OSS and Deaf Developers?

I came across this rant about feature requests in Gnome. I have to confess a sense of siding with the OSS developers on this one.

The author of the editorial doesn’t get into full swing until she quotes the following statement from the Gnome web site:

A feature will be implemented if and only if there is a developer who wants to implement it.

She then extrapolates this to imply that no OSS developer is going to implement a feature request unless she has a personal use for that feature. This reasoning is both myopic and parochial.

The truth is, when you’re working on a product that’s intended for a wide variety of users, individual feature requests tend to be of limited value. That doesn’t mean I’m not interested in hearing your particular feature requests, because that’s not the point. Within the dross of anecdotal feature requests, there’s still the chance that one of them will turn out to be a gem. The truth is, however, that the vast majority of individual feature requests turn out to be bad ideas.

Perhaps an example will clarify the point. A long time ago, a user suggested that we bring back the “Style Area” feature. The feature was, and currently is, a preference in the “View” panel (you set the “width” of the style are). When that width is greater than 0, then Word displays the style of a paragraph to the left of the paragraph. This only works in “Normal” view.

That was the feature request, but it told me nothing about his problem that he was trying to solve. Well, it turns out that he had to review documents to see if they adhered to corporate style guidelines. In order to do this, he needed to be able to see both the style that’s applied to a paragraph and any direct formatting that’s been applied in addition to the style.

Well, in Word 2004, we shipped a feature that solves his high-level problem far more elegantly than the old style area solved it. This is in the “Style” area of the formatting palette. Word lists all the styles, and instances of styles plus various forms of direct formatting, that exist in your document. Each item in the list box is also a drop-down menu that provides the ability to select all instances of that list item’s combination of underlying style plus direct formatting.

With this feature, our user can now solve his particular problem by simply choosing “Select All” in the drop-down menu for combinations that don’t match his company’s style guidelines. He can then correct them all at once. If we’d just given him is direct feature request instead of trying to understand his real problem, he’d still be searching through these documents one paragraph at a time; a task that would have taken him hours to perform on a single document can now be completed in just a few minutes.

So, I think the OSNews editorial is misplaced. The difference between OSS and what we do isn’t in the extent to which we listen to what customers have to say. Rather, the really significant difference is the effort we put into understanding users’ high-level problems. That’s a very costly, and time-consuming effort. It’s not a job that hobbyist programmers, no matter how dedicated, can reasonably hope to accomplish.


Currently playing in iTunes: Lakeside Park by Rush

Comments (43)

  1. Abiword says:

    Call it luck, call it fluke but irrespective of resources some of use are able to take the time to evaluate feedback and try to consider the big picture.

    Quite often I think Open Source gets lucky. Release early and release often makes for a lot of feedback. Better projects make efforts to look at existing solutions and evalate waht works and why. This can provide valuable ideas so long as developers are mature enough to realise competitors have good ideas too and make good use of the idea of embrace and extend. "Managers" (which usually turns out to be anyone on the team who takes the lead and acts with a certain amoutn of maturity) can often get users to explain the real problem they are actually having by asking the right questions. The general caution more importantly the time constraints of developers ensures that features dont get implemented without getting a certian minimun level of review.

    It is worth noting that sometimes Open Source developers are simply good developers who are familiar with what it takes to develop software irrespective of the license it will be released under. Sometimes philosophical disagreements dont make much difference at all.

  2. Rick Schaut says:

    It’s not about luck. It’s not even about competence as developers. It’s about capacity and resources.

  3. bunk says:

    I have a high level requirement that you don’t have to do any work at all to evaluate: I want Word to use the OASIS OpenDocument format as its default format, and a filter for this format to be released for every version back to Office ’97. This will remove the need for everyone I want to communicate with to buy an expensive office suite, I can direct them to a free one instead.

    But as this isn’t really about your *customers* needs, but about securing and extending your monopoly, I expect you to focus on weirdo DRM systems and patenting XML formats instead.

