Using Mozilla in Testing and Debugging

Henrik Gemal writes "In this article I will describe some very cool features in Mozilla which will enable you to quickly find and debug errors in your web site and web applications."

Comments

fr1st ps0t

[usotsuki]

Can it find any bugs here?

-uso.

Re:fr1st ps0t

[usotsuki]

In my hurry, I forgot a quote mark!!!

Check Link

-uso.

I'm a big fan of the html highlighting

[friedegg]

It lets me glace at things pretty quickly to get an idea of what may be wrong, and saves me the step of loading it into a full blown editor. Plus, I can select only part of a document and just view that particular source.

I also like the http header viewer add-on mentioned in the article. I used to have to visit a website and use that to view headers.

Re:I'm a big fan of the html highlighting

[Can it run Linux]

Cool! Is it available for Linux?

Re:I'm a big fan of the html highlighting

[Anonymous Coward]

Woooo boy, with flamebait like that, I'm glad I wear fireproof underwear.

Re:I'm a big fan of the html highlighting

[bugsmalli]

The correct way is of course: var response = true; response = false; the last time I wrote a piece of code like this, I was horny and my girl had a "headache"...

Venkman is good

[digitalgimpus]

This is also a good page to checkout for more info:

http://mozilla.org/projects/venkman/

Venkman is the JS debugger in Mozilla... and it's sweet.

There is also a Netscape made intro that may be helpful to new users:
http://devedge.netscape.com/viewsource/200 2/venkma n/01/

Re:Venkman is good

[Angry White Guy]

So when are Stanz, Spengler and Zedemore coming out?

I must be using Debian for too long...

[Trepidity]

When I read that headline I thought "what's so hard about using Mozilla in testing? Just apt-get install mozilla, same as in woody and sid..."

Re:I must be using Debian for too long...

[Anonymous Coward]

Ha ha! You're so funny. Some people actually get down to the nuts and bolts of things not just apt-get them all away. p0ser.

Re:I must be using Debian for too long...

[Captain Large Face]

Pah! Real men type apt-get install mozilla-snapshot, not mozilla

ok, so....

[Anonymous Coward]

in this new roadmap the mozilla tema put out, where does this stuff fit in? Features like this are decidedly pimpin and mozilla has tons of cool features that I'm always uncovering. But the new roadmap said to focus on phoenix and minotaur. Phoenix is lean and mean and minotaur is email, where does this stuff go? I don't want to lose stuff like this.

paradime change

[Anonymous Coward]

websites & webpages have reached a complexity whereby mistakes are refered to as 'bugs' like as if it were software. It seems the slow drift towards the internet being the computer is slowly happening. Nowonder Microsoft was so afraid of Netscape, though they thought it'd happen much quicker, though it probably would've with the speed Netscape came up with new things (where as MS not having done any real improvements to their browser for a long time).

Re:paradime change

[Blaine Hilton]

I have noticed this too. Another big factor I believe is changes in the shareware industry where a lot of developers are deciding instead of mailing CDs out or providing downloads, they could just offer the same thing on the web. Being web based give the author more control and more flexibility over the software then if it was a "standard" application.

Re:paradime change

[Angry White Guy]

The big problem with web-based tools is that there is greater incentive for the software to become a subscription service, rather than a product. The upside to this is for every web service out there, there is an open source version trying to do the same thing, or a compiled shareware version because somebody is too cheap to sign up for the subscription.

Re:paradime change

[bheerssen]

Complex websites are as legitimate pieces of software as your word processor. They have routines and they maintain state (albeit painfully). They also have bugs. They can have development cycles and release schedules. Just like software. You have to pay attention to threading and memory usage just like real software.

In short, websites are often not collections of html documents comparable to a PDF file, but true pieces of software that require thought and analysis throughout the development cycle. To get an idea of this, download a copy of phpMyAdmin or webmin and have a look at the source. Slashcode is also a good example.

Kaiju

[darkmayo]

I read the headline and thought.. Godzilla vs Mothra...

Testing with mozilla

[mwhahaha]

I normally use mozilla when I'm doing web development, however I still have to run throught he site in IE. Mozilla has great development features, but I have found that IE has bastardized HTML. Mozilla also has it's issues with tables (I'm currently having issues with non wraping text rows) and Horizontal rules (for some reason it just won't display on certain pages). You should see the code to get around the nonwraping text, my god it's horrid. Another thing is that Mozilla's javascript is slightly different than Microsoft's. I have found that IE 5.0's implementation is different than 5.5 and 6.0. Mozilla will also let you get away with certain variable addressings that IE will choke on. Mozilla is great, but you still have to use IE at some point. IE still forces us to do stupid things :[

Re:Testing with mozilla

[FuzzzyLogik]

I just have on my front page, best viewed with Mozilla/Netscape 7 ... with a nice big link to both. Seriously, IE might load faster, but everyone i've converted over from IE, they don't ever close it once it's open... even my parents, but they use Phoenix now not mozilla (though both are installed)... so, really, for me.. i limit who views my site correctly, but seriously, once aol flips over to use the gecko rendering engine in their client... IE might as well go on to the little browser heaven... or hell.. ;)

Re:Testing with mozilla

[mwhahaha]

True, but I'm developing for a company who does e-commerce store hosting. And the site has to be compatible for just about everyone. We always get complaints about IE 5.0 and Mac's IE because of their FUBAR javascript implementations. I am the only one in the entire company who uses mozilla. So they are always coming up with javascript stuff I never see because my browser actually works :]

Re:Testing with mozilla

[tijnbraun]

I still use HTMLtidy to check my pages... I would love to have something like a "stringent" mode while developing web pages (ala browser producer error instead of trying to render the html). A while ago I found out that mozilla can even be more forgiving than IE. There was some weird bug in a parser I was testing, which sometimes resulted in </tr> to be rewritten as </tr or something. Mozilla didn't care. IE was totally confused. (First time I ever found something in html that confused explorer, but rendered ok in moz). Anyways... is such a mode/plugin available?

Re:Testing with mozilla

[FuzzzyLogik]

oh ya, i understand that type of situation. but since my page is just a stupid homepage type deal where i fart around with stuff.. i just say screw the IE people, they're missing out already, so, they won't mind having a fudged up webpage ;) but if you're in an environment where you need to be compatible with IE then ya, you're right, no doubt about it. In fact, i compliment you on making sure it works in both browsers, i wish more websites were like that, even though mine doesn't work well with IE, i just prefer to make it easy on myself since i'm not doing it for a job ;)

Re:Testing with mozilla

[Anonymous Coward]

Is your text not wrapping because it has no spaces? Text in table cells will not wrap unless there are spaces in it. This is intentional.

-g

Re:Testing with mozilla

[Christianfreak]

That's the problem right there: Javascript. The company I work for (small business so this is easier) before I came Javascript was "IT" and slowly I've shown how you can do cool stuff with stylesheets, and how clean pages without alot of stupid javascript effects go a long way to creating happy customers.

Convince them they don't need the JS, much less rely on it for pages to display or navigate properly and life will be much better.

Re:Testing with mozilla

[mwhahaha]

nope, it has spaces and the table was defined with set a width. The table that held the inner table was using proportional width and I know the standard says if you are using percentages, it should not wrap. However if you are using fixed sizes it should wrap. It wouldn't do this in mozilla. So I had to put everything in the in another table. Don't ask me why mozilla wouldn't render it properly. It confused the hell out of me for a good day or so. Now i've got it working, but the html is very ugly.

Re:Testing with mozilla

[mwhahaha]

Ha! I wish it was only display and navigation. It's all form validation :[ So much form validation it makes me cringe everytime I have to add a new form to the system. Evidently we've got a bunch of users who love to do everything wrong and can't follow directions. Bastards

Re:Testing with mozilla

[Anonymous Coward]

If you're using XHTML, and serving it as application/xml+xhtml, Mozilla will let you know when your document isn't well-formed.

Re:Testing with mozilla

[JimDabell]

I still use HTMLtidy [w3.org] to check my pages...

A validator would be a better choice. It's a proper syntax checker, not just a linter.

Re:Testing with mozilla

[Christianfreak]

Oh it was that way here too. And trust me people will turn off Javascript to get around that if they figure out how. Do your form validation on the server side where your customer can't mess with it.

Re:Testing with mozilla

[the-matt-mobile]

I would love to have something like a "stringent" mode while developing web pages (ala browser producer error instead of trying to render the html)

You can use XHTML and then use any XML parser to check your page. If it doesn't parse, it isn't valid.

If you're used to writing nasty HTML like <b><i>example</b></i> then you'll probably complain the XHTML feels a little too stringent, but once you get used to it, designing your page just feels so much nicer.

Re:Testing with mozilla

[4of12]

...IE has bastardized HTML.

That would be "Mr. Bastardized HTML" to you, thank you very much.

IE now sets the standard for bastardization and no weasel browser doing new bastardization of HTML should get the uppity idea that it can do the same.

Re:Testing with mozilla

[ManxStef]

I know just the Mozilla plugin for you:

Checky

A single keypress (F10) will then open the validators of your choice (e.g. the W3C HTML & CSS validators, Bobby, HTMLTidy, URL checker, etc.) in seperate tabs, or windows if you prefer.

