The downside of proprietary data

Mark Pilgrim is a published author, Google employee and long-time Apple Macintosh user and programmer. In the Macintosh universe, he's part of the pantheon: although never an Apple employee, and not quite up there with folks like Andy Hertzfeld, he's nevertheless one of the minor demi-gods of Apple mythology. He also helped create one of the few Mac viruses (the MBDF-A), but co-operated with police on his arrest and paid restitution for the damage done.

Putting aside his checked past, Pilgrim was considered one of the Mac power-user evangelists, so it came an unpleasant shock to the Mac community when he finally discarded his Mac in favour of Linux. There were tears and predictions of doom. Those predictions turned out to be wrong, and Pilgrim is predicting that 2008 will be the year of Linux on the Desktop. (With the sudden expansion of notebooks running Linux, like the EEE, I think those predictions will finally be right. And not before time.)

Not long after jumping ship to Linux, Pilgrim discussed his experiences with long-term data storage, and his frustration with the difficulty of keeping data accessible over a time frame measured in decades instead of months or years. The bottom line? Long-term storage of data is like a series of migrations from data format to data format. Anything which makes that migration harder is going to hurt you. Companies like Apple who don't grok openness are constantly trying to lock people into their products, then change the products. Every time they do that, there's pain and inconvenience for users, and usually the loss of data.

Pilgrim's conclusion is that using open source software and, more importantly, open formats, goes a long way to reducing this problem. You will still need to migrate data from computer to computer (anyone think that the computers of 2028 will still be running Windows Vista?) but the pain will be less.

There’s an important lesson in here somewhere. Long-term data preservation is like long-term backup: a series of short-term formats, punctuated by a series of migrations. But migrating between data formats is not like copying raw data from one medium to another. [...] But converting data into a different format is much trickier, and there’s the potential of data loss or data degradation at every turn.

Fidelity is not a binary thing. Data can gradually degrade with each conversion until you’re left with crap. People think this only affects the analog world, like copying cassette tapes for several generations. But I think digital preservation is actually much harder, in part because people don’t even realize that it has the same issues.

[...]

So if you care about long-term data preservation, your #1 goal should be to reduce the number of times you convert your data from one format to another. You should also strive to increase the fidelity of each conversion, but you may not have any control over that when the time comes. Plus, you may not know in advance how faithful the conversion will be, so planning ahead to reduce the number of conversions is a better bet.

Open source software is not a panacea for this sort of data loss: as Pilgrim discusses, the open source photo-editing software Gimp uses a deliberately undocumented file format that no other application can fully read.

If you care about accessing your data in ten years time, then go read the rest of his conclusions. (And if you care about people accessing your data in 200 years time, print it out on good acid-free paper and deposit it somewhere dry and safe.)

More nonsense about Open Source vulnerabilities

Computer World is claiming that Red Hat Linux and Firefox are "more buggy" than Microsoft Windows.

That at least is the conclusion you are supposed to draw from the article's title, the summary and the opening paragraph:

Windows not that bad after all
By Matthew Broersma, Techworld

Secunia has found that the number of security bugs in the open source Red Hat Linux operating system and Firefox browsers far outstripped comparable products from Microsoft last year.

So they say. But if you read on to midway down the second page of the article, you get a very different picture:

Red Hat [Linux] was found to have by far the most vulnerabilities, at 633, with 99 percent found in third-party components. ...

Windows had only 123 bugs reported, but 96 percent of those were found in the operating system itself.

So let's see how that works. Red Hat Linux, which ships with multiple hundreds of third party applications, almost all of which are non-critical and don't even get installed, has about six vulnerabilities in the operating system. Windows, which ships with a handful of applications, has about 118 vulnerabilities in the OS. According to Computer World, an OS with six vulnerabilities is more buggy than one with 118 vulnerabilities.

Yeah, right. Sure it is. Just how much advertising does Microsoft do with Computer World?

The article goes on:

In the browser field, Firefox led the way with 64 bugs, compared to 43 for Internet Explorer, and 14 each for Opera and Safari.

However, in an examination of zero-day flaws - reported by third parties before a patch was available - Secunia found that Firefox tended to get more patches, sooner, compared to IE.

Out of eight zero-day bugs reported for Firefox in 2007, five have been patched, three of those in just over a week. Out of 10 zero-day IE bugs, only three were patched and the shortest patch time was 85 days.

