Follow Slashdot blog updates by subscribing to our blog RSS feed


Forgot your password?

Comment Re:Not efficient (Score 1) 355 355

1) We didn't know that Mono was the wrong tool for that job before doing the experiment.
2) The install size is very large compared to almost anything you can actually install with a Debian Jessie main packages.
3) The performance is way slower compared to Node.js.
4) The memory footprint is a least twice the memory footprint of Node.js.

Given the results, I doubt that Mono is the right tool for any new projects. It could maybe allow to run historical C# code on Linux that rely on some .NET infrastructure, but this outside of that purpose I don't see any advantage.

Node.js has a clear and big advantage in term of installation size, processing performance and memory footprint. In addition Node.js has the sweet advantage to use the same Javascript language than on the browser client. Even if this is not simple to measure, this feature drastically reduce some kind of "brain footage" needed to deliver a project.

Comment Not efficient (Score 4, Informative) 355 355

I work on a embedded Linux system running Debian Jessie armhf on a Cortex-A5 processor. At some point someone programmed a Web user interface for the system using Mono for Linux. The installation of Mono was difficult, requiring several hundred Mo of space on the filesystem and some trick to get the last package revision. Then the application was started and take all the processing load for almost 4 minutes. At his point it was eating near half the memory available on that embedded system. This was socking, especially for me that like to use qooxdoo for WebUI because it's basically a static file that need no compilation and have a very minimal memory footprint. Finally the guy switched to node.js for the WebUI on that system. The installation was easy, the startup compilation last now less than a single minute and the memory footprint is below 20%, all of that with a more complete demo that with Mono.

Comment Re:x86 (Score 1) 108 108

The agreement is fairly bidirectional and this was a big win for AMD, a kind of life insurance. Even if something go wrong with AMD, there still have the unique value of the cross license granted by the agreement. And given the massive market involved, there will be investors seeking for this kind of value. This is a open gate to a big market. No other open gate to this market exists at this moment.

Comment Re:There's other things they could do (Score 1) 108 108

Yes, but while there try to give usable Linux driver, there completely failed to support a recent leading standard distribution like Debian with a native compiler build system and all the fun and efficient tools. There are stick to an outdated Timesys distribution with an unbelievable obsolete build architecture. There still use static dev over udev, proprietary kernel driver build instead dkms, and no packaging.

There urgently need to evolve from a 'hardware staff that try to code application' point of view to 'software staff use standard distribution to dynamically load hardware capabilities' point of view. There have the advantage on the hardware but obfuscate it by an inappropriate software stack.

Going with AMD is an old idea as Altera was basically a Intel spin-off from the beginning. But this path will expose even more there products to a market where support for standard leading distribution is required and obsolete outdated almost proprietary bullshit rejected.

Comment Masse Storage Device still not back. (Score 1) 106 106

This was the most simple and useful USB function and there removed it. It allowed to play with any computers, TVs, or printers with high speed. Compared to MSD, the "new" protocols are slow and poorly supported.

There removed MSD for the wrong reasons as there is technical solution to provides a snapshot of the internal filesystem as a external view to avoid corruption.The btrfs filesystem support snapshot for example and a gateway could translate the resulting snapshot view in a other filesystem format like FAT32.

Comment Re:Anyon else see the construction? (Score 1) 100 100

Thanks for the offer, but I don't plan to climb so high mountains even if I live in the middle of some of them. I enjoy walking the summer and ski the winter, but no high performance sport for me :-) I will probably take the panoramic cable car with my children someday.

Comment Re:Anyon else see the construction? (Score 1) 100 100

In fact the photo was taken precisely from the lower side of the cable that act as the aerial pylon of the Mont-Blanc Panoramic cable car. If you zoom you can follow almost the full installation but the blurred part before the Gros Rogon pylon buildings.

I wrote Dents du Midi by mistake because there take most the view from my window :-)

Comment Re:Wait, there's a missile silo there! (Score 1) 100 100

No need to be billionaires and no need of helicopter. This is a cable car arrival and the price is far lower than with an helicopter, at least from the Italian side. From the French side the travel is more expensive because this is a long travel including a impressive 5km panoramic track with an aerial pylon.

Earth is a beta site.