Fantastic!

Re:Testing with mozilla

[Alien Being]

Speaking of w3.org, Amaya has some nice features for working with, and validating complicated document structures.

Re:Testing with mozilla

[Spoke]

A validator [w3.org] would be a better choice. It's a proper syntax checker, not just a linter.

What I would love to have is have something like the http header viewer built in so I don't have to pipe in pages to the validator manually.

Re:Testing with mozilla

[pmsyyz]

I would love to have something like a "stringent" mode while developing web pages (ala browser producer error instead of trying to render the html).

I have my local Apache configured to send .html files as application/xhtml+xml so that I know my XHTML files are well formed since Mozilla/Phoenix throw errors if they are not.

Re:Testing with mozilla

[shadowbearer]

Don't you mean MS. Bastardized HTML ?? ;-)

SB

Re:Testing with mozilla

[Anonymous Coward]

You should do both. Client side provides for a better user experience and keeps network I/O down. Server side is for people w/o JavaScript.

Re:Testing with mozilla

[ubernostrum]

As a couple people have already pointed out, if you're serving XHTML with the proper MIME-type of "application/xhtml+xml", Mozilla will show you XML parsing and well-formedness errors instead of trying to render the page; there's a good article on the XHTML MIME-type along with instructions on how to serve that MIME-type only to browsers which understand it over at XML.com.

Re:Testing with mozilla

[Anonymous Coward]

Yes, and I absoluetely love using Mozilla in my development, unfortunately M$IE doesn't even remotely follow W3C standards when it comes to the DOM and event model, so if you DO have to use JavaScript, you end up writing browser-checkers. Using Mozilla you still wouldn't have a clue how it handles in IE. I've even noticed huge differences in how HTML and CSS are rendered.

I wonder if Mozilla's debugging tools (particularly for JavaScript) could be made to mimick IE and all its mistakes?

Re:Testing with mozilla

[WNight]

It's also for security. Never trust users.

Impressive features list

[dtolton]

I have been using Mozilla because I love the tabbed browser windows. I was completely unaware of all the extra features it offers.

The DOM Inspector will be really nice for checking out the rendered structure of a page. I've always had a tough time with this since I generate most of my pages dynamically. In fact most of those tools will be *incredibly* useful in that context.

I have to say I'm really impressed with the progress the Mozilla team is making. For a while there IE was leading the way, now that trend has clearly reversed.

Re:Impressive features list

[FuzzzyLogik]

Me as well man, I never even knew Venkman existed, and i've used Mozilla/Phoenix for the past year and a half i think, maybe longer, since 0.92 i think was when i started using it full time. These little debugging tools are REALLY going to be useful.

Re:Impressive features list

[coolgeek]

I didn't know about the DOM Inspector in Mozilla. I've been using a very handy script from BrainJar called DOM Viewer. The key advantage is it works cross-browser. Also enables you to see what object properties are available on one browser vs. another.

Re:Impressive features list

[eyeye]

I too was amazed at the things in mozilla that I had just never noticed!
Venkman is quite polished too imho, shame I hate JS!

An internet story

[Anonymous Coward]

1) write a snazzy article about website bugs
2) upload article to own website
3) submit link to own article on Slashdot
4) discover the bug the hard way when own website is slashdotted to smitherines within 20 seconds
5) ????
6) Famous!

Re:An internet story

[greenskyx]

7. ??? 8. Profit :pPp

standards compliant so far

[trmj]

I have really been impressed with mozilla since they went fully standards compliant (back in '96 I remember it being all the rage to complain about netscape and how their "netscape-isms" like the , etc tags were ruining HTML).

Let's hope that with these new developer features they continue with this compliancy, and don't go and do what MS did to scripting/programming languages when they released .NET

Re:standards compliant so far

[josephgrossberg]

Care to elaborate what you mean by "what MS did to scripting/programming languages when they released .NET"?

Re:standards compliant so far

[trmj]

MS made their own standards when they released the .Net system. I am hoping that Mozilla doesn't decide that what they like and don't like about a certain scripting language or even HTML itself becomes their "standard" and is allowed.

And while my comment above may have cast me in a completely anti-MS light, I do commend them for getting their act together in the CSS area long before Netscape was able to.

Re:standards compliant so far

[BACbKA]

What I still hate about Mozilla is the (netscape 3 legacy, I believe, or whenever the 1st composer was released) is the "tag abuse syndrome" promulgated by it. The worst thing IMHO is the "indent"/"unindent" buttons which (rather than actually use the CSS indent properties available today) actually use nested tags implying some logic the author doesn't mean. Today's standards are all about the separation of contents and layout, and the NS/Moz composer do it the wrong way IMHO.

Big Advantage

[Dave W]

This is a really significant advantage to Mozilla.

With Mozilla being so attractive to web developers it makes it so much more likely that sites will fully support mozilla and those developers will bring in more users.

Getting developers to use your application is a great way to build market share.

Our own developers tend to use Mozilla as the key browser already, with tests to check behaviour on IE later.

All we need now is full etester type functionality using Mozilla instead of IE (preferably Linux based). We have used many add-ons for JUnit (like Canoo Webtest), but the javascript and DOM support is always the problem. Embedding mozilla might be a better way to go.

Dave

Re:Big Advantage

[barzok]

While I want to agree, I'm afraid I just can't. I'm a developer, I use Mozilla 99.99% of the time at home. Although I do everything I can to stick with the standards and make sure things look/work right across browsers, my clients care more about IE than anything else (because that's all they've ever known, same with their clients) so if IE mangles what I've written, I don't have much choice but to fix it (and probably get what I didn't want in Mozilla, standards be damned). If I'm maintaining an old, horribly broken site, my goal is to just not hose what's already there, and that normally means not coding anywhere near the standards.

Like I said, I love Mozilla, I use it almost exclusively at home, and I would really, really, really like to believe that as long as I code 100% standard code and things check out OK in Mozilla I'm fine, at the end of the day if it doesn't look "perfect" in IE, it doesn't matter how correct it is or how nice it looks in Mozilla.

Mozilla is great for debugging Javascript...

[X_Caffeine]

Many javascript programmers consider Mozilla to have a buggier javascript implementation than MSIE; what they don't get is that MSIE is very good at interpreting buggy code, which leads to... well, bugs and incompatibility. Mozilla forces you to do it the right way, which leads to greater compatibility in the grand scheme of things.

I wish this article had addressed the whole MSIE "document.body" mess, though. The correct DOM equivalent is "document.documentelement", but it doesn't work in MSIE6 unless the document is properly defined with a DOCTYPE declaration (otherwise MSIE is in backward-compatibility/buggy mode).

Otherwise, a really great introduction. I've been using Mozilla to do javascript for months and didn't know most of the data here.

Re:Mozilla is great for debugging Javascript...

[Anonymous Coward]

The proper word is scripter. Damn, you probably call the garbageman a sanitation engineer.

Re:Mozilla is great for debugging Javascript...

[Angry White Guy]

No, they call themselves sanitation engineers, in much the same way MCSE's are engineers.

Re:Mozilla is great for debugging Javascript...

[barcarolle]

Those "javascript programmers" that thinks Mozilla's JavaScript implementation is buggier than IE's fail to understand is that Microsoft doesn't use JavaScript, they use JScript.

Re:Mozilla is great for debugging Javascript...

[Anonymous Coward]

> greater compatibility in the grand scheme of
> things

Sounds great, sucks when you're forced to work the weekend to support the stuff.

Re:Mozilla is great for debugging Javascript...

[X_Caffeine]

You write code with broken table tags, and you're calling me an "asswit"? If all browsers crashed from errors like that the web would be a lot better-coded -- because jerks like you who can't properly write the most basic of languages (HTML) wouldn't be involved.

Re:Mozilla is great for debugging Javascript...

[Anonymous Coward]

Dude, lay off the crack pipe. I don't want my browser to freakin' disappear, which is what Mozilla does frequently. I want an error dialog, or I want to not crash. One or the other.

If browsers crashed from errors, we wouldn't HAVE the web because the vast majority of people don't put UP with that kind of behavior...which, of course, is why IE has all the browser share.

Re:Mozilla is great for debugging Javascript...

[GoodFun!!!!!!!!]

Of course. How could I be so ignorant. No coder worth his salt has ever made a typo. Only a fool would want a browser to actually give a decent error message rather than crashing completely. Never mind the fact that this is a thread about DEBUGGING.

Re:Mozilla is great for debugging Javascript...

[WNight]

Yeah, I'd hate to be called in on the weekend to fix something that tested properly in our browser, yet crashed when it went out to customers because they had a slightly different browser revision.

Much better to have a stricter debugging phase, during regular work hours, and ship a solid product.

Validity checker and indicator

[Alderete]

I'd really like to see a simple plug-in that adds only one visible element to the standard interface, a smiley/frownie face, ala iCab, that indicates whether the HTML of the page actually validates to the DTD declared in the document itself. Clicking on a frownie face would bring up a list of validation errors. This would be a great tool for site developers, making mistakes quickly visible.

