Talk:Comparison of web browsers

From Wikipedia, the free encyclopedia

Former featured article candidate This article is a former featured list candidate. Please view its sub-page to see why the nomination failed. Once the objections have been addressed you may resubmit the article for featured list status.



Contents

[edit] Features

[edit] Konqueror

Since when does Konqueror run without emulation under Windows? Even partially? It doesn't, the only way to get even close is to use Cygwin.

[edit] ALPHA/BETA version features

DON'T add alpha/beta software! SeaMonkey is not released yet as final.

Why not? We already know nearly all the details about the final release. Any information that turns out to be incorrect can be corrected and be up-to-date as soon as 1.0 is released. -- Schapel 23:09, 18 September 2005 (UTC)

Since the information table includes only the latest stable release, should we include features of beta/nonstable releases in the actual tables. I'm all for footnotes, but should the tables say "Partial" or "No" next to the footnote? I believe we should not put "Yes" if the release is not final. KramarDanIkabu 23:56, 15 Mar 2005 (UTC)

We need a "Coming in Next Release" or "Planned" template, perhaps with a blue background. Also, a "Via Extension" one (with a yellow background like partial) would be nice to round it out. 141.157.107.187 21:41, 20 December 2005 (UTC)

[edit] Caret browsing

Gecko-based browsers (Firefox, Galeon, Epiphany, K-Meleon, etc.) have caret browsing, but some disable it by default and doesn't provide the F7 hotkey. It's still available from the "accessibility.browsewithcaret" setting in about:config, so I believe this warrants a "Yes" (or at least "Hard to find, but it's there"). I haven't checked with Camino, so I'm leaving it as is. Edit: Changed to "Disabled".

[edit] Dillo - tabbed browing

I couldn't find any hint that dillo uses tabbed browsing. Change entry to No. --Kubieziel 10:01, 2005 Apr 8 (UTC)

There was an unofficial patch floated for the 0.7.x series that abstracted the page from the window, enabling both tabbed browsing and true frame support, but the project leads didn't accept it. --Kelson 00:18, August 5, 2005 (UTC)

[edit] Email

Should email even be listed on the feature comparison? If it should, Outlook Express should probably be mentioned along with Mozilla Thunderbird but if that route is chosen, I expect that there'd have to be a footnote for basically every browser that doesn't have a built-in email component.

I think either the Thunderbird footnote should be removed (since it's a separate program, and not part of Firefox), or Outlook Express needs to be added, along with any other email programs offered as separate downloads for the listed browsers. --24.218.182.238 03:20, 19 Jan 2005 (UTC)

I think email should be removed. Gruepig 19:03, 19 Jan 2005 (UTC)
I say remove all the footnotes regarding email support. Yes, you can always install a separate email client - that's exactly what "no" in the table means. GregorB 14:56, Feb 4, 2005 (UTC)
I agree... keep e-mail in the table but remove the footnotes. --Ctachme 06:17, 10 Feb 2005 (UTC)
I disagree. At least one footnote should be listed to explain why no email support is listed. Otherwise a reader will see the "No" and assume it's an inferior browser. Alternatively, I think we should remove the entire email column --Will2k 16:14, Feb 11, 2005 (UTC)
But a browser that doesn't support email is inferior (in a literal sense): it can't do what browsers with email support can; again, that's what "no" means. For the record, I think that browser email module is a liability rather than an asset, but obviously many would disagree. That's exactly why I'd keep the column, and let the people judge for themselves. GregorB 14:29, Feb 16, 2005 (UTC)
Browsers and their email capability are an important example of the Law of Software Envelopment. --DLeonard 03:12, 2005 Jun 8 (UTC)
Inferior? A browser doesn't supposed to support email... In the minimal case, it should just be a HTTP browser. --minghong 05:27, 18 Jun 2005 (UTC)
I think e-mail support is pretty self-explanatory, a footnote with the link to a comparison of e-mail clients is a good idea but I don't think it needs the bit about browsers avoiding being e-mail clients. Dyson's Dragon 03:31, 1 December 2005 (UTC)

[edit] Extensions/Plugins

Can we come to a conclusion about whether plugin/extensions "count" for a browser? I'm willing to accept "this browser can do this with a special version" but an extension is not part of the browser and shouldn't be listed. If we start listing extensions for one browser (Mozilla/Firefox), then we really need to have people go and look to see if an extension/plugin is available for every single other browser on the list, and I don't think that's really feasible.

Adding in these extensions either means a lot of additional work to find what does and what doesn't work, or a loss of the NPV as a certain browser (say, Firefox) has features listed that it doesn't support while another browser (say, Internet Explorer) has features listed as missing that can be supported with third-party plugins. (For example, SVG support with the Adobe plugin.)

I think that there needs to be an official decision not to list plugins/extensions for the browser. Maybe an additional section should be added indicating what plugins/extension support a browser offers, but plugins shouldn't be used to artifically make it look like a browser supports something it in fact does not. -- Xenoveritas 21:05, 4 Mar 2005 (UTC)

I think that the current system of saying NO but allowing links to footnotes is fine. That makes it clear the browser does not have the functionality, but the footnote allows it to be known it is possible to add it. --Ctachme 21:41, 4 Mar 2005 (UTC)
I disagree, because the lack of a footnote then suggests that a browser doesn't have a plugin that adds support, which may not be true. Unless someone wants to verify that every single no can't be solved by some plugin, somewhere, then I think it's best just to leave plugins off the page. -- Xenoveritas 01:41, 5 Mar 2005 (UTC)
It is clear that listing of extensions/plugins is a means for fans to promote their favourite browser. In my opinion, we should mandate that only native capabilities are included in the tables and we should remove footnotes. Footnotes about SVG native support coming soon in Firefox/Mozilla might be ok, but notes that Mozilla has a Chatzilla extension for IRC capabilities is clearly trying to compete with Opera or other browsers that have IRC support natively. Where the line gets fuzzy is when extensions/plugins are packaged with the browser (I think Opera does this with Flash and PDF plugins, can someone confirm?). So what should we do about these scenarios? Jeff schiller 6 July 2005 21:17 (UTC)
It's equally "clear" that removing extension/plugin listings would also be a means for fans to promote their favourite browser (it would merely benefit Opera fans instead of Firefox fans). NO with footnotes is the least POV representation: it distinguishes between features which are missing altogether (like page zooming in Firefox) and those for which there are common extensions (like search toolbars in IE), while still identifying native features. The risk Xenoveritas identified above is overstated: any plugin sufficiently widely used to be worth mentioning will have users among those who watch this page. --Haeleth 15:59, July 13, 2005 (UTC)
Agree. And there are extensions made by official bodies for extended features (e.g. ChatZilla) and experimental features (e.g. XForms). --minghong 16:13, 19 July 2005 (UTC)
Something has got to change certainly, the title of the table says features that are availible natively and without third party support. I agree with Jeff schiller on this, firefox and IE certainly have enough plugins and add-ons between them to make the footnotes longer than the actual table. Officialy supported plugins I would say are a different matter and I would say there no different from options that are disabled by default. Dyson's Dragon 03:31, 1 December 2005 (UTC)

[edit] Find as you Type

Just had another thought. Find as you type (FAYT) is a specialized feature found in all Mozilla browsers, Netscape 7.2, and I believe has also been implemented in Konqueror as well. Perhaps this should be added to the list.--Will2k 17:33, Jan 6, 2005 (UTC)

Go ahead and add it for the browsers you know about, that's the only way anything gets done.--Ctachme 21:19, 6 Jan 2005 (UTC)
Indeed, btw Opera has it. -Iopq 15:03, 15 June 2006 (UTC)

[edit] HTML access key

What is "HTML access key"? Samboy 00:06, 22 May 2005 (UTC)

The accesskey attribute of HTML form and anchor element. --minghong 01:45, 22 May 2005 (UTC)
A description of what it does and why it's important enough to be listed along with features like Tabbed browsing and pop-up blocking would be nice. Dyson's Dragon

[edit] Internet Explorer and its language support

Why is IE only listed as supporting English? -- Jugalator 20:50, August 25, 2005 (UTC)

Be bold and correct it. ;-) --minghong 05:04, 28 August 2005 (UTC)

[edit] K-Meleon's crippled Page-Zooming capablities

K-Meleon Version 0.9 has been released with capablities to zoom in and out all elements on some webpages such as yahoo.com by increasing the displayed font sizes, and image dimensions once a button has been pressed on the zoom toolbar. The downside of this feature is that it doesn't work on background images. Because of this problem, should Page Zooming be considered as "Partial", because it works on some sites, or should it be set to "None" until full support is implemented? --Lbmixpro 09:41, Mar 13, 2005 (UTC)
Change it back to partial then. P.S. The page zooming sucks. It takes several seconds to resize the page. Unless page zooming is supported by the Gecko engine, I don't think it will be lightning fast. --Minghong 11:22, 13 Mar 2005 (UTC)

