January, 2007


30
Jan 07

performance tip for gallery 2 + search engine bots

If your gallery 2 install is occasionally getting hammered by search engine bots, the culprit is very likely gallery’s slideshow- search engine bots seem to get trapped in it, and hammer it badly- sapping both bandwidth and CPU (since it hits mysql a lot.)

To deal with it, either disable the slideshow module via the control panel, or use robots.txt to block bots from hitting the slideshows.

This information is surprisingly difficult to find, so I post it here for posterity and for Our New Spidering Overlords.

(Ob. developer/development tangent: it is pretty mindboggling that the gallery developers haven’t figured this out and either made the slideshow bot-friendlier or disabled it by default. I shudder to think what this is doing to people- my load was regularly hitting 18-20 and requiring apache restarts until I figured this out; has hardly crossed 0.1 since then. Thanks to the fine folks at rimuhosting for not throwing me off my shared host before I figured out the solution to this problem ;)


29
Jan 07

totally forgot- gigantic thanks (and two links on post-law school hiring)

When I posted about my summer internship, I forgot to call out the one person most responsible for it happening- Chris Blizzard. Chris started talking RH up at GUADEC, saying ‘I have no idea if we have internships for lawyers, but if we do, you should come work for us.’ And he kept pushing it, and I kept listening, and he kept poking around inside RH. Many moons and some emails later, it turned out that RH did have legal internships, and voila- now I’ve got one. (In case you can’t tell, I’m still totally psyched. :) So thanks to everyone who congratulated me, but particularly thanks to Chris for really being the catalyst that made this happen. You rock, dude, even if you are Typhoid Mary.

Relatedly, Sun’s General Counsel has had a very interesting blog for several months, and he wrote yesterday about their legal internship program. He’s got brief but interesting things to say about how what he sees happening in the legal job market, in terms of experience, skills, etc. Worth the minute or two of your time if you’re interested in the career side of law. Matt Asay also wrote a good post recently, focusing on how terrible the law firm experience is. Everyone should read this- learn something about whether ‘thank you’ or $160K/year is more important, even for the stereotypical soulless lawyer :)


22
Jan 07

I want Linux in my pocket, ASAP, KTHXBYE

Unlike some people, I appear not to have been swank enough to score an N800. This probably indicates good judgment on Nokia’s part, given how little I’ve been able to contribute with my 770. :/ The really unfortunate part, in my mind, is that I’m not that disapppointed. I love my 770, but the complete flakiness of getting internet access with it (even here in New York City where the density of wifi is very high) is very frustrating, and means that I rarely take it anywhere anymore. If I could reliably get internet with the 770/800 (read: cell/EDGE) I’d christen it the Greatest Device Ever; it would always be in my pocket and I’d frequently abandon my laptop. But I can’t, so instead most of the time my 770 is a very nice paperweight- and sits with my other paperweights, on my desk. (Easy/transparent calendar/addressbook data sync, like iPhone promises, might be a partial replacement for always on networking, but that doesn’t appear to be happening any time soon either. :(

So… I’ll probably look at getting an openmoko after my spring exams. The spec looks solid if not great (no EDGE?), my motorola has shitty, shitty software that deserves to be ground into the earth, and an always-ish on data connection sounds like a blessing I can’t get in the N800. The software load will probably be determinant here- how flexible will it be? Will it really sync with my laptop’s calendar/addressbook? But at the moment it at least looks promising in a way that the N800′s lack of reliable network and sync does not.

Tangent: it seems odd that maemo, the platform, won’t be used by the openmoko folks- instead it looks like they’ll be duplicating(?) work by having their own maemo-like gtk-derived base classes. Not sure why this is (perhaps the various proprietary bits in maemo?); it would be interesting to find out, but I’m not holding my breath.


21
Jan 07

Red Hat’s mindshare, and my summer internship

[Edited to clarify who was saying what, and a postscript added to further clarify my position re: developer mindshare.]

Silicon Valley Sleuth said Thursday:

“Red Hat may not be perfect, but the vendor has by far the most developer mind share within the open source community.”

James Governor replied Friday (in his del.icio.us links, which he should make more linkable ;), like Stephen does):

“what might that mean? there is no single open source community.”