It would be an even better tool for standards evangelism if it was included in the default installation of Mozilla/Phoenix. Then you'd turn the entire population of Mozilla users into nitpickers, who would hound site developers about lack of standards compliance.

From personal experience, nothing makes you fix problems faster than users regularly sending you e-mail about things that are broken. So making it obvious when things are broken would lead to more feedback, and more feedback would lead to more standards-compliant websites.

Which would be good for Mozilla, and all other browser developers who work towards standards-compliance.

Re:Validity checker and indicator

[simeonbeta2]

Excellent idea! If i had mod points i'd up you.

This makes complete sense to me based on personal experience. I work for a web shop and some users complain who see javascript errors in the status bar at the bottom of IE. It usually isn't affecting them, but users don't like to see errors!

If the major browser makers would include this feature (even when they still have code to work around non-standard stuff) html compliance would soar.

Even aside from that, having compliance validation right in the browser would be helpful to me as a programmer... I often don't go to the enormous hassle of using the w3c's online validator... a 'right in the browser' check would probably help...

Re:Validity checker and indicator

[bunratty]

I know this isn't exactly what you're asking for, but Checky is a Mozilla Plug-in that will validate the current page when you press F10. It won't help evangelism, but at least it makes it easier for web developers to generate valid HTML.

Re:Validity checker and indicator

[Anonymous Coward]

This request has been in bugzilla for a couple years now. The problem is, the gecko rendering engine wasn't designed to validate pages as it parses them, so in order to get the little smily, you'd have to program a validator into the rendering engine. Doing that would potentially slow the rendering down too.

Re:Validity checker and indicator

[JourneymanMereel]

This was discussed, and even worked on for a while, but was eventually rejected. I go agree that it would be nice to have, however.

http://bugzilla.mozilla.org/show_bug.cgi?id=4710 8

Re:Validity checker and indicator

[d-Orb]

I'd really like to see a simple plug-in that adds only one visible element to the standard interface, a smiley/frownie face, ala iCab [www.icab.de], that indicates whether the HTML of the page actually validates to the DTD declared in the document itself. Clicking on a frownie face would bring up a list of validation errors.

There is a simple solution, and it is to use JS favelets that connect to the W3 server with the URL details. Basically, you need a link to a JS link, such as this: javascript:window.open('http://validator.w3.org/ch eck?uri='+location);void%200 (and I'm just copying and pasteing from my website). You can put this in the navigation bar, and off you go :-)

There are plenty of bookmarklets which are quite useful. Some of them:

• Accessibility favelets
• Slimeland favelets (show tables and divs)

• I am sure there are many more cool uses around

  --
  Free as in freedom

Re:Validity checker and indicator

[ratamacue]

I'd really like to see a simple plug-in that adds only one visible element to the standard interface, a smiley/frownie face, ala iCab [www.icab.de], that indicates whether the HTML of the page actually validates to the DTD declared in the document itself. Clicking on a frownie face would bring up a list of validation errors. This would be a great tool for site developers, making mistakes quickly visible.

That's brilliant. I agree completely.

so...

[wastaz]

So...this is basically an article telling us what we already knew? (Mozilla rocks, IE sucks)

Can I mark it as +5(redundant)? ^_~

Re:so...

[Anonymous Coward]

I think you have that backwards. Mozilla is much nicer on Windows. With the pre-load enabled, it takes less than one second to load. Without, less than three. On the same system under Linux, it takes almost ten seconds to start (yes I have compiled with -02 and -03). Blech. Runs nicely after it starts on both operating systems though.

In this post...

[zurab]

I will describe some very useful hints to consider before submitting a story to Slashdot with a link to your own website:

- Do not run your website off of a Sharp Zaurus;
- If you do run website off of a Zaurus, at least create a Beowulf clus... ^H^H^H^H^H^H^H^H scratch that
- Just don't do it!

Re:In this post...

[Anonymous Coward]

- Profit!

a few criticisms

[mlheur]

My biggest criticism is regarding the source viewer. Yes in IE the default source viewer is Notepad, but that can be changed - there's no mention of that ability in the article.

The other problem with the source viewer is that Mozilla goes to the server to grab the source, not using the exact source displayed on the screen if you're using dynamic server side variables (PHP), whereas IE gives you the source of whatever's in memory and displayed on the screen.

Other than that I prefer Mozilla too.

Re:a few criticisms

[bunratty]

The other problem with the source viewer is that Mozilla goes to the server to grab the source, not using the exact source displayed on the screen if you're using dynamic server side variables (PHP), whereas IE gives you the source of whatever's in memory and displayed on the screen.

AFAIK, this bug was fixed ages ago. Have you tried a recent Mozilla build?

Re:a few criticisms

[javacowboy]

Yes in IE the default source viewer is Notepad, but that can be changed - there's no mention of that ability in the article.

Do you know how? I'm running IE 5.5 and I can't find out how. I can change my HTML editor, Email program, Newsgroups viewer, etc, but not my HTML view under Tools > Internet Options > Programs.

Re:a few criticisms

[platypus]

It isn't fixed 1.3.

Re:a few criticisms

[MarcoAtWork]

Amen regarding the dynamic pages (i.e. pages created as a result of a POST etc.) I have to keep an old NS4.7 around to be able to debug those, which is really stupid.

Whomever originally thought that 'view-source' should fetch the source from the webserver instead of just displaying what's already in the browser's memory should be made to debug dynamic sites for a few years while only having access to this moronic view-source functionality.

Re:a few criticisms

[Dirtside]

I'm using 1.3 (specifically Mozilla/5.0 (X11; U; Linux i686; en-US; rv:1.3) Gecko/20030312), and it does appear to be fixed.

Re:a few criticisms

[ubernostrum]

The other problem with the source viewer is that Mozilla goes to the server to grab the source, not using the exact source displayed on the screen if you're using dynamic server side variables

Using Moz 1.2.1, I'm not seeing this problem.

Re:a few criticisms

[platypus]

Yes, I tried it again it and it indeed does work if the form submission didn't cause a server error. But if it does (which is when I have to view the source for development), mozilla seems to expire the document from the cache and doesn't want you to see the page source without regetting it from the server.
So it seems a) to expire faulty pages and b) not to show expired pages. I think at least one of the two points should be fixed.

Re:a few criticisms

[Dirtside]

How exactly does Mozilla know whether a returned page is "faulty" or not? Maybe you can give me a more explicit example of what causes this, but I've been developing a heavily form-based PHP game for the last several months and haven't seen this bug come up since I was using 1.0. Also, what exactly do you mean by a "server error"? A bug in your PHP code? A bug in Apache?

Re:a few criticisms

[platypus]

So many questions ;).

Mozilla knows if a response it got is faulty when it came with a http status code which indicated something when wrong, a 404 "Not found", a 500 "Internal Server Error", whatever.
That also explains what I mean with server error.
For example go to
http://slashdot.org/this_gives_us_a_404_not_found

Right click, "view page source". Do this two times. Compare the source code.
You'll see the source codes will differ, because of the slashdot ad rotation (I tried, they differed every time). Additionally, if your connection is slow enough, you'll "feel" the delay when mozilla reloads the page.

When coding in Zope (quick plug, try it and compare to php, it's another world, I went this route myself), it's quite easy and handy to include the error tracebacks into the source of your app, enclosed in html comments, so you have to look at the source to see them. Granted, it's easy to not enclose the tracebacks in html-comments, but it's quite handy to be able to hide the ugly messages from customers on a development server.

Mostly this is no problem with zope, since it's transaction aware and rolls back everything (including sql stuff, no more db->commits and rollbacks sprinkled through your app, nice!) on errors, but the fact stays that I want to see the source of the page I'm looking at, not of something which might, or might not, tell me what went wrong.

Oh, and another gripe I have with mozilla is that it doesn't take the content of textboxes into account when you search on a web page.

Re:a few criticisms

[Dirtside]

Ok, you're right, I tested it out, and this certainly does seem to be a bug. Although I might wonder why you would need to see the source of a 404 page, and what would change from one 404 page to another ;) But nonetheless, it's still a bug. I searched on bugzilla.mozilla.org for any View Source bugs like this, and found a couple that mentioned the reloading problem... but none seemed to mention the fact that it only happens on error headers (I was running Ethereal; it showed me the headers, and only when a 404 returns does this happen). So, I added my two cents to the bug:

http://bugzilla.mozilla.org/show_bug.cgi?id=153884

So, hey, there you go. :)

Yay fer verbose article texts!

[American AC in Paris]

from the 404-file-not-*rrrrgk* dept.

American AC in Paris writes "In this thread, we will describe some very cool features in Mozilla which will enable you to quickly find the maximum load of your web site and applications."

load testing

[zzzmarcus]

He obviously isn't using Mozilla for load testing...

15 comments and already slashdotted.

Re:load testing

[HoneyBunchesOfGoats]

Hey, 15 isn't so bad, some go down with only a couple posts on the comments page. I know I sure as hell couldn't afford the bandwidth to withstand a /.ing.

The real article

[llamalicious]

Henrik Gemal writes "With this article I will subject myself to a massive DDoS produced by legions of slashreaders who will certainly turn my webserver to molten slag."