[edit] Opera Adblocking

Opera doesn't have adblocking features, who added this and why? If you need some kind of external plugin, then it really should be Partial, like Firefox. --kyhwana 02:18, 23 April 2005 (UTC)

I don't think there's even a plugin for adblocking in Opera. As far as I know, that entry should be No. --Schapel 12:28, 24 Apr 2005 (UTC)
Opera has a native filtering method, filter.ini. CSS can also be used. And there are plug-ins based on Perl or Lua. 24.132.245.94 14:48, 24 Apr 2005 (UTC)
And here is how you do it: [1]

[edit] Opera Features Bias

The categories for the tables seem largely Opera based. What I mean is, everything seems based on what Opera can do, instead of a summary of what they all can do. Is there anything that can be done about this. I think there should be something about skinning, whether the browser can be skinned, etc. --Zeerus 12:27, Apr 13, 2005 (UTC)

That's because Opera has the most features: it does (almost) everything Firefox does, and a lot FF cannot do. That other browser manufacturors are slow to implement these features is no reason to drop them from the list: Firefox usually copies Opera features shortly after Opera implements them. Of course, Firefox doesn't really support much at all, unless many different "extensions" (i.e. third-party addon software) are installed, with all the risk and annoyance (incompatibility between versions etc.) that brings. --Jordi·✆ 12:39, 13 Apr 2005 (UTC)
Really? These are just some common features of modern browser. I don't think it is Opera-centric. P.S. I'm a royal Firefox user. --minghong 16:13, 13 Apr 2005 (UTC)
Zerus, if there are features you think need to be added, we should discuss them and add them as columns. I tend to agree that Opera is probably the most feature-complete browser out there and the table is just proving that out. Btw, I am an Opera, Firefox and IE user :P Jeff schiller 6 July 2005 21:21 (UTC)
Because Opera has more features. Actually, I think there should be more columns to cover usability testing - clocking users performing various tasks per browser. I wonder which browsers will come out on top. Applesanity 19:28, 30 January 2007 (UTC)
The problem is we can't do it ourselves (as it would be WP:NOR and unverifyable). I doubt any third parties have tested enough of these browsers and published these results, and we can't rely on the claims of browser companies. Koweja 18:38, 28 March 2007 (UTC)

[edit] Pop-up blocking

Shouldn’t their be some sort of qualifier that IE only has pop-up blocking if you have SP2? A lot of people still use 98 and don’t have it.

That's IE SP2, which is only availible on windows XP and server 2003. Of course, every windows browser can get some ad/pop-up blocking through the hosts file.

[edit] Safari doesn't have tabbed browsing

Just to clarify, Safari (The last latest release 2.0.1 /412.5) doesn't have tabbed browsing. N0N4am0r 23:30, September 12, 2005 (UTC)

It certainly does. Make sure you have it turned on in the preferences. —Miles←☎ 14:10, 29 September 2005 (UTC)
This is why Apple should have it turned on by default. Alex 04:15, 2 March 2006 (UTC)

[edit] Speed?

If there information is available, to whatever extent it is available, I think we should categorize the browsers in terms of speed at loading html, etc. Obviously this in part depends on what features people have downloaded, so we should base this information on the base model for the browser. I'll see what NPOV information I can find. Theshibboleth 02:42, 11 September 2005 (UTC)

Problem is that there's no way to measure it impartially, because there's no browser they all work on for one, and two because their speeds are different on each machine anyway. Some people think Firefox is faster than IE, others think the opposite, and they think it for a reason. There's no way to measure it accurately. KramarDanIkabu 03:53, 11 September 2005 (UTC)

Add independent browser speeds tests from http://www.howtocreate.co.uk/browserSpeed.html ?

[edit] Safari and Shiira can zoom in on pages.

Safari and Shiira can zoom in on pages. This function is built into the OS, and does not zoom in on the interface.

[edit] What's Related

Has anyone any knowledge of a What's Related feature that may have been implemented early (or even current) versions of Netscape Navigator and Microsoft Internet Explorer. Some searching has reported that it would show you a list of web sites that people who have looked at the same web site have also looked at. It seems now for a browser to do this, it would have to report what web sites you are hitting to a central place and that would either be taboo in today's anti-spyware or similiar to Googles pagerank thus acceptable.

Internet Explorer's help file has a What's Related entry but it points you to the ability to hit search engines from the address bar.

MozillaZine has this to say: “What’s Related” was a browsing feature first introduced in Netscape Communicator 4.06, allowing users to retrieve a listing of websites similar to the one currently being viewed. This preference actually dates back from that browser—it never did anything in Mozilla, despite having checkbox in the browser preferences.

Mozilla’s implementation of “What’s Related” is dependent upon the appropriate sidebar being open. It’s not implemented at all in Firefox, Minimo, or Camino. See: http://kb.mozillazine.org/Browser.related.enabled --Bill Hamilton 18:36, 16 August 2006 (UTC)chamilto0516

[edit] Opera: Text-to-Speech?

The table "Accessibility features (cont.)" lists Opera as having native text-to-speech capability. I don't see this feature mentioned in Opera's online documentation (help files). Can anybody provide a reference for this? Thanks. -- Writtenonsand 19:13, 9 October 2006 (UTC)

Look here: http://my.opera.com/community/dev/voice/ ale5000 12:43, 10 October 2006 (UTC)

[edit] Standards Support

[edit] Additional Protocal support details

someone might want to add some of these details to the article : [2] porges 03:45, Apr 22, 2005 (UTC)

Any cached page? Can't access the page. Anyway, those should be added to comparison of layout engines. --minghong 09:05, 23 Apr 2005 (UTC)

[edit] Authentication support?

I started researching browser support for negotiated authentication at the SPNEGO page, but now I think it would be better placed here. It's an SSO feature mostly of interest to corporate users at this time. What table would it best be attached to? (Protocols?) --DLeonard 03:05, 2005 Jun 8 (UTC)

[edit] Combining tables

Could we combine the tables Images and Protocols and the CSS, JavaScript etc... "Features" from the features section into a "Standards Support" table. I feel like there are a tad too many tables right now, and those respective tables only have a few members each. I also would add RSS and Atom columns (but I don't really want to take the effort to until it's decided if this is the definitive # of tables). --Ctachme 22:18, 10 Dec 2004 (UTC)

[edit] CSS

Would it be a good idea to add a chunk on CSS compatibility for the latest version of each browser?

CSS compatibility tables can get awfully big, awfully fast. It should be on a separate page if we do create one. On a related note, the present table listing CSS2 support as yes/partial/no is a gross over-simplification. CSS2 is a complex specification, and no browser supports it fully. Singling out IE for "partial" and giving "yes" to virtually every other graphical browser is harsh, to put it mildly. Blufive 23:13, 24 Dec 2004 (UTC)
Since CSS compatibility tables can get really big (see [3] for example), the only way we can add CSS is by a vast over simplification. I feel that in lieu of the facts, the current representation is as close to accurate as possible w/o actually having such a large table. And striking CSS isn't much of an option because it is such a vital part of the web. --Ctachme 04:58, 25 Dec 2004 (UTC)
I agree that any CSS support info on this page (and there should be some) needs to be simplified a great deal for space reasons. So if we go the route of a full-on compatibility table, it should live on its own page, rather than further expanding this one. On my second point, my contention is that listing IE as only having "partial" CSS2 support, while (say) Omniweb gets "yes" (implying full support) is misleading. All the leading browsers have shortfalls in their CSS2 support. IE6 is the "worst" (though I'm not sure about some of the lesser browsers on the list) but it's not terrible; The yes/partial/no categorisation implies perfect/bad/nothing. The reality is more like good/pretty good/nothing. I'd suggest a more sophisticated method (say, percent support) but I have no illusions about how much work it would be to compile and maintain, not to mention the arguments about how to calculate the figures. So, a more sensible option might be to note that they all have problems, and then link out to the hypothetical detailed compatibility grid.
FYI, CSS comparison was already created here: comparison of layout engines (CSS). Help filling in the missing parts. ;-) --minghong


There should be one line telling about the browser's CSS capabilities. CSS1, CSS2, CSS3? Also, some of those features already listed are received with plugins if they are not there by default. That should be noted. --ZeroOne 00:08, 9 Nov 2004 (UTC)

