August 20, 2004
Firefox UI and unrealistic expectations
It cracks me up that people think that they can affect Firefox UI decisions
logical (or, even sillier, emotional) arguments. The whole premise of the
project, as I understand it, is that there is a UI czar (or small group of
like-minded UI grand dukes) who makes decisions which are final.
I've finally been able to pin down what it is that really bugs me
about Firefox, by the way. It's not having a say in the behavior of a product
in which I'm investing a lot of time and effort. Feeling that you're
irrelevant (and easily replaceable) slave labor is no fun. And in the end, it
has to come down to a decision on whether you feel that having your name
associated with the product in question is a good thing. Right now, having my
name associated with Firefox is not—I end up with responsibility (friends
complaining to me about this, that, and the other) without control (I can't
affect any of those things).
Posted by bzbarsky at August 20, 2004 12:04 PM
Now that Asa has WONTFIXed the bug, I'm curious if you have an opinion on either:
A) Sometimes arguing works.
B) The Firefox UI leadership decision process broke down in this case.
I suppose C) A and B should be there too. Oh even D) None of the above.
Remember the title bar change where the product name was removed. That was reversed "through logical (or, even sillier, emotional) arguments."
Bill, I won't wager on what made Asa wontfix the bug; I'm not privy to his thoughts.
Brant, I do not in fact remember it. :)
It's quite likely that there are times when a decision is made without considering all the implications and pointing out those implications will affect the decision. The view-source thing doesn't seem to be such a case, however. I would give much to find out what caused that decision to be reversed.
The title bar change bug was bug 215296. It got up to comment 54 before Dave Hyatt reversed himself. I doubt the change had anything to do with the whining on the bug, tho (excepting perhaps the comment at the URL below, if the author wasn't penning a slightly suspicious-sounding false scenario).
All of these 'so-called' decisions from the top, with no real thought about the ramifications...
Made me FORGET about anything BUT Mozilla. That's all I'll promote....well, maybe Netscape 7.2 now, too!
They (Ben and the Fox) lost a customer...that they will never get back. Sorry. The customers are ALWAYS right, and if you don't treat them as such, you will loose them. I'm NOT saying that you have to do what they say, but make them FEEL like they were part of the process. To dictate, and say that "I know better..." just pisses folks off.
Sorry Boris, I feel your pain.
Boris, the separation from the decision process is typical for a corporate environment. Just go to any company, that has like mozilla a dozens of engineers and ask the engineers how much say they have about theirs management decisions. The viewsource is also a example for the overcoming of the dogfood approach. In the old times the question was, can I do *my* job with the product. This was a important criteria, see the old dogfood in bugzilla. Now the question is more customer oriented, do the users use view source. I clearly could not do my daily job without viewsource. The mozilla management probably thinks the majority of the users can. As long as the mozilla management does not kill Seamonkey, this is just fine with me.
You're not easily replaceable.
The way to look at it is that the product you're working on is Gecko, which is used in way more products than just Firefox. Sure, Firefox is one of those products -- but so it Netscape, and you don't have any word on how Netscape looks either. Not to mention Galeon, Epiphany, etc.
Ben doesn't expect to be able to have much of a say in how Layout does things, any more than you have a say in how the UI team does things. Does that mean Ben doesn't control the product either?
So if your friends think that you are in charge of Firefox, you should explain that you are actually working on a product _used_ by Firefox, but not working on Firefox itself.
I agree with Ian.
If gecko/mozilla/firefox/thunderbird/etc. were all one product, then yeah, you have a point. But they are not.
You have the incredible opertunity to say you've been a key part in all the wonderful Mozilla/gecko based products. That is damn impresive.
Yes, the UI sucks in many of those products mentioned, yes they are improving. But you're not an UI expert and that's not your calling.
The blame goes on others, but in the end, the products still rock and you've been a HUGE part in that process.
Your awesome Boris, don't let some stupid UI bugs get you down. They will play out there way, and if the decisions are wrong, it will be a learning experience for those decision makers.
Look at it this way: they're working on the dashboard of the car, while you are tuning the engine. Which is more important?
