While Mozilla is committed to include only core features into Firefox, I feel frustated with Firefox out-of-the-box features.
To say the truth, I did not wanted to look for extensions too much time as I am a bit lazy. On the other hand, I did not searched also these ones due my point of view, as I felt/thought the expected features have to be into the core. I am not happy with the existing core/extension line.
Let me give an example. Firefox gives the possibility to create multiple tabs. Just fine. Thanks. But when I open a new tab, it is open at the most right position. Quite cumbersome if I am browsing using the most left tab and if there are 50+ open tabs. I am not happy with the core/extension line of my favorite browser ("open tab action"=> core features, "define tab position"=>extension) as these 2 features are so closely related.
OK, one could say core/extension line is a matter of opinion, or taste. So, we could debate until the end of universe about that subject...
Let's imagine 2 ideas in order to try to satisfy, not all, but more people.
PLAN_A : what about proposing multiple distributions ?
OK, let's just look about what happened into other communities.
Just look about what happened for Eclipse and NetBeans.
First, years ago, only the IDE core were packaged and extensions were proposed as... extensions.
Then, NetBeans has packaged core+some web extensions into a specific distribution (core distribution was still available for download, but NetBeans was now proposing 2 distributions: core and core+some web extensions).
People have been very happy with these more complete packagings from NetBeans. And then, Eclipse has been "forced", in order to satisfy users, to propose also multiple distributions: core and, for example, core+some web extensions.
So, for NetBeans and Eclipse, there are core distributions, and more complete distributions, and still, extensions, through multiple update centers (with even different update centers corresponding to extensions with different level of maturity).
Proposing multiple Firefox distributions, from Mozilla, could be an interesting way.
I think this idea is interesting for the following reasons.
1) out-of-the-box features matter as a user experience.
What one may think about a Christmas gift with only part of the gift inside the packaging ?
2) the point (1) is stronger than one may think as people are lazy -or- want to install faster
People are so lazy that Amazon has strongly advertised its 1-click feature for shopping faster.
Other people don't want to build again and again their own distribution (like for NetBeans or Eclipse, years ago), to deal with incompatibilities, when a new version is out.
3) Grandma is on line, so without more distributions, evolution may be too much frozen
As Rob Pike wrote in http://herpolhode.com/rob/utah2000.pdf (slide 20) : "Grandma’s on line. This means that the industry is designing systems and services for ordinary people."
So, you may design Firefox (as a core product) to satisfay 80% of people (Grandma and her friends and her happy family). And that's OK under some metrics : 80% still is > to 20% ;-)
Who matters 20% of people (versus 80%) ? Well, I am fond of the train metaphor. There are the locomotive and coaches. Early adopters+geeks are part (with other people) of the locomotive and 80% sit into the coaches.
So, first, multiple distributions might satisfy different audiences, and then, bring more audience.
3.a) Some think early adopters push, as a locomotive, the following massive adoption move. Why not building a more feature complete distribution for satisfying part of early adopters (belonging to the 20% above) ?
3.b) a more feature complete distribution could part of a testbed for proposing future features and then, enabling, for example, to call for a vote or feed-back.
4) another good signal for the community
While proposing another distribution, Mozilla would include outside Mozilla extensions in such a distribution, and then, it's like recognizing the good work of the community.
They may be different options:
- a more feature complete distribution that would enable all (selected in the distribution) add-ons by default.
- a kind of alternative could be to provide a package with multiple add-ons (choosen according to vote ? popularity ?) and during the installation, to let choose the user for installing/enabling or not the add-ons.
PLAN_B : what about changing the installation process ?
An "multiple distribution" alternative could be, after the end of the installation process, to ask the user different (short and clever) questions about his/her surfing habits and to propose him/her few cleverly selected add-ons.
For example, one question could be "how many tabs are usually used ?" and according to the answers, to propose different add-ons. Another question could be about the user profile: is he/she using Firefox only for browsing or for web development too ?
Following that idea, IMHO it would be a good idea to have the following actions:
- enable to export, after answering the question, the list ot the selected extensions,
- and to enable another user to import such list of wanted extensions during the installation process in order to let a team or a company share more easily at least a common set of extensions. After having imported such a list, Firefox would display the proposed list of options, get automatically information about these options from Internet, and the user would have to select each one in the list in order to finalize the import.
PLAN_B => one has to focus on the installation process. The burden is on the question set, OK, but Mozilla would, then, avoid the burden of selecting itself the extensions for a more feature complete distribution.
PLAN_A => this plan pushes more forward (due to the packaging) the extensions as part of an enhanced experience.
What do you think about these idea ?
"Dominique de Vito" <ddv3...@yahoo.fr> wrote in message