User: Password:
Subscribe / Log in / New account

Ten-year timeline part 4: the end and the beginning needs you!

Without subscribers, LWN would simply not exist. Please consider signing up for a subscription and helping to keep LWN publishing

By Jonathan Corbet
January 30, 2008
When your editor started this series, the idea was to have four installments covering the ten-year life (so far) of LWN. Well, this is the fourth installment, and it gets less than halfway there. This is not, it seems, a topic which inspires brevity. So this series will continue past the anniversary, though your editor anticipates picking up the pace a bit for the second five years. There is less to be learned, arguably, by looking at events in the relatively recent past.

Anyway, at the end of the third installment, LWN had been unacquired by Tucows and was, once again, on its own. The worst of the dotcom bust may have passed, but it was still a somewhat scary environment in which to be attempting to restart a business. It was, in fact, even scarier than we had thought when we so naively set out to show that we could do a better job of bringing in the cash than Tucows did.

  • February 7, 2002: Linus tries BitKeeper at last.

  • February 14, 2002: Sun states that it will "ship a full implementation of the Linux operating system." Dave Whitinger joins

Dave Whitinger was, of course, one of the founders of LinuxToday. He joined LWN with the intent of helping us develop the advertising side of the business. That did not work out as intended, but it is hardly Dave's fault; it was a terrible time to be trying to sell advertising.

  • February 28, 2002: Sun cuts off free access to StarOffice, but we had by then and didn't mind. BitKeeper starts to settle in as the kernel's source management system.

Linus stuck with BitKeeper after his initial trial, setting a number of things in motion. For the next few years, the use of proprietary software at the core of the kernel development process would be a constant source of unhappiness and worry - and, in fact, the story had just the sort of unhappy ending that some observers had feared. But this was also the move which rationalized the kernel work flow and made the whole system scale; the incredible rate of change we see now would not have been possible without it. The use of BitKeeper also made the community aware of what distributed source control could do and, eventually, inspired the creation of a number of free programs with the same essential features. One could say that the community would have eventually developed these systems on its own without the push from Larry McVoy and BitKeeper, and that's probably true. But the fact is: we didn't do it at that time, so we had no real alternative to BitKeeper.

  • March 7, 2002: Martin Dalecki's "IDE cleanup" patches start to raise concerns among kernel developers, who have this strange notion that their disks should actually work. A petition against the use of BitKeeper circulates on the net. Eric Raymond goes around telling the world that the kernel development process is "in crisis."

  • March 14, 2002: Richard Stallman claims that the GNU HURD will be ready by the end of the year. MandrakeSoft pleads for donations to keep the business alive - and LWN does too. Martin Dalecki officially takes over IDE maintenance - and breaks more systems.

We got about $5,000 from our initial plea for donations. It was a real act of generosity on the part of our readers, but one does not keep a business with five employees going for very long with that sort of money.

  • March 28, 2002: The proposed "consumer broadband and digital television promotion act" would require DRM technology in all software which touches digital media. Lineo lays off more staff.

  • April 25, 2002: More BitKeeper flames. Lineo goes through a "recapitalization" effort to be able to do things like pay its employees.

  • May 2, 2002: 1.0 is released.

  • June 6, 2002: LWN switches to the "new" site code. Red Hat applies for a few software patents. ADEOS, a real-time system which avoids the RTLinux patent, is released. UnitedLinux launches. Mozilla 1.0 is released.

It is amazing how many readers hated the new code. Certainly there were a lot of silly things in the initial version of the site; we fixed a number of them in a hurry. Many readers disliked the ability to post comments - often posting comments to that effect. The addition of comments was something we thought about carefully for a long time; we were quite concerned that they could ruin the feel of the site. In the end, it seems, trusting our readers has paid off; the quality of the conversation here is often quite good.

UnitedLinux was a cooperative effort between Caldera, Conectiva, SuSE, and Turbolinux; the idea was to join together to create a common base from which each could then craft a separate product. The effort was never all that successful, and the presence of Caldera would, of course, doom it outright in the end. But it was a big deal at the time. It is interesting to see that Mandriva (despite MandrakeSoft's refusal to join UnitedLinux) and Turbolinux are now attempting a very similar sort of arrangement.

  • June 13, 2002: Secure Computing Corporation claims patents on SELinux.

  • June 27, 2002: The 2002 kernel summit sets October 31 as the date for the 2.6 feature freeze. GNOME 2.0 is released.

  • July 4, 2002: Darl McBride takes over at SCO.

  • July 25, 2002: LWN announces "the end of the road." The "IDE cleanup" patch series (up to number 100) causes system lockups and file corruption. Debian GNU/Linux 3.0 ("woody") is released. Version 1.0 of the Ogg Vorbis codec is released.

By the end of July, we had come to realize that the advertising business was not going to work out for LWN, and we were short of other ideas. The bank account had reached a point where we could not pay even very small expenses. So we concluded that it was time to throw in the towel and try something else - though we had no clue of what "something else" might be. It was with a heavy heart that we announced our plan to shut down the site.

