marahmarie: my initials (MM) (Default)

Firefox actually has the weirdest performance in some areas:

  • clearing comment forms on entry pages (the whitespace between entry and form works as it should in _every_other_browser_; in Firefox the form doesn't clear by more than a few pixels unless I about triple the top margin).
  • arranging cell space on comment forms under 360px width on mobile is not going very well (our comment forms are actually tables nested within tables, which doesn't make it any easier)
  • displaying padding in text fields (for example, the search box in the navigation bar above the header is the wrong shape/size in _every_other_browser_ because Firefox)
  • certain CSS properties/values work in Firefox but in no other browser, which led to this thing today where my comment forms looked like someone shook all my pages really hard and let the form fields land where they would. Such fun.

I think I'd need about a thousand hacks for every possible browser/OS/device configuration to work around some of this, or to simply hack Firefox to allow better display in other browsers. And I could hack Firefox, because there are hacks for it.

The other thing is how needless it is to test page display on multiple versions of modern Webkit (only speaking of Windows browsers). If you check a web page in Chromium or Google Chrome, it seems you've checked it in every modern Webkit browser including Safari for Windows - which stopped at v.5 some years ago - except modern (non-Presto) Opera, which has its own ideas about CSS.

I'm not sure if this applies to Chrome on Android, as well, but as far as desktop testing goes, yeah, it does seem that way.

marahmarie: my initials (MM) (Default)

I can't believe more people don't know you can run Chromium on Windows without having to peruse the dev channels or do funny packaging things to builds or whatever. It was a surprise to me too, and one that I stumbled upon quite by accident.

About ten minutes ago, while I was looking through several weeks worth of change logs on Chromium.org to see why Google Chrome runs just fine now after you delete every Google Update file and folder on your computer (normally Google Chrome won't even start if you delete those) I decided to download the mini_installer.exe out of abject curiosity, and lo and behold, I got a stripped-down version of Google Chrome that runs on Windows - without the Google Updater and spyware-like Google usage statistics collection.

It's just like SRWare Iron, as far as I can tell: no Google Update executables or folders, but unlike SRWare Iron, there are no preloaded German start pages in the Speed Dial window.

Of course, Chromium is pure Google source code no matter how you cut it. So is SRWare Iron; Google Chrome wouldn't be Google Chrome without Chromium source code, unfortunately.

Looking on the Web, I see only one reference to using the mini_installer.exe to install Chromium on Windows, and it's on Lifehacker, which in turn references the Google System blog, and it's tacked to the end of a very long post on how to customize Google Chrome. I don't know why everyone seems to want to keep a non-privacy-invading, non-auto-updating version of Chrome such a secret.

ETA: OK, never mind...Chromium seems to be missing the Google Update features, but not the snooping features. Using SRWare's comparison page as a guide on what to look for, I typed a non-existent URL into Chromium and got a Google Chrome error page, indicating Chromium might be sending DNS errors to Google just as Chrome does. And of course Chromium defaults to the Google home page, just as Chrome does. The other things to look for you can't look for, really, because Google performs those tasks in the background. On the plus side, Chromium doesn't add itself to the Windows Task Scheduler like Google Chrome does, but if you're after privacy you're still much better off using SRWare Iron, after all.

ETA 2: And of course, since Chromium doesn't include Google Update software, reading change logs on Chromium.org doesn't tell me anything, anyway. So now I'm trying to find the version of the code (I mean, the _exact_ version) that Google uses for Chrome, and I can't. I don't know if I'm missing something really obvious, but the latest build of this otherwise craptastic browser is now removing all the Google Update files and folders at uninstall that I talked about it leaving behind in my last post. And I can't find older builds now to compare the latest build against.

ETA 3: File Hippo, of all places, archives all of the old Chrome versions. (http://www.filehippo.com/download_google_chrome/) Back to studying this crappy browser...

marahmarie: my initials (MM) (Default)

While I'm putting together my revised IE 8 Beta 2 post (a process which has given me enormous respect for anyone who takes the time and trouble to do even quasi-scientific browser testing) let's ponder something...

While Chrome's icon has been likened to Simon, Pokemon and web cams, I think it looks like the icon for the Windows Media Encoder. I downloaded said encoder last week to finish installing screen capture software and whaddaya know?

marahmarie: my initials (MM) (Default)

I'm trying out three browsers right now: Google Chrome, IE Beta 2, and AOL 10.1 Desktop. Browsers are an obsession of mine and I like trying out new ones, but getting around all three of them at once is a bit trying especially since I don't actually use any browser besides Firefox. And I want these reviews to be more thorough than ones I wrote in the past - I want benchmarking software to time page loads and to count working and private memory sets so I can come up with more accurate figures. So that anyone who's saying, "Oh, she's just another blogger blowing smoke at us and doesn't know what she's talking about" will stop saying that. Too many browsers, not enough time...