  4. > It’s not about luck. It’s not even about competence as developers. It’s about capacity and resources.

    And luckily, successful open source projects have heaps of all four.

  5. - says:

    "Eugenia doesn’t represents OSS users/developers"

    Lots of gnome developers are people from Sun/Redhat/Novell who get paid to develop what customers want. Gnome developers also don’t care a lot about personal preferences in some places, ej: usability (one of the principles of gnome’s HIG is that usability must not be based in personal preferences but in scientific facts)

  6. Rick Schaut says:

    Bunk: DRM actually fulfills a rather important high level need for a large number of customers, namely keeping their information confidential. Also, the patents on Word’s XML format don’t restrict customers from using that data in whichever way they wish for their own needs–including writing XSLT’s that read/write OASIS OpenDocument format.

    Murray: You do have to have the right kind of resources, and a willingness to use those resources in the right way. A lot of that has to do with the return one gets on using those resources.

  7. Abiword says:

    Rick many people associate Digital Restrictions Management (DRM) with Corrupt Disks and the Entertainment Industry trying to rollback fiar use and the modern day equivalents of taping off the radio or making mix tapes for friends.

    I understand that some people choose to also use the term DRM to refer to document encryption and signing. That is of course very useful functionality but you shouldn’t be too suprised if people think of something a bit different when you describe it as DRM.

  8. Abiword says:

    > It’s about capacity and resources.

    There is a lot of truth in that.

    In case it wasn’t clear my first comment was in reaction to Ricks closing statement:

    > "It’s not a job that hobbyist programmers, no matter how dedicated, can reasonably hope to accomplish."

    It may not be a ‘reasonable hope’ 🙂 but I continue to be amazed what can be accomplished by open source developement. There are all kinds of problems that seem insurmountable but can be broken down into smaller more managable problems.

  9. You’re absolutely right, but where do you draw the line? If you were doing truly human-centered design, you would ask why these documents need to conform to a certain style. And why are they creating the documents in the first place? Why stop designing at the boundaries of Microsoft Word?

    Maybe you’d say that there are just too many processes surrounding Microsoft Word (the forms, the standards, the habits of office workers everywhere) to include them in the scope of your design project. Hundreds of thousands of people use Word in hundreds of thousands of ways, and it’s not the Word team’s responsibility to redesign their business processes. This is true, to an extent.

    But there is an implicit assumption here: that Word is an impartial bystander in these processes, adapting to the context in which it is used. This is not true. I imagine you’ve spent a lot of time observing Office software in use (as I have) and you are very aware of the ways in which people modify their work processes to conform to their software. Word is already designed to create and support certain business processes. It’s just that this has happened implicitly through thousands of assumptions word processor developers have made over twenty years.

    To step back and consider Word’s role in these business processes will take time and money, something Microsoft has more of, and open source developer have less of. However, open source has a silver bullet: OSS developers *are* the customers. They are employees of IBM and HP and Novell, and they are being paid by these companies because the companies are consumers of OSS. As OSS grows, this will continue to happen: McDonalds will hire an AbiWord developer, and that developer will have one foot in the software and one foot in the fast food business. Nike will hire a Gstreamer developer, and that developer will be working on a team to solve Nike’s problems, not Gstreamers.

    Take the user you spoke about in your post. If people were using typewriters to create these documents, his task would be obsolete. Word itself is creating the need for the styles to conform. If you were an employee of his company, you might be more likely to solve his problem where it needs to be solved, wherever in the company that might be. But as it is, you have only one option: turn every problem into a Microsoft Word feature.