James is completely correct that there is no single open source community- there are server people, desktop people, solaris/bsd people, ‘commercial’ open source people, ‘enterprise’ people, community people, etc., and each of those is splintered into a thousand different projects and products.

But here is the thing- Red Hat has a huge mindshare in basically all of those areas. They may not be liked; in many spaces they are being aggressively competed against; and in some places they are losing. But however you slice it, people are definitely aware of RH in virtually every open source community. If you’re an enterprise deploying ‘enterprise’ open source, odds are very high you’re using RHEL, though OpenSolaris is charging hard. If you’re doing web hacking, odds are overwhelming you at least evaluated RHEL or Fedora as a server, even if you settled on Debian or something else. If you’re doing open source desktop deployments, the same (though replace Debian with OpenSuSE or Ubuntu). If you’re an open source community of any stripe, odds are good your software has developers and users who use Fedora. (Mozilla, for example, gains the majority of their market share from Windows, but gets a ton of patches from RH [ed.: see comments], mozilla.org runs on RH, and one of the Mozilla board members is a RH employee.) If you’re a BSD, you’re aware of (jealous of?) of Linux’s (and hence RH’s) market share. If you’re an open source company, you’re envious of RH’s cash flow and growth, and you’re studying (and maybe mimicking) their business model.

So there are many open source communities/spaces, and RH is being challenged aggressively in all of them. But RH is one of the few who can legitimately claim to have significant mindshare, either as a player, tools provider, or role model, in just about every one. So I think the original comment that James quoted is originally correct.

What about me?
One of those spaces I haven’t mentioned yet is the open source/corporate legal space. RH’s leadership on patents has been a model other open source companies should follow, their trademark policy is influential (even if I wish it could be done differently), and they get themselves involved in critical work like GPL v3.

Given that important role, I’m really excited to write that I’m going to work as an intern for RH’s general counsel’s office this summer. I’m sure I’ll be doing the legal equivalent of fetching coffee all summer, and at best working tangentially on the really interesting stuff I’ve mentioned above, but still- it is an interesting and influential place, and I’m excited to get three months to peek inside it, and take away lessons that I can apply wherever I work in the future.

[Tangentially, as an open source-concerned intern, I'm not just excited about how Big And Important Red Hat is, I'm also pleased that they seem to get the little details right. For example, their company code of ethics (which I signed yesterday) explicitly says that contributing to open source projects, even if it adversely impacts the company, is not a violation of the conflict of interest rules. This can't perfectly cover the natural tendency not to bite the hand that feeds, but it is a very solid step other open source companies should emulate. There are other nice, pro-open source/pro-community things scattered throughout their non-compete, code of ethics, etc., but that was the one that really jumped out. I'm always nervous when opening up an NDA or job contract- but after reading through this one I was thoroughly pleased. Kudos to RH on getting the little details right- another reason why I'm excited to work there for the summer.]

Postscript: on developer mindshare

After I originally wrote this, James and Stephen argue that “the linux platform with the most developer mindshare, we argue, is clearly Debian.” I have to disagree. I don’t think that has been true for a long time. Debian is clearly respected, and lot of people (including me) believe that Debian’s health and strong stand on Freedom are critical to the long-term health of Free software. But Debian’s long delays between releases and terrible installer (better now, but still primitive when compared to any modern distro) have been costing it users and mindshare for nearly a full decade now. I used to be a Debian bigot- to install Debian on my first laptop, c. 1999, I was very happy to jump through 1,000 hoops. One of the first hoops was to do a full Red Hat install, because even in 1999 Red Hat’s installer was light years ahead of Debian’s, and I couldn’t convince Debian’s installer to recognize critical bits of my hardware. I haven’t given a moment’s thought to installing Debian since late 2001, when I realized I could get a journaling FS in a RH install and it was likely to be literally years before I would get that in a stable Debian release. (I’m pretty sure that the default installer for stable Debian didn’t recognize ext3/XFS until Sarge, in June 2005; coincidentally the same release that included GNOME 2.x (first released in 2002) in stable.)

