Conversation = Excellence
LibraryThing has always depended on members to set development goals and refine (or ditch) features. But it's amazing how well it's worked with the new "affinities"* feature. We simply could not have anticipated how members would shape our thinking. (I will never ever develop another project in a small, closed group, with occasional trips to watch a "focus group" from behind smoked glass.) We're still watching reactions on the blog, and on a now-130+ Talk topic, but we have some good ideas. When Altay returns from Boston, we'll hammer out changes, including customization of the look, and the ability to turn it off.
I started another thread I want to highlight, about LibraryThing's strategy and a hiring decision for the non-English LibraryThings. Do we hire someone, and what can they do? I hoping the thread gets some traction, at least among the users of our dozen-plus non-English sites. We need a non-English plan.
Part of the problem is technical, starting with better character support. But there's a feedback loop. Right now, the non-English sites can't be the coding priority because they're not contributing as much to our growth, or to our finances. (Not that they're small. Our non-English sites appear to have more action than our largest English-language competitor.) If we hired someone—and had something for that person to do—we'd have a stronger incentive to work on it.
*We called them "affinity percentiles," but it got chipped down nicely by SilentInaWay. Case in point.
I started another thread I want to highlight, about LibraryThing's strategy and a hiring decision for the non-English LibraryThings. Do we hire someone, and what can they do? I hoping the thread gets some traction, at least among the users of our dozen-plus non-English sites. We need a non-English plan.
Part of the problem is technical, starting with better character support. But there's a feedback loop. Right now, the non-English sites can't be the coding priority because they're not contributing as much to our growth, or to our finances. (Not that they're small. Our non-English sites appear to have more action than our largest English-language competitor.) If we hired someone—and had something for that person to do—we'd have a stronger incentive to work on it.
*We called them "affinity percentiles," but it got chipped down nicely by SilentInaWay. Case in point.
Labels: conversation, features, non-English
posted by Tim at 4:42 PM
11 Comments:
It's not just a feedback loop; it's a chicken and egg problem. No one can in good conscience recommend the non-ASCII site, for free or for pay, without the expectation that it will work. And it still does not today.
Tim said: "Right now, the non-English sites can't be the coding priority because they're not contributing as much to our growth, or to our finances."
If there's ever been any statement that made me want to run far away from LT and never come back, this is it. I guess we might as well give up on improvements in overall quality, because you've practically said that the core users don't matter at all. Once we've contributed our $25, it's time for you to move on to the next income source.
I understand that that's how business works, but if it goes too far you might run into problems similar to the ones that amazon has, which you described before: we're contributing tagging data that you can then sell to libraries, which are a better source of growth and income. Admittedly we're getting something out of it, but the situation could still easily be perceived in a negative light. And given how you keep stressing the importance of PR, I'd think you would want to avoid the impression that you care only about money and not about your users.
So, you're saying that the 5% of users who want better internationalization should outvote the users who want authors to work better or wishlists?
You can't have goals without choosing. We have to choose things that help the site. There are two ways to do that—get new users or make existing users happy, so they'll use the site. They tend to happen together, so they're basically the same thing.
But we have to focus on the larger problems, the larger opportunities. We have to get wishlists out the door before we add PubMed searching, Hebrew MARC parsing and so forth.
My goal is to make LibraryThing better, more interesting and larger. I manage two engineers--which will fall back to one soon--and one librarian.
I *have* to make choices. I'm trying to build a compelling case to do more on the European side of things, to add a fractional employee which, with sunk costs and the prospect of support, would move internationalization up the priority list. That's good, isn't it?
I'm actually saying just the opposite. It wasn't the non-English part that I objected to, just the fact that the coding priority is growth and finances. The problem I have with that is that people who have already paid for a lifetime membership (whether they speak English or not) aren't an obvious source of growth or finances anymore.
It seemed like you were already focused on internationalization for months, at the expense of the wishlists and collections that everyone wanted. Then the focus shifted to libraries, again at the expense of wishlists (with collections abandoned completely). And now it's back to internationalization again. I understand that internationalization and libraries offer the most opportunity for expansion--but will the long-awaited features like wishlists ever arrive for the people who have already paid, or do those people just not matter anymore? That's what I'm really concerned about. It seems like the 0% of users who don't use the site are outvoting the users who want authors to work better or wishlists.