    Erik Pukinskis

    <erik at snowedin.net>

    Human-Computer Interaction/Design

    Indiana University, School of Informatics

  10. Bunk says:

    Rick : Somehow I doubt the MS DRM master plan is going work on Office for Mac – and certainly not on current Macs. If it does, its snake oil : you can’t expect people to believe in a cryptography system where attackers are given the algorithm, the keys, *and* the ciphertext? Because all the use cases for Office DRM include ridiculous "you can read this but not print it or copy it" capabilities. This is clearly ludicrous without total control of the hardware and operating system : ie an "Obey Microsoft" switch on the computer. I doubt this will be very popular in Windows land, let alone Mac land. And real tamperproof hardware isn’t going to be cheap…

    And re the xml format: when are you changing the default format to something other than a binary mess?

  11. dave says:

    I take it you didn’t attend Cory Doctorow’s talk at Microsoft on DRM:


    Executive summary: DRM doesn’t work, because if you let someone view or listen to your content, you’ve already let the genie out of the bottle. The best you are going to do is inconvenience the technically naive people who aren’t even trying to rip you of.

    Maybe you should stop expending effort providing your customers with features that not only don’t work, but *can’t* work, even if you think you can make some money from the very real demand for them. That’s the kind of snake oil promises I expect from my spam email, not from MSFT.

  12. Rick Schaut says:

    Erik: Don’t take that single example as representative of the whole gamut of work we do in order to understand users above the level of features.

    Bunk: I’m not aware of any "Master plan" for DRM. As far as Office is concerned, it’s a feature-one that some of our customers have asked for. I have no knowledge of what DRM means outside of Office, and it would be irrelevant to this discussion.

    Dave: I didn’t attend Cory’s talk, but I’ve read the transcript. Cory’s analysis is off the mark when it comes to document-based scenarios within corporations. The DVD scenario that forms the basis of his talk doesn’t even operate with respect to corporate data.

  13. dave says:

    I’m aware of the difference between DVDs and Office documents, and how the social elements Cory talks about do not apply in small scale distributions. However, I can’t see how anything you can do with DRM in Office is going to affect the simple fact that if someone can read it on a screen, they can take a photo with their camera-phone or, if they’re feeling old skool, write it down.

    Am I missing something and/or is MSFT marketing letting the side down in conveying how Office DRM can help customers with "keeping their information confidential" (particularly anything that they can’t achieve by keeping the information private)? Maybe what I perceive as a false sense of security being sold to your users is a result of my own lack of understanding.

  14. Bunk says:


    Its a feature that makes no sense. If it is just signing and encrypting documents, fine. If it includes elements of control *after* decryption, then it can never work. Scenarios "within corporations" are no different than between random people on the internet. You either want control or you don’t : if it is just an advisory " the sender doesn’t want you to print this", then treat it as such rather than peddling snake oil. We should be beyond XOR-ing with Microsoft Barney by now ( google for it).

    DRM in Office cannot work without total machine control. If you have no knowledge of that, how can you state that DRM in Office will actually fulfill a customer goal, rather than curtail a huge number of other customer goals? Eg Having any trust in their computer at all? I certainly won’t trust a DRM’ed computer.

    I think that is very relevant: Your purpose is not to listen to customer goals, it is to find ways to entrench and extend the Office monopoly. Sadly, your particular area of this seems to be to provide evidence that Microsoft does have desktop competitors in the next antitrust trial…

  15. Rick Schaut says:

    Dave and Bunk, I think you completely misaprehend what corporate customers want out of DRM. Sorry, but I’m not going to explain it to you either. I’ll give you a hint, though. Total control over distribution isn’t the point.

  16. Bunk says:

    It must be great to be safe in the knowledge that you can think you won an argument by pointedly not explaining your points and acting smug about it. Not explaining your point of view is something Microsoft does very well: you’ve lived up to your companys reputation by ignoring criticism and assuming you know better. Congratulations, Rick.

  17. dave says:

    Could any other readers shed some light or provide some links on the Office DRM features?

    A quick Google for Office DRM produces the following link, to a MSDN blog, which makes the Office ‘DRM’ sound a bit cumbursome rather than ‘secure’, despite the blog author’s faith that it will all work out in the end. The comments are interesting too.


