What improvements do you want to see? Poll

I'd like to try compiling member views about RSI via a series of polls. Each poll will ask about a different potential feature/improvement; if you want to see that developed, vote yes; if not, vote no.

Please add your own polls too; I'm not going to get to everything myself. Remember that there can only be one poll per message. Edit: To make a poll, put "vote" in html tags around the question.

The usual disclaimer: Tim has said that polls just provide one data point. The polls here will not necessarily have any impact. They do not represent a perfect sampling of LT users, etc. etc.

>14 AnnaClaire: Put "vote" in html tags around the question.

Oh. I thought it was something like that, but had didn't get exactly what.

>18 _Zoe_:
Like what?

>21 AnnaClaire: Like if I had wanted to make one big poll here, where you could choose multiple answers, rather than having a separate poll for each feature.

To clarify, I mean books with a "stopped" date of today or yesterday.

This would show a ranked list of all the books read in the month (based on stopped dates), with cover pictures. The actual number of people who finished the book would be shown too. Ideally previous months' pages would be saved for reference.

So that we can change books from sentence case to title case.

http://www.librarything.com/lttl/

Which would be reflected in places like Recommendations: you would no longer get recommendations for books that you already owned in omnibus form, for example.

So that you could choose to see only books owned by fewer than 500 members, only books with fewer than 5 reviews, and so on.

showing reviews for books that appear on your recommendations list

... so that there is time to finish off the ones that already exist!

This was a tag cloud for a user's books, based on other people's tags, so it wasn't dependent on an individual's tagging practices.

Some of these I vaguely agree with - but only well after all other features have been finished / fixed/ implimented so I've voted no on them.

So that you can select which books you want to see reviews for. No more Pride & Prejudice reviews, for example, unless you want them.

Edit: this is for a more specific version of #31's poll

This is easier to see if you filter by a tag, especially one you don't use much.

>45 keristars:
Um, isn't that the poll in #42?

48> I thought so at first, but when I read it again, it looked like it was asking for a different sort of filter than what I'm wanting.

Edit: or rather, the one Zoe proposed is broad, and doesn't necessarily cover the kind of filter I want, so the additional poll is for the specific case.

I'd say it's a subset of #42, but there's no harm in being more specific.

So that we could look at just those groups tagged "site-related", etc.

Where you could put LT usernames, and then explore books and users by following chains of who recommended interesting books.

Where you could enter Local locations (bookstores or libraries). Especially interesting for books purchased in unusual locations while travelling.

...so that large groups of tags for the same kind of special purpose (like Dewey tags, for example) don't get in your way.

Edited to clarify: Tags would NOT be identically bundled for everyone. You would only bundle your own tags.

So that combined authors aren't always shown separately

>56 _Zoe_:
Not to mention combined books. Or works with no copies.

Since, various checkboxes not withstanding, some things are still just based on All Books.

Exact duplications with no copies are unceremoniously dumped, for starters. And/or zero-copy editions are hidden on a subpage. And works with no copies aren't suggested for combination.

which could be used for things like showing which books you've read on a given list. Once Lists exist.

So that when I enter a starting date, the book automatically becomes part of my Currently Reading collection. When I enter a Stopped date, the book is removed from the Currently Reading collection.

Because these polls don't have a means for showing how important a given topic is for me, I elected to skip polls that are not priorities for me. For those topics / polls, even though I have a preference in most cases, I simply didn't vote.

I realise this is a problematic approach from the standpoint of data analysis, but it makes me feel better about my voting, and I thought I'd state it explicitly as a "full disclosure" measure.

See also, this thread and this tag.

See also, this thread (and also this tag). Specifically, the bilinguals at the end of the first post, and the punctuation stuff in post #4.

>68 elenchus: I've personally been voting Yes for things that I'd be happy to see done right away, Undecided for things that I don't know about or just don't care about enough, and No for things that I'd actually be disappointed to see time spent on now. So basically, if there's ever a case where something is just a low priority, I'll vote Undecided rather than not voting at all.

I think the results should still be relatively useful even with everyone voting slightly differently, though.

Perhaps we could be required to cite a source, such as the front matter of a copy of a book, or the Library of Congress. Further discussion required.

>71 _Zoe_:
That's more or less what I've been doing, too.

Essentially, users could choose certain books to highlight on their profiles: the "living room bookshelf" that people would see when they first stopped by.

Discussed here

> 30

What do you mean by a "unique author page"?

>76 lilithcat:
Splitting identically-named authors onto entirely separate pages, not just sectioning the works.

> 75

Do you mean "a way to mark spoilers in reviews that will be visible to all" or "a way to mark spoilers in reviews that will be visible only to the person who marks them"?

>78 lilithcat: The former; I don't think it would help an individual to mark it as a spoiler for themselves alone, since they'd have to have read the review already!

On a lighter note, I'm almost tempted to put up a poll for "Disclosure of what fertility-increasing additive has been found in the LT Staff water supply."

>80 AnnaClaire: :D

I think more specific bug fixes should also be listed here to give LT staff a better sense of priorities, but I find that I can't actually think of any off the top of my head. Anyone?

I added a specific bug to fix in post #46.

Right, but surely there are more? For now I've changed my general "fix bugs" vote to Undecided, because it seems like that isn't actually a priority after all.

There's a long standing bug tracking thread here: http://www.librarything.com/topic/92113

I haven't been keeping up with it, so I don't know how complete it is, but it would be a place to start.

Thanks, I will start with those. I'm not going to check whether a bug has since been fixed, though, so there may be some non-bugs included here.

Where a deleted Canonical Author makes the author name blank

When using the pink plus, if only a single search result is found

(see post 13 in bug thread for links)

http://www.librarything.com/topic/91861

http://www.librarything.com/topic/93378

http://www.librarything.com/topic/86371

Edited to add: This has been done - thank you Tim!

It's going to be the rare bug that I vote no on fixing it.

97> My understanding is that they do use a bug tracking system internally, but we don't have access to it. Do you mean to suggest a bug tracking system that members can use to enter and track bugs?

This is an obscure, but trivial one. http://www.librarything.com/topic/90658

101:
Yes, because so many bugs are reported in the Bug Collectors group and apparently ignored. They're bumped regularly by members, with no visible response by staff. It's incredibly frustrating. Ideas are proposed in RSI and apparently ignored, there's no way to know if staff have seen them and don't like the idea, or if they're on a hidden list somewhere.

Another incredibly trivial fix. http://www.librarything.com/topic/95046

105: I voted yes, though I would suggest using the same system as used for tracking bugs.