I agree that we need to add CSS support, I just couldn't think of an objective test for CSS support. It's easy enough to say IE has "so and so" CSS1 support and "crappy" CSS2 support, but there needs to be some objective way of demonstrating that. As for plugins, those should be noted, I just didn't add them because I didn't feel like expending the effort to.
i think it might be best if we just added the latest _fully_ supported version of CSS, for example: IE has CSS1 support, but lacks in CSS2, so it should have CSS1 as supported CSS version. and so on. --Ressu 07:47, 18 Nov 2004 (UTC)
Do you think somthing like this little table section down here would be useful? (I haven't actually resperaced the information in the table... as you can see)


  Standards Support
CSS1 Yes Yes Yes Yes No Yes No No No No No Yes Yes No
CSS2.1 Discont. 2004 Yes Yes Yes Yes Yes Yes Yes No No No Yes No No
CSS3 No Yes Yes Yes No Yes No No Yes Yes Yes Yes No Yes
xhtml 1.1 No Yes Yes Yes No Yes No No Yes Yes Yes Yes No Yes
xhtml 2.0 Discont. 2002 Yes Yes Yes No Yes No No Yes Yes Yes Yes No Yes

--Ctachme 11:33, 18 Nov 2004 (UTC)

I don't think there is need to include all CSS standards, as it appears that the newer standards contain the featureset of the previous versions (with little difference) so if a browser supports CSS2 it is bound to support CSS1 (to some extent). --Ressu 20:02, 18 Nov 2004 (UTC)

[edit] Scope

Someone has made changes to the (now former) revision claiming Konqueror, Opera, Safari do not fully conform to CSS (2.1). I've reverted the changes, since I think "Yes" does not by all means imply "full" support. Because NO Browser has full support. The same source he cited implicitly stated this (even for Gecko). If you are going to judge on a more detailed level, do it in Comparison of layout engines (CSS) and please, no bias.
Remember, even Internet Explorer is marked as "partial" so this column doesn't help anybody if all Browsers would show up as "partial". Grey 20:32, 15 June 2006 (UTC)

Hmm, I must have misread the conformance requirements here. I somehow missed the bit about "must . . . render the document according to the media-specific requirements set forth in this specification", I only spotted the parts about parsing and whatnot. That doesn't negate the fact that no existing client conforms fully to the specification; I would be inclined to add some intermediates between {{partial}} and {{yes}} rather than incorrectly suggesting that some existing clients fully support the specification. It's important to indicate that conformance to CSS 2.1 is not yet complete in any browser, just as conformance to SVG isn't. —Simetrical (talk • contribs) 05:15, 18 June 2006 (UTC)
Whereas I agree that it isn't entirely right to call it "yes" and that we should indicate no client does fully implement CSS2(.1), it's diffcult to draw the line. At the Moment, CSS2.1 is Working Draft. The Errata were considered to be CR, but I can't find the note on this atm; they might have removed it. Also, clients intentionally do not comply to certain passages in the specs (and have no intention to change that). As seen in other discussions these are still problems throughout the series and need to be adressed. We could impliment something like "minor" and "major" (like it was done in css comp.), but the colour key imho is very, very important. Having all the Browsers marked "yellow" does not help anybody. I think what could be done is using some orange colour for Trident Browsers and the likes and mark the others as "major" and yellow. I just doubt it's worth the effort. This article is more a summary of the others, so that it doesn't make much sense to be too specific. I'd just leave it as is. Grey 15:02, 18 June 2006 (UTC)

[edit] IDN & Data: URL incorrectly catagorised

IDN and Data: URL are listed under "Protocol support", but these are web standards, not protocols, and should be listed under "Web technology support" (which is crowded as it is)... GregorB 13:51, Apr 29, 2005 (UTC)

That's hard... Any suggestions? :-S --minghong 09:52, 30 Apr 2005 (UTC)
Well... I don't expect "Protocol support" to grow much in the future. On the other hand, "Web technology" will certainly grow (although it looks pretty comprehensive to me now), and we'll probably end up splitting the table anyway. Moving columns is not fun; embedded notes don't make it easier, so I'm not exactly eager to do it. :-( Speaking of which: anyone seen Wikipedia:Footnote3? Automatic numeration would make footnotes much less painful to work with. GregorB 20:09, May 2, 2005 (UTC)
That's a smart idea. Let me try it on a smaller comparison page first, to see if it is good enough to be used on this huge page. You can also take a look of comparison of layout engines and its subpages. We had dropped the use of numeric footnotes already. --minghong 08:26, 7 May 2005 (UTC)
Just tried. It doesn't work when there are more than one reference to the footnote. :-( --minghong 08:29, 7 May 2005 (UTC)

[edit] PDF

I've dropped the PDF column. PDF is not a vector image format. It's a layer on top of PostScript [4]. Also, PDF is supported mostly by its plugin, so it is almost meaningless to compare. --minghong 07:09, 14 Apr 2005 (UTC)

[edit] Support for additional named entities for HTML

While browsing Wikipedia, I noticed a serious deficiency of Internet Explorer 6.0 SP2 (the latest version, I believe). It fails to support the full range of additional named entities in HTML , whereas several other browsers are said to do so. This means mathematical pages on Wikipedia fail to display correctly. Is anyone knowledgable about this aspect of browser functionality to add this element of functionality to the main article? It seems particularly important for Wikipedia users who read and write mathematical pages.

[edit] Unicode URLs

I'd like to add a column specifying if a browser supports Unicode in it's URLs or just Punycode. For example: Safari supports both formats for it's URLs (specific sections of unicode can even be turned off for security purposes). Dread Lord CyberSkull ✎☠ 06:07, 5 October 2005 (UTC)

[edit] XHTML 1.1, 2.0

I've removed the XHTML 2.0 column; the specification for it is unstable, there's no DTD, and there's no information for browser support in the table.

I'm not sure of the purpose of the XHTML 1.1 column; there's a few non-sequitors there (IE/Mac supports XHTML 1.1 but not XHTML 1.0?). Are there any browsers which support XHTML 1.0 but not 1.1? There probably aren't any browsers that support 1.1 but not 1.0. — Miles←☎ 22:35, 19 March 2006 (UTC)

I would like to see the XHTML information broken down even further: XHTML 1.0 Strict, XHTML 1.0 Transitional, etc.

[edit] PICS (Platform for Internet Content Selection)

I'd like to add a column showing support for PICS: http://www.w3.org/PICS/. Self labelling is a healthy act which decreases government regulation by self censorship. I know IE supports it, but can't find support in any other brower. Anyone know?

[edit] Inclusion/Removal of Browsers

I can't believe Arachne (Web Browser for DOS) is not listed. As far as I know, it is still being developed. It's really cool to be able to put an old DOS workhorse on the Intarweb without loading windows. There is also a version for Linux. There was a new release as recently as November 2005 (1.89, GPL).

What about the mozilla foundation's camino? i'm not good with tables, but can someone put it in, it's pretty popular. [maestro] 05:08, 12 Sep 2004 (UTC)

Are Lynx, Links and Dillo major web browsers? I believe more people use Camino than these three together. --Cantus 05:14, Sep 12, 2004 (UTC)

I think Lynx and Links should stay as major text web browsers. Camino and Dillo are rather minor, though. Dreamyshade 06:15, 12 Sep 2004 (UTC)

Links itself is not actively being developed where as ELinks is. Links2 and Links Hacked both have graphical modes. Links Hacked also has tabbed browsing, better support for CSS, JavaScript support, ect in it. Dillo and hacked Links are both often tossed into the Linux Flashcard/Pendrive/mini-Cd Distros and tend to be used in embedded devices as both are very light browsers (Dillo is something like 350 Kb in size).

Cantus you are invited to #wikipedia @ Wikipedia:IRC channels --Yath 06:07, 12 Sep 2004 (UTC)

Lynx is a survivor. Links was a neat hack, often rendering pages better than Lynx, but way too buggy. No wonder links isn't actively developed anymore. Lynx deserves to be in a browser comparison, but shouldn't be in comparison tables for modern features. 65.34.186.143 15:26, 18 Jun 2005 (UTC)

In general, current and modern browsers should not have to share space with old and dead ones. One minor improvement would be to sort all the tables by layout engine. Better would be to exclude all those without CSS2, frames, JavaScript, XSLT, and PNG. (all required for modern web browsing) Maybe also require XHTML, SVG, FTP, and SSL. 65.34.186.143 15:26, 18 Jun 2005 (UTC)

Minor browsers like Dillo is clearly not worth mentioning. Ancient browers like Mosaic and WorldWideWeb, while notable, are too outdated. They can't be compared with Gecko/KHTML. --Minghong 18:33, 9 Mar 2005 (UTC)