    Surely there’s more to it than that? Anyone?

  18. Rick Schaut says:

    Bunk, I haven’t ignored the criticism despite the fact that it’s off-base to the original point. I’ve been circumspect about stating the specific scenario where DRM is useful, and gave a rather whopping hint such that anyone who would care to spend a few minutes thinking about it could figure it out. The irony is that the exchange only serves to underscore my original point.

    Dave, you might try going to a large organization, talk to some of our customers, and ask them why they might want to have the feature. You can’t do sound customer research via Google.

  19. "Rather, the really significant difference is the effort we put into understanding users’ high-level problems." Well I guess you don’t count file compatibility, data integrity, focus, quality, usability, or price as important, so I will stick with better products and save money which means anything other than what Microsoft makes. Personally I like Mellel and Pages.

  20. Rick Schaut says:

    Benjamin, just out of curiosity, by what objective standard do you conclude that I don’t care about file compatibility, data integrity, focus, quality, usability or price?

  21. Well, for one thing I agree with the Existentialists in that no one can be objective. What I can say is I have used Word, Excel, Powerpoint, Publisher, WordPerfect, Lotus WordPro, Lotus 123, OpenOffice.org (mainly writer), Pages, Keynote, and Mellel. The last 3 were the best by far, but OpenOffice.org is good for Windows or Linux users. Publisher was ok. The Microsoft programs especially Word and Excel crashed very frequently as well as Windows. I was also very disappointed by the graphical quality and the HTML output of Word. Classmates were unable to open Word documents with products like WordPerfect or lower versions of Word. The price is more than I can afford especially for what it does at the quality you get and how much effort it takes to even get acceptable results. As far as focus, Mellel as well as iWork are examples of great focus – iWork is great with multimedia – how it integrates so well with iLife, the ease at which you can create graphical documents from already designed layouts, the highest quality output in the industry standard PDF format and Mellel is great at academic and technical writing with excellent support for any language, footnotes, bibliography, and all sorts of structured document markup – like LaTeX but without having to mess with code. Another big sign of sloppiness is how much memory Word takes compared to Mellel: Mellel takes 1 MB of RAM, Word takes over 40 MB. Another is how big the files are – if you take an average Word document and open it up with any degree of success with OpenOffice.org (and the only reason why you can at all is because they reverse engineered it) and then save it, it will be 1/3 the size roughly. The fact that Microsoft produces such low quality products drove me towards Linux at first an then to Macintosh when Jaguar came out. When I said you I meant the development effort at Microsoft. You are one of the developers, so I expect that you feel somewhat responsible for what you produce and who you work for.

  22. Bunk says:

    DRM is "control after decryption". This is ridiculous and won’t work without total machine control. If it is advisory, "the sender does not want you to print this" then it isn’t secure and is not really DRM.

    You have not explained any scenario that doesn’t fit into these categories. I imagine that you are thinking of an advisory system that is trivially broken : this is not really DRM, and if you are marketing it as "control your documents", it is snake oil.

    Please answer explicitly rather than snidely this time. You aren’t doing yourself any favours with this attitude.

  23. Rick Schaut says:


    Thank you for the feedback, and I say that in all honesty. I much prefer a more carefully-considered criticism like your second comment to your first comment, which struck me as being mostly snide.

    I’ll take issue with the Mellel-Word memory usage comparison for a couple of reasons. First, I have no idea what number you’re referring to, but Word’s running VSIZE is about 217 MB with an empty document. By comparision, BBEdit’s VSIZE is 153 MB with an empty document. Secondly, a single memory size comparison isn’t all that meaningful. How do they compare, for example, when each has a very large document open? What about the number and size of graphics in the document?

    I’ve yet to find a really objective measure of "quality" that allows one to say that one software product is of higher quality than another. One has to account for usage patterns and feature sets. It turns out that "quality" ends up being a somewhat squishy subject.