It's going to be the rare bug that I vote no on fixing it.

That's what you'd expect given that pretty much everyone said bug fixing should be a priority, but the bug votes so far don't actually reflect that.

This is in the bug-tracking thread but for some reason not in the list of the first message, so it's not yet here either. Here comes:In short, CK orig. publishing date -field in Catalogue view shows anything in brackets as double.
E.g. like this for one of Burroughs's Mars-books:
1964 (John Carter of Mars) (John Carter of Mars)
1942 (Skeleton Men of Jupiter) (Skeleton Men of Jupiter)

or

1936 (orig. eng.) (orig. eng.)

Instead of just hiding them. c.f. http://www.librarything.com/topic/98378

Just to add to 109, I've suggested returning a 403 code when a "marked as spam" work or user is viewed. That was the search bot will no longer index those pages:

http://www.librarything.com/topic/98378

I don;t know how to do the polls in the forums here.

edit: Oh wait, that what rorrison linked to. Got my thread numbers confused. Too many windows open.

To create a vote, enter this:
<vote>Text to appear in green goes here</vote>
to get this:

I.e. If I type in &lt; to get <, then edit the post, the edit box should contain &lt; instead of <, so that I don't have to change it back before saving the post.

Alter the "Date Stopped" field so that when displayed in catalog view it allows more than one date to be shown. Currently it only displays the most recent date.

Here's a summary, up to the point where I started listed bug fixes.

Improvements are classified into five groups:
Ultra-Popular and Uncontroversial
Popular and Uncontroversial
Popular But Controversial
Relatively Popular and Uncontroversial
Relatively Popular But Controversial.

"Uncontroversial" means three times as many people for as against; otherwise it's controversial. Yes, this is arbitrary.

Popularity numbers will be revised as people continue voting, but currently Ultra-Popular is more than 30 in favour, Popular is 20-29, and Relatively Popular is 15-19.

109: This is only feasible once stuff that isn't spam no longer gets marked as such. So the whole anti-spam feature has got to work properly in the first place.

v

http://www.librarything.com/topic/85193

117: Actually I would think that if a work was incorrectly marked as spam was returned back to normal, the 200 result would return and it wouldn;t be a 403. That would be common sense.

Holy bejeezus! This is a valiant effort on your part, Zoe, as well as the other posters that have contributed. I've wanted a bug tracker for a long time, even if LT staff take no part in it. At least that way, we could track bugs and features in a better fashion than Talk (which is fine for discussing, but poor for tracking - so this would supplement rather than replace talking about bugs and suggestions in Talk).

I've considered setting up a (unofficial) project for LT on one of the free web-based bug tracking sites. The main stumbling block would be privacy, as plenty of people wouldn't want to give their email addresses to a 3rd party. It would be much nicer if LT hosted it so the accounts would be linked. We're already half-way there with all these polls.

What I'd really want is an extremely simple bug system. The LT site is a lot simpler than deployed software. You only have a few different aspects of a "version": which language site, whether it's the official live site or a site only LT staff can access (useful for when Tim says they've fixed it on the dev site but haven't rolled out yet). No need to bring in all the complexity of a lot of bug tracking systems. Just this, a way to vote on importance of an issue and see which ones you are watching, ability to add comments to the report, and ability to tag it as bug report or feature request.

121> You'd need some way to consolidate duplicate or nearly duplicate bug reports. I suppose minimally you could just close the duplicate reports with some text that points to the main report.

122> I wasn't fully listing all the features I'd want. Marking as duplicate is a pretty basic one, yes.

122: Most tracking systems have such a feature built in. I know on Trac, the system we use, we call it a dupe ticket.

You wouldn't want to automate that for the same reason why automated combining is looked down around here. It's not always right.

So users could annotate their own tags. Interesting to be able to view the anonymized notes on the tag page, too!

This is frequently requested. A set of 3 fields should do it: "checked out date"; "borrower"; "returned date".

Allow users to see the history of changes (and revert) to individual records in their catalogs. I know this is a huge pony! More like a giant winged-unicorn-stallion!

So people can easily manage / monitor / maintain groups they began.

For example, in addition to number of copies held (current default), sort by average star rating; original publication date.

For example, editors, "characters", "relationships".

Another frequently requested feature. One implementation might be allowing us to add items in one field (1, 2, 3; v.1n1, v1n2, v1n3, v2n1; etc.) and a "x" in an associated field.

Another frequently requested feature. This could range from simple add-ons (adding "producer" or "director" to author roles; adding a media field) to complex "DVD-Thing" solutions.

Suggested here.

See here for some discussion.

Most works will never have their CK data filled in. But the original publication date could be estimated by taking the earliest publication date associated with any book in a work. CK would override the estimate.

Estimated OPD could be displayed on work and author pages, used for filtering recommendations and many other nifty features.

18, 21> It would have been useful if these votes could have been set up with the options "right away", "soon", "eventually", and "don't care".

143>

And "Please God Never", which would be useful in the case of, say, #138, where you feel that the proposed "improvement" would be actively detrimental. (Many people would probably feel the same way about, say, review comments.)

I've updated the summary, up to post 123. At the end it was becoming clear that there just hadn't been enough voters yet on the more recent polls.

Ultra-popular is now 55 or more in favour, popular 30-54, and relatively popular 25-29. Interestingly, it seems that the middle categories on my list are disappearing; the proposals with at least 30 in favour tend to be uncontroversial, while those with 25-29 in favour are much more likely to have a significant number opposed.

Why was fuzzy dates removed?

As the number of votes increased, I also increased the thresholds, and it no longer made the cut (though it's very close--just one vote away).

I asked the question about curmudgeonly LTers because it is clear to me that a lot of the polls thus far suggest a high priority on cataloging improvements, and a low priority for fun gee-whiz features or major new directions. This seems characteristic to me of curmudgeonly old-timers .... (I include myself in this category, for the most part).

Maybe its also a reaction to the large number of gee-whiz features lately at the expense of basic cataloging improvements (or I could say major cataloging deficiencies).

I'm actually pleasantly surprised by how un-curmudgeonly the results are. Yes, there are a few basic cataloguing things in the lead, but there's also a lot of support for features that relate to using LT as a reading tracker, a direction that's largely been neglected in favour of ownership. I'm happy to see that the option to connect Currently Reading to the Date fields is currently popular and uncontroversial, and that the reading timeline is relatively popular too. There are also a few pretty minor but interesting and new things on there, like Featured Bookshelf and Books Read This Month--these are fresh ideas that haven't been sitting around in RSI for ages already. Basically, I'd be happy if LT took the whole list of popular and relatively popular improvements as a development guideline for the foreseeable future; I think the voting has reflected a decent mix of cataloguing improvements, social features, and fun little things.