Dillo is worth mentioning; it's one of the two graphical truly lightweight open source browsers for Linux (the other one being Links). Now, it is true that Dillo has some rendering issues with tables that web designers should just ignore, but the last release of Dillo was on January 11 of this year, so the program is being regularily updated. I think that open source browsers can be considered for removedal (or not listed in the first place) when they haven't been updated for a year or two; there's a reason we don't have BrowseX or Grail listed here, for example. Samboy 20:10, 9 Mar 2005 (UTC)
The point is that browsers like Dillo and X-Smiles is missing too much fundamental features, making them nearly not usable. Mosaic and WorldWideWeb, other than historial influence, are also as not usable as Dillo and X-Smiles. --Minghong 10:12, 10 Mar 2005 (UTC)
I think the comparison should make that clear; I don't think we should just remove them from the list as long as the free software writers continue releasing new versions of the software. Samboy 10:17, 10 Mar 2005 (UTC)
I don't think it is a problem having outdated or minor browsers listed here. Paul August ☎ 20:31, Mar 9, 2005 (UTC)

I think in browser comparison are many obscure, non-modern, non-used and prehistoric browsers - pointlessly :( Now in this comparison is everything hardly finded. I think rather we should comparise actual, modern, well-knowen, latest vesions of browsers (IE 6.0, MF 1.0.x, O 8.x, N 8.x, K 3.4.x, S 2.x; maybe even Epiphany, K-Meleon or Galleon). Anything more is nonsense. Every (obscure, prehistoric and modern) browser on the world belong to the list of browsers, but no in this comparison. --Ptomes 14:44, 20 Jun 2005 (UTC)

I asked this question before and the answer was no. --minghong 1 July 2005 10:23 (UTC)
I don't know... I agree that there are just too many browsers that aren't appropriate for comparison. But really, the only way I could think of fixing this would be two separate pages. One with just the 6 major browsers, and one with the more obscure browsers. --Ctachme 3 July 2005 00:47 (UTC)
I like this idea. But main article must be that with major browsers with link to another comparison with more browsers. --Ptomes 3 July 2005 06:23 (UTC)
I agree with the general idea that too many browsers are listed and that we might do well too differentiate the major browsers: IE, Firefox, Opera, Safari, Konqueror, Netscape from the rest. But if we choose to remove any browser we'll just get fans of that browser in an uproar. I think defining certain criteria to be on this list might be a good idea: a) only the latest version of any particular browser is listed (obvious), b) browser has had SOME type of update in the last 24 months, c) has an established online community (need help with this definition, at least two independent fan sites?). Jeff schiller 6 July 2005 21:25 (UTC)
If we set any sort of standard we may end up with more browsers on the list and not less. We need a set of standards that will allow you to use 95%+ of the web. Fancy visual things like CSS are not required. Dyson's Dragon 03:31, 1 December 2005 (UTC)

Netscape Communicator is missing from the browser list. Someone trying to find the features of Netscape 4.8 will wonder why that version of Netscape isn't listed. -- Schapel 05:06, 29 December 2005 (UTC)

Just in case, here are two sites with lots of info. Mac Broswsers and Evolt. Alex 04:19, 2 March 2006 (UTC)

Iceweasel and Gnuzilla are missing from the table. 70.111.218.254 02:38, 25 October 2006 (UTC)

[edit] Other issues

[edit] Cost of browsers packaged with OSes

At the risk of dragging this issue around the block a few more times, I am having trouble finding a free download of MSIE on the IE page. I see service pack downloads, and cheap CDs with service packs, but no browsers. Can someone find the free MSIE? --Yath 04:08, 12 Sep 2004 (UTC)

MSIE for Windows would be an upgrade unless for versions of Windows pre-98, which I'm pretty sure IE isn't made for anymore, so...
James F. (talk) 04:16, 12 Sep 2004 (UTC)
If you look at the IE page you linked to, click on the Get Internet Explorer 6 Service Pack 1 (SP1) link. On the subsequent page, it says "This version includes a full installation of the Web browser and the most recent version of Outlook Express, the e-mail client that is included with Internet Explorer." It seems to me that this is indeed a full browser download and not a service pack download; it's a bit confusing because the words "Service Pack" appear in the name. --timc | Talk 13:43, 15 Sep 2004 (UTC)
I don't think you can obtain Internet Explorer 6 SP 2 separately, only as part of Windows XP SP 2. If this is the case, then truly IE isn't free—you must pay for Windows XP to get the latest IE. Safari and all the other browsers listed as free can be downloaded and installed without paying money, so I think the cost IE should be marked as "part of Windows" -- Schapel 01:50, 17 August 2005 (UTC)
I'm pretty sure the license for IE forbids you using it without having a Windows license, so whether or not there is a place you can download it from isn't that relevant to the cost involved in using it legally. You may have a non-Windows platform that is capable of running IE (such as Wine on Linux), but you may not use it without getting a Windows license. --Ecoffey 11:44, 27 September 2006 (UTC)
I believe you can't say IE is free. It's bundled with windows and that isn't free. MS might say that Windows itself costs $XY and IE is just a free bonus, but that can't be farther from true. The cost of windows package is a sum of all its components, including Windows Media Player and IE. The problem is we don't know the price for each product separately (there's no retail price for IE). I suppose it'll be fair to put a question mark instead of "free" and provide a footnote explaining the situation. Jancikotuc 16:13, 25 October 2006 (UTC)

Someone removed the cost of Mac OS X now (Safari's price was listed as "Part of Mac OS X ($129)", I believe). I agree with this to some extent, although I'm not sure if it shows that it's not free good enough... But if it's agreed that it should go away, what about IE? --Spug 07:44, 2 Nov 2004 (UTC)

I think entries here for Safari and IE should mention the price of the OS, because the fact is that many internet users are running outdated versions of these browsers for the sole reason that they haven't paid for an OS upgrade recently. If these browsers don't have a price listed, I don't see why we should bother to have a price row in the table at all. I'm not just saying you have to have the OS to run the program; that obviously applies to any browser, and doesn't mean the price is relevant. The reason it is a relevant issue with Apple and Microsoft is that relatively recent versions of their OSes (which are still in wide use) can't run the latest version of these browsers, and the upgrades cost money. I think the prices are important facts to include in the article, but in the interest of consensus I won't revert my addition of the Safari price again until others have had a chance to respond here with reasons not to. One further piece of evidence here is that apple.com/safari has a "Sneak Preview: Tiger" link that references the upcoming version of the browser and OS - which again will cost $129. I use Safari today, and I will have to pay this price (or be a criminal/pirate) to use the latest version of it in the near future. When comparing web browsers, the $129 price of Safari is relevant and should be included here. ~leif ☺ HELO 07:31, 4 Nov 2004 (UTC)
In response to the query at my talk page, I don't plan to do any more editing or reverting of this article. func(talk) 04:45, 6 Nov 2004 (UTC)

What about a footnote at the end of the table? Even though the amount shown should make it obvious, it isn't nessesarily clear that the $129 in "Part of Mac OS X ($129)" refers to the cost of the OS rather than the browser. In a big long table like this, (which doesn't fully fit into my browser window), we should look for ways to make the table smaller, I think. ______

Agreeing with this point, I have edited the mention of "free" for the browsers IE and Safari as being 0*, with an additionnal note at the bottom of the table indicating that you can't have the latest version of the browser legally without paying any fees. --Dalzhim

I'd have to say 'part of O/S' would be the best label. Integrated-IE is only availible for windows so you wouldn't be considering it if you didn't already have the operating system so it is free. Presumably it's the same story with safari. And if IE for mac is free I'm not sure the comment about buying the OS pays for IE is valid.
As to the upgrade situation I feel it's a non-issue. If you want to get the latest version of a non-integrated free browser and it only runs on a newer OS you'll still have to buy the OS. Dyson's Dragon 03:31, 1 December 2005 (UTC)

[edit] Discontinued/Dropped

Shouldn't MSIE have "Discont. 2001" in the box for Windows? It's no longer available and has since become part of the Windows OS. Nicholas 21:53, 6 Oct 2004 (UTC)

Maybe if we had different boxes for WindowsXP, Windows 2000, Windows 98 etc. However Microsoft are still creating new versions and releasing them. Okay the most recent one was part of a service pack, but it is still being updated and made available to Windows users (granted only a subset of them). AlistairMcMillan 13:16, 7 Oct 2004 (UTC)

For several entries it says "Dropped", but there is no definition of what that means. For example, it says IE for UNIX has been dropped, but under IE for Mac, it still says "yes" for both OSX and OS9, while according to this article (dated June 13, 2003) "Microsoft on Friday officially dropped development of Internet Explorer for Macintosh. The company's Macintosh Business Unit (MacBU) will release an update for the OS X version of the browser on Monday that will have security and performance enhancements, but no new features. Microsoft said it would continue to support Internet Explorer 5 for the Mac for the foreseeable future, but development of the browser has been discontinued. Any future updates to the browser will only contain maintenance or security fixes, no new features. "Support for IE 5 is not ending, but we will no longer be developing future versions of Internet Explorer," Jessica Sommer Product Manager for MacBU told MacCentral." Please explicitly define dropped on the page and if applicable, update the IE for Mac entries. -- Limulus 10:31, 25 July 2005 (UTC)