What happened next is that our donation box, which had sat mostly empty after the initial announcement, was suddenly topped up to the tune of about $35,000. Many of the donations came with notes to the effect of "use this to throw a big party." This, shall we say, got our attention. We decided that, just maybe, the subscription idea was worth a try after all, and decided to make a go of it. It was not the end after all.

  • August 1, 2002: A new beginning. HP tries to use the DMCA to shut down disclosure of security holes.

  • August 15, 2002: Distributions from MandrakeSoft, Red Hat, and SuSE are certified to be compliant with the Linux Standard Base.

This was when our credit card merchant bank at the time decided that all those donations might just be fraudulent. So they seized the money back out of our bank account. That, too, got our attention. It took a few months and some lawyer time to get the money you all had sent in our direction; during that time, it was money from PayPal (the subject of everybody else's horror stories) that kept the lights on while our main source of cash was blocked.

Needless to say, we got a new merchant bank, which we still use to this day. The new bank exhibits a rather higher clue level than the old one did, but we also learned a valuable lesson: don't mess with the credit card money pipeline. Every now and then, somebody asks why we don't accept pure donations; this is why.

  • August 22, 2002: Martin Dalecki quits and the entire series of 115 "IDE cleanup" patches is deleted from the 2.5 kernel.

  • August 29, 2002: British Telecom's attempt to patent the web dies in court. The BitKeeper license changes. Caldera becomes the SCO Group.

  • September 12, 2002: Some patches get dropped after Linus starts running his mail through a spam filter.

It's hard to believe that, only 5+ years ago, somebody with an email address as well distributed as Linus's could get by without spam filtering. There are a lot of free "productivity" applications, but, arguably, few have actually increased productivity to the extent that SpamAssassin has.

  • September 26, 2002: The first development release of the "Phoenix" browser is announced. UnitedLinux upsets the community by releasing a closed beta.

Phoenix was the Mozilla Foundation's answer to (relatively) lightweight browsers like Galeon, which had managed to turn the Gecko engine into something which was truly usable. The Phoenix browser proved popular, and eventually became the tool now known as Firefox.

  • October 3, 2002: The first subscriber-only weekly edition. Eldred v. Ashcroft is argued in the U.S. Supreme Court.

Eldred v. Ashcroft, argued by Lawrence Lessig, was an attempt to roll back copyright extension in the US; it eventually was unsuccessful. To this day, there still has not really been a successful challenge to the extensions to copyright passed over the last few decades - though some especially nasty attempts to make things even worse were defeated.

With the October 3, 2002 edition, LWN adopted the new policy of requiring subscriptions in order to read our original content prior to the publication of the weekly edition. That policy has stayed essentially unchanged since then, despite the occasional temptation to increase the subscriber-only period. Subscription rates have also stayed unchanged, even though raising them is also tempting.

Subscriptions have certainly been successful, in that they have kept the operation going in the years since then. And there is a real joy associated with being truly answerable to our readers instead of advertisers. Nonetheless, it is a challenging business; people do not like to pay to read web-based content. The fact that so many of our readers are willing to do so is most gratifying. Trends in other parts of the net are moving away from this approach, though, with formerly subscription sites moving to pure advertising models. So it will be interesting to see how it all plays out in the future.

Meanwhile, next week's installment will look at how things went for Linux (and LWN) starting toward the end of 2002. Stay tuned.

(Log in to post comments)

Ten-year timeline part 4: the end and the beginning

Posted Jan 31, 2008 1:40 UTC (Thu) by mbcook (subscriber, #5517) [Link]

It's hard to believe that I've been reading LWN for so long. Slashdot recently had it's
anniversary and  
I had the same reaction there.

I had been reading LWN for a couple of months, maybe a year, when the subscription thing came 
down. I don't remember if I was reading before all the financial trouble started. I think I
was, but I'm 
not sure. It may have been the reports of the financial trouble that took me to LWN in the

I read LWN for a while after the subscription thing happened, having to wait for the content
expire and become free. After a while (I don't know how long) I signed up. Here I am today,
just renewed in the last two months.

I love the Kernel page. I've learned so much. That and Kernel Traffic have allowed me to learn
much about Linux. Of course, Kernel Traffic died ~25 months ago... leaving just LWN.

Ten-year timeline part 4: the end and the beginning

Posted Feb 1, 2008 2:35 UTC (Fri) by jengelh (subscriber, #33263) [Link]

___Kerneltraffic and LWN have vastly different writing styles. I actually preferred KT over

___KT picked up the interesting threads on LKML, and made an interesting read solely based on
developer's citations, mostly direct speech. It was an art of taking snippets from everywhere,
and pasting it together again in a coherent fashion. This gave it a feel like you were
listening to a parliament discussion -- judgement: "raw material".
The last KT#335 issue covered 14 threads.

___LWN on the other hand uses the usual "newspaper" style. (This week: mem_notify and the blk
rq api.) The Editor makes himself familiar with the topic and then writes down his findings
from his viewpoint. Standard (e.g. "cups slow on linux-2.6.24") and philosophical
(hypothethical "libata does not power up my usb coffee pad, who broke it?") matters don't get
much attention, while really technical things are explained in a way so that people who are
not a master of the particular subsystem still get to understand what it's about -- judgement:
___Sometimes, there are only two [longer] stories, sometimes there are four stories in the
Kernel section of the Weekly Edition (e.g. 2006-01-17) - it kinda fluctuates. (Probably
related to LKML I guess, everyone's busy with merging for 2.6.25 right now.)

