On Mon, 31 Oct 2011 23:12:56 -0400 Youness Alaoui
<kakar...@kakaroto.homelinux.net> said:

> On Mon, Oct 31, 2011 at 9:38 PM, Carsten Haitzler <ras...@rasterman.com>wrote:
> 
> > On Mon, 31 Oct 2011 17:19:17 -0400 Youness Alaoui
> > <kakar...@kakaroto.homelinux.net> said:

> Ok.. so I don't understand, you're saying you'll get devs when it shipped
> as the default WM for a major distro, but you say it won't happen (kind of
> expected), so you're basically saying that the project is already doomed?

no - there's another plan... why do u think there's so much focus on efl for
"embedded"? :)

> Nope, not a rumor I'm making up, but a rumor that I heard. I don't know
> (never bothered to ask) about any details to be honest, but the rumor was
> basically that indeed there was bad blood, and that you got fired. Never

yes - fired because i had accepted a competing job offer (pending legal visa
paperwork of course). h1-b visas at the time were non-transferable. you had to
re-apply for each new employer. i couldn't sensibly quit while waiting - it
could take anywhere from a few weeks to months to get a h1-b and if it took too
long i'd have to leave the country. thus i HAD to "wait it out".

> heard of the job/visa thing though, the rumor states that your vision and
> the vision of gnome/redhat/whatever was diverging.. you wanted to customize
> everything and they wanted very little customization but for it to do the

yes. i wanted customizations for everything that at all made sense to customise.

> right thing by default. Can't remember though if the rumor ended with "then

i don't see any problem with having BOTH. you have a DEFAULT config that's good
for most people - and those that disagree can tweak. i still totally disagree
with the model of "no options, no tweaking, just 1 way of doing things for
everyone". it's good that we diverged.

> you got fired" or "then you left because of this disagreement".

which matches pretty much what i said - but i gave the details now. i was
brought into a room with eric troan (then head of engineering) and asked if i
had another job accepted/lined up (don't remember the exact words - this is a
long time ago - but conversation to that effect). i said yes and was fired on
the spot. yes - bad blood. i wasn't employed by someone else yet. of course
that'd wait until i had done my notice period, but redhat decided it'd be good
to fire me on the spot. it actually was a good thing. i got out of north
carolina to silicon valley and man was silicon valley nicer. a real city (san
francisco), and absolutely awesome offices at VA Linux. fantastic equipment,
network and so much more.

some history - VA linux talked to me while at linuxworld in san jose. the
simple version was "do you want to come work here in silicon valley on e and
make an awesome desktop! we actually make hardware and desktops so this will
ship". i had the chance to ensure the hardware was good and would run e well.
somewhere along the way opengl was becoming a thing - it was bad grantham that
pointed me to trying to use opengl for 2d. and it mostly made sense. all the
advanced features were now going into 3d, not 2d accel. all the things i really
wanted. i started toying with gl for 2d - frankly a lot of stagnation on e is
due to this whole direction. i was writing a filemanager for e16 - efm... used
imlib2 and lots of alpha channels everywhere. gl made so much sense. but at
valinux we had precision insight as aprt of the company - the guys working on
drm/dri originally. another guy worked on the nvida 2d drivers. he now works at
nvidia itself. it all made so much sense and this was heading towards having a
desktop with specifically chosen 3d hardware with good support that we could
accelerate a fancy ui experience via opengl... at the time though opengl was an
iffy prospect. so many things it just couldn't do right to get "pixel-correct
2d". i tried playing with all sorts of games to get it to work. also back then
max texture sizes were often 256x256 which didn't help - we had to tile stuff in
the original evas code for opengl. but i couldn't commit to opengl totally. it
wasn't quite there yet hardware and api-wise. evas HAD to have its software
back end too. and to this day it still does and its maintained and optimized.
anyway in 2001 or so the dot-com bubble burst. all the companies riding high on
inflated stock prices - like va linux, had a hard time, and then no one was
buying servers from va because of the dot com bust. that was their main income
stream. va basically crumbled. sure - given a few years and maturity there'd be
a solid smooth linux desktop with gpu acceleration, but it didn't happen. the
market crash killed that. e development also took a tumble along with this. but
i realized the desktop was way too hard a "win". and just about then the first
ports of linux to the 206mhz strong-arm (64m ram. 16m flash, qvga) ipaq turned
up... and i went "oooh... so this really CAN work on things like pda's... hell
THIS is where the future is". ever since that has been a primary motivator for
me. anyway back to back then...

yes - enlightenment and gnome were diverging. i was hoping we could work
together but for 10 months or so gnome didn't want to have squat to do with a
wm and by then i'd made my mind up to just go my way anyway. if that's
stubborn, i guess you can call gnome stubborn too, and kde, and redhat, and
pretty much every project, company and person on the planet. it was a
difference of direction about something i was pretty familiar with by then as i
was writing wm's.

remember back in 1997 and 1998 or so... gnome was now mimicking everything
windows did slavishly "omg! windows has dcom/com - we must have that too" for
example - thus gnome decided it'd do corba. corba required this ORB called
mico. it was a c++ monstrosity. it took HOURS to compile. most people couldn't
compile it - they literally didn't have enough ram. they'd end up swapping
forever without enough ram. redhat started doing daily builds as we were one of
the few people with good enough equipment (p2-300's with 64m or 128m ram from
memory) to actually compile it. corba was the "ipc" layer. that was part of
gnome's "direction". i have to say that frederico mentera a stellar job making
an orb (orbit) finally that was not bloatware from hell. but that was well into
the project and things like mico had me wondering about gnome in general. you
know gnome was meant to be a desktop written entirely in lisp? early on? sure -
reality turned out that wasn't a happening thing... - anyway. there's a lot of
history there

even better was that redhat forgot to remove me from internal mailing lists for
a week or so after i left - so i was still getting internal mails. the team
leader for red hat labs (mike fullbright) did a nice bit of back-stabbing which
i nicely got all the content of. if you're going to say nasty shit about me...
have the balls to say it to my face and/or in public. that of course sealed the
deal with bad blood. sure!!!! i was going to help my former employer who fired
me (though i was leaving anyway, and would have kept good relations) and was
stabbing me in my back with their pet project (gnome). i never removed support
for gnome, i just stopped making that the default out of the box configuration
and setup. gnome would have to do WORK to configure e the way they wanted. they
decided they didn't want to. also note that back around 97/98/99/2k/01 mandrake
was the other major e author - so it was not just me here. we both agreed on
this. we cut loose and just went off to do e16 and then i did imlib2, efm (you
never saw the early efm back in 2000/2001 - that efm gave birth to evas). it is
thanks to this split that efl is anything like what it is today.

> Thanks for the history lesson. (and no, I'm not being sarcastic, I don't
> know much about the history of gnome/WM/linux/etc..)

cool.

-- 
------------- Codito, ergo sum - "I code, therefore I am" --------------
The Rasterman (Carsten Haitzler)    ras...@rasterman.com


------------------------------------------------------------------------------
RSA&reg; Conference 2012
Save &#36;700 by Nov 18
Register now
http://p.sf.net/sfu/rsa-sfdev2dev1
_______________________________________________
enlightenment-devel mailing list
enlightenment-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/enlightenment-devel

Reply via email to