Doesn't Really Matter

[Anonymous Coward]

The only concern the developers have is "How does the web site look in IE". Any other browsers are used by dateless geek loosers anyway.

Some people never learn.

[ca1v1n]

I was about to express my pity for the poor guy whose overloaded server is hosting the article, when I realized that he's the one who slashdotted it!

Feature requests

[robbo]

I agree that HTML highlighting is great, but I've been meaning to post a feature request on bugzilla for some time: let me view-source in a tab, rather than open a new window. That should be really easy to do. I'd also love to have keyboard navigation between tabs (ok, it's probably there already, or at least in Phoenix, but I haven't found them yet..)

Re:Feature requests

[Fuzzle]

Chimera...err...Camino for OSX does source in a tab. Phoenix Tab Switching is CTRL + Tab or CTRL + SHIFT + Tab depending on which way you want to cycle.

Re:Feature requests

[Cainam]

It's already there. Ctrl+Tab cycles through the tabs. Ctrl+PgUp and Ctrl+PgDn walk forwards and backwards through the tabs.

Re:Feature requests

[slothdog]

I'd also love to have keyboard navigation between tabs

Ctrl-1,2,3, etc will switch between open tabs, at least in Phoenix.

Re:Feature requests

[Koyaanisqatsi]

Keyboard navigation between tabs: CTRL+Tab, CRTL+Shift+Tab

There some (rare) instances in which it fails, that depends on the contents of the page. I haven't researched it enough to fill a bug report yet.

Re:Feature requests

[stevey]

Keyboard navigation in tabs for Mozilla is Ctrl+PageUP/PageDown.

There are a couple of issues, if you have a n open tab which contains a site which failed to load Ctrl+W doesn't seem to close it, for example.

I'd love to have a gui for remapping keys under Mozilla, I've seen a guide to remapping some keys, but I've never managed to remap 'next tab'/'previous tab'...

Re:Feature requests

[commodoresloat]

But no keyboard tab switching in Camino :(

Re:Feature requests

[seanmeister]

"view-source in a tab, rather than open a new window"

I'm pretty sure that MultiZilla will let you do that.

Re:Feature requests

[jim3e8]

Yes, there is. Use Command-{ and Command-}. Works great.

Re:Feature requests

[TREE]

Already there, use it all the time: control tab or control pageup/pagedown.

Re:Feature requests

[josephgrossberg]

FYI, in Mozilla, they have a different effect:

CTRL-1 launches a new browser
CTRL-2 launches a new mail app
CTRL-3 launches a new chatzilla

You've gotta use CTRL-PageDown and CTRL-PageUp

Re:Feature requests

[edbarrett]

I don't know if it's a function of Multizilla or not, but I can't get view source to not open in a new tab.

Re:Feature requests

[data64]

I would really like to have it open up the source code in the editor of my choice => Emacs rather than a Mozilla window or tab.

This would give me color coding, powerful search
capability and being able to easily correct the insane formatting you find on on most generated webpages.

Re:Feature requests

[PunchMonkey]

And CTRL+Tab / CTRL+SHIFT+Tab to switch a la ALT+Tab (at least in Windows anyways)

Re:Feature requests

[Idaho]

I'd also love to have keyboard navigation between tabs (ok, it's probably there already, or at least in Phoenix, but I haven't found them yet..)

Try Ctrl-PageUp and Ctrl-PageDown.

Has been working for ages...we're talking pre-1.0 here (I guess since 0.9.something).

Apparently there should be better documentation for Mozilla, as it took me ages to find this out as well (thanks to the guy who posted it on slashdot sometime, that's when I discovered it..so now I'm keeping up the tradition :-)

AFAIK there is no quick way to see all the keyboard combinations, let alone the ability to change them to something else...

Hmm..I see a RFE coming..or does there already exist one, regarding this problem?

Re:Feature requests

[Anonymous Coward]

In both Mozilla and Phoenix, you can use Ctrl+Tab and Ctrl+PgUp/Dn as well. Personally I use these more often, as Ctrl+1/2/3/4 is a little tough on a laptop keyboard.

Re:Feature requests

[An Onerous Coward]

CTRL+tab and CTRL+Shift+tab will also work.

I would be highly in favor of Mozilla adopting Phoenix' usage of CTRL+[123]. Me like.

Re:Feature requests

[arvindn]

Ctrl-PgUp: previous tab
Ctrl-PgDn: next tab

Works in both mozilla and phoenix

Re:Feature requests

[Khazunga]

I'd also love to have keyboard navigation between tabs

In Linux, you can use Ctrl-Tab. It doesn't work on Windows, but it must surely be a pref whose default is platform-dependant.

Web Development Bookmarklets

[jesser]

Web development bookmarklets:

• shell (type JS statements to evaluate them)
• onerror status
• onerror alert
• test styles (type CSS rules; it applies them immediately)
• zap style sheets
• view style sheets
• view scripts
• view variables
• generated source
• partial source (not as good as "view selection source" in Mozilla's context menu)
• show blocks
• ancestors (makes status bar show what you're hovering over, in the format "BODY > DIV#content > DIV.blog > DIV.blogbody > P")
• make link (create HTML to link to a page)
• show named anchors

You can do many of these things with the DOM Inspector or JS Debugger, but boomkarklets usually require fewer clicks and are easier to learn. All of these bookmarklets work in Mozilla, and many of them also work in IE or Opera or both. Web developers might also find these validation bookmarklets and keywords bookmarklets for scripters useful.

Re:Web Development Bookmarklets

[ManxStef]

Another cool set of tools are the Mozilla Sidebars available over at Netscape's DevEdge.

Basically they add quick references into the sidebar for a variety of official standards as CSS2, CSS2.1, HTML 4.01, DOM 2, XSLT 1.0 Reference, and the Gecko DOM Reference.

Now I've installed them I use them all the time; and to think for ages I thought Mozilla's Sidebar was useless! Very handy, it is.

Re:Web Development Bookmarklets

[sepluv]

Also see www.favelets.com.

Done but with errors

[StrandedOrg]

You would think a site about debugging wouldn't have any errors on it... oh well.

Must be a pretty big lizard...

[cscx]

Bite your head off, man.

Re:Must be a pretty big lizard...

[Anonymous Coward]

the mod who didn't catch that reference is *so* kicked off the team.

Variable timeout?

[alispguru]

What I could really use is a browser that let you set the timeout (when waiting for an HTTP transaction to come back) to something large, or better yet to turn it off completely.

I write web applications in Common Lisp, so when developing I have the Lisp top-level open and running. Errors on the server side pop up a Lisp debugger on the thread doing the transaction, I can poke around in the stack, figure out the problem, even fix it and continue - but only if I do it quickly, before the browser decides it's waited on me long enough and closes the connection, which causes a broken-pipe error on the server side and can clobber my debugger session.

So, anybody know how to make any decent browser never time out? Mac OS X browser preferred, but I'll take Linux or Windows in a pinch.

Re:Variable timeout?

[Anonymous Coward]

To make a browser never time out, you'd first have to convince people to stop using TCP/IP for sending data between the server and your machine.

Re:Variable timeout?

[the_consumer]

try taping down F5 ;)

Re:Variable timeout?

[Anonymous Coward]

Maybe if you remove the timeout code and recompile the browser?

Re:Variable timeout?

[bunratty]

Try going to the URL about:config and changing the preference network.http.request.timeout to a large value.

Re:Variable timeout?

[fish waffle]

So, anybody know how to make any decent browser never time out? Mac OS X browser preferred, but I'll take Linux or Windows in a pinch.

wget might do what you need. The -T option lets you play with the timeout if 15 minutes isn't enough.

Re:Variable timeout?

[alispguru]

Try going to the URL about:config and changing the preference network.http.request.timeout to a large value.

So, I mung my preferences.js file, adding something like:

user_pref("network.http.request.timeout", 3600);

I'll try it. Thanks!

Re:Variable timeout?

[mmcshane]

So, anybody know how to make any decent browser never time out? Mac OS X browser preferred, but I'll take Linux or Windows in a pinch.

In any gecko browser (including chimera/camino), set the network.http.connect.timeout property to an integer (number of seconds). Default is 30.

You can access this property through the about:config url or add the line

user_pref("network.http.connect.timeout",999)

to your user.js file

Re:Variable timeout?

[poot_rootbeer]

So, anybody know how to make any decent browser never time out?

1. Download the Mozilla source
2. Find the code that closes the connection when a timeout interval has been reached
3. Comment it out
4. Recompile it
5. Run

(You'll notice I've left
6. ???
7. Profit!
off the list, since those are already part of the standard Netscape/Mozilla business plan)

Re:Variable timeout?

[Gerv]

You can configure the timeout on IE by editing the registry - Google for ReceiveTimeout. It's almost certainly also possible by setting a Mozilla pref; or, if not that, then a recompile. Lastly, you could get your server to use HTTP 1.1 persistent connections, by not sending Connection: Close on a connection opened by the webserver. They'll stay up for hours.

Gerv

Re:Variable timeout?

[jesser]

Or you can right-click on the pref in about:config.

Re:Variable timeout?

[eyeye]

Are you seriously suggesting that someone recompile mozilla for something that is changeable in user prefs?