___I dunno if LWN has the necessary manpower for it, but getting to read something KT-style
would be great. Does not need to be a dozen of threads, but maybe two, three threads that did
not make it into a "newspaper" article. Spot on: the ext3 feature freeze, er, ext3 freeze
feature thread, softpanic v2, and perhaps the ext4 list could have been showcased KT-style for
this week.

Ten-year timeline part 4: the end and the beginning

Posted Feb 1, 2008 2:38 UTC (Fri) by jengelh (subscriber, #33263) [Link]

s/this week/week of 2008-01-24/
And pre-chewed was meant to be positive. There are a few areas in the kernel I do not want to
tamper with, so reading about it in this form is welcome

Ten-year timeline part 4: the end and the beginning

Posted Jan 31, 2008 1:43 UTC (Thu) by rahvin (subscriber, #16953) [Link]

People are willing to subscribe because unlike sites like Slate there is real journalism on
LWN. In an era where Journalism has died and most publications are nothing more than slightly
upscale tabloids LWN has stayed with real unbiased reporting of facts with the occasional
editorial. As a highly technical publication covering such a broad range of the Linux world
it's an essential source of information, because as your other co-founder said, not everyone
can subscribe to 150 newsgroups and mailing lists and sort out what matters. Keep up the solid
commitment to real journalism and keep the in depth coverage of Linux and I think you can even
grow the subscriber base as Linux adoption continues. 

Ten-year timeline part 4: the end and the beginning

Posted Jan 31, 2008 16:39 UTC (Thu) by kirkengaard (guest, #15022) [Link]

Agreed.  LWN is worth my subscription (or lump sums, depending on my income at any given time)
exactly because of the high S:N here.  Even when I've read the news which is posted here on
its original pages, a) it is noteworthy that LWN mentions it, and b) the comments from readers
here are usually worth a better look than some other news outlets.  High-signal, high-clue
newshounding and editing, with higher-than-average-clue readership.

Consider that no other place would be likely to get away with such a text-centric design, and
still draw readership.  It's the quality of the product over the packaging.

Ten-year timeline part 4: the end and the beginning

Posted Jan 31, 2008 21:04 UTC (Thu) by nix (subscriber, #2304) [Link]

Adding piles of low-info graphics would make the presentation *worse*, so 
I don't buy that.

(I just buy a subscription.)

Ten-year timeline part 4: the end and the beginning

Posted Jan 31, 2008 17:00 UTC (Thu) by iabervon (subscriber, #722) [Link]

Actually, I think that people subscribe to LWN in large part because there's real tech
writing. The journalism is a good thing, too, of course, but I think that keeping your editor
writing documentation for 2-3 kernel features each week would be worth indulging his choice of
business even if it weren't beneficial on its own merits. On the other hand, I think that LWN
is in a better position to do good journalism because it has an secure value in the technical
articles independent of the popularity of the news it reports.

Ten-year timeline part 4: the end and the beginning

Posted Feb 1, 2008 13:32 UTC (Fri) by stevem (subscriber, #1512) [Link]

Seconded! Great news summaries, plus good technical articles. Keep them coming please.

Ten-year timeline part 4: the end and the beginning

Posted Jan 31, 2008 17:34 UTC (Thu) by graydon (guest, #5009) [Link]

Agreed. This is the only website I've ever paid for a subscription. LWN seems to be one of the only sites produced by people who understand the value equation of time and editing.

Most sites that try to offer subscription services try to wow their readers with "lots of content" available through them. But this gets the equation backwards: if I pay for access to their site, I'm paying to have even more of my time spent digging through stuff on the internet. The internet is already packed with free stuff to dig through! What I'm short of is time.

If an editor manages to take some portion of the internet and faithfully, carefully reduce it to something much smaller -- reduce some important firehoses to a 7-page weekly summary -- they're performing a time-saving service that is worth money. If they do it in a timely fashion, and the information is time-sensitive, then a simple time-locked subscription system is perfectly reasonable.

Ten-year timeline part 4: the end and the beginning

Posted Jan 31, 2008 21:05 UTC (Thu) by nix (subscriber, #2304) [Link]

What he said.

Ten-year timeline part 4: the end and the beginning

Posted Feb 1, 2008 12:42 UTC (Fri) by N0NB (guest, #3407) [Link]


At this time I plan to continue to subscribe to LWN and drop Linux Journal when it comes up
for renewal.  Linux Journal no longer caters to the home/hobbyist user but I renewed when Nick
Petreley took over as editor.  Now that they've quietly parted ways, I think I shall too.
That is not an easy decision as I've been an LJ subscriber since 1996.  

I'm not saying that LJ's content is poor, I just don't relate to it very often.  For example,
Reuven Lerner writes an excellent column, but it's wasted pages to me as I don't do enterprise
type web development.