I don't agree that getting new users and making existing users happy are necessarily the same thing. This clearest example that I can think of is the person who had entered a significant Esperanto library (I think it was Esperanto), but deleted it because search etc. didn't support it. It seems counterproductive to work hard to attract people to a site that doesn't make them happy enough to stay.
I wonder whether a similar thing could happen with academic libraries, when they notice that LC call numbers don't sort in the right order. Presumably it would be fairly simple to fix, especially if someone who was paying thousands of dollars every year complained about it, but it's the kind of thing that would make me have serious doubts about the whole product.
I know that making choices isn't easy, and I'm sorry that I'm angry. It's just extremely frustrating to see the constant emphasis on growth and to feel that the current users are the least important.
Okay, I understand you better now.
I think there's a very valid criticism there. I certainly expressed it that way. But I know that happy users translate to growth better than anything else. It's not like we've been doing real marketing or PR. We have developed new things for the site, like Suggestions, which alienates some users. But the site needs to grow into what it should be. The same case was made for not adding any social features. "I didn't join to socialize" was the call. But those features have grown the site to where it is, and, frankly, many of the anti-social users have ended up socializing a lot.
>I wonder whether a similar thing could happen with academic libraries, when they notice that LC call numbers don't sort in the right order.
Well, we aren't giving libraries that sort of data. It wouldn't make any sense. Nor would they take it. Libraries don't want it. We know that, although libraries want to use LibraryThing as their catalog, it isn't strong enough. Libraries spend millions of dollars to get junk, but it is reliable within its junkiness, and a tech support person on call 24/7 and there's a salesman in your area.
Look, I don't want to whine about how difficult things are, but I do want to lay out why some problems are hard.
LCC sorting is a good example, actually. LT always prioritizes the users' actual copy and their ability to edit the data. This means we can't rely on valid data, but have to keep separate validated forms at the same time, or go through every book creating such forms before we can act on any of the data in a sort or search. Doing that should be part of a rewrite of the whole sort and search structure, not a hack stuck on.
Notable, as much as I want to work it out, LCC sorting is not called for by that many people. And, lest we forget, LT is almost the only site of its kind that even *has* that data, let alone sorts it correctly. Shelfari now does show Deweys, because Amazon added them to their feed. Dewey is also not trivial to sort. Their sort isn't broken, because they *have* no Dewey sort. Indeed, they have only four possible sorts. And you can forget editing Amazon's DDCs.
The same considerations apply to other changes, like the author issue. If LibraryThing, like Shelfari, only used Amazon data, and didn't allow editing, it would be relatively easy to change how authors were handled. There would be limits—the data is ratty, Amazon doesn't disambiguate, and there's not even any concept of first names and last names--but whatever changes could be made, could be rolled out across everyone's books instantly. The choice to keep 14 million records, from 70 sources and with full user editing complicates things. But it also gives us a firmer ground to build on.
I also urge you to appreciate that the character set issues aren't trivial either. Spend some time on the Library of Congress catalog. The LC is effectively the national library. They 4,000 people and one of the top online catalogs out there. And while searching works better, character display is at least as screwed up as LT. And the LC has all of its records in a single format. We draw from 70 libraries and multiple formats. Danish libraires, for example, have their own format, "Danmarc" with their own character set.
Lastly, LTFL is in a special category. The desire there was very strong. And it fits with LibraryThing's basic identity. We use library data. The registered librarians on LT hold almost a million of LT's books. We are accepted into the library community--talks, blogs, etc.--and we are grateful about it. Putting some LT features in a library catalog lines up with that, gives libraries something back for all the interest, support and data they've given us. And, frankly, it may well be a huge financial boon to us. Libraries are not rich, certainly. But the are large institutions with considerable spending on technology. They often spend literally *millions* on their crappy online catalogs. LibraryThing can charge hundreds or low thousands--less than other enhancements--for a library to turn their catalog into something much more exciting. The development work on LTFL is mostly done. The revenue goes to the site generally. If LTFL takes off as I expect, I think it'll allow us to add a number of people. That's something I *have* to go for.
Anyway, we've heard that users are frustrated by the pace of development. We are picking it up, and will be balancing "new" with "old."
Thanks for the response. I do appreciate the fact that a lot of the problems aren't trivial, and I really hope that LTFL is hugely successful and makes a lot of money for the site.
I know LT is much better than Shelfari or any other cataloguing sites, and that's why I care so much about how LT progresses. I'm excited to hear that you'll be picking up the pace of development.
