

Chrome: On Windows, Mac, or Linux? Google Chrome supports most popular plug-ins, including Adobe Flash Player, Adobe Reader, Java, ... (c)

Safari for Mac OS X: Enjoy running Java applets on the web thanks to Java support in Safari. (c) (no plugin, just JRE?)

Safari for Windows: For a richer browsing experience, you may want to install the following plug-ins: ... Java (c)

Konqueror: The Java support relies on the KDE Java Applet Server( KJAS ) ... You do not need a complete JDK for KJAS to work- you just need a JRE. (c) (no plugin, just JRE and KJAS)

The Java Plug-in: Java Plug-in enables applets written to the Java SE Platform 6 specification to be run in Mozilla and Internet Explorer web browsers. Java Plug-in comes with the 6 JRE/JDK. (c)

Therefore these two ways are about the same for users: using a 3rd-part JRE directly or using a 3rd-part JRE through the plugin supplied with JRE. Compare with 3rd way in Nexus: it has never had the Java support and has the same No in table as modern browsers. —Preceding unsigned comment added by 80.70.236.82 (talk) 17:01, 17 January 2011 (UTC)

The only browsers I can see that you're saying work without built-in or plugin support are Konqueror and Opera. I can't load the page you gave for Konqueror. The page for Opera you list states Opera uses the JRE directly, but only on Linux. If you'd like to add a footnote that Opera on Linux uses the JRE directly and not the plugin, that's fine. I'll change the table to reflect that IE no longer has built-in Java support. I'll also remove the from SUN qualification from the statement about Java plugins, because there are plugins from other sources now. -- Schapel (talk) 14:52, 18 January 2011 (UTC)

Now You can see the nonsense of this approach: all browsers has No excluding Safari with a noted restriction. Would You use the same without-a-plugin rule case we add the Flash support column? —Preceding unsigned comment added by 80.70.236.82 (talk) 07:57, 19 January 2011 (UTC)

No, because no browsers support Flash without the plugin. It would be silly to have a Flash column for this reason. We do it for PDF support, because some browsers do support PDF without a plugin. The article seems to be consistent with the way it compares browsers. If you have some sort of rationale for changing the way the tables in the article are constructed, let's have it. -- Schapel (talk) 17:06, 20 January 2011 (UTC)

That's about data, not tables. The article is for users, not only compare geeks like us. I believe they do interested in the ability of using a feature rather than in internal support details. Having some JRE preinstalled (most cases) user can install a modern browser and have the Java support enabled without extra fuss (because of Java plugin supplied with JRE). Or he can install a not so modern browser and get no Java support at all. Why would he need a table with No for all these cases?

My suggestions are higher: drop the native support restriction and check the ability of browser to run Java applets with JRE installed. Does it (pre)installed or not depends on the OS (incl. release version). Some browsers require an additional plugin instead of the "Java Plug-in" from JRE and should be marked accordingly. —Preceding unsigned comment added by 80.70.236.82 (talk) 09:18, 21 January 2011 (UTC)

Spatial navigation feature of web browsers

As the creator of TVFriendly, I have conflict of interest regarding this article, so I leave the decision whether to include this adding up to others. The adding that I am suggesting is about the (absence of) Spatial Navigation feature in IE and in Firefox that can be achieved by using an extension like: TVFriendly (www.tvfriendly.com). In the table: “Accessibility features (continued)”, in the rubric for the feature “Spatial Navigation” – for the: “Mozilla Firefox”, It is written “No” with a note [note 14] that tells about some special Firefox builds with the spatial navigation features. I think that it can be more informative if the note will tell also about the extension. Sincerely, Tal Talbob (talk) 16:47, 2 June 2011 (UTC)

Web-gl support

Web-gl should be added in the list for comparison. 83.108.198.50 (talk) 20:37, 12 July 2011 (UTC)

would you bother elaborating on adding yet another security hole? — Preceding unsigned comment added by 85.130.6.148 (talk) 00:12, 23 July 2011 (UTC)

Mobile browsers missing

Browsers like iOS Safari (different from desktop Safari (web browser)) and Android (operating system) browser are missing. They share the same rendering engine (WebKit) but the user interface is quite different. — Preceding unsigned comment added by 88.27.29.10 (talk) 10:20, 7 September 2011 (UTC)

Yeah, the should create a Comparison of mobile web browsers, but starting a new page is highly hard since we should start another comparison with unreferenced facts (in my eyes). we should only create (new) pages with references for ALL features... (in my eyes). mabdul 19:30, 27 September 2011 (UTC)

Browser stats conflict

Being a web developer, nearly everyone I know goes by w3schools as the best browser stats, which completely disagrees with your data. Did Microsoft pay you guys under the table or what, lol? I hope you review the best source for stats, http://www.w3schools.com/browsers/browsers_stats.asp — Preceding unsigned comment added by 68.183.149.24 (talk) 17:24, 27 September 2011 (UTC)

Those stats are only for the w3schools server itself. The stats for that site are not typical, because the visitors are not typical users. You can find more explanation at usage share of web browsers. Oh, and assume good faith. Thanks. -- Schapel (talk) 18:53, 27 September 2011 (UTC)

Do they or don't they support Java?

I'm looking for comparative info browser-to-browser. I cannot find any USEFUL info about Java support. Is this because too many people are firmly welded to Wikipedia conventions? GET OUT OF THE SQUARE. As it stands, this page is useless to me and many others. The comment above from Schapel on 14:46, 17 January 2011 (UTC) in Java_applet_support is not helpful, it sidesteps the object of an encyclopedia, which is to give or clearly reference all available information. So, does any given browser support Java in any way at all, or does it not? 122.200.166.205 (talk) 02:53, 30 December 2011 (UTC)

No, the object of an encyclopedia is not to give or clearly reference all available information. There are many things that Wikipedia is not, including an indiscriminate collection of information, which you seem to think it should be. As to your immediate need, to get Java, you install the Java plugin and then Java works. -- Schapel (talk) 13:10, 30 December 2011 (UTC)