LWN typically has a lot of useful information that even I can relate to.  It's timely and
informative.  LWN doesn't waste its time trying to convince me to use some web development
acronym of the week.  Instead it focuses on Linux kernel development in particular and the
Free Software community in general and those are of most importance to me.

LWN gives me consistent news about the Free Software community and that's what I value most.

Horse's mouth

Posted Feb 2, 2008 23:08 UTC (Sat) by man_ls (guest, #15091) [Link]

Agreed again. It's also the only magazine I subscribe to, great value for the money.

I actually do enterprise web development, and use some of those acronyms in my daily routine through enterprise-style chaos, so I don't relate to lots of LWN content. However, I value solid software engineering in kernel development as in any other field. Lots of ideas are usable in any kind of environment.

Plus, my favorite point: LWN is the only place I know where the news meets the newsmakers. You can read our editors' take about something and then read a comment from the horse's mouth, be it an average Debian developer or Linus himself.

Ten-year timeline part 4: the end and the beginning

Posted Feb 4, 2008 15:25 UTC (Mon) by filipjoelsson (guest, #2622) [Link]

LWN is the only computer related mag (on- and offline) that I bother paying a subscription
for. The reason is that all the rest of them have become product review or preview magazines,
where even most of the articles try to sell me gizmos or software. Oh, they usually have
"developer pages" too - of the learn php/python/gimp/something else in 5 installments,
blatantly ripping off some online tutorial (and usually doing worse). And to top it off, they
have some month old news.

All in all, LWN is the only place writing about what's interesting in a timely fashion,
cutting the crap and the disguised ads. Which makes it well worth my money. :)

BTW Did you actually throw that party? I believe I explicitly tagged my donation as "for
beer". Was it good? ;)

Ten-year timeline part 4: the end and the beginning

Posted Jan 31, 2008 2:48 UTC (Thu) by AdHoc (guest, #1115) [Link]

LWN is the only web content I've ever paid for. But it is certainly worth it.  Here's hoping
for another ten years :)

Ten-year timeline part 4: the end and the beginning

Posted Jan 31, 2008 13:12 UTC (Thu) by Kluge (subscriber, #2881) [Link]

It's the only online publication I've paid for too.  And well worth it.

I can't remember what I thought about comments at the time, but now they're one of my favorite

Ten-year timeline part 4: the end and the beginning

Posted Jan 31, 2008 16:06 UTC (Thu) by jzbiciak (subscriber, #5246) [Link]

I know I've been a fairly large fan of comments since the beginning.  

I guess there was always some risk of trolls and flame wars, but with the 1 week delay on
original content, at least those of us with some level of commitment to the site manage to
have reasonable discussions on those articles before the wider audience arrives.  In practice,
the signal to noise ratio has remained fairly high on all pretty much all comment threads I've
seen.  This hasn't turned into Slashdot, or worse, FARK.  :-)

Ten-year timeline part 4: the end and the beginning

Posted Jan 31, 2008 3:10 UTC (Thu) by paulmfoster (guest, #17313) [Link]

I remember when LWN first hinted it might shut down. Knowing Robin Miller 
(, Sourceforge, etc.), I contacted him about the possibility of 
those guys absorbing LWN. He liked LWN, but I think he believed absorbing 
LWN wouldn't happen unless Corbet and company were fully on board with it.

In any case, even though there are plenty of other (Linux) news sites out 
there, LWN is by far the best, and the only one I do or would pay for. I 
hope it continues to pay for itself in the future.

Darl McBridge / SCO / Caldera

Posted Jan 31, 2008 3:23 UTC (Thu) by ctg (guest, #3459) [Link]

"Caldera" is used throughout the rest of the 2002 references, except for 
the one for Darl McBride.  It gives the impression that he became 
CEO of SCO (as in old SCO), rather than Caldera.  And I'm pretty sure he 
took over from Ransom Love at Caldera, before they renamed themselves the 
SCO Group.

It might be worth mentioning when Caldera took over SCO, and when the 
ancient UNIX stuff was released.

Ten-year timeline part 4: the end and the beginning

Posted Jan 31, 2008 4:10 UTC (Thu) by cventers (guest, #31465) [Link]

Nonetheless, it is a challenging business; people do not like to pay to read web-based content.

You're right, and I'm no exception. But LWN is an exception to my policy, and that's because the content is well-researched, in depth, and the presentation is excellent. That's very hard to find in today's publications.

By the way, I enjoy the ability of a "Project Leader" to turn off ads, but I wish it were broken down into two options. I don't personally want to view the inline flash ads as they can be very distracting when trying to read the content around it. But I do like to see the text ads on the left nav pane. I've got both turned off right now... but if there were an option, I'd turn said text ads back on.

Ten-year timeline part 4: the end and the beginning

Posted Jan 31, 2008 7:25 UTC (Thu) by ebirdie (guest, #512) [Link]

I want to express my gratitude toward as well. I have been reading since 1998
almost weekly and I fully agree with other comments here that maintains remarkable
journalism among not only linux journals but the journals I follow, of which couple are very
high profi(le/t) national publications.

