Bashing Chrome has become a kind of hobby for me, but I’ve begun to feel bad about it–there are quite a few really nice features. One of these is how it handles JavaScript. So I ran a benchmark to test it against Firefox 3 and Internet Explorer 7.

The Test

The test is simple: Open up this link (caution: It will crash your browser) from an email I sent to myself, open a new tab, switch between tabs a bit, and then close the problematic tab. Simple.

The Results

Firefox: Firefox started great–I was able to open a new tab and conduct a Google search after the crash page loaded. However, the browser crashed when I tried to close the crash tab. 3/5

IE7: By far the worst–as soon as I opened the link everything crashed, and closing it was even harder than closing the dead Firefox. 0/5

Chrome: I was hoping that I would get an infamous “Sad Tab.” I was wrong. The crash tab did nothing to affect performance–I opened a new tab, used it, switched back to GMail, closed the crash tab, and nothing happened. It was very exciting. 5/5

Conclusion

So this pretty much says what we knew already: Chrome is great at JavaScript, Firefox is great at everything else, and IE7 is terrible.

But I will be  very excited to either have Chrome adapt the usability of Firefox, or Firefox to adapt the solidity of Chrome–the utter nonchalance with which Chrome handled the test is still blowing my mind. I’m pumped.

I’ll tell you right off the bat–this is not an “article”, just a post. Now that we’re clear:

I take back a lot of what I said about HTML, CSS and the W3C. It’s all a good idea, and I now realize that. That said, I’m still a bit touchy about the Validator. Want proof? This page was validated in 2008, but running it through the Validator again brings up four errors and two warnings. I’m sure they have their reasons, but I still feel that my anger is justified.

All of that said, I’ve been redoing my website to:

• Use HTML for page logic.
• Use CSS for page looks (a challenge, considering that as of this morning I didn’t know CSS at ALL.)
• Make the little navbar thingy towards the top a lot easier to maintain.
• Pass the Validator (OK, OK, maybe Tidy HTML can help a bit.)

I also took advantage of the shuffle to rewrite the content a bit.

So far, I’ve finished one page, which was based on this one, but will now double as my index too.

A few notes:

• All the CSS on it was whatever I could hack in a few hours using Google, part of this, scientific experiments with simple pages and Tidy HTML and logic. I have no idea how ‘nice’ it is (in terms of coding style) but it appeases my eyes, the Validator, and the end product, so I’m not complaining. I do not recommend this method of learning for the faint of heart. (Yes, I am going to do a proper tutorial this summer, I just wanted results NOW.)
• I hacked together some Javascript for my navbar. This was accomplished by: Looking at sample code, and writing like a Pythoneer while using the Java syntax I’m learning in school.
• So far, the Validator is happy!

Some of you might be wondering what happened to my colors. Answer: I’m colorblind, and as such have absolutely no sense for what looks good on a page. I began to think that all the blue might be a bit overwhelming, so I’m going old-school with the black-on-white. If you have any opinion on this, please relate it in the comments. I would really appreciate it.

I’m a bit miffed that this means absolutely nothing to people looking at my site with polite browsers, but I guess this is what they would call a Black Triangle. It makes me feel really proud of myself though, which is one of the reasons I love coding.

Goal: Finish the rest of the pages by the end of this weekend, and move everything from my ‘experimental’* subdomain to the real deal.

*That thing is great for organization.

I’ve been making an honest effort to use Google Chrome.

Really, I have.

It seems like such a better idea–one process per tab, faster Javascript, easy interface, incognito mode, integrated search from address bar–it’s positively beautiful on paper.

But I just can’t accept it as a ‘serious’ browser.

To help you understand this, let me walk you through a typical weekend Firefox session:

Open Firefox, all the tabs I was looking at last time are up. Finish reading/digesting them, and close them as I finish. Check my mail, Twitter and Facebook, writing when necessary. Check the stats for this blog, as well as my ad performance. Read everything new on Google Reader, and if I want to read some more, do a bit of Stumbling. If a new term comes up in Reader or SU, look it up on Wikipedia. Write a post for this blog or my other one, if I have the time, and maybe check the Programming section of the Ubuntu Forums–either I learn something or I can help someone out. If I still want something to do, then I open up either my C or wxPython tutorial, and start learning.