The stuff that comes out of mozilla.org would be nothing without the layout code and tens of other backend subsystems, and people like you are unsung heroes in the same way that soccer defenders never get the same praise as star forwards.
But although normal users don't know, *you* know, and the people who count know too. That should be satisfaction in itself... :)
Prepare for another one: stylesheet switcher is going to be gone in Fx 1.0.
"In the old times the question was, can I do *my* job with the product"
There are many very well designed products that still fulfill that criteria. See Jira or IDEA.
Ok these examples are not targeted to main stream.
I believe that to be a proof of success. If you can't use the tool you build, then you should maybe question its utility.
I'm sorry that you're feeling disheartened about the great work you do for the mozilla project. Your work on fixing bugs and the intelligent comments made in bugzilla are outstanding. I never really liked the way some of the mozilla guys just decided to go off and create phoenix etc., it seemed to take some of the community spirit away from the project. I'm very pleased that seamonkey will continue and I hope you will continue to work on the project. I guess there are many people who appreciate your work and the time that you give. You have certainly helped to make Seamonkey a great product. Thanks
bz, there's no mystery to Asa's WONTFIX'ing the bug: mscott, dveditz, hyatt, p_ch, Ben, and I all jumped on the bad decision as soon as we learned of it.
See my blog item posted just now for more.
I think the disconnect between all the developers who make the pieces of a browser and the few who control how end-users percieve the technology is problematic. It is probably also inevitable.
Part of it is a confidence issue. The frontmen of Fx need to acknowledge that they wield significant control over the project and they need to give back-end devs some level of confidence that they are making reasonable decisions. Often, this confidence happens by trying to coopt the backend devs into some of the larger decisions before they annouce them. Even this minor flare-up would have been diffused if someone would have bounced the thought off a handful of devs individually first (and they would have given the three or four good reasons not to do it without having to make an emotional debate out of it).
As I see it now, they talk from a purist standpoint of what is best for users. This is all moot if they start driving contributors away from the project (not to mention the "best for users" arguments is sometimes not backed up with anything better than "I know Joe Hayseed").
UI czars taking oukazes based on what knowledge of the mass market, what knowledge of the corporate market, inside and outside of US, what knowledge of users' practices (laptop vs. desktop, modem vs. dsl/cable, corporate vs. home) depending on geography, what knowledge of preferences depending on user's gender, what users' expectations poll, ... ?
Come on, they do exactly like we all do, like I do with Nvu: we try to come up with the best UI, based on our own experience and hopefully cleverness. We make decisions alone, because we don't have the means a BIG corporation can afford. And sometimes we fail. There's nothing more. But nothing less. We're all only humans...
Daniel: I agree about ukases (haven't used that word in a while -- fun, Russian/French! Do you spell it differently?) based on some divine, or at least unsupported, knowledge of end user needs are bogus. Firefox has been designed not by such edicts, but by simplifying Mozilla's browser code, studying IE, Safari, and other browsers, and thinking innovatively about the problems browsers try to solve.
BTW, Firefox has had a bit of formal usability testing lately, but we haven't seen results yet. And when we do, it will be tricky to reason from those results to claims about "end users".
The bigger point, which I make in my blog, is that Firefox should not cater only to one (wrong, over time) crabbed notion of "end user". As Boris and others say here, it must please core developers, even if it requires an extension (domi) here or there.
Boris, hope to see you on IRC soon.
All-in-one response to comments coming up (I was away for a week, and come back to all these neat-o comments!):
Bernd: agreed that that's the general aura I'm feeling. And I don't like it one bit. There's a reason I'm not in the software business to start it.
Ian: Some people absolutely want a say in how layout does things. See the debate in the "all localization should happen in toolkit/" bug, for example (no bug number offhand, sorry). I'd feel a lot better about "working on Gecko" if Firefox advocates (and developers) didn't spent quite so much time trash-talking other Gecko-based browsers (to the point where I fully expect most of them to disappear in the near future).
Brendan: Thanks for the clarification on the view-source thing. One note, though, is that I don't much care about view-source Firefox per se. It was thinking about _why_ I don't care that made me come to the realization described in my second paragraph.
Rest: Thanks for the encouragement. :)