|Was Firefox 3 truly ready for release?|
How is the decision made?
| 1:21 pm on Jun 19, 2008 (gmt 0)|
Based on my review of bugzilla reports that were still unconfirmed and/or still unaddressed, I am puzzled by what criteria the creators of FireFox used to decide that it was ready for a final release.
If they are counting on the auto update feature and taking a "fix it later" approach, that's a terrible policy and a big mistake.
FF's share of market is sufficently large now for it to make life very, very difficult for developers if a new version is launched prematurely without thorough testing.
Does anybody have any idea about this process?
| 1:35 am on Jun 20, 2008 (gmt 0)|
The best place to find out is at the developers site.
But I never heard of a 'perfect' product launch :)
| 1:38 am on Jun 20, 2008 (gmt 0)|
Not sure about the product itself - but their servers certainly weren't ready!
| 1:40 am on Jun 20, 2008 (gmt 0)|
Did they do their world record thing? Or did the servers let them down?
| 3:36 am on Jun 20, 2008 (gmt 0)|
Over 8 million downloads (pre-audit figure is 8,290,545) in twenty-four hours. They got the servers and network fully-operational one hour and 16 minutes after the announced starting time.
Firefox/3.0 gained a market share of about 4% within 24 hours -- pretty impressive since the auto-updater hasn't been told to update older copies yet, so this figure represents people who actually went to the site and downloaded the new version.
There's no such thing as a perfect product at launch. When I was a computer designer, a sign in our drafting department said it all: "There comes a time in the life of every project when you must shoot the engineer and go into production." Even as one of the engineers the sign threatened, I found it quite funny and agreed (in principle, at least). Seriously though, Mozilla classifies bugs as "blocking" and "non-blocking." Serious problems are classified as blocking, and will hold up a release until fixed.
| 8:29 am on Jun 20, 2008 (gmt 0)|
|There comes a time in the life of every project when you must shoot the engineer and go into production. |
Reminds me of a writer named Ed Hoch, who died recently; he tried to write novels, but by the time he got to the end, he'd had so many new ideas for plot and character, he had to rewrite. In the end, he stuck to short stories, and produced several a month consistently for decades.
If he'd persevered with the novels, he'd never have got into print!
[edited by: Quadrille at 8:30 am (utc) on June 20, 2008]
| 9:45 am on Jun 20, 2008 (gmt 0)|
People expecting perfection (or six-9s-reliability or whatever) in any consumer-priced (or free!) product or service are going to be disappointed, else it's gonna cost MUCH more than they want to pay.
| 3:07 pm on Jun 20, 2008 (gmt 0)|
Disagree with you on the "quality means extra cost" issue but that's for another thread on another site on another day.
And so, except for the idea of "blocking" bugs that will hold up a release, nobody posting here so far has the foggiest, I take it.
You see, to paraphrase comedian Billy Crystal's line from the movie The Princess Bride: "there's dead and then there's really dead".
Meaning that there's "ready to release" and "really ready to release" meaning that it's a very subjective call and there is a lot of trust being placed in those who make that decision.
One thing that bothers me is that Mozilla is not beholden to paying customers. Corporate sponsors, yes - the rent money has to come from somewhere - but if they screw up, what happens? Nothing, I would imagine.
My concern goes beyond "bugs".
For example, I think they made a terrible mistake in implementing FF3's new Zoom feature as "sticky". In other words, if for some reason the user zooms up (or down) on a page, the next time the browser opens that page, it zooms back to that level automatically with no indication to the user whatsoever. (And no way to detect it, via script, as far as I can see and I've checked. Not so in Opera, for example.)
Newsflash: A lot of users hit Ctrl+ or Ctrl- purely by mistake. Plus there are keyboards marketed that actually make it easier for users to zoom up or down by mistake. Further, beyond stickiness, once you zoom a page on a site, the browser, unasked, applies that same zoom level for the other pages on the domain. In other words, it's suddenly like a cookie has been set to zoom every page on the domain.
Who the hell asked for this? (And it's got nothing to do with accessibility, don't get me started on that.)
This adds a huge level of uncertainty which leaves authors completely in the dark as to how their content is actually being viewed.
It's a radical departure with real repercussions. Was it adequately tested?
I haven't the foggiest. Good luck to us all.
| 3:49 pm on Jun 20, 2008 (gmt 0)|
That may be a design error - and all changes bring out the reactionary in someone, somewhere. But not a bug.
You don't seriously believe that release should have been held on that basis, surely?
New features do take some getting used to - but if the advantages are big enough, people will do it.
I wouldn't part-ex my new FF for the old one ... and I certainly wouldn't part-ex FF3 for IE anything-on-this-planet.