During the recent year I have started to respect A LOT that there is this weekly edition. It
makes me calm down and not chase for information or keep up with the information flow many
sites like slashdot have. Weekly Edition is there as told to be and contains essentials which
I find very relaxing experience among the net publications these days.

As I work as an IT manager and system administrator, what I am frequently missing is a
directory of FOSS service partners. What I'm trying to suggest that advertizing on
could be tied to some kind of FOSS service provider directory, and of _international_ nature
please. There is one a kind of here at The directory and "persistent ads" be seen
by advertizers as an extra service and increase their intrest. From a subscriber aspect this
would make ads more appealing as well like "I saw that ad in beside that news bit,
have to go and check".

Anyhow I find the FOSS service provider business is still badly behind non-FOSS or in places
non-existent, at least here in middle Europe (which depends which way you look at the map
;-)). There is still too much product oriented FOSS-business and their advertizing, I think.

Ten-year timeline part 4: the end and the beginning

Posted Jan 31, 2008 11:08 UTC (Thu) by pointwood (guest, #2814) [Link]

"The addition of comments was something we thought about carefully for a long time; we were
quite concerned that they could ruin the feel of the site. In the end, it seems, trusting our
readers has paid off; the quality of the conversation here is often quite good."

I think the quality of the comments is largely because of the quality of the content. The
content on LWN is well researched and rarely good candidates for flamefests. It's a testament
to how good LWN really is. 

Here's to another 10 years of LWN!

Ten-year timeline part 4: the end and the beginning

Posted Jan 31, 2008 14:47 UTC (Thu) by AdHoc (guest, #1115) [Link]

Another bonus from the comments is seeing kernel developers (or other OSS developers) respond
to stories they are involved in. It's great seeing Ingo Molnar and others add to the

Though, sometimes I wonder how many of them read just to see if they've made Quotes of the
Week :)

Ten-year timeline part 4: the end and the beginning

Posted Jan 31, 2008 16:44 UTC (Thu) by kirkengaard (guest, #15022) [Link]

Definitely a plus - you're read and responded to by the people involved in what you report.
How many 'industry' publications get that kind of regular 'industry-insider' feedback?

Ten-year timeline part 4: the end and the beginning

Posted Jan 31, 2008 16:29 UTC (Thu) by jzbiciak (subscriber, #5246) [Link]

Agreed.  I signed up for a subscription not long after they switched to that model.  I
recently moved up to the Project Leader level, because I can justify the expense.

The insightful articles here, especially on the weekly kernel page, are what bring me back
week after week.  I've learned so much about the kernel's internals and the challenges kernel
developers face reading the detailed articles there.  The ensuing discussions are also
enlightening.  If there's something the article missed, often it's captured in the comments
that follow.  I know I've asked a few questions here and have been thankful for the responses.

My day job had nothing to do with Linux, and in my personal life I haven't even compiled my
own kernel in probably 5 years.  (I've been using Linux for over 14 years now, FWIW.)
However, lately, I'm finding all the stuff I learned here and a couple other places (Kernel
Traffic and KernelTrap come to mind) are becoming more and more useful in my professional
life.  I'm a CPU architect these days, and knowing how the Linux works is very important to
architecting the CPU correctly if we expect to run it in any reasonable form.  My fellow
architects at work who haven't followed Linux like I have in their "copious free time" are at
a disadvantage.  :-)  Maybe I can get my employer to pick up the tab for my subscription now.

LWN holds a distinction over the other two sites I mentioned, since it goes further to
actually digest the content and present it as a coherent whole.  Kernel Traffic, while
interesting in its own right, often put me to sleep unfortunately.  It was too much raw
information.  KernelTrap has been more dynamic, but it still focuses mostly on bringing
interesting threads from LKML and other lists to my attention more than providing actual
original articles exploring concepts in depth.  LWN ends up being much more educational as a
result.  Articles such as the RCU series and detailed explanations of the various schedulers
and so on... I don't know where else I would find that!

Here's to another 10 years!

It's worth writting a book

Posted Jan 31, 2008 11:40 UTC (Thu) by alejluther (subscriber, #5404) [Link]

Hi Jonathan!

Great stuff, as always I should say.

I think it would be interesting to write a book about your 10 year experience. You have been a
good witness for this incredible epoch and have the skills needed to do it.

I'm reading "Bill & Dave" book from Michael Malone about HP history and he talks about how
Silicon Valley was founded, first companies, movements in the market, ... I think you could do
something similar or at least I'd like to read it :-)

Congratulations and best wishes 

Ten-year timeline part 4: the end and the beginning

Posted Jan 31, 2008 15:44 UTC (Thu) by michaeljt (subscriber, #39183) [Link]

I suspect that the LWN editors might be able to make a reasonable amount of money as
consultants - after all, they are presumably among the best informed people around on Linux
and Free Software, both of which are currently hot business topics.

