The end of LinuxDevices? [LWN.net]

Another source of info

Posted Feb 4, 2012 21:06 UTC (Sat) by david.a.wheeler (guest, #72896) [Link]

For Linux and open source software specific information, you might check out lwn.net. I hear it's pretty good :-).

Another source of info

Posted Feb 4, 2012 23:07 UTC (Sat) by SecretEuroPatentAgentMan (guest, #66656) [Link]

Sure LWN is good, it is my primary source of news for open source software. Still, for embedded systems and hardware I followed LinuxDevices and still follow OpenCores plus EETimes, various Usenet news groups and some newspapers. There just doesn't seem to be an LWN equivalent for hardware, unfortunately.

Another source of info

Posted Feb 4, 2012 23:42 UTC (Sat) by mtaht (✭ supporter ✭, #11087) [Link]

perhaps this could end up being an opportunity for lwn. It would be nice to have one stop shopping for embedded too!

I note, however, that embeddedlinuxdevices tended to be far more of a press-release thing than the in-depth thing we get from lwn...

Another source of info

Posted Feb 5, 2012 3:01 UTC (Sun) by karim (subscriber, #114) [Link]

Sure, but there's a place for that too.

Maybe a "hardware news" section right here on LWN?

Another source of info

Posted Feb 5, 2012 8:01 UTC (Sun) by xxiao (subscriber, #9631) [Link]

linuxjournal used to publish embeddedlinux journal, which was _great_, not sure why they stopped it, it's pretty in depth then(around 2000/2001 time frame).

time for an embedded linux site!

Another source of info

Posted Mar 2, 2012 4:18 UTC (Fri) by btraynor (subscriber, #26672) [Link]

Perhaps an addition to elinux.org is in order.

Another source of info

Posted Feb 5, 2012 14:30 UTC (Sun) by clump (subscriber, #27801) [Link]

Sounds like you should volunteer.

Another source of info

Posted Feb 5, 2012 22:08 UTC (Sun) by SecretEuroPatentAgentMan (guest, #66656) [Link]

Since I have worked for some of the companies that frequently appear in these news I hope for understanding that I cannot be expected to be entirely objectively in such writings. It would therefore be best I didn't write about it.

What I can do and what I did, was to list a few of my sources for embedded related news.

Another source of info

Posted Feb 6, 2012 2:27 UTC (Mon) by josh (subscriber, #17465) [Link]

If you have concerns about your own personal objectivity, you could still forward interesting hardware news links to LWN, relying on the objective reporting of publications like those you mentioned.

But in general, many people in the Linux kernel community work for various companies, and nonetheless manage to objectively deal with contributions and discussions with their competitors. (Also, you said "worked", the past-tense of which suggests you may not anymore, which would make a potential conflict of interest even less likely.) Given that, I see no reason why you couldn't sensibly provide good reports on the state of open hardware and embedded Linux, if you had the time and inclination to do so.

Ruling out anyone with experience in the field doesn't seem like a recipe for good reporting. :)

Another source of info

Posted Feb 8, 2012 15:42 UTC (Wed) by SecretEuroPatentAgentMan (guest, #66656) [Link]

Objectivity cuts many ways. It is possible that in the line of my work I get to know things that I cannot report and also would make general reporting awkward. Also if I report on news others will question my objectivity for the same reasons, moreso since I prefer not to name my clients, which in any case will change over time and require clearance work due to professional ethics.

It is true that some of the big names in embedded no longer are my clients, nevertheless some are considered attractive as clients and it is therefore quite possible I will again work for these companies.

Another source of info

Posted Feb 6, 2012 16:00 UTC (Mon) by sjj (subscriber, #2020) [Link]

There is no such thing as "entirely objective". Except maybe a book of the digits of Pi or somesuch. Not all that interesting.

Having worked for some companies in the field would actually be a huge plus, since you have a much better grasp of the players and their goals.

Just be open about your biases, if any. In this age when the internet is huge dump of opinion media, most intelligent people have learned to figure out where writers are coming from.

Another source of info

Posted Feb 6, 2012 8:58 UTC (Mon) by ebirdie (subscriber, #512) [Link]

phoronix.com does run a steam of news, benchmarks, reviews etc. about hardware with dedicated Linux/OpenSource angle.

I'd love to see the idea of hardware section in lwn.net, although getting one to the depths of Kernel pages & Co will not come easy or may be even impossible for most of the time. Reading and aggregating open source code and open discussions is a whole different game than getting your hands on a piece of supposedly open "linux" hardware and getting a grip of it.

Michael Larabel of the Phoronix has done a lot to automate hardware testing by developing an "open" benchmarking suite and easing the pain with the software stack by standardizing on Ubuntu. Phoronix also seems to have heavy emphasis on graphics hardware and on its software stack, but ie. networking seems to be out of personal intrests for Larabel. It seems that Phoronix is pretty much a one man's project.