Perhaps the summary should include an Unpopular category, too: at least 30 against.

>151 SchanleyMedia: Done:

Comments on reviews
Show books removed from Currently Reading in Connection News
Ability to sort Local bookstores by number of members who have favourited them
Catalogue field for price paid
Ability to upload more than 25 photos

I have to admit, some of these are pretty inexplicable to me. Who knew sorting venues by how favourited they were would be one of the top most disliked ideas? Likewise, allowing more than 25 photos seems pretty harmless; storage is cheap and it's not a matter of programming time. Oh well.

I don't understand the first, but for the second, LT isn't flickr.

>153 jjwilson61: Nor is flickr LT; we may not want to connect our various accounts. I debated for a long time about how to share pictures with LT friends who wanted to see them; ultimately it came down to giving out a link via private profile message only to people I trust, or uploading a very limited selection to LT itself that I could then show to everyone. I chose the latter as being more inclusive, and I think allowing people to do more of that fosters a better social atmosphere on the site. I'll probably start a new thread about this tomorrow, after I come home from a gathering of LT-ers with (hopefully) plenty of pictures but nowhere to put them.

152> I have to admit, some of these are pretty inexplicable to me. Who knew sorting venues by how favourited they were would be one of the top most disliked ideas?

I think it's just something so niche that a lot of people see it and think "okay, that seems random, no way that's anywhere as important as authors having their own pages." It's not that I hate the idea, I just think it's WAY down on the priority list.

>155 brightcopy: Yeah, probably. But at the same time, everyone is theoretically in favour of completing partially-started features, and I think Local is one of the biggest incomplete features out there: it's a powerful structure that they spent a lot of time on, but it doesn't really do much yet.

154> Or maybe its because the personal photos were only added because Tim said it was part of a big change to the way pictures are handled leading to improvements in cover pictures. Yet after the photo bucket feature was added Tim seems to have forgotten about the rest of it. So maybe its sour grapes over Tim once again short-shifting catalog features for social features.

Specific (and I think important) example of #7's poll for 'finish half-finished features' - I think there's someplace else people want it too, but those two get a lot of requests. I really really want it in the tab-delimited export...

On the Your books page:

Would have to probably set it as "earliest date" when multiple dates are added.

Either for a specific date or for a range.

Edit: for example, search for books read in 2008.

What date fields? Reading or publication?

>167 Nicole_VanK: I was thinking reading, but they should really all be searchable.

{humor} Vote: Limit the number of polls one person can post in a day

Yes/No/Undecided: {/humor}

:D

#26 Like the idea but not that it's based on stop dates - OK for those that use them but not for those of us that use Currently Reading collection rather than the date fields. I get that some people use exclusively stop dates. Others of us use exclusively the CR collection. If you're going to have a CR feature or Recently Read feature it either has to be based on both or neither, otherwise you just exclude/marginalise one group of users and the whole feature is pointless. I can't express that as a poll because they are shallow & unidimensional.

(This kind of horror thread is why I still think polls should just be banished to a lower dimension of he'll)

Can't edit previous post on iPhone but that was supposed to be 'Hell'

>171 klarusu: The thing is, removed from CR doesn't necessarily mean read. I listed both dates and removed from CR as options in #23/24, and the CR one is doing considerably worse. I did suggest a way to link CR and the dates in another poll, and that's currently doing well, so it doesn't automatically have to be either/or. But ultimately, I'd rather have accurate data from a smaller group of users (people who actually say that they completed a book on a given day) than bad data from a larger group of people ("Books Read" based on both books actually read and books abandoned).

Also, all features "marginalize" the people who don't use them. I don't see that as a problem. I'm not insulted when LT does something with the Bookcrossing ID field, and I don't use it. If they can make an interesting BC homepage module, great. I don't buy the argument that nothing should be done because not everyone will use it.

If they can make an interesting BC homepage module, great

Now *there's* something I'd love, but I'm sure I'd be completely driven out of LT by the "curmudgeons" if it ever came to pass! ;)

#174 What's a 'BC' ? as in homepage module? Not enough caffeine yet?

BookCrossing, a book swapping site which is somewhat integrated into LT already since the BCID field in the catalog stands for Book Crossing ID.

ah thanks jjwilson

and I thought the US military uses too many acronyms...

> 178

Hey, thanks, lquilter! :)

182 > I live to serve.

I really think that they should take a month off from new features and work on outstanding bugs and reports. My voting up above reflects that.

edit: Just to clarify, I don't like Lobster. Whenever I went out of town is when my folks had it. Found the cooking method to be distasteful.

> 176: Ah, thank you. I was thinking along these lines: http://www.librarything.com/work/788140

> 176

FWIW, Bookcrossing is *not* a book swapping site. It's a book tracking site, kind of like Where's George?, only for books.

I don't really know what this would be, but it's been discussed before; see the LT wiki.

Copyright registration / expiration / status information for each country as findable.

LTwiki

see ltwiki

Even given the caveats, I think this thread has been useful as input for prioritizing by interest, since features can be more-or-less directly compared in one thread rather than separate ones...one sample of users rather than many (at least to some approximation, given that some people might vote early and not return for later polls.)

> 185

And old Lifetime members should get two free lobsters!

Lobsters: Love eating but not cooking them. Around here go to the supermarket, for $5.00-$7.00 you can get a good sized one and they will steam it for you.

Printouts at present are limited in a lot ways: the width of fields depends on the header, so, for example, Original Publication Date is a very wide field with (mostly) four-digit dates in it; every page is cluttered up with an annoying green-text paragraph even if the selected fields have no green-text data, etc.

Is there a thread somewhere listing unfinished features?

"Original" as in first publication in original language, that is.

There is an original publication date field in Common Knowledge, but not original title.

Forgive me if there is a way to do this that I haven't figured out yet. I'm new here. I'd like to see a bulletin board where all reviews are posted by time, just like these discussion boards are. Of course, people would need a way to opt their review out. It would be good for people like me who don't have a lot of "friends" yet and sometimes review rare books. Also good for people who are eclectic and want to be exposed to something new.

This has been mentioned before (no idea where) and as I recall Tim thought it would be nice...but it hasn't happened.

brightcopy's Greasemonkey script to identify Amazon images is kind of the inverse of this.

