Profile

Cover photo
Edward Hervey
15,387 views
AboutPosts

Stream

Edward Hervey

Shared publicly  - 
 
 
Doesn't anyone else in the open source community feel concerned that a lot of the discussions we have on G+ are owned by Google and can't (in as far as I know) be archived/retrieved by 3rd parties. I'm beginning to feel very uneasy about some of the info in here.
1
Marc Ordinas i Llopis's profile photo
 
Yep… We seem to have accepted G+ in good faith given Google's past record on data availability (Google Takeout) and use of open standards (gtalk based on XMPP). But now Google seems to have fallen into the 'to beat Facebook we have to be like Facebook' mindset (which I think is completely wrong), closing our data, eschewing open APIs and standards, etc.

The big question is: Is it possible to have a viable social communication tool, more advanced than email lists, based on open standards and open data? So far the answer is no :(
Add a comment...

Edward Hervey

Shared publicly  - 
 
True .... sadly
 
Truth
12
4
Kenneth Rohde Christiansen's profile photoAlexandre Prokoudine's profile photoGlenn Pierce's profile photoLuca Barbato's profile photo
2 comments
 
Not quite.
Add a comment...
2
Pēteris Krišjānis's profile photo
 
Just awesome human being, my respect for him just went sky high. Maybe there's reason no matter what he's playing, we can relate to him.
Add a comment...

Edward Hervey

Shared publicly  - 
 
 
En quelques heures, 9.000 citoyens ont signé l'appel au président de la République pour une grande loi de moralisation de la vie publique, sur http://www.moralisation.fr. Soutenez et relayez cet appel autour de vous. Ensemble, obtenons la moralisation !
 ·  Translate
1
Edward Hervey's profile photoJean Chalard's profile photo
3 comments
 
Je n'ai pas dit que je n'aimais pas la proposition.

Mais je n'approuve pas ses façons de faire. Que sa proposition de proportionnelle existe depuis longtemps ou pas ne change rien. La vérité, c'est qu'il saisit la chance du scandale actuel pour faire revenir sa proposition de moralisation sur le devant de la scène, qu'il sait que les gens vont signer à cause du scandale récent. Jusqu'ici, je n'ai rien à redire, je trouve que c'est à la fois une bonne chose, une bonne proposition, et que c'est bien joué de sa part. Il a raison de saisir l'occasion, et il le fait de la bonne façon.

Mais il en profite pour faire passer au plein milieu une autre proposition à lui qui n'a rien à voir avec le sujet qui nous occupe. Et ça je suis désolé, c'est malhonnête.
 ·  Translate
Add a comment...

Edward Hervey

Shared publicly  - 
 
I couldn't agree more with that analysis. Don't be afraid to say "I don't understand".
Took me a few years to actually do that, but I have no problem whatsoever doing it nowadays.
2
Add a comment...

Edward Hervey

Shared publicly  - 
 
 
Busy day for Wayland today.

+Matthias Clasen created the feature page for the Fedora + GNOME Wayland transition: https://fedoraproject.org/wiki/Changes/Wayland.  This is part of the process of getting features into Fedora, so it's exciting to see the page go up.  I'll be at GUADEC in Brno this week and we'll have a panel discussion about Wayland and much more.

+Rob Clark reports that Weston now runs on Freedeno, the open source driver for Adreno GPUs: http://freedreno.github.io/.  G+ posting should be just around the corner.

Alex Davy has been getting hybrid GPUs support working, Experimental Prime support on Wayland, that is, a client running on an AMD HD7730m discrete GPU, rendering into a window, then composited by Weston running on the Intel GPU.  Thanks to the DRM PRIME infrastructure by +Dave Airlie this was surprisingly easy.  Still experimental work, many open issues, but very cool.

Last, but not least, +Carlos Garnacho just announced a new experimental, low-level toolkit, Mechane (http://blogs.gnome.org/carlosg/2013/07/29/introducing-mechane-guadec/) - written for Wayland initially.
1
Add a comment...

Edward Hervey

Shared publicly  - 
2
1
David Zeuthen's profile photoMarc Ordinas i Llopis's profile photo
 
Incredible! And they just kept on defending their "methods" on the comments :-P
Add a comment...

Edward Hervey

Shared publicly  - 
 
 
This is definitely one of the coolest projects I've ever stumbled upon. Congrats +Andrew Kelley. Great stuff!!
3
1
Måns Rullgård's profile photoMarcelo Hashimoto's profile photoBastien Nocera's profile photo
2 comments
 
Yeah, but a NES on my computer is more interesting than running a program (I could already run on my machine) on a machine I'd never own.
Add a comment...

Edward Hervey

Shared publicly  - 
 
 
The BERG x Google Creative Labs Connbox ... with a software stack from everyone's favourite open source consultancy :)
1
1
Jon Steer's profile photo
Add a comment...

Edward Hervey

Shared publicly  - 
 
I finally got a splitter and 10m extension coax cord for DVB-C in the office so I can finally do DVB testing/hacking on a more regular basis (as opposed to going in the living room with a laptop).
A few items I'll be reviewing and working on next (some are reviewing pending patches): Port to Linux DVB API v5 (for cleaner code and support of more frontends live DVB-S2), faster tuning support (i.e. kill most of the usleep in the code and replace with poll), move more code to a helper thread (for stats and status querying for example), experiment with injecting PMT info from source (to speed up setup of downstream mpeg-ts parse and demuxer).

Once all that is done, I can finally start working on the "reverse QoS" system. Right now elements that estimate remote clocks (needed for RTP and DVB) modify the outgoing buffer timestamps (so they synchronize locally at the same speed as they were produced remotely) . While that is fine for playback use-cases, it completely screws up scenarios where you want to do transcoding or the live stream or any other transformation (you end up with bogus timestamp deltas in the transformed streams).
The idea with reverse QoS is that instead of modifying the timestamps you send downstream at regular interval the information needed to correct synchronization for those streams (running-time delta to apply and rate modifier (> 1.0 being synchronize faster, < 1.0 slower). Synchronization algorithms (and clock users) can then use that information to correct the synchronization times.
2
Robert Swain's profile photoZaheer Merali's profile photoEdward Hervey's profile photoLuca Barbato's profile photo
6 comments
 
Good stuff Edward, about time they get attention! 
Add a comment...

Edward Hervey

Shared publicly  - 
 
 
Mission, meritocracy, ecosystem, interdependence, platform, concentrated investment, coordinated changes, energy, stakeholders, divisive leadership, exciting and energetic, cloud, tablets, enterprise computing, cadence, agile development, innovate, platform, vision... BINGO!!!
1
Add a comment...

Edward Hervey

Shared publicly  - 
 
 
Mission, meritocracy, ecosystem, interdependence, platform, concentrated investment, coordinated changes, energy, stakeholders, divisive leadership, exciting and energetic, cloud, tablets, enterprise computing, cadence, agile development, innovate, platform, vision... BINGO!!!
1
Add a comment...
Basic Information
Gender
Male
Links
Other profiles
Contributor to