Another source of info

Posted Feb 6, 2012 13:40 UTC (Mon) by dbnichol (subscriber, #39622) [Link]

It may be worth noting that many people actually involved in the Linux graphics software stack think that Phoronix is garbage and that the purpose of its content is mostly to drive page hits. I don't think you'd see kernel folks talking the same way about lwn.

Another source of info

Posted Feb 6, 2012 16:35 UTC (Mon) by cmorgan (guest, #71980) [Link]

I've always though Phoronix did a good job with their testing and reviews. Do the Linux graphics guys not like the results or methodology? I've found the reviews and testing pretty thorough and most of them are automated so they can be run across dozens of kernels/distributions.

Another source of info

Posted Feb 6, 2012 16:44 UTC (Mon) by bronson (subscriber, #4806) [Link]

I think it's mostly the breathless, sensationalist headlines, the awful ads heaped everywhere, and the fact that it takes 15 clicks to read any single article.

Phoronix's content has been good but its presentation is a little tough to swallow.

Another source of info

Posted Feb 6, 2012 17:26 UTC (Mon) by josh (subscriber, #17465) [Link]

Mostly, that Phoronix rather quickly seizes on random development discussions and makes them ridiculously visible while still in progress. It becomes difficult to have speculative discussions on mailing lists when a random comment from a developer at $FOO_ORG of "Could we consider doing $BAR; what might it take to do so?" becomes a headline "$FOO_ORG planning to do $BAR!", frothing up throngs of angry and uninformed users (and uninformed managers of the developers having the discussion).

In some cases Phoronix has useful reporting, but it frequently includes a heavy dose of speculation and sensationalism.

Another source of info

Posted Feb 6, 2012 17:58 UTC (Mon) by Cyberax (✭ supporter ✭, #52523) [Link]

Yet it's quite often useful.

I've stopped following several mailing lists because Phoronix usually reports interesting news from them.

And their benchmarks are quite interesting in themselves. It'd be great if kernel developers could read them as well.

Another source of info

Posted Feb 6, 2012 18:26 UTC (Mon) by josh (subscriber, #17465) [Link]

Yet it's quite often useful.

I agree, and I said the same thing.

I've stopped following several mailing lists because Phoronix usually reports interesting news from them.

I don't trust the quality of their reporting enough to do that. I expect them to report on whatever generates the most noise, while ignoring other interesting threads. By contrast, LWN's reporting *has* made it nearly unnecessary to read LKML, but LWN has demonstrated a much better filter for interestingness, and LWN doesn't feel the need to apply sensationalism or otherwise kick hornet's nests.

And their benchmarks are quite interesting in themselves. It'd be great if kernel developers could read them as well.

That I'll agree with. While Phoronix's reporting about regressions (in performance, power usage, or otherwise) contains more than the average amount of sensationalism, at least there it serves a good cause. That said, I think the work by kernel developers on regression handling serves the same function without the drama.

Another source of info

Posted Feb 6, 2012 17:27 UTC (Mon) by SEJeff (subscriber, #51588) [Link]

The reporting _quality_ of phoronix vs lwn is completely different. You look at some of the editors here @ lwn such as Jake or Jon... They will research a topic, dive into the source code, learn how things work, and *then* write about it.

Phoronix is more of a site that does pre-canned tests with default OS / software configurations and publishes pretty graphs. Michael also quite obsessively follows git and various news sources for a lot of the linux graphics stack. He does great reporting for those areas, but as far as quality, he just isn't in the same league as lwn.

Note that I'm a lwn subscriber and avid reader of phoronix. I think Michael does a good job of following the really obscure stuff about Wayland, the Intel Linux graphics stack, kernel mode setting news, etc. However, phoronix is not lwn and lwn is no phoronix. It isn't fair to compare them.

Another source of info

Posted Feb 7, 2012 16:15 UTC (Tue) by ebirdie (subscriber, #512) [Link]

I fully agree on the quality, but my intent wasn't to compare them on quality. First I just wanted to mention Phoronix as one service to check as a substitute to LinuxDevices, if the LinuxDevices will die. Speaking of embedded hardware Labarre lately tested and compared couple ARM development boards, which were very interesting as compared to usual desktop gear. This was a micro comparison of Phoronix vs. LinuxDevices content. :-)

Secondly what I was after to express, was that in the wish of a hardware section at lwn.net, I think, it is unrealistic to expect the same level of quality as we are used to in kernel, security etc sections of lwn.net. In part this comes from the fact how most hardware is put up, which we are very aware of. As such hardware is much worse to research compared to grinding problems, concepts, organization and chaos in open software code and putting up the material as spicy articles as lwn.net has done. Thus hw digging means more work and effort to produce articles. And that is very important to acknowledge for a business. What one can put up from a work of four hours with hardware? I think the ratio is much worse for hardware than for open software. So it does matter.

To improve the ratio one has to develop methods, standards and tools to research hardware. Maybe there is something to do in open source community in this respect. That is also where Michael Labarre has done something. I think, what he has accomplished is very interesting. But on the other side I see his accomplishments as one man's project, where isn't much room for participation and/or for a community to grow up from. In this respect I see lwn.net in better shape with guest writers, for one example. My vague perception is that subscribers are one community around lwn.net. But in a sense there hasn't any community grown up from lwn.net either.