You got that? The shortest time IE was vulnerable to known security bugs was nearly three months, compared to just over a week for Firefox.

But IE only looks as good as it does because ActiveX bugs are counted separately: IE had no fewer than 339 ActiveX bugs in 2007. If you include them in the count for IE, as you should, then you're comparing 382 for IE versus 64 for Firefox.

You almost -- almost -- have to admire the journalist's gall in trying to push a whopper of this size. Sadly, this sort of behaviour is very common: half-truths and deceptive statements in paragraph one, the actual facts buried deep in the article. That way you're not lying, because all the facts are there.

The people doing this know that there is a strong correlation between the number of readers and how close to the top of the article: for each extra paragraph you bury something under, you reduce the number of readers by a surprisingly large percentage.

I've written about the tendency of the IT press and security industry to make misleading if not dishonest comparisons between Linux and Windows before.

Would you invest in this company?

Anti-virus computer software company MacAfee has recently warned investors that they enter into legally-binding agreements without understanding those agreements.

Naturally McAfee didn't quite put it that way. Instead, they warned investors that their ability to "commercialise products" based on open-source software -- that is, software which other people have created and published under an open-source licence -- might be at risk if they are forced to obey the licence, since they're not sure what their obligations will be.

Well, here's a thought. Maybe they could read the licence and find out? There are poor quality licences that are ambiguous and incoherent, but the GPL is not one of them. And despite McAfee's wishful thinking that the GPL has never been tested in court, it has, successfully. The reason the GPL rarely makes it to court is because the infringers generally settle out of court and promise to obey the GPL.

It really shouldn't be that difficult for McAfee and others like them. If you want to use open source software in your products, then you have to follow the rules in the licence you have to use that software. That's no different from closed-source licences you might get from Microsoft or any other software company. If McAfee warned investors that they might not be able to use Microsoft's software in their products without obeying the agreement they have with Microsoft, everybody would laugh at them. But put open source in there, and suddenly folks are bewildered -- do they have to obey this agreement or not?

I'm not the only one wandering if McAfee's comments indicate that they are infringing the GPL. Zdnet's Dana Blankenhorn also suggests they're asking to be sued.

McAfee also raises the specter of open source software infringing other copyrights or patents. Naturally it is difficult to tell whether open source software infringes, but that's because recognising infringement is very difficult. Every sizable software project will invariably infringe patents, because the software patent system is seriously broken. Whether software infringes patents has nothing to do with whether you can see the source code or not. It isn't even a barrier to whether patent holders will find out about the infringement. If McAfee were honest, they'd warn their investors that to the extent they licence closed-source software from third parties, they are at greater risk because they have less ability to recognise patent infringement.

Open Source software in education

A fascinating article about the increasing use of Open Source software in education, including a nice list of useful Open Source applications that can be useful for schools.

The article also includes this quote, from Mark Driver, research vice president at IT industry analyst firm Gartner:

The community of a large open source project is better than the knowledge base of any commercial vendor, including the world's largest vendors like Microsoft and IBM. Because of the number of people who work on these projects, the sheer power of the knowledge base is unrivaled.

Goodbye XP

After months of planning procrastination, Mrs. Impala's long-suffering Windows XP computer has been cured of atherosclerosis and Alzheimer's disease by an upgrade to Kubuntu.

I've got more experience with Red Hat and the Fedora Core series of distributions, and last time I played with vanilla Ubuntu, I was seriously unimpressed. But Kubuntu seems pretty impressive, and KDE doesn't dumb everything down like Gnome seems to do. And WINE installed flawlessly the first time, unlike my experiences under Fedora Core 5.

And naturally, there was no product activation and we didn't need to register the software. I can change hardware in the PC as often as I like with the operating system deciding that it has been installed on a different PC and refusing to run.

I'll report back after Mrs. Impala and I have had a chance to give it a solid workout.

Update Monday, 13/3/07: Seems I'm not the only one ditching Windows XP for Ubuntu (with or without the K). So is the French Parliament, which is purchasing 1,154 new PCs running Ubuntu.

The market for free software

Software developers are sometimes concerned that open source software will lead to the collapse of the market for software. If everyone gives software away, how can the programmers make a living?