Now let’s consider Chrome:

• Open up the tabs from last time: Check.
• GMail/Twitter/Facebook: Check.
• Blog info: Check.
• Reader: Check.
• StumbleUpon: Fail.
• Wikipedia: Fail (this will be explained.)
• Blogging: Check.
• Ubuntu Forums: Check.
• Programming tutorials: Err…I program in Ubuntu, and use Chrome in Windows, so Fail.

Besides StumbleUpon (and Wikipedia(?)), everything looks fine. So what’s the problem?

Ease of use.

But Timmy, Chrome is ridiculously easy to use–Google has long been famed for its superior design, interface and usability!

So I present my humble list of Five Usability Fails on Google Chrome:

1. The Omnibar. I think Google’s trying to one-up Firefox 3′s Awesome Bar with this. I think they failed. Firefox does a superior job of searching ALL parts of a URL and putting the MOST relevant one on top. I have mastered the <Down><Enter> hand flick. Chrome tries, but it seems to weigh the first part of the URL a LOT more. Take this blog–I access it frequently, and in Firefox always used “Ran…” to do so (it found Rannsaich mo Inntinn) Even though I acces it a lot in Chrome, Google puts Rand McNally (rand.com) as the first result. I haven’t even been there! This is just Google outsmarting itself.
2. The search box. Chrome doesn’t have one. Fail.
3. Number two cheated. If I want to search Google in Firefox, I hit <Ctrl><K>. If my cursor is in the location bar, I hit <Tab>. If I want to search Wikipedia, Ninjawords, or BibleGateway, I use <Ctrl><Arrow Keys>. This is exceedingly easy and fast. What’s more, if I’m Feeling Lucky (say I want the Wikipedia page on Agile Programming), I just type “wiki agile programming” into the location bar, hit enter, and BAM! there it is. This has about a 97% success rate. Good enough for me. Chrome, on the other hand, only lets you use ONE search engine, and you can NEVER feel lucky. What’s more, to access ‘real’ webpages (see #1) you have to scroll through the search entries in the drop-down list.
4. Chrome. As in, the decoration around the browser, not the browser. I feel like Firefox provides a lot of power with its UI–navigation, bookmarks, tabs, extra third-party toolbars, menus, etc. If I feel cramped for screen space, guess what? I hit <F11> and it all goes bye-bye. Chrome is stuck in this middle ground of having less power than FF’s complete chrome, and less screen space than FF’s fullscreen. Why can’t Chrome have a fullscreen too? I have no idea. Feature request!!!
5. Tab layout. Call this minor, but it really bugs me. I have grown fully accustomed to new tabs being opened at the far right. Maybe it’s just the way my brain works, but I’d much rather see things chronologically instead of topically. I’m probably crazy, but the fact that Chrome opens links in adjacent tabs drives me cuckoo. They could at least make it optional….

I know that I could learn to live with a lot of these, but why? Firefox works fine for me, and I see no reason to change.

I really need to vent this:

Back in the day, when I learned HTML, I was taught that “Hello World!” was nothing more than:

<HTML>

<HEAD>

<TITLE>Hello World!</TITLE>

</HEAD>

<BODY>

Hello World!

</BODY>

</HTML>

So when I made my extremely simple Python page, I just did that. I obviously had to throw in some <a href=”http://whatever”></a>’s and some <em>’s (I actually thought I was being a good boy, using <em> instead of <i>) but it’s really a simple webpage.

So just for fun, I decided to run it through the W3C validator.

And it coughed up some errors. I have now looked through most of them, and all of them (with the exception of the DOCTYPE one) seem extremely arbitrary. And I would NOT like to use a stylesheet, I’ll get by just fine with my <FONT size=”4″>, thank you very much.

Now I’m sure that the W3C has legitimate reasons for these standards, but I just get the impression that they create new standards out of boredom, not necessity. Less is more and if it ain’t broke–don’t fix it for crying out loud!

I can hear your brilliant counterargument unfolding. “Timmy,” you say, “why don’t you just shut up and use Tidy HTML Generator? W3C had the good grace to include it, lest people greater than you are confused by their cryptic error generation.” to which I reply, “Take a  look at what Tidy suggested”

WHAT IS UP WITH THAT!?!

OK.

</rant>

I can breathe now.

Seriously though–why?

I think GMail has made an error.

No, it’s not a big error, and yes, it’s still my favorite web-based email service.

But they still messed up with GMail Labs.