Have you ever compiled mozilla?

anyone?

[caino59]

have it in their cache?
ill gladly put up a mirror...someone send me the files ;o)

Re:anyone?

[caino59]

lil late

but here is one anyways.

clicky

(and yes, i know it displays incorrectly...deal with it)

Re:Feature requests OT

[schmink182]

Navigation between tabs:

• Forward - control + tab
• Backward - control + shift + tab

full text

[Anonymous Coward]

Using Mozilla in testing and debugging web sites
Mozilla is a great tool to use in developing web sites and web applications. Not as a development tool itself, like an editor, but as a testing and debugging tool. In this article I will describe some very cool features in Mozilla which will enable you to quickly find and debug errors in your web site and web applications.
JavaScript Console
A lot of the errors found in todays web pages and web applications are caused by JavaScript errors. Most of the time they're very simple errors. This is in my opinion the most common reason why sites doesn't work in Mozilla. But these errors could easily be avoided. With Internet Explorer you are, if you have set the correct setting, presented with an almost useless dialog that "page contains errors". The dialog doesn't let you copy the error to the clipboard for starters. If you want better debugging in Internet Explorer you can install the Microsoft Script Debugger which is a debugging environment for scripting in Internet Explorer.

Picture 1: JavaScript error in Internet Explorer
With Mozilla on the other hand you have the JavaScript Console. All JavaScript errors are logged here. So if you keep the JavaScript Console open while testing your site you can on-the-fly see if there are any JavaScript errors. An indispensable tool for developing web sites and web applications.
The JavaScript Console reports the error and the filename and the line number. Furthermore the context of the error is shown. This makes it very easy to get a clue about where the error is and what caused it.

Picture 2: The Mozilla JavaScript Console with errors
You can right-click on each error and copy it to the clipboard. The JavaScript Console could still need a lot of improvements. You can't save all entires to a file and it has problems with wrapping.
The JavaScript Console can be started via Tools -> Web Development -> JavaScript Console.
JavaScript strict warnings
JavaScript strict warnings are messages that are produced inside the JavaScript Engine, which is in the core of the browser. JavaScript strict warnings are produced in all browsers. In both Mozilla and Internet Explorer and Opera. But only Mozilla shows them. JavaScript strict warnings are warnings from the JavaScript Engine about some mistakes in the client side JavaScript code. These mistakes, unlike JavaScript errors, do not stop the execution of the web page. But they do slow it down a bit, since they produce an exception inside the JavaScript Engine.

Picture 3: JavaScript strict warnings
In other browsers than Mozilla these exception are not available to the developer but with Mozilla you can access these warnings. This puts you in the driver seat for making 100% valid JavaScript code!
A common mistake in JavaScript is to declare a variable twice:

var response = true;
var response = false;
This will produce a JavaScript strict warning saying

"redeclaration of var response"
The correct way is of course:

var response = true;
response = false;
The JavaScript Console can be enabled in nightly builds via Edit -> Preferences -> Debug ->. If you run a official release you can use the javascript.options.strict pref which can be set by entering about:config in the Location and hitting enter.
More info...
Tackling JavaScript strict warnings

Cookie Control
Most web sites and web applications nowadays are using cookies. Debugging cookies can be a problem. But not if you use Mozilla. If you're using Internet Explorer the only option you have from within Internet Explorer is to delete all current cookies. If you want to delete all cookies from a specific domain you have to manually delete the Internet Explorer cookie files which are located in the %USERPROFILE%\Cookies directory. Since the files are in a unknown text format I'm not sure it you can delete or edit specific cookies from a site or domain.

Picture 4: Cookie Manager in Internet Explorer
With Mozilla it's all

Something I miss when testing websites locally...

[starvingartist12]

One of the reasons I switched from Netscape Navigator 3 to IE 3 was that when I viewed the source of a website on my local hard drive (ie: testing and debugging), IE would open the actual file in Notepad (or any other editor), while Navigator would open either a non-editable page source window or a cached version of it in Notepad (no, I'm not going to use Composer).

The same is kinda true with IE6 and Moz today. IE6 lets me move around my local prototype website and click on a large Edit button. This simplifies editing static html pages for me.

But hey, I still think Mozilla is great and invaluable for testing and debugging code. The Javascript Console mentioned in the article has saved me tons of time. I totally recommend it as the first thing to use to check for scripting errors.

One final though... IIRC, IE5.0 has had a View Partial Source tool available as part of a powertoy (er, Web Development Accessories) for web developers.

Mozilla keys

[Odds]

For mozilla:

• Forward: Control-PgDown
• Backward: Control-PgUp

Re:Something I miss when testing websites locally.

[bunratty]

If you try File | Edit Page in Mozilla, I think you're in for a big surprise!

Also, CSS debugging help

[robbo]

CSS validation would be cool (of course there's always w3.org)

I'd love to see something that helped me with CSS layout- a way to put big bright borders around divs and highlight their containing blocks, etc. I *don't* want that in composer, mind you, because I prefer to play with the raw source in an editor and reload the page to see how it looks.

Re:Also, CSS debugging help

[barcarolle]

Actually, that's what I always hoped Composer would implement. WYSIWYG editing of CSS2 fundamentally involves layout of boxes in different display modes such as block, inline, and the table and list modes. Unfortunately, Composer (and all other WYSIWYG page editors) adopt a word-processor motif for editing web pages that poorly corresponds to the actual markup and style information being constructed.

Re:Also, CSS debugging help

[grayrest]

Use Jesse Ruderman's test styles bookmarklet. It's about the coolest thing ever. If you combine it with a custom keyword and the the DOM Inspector and you can debug any CSS problems in no time.

Re:Also, CSS debugging help

[benw01]

I'd love to see something that helped me with CSS layout- a way to put big bright borders around divs and highlight their containing blocks, etc.

The test_styles bookmarklet might fit the bill - it pops up a little window where you can type in CSS rules and have the page you triggered it from dynamically update based on the rules you enter. I put this bookmarklet into my Personal Toolbar Folder in mozilla, so it's just a click away.

There's HEAPS of useful bookmarklets linked off that page too. The javascript shell is amazing.

Re:Also, CSS debugging help

[crazyj]

You might want to check out the Show Divs bookmarklet.

iCab, yes

[ianscot]

I've used iCab in this way, to test sites under development. The face that shows you errors with a click is simple and effective. Heck, DreamWeaver doesn't show you errors so easily, maybe because its own scripting "behaviors" would cause so many.

A comparable plug-in would be cool, and would underscore Mozilla's standards-compliant MO. (That's its M.O.zilla, citizen.)

The iCab face is dang particular, though, and it seems like your page's rating has to do with the number of standards violations. You don't find one site in 50 that get a real smile out of the thing. (Cranky bugger.) I'd rather the rating had to do with the seriousness of errors -- you have an onLoad to an undefined function, BIG FROWN for you. The difference between one and ten missing Alt tags isn't that big a deal by comparison.

Not sure how to practically have the browser make judgment calls about seriousness, though...

Re:iCab, yes

[Ed Avis]

A missing alt attribute isn't an HTML error - you can write valid HTML without alt attributes. It's more of a warning from some lint-type tool. A smiley face should warn about things that are definite, unarguable errors, like HTML that doesn't validate. Really this is a yes or no question.

No wonder mozilla is so bloated.

[Anonymous Coward]

If youre going to test websites, use a proper ide. Such as dreamweaver or quanta, even vim/emacs are more suitble for the job. Then you call your external browser (or embed its rendering engine in the ide) to test it.

Mozilla dosen't even support editing in "view source". Other browsers let you call your own external editor, but not mozilla!

The truth hurts

Re:No wonder mozilla is so bloated.

[Christianfreak]

Interesting? Everytime a Mozilla article pops up so does this comment. Redundant is more like it.

But I guess however I'm in a troll feeding mood. So lets examine a few facts:

A) Mozilla is 4 apps in one. Some people like this, some people don't. If you don't like it then use Phoenix or one of the other stand-alone browsers

B) For an app that is four apps in one, Mozilla has a relatively small foot-print. I've got several tabs open and the mail client right now and its only using 48 megs (mostly just idle in swap) that's on Linux.

C) IE (which I'm assuming you run) starts up faster than Moz, why? because IE starts up with windows so you actually run it all the time. Besides in Windows you can run mozilla in your sys-tray so that you get a fast start up just like IE.

D) I don't have turbo mode on (I don't even think its availiable in Linux) and still Moz starts up very fast on my 1 Ghz Athlon

And E) a little extra code to tell you where Javascript errors wouldn't really be creating much bloat even if there was bloat in the first place.

Mozilla does have a way to edit source, its call 'Composer' which is an HTML editor. View Source isn't for editing files, and the fact that MS decided to use the wonderful invention of Notepad for their view source has always annoyed me.

And on a side note as a web developer, Mozilla with its web development tools plus Vim is far easier to use and more powerful that DreamWeaver or anything else out there IMHO.

Re:No wonder mozilla is so bloated.

[Anonymous Coward]

Mozilla has a nifty little bug/feature that lets you edit in an external editor. Try prepending view-source:view-source: to the URL. That is TWO view-sources. It will then use your majick file, or file extension map to open a suitable editor.