A *lot* of the improvements put into polls in this thread are small things, or very specific things, things only a small percentage of users would actually use. A lot of them are things I wouldn't use, even tho I do consider myself a "power user". I have answered "undecided" for everything I really don't care about, and for most of those things I would really like the see other, more important, site-wide things done/finished first.

For example, anything that has to do with CK is low on my personal "want to see" list, because honestly, a minority of users actually look at and/or use any of the CK. Something that would benefit more people, like the long-asked-for better way to add a book directly from the work-page, is much higher up on my list.

>201 Heather19:

I have a similar view. I'm considering converting my 'undecided's to 'no', just so the the 'yes' votes get that slightly higher emphasis.

I've only voted on the ones I feel strongly about. I don't want to vote no for other things, because I know other people really want them. Maybe I should go back and put undecided in just to show that I looked at them.

If you want to play around with the results, the data is available here in csv format. I've double spaced it so that it looks reasonable on the wiki, so you'll have to delete the blank lines. Also, you might have to play around with the quotes first, but I removed all commas that may confuse a spreadsheet.

ETA:
Rats, I forgot to take my voting record out.

I don't have time to edit the summary now, but if I did I would use cutoffs of 40, 50, and 80(?) in favour. I'm encouraged to see that the very last feature suggested would make the cut.

Thanks, Noisy for putting up these results in delimited form for analysis.

I think it really stimulates discussion to look not at the things people are FOR, but what they are against. In a dream world, we could have all of the bug fixes and new features we could dream, instantly. In the real world, developer time is finite. Voters could have two reasons for voting against a feature: either they see it as potentially directly harmful to the site (the truly controversial), or indirectly harmful because time spent on that feature is time not available for other much more important things.

I sent Noisy's data to a spreadsheet and created a column for the ratio of "no" votes to "yes" votes. Only 48 of 129 proposals had more "no" votes than "yes" votes. But even 81 proposals could keep the developers rather busy for a while, particularly given the complexity of some of the most popular ones (e.g., separate author pages and the "contained in" feature.) Something's gotta give.

What might explain proposals currently opposed by a rate of at least 2 to 1? There are only 21 proposals so opposed. Very few are actually controversial in the sense of generating heated debate here (for instance, comments on reviews 2.8:1 against, or supporting other media, at 5.6:1 against). These are old, long-standing philosophical debates, and new RSI discussions or polls are unlikely to add new or different insights. Perhaps a perceived threat to the feel of the site may have contributed to the least popular proposal (13.5:1 against for uploading more than 25 photos; "LT isn't flickr") and an often-suggested option for friends-only access to the catalog (5.8:1 against; arguably making LT less "open").

OK, I'm pretty sure the free lobster (2.7:1 against) would have some serious ecological and financial consequences. No further debate needed for that one either.

Still, many of those voted down weren't necessarily bad ideas in themselves, but they were "ponies" for a niche audience. For instance, the intersection of people who want to play games on LT and those who care about CK seems pretty small, not worth developer time to most voters. Now, I personally like CK, but I acknowledge that it's not critical for most people. I'm not going to keep resurrecting the issue. Similarly, I see many talk posts about people using the "unneeded" BookCrossing field for something else, so it didn't shock me that developing a BC module was opposed more than 10 to 1. And, cool as it may look, how many people, realistically, are going to scan their book spines (against 3.4:1)?

Crowdsourcing can feed *some* ponies. The more that's accessible by API, the more modules techie LT-lovers can extract/mash on their own, and then share with others.

Some features, though, may be inherently more a drain on system resources than they are worth, even if volunteer or developer time is spent to create them. Various options for sorting results fall into this category. I *loved* affinity, but it was crushing the server, and I have to accept that reality. Similarly, everyone has their own pet field they'd like to add (or favorite--which is the same as adding a field), but adding "just one field" multiplies the work the server must do much more than most people realize. That's just fact.

I wish that before anyone suggested an improvement or a bug fix they'd be forced to rank that request on a long list of everything else that's been requested. Recommend Site Improvement: More patient members, with an understanding of economics. BUMP! NEW THREAD!

...like 'see yours' on the author page.

http://www.librarything.com/topic/95602#2106553

#206
I wish that before anyone suggested an improvement or a bug fix they'd be forced to rank that request on a long list of everything else that's been requested.

Isn't that what people have already done in the List your top 5 improvements! thread?
http://www.librarything.com/topic/92742

The problem for me is that the priority I would give is conditional on the time it would take: I'd really like separate author pages for split authors, for example, but if 5 or 6 annoying bugs could be fixed in the same time, then I'd give a higher priority to fixing the bugs.

206: Excellent analysis! Almost all my no votes were on things that I felt were marginal interests but would require large system and/or programming resources. Of course, not knowing exactly how LT is structured, sometimes I'm guessing at the development costs.

Perhaps, if TPTB are tracking this post, they could take the proposals that generated positive votes and split them up for us by how much they would "cost" in terms of limited resources. Perhaps, knowing the cost, a second round of voting would come out differently.

>206 SchanleyMedia: Thanks for that thought-provoking post, caffron. There's one more thing that I think bears mentioning, which is the value for development time spent. Some of the seemingly trivial, low-priority improvements are also very easy to implement, while some of the great and important things may require months or even years of effort. It's not always a case of weighing A against B; for the time-cost of implementing A we might be able to get B and C and D and so on all the way to Z. By implementing a large basket of smaller improvements that offers something for everyone, people might end up happier overall than if they implemented one basic improvement that everyone agreed on.

Edit: which is basically what Eveleen and aulsmith said while I was typing this.

I don't think the point of this thread is to come up with a dictated list of things for staff to do, but rather just to provide them with an idea of user desire for the various features. In the final analysis it will be up to them to consider available resources, difficulty of implementation, and whatever other factors there are. All we can do here is say what we want; I don't think it's necessary for us to consider internal factors. I did not vote no to anything based on my perception of the work involved.

This exercise wasn't driving by staff asking us to prioritize their work for them. I don't think a second round of voting is necessary, they'll do that -- we can say what we want, and they'll decide what they do. There shouldn't be any expectation that whatever prioritized list we come up with will have any relationship to what's actually implemented when.

>212 r.orrison: Well, we can certainly hope that what they do will be strongly influenced by member priorities. There are plenty of popular features for them to choose from.

Inspired by caffron's comments, here's a list of the improvements ranked by difference in users for and against.