I recommend that the term dropped be defined as the latest stable version of the browser does not work on the operating system without emulation, although an older version is available that does. -- Schapel 12:36, 25 July 2005 (UTC)
How about simply, "the browser was, but is no longer being officially developed for that OS"? If we set this as the definition, IE for Mac is "dropped". Also, this could leave the table open in the future to entries such as "unofficially" (e.g. if someone compiled Firefox for OS9). -- Limulus 23:56, 11 August 2005 (UTC)
That's the problem with that definition of "dropped". If a browser is no longer being developed, it will be listed as "dropped" for all OSes it ever ran on, with no indication if the latest release, perhaps still available and used by millions (e.g. IE for Mac and Netscape Navigator), will run on a particular OS. Another problem with the definion is that it depends on what your definition of the term "officially" is. If the BSD builds of Firefox are contributed by a third party to Mozilla, does that mean Firefox is not "officially" developed for BSD? -- Schapel 00:55, 12 August 2005 (UTC)
How does that imply that it won't run on a dropped system as of the last version numebr shown? Also, regarding BSD, I would say so since [the Mozilla download page http://www.mozilla.com/firefox/all.html] only lists Windows, Mac and Linux. --Limulus 18:55, 15 July 2006 (UTC)
To fix the BSD issue, I've changed the wording to read "The operating systems the browsers have been compiled to run on without emulation" thus for open source projects, an unofficial build can be considered :) For closed source projects, like IE, note that it was compiled against MS libraries and not Wine libraries, so it should not be considered for inclusion under the Linux column ;) -- 70.64.95.83 05:01, 16 July 2006 (UTC)
I've taken the opportunity to change the tables in the imho only sensible way: The way Schapel described it. I also changed the definition of "dropped" for the legend needs to be consistent with the data. I merely copied the definition out of this discussion page, though, so credit goes to Schapel. Grey 20:25, 20 June 2006 (UTC)
IE for Mac has really been dropped and needs to reflect that :) I will alter the definition slight to try to accomodate your concerns. --Limulus 18:55, 15 July 2006 (UTC)
I decided the beast way to deal with this was to make a new option, "Terminated" :) It means that the most recent version of the browser is very much the last version of the browser ;) Note that dropped is still valid. --19:49, 15 July 2006 (UTC)

From Microsoft's IE for Mac page: "In accordance with published support lifecycle policies, Microsoft will end support for Internet Explorer for Mac on December 31st, 2005, and will provide no further security or performance updates." -- Limulus 08:20, 19 December 2005 (UTC)

Oh and could someone explicitly define "Partial" too? :) -- Limulus 10:04, 19 December 2005 (UTC)

For the time being I'm just going to write "Partial indicates that the browser lacks certain functionality compared to versions for other OSs." If you have a better definition, please change it ;) -- Limulus 04:34, 20 December 2005 (UTC)

I'm having a lot of trouble finding information about IE for UNIX; I managed to find cached copy of the old MS site before it was pulled, but it doesn't give much info... Was the UNIX version really based on Trident, or was it based on Tasman? AFAIK, OSX is a UNIX derivative, but OS9 wasn't... -- Limulus 03:05, 20 December 2005 (UTC)

I started an Internet Explorer for UNIX page; based on what I've found so far, it was a direct port from Windows IE 4 and 5 (unlike Internet Explorer for Mac which started out with IE 2, so prior to Trident, which wasn't introduced until IE 4), so I'm fairly convinced it was Trident in IE for UNIX... -- Limulus 04:50, 21 December 2005 (UTC)

[edit] Download managing

As all (or most) browsers can download file, what's the point of having this column? Traditional single-window download window can also be considered as "download manager". I guess dropping this column and have some other distinguish features instead. --minghong 19:23, 28 Apr 2005 (UTC)

Download manager is what download manager does. Quote from the article:
  • Pausing the downloading of large files.
  • Resuming broken or paused downloads (especially for large files, like Linux distributions).
  • Downloading files on poor connections.
  • Downloading several files from a site automatically according to simple rules (file types, updated files, etc.).
  • Automatic recursive downloads (mirroring).
  • Scheduled downloads (including, automatic hang-up and shutdown).
  • Searching for mirror sites, and the handling of different connections to download the same file more quickly.
MSIE, for example, can't do anything from the above list, so by any meaningful definition it doesn't have a DM. Firefox can only pause and resume downloads, but not across sessions; does that qualify? I think DM is an important feature, and that it should be listed, but we should be more strict concerning its definition. GregorB 14:46, Apr 29, 2005 (UTC)
Agreed. MSIE gets a no. Mozilla could probably get a partial with a footnote. If there's ambiguity in the definition, we call the column "robust download manager". Is there an article for download manager?--Will2k 15:01, Apr 29, 2005 (UTC)
Yes, it's linked above. GregorB 15:52, Apr 29, 2005 (UTC)
Yup. Go ahead: add a footnote explaining our definition of "download managing" (1. cross-session; 2. pause/resume). Fragment download may not be a MUST for download managing. Add a partial footnote for Moz if you want. That would be great. --minghong 09:48, 30 Apr 2005 (UTC)

[edit] Mac OS9

There are still plenty of OS 9 users on the Web (those who can't afford OS X or don't want it). So, let's keep the OS 9 column. --KelisFan2K5 21:42, 13 Mar 2005 (UTC)

[edit] Multiple Netscapes

There has got to be a way to clarify the difference between Netscape, Netscape Navigator, and Netscape Browser. I can figure it out since I've been following browser development since 1995, but even I had to think about it when looking at some of the tables. I do realize that the dates are described in the General table, but given that the average user only knows all three as "Netscape" (and NS4 was referred to, in full suite form, as Netscape Communicator) it seems confusing to list them only with the current names. Unfortunately, the best I can come up with is to add "1-4" "6-7" and "8+" to the labels, and that just seems too cluttered. --Kelson 00:28, August 5, 2005 (UTC)

Yes, there can be only one = Netscape Browser. All other is old obsolete versions not supported anymore. Another core, name change? So what?

They're three different browsers which have the same name. There are still people using the old Netscapes (just look at browser stat sites), and especially the new Netscape Browser with its Unholy Alliance of Gecko and Trident cannot and should not be compared with other variants of Netscape. I hope the note and version numbers I added clarifies the matter some. Jordi·✆ 19:28, 7 August 2005 (UTC)

[edit] Operating System support

Why Windows has to be the first column in the operating systems support table? Why not alphabetical order? utcursch 06:27, Dec 27, 2004 (UTC)

No reason, got ahead and change it. And change the other comparison pages too since I think they may be the same. Good point. --Ctachme 21:13, 6 Jan 2005 (UTC)

WorldWideWeb (NEXTSTEP only) - Is not listed as being avalible for unix, NEXTSTEP is a unix. Therefor shouldn't this be a Dropped ( Release Version ) or a Yes in the MacOSX tab if it still exists ?

[edit] Security section based on unpatched Secunia Advisories

What do You think about add "Comparison of Security" based on Secunia Advisories: www.secunia.com

Secunia advisories marked as "Unpatched": MSIE: 19 Opera: 0 FireFox: 5 ...

(This is actual data by the way.)

Object: Secunia doesn't cover every single security issues (e.g. how can Opera has zero security holes? that's highly impossible). Also it is hard to compare. Showing the number of holes are not meaningful. --minghong 08:36, 12 May 2005 (UTC)
Well, Opera is extremely fast about fixing security issues, and historically has had very few. So "zero (known) security holes" is very possible. Still I object, as the number of open exploits is really not meaningful. For example MSIE has 19 open issues which are in many cases months old, which means that they are very likely to be exploited. Mozilla/Firefox/etc. has five, most of which are trivial, and are very recent. This means they are far less likely to be exploited than the old MSIE ones (which also affect other software). Jordi·✆ 13:28, 21 May 2005 (UTC)
Compare for yourself:
Perhaps we should include in this article links to an article on computer security. Jordi·✆ 13:38, 21 May 2005 (UTC)
Using Secunia for security comparison isn't a bad idea (I do that all the time on my blog ;), but you need to be careful in how you present the data; consider that 5 'non-critical' issues do not trump a single 'extremely critical' one. I would say that if you present the data, either explicitly list the number of flaws at each level in a table (to avoid an 'apples and oranges' problem) or just list the worst level that currently affects it (e.g. for Firefox right now that's "Less critical" while for IE its "Highly critical"... you could even use a color-based scheme similar to Secunia's; blue=none, dark green=not, light green=less, yellow=moderately, orange=highly and red=extremely critical). Also, you can get around the 'zero holes' problem by explicitly stating that the table covers "publically known vulnerabilities", e.g. it could be that Opera is sitting on a security issue or two that it won't make public until it gets fixed. --Limulus 10:16, 25 July 2005 (UTC)