    All that said, I’m glad you’ve found products that work for you. It’s perfectly fine with me if people purchase another product than mine because that product works for their needs.


    Sorry, but you’re missing the entire point. Your DRM argument is, well, bunk. Either DRM works, in which case the user benefit is obvious, or DRM doesn’t work in which case you can’t reasonably claim that the purpose is to lock users in to Microsoft’s products. You can’t make that argument work both ways, no matter how hard you try or how often you accuse me of being smug and/or non-responsive.

    You’ve inferred a motive based on a self-defeating argument when, in fact, a far more simple motive exists for why DRM was added to Win Office: that one or more rather large customers told us that they wouldn’t buy our product if DRM wasn’t there.

    Now, I claim that there is a real scenario for which DRM is a viable solution even if it can be by-passed, and I also claim that one can figure out what that scenario is if one actually thinks about the kinds of ways by which information "leaks". Neither of those claims is necessary to counter your rather obviously flawed argument.

    I point to the existence of this scenario not in an effort to be snide or smug. I point it out because the issue is my initial statement about a fundamental difference between a well-funded, proprietary development process and a not-so-well funded, open source development process. The fact that you have yet to tumble upon the specifics of the scenario where DRM actually works serves to underscore my point.

  24. Bunk says:


    I’m come to the conclusion that you are blinded by your pay check. DRM can work but *only* when you have control of the hardware. Of course it can fulfil a customer need. So do race targeted viruses – a lot of customers will pay a lot of money for them. But *in fulfilling* that customer need, both of these things trample upon many much more significant needs and rights.

    If this is just a trumped up version of the passworded locked areas on Excel, which as far as I can tell is what you are saying, it is very hard to see what it can acheive: it is being marketed as doing a lot more than it can. If it is a "word undo history scrubber" it is not DRM, its a bug fix. If it is an advisory "please don’t print/email this", then again, it is being pushed as far more capable than it is.

    Until your claims above, I believed MS was aiming for total machine control (TCPA etc). If not, you is selling snake oil. So there are two scenarios:

    1) You are selling snake oil.

    2) You are going for a total machine control solution, which will horrifically distort the relationship between a machine and its owner. At the same time, some people will like features of it : those people will not include me or anybody who cares about social consequences.

  25. Rick Schaut says:


    First of all, you’re conflating two rather different issues. Your initial argument with respect to DRM was about "securing and extending [a] monopoly." This can’t possibly be the case, as my last comment shows. If DRM can be circumvented, then there is no possible way to use DRM for the purposes of "securing and extending [a] monopoly."

    That only leaves one possibility: that we’ve done DRM in response to customer demands. The mere fact that you don’t see how customers might actually want DRM, even if it can be circumvented by a number of different means, doesn’t change this.

    As for snake oil, the only way we can be selling snake oil is if one assumes that the people who’ve been asking for the feature are unaware of the limitations of software-implemented DRM. Assuming that money-paying customers are ignorant, while quite possibly convenient for pejorative rhetorical purposes, is never a good idea when your livelihood depends on selling your products to them :-).

    Now, do you really want to continue to press an argument based on a set of assumptions that, demonstrably, have no connection to reality? Or, are you going to think about customer scenarios under which a potentially circumventable DRM might still be a desireable feature? In an environment where the most significant deterrent to "leaking" confidential information is losing one’s job and potential civil penalties, how do most "leaks" occur?

  26. Bunk says:


    Separating out two scenarios and asking which one you are talking about is not conflation. Look it up.

    So, you admit that your solution is an advisory system, "the sender does not want this printed"?

    Your marketing needs to make it very clear that this is *purely* advisory. It has practically zero security value against an *attacker*. DRM strippers will be out in short order after release : customer needs work both ways. Will your employer pull an Adobe here?

    And, to utilise your favorite rhetorical device, surely even one so entwined as yourself can come up with a scenario in which dodgy laws like the EUCD & DMCA, plus an encrypted file format, lead to "securing and extending [a] monopoly"? Are you honestly surprised that people don’t trust your employer in this situation?