I suspect your post should have been headed "Features I don't like about the new FF" - in which case I'd have left you alone!
| 6:24 pm on Jun 20, 2008 (gmt 0)|
All depends on how you define bug. (I'd be interested in your definition, now that you're not leaving me alone.) ;)
Actually I do believe so, yeah. Why not? Why not wait for the results of a usability study before going off half-cocked? Whatever happened to empirical evidence as opposed to "best guess".
It's about respect for those who create content and the people who want, mostly, to view it the way its author conceived it.
Should there be options? Surely. But it's only polite to ASK.
And there actually seems to have been some debate within the developer team about the zoom feature because the initial betas did away with the Text Only Zoom and then it was put back in.
Is a feature that's poorly implemented "buggy" in your book or are we operating under some other definition.
Frankly, I'm tired of being a test subject aren't you?
And hey, my main point was only - what user testing was done? That's all. And what criteria was used to determine whether the product was good to go, that's all.
And hey, by the way, overall I'm very impressed with FF3 so far.
| 9:12 am on Jun 21, 2008 (gmt 0)|
1. FF is free
2. There's no law says you have to download it. It's a choice.
3. They have a thriving community; you could join and raise objections to launching before you think they are ready.
4. I'm no techie, and have no definition of bug - but I've always thought it was something that failed to work properly and needs fixing, rather than something that didn't do quite what I would like it to; with new software, there is no 'one right answer' - and from what you've said, there's nothing that needs 'fixing' - unless it turns out that your preference is a general preference.
But the key is #3. FF is open source. It is only ever as good as the community that created it. That community is obviously happy, and are sharing the fruits of their labors for free. Seriously, if you feel strongly, the very least you can do is tell them. Feedback is all, in open source.
I really don't feel like a test subject. With all new software, there are things I like, and things I don't. Sometimes, it's enough that I change suppliers. It's always a risk - but if it's free, the risk is mostly time.
And however much testing they do, there's ALWAYS surprises - ask Bill Gates about that ;)
| 12:05 am on Jun 22, 2008 (gmt 0)|
All points well taken, Q.
I have gone thoroughly through the Mozilla site and found no concrete answer to my question. Some general principles, blah, blah, blah, but I will, however, press further.
I will also be posing the question to the IE8 team, as well. I'd like to know what their criteria are - if I can get an answer. (They will, at least, say whether they are allowed to answer such a question, or not.)
The only comment re your post that I'd like to make is that the word "Free" can be very misleading. There is far more at stake than someone's time. "Switching Costs" can be very steep indeed. And how would you feel if a popular browser got released that in some way screwed up your site when viewed within it, without warning?
Just because something is offered to you at no charge there is still an obligation on the part of the manufacturer that the thing is fit for the purpose and won't cause you any damage. Speaking as someone who is about to market a software product myself, with some licenses offered at no charge, it can be argued that "free" is an enticement that carries with it even more obligation than something sold for a price.
FF, as just an example, is still released under a license and on install the would-be user still has to click-off on a Disclaimer Of Warranty, Limitation Of Liability, etc...
As far as the "it's a group effort" part, I'm a rather jaded fellow. I notice that there's still a paid full-time staff at Mozilla and no doubt in my mind that the big decisions are made, as they always are in human organizations (non-profit or not), by a select inner circle. It doesn't seem to be a democracy. There is no voting that I'm aware of. And I'm sure the money donated by Google and Sun Microsystems never enters anybody's minds, ever.
Anyway FF3 looks pretty good to me, like I said. The only problem I'm having is: where the heck did DOM Inspector go? With FF2 you got prompted to install it or not along with FF, but not with FF3. Any idea what's up with that?
| 12:35 am on Jun 22, 2008 (gmt 0)|
"FF, as just an example, is still released under a license and on install the would-be user still has to click-off on a Disclaimer Of Warranty, Limitation Of Liability, etc... "
Exactly; and no-one reads such things, do they! (Well, I very rarely do!).
I think that their duties are no different to anyone else's - but I also think the user should be more careful. Common sense says to me that if I got something for free, and it went disastrously wrong, I'd have little recourse to compensation ... whereas if I paid hundreds to a multinational corporation, I'd expect to have some chance of compensation.
Caveat emptor - "Let The Buyer Beware" applies as much, if not more, to free stuff.
Ultimately, it's about trust ... and I suspect most of us are quicker to trust multinatinals than freebie firms.
personally, I've a lot of trust in open source ... but I expect quirks to come as standard, and I don't expect the 'finish' to necessarily match a multinational's.
But that's just me, and we all choose our standards, I suppose.
BTW, I'm not a lawyer, etc., and I certainly don't expect the small print to match common sense :)
[edited by: Quadrille at 12:44 am (utc) on June 22, 2008]