Some reactions so far

Jon, don't worry about "not offering enough extra" for higher subscription rates. There are many of us who value the LWN content, and want it to continue (and expand!). If paying more will increase the chance of LWN's longterm survival, that's extra enough. And nobody is forcing anyone to pick the higher level; think of it as PBS or NPR donations: pick the price that correlates with the value you think you're getting.

And on that note: a few days ago, I renewed my subscription. I'm pretty sure that it did not offer the chance to change subscription levels, and as many others have commented, I'd just plain forgotten that there was a higher level. I've fixed that today, but it might be worth adding that option directly on the renewal form. I picked the medium level initially, because I wasn't sure how it would work out, but I'm glad to kick in more now.

I have certainly wished for something like this in the past. I know of a few people who I think would probably see the value of LWN after just a few timely reads of the section they are most interested in.

Even just a form to email a link when that page goes to free access would be helpful, for those articles that are not terribly topical (everything but the security page, really). In fact, I've often thought "Hey, my friend needs to see this, I'll email a link next week". But I usually forget when next week rolls around. Have the server remember it for me, and I'll probably use it often.

Subscription levels: I suppose we could always implement a higher level, though I already feel bad that we don't really have much to offer the project leaders now. Somebody asked about distribution: approximately 18% starving hacker, 76% professional hacker, 7% project leader.

I'm frankly proud to be a "project leader" subscriber. When I chose to subscribe I thought this site's content was easily worth $10 a month, and I was by no means long on cash at the time. If you consider a higher tier of subscriptions, allow me to make these suggestions:

Starving Hacker International (intended for readers in countries where $5 USD per month is still a substantial amount of money): $2.50
Starving Hacker: $5
Professional Hacker: $10
Uber Geek: $20

That gets out the idea that $2.50 levels are geared towards those type of people specifically unable to afford $5 a month because no one in their area makes enough to pay that.

Early on, the starving hacker and project leader levels had almost equal numbers of subscribers, but that has changed over time.

Is that because you've gained starving hackers at a faster clip than project leaders, or have project leader subscriptions fallen? Here's one option to increase your number of project leaders: LWN mailing lists.

LWN mailing lists could be as simple as an e-mail on the most recently published article as they came up. This could even be offered to the $5 professional hackers. $10 memberships would also get access to LWN mailing lists on a variety of topics, like say desktops, servers, clustering, or just one general LWN-discussion mailing list. I suspect that your $10 subscribers are among the best and brightest in the community, and having such a mailing list to ask questions that don't easily fit in other places is a valuable tool.

For at least a couple years now I've been a subscriber to the slackware-OT mailing list, a mailing list that the best and brightest Slackware users of the newsgroup alt.os.linux.slackware have all subscribed. It is undoubtably the best all-around reference for any linux questions I have ever had. Such a mailing list that only LWN $10 subscribers could join would likely be a major boon for them and encourage higher fee subscriptions. You could keep this in everyone's mind by including sumaries of important discussions on the mailing list in the weekly paper just like you do for the LKML.

Some way for subscribers to pass out "read this article free" tokens. That's a new one, we'll have to think on how it would be implemented. Not sure how useful it would be without the ability to post to a mailing list, though.

I have a suggestion to address both these concerns: give each restricted article a link with an embedded token that only LWN could convert back to a subscriber account number. A subscriber could right click, select "Copy Link Location" (or equivalent) and then paste into an email client. Anyone who clicks such a link would get the article, even while the restriction is still in effect. How many times can such a link be clicked? As many as bandwidth and server capacity allow, so mailing lists present no special problem. How many of these could an individual subscriber issue? As many as he or she wants to use.

Why bother limiting these things? Would having a token link accepted by Slashdot really be a bad thing? Sure, one could post such links for every article to some public place. But one bent on mischief could also post the article contents instead. At least when a subscriber abuses link tokens the editors can easily detect it and take disciplinary action such as suspending the privilege for that account. The point of restricting content at all is to give casual readers an incentive to pay. Do they lose this incentive just because one of us occasionally sends links to the most interesting and timely articles?

Actually, getting links from friends who subscribe is a better incentive than access to undifferentiated week old articles. Subscriber links will be sent right away by people in a position to know what that particular reader is likely to find interesting. Doing things this way empowers subscribers to be better evangelists. Considering the torrent of replies every article on the status of LWN recieves many subscribers are eager to do this kind of work. When subscribers have the power to decide which articles should be more widely read the editors can remove themselves from pointless marketing discussions.

Where is the downside? This could be done on an experimental basis. Repealing the feature if subscriber totals begin to decline would be simple. Do you trust us enough to let us try?

More grumpy editor articles. We want more content like that, but those articles are a lot of work.

That's the wrong way to think about it. I am certain that a large fraction of the Lwn-readership send you money not so much because it gives them any tangible benefit as simply because we *like* what you are doing and want you to *keep* doing it.

For me, any advantages, time-limits, one-time-links whatever are completely irrelevant to if I subscribe or not. What is important is that I want to support what you do. I did so by sending you money as a donation even before subscriptions existed, and I'd keep doing so in the future even if *all* advantages for subscribers where removed.

If someone *wants* to send you more money every month (seems to be the case based on a few comments here) then I don't see any reason whatsoever not to allow that. You need the money, and there's people who are happy to give it to you, where's the problem ?

It's not as if those people aren't perfectly aware what they're doing with their money. They're a) increasing the chanse that Lwn will stay around and b) helping people who do an excellent job get paid more in line with what you deserve.

The truly distinctive and valuable readings have been the weekly Kernel pages. Therefore I have contemplated subscribing for years. Years ago I even tried to get our small company to subscribe, but the value of linux was not yet clear to magagers back then. I might try again now, when managers are much more 'linux-prone'. Also, the salaries of developers here in Eastern Europe are growing so that the subscription would be well affordable already.

I think there may be many potential subscribers out there in a similar position due to the growth of linux use and linux-related jobs. And also because of continuing economic growth in quite some areas of the world.

1. Please extend the access restriction to the weekly Kernel pages. It should become at least 2 weeks, but I wouldn't mind more.

2. If making more subscription levels, please keep also the current lowest-fee level present as salaries of developers may well differ an order of magnitude between countries.

3. Don't bother with translations. After all, we non-English developers want to be able to read the cool comments in kernel ourselves, don't we.

It would be interesting to hear little bit about the statistics of guests as well. How many of them are regular readers (log in more or less regularly)? How many of them are there compared to subscribers?

The Grumpy Editor articles are very good ones compared to many of the software comparisons and overviews available on the net. But to be honest, most of the time they just save us just time and hassle to learn these things ourselves. The weekly Kernel pages, in contrast, provide overviews that are available nowhere else and often give insight to things which weren't graspable to a mere mortal developer without guidance. I have enjoyed also very much the Kernel articles written by other authors, the well-known kernel developers. Shortly, every effort put to extend and enhance Kernel pages would make LWN even more distinctive and would bring you more readers and of course more subscribers.

Jon, I would really like to thank you for all that excellent work, and also for the LDD work. In not so distant future I will subscribe, for sure.