Re:No wonder mozilla is so bloated.

[Anonymous Coward]

'Mozilla dosen't even support editing in "view source".'

It's VIEW Source. Not EDIT Source silly.

Re:No wonder mozilla is so bloated.

[Anonymous Coward]

You are a fucking moron. Can you step through javascript in Dreamweaver? I don't think so. If you need to edit the source code, just open it in composer, which gives you 3 different options on how you want to edit the html/js. Have you even tried using Mozilla, if not just shut the fuck up.

Nothing beats....

[Richard_at_work]

print or echo. or print_r, or vardump. or...

I for one, have never found myself doing any debugging in the browser past printing to it from php. Thats how i debug, simple. My html is rarely at fault, and if it looks good in IE or Phoenix, then it rocks. I dont do javascript. I dont do stuff client side. Theres nothing for the client to debug, job done.

As an aside, this article struck me as less of a "handy things in mozilla" and more of a "oooh look at what mozilla does over IE". It really struck me as that, another flag waving rather than truely informative.

Re:Nothing beats....

[thelexx]

You must not create web apps that require much data entry then, because pushing all validation to the server side is simply retarded. Regardless, the fact that the article was of no use _to you_ doesn't make it propaganda.

Re:Nothing beats....

[Richard_at_work]

I design and write a couple of apps that are heavy on end user input, and i do ALL validation server side. Its faster, less code to send client side, I have a guarenteed enviroment to validate in, AND I CAN TRUST MY VALIDATION 100%. Anything that dynamically alters client side information/input is asking for trouble.

I didnt say the article wasnt any use to me, i found it informative. It still came across as flag waving tho, and i deplore that sort of thing.

Re:Nothing beats....

[AeiwiMaster]

I agee with you fully ?

But I just made a page where that didn't work well
check
http://all-technology.com/eigenpolls/ atsbooks/

Knud

Holy crap, that's an *excellent* idea

[revscat]

Anybody know how to go about requesting this feature? Or, better yet, know how difficult this would be to implement? This kind of changes seems rather fundamental, and I would imagine difficult to do. However, Moz seems pretty well architected, so I wouldn't be surprised if this turned out not to be the case.

Man. If I could just see an icon that indicated valid/invalid documents, life would be so much easier.

Re:Holy crap, that's an *excellent* idea

[bunratty]

Take a look at Bug 6211 in Bugzilla: http://bugzilla.mozilla.org/show_bug.cgi?id=6211

One more feature request: IE emulation

[robbo]

Don't laugh, I'm serious. It would be nice to see that how page I'm working on renders in IE without switching OS and browsers. Most of the layout bugs and standards-defiance in IE are well documented and it shouldn't be too hard for Moz to behave like IE if the user so desired.

I know, I know, I should post these requests on bugzilla..

Re:One more feature request: IE emulation

[barcarolle]

Think about how much time they would have to spend trying to reverse-engineer IE's quirks. They'd have no time at all left over to actually develop the rest of Mozilla. However, there's nothing stopping a group of outside developers from developing a Mozilla "add-on" that does this. Mozilla's open source, remember.

Re:One more feature request: IE emulation

[duffbeer703]

If you just clone IE, what's the point of having Mozilla?

The whole point to mozilla is to compete with IE. They are producing a superior, standards complaint browser that will eventually force microsoft to clean up their act.

Re:One more feature request: IE emulation

[robbo]

Fair enough, but the reality for most web designers is that IE represents ~96% of their audience. From a debugging point of view, I *need* to open a page in IE, just to check that it lays out properly. Wouldn't it be nice if Mozilla obviated that need? I don't think it should be a default setting, but something the user can switch on when needed.

Re:One more feature request: IE emulation

[Anonymous Coward]

Most of the layout bugs and standards-defiance in IE are well documented and it shouldn't be too hard for Moz to behave like IE if the user so desired

Really? You don't think it would be hard to re-implement an entire rendering engine? I guess you don't program.

On the other hand, on Windows, it is possible to embed to a browser fairly trivially. I doubt that would go over well thouhg, since it would exclude all of Mozilla's other target markets.

Re:One more feature request: IE emulation

[Anonymous Coward]

I see on your website that you have done some programming. But still, I can't believe you seriously meant that implementing an entire other rendering engine and every single quirk, would be not too difficult? Really?

And I meant to say that its fairly trivial to embed IE's browser on Windows, not any browser.

Re:One more feature request: IE emulation

[robbo]

Well, if you put enough monkeys at enough typewriters.. ;-)

I haven't looked at the rendering engine code since the early gecko days, but you can bet that there are a lot of quirks in there to get pages on major sites that were designed for IE to render as the designers intended. Gecko may be standards compliant, but for ages the Debug menu had a long list of major sites that were important to render correctly.

While I agree with you that it's probably not as easy as I suggested, I do think that the moz developers have a good sense of the major IE non-compliances and how they can be reproduced.

Re:One more feature request: IE emulation

[mckayc]

Read the post, moron, he said he would like to have *OPTIONAL* IE emulation for TESTING PURPOSES. He didn't say he wanted to make it the default rendering mode for Mozilla or would like Mozilla to be IE.

Re:One more feature request: IE emulation

[barzok]

So you want bug-for-bug compatibility with IE? I doubt all the IE rendering bugs and oddities are even documented. Sounds like the same struggle WINE has to go through with the Windows APIs.

Re:One more feature request: IE emulation

[duffbeer703]

if you can run ie in mozilla, fagboy, ie becomes the most common denominator.

Re:One more feature request: IE emulation

[Anonymous Coward]

Yeah, it would suck if that were to happen. God forbid IE should ever become the most common browser.

BTW, it's "most common" or "least common denominator".

Re:One more feature request: IE emulation

[robbo]

Certainly not bug-for-bug, but perhaps at least the major layout bugs that affect flow, div widths, etc. (and maybe some of the JS bugs, but I think that's less of an issue).

I think the nice thing about the standard is that document flow is pretty linear, so at each element moz can simply ask "What goofiness would IE do here?" Obviously IE was designed with the spec in mind and the spec practically defines the algorithm that you need to use for layout. But, as they say, the devil is in the details.

Re:One more feature request: IE emulation

[mooman]

But given the fact that there are some significant differences between IE versions, which one would you emulate? All of them? Jeez.. talk about bloat, all for the purpose of emulating quirks!

And we won't even get into IE differences between platforms either...

Personally, one of the singularly biggest and best features of Mozilla is that it plays well with others. You can have multiple versions of it installed at a time on one box. Any webdev worth their salt will have a copy of every major browser that intend to support already installed on their box(en). It's been one of my biggest complaints about IE since the IE3 days. I can only have one IE version on a box, usually without even the means to uninstall it... Makes testing cross-platform a pain in the @ss...

Re:One more feature request: IE emulation

[barzok]

How do you account for things like the various ways IE calculates widths? IIRC, some versions use "width" to mean the whole element, padding, margin, border and all, while the spec (and maybe a version of IE) says "width" is the width of the actual content.

I think IE was designed with the spec in mind more from the standpoint of "this is what the spec says, how can we screw with it enough to not be noticed but still mess things up?"

Re:One more feature request: IE emulation

[robbo]

Yes, well I'm still waiting for an IE release on my os of choice.. ;-) If I want to test something on IE I have to reboot. Even harder when the page I want to look at is hosted by the box I have to reboot.

Re:One more feature request: IE emulation

[robbo]

I think the width bug is the easiest to emulate. (iirc, it's only IE 5 that has this problem, they fixed it in 6). It's quite simple to adjust your width computation based on the spec or based on IE's broken idea of what it means..

Re:One more feature request: IE emulation

[Anonymous Coward]

*sigh*

Mr. mckayc ... if you're trying to convince the orignal poster that they've made a mistake, insulting them is probably not what you want to do. I've yet to see someone convinced of a point when you call them a moron. More often, they just go on the defensive, and irrationally refuse to acknowledge your most informative points.

I'm sure this is possible, but...

[Anonymous Coward]

how do I make it so Mozilla opens an HTML editor other than Composer?

Re:I'm sure this is possible, but...

[jcupitt65]

you can click on the "html source" tab at the bottom and just use it as a notepad-alike

Opera's handy access to alternate rendering modes

[wfmcwalter]

Opera has some one feature that (as far as I can tell) mozilla doesn't. It has a really easy way to change the rendering enging to run in a number of different modes, including small screen, text, accessibility, high contrast, and the rather fun CBM=64 lookielikie mode. Mozilla can do some of this, but Opera's handy menu access (view->style->usermode->xxx) makes testing the many pages of ones website for accessibility quite easy.

I'd really like a "tandem" mode, where the browser would automatically open each page in both normal and accessibility, or normal and text-only, modes (in two parallel windows, naturally).

In testing my own website for IE6, Mozilla-1.3, and Opera 7, I seem always to find the same thing:

• Opera is the standards fascist
• Mozilla is the make-it-work-somehow guy
• and uniformly, IE is the "problem child"

Also gave them

[Archfeld]

the ability to claim that the reason it was not working was the WEB and not their crappy code or poorly designed sites. My company has been migrating heavily to the intranet/internet, and then looking stupid when a 3rd party router somewhere cuts off all our customers. Someday we are going to find all the left brains that were removed from our managers heads...