The usual assumption behind these fears is that software is sold. The image these developers often have in their mind is that of proprietary, shrink-wrapped software like Microsoft Office: the end-user pays money for a copy.

But in reality, software for sale in that fashion is only one small part of the software marketplace, and a very small part at that. The fears of developers seems to be mostly unfounded. IP-Watch has details about the size of the software industry:

In the software market, by far the most money is made in services and the development of tailor-made software. In the EU and in the US, under one fifth of software investment is in (proprietary) packaged software; the rest is in custom software and in-house software.

In terms of jobs, firms selling proprietary packaged software account for well below 10 percent of employment of software developers in the US. Custom software developers and service providers account for about a third. But the majority of programmers work for "user" organizations such as banks, the retail and manufacturing sectors and government.

[...]

Most organizations - and a vast majority of programmers - make money selling their time spent writing or supporting software, but not selling the software itself. This is in fact the economic model of free software: sell potentially everything other than the software itself. The report shows that it is the proprietary software industry that is an anomaly in today's software market, with which the economics of free software is more in tune.

More information can be found in a recent report by the European Commission, the FLOSSIMPACT report.

The EU found that free and open source software is already a huge part of the IT landscape: to replace the free software in common use in the IT industry would need 160,000 person-years costing 12 billion euro or more. IT firms have invested an estimated 1.2 billion euro to develop free software. For every euro spent on open source by industry, open source developers have effectively donated nine dollars worth in their own time and effort. That's a huge return on investment.

The Internet is a great example of how this works in practice. Most of the technologies used in the Internet were built, not by corporations or even governments, but by software developers and programmers who had an itch to scratch. They built the tools for free, not necessarily because of altruism but because they needed the tools for their own use, and getting the tool was more important than waiting for some Good Samaritan to offer to pay them for it. By giving the tools away for free, they encouraged other people to use the same technology. This is an especially powerful factor in the rise of the Internet. Networks like the Internet are useless if only one or a few people use it: the real value from the Internet comes from network effects of many users. Fragmenting the network into mutually incompatible pieces is a losing strategy, and anything which discourages more people from adding themselves to the network is going to keep it fragmented.

While the free and open source status of the Internet encouraged connectivity, commercial IT companies were building fragmented, mutually incompatible, proprietary networks, such as CompuServe, GEnie, Prodigy, and AOL. Of these, only AOL has survived, and only by reinventing itself as an Internet Service Provider. Similarly, both Microsoft and Apple tried to create their own networks, MSN and eWorld. eWorld disappeared after only a few years, and the "MicroSoft Network" became a mere part of the Internet.

While all these proprietary, expensive networks were collapsing, the Internet, built cheaply from free software, was going from strength to strength. Today, companies like Microsoft and Apple who invested nothing in the creation of the Internet, have made the free software behind the Internet into critical parts of their business plan. Outside of the IT industry, the Internet is now essential to companies in hundreds of industries.

Wikipedia 1, China 0

China has demanded that Microsoft, Yahoo and Google all censor their web applications or be banned. All three software giants -- reluctantly in the case of Google, seemingly willingly in the other two cases -- complied.

China also demanded that Wikipedia censor itself for Chinese viewers. Wikipedia refused, and the powers-that-be in Beijing responded by blocking access to Wikipedia from within China.

Or rather, they tried to block access. But it seems that China needs Wikipedia more than Wikipedia needs China -- after barely a month, China has stopped blocking Wikipedia, instead concentrating on the cat-and-mouse game of trying to block only certain sensitive (that is, embarassing to the political leaders) pages, like those on Tiananmen Square.

China is desperate to catch up to the West, and that means accessing our knowledge banks, especially the Internet. If China can't afford to block Wikipedia, they certainly can't afford to block Google, Microsoft or Yahoo -- let alone all three. China is bluffing with a pair of twos, and unlike Wikipedia, the three software giants didn't have the cajones to call their bluff.

GPL wins another court battle

So much for all those people who insist that the GPL has never been validated in court. The GPL has just been upheld again, in Frankfurt, where D-Link got slapped for breaking copyright law by failing to live up to the GPL.

On September 6, 2006 the district court issued its judgement, confirming the claims by gpl-violations.org, specifically its rights on the subject-matter source code, the violation of the GNU GPL by D-Link, the validity of the GPL under German law, and D-Links obligation to reimburse gpl-violations.org for legal expenses, test purchase and cost of re-engineering.