How reliable is Secunia? Wouldn't quoting from several sources be better? Dyson's Dragon 03:31, 1 December 2005 (UTC)

There are also public vulnerabilities that have not been acknowledged by the vendor. I am not sure if Secunia lists these, but in fairness they should be. See http://news.yahoo.com/s/cmp/20060308/tc_cmp/181501722;_ylt=AhJkTGJ9tXH5Y5Q8lrKhf1kjtBAF;_ylu=X3oDMTA5aHJvMDdwBHNlYwN5bmNhdA--

[edit] Web browser or web browser

This is an outstanding article, short and to the point, and I'll strive to get every browser article to link to it. However, I actively wonder why the "W" in "web" is capitalized in the title? The term "web browser" is an improper noun like any other. You capitalize the "W" when referring to the WWW as a proper noun, not when you're merely using "web" as an adjective. --Ardonik.talk() 01:40, Sep 15, 2004 (UTC)

I think that "Web" is capitalized because it is part of "World Wide Web". On the other hand, web browser uses lowercase "web". Personally, I think that it is a proper noun like "Internet". Since web browser uses a lowercase W, this article probably should as well. --timc | Talk 13:39, 15 Sep 2004 (UTC)
I can't copy & paste back to Comparison of web browsers because that would destroy attribution. Would anyone be opposed to a move? (Is a move even possible when the destination already exists?) --Ardonik.talk() 17:58, Sep 15, 2004 (UTC)
  • All relevant browsers seem to already be linked. --Ardonik.talk() 01:44, Sep 15, 2004 (UTC)
When the target page has a history it has to be deleted to make way for the move. As the article on web browser does indeed use a lower-case W - this is where the page should be until that changes. (It should be possible to move it back and forth now - provided no one edits the redirects). ed g2s • talk 18:33, 15 Sep 2004 (UTC)
Thanks for performing the move. --Ardonik.talk() 19:33, Sep 15, 2004 (UTC)
Gah gah gah gah gah.
See [5].
I will move this back, and correct Web browser, in about 12 hours, unless someone gives me a reason not to.
James F. (talk) 21:48, 15 Sep 2004 (UTC)
By all means, that appears to be the correct way. My move was only to create consistency with web browser. ed g2s • talk 21:54, 15 Sep 2004 (UTC)
Do not revert. When you're talking about the World Wide Web, of course you capitalize, but you don't capitalize the "W" in "Web" when it is used as an adjective, any more than you'd capitalize the "f" in "french bread." Have you seen the terms "web spider" or "web crawl" use capitalized Ws? What makes "web browser" so special? I think you're almost totally in the wrong about this. --Ardonik.talk() 00:55, Sep 16, 2004 (UTC)
Web in this context is not used an adjective, the grammatical structure is like "France lover" (lover of France) rather than "french lover". There is an issue however of whether you are referring to a web in general - or the Web (WWW), depending on which you then interpret so "web browser" has a different meaning to "Web browser", the same way a web is different from the Web. The question is then which one best describes these programs? ed g2s • talk 02:51, 16 Sep 2004 (UTC)
I suppose that, yes, this page could be about "web browsers" - that is, programs that work on HTML in general, and not necessarily over the Internet. However, this page is in reality about "Web browsers" - that is, "web browsers" which do use the Internet.
Yes, I have seen "Web browser", "Web surfing", "Web spider", and so on, and I pity you for the lack of grammar amongst the people around you that you have not.
I will hold-off the move, however, whilst discussion continues on apace.
James F. (talk) 09:20, 16 Sep 2004 (UTC)

[edit] What is Page Zooming?

I would like to know. I changed Page Zooming to Full Screen, but was reverted with the argument that it was something entirely different. I need some explanation. —Cantus…☎ 04:40, Feb 10, 2005 (UTC)

Browsers like Opera have the ability to Zoom into all of the elements on a page. I.E. instead of just making the text bigger (like most browsers do), it makes EVERYTHING bigger, like Images too. Maybe we should have a footnote explaining this. --Ctachme 06:16, 10 Feb 2005 (UTC)
Ctachme explained it well: Opera's zoom zooms in everything, including images, border widths and thicknesses, form buttons, scrollbars, line-based plug-ins (Flash and SVG, some QuickTime) etc.. MSIE/Mozilla's zoom just picks a larger text size.
As far as fullscreen is concerned, this is also confusing: while all (nearly?) all graphical browsers have a 'fullscreen' mode which hides interface elements, as far as I am aware only Opera actually uses a presentation CSS media profile in its fullscreen mode. User:Ana'rion/sig 18:02, 11 Feb 2005 (UTC)

What is "Full Page Zoom"? If it's expanding the window to occupy the entire screen so ONLY the webpage is displayed, Firefox can do this as well.--Will2k 17:23, Jan 6, 2005 (UTC)

When I added "Full Page Zoom" I meant the ability to have the page itself zoom in and out in size. Most browers only zoom the text, not the images and styles, as is possible in Opera. However, I see how this could be ambiguous, and if you could think of some better title for that feature, change it. As for Full screen, that would be a cool feature to add too.--Ctachme 21:19, 6 Jan 2005 (UTC)

[edit] Page Layout Discussion

Because of it's length I felt that the discussion about the Page layout should be moved to a seperate page.

[edit] Validity

For the chart for critical problems on the browsers, the info for IE is very old, 2003. Microsoft has released several patches for IE6. Please update the numbers. 128.6.175.40 14:34, 16 December 2005 (UTC)

Why don't you update the numbers? -- Schapel 20:24, 16 December 2005 (UTC)

[edit] iCab and Gopher

The table says it does, but iCab does not list that in its preferences of protocols it can handle, it opens in a different application. Benn Newman 17:10, 29 December 2005 (UTC)

Same for elinks, it supports using an external program as best as I could tell. Benn Newman 04:31, 2 January 2006 (UTC)

[edit] Vulnerbilities in IE

There is a huge disparity between the vulnerbility count in this list and the one in the actual pages. For example, the SecurityFocus for IE6SP1 lists 179 vulnerbilities while Wikipedia lists 92. With such a large disparity, and because Wikipedia's count hasn't changed much lately, I thought it likely that there must be some subtlety in the count. Could someone with more knowledge double check to make sure this is correct, and then change if appropriate (I changed all the other browsers already). Telso 22:39, 17 February 2006 (UTC)

Maybe could we have the percent market share or number of users of each browser? I think that might be informational alongside the number of voulnerabilities. Whatcha think? -Mysekurity[m!] 06:17, 3 June 2006 (UTC)
"There is a huge disparity between the vulnerbility count in this list and the one in the actual pages. ... I changed all the other browsers already" The problem with your statement is that you don't know what are you writing about. The table of vulnerabilities should not count fixed vulberabilities, but only those known and unpatched. --Ptomes 09:05, 10 June 2006 (UTC)

[edit] Layout suggestion

Why do all the (very many) features available only with extensions/plugins get "no" in the main box and then a footnote mentioning it? It would seem more sensible to color the box yellow or orange and have the word "plugin" or "extension" plus a link to the item in question. After all, part of browser functionality is moddability. (Or maybe I'm just a Firefox fanboy.  ;)) —Simetrical (talk • contribs) 03:50, 27 February 2006 (UTC)

If some way can be found to do that, I'm in support of it. A yellow/maybe/sorta is clearer, IMO, than a green/yes or red/no with a footnote. ⇔ ChristTrekker 18:34, 22 December 2006 (UTC)

[edit] Update on MSIE 7.0

I find that this page is not updated as Microsoft Internet Explorer entries everywhere correspond to MSIE 6.0

Now as MSIE 7.0 is also in the market, shouldn't its data be included in the list? One example that I can give is that MSIE 7.0 has tabbed browsing support, while here it says not supported. -Ambuj Saxena (talk) 11:57, 5 March 2006 (UTC)

[edit] This article is being considered for deletion

consider improving the article so that it meets the Wikipedia inclusion criteria.