Or then again, perhaps you are doing it already :)

Ten-year timeline part 4: the end and the beginning

Posted Jan 31, 2008 17:19 UTC (Thu) by iabervon (subscriber, #722) [Link]

I'd still argue that the outcome of the whole BitKeeper thing was more what Larry McVoy seemed
to fear than what anybody else seemed to fear. BitMover worked out a lot of the fundamental
misunderstandings in version control through a lot of effort and made the results clear to a
programmer who had written off the entire concept before, but turned around and, in six weeks,
wrote a better replacement, at a cost to the community of a slight delay in the release of
2.6.12-rc1. I don't think it's a coincidence that the only person able at that point to write
a better version control system than BitKeeper was also the only person trying to write a
version control system who had extensive experience using BitKeeper.

On the other hand, the middle period, where people who could abide the license were using
BitKeeper, was awkward, because not everybody had perfect access to the development process.
So I'd say that 2002 started a hard time (albeit less hard than the time before without
significant tool support) that turned out very well in the end, rather than a good time that
went sour eventually.

The BitKeeper episode

Posted Feb 1, 2008 1:46 UTC (Fri) by giraffedata (subscriber, #1954) [Link]

I agree.

It's easy to get the wrong idea about the Bitkeeper story in retrospect and believe it was an I-told-you-so event. Actually, I remember Linus countering some of the anti-Bitkeeper argument in the beginning by saying, "Worst case, it gets taken away from us and I go back to what I'm doing now or write a replacement." The outcome was planned for.

The real opposition to Bitkeeper wasn't the practical concern of the license being withdrawn, but the political statement of the quintessential open source project endorsing closed source software.

Ten-year timeline part 4: the end and the beginning

Posted Jan 31, 2008 22:01 UTC (Thu) by socket (subscriber, #43) [Link]

I'd like to echo the sentiments of others who have said that LWN is well worth the

In fact, LWN is often the standard I use when I feel it necessary to compare journalistic
quality.  I wish there were someplace I could find local and national news that comes even
close to the quality that LWN produces on the rare "off" week.

Ten-year timeline part 4: the end and the beginning

Posted Feb 5, 2008 16:04 UTC (Tue) by NAR (subscriber, #1313) [Link]

Subscription rates have also stayed unchanged, even though raising them is also tempting.

Actually from Europe it's not really true, because the same 10$ worths a lot less today than 6 years ago, so the LWN subscription is a lot cheaper today: the project manager level was 2453 HUF back then, now it's 1735 HUF - and we've had about 5-6% percent yearly inflation, so it's really a bargain.


Ten-year timeline part 4: the end and the beginning

Posted Feb 13, 2008 13:15 UTC (Wed) by ekj (guest, #1524) [Link]

True. For example, from a Norwegian perspective:

$10 in 2002 was 67kr - which was aproximately the net pay from 45 minutes of working a
beginner-level programming-job.

Today $10 is 55kr, which is aproximately the net pay from 20 minutes of working a
beginner-level programming-job.

The price of Lwn measured against what matters, salaries, is more than -halved- in these 5

Even if we ignored the increased living-standard and counted only the dollar-fall and
inflation, it's still a price-fall from 70nok to 55nok in 6 years, a significant decrease.

Besides, LWN has "starving hacker" modes for those who have problems paying the "normal"
price, and that is dirt-cheap, the price of a hotdog in Norwegian terms, literally (actually
in many locations the hotdog may cost more)

boosting revenue with coupons?

Posted Feb 7, 2008 7:00 UTC (Thu) by jabby (guest, #2648) [Link]

While reading this story an idea occurred to me for how to increase the number of subscribers.
Has LWN ever considered offering subscription rebates/coupons via major Linux/FOSS-related

I'm not a businessperson so I don't know all the details of how such things work, but I
imagine that some vendor in the Linux/FOSS gig could include with their offerings a coupon for
a reduced price or limited trial subscription to LWN.  They could stipulate that the coupon
would only be valid for *new* subscribers (not sure how you could enforce that, though).
Anyhow, some fraction of customers use the coupon and try out LWN.  Some fraction of those
become new reader-subscribers of LWN.  For each of these "coupon accounts" that convert to
paid subscribers after the trial period runs out, the company that was so kind as to include
your coupon could get a one-time kick-back.

I'm sure more business-savvy people will poke holes in my idea and find all manner of
weaknesses.  I can also appreciate the fact that these new readers may be some of the least
reliable for renewal.  (But then one also has to consider the ripple effect of at least
spreading awareness among a demographic who are already spending money in the "Linux market.")
Anyhow, if the problems/weaknesses can be overcome, it could possibly broaden the readership
and provide a boost to LWN's revenue stream.

Best wishes for another 10 years,


boosting revenue with coupons?

Posted Feb 7, 2008 14:25 UTC (Thu) by Duncan (guest, #6647) [Link]

The idea is relatively common in industries such as entertainment and 
hospitality, but often works somewhat different than you outlined.  As you 
outlined, one party ultimately ends up paying hard cash as a kickback.  
The way it often works in E&H is as a "coupon exchange".  That is, an 
entertainment facility will ship coupons to various local hotels/motels 
and restaurants to distribute there, which in turn ship coupons back to 
the entertainment facilities for distribution there.  The cross-promotion 
thus costs little but the printing and mailing costs, but tourists and 
hungry restaurant patrons pick up and sometimes use the entertainment 
coupons, while after their entertainment, patrons there are often tired 
and hungry and find the restaurant and/or motel coupons useful as well.  
As an additional bonus, all participating establishments are seen as 
active participants in the local community, boosting trade not only for 
themselves but for others as well.  Of course, the usual government and/or 
chamber of commerce sponsored natural and man-made attraction tourist info 
fits right into this model, with the various coupons generally displayed 
right beside the various informational pamphlets.  Ideally, tourists take 
some of them home in their souvenir packets and end up sharing them with 
others, who come the next year...