More details here.

Video editing badness

The Linux video editing software Kino doesn't support AVI files natively, it works with camcorder DV files. However, it will import AVI files and convert them into DV format. I have had problems with Kino importing a 1GB AVI file.

I'm not specifically upset that after 15 minutes of processing, the temporary .DV file it created had expanded to 6GB. These things happen -- some data formats are bigger than others. I'm not even upset that processing hadn't finished -- some things take time.

But it is absolutely unforgivable that after not just clicking the Cancel button, but having quit the Kino application, the data import was still churning away in the background. What sort of jerry-built, buggy piece of crap software leaves processes running after you've not just explicitly said "Stop that!" but even quit the application?

I miss the days when Linux programmers actually had a clue. If your application launches a thread to run a job, and the user says cancel the job, CANCEL THE JOB. You don't need an IQ of 168 to know that.

As it was, I was lucky that I knew what was happening. I had launched Kino from the command line, instead of from a menu command, and the status messages were flying thick and fast in the CL window. After much to-ing and fro-ing, from process manager to command line and back again, I eventually discovered that the process in question was the ffmpeg library, and was able to stop it.

The lesson from this is not that Linux isn't ready for use on desktop PCs: this sort of behaviour is no different from what goes on under Windows, except under Windows it is even harder to track down the rogue process. I've resorted to reboots under Windows to stop software running. The lesson is that a colourful animated user interface does not make quality software. I wish application developers would spend more time on getting the basics right.

I'm shocked

You can knock me over with a feather.

As a Linux and former Macintosh user, I'm quite used to being sent proprietary, Windows-only file formats that can't be read except by specific, commercial software -- although it must be said that over the last few years, Linux software has become very good at coping with all sorts of secret file formats. It's been a while I've come across a file I wasn't able to open under Linux.

And then the other day, I received an email with an attached .mht file, and neither Kmail, Firefox, Konqueror or Mozilla seemed able to deal with it correctly.

That's not the shocking thing. The shocking thing is that .mht files are a standard, open file format, with a RFC from 1999 specifying the format: HTML plus external resources such as images, in a MIME encoding. It is simply a MHTML file. Essentially, it is a web page, plus all its images, sounds or other extras, in a single file.

Internet Explorer has supported MHTML in the form of .mht files for years; Opera has recently added support for it. Konquorer does something similar, except it puts the files in a compressed tar ball (.tar.gz or .tgz) and calls it a .war (Web ARchive) file.

As far as I can tell, this is a case where Microsoft has actually done the right thing, using a standard, open file format, and the Linux world is lagging behind. Shocking, but true.

France sticks it the public and Open Source

The French Constitutional Court has just given the public and Open Source developers a fist to the face with their decision about the French DADVSI legislation.

The decision effectively blows away principles of fair use, criminalises even private copying of legally purchased works, and makes it virtually impossible for developers of software to go into competition against any software company that uses Digital Restrictions Management software.

French consumers already pay a levy on recordable media for the purpose of compensating copyright holders for copying.

The law effectively turns the French police and courts into private enforcers for DRM software producers, and allows corporations like Vivendi-Universal and Apple to use the threat of criminal charges against competitors. It makes French criminal law reinforce anti-competitive behaviour. It is especially a threat to Open Source developers.

It effectively outlaws peer-to-peer software, even if used for legal purposes, merely because it could be used to transmit copyrighted works. Arguably, the wording of the legislation would equally outlaw email or Internet or any other method of transfering files.

Worst of all, inexcusably after the Sony rootkit debacle, the legislation makes no provision for allowing the removal or bypassing of DRM even in the event it behaves as malware or threatens lives.

Microsoft loves Open Source now

Microsoft was virtually the last major IT company to discover the Internet (despite what Bill Gates would tell you in the second edition of his autobiography). Now, after years of describing Open Source as a "cancer", Microsoft has discovered that they and Open Source are bestest buddies.

Director of business development, intellectual property and licensing at Microsoft, David Kaefer, said open source had bolstered innovation in a distributed fashion.

He called the open source software movement a very powerful force in the industry.

"I think one of the exciting things about the open source software movement is it actually brought together a very distributed group of developers," he said, speaking at Business of Innovation,a Valley Speakers Series event held at Microsoft's Silicon Valley offices.

