[elinks-users] [ANNOUNCE] ELinks-0.9.0 (Skyrider)

[Petr Baudis]

  Hello,

  you are probably quite surprised. You ought to ;-). Well, I'm trying
to push a suprise for each Christmas release, and I hope I managed to do
a good one this time. Hold your breath, first some blurb for those who
did not follow the Skyrider development and want to know what does it
bring them compared to the Iceberg (0.4).

  The year has been long and brought a lot of very interesting things. I
will focus only on the most important things, but there is of course a
huge number of smaller changes, the codebase went through big changes
etc. So, here is a 0.9.0, pleased to meet you: tabs support, background
document colors (it looks really cool ;), local CGI support, 256-color
terminals support, builting user prefixes support (URLs like 'gg:google
query' or 'bug' for ELinks Bugzilla), plethora of new managers (from
cookie manager to cache manager), localization now through gettext etc.
You do not get ELusive, as was originally promised. That is post-1.0
stuff (basically 2.0 stuff), you can read further about that below.

  When I reflected the life of ELinks so far, we'd just met 0.5, really
long before 1.0, and our schedule would be really planned only for 0.6
or 0.7 and maybe something for 0.8, but really what would qualify for
1.0? It could very well be that the whole ELinks life was spent in the
0.x series, and that would be a pity.

  On a different note, ELinks is a text browser. It is certainly not
intended to stay so forever, and we have those grand plans for the
future, but in the heart it is still a text browser for now. And our
mission for now should be to make it THE killer text web browser (that
means, can do almost everything important all the other text browsers
can, and something more). And that's the goal of 1.0.0.

  My main intended feature goal for 1.0 is cursor routing, but there is
a number of smaller glitches still unresolved. Bugs in the scrollable
widget to start, through SMB and weird encoding handling, to sane cache
and frames handling (ports from Links2?). The main code change should be
the event system massive cleanup and sanitization, variously
interdependent with the cursor routing etc. Jonas will be the main
authority over this; although I will be the Reverter, I hope I won't
need to use it anymore and Jonas will be reasonable (he seems to do
better and better).

  In fact, originally we (me and Jonas) planned this to be 1.0.0, but
because of the schedule changes, this didn't work out. And it's probably
better, 1.0.0's release date is not set (but I hope for it being
sometimes in January or February), so it will be able to get as much
testing as necessary, and we'll be able to ensure that it is really
worth its "stable" sticklet.

  So we chose 0.9.0 and now will make our way through the planned stuff.
And this time, I want to _focus_ on the goals, actually. That doesn't
mean anything else isn't going to happen, but I'll ponder twice, and try
not to let people do other stuff. We need to first get ourselves to a
good shape with certain good set of stuff without major glitches before
we will be able to well advance further.

  That means no ELusive, guys. No graphics, no javascript, no this or
that (no Big Modularization neither). Also no Unicode --- the others
have it, but it would be too dangerous to introduce now and could
introduce too many bugs. That means people interested in that are
prohibited to do anything. In fact, they could likely even get their own
CVS branch where they can cook their stuff. However, it is just not
going to happen in 1.0.0.

  So, well, 0.9 line will be atypical - there will be no stable/unstable
distinction, everything will happen on a single branch (if you were
attentive, you must have guessed this already; the separate branch would
be here already). This should help to ensure that we won't drift off so
much from the reality, Jonas and Zas won't start up some huge cleanups
which will break stuff for months or so ;). I hope the 0.9 line to
advance fast to 1.0.0.

  After 1.0, things are more blurred now, but I think that we will adopt
the popular concept of odd/even branches meaning unstable or stable. I'm
already really fed up with this pathetic ridiculous -pre scheme, so
*that* one is not going to return. I hope us to go fast, and I also
think we shouldn't fear too much the majo versions. I see ELusive a
change major enough, as well as graphics support and modularization
(javascript wouldn't be so revolutional in such a context, so I'm not
sure if that one would deserve new major number). But this is just
plans, we will see yet.

  Don't worry, be happy. I think we have a bright future ahead.


  Pah, what a pityful text I just produced. I knew I should've prepared
this in advance, but I was so short on time... :-(( Please forgive me
for my poor style late in the night after the tiresome day. And I hope
that the reality won't reflect style of this... this... thing.

  The ELinks itself can be found at http://elinks.or.cz/.

-- ELinks commit statistics for the 0.5pre branch:
Total commits: 6051
pasky   Petr Baudis                    2082 commits  34.4%
jonas   Jonas Fonseca                  2065 commits  34.1%
zas     Laurent Monin                  1459 commits  24.1%
miciah  Miciah Dashiel Butler Masters   171 commits   2.8%
kuser   Karsten Schölzel                 97 commits   1.6%
fabio   Fabio Bonelli                    77 commits   1.2%
witekfl Witold Filipczyk                 66 commits   1.0%
vinnui  Varga Balázs                     23 commits   0.3%
samek   Peter Samek                      11 commits   0.1%

  Meet us at #elinks @ FreeNode (http://freenode.net/)!

  Merry christmas,

-- 
 
				Petr "Pasky" Baudis
.
He who re-invented the wheel, understands much better how a wheel works.
.
Cool crap: http://pasky.ji.cz/