Actually, LWN sort of does this already, informally.  I discovered LWN 
back shortly before the subscription thing began, as a result of finding 
it a site choice in knewsticker.  LWN in turn covers KDE and points to 
stories of interest on the dot (, KDE's news site).  I'm not 
sure how much of this is just covering the territory, however, and how 
much might be deliberate, but it's worth trying a bit more of the 
deliberate stuff. =8^)

(As for me, I'd love to subscribe and did for awhile.  However, my 
conscious began to prick me as I realized I was making demands of folks 
like ATI to provide open source drivers or at least specs for their stuff 
before I sent any more of my money their way, yet LWN, who had made the 
promise to open their own code years ago, had yet to do so and was still 
getting my money.  Thus, it became a personal integrity issue, and after 
letting the LWN folks know why I could no longer subscribe, I had to let 
my subscription lapse.  LWN still has that promise to open their 
code "once it gets a bit more ready".  During the intervening years, Sun 
has promised and almost completed opening up Java, and ATI, after itself 
promising and failing to deliver for years, has begun opening its hardware 
specs to the community as well, yet, despite what I must assume are the 
best intentions, LWN continues to prevaricate on its own promises.  When 
as a FLOSS community proponent I can as a man of integrity again sponsor 
this otherwise great beacon of the FLOSS community with my own 
subscription, I'll be /extremely/ happy to do so -- as happy as I'll be 
when I can finally install a fully approved and fully open Java, and when 
I can purchase an ATI card once again well supported by FLOSS drivers.  
Meanwhile, I won't purchase cards that won't work well with drivers I can 
actually install to run them (EULAs may be legal here and if I can't agree 
to them, I can't legally run the software they cover), I can't install a 
partially proprietaryware Java, and I can't fund an LWN running on code it 
has continually pledged to release for years, yet has failed to do so.  If 
others find themselves free to do so, I wish them well, as I really do 
enjoy LWN and would hate to see it die, but funding it with my 
subscription money is something I simply cannot do under the current 
circumstances, however I might otherwise love to do it.  Hopefully someday 
the issue will go away and I can once again fund with my own subscription 
a beacon of the FLOSS message, made even more powerful a beacon as it 
finally walks its own talk.)


boosting revenue with coupons?

Posted Feb 7, 2008 16:16 UTC (Thu) by bronson (subscriber, #4806) [Link]

I applaud you Duncan for taking this stand.  I hope our editor is aware.

Gosh, the site code has been promised since at least 2002...  That's some mighty vapor.

19 Nov 2002:
All versions:*/

It sounds like Jon needs a deadline.  :)  Come June 2008, if the LWN code isn't released,
guess I'll cancel my subscription.  And, as soon as code is available, I shall re-susbscribe
with relief.

Jon, don't worry about supporting it...  An anonymous code drop is fine; support can grow
organically if desired, or can remain nonexistant if not.  I agree that potential security
holes need to be plugged first!  Hopefully the white hats works faster than the black hats.

LWN sources

Posted Feb 9, 2008 14:07 UTC (Sat) by Duncan (guest, #6647) [Link]

Jon's aware, as he wrote me a nice note after I inquired as to the state 
of the code release as my subscription was about to expire.  He had in 
fact made some progress toward that end not long before, I forget the 
details, but he explained  (and I hope I'm not taking liberties with the 
paraphrase) that security audits and the like cost money, and the code had 
been back-burnered as LWN had had to worry about just staying alive, 
implementing the subscription thing, etc.  He also wanted to be able to 
support it and the like, which is nice, but as with you, my feeling is get 
the code out there and if there's interest, support from somewhere will 
come.  If not, well, at least the code is out there for anyone interested, 
something that can't be said about it now.  It's also worth noting that 
should LWN go under ($DEITY forbid!), if the code's already out there, it 
then has a chance to live on.  If it's not out there, well, it never had 
the chance.  Think about all those failed companies over the years and 
what might be if they'd all made their source public, so at least others 
could build on it, or learn about the failed efforts before duplicating 
them.  That's one of the bonuses of FLOSS that LWN would currently be on 
the wrong side of, if it did go under.

I've absolutely no doubt that Jon/LWN does intend to release it, but like 
all those things we all have stacked up waiting on "round tuits", many 
wait years, and others never happen.  That's certainly understandable on a 
human level, but understanding it doesn't get people closer to having the 
actual code available to work with, and well... when ATI and all the 
others have been condemned all these years, and some of them have turned 
over a new leaf and are cooperating with their respective area FLOSS 
developers now, but LWN, part of the FLOSS community, continues to 
promise... it's just embarrassing!