Definitely agree that features aren't equally valuable to members for development time spent. Related to that, features aren't equally valuable to developers in terms of motivation/feeling of accomplishment. If Tim et al. spent all of their time squashing bugs, they'd probably be so demotivated that they'd cut their workdays in half, thus yielding a net decrease in output for us. Little features make them feel good, particularly when they get lots of positive feedback on a creative idea that members didn't ask for. Some of the more popular items on this list came from creative, unexpected features that members liked, namely Overcat and Tag Mirror. I think that overall the staff do a pretty good job of balancing the new with fixes. There are some bugs/incomplete features near the top of the list that are a bit overdue, but overall I'm glad that the site is adventurous and generous enough to have lots of optimization left.

So that reviews
aren't all
Squashed
together in one
annoying
column like
this

so that if you get to page 3 of Talk, and read a thread, clicking on Talk will take you back to page 3, rather than page 1 and forcing you to wait for 2 more pageloads.

#206

I see a 5.6:1 against as not controversial at all. There is a pretty strong consensus against. A controversial (in terms of generating a lot of heat in long drawn out debate) are those issues where there are pretty equal numbers strongly against and strongly for. Because of the limitations of the polls we don't know numbers strongly for / strongly against. A "sure, why not?" vote is treated as equal weight as a "wow, that's a must" vote. Having said that I think that this process has resulted in data which can be taken as a 'guide' to help the general direction of development but not to strongly direct priorities.

On ranking bug fixes on priority I could see a proportional style vote (maybe cumulative, maybe STV) to decide which get fixed first. However in general I don't think enough people would vote to be worth setting that up. Most people don't get involved in these discussions and Tim also has to think of those.

219> Caffron said controversial in the sense of generating heated debate here...

So when he said controversial he was referring to the heated debate in Talk and not to the vote.

Okay, here comes the inevitable griping about the poll phrasing. :D

I think it's a bit of a problem that the polls didn't define what Yes, No and Undecided really meant. At least, a problem for really drawing a conclusion. If we'd had something like this:

Yes means you like the idea and want it implemented in the same priority as everything else you voted Yes on.

No means you don't like the idea at all.

Undecided means you like the idea and want it implemented, but after all the ones you voted Yes for an at the same time as all the ones you voted Undecided for.


Of course, this still leaves out some options, like if you're actually undecided on whether or not you want the thing implemented. It all goes back to the suggestion of letting polls allow other answers, so you could have:
Top Priority, Secondary Priority, Never and Undecided

Though given Tim's lack of love for polls about features (unless he's trying to justify a feature/implementation HE likes ;) I doubt that one will rank highly on his own priority list.

> 221

That leaves out a major option, namely, "I don't care", which is not the same as being undecided.

>222 lilithcat: Honestly, I'd use "I don't care" fairly often. I suppose abstaining from voting is "I don't care," but you don't know your voter pool so you cannot get that out of the data.

222> I consider that on par with not voting. ;)

ETA: If you actually HAD an "I don't care option", then you'd just not know how many people clicked "I don't care" and how many people read the poll but didn't care enough to click "I don't care." So yeah, I see it as a bit of a moot point.

Heartily agree with #222-definitely need "I don't care, what difference does it make, or no opinion" as opposed to undecided.

Most of the time I voted undecided because I didn't care about the issue one way or the other. This didn't seem ideal to me, but I wanted a way to distinguish "no, I don't like this idea" from "no, I don't care about this idea." Ideally, I would have liked an "I don't care one way or another about this proposal" option. Occasionally I was truly undecided because I didn't know enough about the issue.

I've only used undecided when I really was undecided (i.e. good and bad aspects basically evenly matched in my mind).

If I didn't really care and saw no significant downside, I didn't vote.

(FWIW, there were two or three where I voted NO on ideas I would have liked/found useful for myself, but I saw a significant down-side for the site/community/members/staff.)

229> That's going to take some clarification. As it stands, I don't understand how this is any different than the RSI forum, with it's threads and posts on them.

>228 _Zoe_:
I'm probably being dense, but what do you mean by "work on Legacy Libraries"? Some sort of increased functionality or connectivity related to the LLs?

232>

http://www.librarything.com/more/projects

"Legacy Libraries" are the libraries of famous dead people, entered into LibraryThing by dedicated members working from published catalogs, archival lists, auction catalogs and personal inspection.

Legacy Libraries offer unique insights into booklovers of the past and their times. Once in, LibraryThing makes it easy to compare your library to theirs.

Over 100 have been completed with another 50 in progress, and include historical figures (Thomas Jefferson, Marie Antoinette), authors (Dante Gabriel Rossetti, Astrid Lindgren), and modern celebrities (Jackie Gleason, Tupac Shakur). There are even a few institutions (The White House Library), and even boats (The H.M.S. Beagle). Members work on libraries by themselves or in groups.

>231 brightcopy:, 232 Both were a somewhat tongue-in-cheek comment on the fact that Tim has been posting about Legacy Libraries throughout the day, initiating a project to catalogue Napoleon's library and committing to doing some of the cataloguing himself (he also posted in the ER group and Ancient History).

For #229, though, I do think it would be worthwhile if Tim read all the threads and at least said something like "Interesting, but not a priority" or "No way" or whatever.

234> Cheekiness is all well and good, but I have to say if you go down this path I think you'll defeat the entire purpose of your thread.

>233 brightcopy:
Sigh, yes, I know what Legacy Libaries are. That wasn't the question. What I didn't know was what Zoe meant by "work on" (silly me, I should have recognized the snark was meant there too).

>235 brightcopy: I don't think the motivation for including a poll will impact the way people vote. They're still polls in the list of polls.

What I mean is that if this drifts from the original point to just a way to poke fun at Tim, I (and I have to assume others) will just ignore the thread.

236> Sorry, I misread. No offense intended.

>238 brightcopy: Well, I'd say the major data collection is done anyway. There's certainly no point in continuing if Tim can't be bothered to look at the thing.

"can't be bothered to look at" seems rather harsh and makes assumptions that may be unfounded. Was this a jokey phrasing for "does not comment on"?

>240 TheoClarke: No, it wasn't a joke at all. I have no reason to think he looked at it.

I might be less harsh if he weren't simultaneously ignoring a request to comment on an unusual account deletion that had generated a bit of debate. I don't expect him to deal with things like that on weekends, but it would nice to see that site issues were getting addressed come Monday.

(sniff sniff squeeek)

For the record:

1. I had Liam from Friday to Sunday, so no commenting.
2. I've been waiting to comment before I could read it. I've done so.
3. Zoe has been exiled to Elba.

On balance, I think it's a very good list, and one I'll take to heart.

