Slashdot Mirror
Places Feature Cut From Firefox 2
segphault writes "Apparently, the new bookmark and history system (called 'Places') scheduled for inclusion in Firefox 2 has been removed from the roadmap and disabled in the builds. An article at Ars Technica discusses some of the implications: 'Since Firefox 2 (and all alpha builds from here on out) will use the conventional bookmark system, those of you that have been using Firefox 2 alphas (the Gecko 1.8 branch) will have to export your bookmarks to HTML in order to preserve them. As a Firefox user and a software developer, I am personally very disappointed with the removal of this innovative feature.'" Update: 05/01 01:16 GMT by Z : Ars link updated.
This feature was cut from Firefox2 because it was unpolished and unacceptably buggy still. It is now on the "trunk" for inclusion in Firefox3, so it's still on the roadmap.
In fact, it remains enabled on the Trunk nightlies for Firefox3.
Only one number off. So close, and yet so far.
1 .html
Features cut from Firefox 2:
http://arstechnica.com/news.ars/post/20060430-670
Slashdot gets worse every day... Pipedot: News for nerds, without the corporate slant
The correct arstechnica link is here: http://arstechnica.com/news.ars/post/20060430-6701 .html
The alpha automatically imported all data from the old bookmarks.html into the Places system. Firefox converted from IE to Mozilla's bookmarks.html already.
The later builds of the Alpha (last week or so) all included an Export functionality to dump your Places DB into a bookmarks.html file again for this next build. You can still download those builds if you need to export your Places DB to the old Bookmarks.html format.
In case someone is looking for more information about the actual implementation of the "places" concept: http://wiki.mozilla.org/Talk:Places:Design_Overvie w
http://wiki.mozilla.org/Places:Design_Overview
I tried the Firefox 2 beta briefly and wasn't impressed. There's very little in the way of real differentiation from 1.5 and 1.5 had very little differentiation from 1.0. Prior to this improvement was obvious, now it seems like there are a few cosmetic and stability/security changes but nothing serious. If you take out Places for 2.0, what's really left? The close button'll be on the tabs, but that seems about the only user-visible improvement.
Just like Microsoft, Firefox developers have gotten stuck on the feature-creep treadmill. Instead of fixing incessant crashes and debilitating memory leaks, they add more whiz-bang features to compete with the "enemy". Instead of adding features to make their browser more robust and responsive, they add more crap to make it bigger, slower, and buggier.
Firefox is no longer about doing the right thing. It's now all about one-upping Microsoft at their own stupid game, and the users are suffering for it. Open Source developers, apparently, are no more ammune to this competition attitude than the proprietary vendors. There is no longer anything special about Firefox. What's more, they suffer from the syndrome many open source projects suffer from, which is that they prefer to work on the "interesting" bits, rather than spending time adding some polish to make things work WELL.
The other day I decided to try Epiphany instead of Firefox. It is much "snappier" than Firefox, has a smaller memory footprint and has a smarter topic-oriented bookmark system. Those who are disappointed about this functionality being removed from Firefox should seriously consider Epiphany.
Disappointing, yes, but this is what makes excellence in software. They recognized the problems, realized the time it would take to fix, and decided on a "better safe than sorry" approach. This will make the eventual release of "places" that much better!
I really like the way Epiphany handles bookmarks with it's "Smart Bookmarks" features... Basically, you just tag your bookmarks with arbitrary tags e.g. "Work", "Sport", "Geek", and you can search for them dynamically.
I would like to see an extension of this (and I know work is in progress)... With meta-tagged files. God knows why browsers do not store bookmarks as files in a "Bookmarks" folder.
Because forking it won't get code written any faster? It's not as if forking magically gets stuff done...
Their official reason for disabling Places amounts to "either we kill this, or no new Firefox for everyone". They chose to release something with the other changes rather than wait.
Basically, in Netscape 4, bookmarks were stored in a quasi-HTML file, and history in a DB file.
In Mozilla, bookmarks are stored in a XML-that-almost-look-like-HTML format, while the history is stored in the most insane file format ever devised by mortal mind. It's called MORK. Remember that name. Remember it well. (Seriously, take a look at your history.db. It's a text file. It really is. Or it might look like one from a good distance.)
While in the new grand concept, everything is stored in a SQLite database - simple, well tested, portable, efficient, doesn't make Firefox much bigger than it already is, and above all, programmer-friendly file format that isn't causing peoples' brains to ooze out of their ears when they try to figure it out.
- Places is still in the roadmap, just not for v2.0. (maybe 3.0 if not earlier)
- Places was too buggy to work with. Nightly testers report far "too many" bugs with it... even if they were fixed, imagine all those bugs that would be uncovered if used by the masses (nightly tester build bugs are a good indication of how many bugs will be found if open , it's somewhat proportional).
More to read at MozillaZine
I guess I am a little confused what the rush is. Can't they just hold the release until they get this feature correct? It is not like they are selling a product and need the churn to make revenue.