It’s a great idea, so far it’s giving off great fruit, and I would be a sadder person today if it did not exist. That said, I was kind of interested in how to develop for it. So I went to the Labs portion of my Settings page, scrolled to the bottom, and was told:

Looking for something we don’t have? Suggest a Labs feature or join the Gmail team and write your own!

Lame! There are probably hundreds of people who would love to get a development kit from GMail and write their own Labs features, host them externally, and suddenly POOF! GMail would have hundreds of plugins, and it would suddenly be just as awesome as Firefox or gedit or vim! I’m not saying GMail needs to release its source. All I’m saying is that they should release enough of it for developers.

FAQ

Q: How would this help Google?

A: It would make GMail much more powerful, giving it an edge over the competition.

Q: How would this help me?

A: It would provide you, the GMail user, with a much more powerful interface.

Q: Why do you keep using the phrase “much more powerful”?

A: Because it sounds better than “powerfuler”.

Q: That was a terrible answer.

A: And that’s not a question. Are we even?

Q: But, couldn’t you just submit a feature request?

A: Yes, but it’d have to be an exceptional idea for the official GMail staff to make it. Opening it up would provide a lot more people willing to code, which would make the good ideas happen too.

Q: What if this crashes GMail?

A: They already have the option to open GMail without Labs.

Q: What about privacy?

A: I’m sure there’s a way to give developers enough control to develop, but not enough control to feed them your personal information.

Q: What about Better GMail/Greasmonkey? (Thanks, Peter)

A: First of all, native is better than third-party. Second of all, a LOT of end users don’t know about/care about Better GMail/Greasmonkey. Third of all, Better GMail/Greasemonkey only works on computers you install it on, while what I propose would be universal for your account.

Since I’m a webmaster, I obviously need to upload things. When I was on freeservers.com (very bad), and Geocities (pretty bad), there was a little web-based ‘File Manager’, which did all the uploading and manipulation of my files. It was bad. Slow, non-intuitive, clumsy, you name it.

So one of the many reasons I’m now on andrewmin.com is that I can use FTP (File Transfer Protocol) to upload things. Dreamhost has ‘WebFTP’, which is marginally better than the ‘File Manager’s. So I just use (much, much) better third-party apps to do it instead. I’ve used three. Filezilla, and Whirlpool (just kidding, Filezilla, gFTP and Nautilus–I felt like messing with the people who don’t read the title) (Note to self: If FLOSS app is ever made, name it Whirlpool )

(Edit: I actually made Whirlpool….)

In keeping with the style of this blog, I’m just going to spill my thoughts out (in a concise way, I hope (!) )

• Nautilus is beautiful.
• Filezilla is functional. Sure, it has a GUI, but it has a GUI to be functional, not to be pretty.
• gFTP is very similar to Filezilla, except you might convince me it’s prettier. It’s probably personal preference.
• Assuming Nautilus is your File Manager, it’s very nice–the FTP server is just like another drive. It even shows up when you hit ‘Browse’ in something.
• This means that you can open up a webpage in your IDE, and when you hit save it’ll be uploaded.
• I haven’t clocked it, but I think Filezilla and gFTP are faster than Nautilus.
• Filezilla works in Windows.
• Nautilus and gFTP don’t.
• You’ve probably noticed by now that Filezilla and gFTP are very similar. The three main differences are:

1. Operating System compatibility.
2. The way you use it. Filezilla is drag and drop–very easy and intuitive. gFTP is a bit more complicated–you select a file(s), and then hit the “->” button to copy it to the server. Which is actually safer….
3. gFTP has a really easy interface for getting files off “FTP, FTPS (control connection only), HTTP, HTTPS, SSH and FSP” websites (quote from the official site). It even has preloaded bookmarks of common ones–such as Debian and SuSE.
   

• If you have this much time to spend reading about FTP clients, than you’re probably fine with either one

Summary

If you have Windows, use Filezilla. If you have Linux but not GNOME (and use Thunar or Konqueror or something) don’t mess with Nautilus, just investigate FTP with your own File Manager. If you do have GNOME, then use Nautilus, unless you’re the type to upload a LOT of files at one time, infrequently, or you just want a slight edge in speed.

As for Filezilla vs. gFTP…make your own decision. If FTP managers were more popular in general, that would probably be a flame war comparable to GNOME/KDE or Vi(m)/Emacs.

In short, this pretty much boils down to appearance vs. function, which I can’t hope to compare

The solution? Do what I do–use both. Nautilus for editing, Filezilla/gFTP for adding.