It has some biases. Easy-to-understand or -explain stuff gets some unearned votes. (For example, "Allow YourBooks column widths to reflow properly," begs the question whether they do reflow "properly" or not, which is highly contingent on your browser, OS and so forth, and on your taste for what "properly" means, but most will assume is general.) And I'd love to see--and don't know how to construct--polls that consider trade-offs and costs. It might be enough to give everyone $100 and then put prices by all the features. To spice that up, you'd have to spend some of the money to keep current features, if you wanted to. To reflect the situation better, certain features would make other features cheaper, and others more expensive. (For example, cataloging improvements make other cataloging improvements more expensive, but privacy would make others more expensive.)

The account-deletion issue is interesting. I'm waiting for Dan to return tomorrow to comment on it. In general, we DO take a different attitude towards spam. Fighting it is a serious job, taking up a lot of time. We have to be somewhat brutal, I'm afraid, and will make mistakes. Generally, everything can be reversed, but Dan used the rarely-involked "obliterate" feature here--which can't be completely reversed. So I need to figure out what's going on with him.

Vote: Tim gets to have a life

Yea or nay.

>242 timspalding: Thank you; I actually feel much better now :)

Does anyone have a suggestion for how to handle the relative importance issue? For example, "Talk remembering which page you're on" is something I would anticipate wasn't an issue for many users. I myself never use the next button, and I am in a lot of groups. Perhaps I'd use it more if I were in certain types of groups I'm not in, or if I visited the site less frequently. I'm not sure. In any case, it seems to me I can read the "pro" people as being a mix of people who think it critical and people who think "Well, sure, that sounds nice and someone must want it."

Does anyone have a suggestion for how to handle the relative importance issue?

Yup, way back in Message 18:

polls that allow other answers than yes/no/undecided

;)

In general, though, I'm not sure that people are voting yes just because they think a feature will be useful to someone else. Based on the comments above, it seems more common to choose "undecided" for features you don't personally care about. In my ranked list, Talk remembering which page you're on would be #75 with a score of 6, which really isn't very high. Of course, there aren't that many votes yet for that poll, but some later polls are already way above that.

Polls are eye-candy. People take them. I do not believe for a second most people who've taken the polls have also read all the words.

To add to 245, the best system would actually impose the result. That is, if you voted for something, you'd get it before other people. If you didn't vote for it, you'd be stuck without it. If you didn't "pay" the $5 necessary to sustain a certain feature, it would be taken away from you--and available to others :)

To be clear, I think some things stand out very clearly. That's more important than getting everything right.

>242 timspalding: Thanks for the response, Tim (which I almost missed because it was an edit of post I'd already read!)

3. Zoe has been exiled to Elba.

Will you catalogue my books for me while I'm there? :D

Easy-to-understand or -explain stuff gets some unearned votes.

I don't know, I think there's something to be said for implementing a feature that's easy to understand. Something really obscure and hard to explain is just much less likely to be a crowd-pleaser.

And I'd love to see--and don't know how to construct--polls that consider trade-offs and costs.

I think your idea of giving everyone $100 is a good one. I see this as a possible outgrowth of the List feature, though I hadn't previously considered allowing people to vote multiple times for the same item in a group-constructed list.

To spice that up, you'd have to spend some of the money to keep current features, if you wanted to.

I think this would just dilute the value of the data. People are always going to be much more attached to existing features than to new ones; I don't think Tag Mirror would be so high up on the list if you hadn't given and then taken away. Features should only be removed as a last resort.

That said, I'd still gladly be rid of Will You Like It.

That is, if you voted for something, you'd get it before other people. If you didn't vote for it, you'd be stuck without it. If you didn't "pay" the $5 necessary to sustain a certain feature, it would be taken away from you--and available to others :)

This could make for some interesting real-life pricing. There are certain things that it would be reasonable to pay for, like photo storage.

Or have you considered having a donation contest, where you set a price for your time and people can donate real money to the features they want developed? That could be fun.

1. Page speed, catalog loading speed, search speed, etc.

>252 timspalding: I'd definitely vote for caching more. I'll take a six-month old Tag Mirror over none at all, thanks :)

As a for-example of how I take these things:

1. People express idea that current speed is okay, but no slower.
2. LT is often slower than other sites.
3. People who think it slow have left. People who stayed have gotten used to it. People cope by not visiting certain pages that are slow. I bet they don't even realize they cope this way.
4. Studies show a REMARKABLE link between site speed and people's willingness to use a site. Slow pages are the kiss of death.

>251 timspalding:

This has too many questions in it - I might vote yes on one part, but no on another.

>254 timspalding: I think there's another issue that's closely connected to speed, namely efficiency. Having to load two pages where one would do is effectively the same as making each of those pages twice as slow. Unnecessary page loads should be avoided at all costs.

I introduced my brother to LibraryThing a month or two ago, and it was pretty embarrassing. "What do you mean, I have to load a separate page in order to search?" "Why can't I add multiple books from this series by checking a bunch of boxes on one page, rather than searching for each one individually?" I suspect that a lot of users give up because of the clunky UI before they even get to the point of realizing that some pages might be slow.

Also, I can't personally think of any pages that I avoid because they're too slow. I do hate using the search page and the edit books page, though, for different reasons.

256: Totally agree. Remember our arguement/debate over the removal of the Tags tab, Tim?

LT's speed is fine, fast, actually, compared to some of my other usual sites (this may just be my personal experience, tho). But *relative* speed, ie how long it takes to get to a buried feature like tag-combinations for example, is embarrassingly slow.

And as much as I voice the whole "new features and exisiting bug-fixes do not have to be mutually exclusive", I'm starting to cave on that. There are WAY too many long-standing bugs and half-done features that have been stalled for way too long. I'd rather all those get taken care of and not have ANY new features for the next 6 months. .... I can dream, can't I?

In response to 256:Where Generic editions allow you to quickly add a book that consists of only title and author, copied from a work. In addition to the usual wishlist purpose, this could also be used to add multiple books from a series. You could even modify the Add Books page to have Title and Author boxes, then two buttons: Search or Add. Search would do what it does now, Add would add a generic edition.

The poll is: should something like this be done specifically to make it easier for new users to add books.

I definitely have to second Zoe's point about really noticing the slowness of LT often through inefficiency. One place this constantly bugs me is the series page, where I wind up opening five different windows just to try to find out if a series stays good over time or goes downhill. This is one of those "work smarter, not harder" suggestions that have been around forever. The data could even be cached on a weekly basis - so no real argument about it bogging down the series page.