Anyway, I had asked if there was somewhere I could send donations 
earmarked specifically for audit or whatever else it took to open the 
code, but that really didn't get anywhere, and with LWN having the CC 
processing trouble with donations history it does, and what might be only 
a single person's small donation, I understand the reluctance.  
Realistically, my donation alone would be hardly more than symbolic in any 
case.  If people knew my budget... but I still like to do what I can.  
Even if it took writing an old fashioned check instead of a CC 
transaction, I'd do it if there was a place to send it, symbolic tho the 
act in itself might be.

I should also mention that I'm NOT a web developer or anything, either, so 
don't anticipate ever having a personal use for the code, except perhaps 
as a user of any sites other than LWN that adopt it.  No, for me, it's the 
principle of the thing.  I can't do a check or CC transaction with LWN 
under the current circumstances any more than I could to MS, for what 
amounts to the same reasons.

Anyway, if you do a deadline thing, be prepared to live without a 
subscription for awhile, as I have had to do.  The feeling is a bit weird, 
as besides reading late, while one can still post on the weekly edition 
and special features, it's a week late, and most of the action has already 
happened.  One ends up feeling much like those "observers without a vote" 
in various forums must feel much of the time, especially if they are 
additionally restricted to not seeing the work in progress, only the done 
deal.  Yes, the subscription puts one much more in the "know" AND "do", as 
one can watch (and affect) events as the actually unfold instead of after 
the fact (and an LWN subscription affords that opportunity, even for those 
who don't happen to have coding skills and would otherwise be just another 
in the thousands or millions of users).  I'd absolutely have a 
subscription if I could, and I really do miss it, but there comes a point 
when one simply cannot act out of line with one's own principles, no 
matter how convenient it might be, and how much one otherwise wishes to 
contribute to and participate in the (nominal) message, and I simply 
reached that point.

Anyway, it might be worth writing a simple note to Jon mentioning your 
intent, before both you and he are up against a deadline that there's 
simply no way to meet.  Maybe the two of us will get lucky and 
something /will/ happen. =8^)


Source code release

Posted Feb 9, 2008 16:11 UTC (Sat) by corbet (editor, #1) [Link]

If the value of LWN is in a release of a bunch of special-purpose site code rather than in what we write, then perhaps you are indeed better off canceling your subscription.

It's still on the list. I just have never been able to justify spending a week or three on that rather than on trying to keep LWN alive.

One idea that crossed my mind too late last year was to see if we could get a summer of code grant for somebody to do that work. This year I plan to be a bit more on top of things, but can promise nothing.

Source code release

Posted Feb 9, 2008 20:10 UTC (Sat) by bronson (subscriber, #4806) [Link]

The value in LWN is everywhere.  It's in the articles, it's in the comments, it's in editorial
decisions, its in the editors' presence at conferences, and yes, I think it's in the code too.
If it weren't in the code, wouldn't LWN have switched to Drupal or $CMS by now?

My prediction: the 2010 upsurge in Linux adoption will bring 5X as many subscribers, our
editor will be working even harder just trying to keep the servers up and his readers
placated, he will tire quickly when the Yahoogle browser crushes Linux whole in 2015, and he
will retire to a large house in Aspen a rich rich man.  And the LWN code will still not be
released.  :)

Jon, I kind of hope you never find convenient week or three to polish the code the way you
want.  May you always have more meaningful things to work on.  Still you might do what Linus
did: quietly drop it on an FTP server somewhere, unpolished and full of warts.  Or, at least
reword the FAQ to reflect current realities?

Thanks for LWN, btw.  It's the single most fascinating news site on the internet.

Source code audit?

Posted Feb 28, 2008 23:48 UTC (Thu) by Max.Hyre (guest, #1054) [Link]

Should you consider signing on a few volunteers to go through the code, checking for vulnerabilities? I'm willing, have a couple decades of programming experience, and I bet I'm not alone. It would be worth it to me simply to get a few more subscriptions for LWN. :-/

Of course it couldn't be allowed to add to your workload or your worry level.

As for myself, I will take a back seat to no one¹ as a believer in Free Software, and can't see how it conflicts with being a subscriber. Besides, the code isn't being distributed, so even if it were GPLv2, there'd be no breach of license.

¹Well, there is RMS. :-)

Ten-year timeline part 4: the end and the beginning

Posted Feb 13, 2008 22:14 UTC (Wed) by dyork (guest, #2819) [Link]


Wow!  This series of articles is definitely a trip through the past! So many names of people,
companies and events that have now faded into seemingly-distant memory.  I remember starting
to read LWN back in your early days and continuing to read LWN (and advocate to others that
they should read LWN) all through the years of building LPI and moving on to e-smith and other
ventures. Thank you for writing this series of articles.

And thank you for continuing to produce LWN despite all the challenges.  You still are the
best source to me for providing *context* to the stories that are going on out there.
Whenever I want to know what is happening within the world of Linux and open source, LWN is the
first site I head toward.


Copyright © 2008, Eklektix, Inc.
This article may be redistributed under the terms of the Creative Commons CC BY-SA 4.0 license
Comments and public postings are copyrighted by their creators.
Linux is a registered trademark of Linus Torvalds