In fairness, some of the Shared Source licences are quite reasonable, and some even qualify as truly open source, but with Microsoft you always have to be on look out for the bait-and-switch. Just because a Shared Source licence is described by Microsoft as "open" doesn't mean it is.

25% of web users now using Firefox

Recent statistics from W3Counter show that Firefox's marketshare has now reached 25%. The top five distinct browsers are:

1. Internet Explorer 6.0 with 66%


2. Firefox 1.5 with 21%


3. Firefox 1.0 with 4%


4. Internet Explorer 7.0 with 1%


5. Safari 2.0 with 1%

The statistics are based on over one million distinct visits to more than one thousand different websites.

What I find amusing is that Netscape Navigator has completely fallen off the radar, with an insignificant marketshare, and yet corporate and government websites (always assuming they even acknowledge browers other than IE) continue to act as if Netscape was still important.

This also shows than any web designer who codes purely for IE is effectively treating one third of all users as unimportant. Maybe if IE held a 98% marketshare could that attitude be defended, but it certainly can't be now.

One last factor that should be considered... many web users set their browsers to identify itself as Internet Explorer. The main reason for doing this is to work around obnoxious websites that would work perfectly with multiple browsers, but explicitly check for non-IE browsers and kick them off the site. (Your browser is not recognised. Minimum system requirements are Internet Explorer... sort of nonsense.) So that figure of 25% for Firefox should be considered a minimum.

Link to the W3Counter report here.

The Baen Free Library

A few days ago, I wrote about the death of science fiction editor Jim Baen.

This may be an appropriate time to mention the Baen Free Library, started when author Eric Flint decided to put his money where his mouth was and offer one of his books as a free, unencrypted, unprotected download:

There was a school of thought, which seemed to be picking up steam, that the way to handle the problem was with handcuffs and brass knucks. Enforcement! Regulation! New regulations! Tighter regulations! All out for the campaign against piracy! No quarter! Build more prisons! Harsher sentences!

Alles in ordnung!

I, ah, disagreed. Rather vociferously and belligerently, in fact. And I can be a vociferous and belligerent fellow.

Flint, like Baen, didn't do this because he is a tree-hugging hippy. Their political philosophies are almost polar opposites: Flint has been described as a socialist, Baen as a conservative, but both agreed on the importance of individual liberty and the profit motive. Flint wrote:

And it is that word of mouth, percolating through the reading public down a million little channels, which is what really puts the food on an author's table. Don't let anyone ever tell you otherwise.

Think about it. How many people lend a book to a friend with the words: "You ought a read this! It's really terrible!"
[...]
And, just as important — perhaps most important of all — free books are the way an audience is built in the first place.

Even if you're not a fan of science fiction and fantasy, check out Flint's updates on the progress of the free library.

Office finds a place in the Creative Commons

C|Net News is reporting that Microsoft has teamed up with the non-profit Creative Commons organisation to provide a free tool that will let people attach a Creative Commons copyright license to their Office documents.

Report.
Image only.

There are many reasons why reliance on closed, proprietary, secret file formats like Microsoft's .doc is a bad idea (e.g. vendor lock-in, sudden obsolesence at Microsoft's whim, etc.), but the reality is that .doc is a de facto "standard". (It isn't really a standard as such -- have you ever tried opening your old Word version 1.0 documents?) At least for the short term future, people are going to continue using Office, and this has the potential to introduce, and give Microsoft's Seal of Approval to, the Creative Commons to millions of people who otherwise would think that anything less that All Rights Reserved is an open invitation for the communists to take over and start eating babies.

Flickr gives full access to competitors

BoingBoing is reporting that Flickr's co-founder, Steward Butterfield, has promised that Flickr will allow any of their commercial competitors full access to their programming interfaces to help customers switch away from Flickr -- but only if the competitor has to offer the same deal.

This is fantastic news for openess and competition. Flickr promises that copyright to all photos placed on their website by users remains with the user (assuming the user owned the copyright in the first place). That promise that Flickr won't claim ownership of your digital content is meaningless if there is no easy way to move your content to another service.

(Imagine that you put your property in storage, but then discovered that you couldn't get it back out of storage. There is a word for that -- theft.)