260> Actually, that's one of the downsides of generic editions. If just title and author editions were available, I'm afraid a lot of new users would go that way (and not bother with editions). Which means a lot of non-data. Of course, the alternative is random (and therefore bad-data) editions...

But I'd love to have generic editions for myself, for wishlisting and ancient books (ones I read as a kid, but have no idea what edition - same idea as fuzzy dates). I didn't vote.

Wow. I've only posted three polls and Tim commented on two of them.

Definetly another+1 on the efficieny angle. I do avoid parts of the site because of long load times though - like I avoid page 2 of talk because I know how much faff it is to read more than 1 thread on it.

On page 2 of talk I always use a right-click and "open in new tab" to read threads.

like I avoid page 2 of talk because I know how much faff it is to read more than 1 thread on it

I don't understand that: surely you don't load open the thread in the same tab the list is in, do you? No wonder it's slow! I just load up the topic list, middle-click each thread I'm interested in reading, and then either close that tab or switch to the next one, read the first thread, close that tab, read the next thread, etc; so it's loading the talk pages while I'm still looking at the list...and if you don't close it, you're back where you left it after the last thread.

About talk pages remembering and who actually wants it... " Perhaps I'd use it more if I were in certain types of groups I'm not in, or if I visited the site less frequently."

I need this to use the all messages/groups view. If you aren't sorting by your groups there are a LOT of threads to go through to find something you might want to see. I may want to read one 75 group challenge thread without joining the group and seeing a million, so I flip back and forth between my groups and all groups to make sure I'm not missing out any good threads by not joining/watching it all.

I'd say this feature is actually more necessary for people with fewer groups because they are more likely to want to view All Groups rather than just theirs because theirs might be rather limited. I open new tabs currently to make it work but it's nice not to need workarounds.

Edited to add: Also using an old IE at work reminded me of how painful it is not to have tabs and how much harder life is without the work-around.

#266

I use other tabs as well however it is a coping mechanism that some are not going to arrive at, and which some will always feel uncomfortable with.

262>

I voted "no" on this not because I think buried features should stay buried, or because I think the current UI is perfect, but because knowing Tim's philosophy, any UI work would hide more features and make things worse. So I'm reluctant to open that can of worms.

I just wanted to pop in to thank everyone who has taken the time to put together (and do some basic analyzing of) this list of polls. I've found it helpful to see what is 'on the table' and what other people value as compared to what I value.

The one thing I've felt strongly enough to comment on thus far, though, is the last set of polls - LT's user experience is frankly a mess, it doesn't reflect its power and the awesome things it offers, and it's sometimes painful to so much as add books.

I would spend probably my entire $100 on reworking aspects of the UI - from menus and tabs and a search box on every page to adding books with fewer clicks - if we did that! (I like the idea of the money thing to gauge relative importance btw Tim!)

Here's a new thread to try out this money idea.

I avoid clicking on a work link in a talk message because of the time it takes to bring up that page. It's only slow the first time such that subsequent links to different books aren't slow but I know I avoid that first hit because it takes so long.

I just tried it and the first book took about 13 seconds but the second took only 3 (that's with me counting slowly under my breath so it's by no means exact). So it seems the slowness isn't in the time to access the work record but in something else that's the same across works.

This could include several different ways of doing it, including putting the rating on the main series landing page, having separate tabs that might show you multiple data items across series such as rating, year published, etc., or yet some other way of doing it. It also includes finding the best way of doing it without slowing things down. Though as I mentioned earlier, series are already slower to the user when they have to open new pages for all the works they want to see the rating on.

I've voted on every one of the polls here.
FWIW The above (274) is the only new feature (i.e. something that is not "finish part finished stuff/fix bugs") that I've voted "Yes" on.
So good job there then.

I'm pretty ambivalent about ratings for works on the series page, but only because I think the series page itself needs a lot of improvement. I feel like there needs to be more CK attached to series than the description (a checkbox for "on going" and a box for "number of works currently comprising series" are two that come to mind right away and which can't be surmised from looking at the list of works within the series - this can be included in the description, but it'd be nice to have dedicated spots to standardise the information), plus indication of whether or not the related series are on different language-sites (which has a poll up above), and perhaps one day, as a form of container relationships, series reviews and ratings.

I'd be interested in seeing the ratings of each work on a series page if it's part of an update that gives expandable work information/ck, but otherwise, I'm not sure that it's all that important to me. And, to be honest, series page revamp is pretty low on my wishlist, excepting the language-site indication for related series, which is in my top ten this week.

The (admittedly limited) results of the other thread have made me curious....interpreting "a few" relatively loosely, maybe up to 5

Deleted (See post # 283 below)

Thanks, lquilter!

281 appears to be a duplicate of 75.

>282 _Zoe_:

Something like FB's wall-to-wall option?

(Oh, you're right, there is a page that I avoid because of slowness)

>293 _Zoe_:

Apparently splitting friends into groups has been something of a failure on Facebook. Few do it. It's too "work"-y. We do have the ability to make new categories--you can call some "Sort of Friends," "Friends with Benefits," etc.

>294 timspalding: No, that's only for Interesting Libraries and Private Watch List. If I want to accept a friend request from someone, they have to go in that one single Friends list. So if I've decided to use Friends for people I know in real life, everyone else is out of luck.

I'd much rather know about a "new feature" when it's in it's planning stages, and have us all tell him "yunno, that's not really something we care about too much", then have the LT admin take so much time developing this mystery-feature and it turns out it's barely even used. Which has happened before.

While generally agreeing with you, Heather, I can't help remembering all the times that Tim has asked for input about a planned/possible feature, and gotten so much conflicting and often acrimonious response that the project sank without a trace. :(

This is actually sort-of in progress, there's a LibraryThing for Libraries app which includes a lot of the ground work for a LibraryThing app.

Yes, http://m.librarything.com exists, but I'm thinking of something a bit more functional. Amazon is a good example, or Marks and Spencer in the UK.
Edited: It's http://www.librarything.com/m/

The very fact that a number of different apps are wanted to serve the same purpose mean that they would take up a lot of time that I, for one, would rather see spent on the main, big screen version(s).

You've detected the bias in my questioning! The first three are unnecessary if the fourth is done. On the other hand, how the development effort and potential number of users compares for the four cases is unknown.

The ability to specify on a work (not a book) who the author is, overriding the system selected author.

I can't help remembering all the times that Tim has asked for input about a planned/possible feature, and gotten so much conflicting and often acrimonious response that the project sank without a trace. :(

I think polling would help here. Sometimes there are really just a few people who strongly object, or lots of people objecting to one small aspect of the feature, and Tim gets the impression that the whole feature is bad.