After a while back I saw a news bit, where Corbet was after a new editor in staff. Maybe Corbet could go after the outing editor-in-chief of LinuxDevices? Well that is just an wild idea or wishfull thinking out loud for getting a new and resourced hardware section.

Anyway, my assumption is that for making a hardware section happen, it needs to be resourced many ways. I assume Corbet & Co already has a harvesting machinery for gathering bits and pieces. The machinery might be extendable for production purposes of a hardware section. What else resources might be needed? Another hardware database? There already has been many sites with hardware databases to promote and share information of "Linux compatible" hardware. I think yet another database isn't the right content. What about performance benchmarking? That is where Phoronix and numerous others with Windows tools already are, so very likely not.

I think lwn.net already has had decent flow of bits from hardware land, which could be a base for the hardware section. The format of the hardware could follow the format of the security section, as an example. One article like Aaron Seigo's KDE Plasma Active tablet project is a sure fit. Firmware issues are too. And then few information bits about announcements in hardware land, including chips, where the manufacturer mentions any kind Linux support, open documentation or firmwareless design. No further evaluation needed in the briefly listed announcements.

Well I hope my comments aren't felt offensive, or something alike, by Corbet or anyone else as I usually say something in line as thinking out loud I somehow manage to put people off way too often although it is not my intention. :-(

The end of LinuxDevices

Posted Feb 23, 2012 22:43 UTC (Thu) by johnnyangel (guest, #83089) [Link]

Having just caught up with this thread, I thought I'd chime in. I'm the former LinuxDevices.com "Editor-in-Chief" -- sounds grand, doesn't it? In fact, at the end we were down to just myself and one freelance writer, both of us trying to update three different websites.

Since QuinStreet acquired Ziff Davis Enterprise at the beginning of February and immediately laid me off -- along with many others -- LinuxDevices.com, WindowsForDevices.com, and DesktopLinux.com have been in, shall we say, suspended animation. Many people have written me to ask about the future of the site archives, at the very least, and some have also kindly volunteered their services in somehow continuing LD.

In fact, it has become a challenge to respond to all the well-wishers individually. Perhaps this post can partially suffice: All I can really say is that I've heard nothing from anyone at QuinStreet about the sites' future, nor do I even know anyone there to ask. If I hear any news, I'll post it on my blog and Twitter feed.

Regarding continuing LD's efforts on my own under another name, all I can say is that I've had many ideas over the years for projects -- websites, books, board games, and a host of other things -- and I continue to have 'em. But, I'm a mere writer and editor who has little interest in handling money (even making it isn't of much interest, as my present and former wives would sadly agree), selling ads, or meeting a payroll. And, I need paying work too; if I'm going to write for free, I'd probably do so regarding one of my hobbies (for instance, antique toy cars, music, or audio).

That likely means the LD tradition will have to be carried on by others, and other sites -- at least for now. I'm interested in compiling a list of recommendable sites as a service to former readers, and invite your submissions (which can be sent via comments on my blog, www.gadgetsense.com, or to edit at angel.org).

Meanwhile, I'd like to defend LD against the LWN commenter who characterized the site as "a press-release thing." It's true that LD was product-news driven, but we certainly did not reprint press releases (unlike many other sites), did extensive work to create original tables or otherwise listing product features, and always maintained an independent and skeptical approach -- as I think a fair analysis of any of our stories would prove.

As a former reviews editor, product tester, member of the InfoWorld review board way back when, etc. etc., I'm the first to agree that a *single* story on LD per day that incorporated actual product testing or at least more detailed analysis would have served readers better than four or five stories listing new products -- even if we did our best to deconstruct them. Unfortunately, management didn't see it that way, and we were under pressure to produce quantity, not quality.

Plus, there was absolutely no budget to commission outside articles, nor even enough time, at the end, to edit white papers that were offered to us gratis.

It should be noted that Rick Lehrbaum, who founded LinuxDevices.com, sold the site to Ziff Davis Enterprise in 2004. He stayed on through the end of 2007, but after his departure, management lost interest and slashed the staff. Not long before the end, I really had no idea what the company expected from me and I didn't even have a manager whom I could ask.

None of the above is intended to bash former colleagues at ZDE, who were all as over-stretched as I was. The company had simply taken on more than it could digest, and I was working in one of the forgotten outposts of the empire.

As an aside, I'm interested to have read other LWN comments from back in '04 suggesting that the sale of LD to ZDE meant the site would inevitably be selling out to Microsoft. It's true that LD's sister site WindowsForDevices.com was subsidized by Microsoft advertising, which enabled us to -- independently but relatively slavishly -- cover Windows Embedded operating systems and related topics.

But, Microsoft never interfered editorially in the coverage on either site. And ironically, because most articles we wrote regarding hardware were "cloned" to both Linuxdevices.com and WindowsForDevices.com, Redmond was actually facilitating expanded Linux coverage.