And meanwhile I'm sorry to be so thin-skinned. I do understand there's frustration in various quarters. To some extent, that's just a result of making choices and not having a limitless supply of programmers and money. But to some extent it's bad choices we've made. We heard. I'm not guaranteeing we'll made the right ones, but we're going to be ramping up, so we'll make more of them.
Tim, thanks for the long explanation. As someone who has complained occasionally, I am glad to get a better understanding of your point of view. But I think that before trying to promote the site seriously in the non-English-speaking world, what you really need to do is get a coder - whether staff or consultant - who has experience with software internationalization and can work out how the core functionality can be improved to make it adequate for international use. You've gotten a lot of comments in Talk about what isn't working, which gives you some goals, but that's not the same as someone who can look at and work on the code with these issues in mind.
The basic problem is that too many things throughout the site - not just the character coding issues but also the grammar-insensitive translation system, search only working for the most common version of the work, American-style date strings, onomastic assumptions, etc. etc. - simply aren't going to be very attractive right now to people who want to use the site in other languages, even if the range of libraries for importing the books were good. Getting a system in place that can handle these things would probably reduce problems (and thus costs) in the long run - not just for multilingual use but also because some of it would involve modularizing stuff that is at present probably deeper in the code than it should be regardless of language issues.
I realize it isn't a small task, and I can see how it hasn't happened - it's not the sort of thing that can be done in a day or two between conferences and server problems. Clearly most of the coding time, especially your own given the diverse demands on your time, needs to be devoted to short-term tasks like chasing bugs and doing small features, but the longer that work on the fundamentals is put off, the worse a chore it will be. It's a basic problem for any system that has evolved incrementally; there comes a point when some basic architectural changes are needed to support improved functionality.
So I hope that LTFL does produce a nice revenue stream that you can use to hire someone specifically to do a thorough overhaul of the basics, and I really hope that the person you hire is someone familiar with internationalization issues as well as cataloging problems in general. Until then, I think efforts to expand within the English-speaking world may be more fruitful.
It's not done yet, but see http://www.librarything.com/talktopic.php?topic=12386 for the new search.
Thanks for the explanations it is always pleasing to know that the owner is hearing and responding to us over at talk - which as you say is a sideline from the main purpose. Everyone always wants everything yesterday, explanations and justifications from the top are excellant ways of mitigating our frustrations.
专业的翻译公司
印尼语翻译
阿拉伯语翻译
马来西亚语翻译
食品翻译
技术翻译
化工翻译
在深圳投影仪公司中领先
深圳翻译公司提供
英语翻译
日语翻译
深圳德语翻译
深圳俄语翻译
藏语翻译
韩语翻译
泰语翻译
越南语翻译
维族语翻译
瑞典语翻译
农业翻译
汽车翻译
通信翻译
英译汉，汉译英
汉译日，日译汉
西班牙语翻译
葡萄牙语翻译
能源翻译
乌克兰翻译
医学翻译
外贸翻译
食品翻译
保险翻译
化工翻译
农业翻译
汽车翻译
通信翻译
建筑翻译
物流翻译
工程翻译
深圳翻译公司
深圳翻译
翻译公司
法语翻译
德语翻译
俄语翻译
藏语翻译
韩语翻译
泰语翻译
深圳翻译公司
深圳英语翻译
深圳日语翻译
深圳法语翻译
>菲律宾语翻译
意大利语翻译
希腊语翻译
柬埔寨翻译
土耳其语语翻译
越南语翻译
印度语翻译
地质翻译
合同翻译
法律翻译
机械翻译
标书翻译
瑞典语翻译
印尼语翻译
阿拉伯语翻译
马来西亚语翻译
菲律宾语翻译
深圳意大利语翻译
爱尔兰语翻译
柬埔寨翻译
土耳其语语翻译
西班牙语翻译
葡萄牙语翻译
外语翻译
小语种翻译
广州深圳同声传译|
证件翻译
合同翻译
金融翻译
新疆租车
珠海翻译公司|
佛山翻译公司|
惠州翻译公司
汕头翻译公司
三亚翻译公司
东莞翻译公司
深圳投影仪租赁|
口译|
深圳陪同翻译
的翻译公司欢迎您的访问！
翻译公司|
北京翻译公司|
同声传译|
同传设备出租
会议设备租赁
同声传译
同传设备出租|
法律翻译
机械翻译
标书翻译
医学翻译
外贸翻译
上海翻译公司|
北京翻译公司|
广州翻译公司
Post a Comment
<< Home