This doesn’t even consider Debian’s constant internal bickering, which I’m willing to ignore when choosing an OS, but which generally has made their developer community the laughingstock of other large free software projects. Debian’s developer community is the retarded boogeyman the rest of us want to avoid being (and I say that as a GNOME developer, where talk first/develop later has become a scary norm. GNOME still isn’t that bad.) I think that has cost them a lot of mindshare, at least among knowledgeable/’old hand’ free software developers.

So… I love Debian, and think it is critical to the long-term health of the Free Software ecosystem. I hope it always has users, developers, and mindshare. But I haven’t considered installing it in ages, and I think I’m in a large majority there. If even a complete Free Software bigot like myself won’t install it, I can’t believe that it has dominant developer mindshare. I’m pretty sure that position still falls to Red Hat, by virtue of their broad user base, broad commercial uptake, and freedom, ease of use and up-to-dateness for non-commercial (read: community) users.

(Worth noting, by the way, that Ian seems to agree that RH is #1 in mindshare.)


16
Jan 07

1201 rulemaking talk

Am currently attending a talk on the DMCA 1201 rulemaking. I’ve become just too cynical a bastard to really attend these things; the first speaker here is a shill for the content industry (albeit a very bright shill), and I’m having a hard time not laughing out loud at some of his claims. (He’s riffing off the New York Times article I blogged about the other day, claiming that it justifies DRM, despite the entire thrust of the article being about how bad DRM is for users.) So… argh, I had intended to liveblog it, but it just pisses me off too much. Hopefully the other speaker (Jonathan Band, who it sounds like is a skeptic) won’t piss me off quite so much.

[Later: Band's focus was on how DMCA creates antitrust-like problems, specifically tying, and how DMCA interferes with legitimate uses (similar to my rant about why DRM and fair use are fundamentally incompatible.) In discussing tying, he focused on Lexmark and similar cases, staying away from the iPod-market protection discussion. In discussing legitimate use, he focused on the 1201 exemption process for the academic fair use of films (which Bill McGeveran discusses here.) Apparently the MPAA actually had the gall to suggest that film professors should use a video camera to tape DVDs off a big screen. Nothing new for me, though parts of it were probably new to most of the audience.]

[Later, again: one thing that was new to me came out during the last question. Both professors agreed that the procedures established during the 2006 rulemaking will allow for more targeted exceptions, which they think over time will mean more exceptions, despite the opposition of the RIAA/MPAA.]


16
Jan 07

yummmmmm.

Am still practically in food coma from Krissa’s butternut squash and brown rice risotto:

I am very, very spoiled.


14
Jan 07

yummilicious snack from a NY farmer

Easy, easy totally delicious snack: wheat thins, small chunk of cheese, dab of the plum chutney from Beth’s Farm Kitchen. Hint for classmates: you should be able to find Beth’s selling their delicious wares at the Broadway and 114th farmer’s market on Thursday and Saturdays. Everyone else: the magics of the intarwebs can get this yumminess delivered.


14
Jan 07

new research on motivation and money

I haven’t had a chance to read the whole thing yet, but those who were interested in my past post on intrinsic motivation might be interested in this study on the psychology of money, from Science late last year. Apparently even the mere mention of money can make people less helpful- “Reminders of money, relative to nonmoney reminders, led to reduced requests for help and reduced helpfulness toward others.” There is also a related article which gives some context. Note that the survey was performed on undergrads in Minnesota and so probably has significant cultural biases; it would be very interesting to see a cross-cultural replication of the survey methodology.

[Ed.: this is really quite interesting; the paper is more nuanced than the bit I originally came across- the core of it is that reminders of money makes people feel more independent- not just less likely to give help, but also less likely to ask for help. The report ties this into a number of things, including self-image of those who are unemployed. Well worth a read.]


13
Jan 07

nice Times article about apple’s comfy handcuffs

A nice refresher, from the Sunday NYT no less, about how iTunes and DRM hurts perfectly innocent customers, fails to stop piracy, and reduces competition. Bonus stat: emusic implies every music label except the big four now allows at least some distribution of tunes without DRM.


11
Jan 07

a couple quick evening links- hiring, a job, and cool profs


This work by Luis Villa is licensed under a Creative Commons Attribution-ShareAlike 3.0 United States.