Flickr's move will increase the pressure on other commercial content-hosting services to make good their promise to customers that the host won't snatch possession of their content from them. And because it will be a reciprical relationship, open competition will be encouraged, allowing the marketplace to choose the best services.

Why your business will be using Open Source

Bernard Golden writes for CIO Magazine:

The first reason you need to begin using open source software is that IT budgets suffer from two simultaneous imperatives: low-growth and increasing demand. One important way to respond to these imperatives is to lower your cost of delivering technology. Open source can be an enormous help in this response.

The second reason for using open source is that the software industry itself is undergoing change and will increasingly resemble the open source model: the software itself being freely downloadable, but with far fewer ancillary services delivered by the vendor for free. [...]

Now, let's discuss the third reason you need to jump on open source -- and this one extends the impact past the confines of the IT organization: open source can offer competitive advantage to the overall organization -- in other words, open source can help businesses perform better financially.

Golden goes on to describe the economic forces that make Open Source a huge competitive advantage, and warns that businesses ignore it at their peril:

The counter-argument to the cost advantage of open source software is that the cost of licenses is unimportant in the overall budget of IT. [...]

This reminds me of the dismissivenss that Detroit used to evince toward Japanese automakers. It used to take US car manufacturers two weeks to perform annual model tooling changeovers. Toyota figured out how to do it in less than a day. So what, was the attitude of US automakers. The cost of model changeover is peanuts compared to everything else.

But Japanese car companies continued to incrementally improve their cost structure -- quicker changeovers, less inventory via JIT techniques, lower manufacturing cost by creating option bundles. One day the US industry woke up and Japanese makers had an unbeatable cost advantage; thirty years later, the domestic manufacturers are still trying to catch up.

Golden gives figures for the cost of licences: 10% to 20% of the cost or individual project, perhaps 6% of the entire IT budget, and warns that many people (especially Microsoft) argue the cost of licences is insignificent in the grand scheme of things.

Who's drinking that Kool-Aid? 20% of the project budget is insignificant? On what planet?

Failure to take a 6% saving is mathematically equivalent to a 6% increase. (Note: the two figures aren't quite equal -- it is actually equivalent to 6.38%, because the base is smaller. But let's not quibble about a few tenths of a percentage point.) So let's do a thought experiment.

Imagine that your head of IT goes to the company accountant and demands a budget increase of 6% "because it's an insignificent amount, you won't miss it." What reaction do you think he'd get? If your IT department is one Windows 98 PC, the accountant probably won't blink. If it is 3,000 PCs and 100 servers, well, blinking won't be the only reaction.

I'm sure that a good IT manager will be able to justify that 6% increase. Sometimes you really do need to put more money into IT. And sometimes the right place to spend it is on software licences. But to say that cost savings of 6-20% are insignificant is sheer poisoned Kool-Aid drinking insanity.

The Hive Mind is stupid and boring -- why do we love it so?

Jaron Lanier, writing for The Edge, writes about our infatuation with the Hive Mind.

As somebody who believes strongly in the advantages of such examples of collectivism as Wikipedia, Google's page ranking, the Open Source philosophy, and the benefits of free and open competition, Lanier's argument appeared provocative (if not downright ignorant) to me -- at first glance. But after getting past the first part of the article, and actually reading it through to the conclusion, I realised that he wasn't arguing against the wisdom of the crowd, but warning that crowds can easily turn into mobs, which are notoriously stupid. (The intelligence of a mob, it has been said, is that of its stupidist member divided by the number of members.)

There are many examples of collectivist behaviour where the crowd is more intelligent than the smartest individual. It is the basis of capitalism: the free market is often better able to allocate scarce resources than the wisest autocratic ruler, let alone some out-of-touch bureaucrats. It explains why the scientific consensus is so often right, when individual scientists so often have feet of clay and irrational foibles.

But likewise, crowds are not always wise, the free market is not always clever (witness Tuplip Mania, the South Sea Bubble, and the Dot Com Boom), and sometimes it takes an individual genius to shift the scientific consensus. The Hive Mind (or the Borg Collective if you prefer) is dumb precisely because it is made up of many individuals with but a single opinion. Crowds are wise when they contain opinions and free competition. Lanier's article, despite its title and provocative opening, is not so much warning that crowd decisions are stupid as reminding us of that sometimes they can be, and that we, as individuals, need the sense to tell when to listen to the crowd and when to seek out individual experts.