Mozilla vs IE

[Tolkien]

yaddi yaddi yadda

Although IE is buggy as all hell, it's got a couple things that I wish Mozilla had, like when you middle click and can go up and down a 10k text file in 2 seconds flat, though I love Mozilla's tabs et al, I hope Mozilla developpers impliment the middle-click scrolling soon, most every windows app can do it except Mozilla!

Re:Mozilla vs IE

[Yunzil]

I hope Mozilla developpers impliment the middle-click scrolling soon,

In the meantime, there's this

Re:Mozilla vs IE

[Tolkien]

I wandered mozdev.org for a minute because the given address didn't work, and then I found http://autoscroll.mozdev.org.

Thankya for the tip-off! Now I have the best of both worlds in Mozilla :D

Re:Feature requests OT

[Naikrovek]

+ and + are easier to remember and easier to type.

fark

[Naikrovek]

[CTRL]+[PGUP] and [CTRL]+[PGDN] are easier to remember and easier to type.

dremweaver

[b17bmbr]

import com.suit.flameproof;

if you do serious web dev., then DW is far from expensive. and it will generate correct html, even if you use lots of fireworks dhtml, or layers. it will do all your checking, and it can check for browser differences. no, it's not open source, but at least DW is very platform nuetral. it does CF, asp, php, etc. yes you need windows or a mac, and that is a drawback.

Tabs? Look At Mozdev.org

[imtheguru]

While Moz developers add your feature request, you could look at the Multizilla project at mozdev.org provides a toolbox of options. Tabbed "everything".

Anyone else get this?

[Micah]

As others have mentioned, Ctrl+PgUp/PgDn cycles through the tabs.

But often when I press those, or Ctrl-T to open a new tab, it often does the action more than once. Sometimes up to five times!

Like, I'll press Ctrl-T and up to five new tabs will open. Or, if I have 4 tabs open, and I press Ctrl-PgUp, it will cycle through ALL of them and go back to where I started!

It seems quite random. Sometimes it will correctly only do it once. Sometimes twice.

It has happened at least since Moz 1.2, and maybe earlier, and also in Phoenix.

And those two operations are the ONLY keys that do anything like that. Even Ctrl-W (close a tab) has never registered more than once (thank goodness!!!).

Re:Anyone else get this?

[TobiasSodergren]

I experienced this, but only in redhat 8.0. Since I switched to mdk9.1, it stopped. I was using the same version of mozilla in both distributions.

Another thing happening to me in rh8 was that when erasing text in a textbox in mozilla, it didn't stop erasing text when i release the backspace button. It was like mozilla was slower than my keyboard repeat rate, and the keyboard buffer was filling up. This does not happen in Mandrake 9.1 either.

Re:Anyone else get this?

[Micah]

Interesting. I *am* using Red Hat 8, and *sometimes* I may have experienced your text-erase thing. I think that might only happen when Mozilla is unusually busy, like with an intensive Flash animation or something.

Re:Anyone else get this?

[juhaz]

I had this - on RH8. Seems like there's something here...

It mostly happened with galeon and not Mozilla (but then again, I didn't use Moz very much) and it DID sometimes happen on ctrl-w as well, you can guess how damn annoying that is.

Haven't seen it happen on RH9 ... yet.

Re:Anyone else get this?

[Micah]

Haven't seen it happen on RH9 ... yet.

Cool, maybe that's the excuse I need to upgrade. :) I planned to do so, but I tweaked my RH8 so much I'm afraid an upgrade would screw me up.

Re:Anyone else get this?

[BrokenHalo]

Nope. I'm using the Dropline build with Slackware, and it works flawlessly.

Live HTTP Headers??

[chickenbird]

Wow, it said you could go to Tools -> Web Development -> Live HTTP Headers to see the HTTP Headers, but there is no such item in my Mozilla 1.3 Tools menu.

Hmmm, looks like they haven't implemented it for Macintosh versions.

The other nice thing to have would be an item in the DOM inspector that would show you the XPath for the selected node.

Mozilla's gratuitous changes drive me nuts

[OYAHHH]

I'm,

A big fan of Mozilla, but I do have to admit that I would prefer it if things remained a tad bit more stable.

For example, from the 1.2 to 1.3 release of Mozilla the "New Tab" popup menu item moved from the 0 (zero) position in the popup menu to the 2 position.

From a day to day useability standpoint it's annoying for the menu's and the like to change around but just try to write certain automated test programs with that sort of thing going on.

I know that Mozilla is usually advertised as "test platform" but that doesn't mean that it also should serve as a point of frustration for those who would like to be able to count on a feature existing from one dot release to another.

Other than those sorts of things I love the darn thing.

Over...

Re:Mozilla's gratuitous changes drive me nuts

[Gerv]

just try to write certain automated test programs with that sort of thing going on.

Are you doing automated testing of web apps? What tools are you using? I've been looking for something that does that...

Thanks,

Gerv

Re:Mozilla's gratuitous changes drive me nuts

[jesser]

That change wasn't "gratuitous". It was made to reduce the chance of accidentally clicking "close other tabs" while trying to click "close tab" (in bug 191826), without removing the "close other tabs" command completely (which was the initial fix, I think in bug 103354). If jag decides the change didn't help enough, he'll try something else.

This type of change temporarily sucks for people who download every version of Mozilla, but it's better in the long run and it's better for people who only use major versions (such as Netscape).

Re:Mozilla's gratuitous changes drive me nuts

[cymen]

I agree on the "open in new tab" menu change. My brain had that menu memorized so I never had to look at the menu. Little changes like these are annoying -- the last one is new tab opening the home page instead of blank like normal. I sometimes wonder if these little annoyances aren't the result of political infighting among the Mozilla team. I might be dead wrong but I find it hard to imagine any other explanation.

Re:Mozilla's gratuitous changes drive me nuts

[skt]

Well, I think you answered your own question about mozilla being a technology test. You can't be on the bleeding edge of browser technology and have stable platform I am afraid.. netscape 7 is more polished if that is more important to you. Those things are changed because they actually improve usability.. the old setting was found to be bad, and so they improved it in the next version.

Re:Mozilla's gratuitous changes drive me nuts

[friedegg]

While I'm with you on the tab menu (it took me many many uses to re-memorize it), I like the new tab opening my home page (Google.com). I always wished it would do that before, and it started doing it, and I was happy. But I don't see any reason this option couldn't be in the "tabbed browsing" prefs.

Re:Mozilla's gratuitous changes drive me nuts

[ubernostrum]

I know that Mozilla is usually advertised as "test platform" but that doesn't mean that it also should serve as a point of frustration for those who would like to be able to count on a feature existing from one dot release to another.

Considering the change you're complaining about was made for usability reasons, do you really want Moz to keep a bad usability situation just so you don't have to make a small change in automated tester?

Keyboard Navigation Between Tabs

[mhesseltine]

Ctrl+PgUP / Ctrl+PgDown

Don't feel bad, someone pointed it out to me also.

Re:Live HTTP Headers?? - it's an add-on

[Tolkien]

It's an add-on, and the link to get it was posted earlier.

JS & JSP

[javacowboy]

Two Things I like about debugging apps in Mozilla:

1) JavaScript debugger. It's WAY better than IE and far more specific. IE will give you a cryptic error message telling what line of code generated the error, but in what file? (js, html or jsp?) Mozilla's JavaScript debugger will tell you the filename, method name AND line number.

2) JSP debugging. If my JSP page is broken in IE, IE will simply give me a page not found error. Mozilla will point out the line number and the part of the line that's causing an error, as javac would when I'm compiling a Java file.

Re:JS & JSP

[JJGreenaway]

> 2) JSP debugging.

It won't be Mozilla doing that. It'll be the server. IE by default hides those error messages though. Try going to IE -> Internet Options -> Advanced and uncheck "Show friendly HTTP error messages".

Re:Opera's handy access to alternate rendering mod

[Anonymous Coward]

Opera is the standards fascist

Opera 7 is definitely better than 6 in this regard, but there are still a few things it just doesn't do right. For example, alternate text when an image loads should follow css rules. Instead, in Opera, it's just black text... which doesn't work if you're using a black background.

Oh sweet irony...

[decepty]

As I loaded the article on my computer at work (IBM NetVista P4 - Win2k)using IE I got "Error on Page" on the status bar. Oh the irony of it all...

Opera feature (offtopic)

[illsorted]

Opera provides a shortcut (Ctrl+Alt+V) that uploads the source of the active frame to the W3C validation page automagically and displays the validation results.

Sweet sweet irony...

[decepty]

Trying to access the page with Internet Exploder from my desktop at work (IBM NetVista P4- Win2K) I managed to get an "Error on Page" notice on the status bar... Yes, I concede, Mozilla is a far superior browser.

Re:Feature requests OT

[Captain Large Face]

In additional to the parent, CTRL + Page Up can be used to go forwards, and CTRL + Page Down backwards.

