Opera: Firefox User Figures 'Inflated' 810
Anonymous Coward writes "ZDNet
notes, 'The chief executive of Opera Software claimed on Monday that the market share figures for Mozilla Firefox are inflated, due to its support for link prefetching" In addition, "Opera has a better caching mechanism so it doesn't access Web sites as often as other browsers" and "Opera is configured by default to identify itself as Internet Explorer' "
Re:Now why? (Score:3, Informative)
Re:Link prefetching? (Score:5, Informative)
Firefox only prefetches links when the links are marked on the previous site as "hey you might want to prefetch this".
Specifically on Google, only the top result is marked as prefetch. And even then, only when Google has determined that most visitors would choose it. Google has said that they mark it for things such as prefetching cnn.com when somebody searches for "CNN".
The article states: "...which means that Firefox will pre-load the top search results into its cache." which is innacurate. It should say "...which means that Firefox will pre-load the top search result into its cache when Google thinks there is a very high probability you will visit the first result."
Currency exchange rate calculator and foreign exchange converter [coinmill.com]
Re:Opera versus Firefox (Score:3, Informative)
Relying on the headers the browser sends to figure out stats on browser usage is a dicey proposition at best. A statistically significant survey of Internet users would yield better results, but who's going to pay for that?
EricDescriptions of my books [ericgiguere.com]
Re:Link prefetching? (Score:2, Informative)
So, whining about a feature that the website has to turn on is kinda pointless. You'd think the site would figure out how to count visitors BEFORE turning on prefetch right?
Read the FAQ [mozilla.org] and see what's really going on.
Re:Irresponsible as hell (Score:2, Informative)
Re:Caching (Score:3, Informative)
Easy answer. (Score:1, Informative)
The people at Opera were trying to sell their browser not enforce standards or change web designers. That meant that it was Opera that had to bend to be compatible with the sites not the other way around. Now that Opera has a market share of relative significance they may choose to try to change the world but, I'll bet that they just stick to selling browsers.
If you want your browser to work with 100% of the world's websites, the user agent better say Internet Explorer.
Re:Irresponsible as hell (Score:5, Informative)
Re:Irresponsible as hell (Score:2, Informative)
Re:he may be right, but (Score:5, Informative)
I reverted to using Firefox for a while, but now I'm happily back to using Opera. Been a happy, registered user since 5.0.
Re:damn the mouth-breathing majority!!! (Score:5, Informative)
Isn't that fraud?
No. Fraud is about using lies for direct financial gain, and requires specific intent. Opera identifies itself as IE for interoperability purposes, something that "modern" tech laws (such as the DMCA) protect.
Plus, the whole point of the www is that it is browser independent. So this is unstandard behavior, and should be shunned(2).
I'm sure Grandma will think it's great that her bank and realtor websites don't work because Opera is taking a stand.
The real blame for this lies first in Netscape (which extended the web in many incompatible ways, but at least worked on every OS) and later in Microsoft (who used Netscape's tactics to sew up the web). If Tim Berners-Lee was dead, I'm sure he'd be rolling in his grave. Instead he's had to settle for being alive and helping correct this nonsense. [w3.org]
Re:Hey (Score:2, Informative)
http://www.opera.com/swim/ [opera.com]
Re:Irresponsible as hell (Score:3, Informative)
Re:Double-click (Score:3, Informative)
Really? IE doesn't. In fact, there's an MSDN page that describes the ondblclick event that fires when someone double-clicks a DHTML element:
http://msdn.microsoft.com/workshop/author/dhtml/r
The only reason I know this is because we actively use this event at work (Don't shoot: it wasn't my idea).
--
Mando
Re:Yeah well... (Score:3, Informative)
Basically, Opera always used to put Opera in it's UA string. They would constantly get blocked from sites for no reason (if you used proxomitron and blatently lied to the site to get past the checker, it worked perfectly).
Opera's customers asked for a way to hide that they were using Opera (check the forums, this went on for years). In Opera 8, you now are able to hide what browser you are using completely. Heck, Opera takes care of some of that automatically for you with auto updates.
Re:This is Interesting (Score:5, Informative)
I agree that Opera is a decent browser and they've been decent for a long time. I just don't want to pay for a browser or be forced to view advertisements. And thanks to Firefox, I don't have to.
My only complaint is that Firefox seems to run painfully slow on OSX.
Re:This is Interesting (Score:3, Informative)
Link Prefetching Hype (Score:1, Informative)
See the Mozilla site for more information:
http://www.mozilla.org/projects/netlib/Link_Prefe
Only Google Web Accelerator tries to use prefetching to fetch several(maybe all) links on a page, and it is available for IE and Firefox. It also uses a Google proxy server to cache pages, so perhaps not all hits will show up in a website's weblogs.
Its bullshit to claim that prefetching is causing inflated numbers, because websites that use prefetching also have the means to identify prefetch requests and can devise a way to take them into account when analyzing logs.
Re:he may be right, but (Score:5, Informative)
Under and over rated, to my understanding, are the generic +1 and -1 options. It covers all the reasons you might want to adjust a comment but don't have existing options for, like 'Completely wrong' or 'Proper use of iambic pentameter in a ruby script'. It never gets listed as the reason though -- if I modded you under (or over) rated right now instead of commenting, your score would go up or down, but no word would appear next to the score.
Re:Irresponsible as hell (Score:5, Informative)
despite deflated numbers . . . (Score:2, Informative)
Re:he may be right, but (Score:3, Informative)
Because, as was discussed here recently, there have been sites (mostly owned by Microsoft) that actively sabotage pages sent to Opera users, by using CSS that cause the content to be garbled. It was shown that only requests with "Opera" in the ID string had this problem. (Changing it to "Oprah" made the pages display properly.
I haven't (yet) read about this being done to firefox. Anyone know if it's happening anywhere? I wouldn't be surprised; I just haven't heard about it.
There certainly are pages that are only delivered to clients that identify themselves as "IE". I've seen such pages, and verified that an IE ID is needed to get them. I have a little perl web-page tester that has an ID-string command-line arg for exactly this reason.
Now, if this were due to bugs, or were otherwise inadvertent, you'd expect it to effect all browsers equally. But it doesn't. So far, the only cases I've seen are pages that require an IE browser ID. I've never seen a page that requires some other browser's ID. I'm not saying they don't exist; I've just never stumbled across them. This tells you exactly why you might want your browser to masquerade as IE from time to time.
I've read that IE identifies itself as "Mozilla/4.0 " because there were some pages that required the Mozilla ID to function properly. Maybe this is true. I've never seen such pages, but I suppose they could exist.
Re:he may be right, but (Score:2, Informative)
Yes, you certainly are. If I tried to take away the ability for workers in my company to do useful work from anywhere with just about any PC, I'd be fired and reviled by hundreds of otherwise well-meaning folks.
X windows was not the answer to ubiquitous application mobility, client/server wasn't the answer, nor was Java or ActiveX. Web-based apps seem to be a big step in the right direction.
Re:This is Interesting (Score:2, Informative)
I landed a patch yesterday that significantly speeds things up (we now use use CFRunLoop instead of Carbon Events, in case you can understand that). Huge difference, especially with plugins. It'll be in Firefox 1.1. This one patch makes a huge difference.
-Josh Aas, Mozilla Foundation Mac developer