Hmmm, I would consider doing so, if only I could find what the Wikipedia inclusion criteria are! This article does not seem to fit the description of what Wikipedia is not in Wikipedia:Policies_and_guidelines. Maybe someone could clarify why the article is being considered for deletion, or how to improve the article so that it would not be considered for deletion? -- Schapel 16:57, 7 March 2006 (UTC)

Ah, but isn't that the point? This article can never meet Wikipedia inclusion criteria, which is why it should be deleted. —donhalcon╤ 17:42, 7 March 2006 (UTC)
You've completely lost me. What is the Wikipedia inclusion criteria? I can't find it anywhere. All I can find is Wikipedia:What Wikipedia is not, and this article isn't any of the things mentioned there. Can you give a simple, reasonable response to the question Why is this article considered for deletion? -- Schapel 18:30, 7 March 2006 (UTC)
His reasoning can be found at his user page. --Kamasutra 18:44, 7 March 2006 (UTC)
Yes, I just found his opinion that comparisons are not encyclopedic. There doesn't seem to be any sort of consensus on this point, because I can't find anywhere else that states that comparison pages don't belong in Wikipedia. Perhaps we should reach a consensus on this point first before doing anything else. -- Schapel 18:53, 7 March 2006 (UTC)
I'm not saying that all comparisons are unencyclopedic, mind you. Just the stand-alone tables that don't have the kind of context you would expect to see in a professionally edited encyclopedia. —donhalcon╤ 20:17, 7 March 2006 (UTC)
That's what I was just thinking — these tables should probably be put in the web browser article in sections entitled something like matrix of technologies supported by various browsers and matrix of security vulnerabilities in browsers. The General information, Release history, and Operating system support information probably belong in the Internet Explorer, Mozilla Firefox, etc. articles. And then when the information is put into the approapriate places, then this article should be deleted -- Schapel 15:52, 8 March 2006 (UTC)
It appears that if you read donhalcon's userpage, he has given up on his campain against comparisons and decided to leave wikipedia altogether. Perhaps it would be worth reverting the proposal to delete? This is a highly seen (and even a former featured page) --ZacBowling 02:50, 9 March 2006 (UTC)
It sounds like he wanted to boss around others to make the changes that he wanted, rather than do the dirty work himself. We should just remove his proposals for deletion. I'll try to get a real consensus on what we agreed upon (articles should not consist solely of tables -- tables should be sections in legitimate encyclopedia articles) and make it official policy. Does anyone know of an existing official Wiki policy like that? -- Schapel 14:59, 9 March 2006 (UTC)
There is no such policy as far as I know, but if you want to propose anything do so at Wikipedia:Centralized discussion. He was correct in regards to the comparisons not being encyclopedic, but mass deletion requests was the wrong way to go about it. For tabled ones in particular, such as this, there is conflict with at least one of Wikipedia's policies which I found. Many of the attributes are not verifiable via WP:Verify, and hence one must resort to WP:NOR. Your proposal alone would not rectify this and would unfortunately serve to make the Web browser article longer and more unwieldy. A better solution, I think, would be to put each browser's attributes into its own respective article AND remove all unverifiable information. This seems to be a rather daunting task and one would soon find that there is relatively little verifiable material due to unthorough, and in most cases just plain poor, documentation. --Kamasutra 17:20, 9 March 2006 (UTC)
If you look at Wikipedia:Arbitration_policy/Precedents#Decisions_on_case_acceptance, the arbitration policy does have a clause to close a subject if the person leaves and doesn't contribute to wikipedia. Maybe it might fall under that? --ZacBowling 07:57, 10 March 2006 (UTC)
Surely we can verify which of the major popular browsers support which features, with no original research. I added the TIFF column to the Image format support table, and did so with no original research — I found which browsers supported TIFF using web searches. For the more popular browsers, we can find the data in these tables in any number of printed books on web browsers. I think that if you attempted to verify the data in the tables, you could easily verify it for the major browsers. A better solution, I think, would be to remove the minor browsers whose features are poorly documented. -- Schapel 16:37, 10 March 2006 (UTC)

Please keep "Comparison of" articles in Wikipedia! They are of great use to wiki users/readers. Forcing the info to be split up would be a dis-service, inefficient, and force fragmentary conflicting POV pieces, with no attempted summary/neutral ground. (And don't force such articles to have weird names that people would not find/understand.) 69.87.204.177 23:25, 20 June 2006 (UTC)

[edit] IE is paid for by the purchase of the OS

Is this necessary? It seems that you have to purchase an OS to use a browser, so that any browser that is utilized under only one OS has a prerequisite for use that OS.

It can be downloaded for free, so it's free. It just won't work on something other than Windows Sceptre (Talk) 20:00, 20 March 2006 (UTC)
It'll also work on ReactOS (for a suitably broad definition of 'work') Wayne Hardman 22:06, 12 April 2006 (UTC)
"It can be downloaded for free, so it's free." Have you read the EULA?

[edit] Normalizing Vulnerability Data

In order for the vulnerability data to be considered strong from an argument standpoint, the data should be normalized such that readers can perform an apples to apples comparison. Each vulnerability category should be adjusted to reflect the number of users using the given browser for a "Vulnerabilities per User" statistic. After all, it is the open source mantra that "Many eyes make all bugs shallow" and I would venture to say that having many users is the equivalent of many eyes seeing the surface of the app, so it would also be fair to assume that there is a marked increase in vulnerability discovery capability, quickly inflating figures for more popular browsers.

Rubbish. I thought it was long ago settled that "vulnerabilities per user" is reducible a mere appeal to popularity fallacy. Sometime back when it was cool to trash MS, I think. We're meant to be comparing the software itself, not who uses it or how many there are of them. Wayne Hardman 22:16, 12 April 2006 (UTC)

[edit] Bias Towards Open-Source

Because of the Yes/No template usage on the Open Source column, this article appears to have bias towards open source. The Yes/No templates highlight Yes in green and No in red, seeming to indicate that non-open source implementations are somehow inferior. Though this may not be the intent, it seems to highlight this feature as somehow positive or negative. Wikipedia should not choose sides in this issue.

Good point. Would it be an adequate solution to simply use "Yes" or "No" without using the template? ⇔ ChristTrekker 18:30, 22 December 2006 (UTC)
Why not use this space to specify the license, linking to proprietary, GPL, and so forth? A column indicating whether the source is open or closed is going to come off with an OSS bias no matter what. Simply linking to the article describing the license style gives information that is both more neutral and more accurate. —ptk✰fgs 04:35, 3 February 2007 (UTC)
I'm all for that, as I noted in the layout engine article. I think it would be best to remove the column entirely, since there already is a column for the license. --Grey 23:46, 3 February 2007 (UTC)
As per discussion on the template page, green merely indicates "yes" and red merely indicates "no." There are NO value judgements--only presentation of facts. As such, I've restored the use of temlates. --Karnesky 00:13, 13 March 2007 (UTC)
This information is clearly misleading. Whatever you conclude it SHOULD mean, to most people, green will mean "good" and red "not good". That's like changing traffic light laws so red is "go" and green is "stop". Also, the column "Open Source" is unnecessary. More detailed and accurate information is given in the "license" column. The fact is, people will judge the products according to the colors. I will remove the templates once more, and what's more I will remove the "open source" column, since there has not been one argument/reason why it should be kept (even if I'll be doomed forever). I would ask you to give a good explanation before reverting anything. --Grey 01:24, 13 March 2007 (UTC)
I strongly object to removing the column for this particular page, as it clarifies the multi-licensed products. As for "bias," please take up your case on Template talk:Yes‎. Most other comparisons ONLY use green for yes and red for no (including for "free software" or "open source" columns). If you see bias, it brings into question the use of the template for every other feature. You're making am implicit value judgement that browsers SHOULD HAVE download managers, password managers, form managers, spell checkers, etc. Others would want a browser which is lightweight and didn't have these. "Open source" is no different than any other yes/no column. --Karnesky 01:34, 13 March 2007 (UTC)
For that column, the licsensing column gives much more and more detailed information. And without it, the licensing models are not less confusing than with it. I therefore propose that we add explanatory information to the licensing column. And yes, it is an implicit judgement that browsers SHOULD HAVE these features. Why do you think proprietary technologies have not been listed? Because the editors are of the opinion that they shouldn't be (among other reasons, on the grounds that usually only one layout engines possesses these "features").
On that template page, take the example of the "non-poisonous" fish. The wording of the column, "non-poisonous" or "poisonous" expresses a view on the "qualities" or "benefits" of a particular kin of fish (and thus that another has "less" benefits etc.). Now, with "poisonous", the majority of mankind prefers not to be poisoned, thus this expresses quite a broad view. With "open source" the plus/minus is highly disputed. Sympathising with one or the other would be POV.
Maybe a link to a "feature creep" article would help overcome this bias? I won't revert your changes again (at least not for considerable time). I don't think the templates belong there, though, and I am positive someone else will change it soon enough.
--Grey 03:01, 13 March 2007 (UTC)
I'd support adding better explanations for the licensing column (particularly for all multi-licensed products). Can you do that BEFORE deleting the open source column? If it is clear enough, I wouldn't be opposed from just removing the "open source" column as you propose.
I still disagree that we should be saying that browsers should have particular features. Certainly, it seems there is no consensus at all to have bittorrent support or an integrated email/newsgroup client.
I'm not sure which proprietary technologies you are referring to. If only one or two browsers supports them, they'd clutter up the table if they were whole columns. I think an "other" column would be perfectly acceptable.
The poison fish example is trite & there seems to be consensus that largely rebels against that single wikipedian's view. Feel free to comment on that page, though.
Karnesky 03:15, 13 March 2007 (UTC)
Regarding extending the license column before removing the other... agreed. I wanted to do something about it anyways. Any suggestions? I thought that "tri-license" could be a footnote explaining what is going on. Generally, we could make the licenses footnotes and quickly explain them in the article. Or something like that. What do you think?
Generally, though, I think features are considered to be something good. It's the side-effects of implementing them (e.g. size, complexity) that are considered "bad" by people. These should be pointed out. I wonder how that format would look like, though. Many of these are hard to be verified, also.
Concerning proprietary technologies/features... they're often covered by patents, copyright, a ristrictive license (including GPL), which prevents competitors from implementation. That's why there are web standards... well, in theory.
The fish example was nothing more than an example. I'm not in favor of "but-not" etc. but my intent was to leave that discussion there, not here.
--Grey 03:37, 13 March 2007 (UTC)