He's not kidding, take a look at my history.dat file:
// (f=iso-8859-1)R eferrer)n t)(87=Name)
: //www.google.ca/)(4B70=1146443053431000)(4B71
=http://www.google.ca/search?hl=en&q=slashdot&btnG =Google+Search&meta=)d ot.org/)(4B77=slashdot.org)3 0/2128229.shtml)o rg)(4B7C8 4502&cid=15234094)^ 4B72)]^ 4B71)(^87^4B75)]^ 4B77)(^87^4B78)]^ 4B7B)(^87^4B7C)]^ 4B7B)(^87^4B7F)]}
// <!-- <mdb:mork:z v="1.4"/> -->
< <(a=c)>
(8A=Typed)(8B=LastPageVisited)(8C=ByteOrder)
(80=ns:history:db:row:scope:history:all)
(81=ns:history:db:table:kind:history)(82=URL)(83=
(84=LastVisitDate)(85=FirstVisitDate)(86=VisitCou
(88=Hostname)(89=Hidden)>
<(4B6E=LE)(4B6F=http
=google.ca)(4B72=G$00o$00o$00g$00l$00e$00)(4B73
(4B74=1146443064149750)(4B75
=s$00l$00a$00s$00h$00d$00o$00t$00 $00-$00 $00G$00o$00o$00g$00l$00e$00 $00S\
$00e$00a$00r$00c$00h$00)(4B76=http://slash
(4B78
=S$00l$00a$00s$00h$00d$00o$00t$00:$00 $00N$00e$00w$00s$00 $00f$00o$00r$00 \
$00n$00e$00r$00d$00s$00,$00 $00s$00t$00u$00f$00f$00 $00t$00h$00a$00t$00 $00m$00\
a$00t$00t$00e$00r$00s$00)(4B79
=http://developers.slashdot.org/developers/06/04/
(4B7A=1146443070774750)(4B7B=developers.slashdot.
=S$00l$00a$00s$00h$00d$00o$00t$00 $00|$00 $00P$00l$00a$00c$00e$00s$00 $00F\
$00e$00a$00t$00u$00r$00e$00 $00C$00u$00t$00 $00F$00r$00o$00m$00 $00F$00i$00r$00\
e$00f$00o$00x$00 $002$00)(4B7D
=http://developers.slashdot.org/comments.pl?sid=1
(4B7E=1146443122321625)(4B7F
=P$00l$00a$00c$00e$00s$00 $00F$00e$00a$00t$00u$00r$00e$00 $00C$00u$00t$00 \
$00F$00r$00o$00m$00 $00F$00i$00r$00e$00f$00o$00x$00 $002$00)>
{1:^80 {(k^81:c)(s=9)[1(^8C=LE)]}
[3ED4(^82^4B6F)(^84^4B70)(^85^4B70)(^88^4B71)(^87
[3ED5(^82^4B73)(^84^4B74)(^85^4B74)(^83^4B6F)(^88
[3ED6(^82^4B76)(^84^4B74)(^85^4B74)(^83^4B73)(^88
[3ED7(^82^4B79)(^84^4B7A)(^85^4B7A)(^83^4B76)(^88
[3ED8(^82^4B7D)(^84^4B7E)(^85^4B7E)(^83^4B79)(^88
That's one way to kill interoperability.
Because marking a post 'redundant' is a punishment. That's the system that we have. A person trying to help out the community by providing a proper link shouldn't be punished and their comment doesn't need cleaning up. We do have to assume the intentions of people and that's why this post shouldn't be marked redundant (at the time they starting typing, the other 'redundant' post didn't exist). If it's rated high when a previous post does the job, then you can rate it 'overrated'. That follows the spirit of the rules, IMO.
In times like these, it is helpful to remember that there have always been times like these. - Paul Harvey
All I want, and I'm betting so do a great deal of other people who work with the web, is a browser that follows the standards for HTML, XHTML, CSS 1 & 2 (maybe even 3), Javascript, and DOM.
Extra features are nice, yes, but the top priority should be putting out a browser that follows the standards, first and foremost.
What good are extensions and themes and fancy bookmarking tools if the core program for seeing information on the web cannot render pages which have been correctly created?
His name is Robert Paulsen...
Sorry for shouting, but I'd be happy if they did *nothing* but fix the memory leaks.
Memory leaks are unforgivable.
Ian Ameline
I have an old Thinkpad 760, but it won't run any of the new distros. I used to be able to run OpenBSD 2.something on it with acceptable speed, but XFree86 made point revision and it stopped being reasonably snappy. Running Firefox on any modern distro, BSD, Linux or otherwise, is painful.
However, I can run Win98 on it with little trouble. Is that a good thing? I don't think so.
Potato chips are a by-yourself food.
Putting close buttons in individual tabs is nothing but evil, wrong and stupid.
One mis-click on a tab (which is very common when managing a dozen or so tabs) and you've just closed an important page with no confirmation dialog.
See http://bugzilla.gnome.org/show_bug.cgi?id=335453 for the current gnome-terminal fiasco.
Just don't do it.
"Nine times out of ten, starting a fire is not the best way to solve the problem." - my wife