  27. Rick Schaut says:


    I said that you were conflating issues, not scenarios. That you only admit two scenarios is generally known as the fallacy of the excluded middle. You are, nonetheless, conflating the question of the utility of DRM in Office with the issue of "securing and extending [a] monopoly." That’s a quote of your words.

    The rest of your comments are predicated on this fallacy, and, therefore, don’t merit any kind of rhetorical response.

    And, sadly, I’m not surprised that people don’t trust Microsoft. I wonder how many times I’ll have to repeat the fact that DRM exists in Office because some rather large organizations said they wouldn’t buy the product unless we added it before you’ll just flat out accuse me of lying about it.

    There are those who won’t trust Microsoft no matter what we do. You’re doing an excellent job of convincing me that you belong to that group.

  28. Bunk says:

    I have never denied that some customers have asked for DRM. I have merely said that fulfilling this need necessarily infringes upon the rights of other customers. A moral stance would be to refuse the first request no matter how great the payoff. Will you add political speech monitoring tools to Windows if a big customer asks for it? The fact is that your proposed DRM solution *can* be used for securing and extending a monopoly. Why, given past behaviour, would we not expect that to be exploited?

    And as to when I would trust MS : I will trust that MS will stay within contracts that I have the legal and financial capacity to enforce.

    To be honest, that is not much. I *hope* it will improve, but at the moment, while every MS employee blindly cheerleads every significant that the company makes, how can you expect the situation to improve.

    So far you’ve done a great job of assuming your customers are idiots. Long may it continue.

    I think I’ve wasted enough of your time and mine on this, we aren’t going to agree (unless MS drops DRM due to consumer backlash and you slavishly toe the party line. Could happen, I guess.)

  29. Rick Schaut says:


    Hate groups can use Word to promulgate ideas that lead to violence in various forms. By your reasoning, the fact that Word can be used for this purpose implies that we should stop selling Word altogether.

    As for "political speech monitoring tools," exactly how might such tools differ from SPAM filters?

    You can certainly accuse me of being a sycophant, but one would hope that you’d come up with a far better argument upon which to base that accusation.

  30. Bunk,

    If companies don’t believe in using DRM, they don’t have to use it.

    For those companies to whom DRM is critical to their business practices, Microsoft would be silly not to give them the functionality that they’re wanting.

    Microsoft is in the business of selling platforms. Windows is a platform. Office is a platform. Our customers desire that those platforms have certain features, including DRM (in both Windows (Media) and Office). They have different reasons for asking for the DRM functionality, but they have a legitimate right to ask for DRM functionality.

    Microsoft has two choices. #1: They can say "DRM is evil" so we won’t allow you to do DRM. Or they can say "We’ve got a solution for you".

    Which of the two is likely to result in customers buying our product? And which is likely to result in the customer going to a competitor who doesn’t have the same issues with DRM?

    The customers who want DRM have decided that they have a business NEED for the functionality. They’re going to get that functionality whether you want it or not, the ONLY question is if they remain your customers or if they go and become someone elses customer.

  31. Roy says:

    I really enjoyed this blog until I got to the end. The last paragraph is a cheap shot at OSS. You charactarize OSS developers as hack amateurs (hobbyists). This is both arrogant and foolish. It ignores the fact that many key OSS developers are paid professionals working for OSS based companies, while other are professional software developers working on OSS in their spare time. After spending some time parusing the gnome developer mailing lists, I’ve been impressed by their efforts to understand the high level problems they are trying to solve.

  32. Rick Schaut says:


    No slight was intended. As I’ve mentioned in other comments in this thread, the issue is really a matter of having the resources available to conduct market research. You need an income stream from the software in order to do that, and the only model that provides a sufficient income stream would be one akin to Apple’s model.

  33. What people ask for and what they want are not always the same thing.