(arbitrarily starting a new, unindented block for readability)

One suggestion I have for improving readability is that any proprietary products which have some material under an open source license should be listed as "proprietary," but with a footnote. The footnote for any multi-licensed products should describe which licenses apply to which parts of the code. I believe that some of these have EVERYTHING is under the user's choice of one of three licenses & that some of libraries/renderers/other backends which are open, but front-ends which are proprietary. --Karnesky 04:26, 13 March 2007 (UTC)

So Safari would be "proprietary"? I don't know if I 100% approve of this. If we were to do it, though, it would be smart to arrange the footnotes according to the rendering engine (atm only Gecko and KHTML/WebKit, imho). Also, what about the "protected" pictures in Firefox (i.e. the reason IceWeasle (sp?) exists). Does that make it proprietary? --Grey 00:17, 14 March 2007 (UTC)
Safari IS proprietary, yes. That is why there is a "no" in the "open source" column and why the wiki article is not in the free software category. Yes, the backend (webkit:webcore) is open, but the rest is not. Comparison of layout engines already does a good job of explaining what is or isn't free/open source.
Regarding Mozilla Firefox, trademarks are not generally seen as upsetting the free/open source nature of a product. Furthermore, the Mozilla Firefox source distributions are shipped with a Makefile which allows them to be compiled with generic branding, making the whole issue moot. --Karnesky 00:27, 14 March 2007 (UTC)

[edit] Acid2 test

Would the addition of a table showing when the acid2 test was pass could be interesting?

Acid2 is a theoretical concern. Nothing under consideration here would "pass" it. Indeed, it isn't meant to be passed, since Acid2 is broken by design - it's as much a feature test as it is a test of how a browser handles broken coding. That leaves us trying to decide what degree of failure is worth what merit, which is untenable. Wayne Hardman 22:09, 12 April 2006 (UTC)
But Opera passes the Acid2 test... -iopq 04:19, 13 March 2007 (UTC)

[edit] Opera userJS MathML support

http://my.opera.com/userjs/forums/topic.dml?id=124033 -Iopq 12:55, 25 April 2006 (UTC)

Yes, this is noted in the table. It still doesn't mean Opera supports MathML. In other places in the tables where you need a plugin or extension to enable a feature, a No is used to indicate that the browser does not natively support the feature. -- Schapel 13:21, 25 April 2006 (UTC)
True, but it usually has a footnote noting that it is possible with an extension. Someoneinmyheadbutit'snotme 17:57, 27 April 2006 (UTC)
True, but the entry for Opera's support for MathML already has such a footnote, so there is nothing to change. -- Schapel 01:11, 28 April 2006 (UTC)

[edit] SecurityFocus links in Vulnerabilities section

Please do not manually change the link to SecurityFocus data. For example, some editors try to type in the latest version of a browser directly into the URL, and when they follow the link they see the message "No matching vulnerabilities found". This does not mean that there are no known vulnerabilities in that version of the browser. If the text "Select Version" appears in the Version pull-down menu, that means that there is no information on that version of the browser. Select the latest stable release version of the browser from the pull-down menu, which is the most up-to-date information SecurityFocus has on record, and use that link and the information provided through the link. -- Schapel 13:22, 26 May 2006 (UTC)

[edit] Discussion page mass edit

I have edited the discussion page in order to restore the initial order. Please use three equal signs (===) to create sub-categories rather than categories. Not sure if "this article is being considered for deletion" deserves an own category; I'll leave it as is. Grey 18:37, 8 June 2006 (UTC)

[edit] "Criticisms of Opera" in See Also

Under See Also is: "- List of web browsers / - Comparison of layout engines / - Criticisms of Opera". Two links to lists/comparison pages that apply to all the browsers here, and one to a criticism of one particular browser? Rather biased, isn't it? If you're going to have links to criticisms of the browsers, have one for each of the browsers (or at least, all the main ones); if not, I say delete it. Simxp 23:55, 12 June 2006 (UTC)

[edit] "Microsoft Internet Explorer" vs "Internet Explorer"

Just a small detail, but I couldn't help but notice that Microsoft Internet Explorer is listed simply as "Internet Explorer," while other browsers (i.e. "Mozilla Firefox" and "Netscape Navigator") are listed under their full names. Should "Internet Explorer" be renamed to "Microsoft Internet Explorer" in this article, or should "Mozilla Firefox" be listed as "Firefox" and so on? Rob T Firefly 06:33, 2 July 2006 (UTC)

It is to be renamed Windows Internet Explorer in version 7. bruce89 20:01, 21 July 2006 (UTC)
As if simply having "Explorer" in the name didn't confuse it with Windows Explorer already... ☹ That said, I agree with the point being made. If we are going to list the "make" and "model" of some, let's do it for all. ⇔ ChristTrekker 18:28, 22 December 2006 (UTC)

[edit] Unclear referencing

Throughout the document alot of the tables are marked with a little cross, and further details are shown underneath the table. Is there no way different symbols can be used for each mark? At the moment it is often impossible to tell which note corresponds to each mark! 86.3.170.69 17:59, 16 September 2006 (UTC)

[edit] Firefox 2.0

I was just wondering if the table has been updated for 2.0 yet. Someoneinmyheadbutit'snotme 16:49, 28 October 2006 (UTC)

[edit] Java

In all other columns I can find "Available with...", but not for Java. I am not sure about it, if Java is included in the install package from all other browsers with java support, but it's not included in any IE package and must downloaded from SUN. (Except the an old version where this MS java virtual machine was included) But only with Internet Explorer java is not available. If all browsers need a SUN download, maybe a footnote in the header would be nice. --Lastwebpage 16:08, 8 April 2007 (UTC)

[edit] comparison of mobile browsers?

Any plans for a similar comparison page for mobile browsers? Right now there doesn't appear to be a central source of their requirements and compatabilities.

Well, you're kind of right; this series lacks a comparison of mobile browsers. If you are in apparent need of info try these two links. If you read all the comments and see all the screenshots (link #2) you can find out quite a bit of information. If you're up to developing your site for devices I would also recommend reading the full page found at link #1 (for a quick overview scroll down a bit to the browser listing). Be sure not to miss the phone emulators linked to further down in #1. Grey 00:30, 7 August 2006 (UTC)
  1. http://www.howtocreate.co.uk/tutorials/css/mediatypes
  2. http://www.howtocreate.co.uk/operaStuff/devices

[edit] FTP upload/download file/folder

The FTP column needs more than "yes/no". On Firefox 2.0.0.1 I cannot find any way to FTP upload even a file. (There is a FireFTP extension.) Same on Opera v9; download only. Yet even old IE5.5 is very nicely integrated with Windows Explorer, and can easily FTP upload not just files but also whole directory structures. 69.87.203.244 02:01, 12 January 2007 (UTC)

[edit] Where is E4X and JavaScript 1.7

Missing info, simply.

--itpastorn 21:50, 27 February 2007 (UTC)

JavaScript 1.7 doesn't have a formal specification and thus is not supported by anything except Gecko (it is a proprietary feature). E4X is an Ecma standard, but isn't implemented anywhere else, either (could be listed, though). --Grey 01:15, 2 March 2007 (UTC)

[edit] IE6 and IE7

Should there be separate rows for IE6 and IE7 (like there is for IE for the Mac)? Both are very widely used so the IE6 information is still relevant, and they have a very different features, rendering abilities, and platform compatibility. Koweja 18:47, 28 March 2007 (UTC)

Retrieved from "http://en.wikipedia.org../../../c/o/m/Talk%7EComparison_of_web_browsers_e5a1.html"