Sounds like the quirks mode

[jeti]

Mozilla has a "quirks" mode that is trying to be as forgiving as possible.
But as soon as you properly define the DOCTYPE, Mozilla only renders correct HTML.

Re:Sounds like the quirks mode

[bunratty]

It's not that simple. You have to specify certain DOCTYPEs (e.g. HTML 4.01 Strict) for Mozilla to use Standards mode instead of Quirks mode. However, Standards mode does not mean that Mozilla recognizes only valid HTML -- it only means that Mozilla follows the standards and the HTML standards say nothing about how to handle broken HTML. Mozilla will still try its best to make sense of tag soup in Standards mode.

If you want Mozilla to render only valid code, use XHTML with a proper MIME type. If the XHTML is not valid, Mozilla will display only an error message. But using XHTML with the proper MIME type will cause IE to display the XHTML code without rending it, so you can't do this for Internet accessible web pages.

More on Cookie Handling

[Alan]

The article notes that the cookie allow/deny dialoge is almost identical between the two browsers, but misses one huge plus for mozilla. The IE cookie confirm dialoge doesn't save it's state. I personally browse with the cookie set to "ask me" each time, and I look at the cookie that is being set, make sure my IP address or other personal-looking information is in there, and allow or deny.

In mozilla the "more" dialoge starts up open if it was open last time, but the IE dialoge always starts closed, so I have to hit "more info" each and every time. Because of this mozilla is a big winner there for me, just from this one small detail.

The nay sayers will say "no one does that", but I say that for the minority of us out there, it *does* help, and the majority will never see or be affected anyway...

Re:More on Cookie Handling

[gemal]

I've added this to the article. Thanx for the info! Henrik Gemal http://gemal.dk

Gotta Use It

[E-Rock-23]

Being a small-peanuts web designer, it falls on me to debug in quite a few browsers. Mozilla and Opera are my "primary use" browsers, and most sites I design (1, 2, 3), I debug using those two, IE (which I hate to no end), Netscape, and if I can get my hands on someone's account, I'll also see how it looks under AOL's browser. Gotta figure that most people are using IE (which likes to use its own settings for viewing text, bad M$), a nifty chunk uses Netscape, and of course we of the nerd set that like to use other options.

It's all about having your site viewed by everyone, and making it look as good as it can for any browser.

Still room for improvement.

[AeiwiMaster]

1) If mozilla could do for HTML what it does for javascript.
Have a live console which shows the HTML errors
this would be very useful for web development.
I do know about the online validators but
normal development take place behind a firewall
using dynamic server-side scripting,
so every time your change the state of a page you
would have to save it and upload it to the validator.
a very slow process.

2) In the cookie tool i need a function
which remove the cookies from the current site.
So you don't have to look trough a long list of cookies
every time you test your cookie code.

3) In view source i need a function which could auto format
the HTML code.
Many auto generate pages have there HTML in one long line.
If Mozilla could auto format the code it would be useful.

4) Finally I would like the spell checker to automatically
to mark every misspelled word on the current web page.

Re:Still room for improvement.

[gemal]

2) it's very difficult, since remote images, stylesheets ads etc could also set cookies. So it's kind of complex.

But it could be cool!

Re:Still room for improvement.

[elemental23]

I do know about the online validators but
normal development take place behind a firewall
using dynamic server-side scripting,
so every time your change the state of a page you
would have to save it and upload it to the validator.

I had that exact problem until I installed the WDG HTML Validator on my development server (if you're using Debian, just do "apt-get install wdg-html-reference").

When I have something in development, I add a bit to my global footer saying something along the lines of:

if (mode == "dev") {
print "<a href=\"$WEB_DEV_SERVER/validator/validator.pl?url= referrer\">Validate this page</a>";
}

It's extremely handy to have around.

Opera anyone?

[lpret]

Honestly, how many opera users out there are quietly giggling and counting how many ways there are to switch between tabs

• Ctrl+Tab
• 2
• Alt+Pg Down
• Ctrl+F6

And if we can use our mouse, we can right-click and use the scroll wheel to bring up the little menu. Please -- save yourself some time, get opera.

Re:Opera anyone?

[dvdeug]

how many opera users out there are quietly giggling and counting how many ways there are to switch between tabs

Because that's exactly what I need - four different ways to switch between tabs.

Please -- save yourself some time, get opera

And instead of adding a minor improvement to an existing browser, or reading the documentation, I should go to a completely different, proprietary, browswer. No thanks.

Re:Opera anyone?

[I. M. Bur]

* Ctrl+Tab

Works in Phoenix.

* 2

Ctrl-2 works in Phoenix.

* Alt+Pg Down

Works in Phoenix.

* Ctrl+F6

Works in Phoenix, although it does something else ;P

Again, you were saying?

DOM inspector

[barzok]

When you click on a node in the tree view of DOM inspector, it flashes a border around the element in the rendered view.

Re:DOM inspector

[Anonymous Coward]

Mod this up! This is possibly the easiest way to do it, I use this feature all the time

autoformatting source

[KnightStalker]

http://mozilla-evangelism.bclary.com/sidebars/

The sidebar here will (among other things) auto-format the HTML for the current state of the page, after any Javascript document.write() or DHTML operations. It's rather slow though.

Re:Opera's handy access to alternate rendering mod

[ubernostrum]

Opera's nice, but Opera 7 has two show-stoppers for me. First it doesn't have Mozilla's standards-fascist XHTML rendering (if I serve Moz XHTML with the correct MIME-type, it'll validate and tell me where any errors are).

Second, it has a behavior that's personally annoying to me; Opera emulates the Windows "single click in the address bar selects the entire address" behavior, which is frankly stupid under X where that has the side effect of overwriting the clipboard.

Re:dreamweaver [SIC]

[mooman]

As someone who has used both DreamweaverMX and ColdFusion Studio, lemme say that I absolutely can't stand Dreamweaver.

My office tried to "upgrade" us to the latest Dreamweaver and I found so many things that I consider unusable that I told them where they could stick it and I went back to using CF Studio. If you're curious, here's the list I started compiling about my beefs:

1. The ability to detect when someone else changes an open file is broken. Even losing focus triggers it now. Had to disable entirely.
2. Lost ability to have editor reopen all the files you were previously working on
3. Lost row/column indicator for current cursor location (but still got the oh-so-useful download timer instead which only measures raw bytes - almost irrelevant in our complex CF docs)
4. Lost the word wrap column reminder (cosmetic, but was nice to have)
5. Lost the entire icon bar that had the indent/outdent buttons
6. Can't edit/change/move the individual tabs of code bits to insert [EDIT: later found that you can. Must manually edit an XML flatfile and a TXT file to make any changes]
7. The process of forcing a case change to your tags now requires you to make a setting in your preferences, save your file, close your file, reopen the file, then turn that preference back off. It used to be a single menu option with 1 click.
8. They removed the ability to toggle the sidebar file explorer on and off with F9. I found a kludgy workaround (and had to remap F12 to F9) but toggling it now removes *all* the panels, including the insert one, not just the sidebar.
9. File->Open only lets you select and open up one file at a time
10. If the sidebar file explorer is open, and you switch out of back to DWMX (with alt-tab or taskbar), it freezes for about 5 seconds to render the screen, apparently from the tree structure. If you close the file explorer, it's instant.
11. Built in reference lists CF tags, but not CF functions. huh?
12. To research CF functions, now have to use Winhelp.
13. Loading the list of keyboard shortcuts locks up DWMX for nearly a minute
14. The "Answers" panel has an "update" button that you apparently can't configure a proxy for, so you can't do the updates.
15. They removed "Extended Find and Replace" but put in a sort of enhanced "Find". But now missing ability to search all open files!!
16. "Find and Replace" also no longer lets you make changes just to the currently selected text.
17. When you close an open document in DWMX, it doesn't revert to the next prior doc that's open.. it jumps to another arbitrary one?

LOVE Mozilla 1.3 on Redhat 8 but...

[Anonymous Coward]

... Java just does not work. Moz 1.1 Java worked fine, but now it says I need to download the plugin. I already have the plug-in. It's installed and symbolically linked to the mozilla 1.3 folder. What do I do? Anyone else having trouble upgrading to Moz 1.3 on Redhat 8?

BTW, I think the anti-aliased fonts in 1.3 are kick ass. If I could run Java, it would be my browser 100% of the time...

Re:dreamweaver [SIC]

[b17bmbr]

there are some gui f*** ups in MX versus 4, which i had prevously used mostly on the mac. though i am unfamiliar with CF, i actually do most of my editing(php, perl, mysql) in vim and dwmx will let you open files in an external editor.

i think the point is that dw has lots of tools built in that save the developer time. like any IDE, there are some great things, some crappy things. the biggest problem isthat it is not a cocoa app, but carbon, as far as OS X is concerned. i understand that you can dock all the windows in, err, windows. dw probably generates the best standardized html, far better than FP or go live. excessive, yes, but i have rarely had dw pages fail to render correctly.

i usually will actully do most my work, especially in data driven sites from quanta gold or just vim. but, when i bring my ibook into a shop and set up a site locally, dw is far more impressive. so, if spending a few hundred gets me more work, oh well.