I'm thinking of the initial Currently Reading features, where some people really didn't want the connection to the Date fields, and rather than making that connection optional, Tim scrapped the whole feature. A poll might have clarified that the feature as a whole was good, that there were plenty of people who did like the Date connection, and that simply making the Date connection optional could have kept everyone happy.

Likewise, I remember Tim saying that there was a lot of negative feedback to the introduction of the new homepage, when really everyone was complaining about the loss of the Profile tab.

Clarifying user concerns with polls could avoid a lot of these problems. Then we can focus on the issues that really need discussion because a consensus hasn't yet been reached about the best course of action: for example, it had seemed that users wanted some sort of improved book discussion feature, but opinion seems to be split between new book discussion forums, new book discussion threads, and improved touchstones. More talking rather than just polling is needed there.

Zoe, Tim and users who are interested in polls and features.

I enjoy participating in giving Tim and the staff information regarding the features that continue to help make LT one of my favorite websites.

Polls can be helpful with figuring out in what direction people are thinking, for sure. However, this site has a very diverse group of people and many are simply here to use the site and don't have any interest in development and how decisions get made. They just want what they want when they want it like people all over the world.

I think what I am hoping to mention and ask others to consider, is that many, many LT users and members have never done polling (on the design end of it), and won't participate or submit submit ideas for site improvements, etc.

It might be worth considering that an interested group of knowledgeable communications, research and development people can work together (as volunteers) to submit polls for site improvements. Maybe having people from several countries represented would be a good thing.

Readers *can* be very independent for sure OR they simply follow the crowd. Many would be happy to not have to even think about LT, just use it, and for others of us, it is central to our lives. Maybe something like a think tank, which I know is a very dated idea, but might be good in the end.

Tim, this is your baby and your business. What would help you decide and focus where to put the time and resources that you have to getting the most bang for the buck?

Zoe, and the other movers and shakers here, Thank you, thank you. Your enthusiasm and energy are appreciated. R&D always seems to be the key to having a business work for the "customer/public".

I would love to see what some of you can come up with yourselves in order to keep this RSI viable; knowing, going in, that many LT users will not even know that it is going on behind the scenes.

What do you think? What kind of time, brain power, perspective and wit do you wish to offer as a "customer/consumer/member" of LibraryThing?

Just my thanks and *two cents worth*.

As much as I am entertained by the polls, I do want to comment on the idea that users know what they want and what they would use.

The very first video tape machines (notice, I did not say "recorders") sold for home use were playback only. That might seem absurd now, but at the time, focus groups of consumers could not imagine why they would ever want to record a television program. Obviously, once the device was in the hands of the public, users came up with all sorts of reasons to do just that.

The same phenomenon exists with LT features. Many people cannot imagine why they would ever use something (what was I going to do with collections that I couldn't do with tags?) but once it is available, it gets used. Or not. I'm sure there are features some people lobbied for that turned out to be not so exciting after all.

The wisdom of the crowd does not necessarily apply to evaluating something never before experienced.

@309

You do have a point, but I think that the group of unaware future users will have a much larger proportion voting undecided (as opposed to no) than the group of aware future non-users.

And from a comparative perspective your point is not as important either. Even if I will eventually like feature A my "no"-vote might express the fact that I rather want to see feature B and C (as soon as possible) than feature A. That is, I am not necessarily saying I don't want A even though I vote "no", I am saying it should not be prioritized.

I just have to say that I have voted in most polls, and read all the words.

I voted yes for what I want, no for what I don't and undecided for those that I don't care about or don't understand what is being suggested.

In general I care more about fixing Bugs/finishing current features than adding new features.

I do care about speed of the site, and do find it slow now. I am not in favor or more caching, too much is cached now for too long. What is the point of collecting/displaying info that isn't accurate or complete ?

We have a big database change to the core personal books structure coming on Sunday night--when we will go down for a few hours. I hope it will help on speed, although it may require two Sundays. In tandem with it, John and I have improved speed tracking. We are thinking we're going to bring the speed stats page live for users to see.

Does this big database change involve finishing the Other Authors feature, by any chance? Or maybe Fuzzy Dates?

No, it's a technical change--designed to reduce the size of the data, and in this and another way make it faster and therefore more scaleable. I will also be taking one of the table "shards" and testing another performance improvement, which can be applied to the others or pulled back, depending on whether it helps or hurts.

While a majority of users may feel the site is fine as it is (250 ff.), I feel it is too slow, especially for users with a lot of books, and, even if it weren't, keeping it from slowing down, which people did say they wanted, requires constant optimization as the site grows.

I'm currently working on venue images, one of the long-standing incomplete features.

Is Other Authors the next long-standing incomplete feature, then? (It seems to me to be one of the longest-standing ones, and it isn't exactly getting any shorter.)

(I vote for Same-Name Author Splits as the long-standing incomplete feature I'd most like to see finished.)

316: agreed!

316, 317>

Me too, but before even that I vote for a round of serious bug-fixing. Things like the giant "No title, no author" work and the occasional work getting misclassified as spam really need to get fixed.

315: It just occurred to me that since there is no functionality to Other Authors, that it really doesn't matter that long-term members have Other Authors in two different formats. If LT ever introduces functionality, we can press for this upgrade then.

>319 aulsmith:
The only reason why I'm making a fuss now is that I once tweeted Tim about it: his reply left me with the impression that it would be harder to fix if left 'till later.

The Other Author issue is, I think, going to be resolved when we allow switching and re-parsing of records.

I'm uncertain if other authors needs to be involved in the splitting or not. It's fairly complicated, frankly, and that's why it hasn't been done yet.

Tim, send me the $100. I'll keep that, you can decide what to do with the site. ;)

I think I'm one of those devoted marginal users. I love LT as it is right now, which is not to say that I won't love it as it evolves, unless it becomes like certain other social book sites which I've left because they did not allow me to control my library and the details in it. What I love about this site is that it is flexible and there are many depths of usability here. If I ever get my life back as it once was, a lot of the features up above may become important to me, but until then, I can just exist here happily reviewing, adding to my catalog and chatting with other book lovers in my groups. That's why I voted undecided on a lot of the questions, they were either beyond my ken, or not important to me.

I.e., only show the same thread once.

>323 _Zoe_:
Could you clarify that?

Undecided on 323 because I don't think it would be necessary if searching Talk *worked*, so that you could actually just search "My Groups".

When you search in Talk, it only shows 100 results, so you can't find older posts.

>326 lorax: I usually search within a single group, and the restriction is still annoying there.

Yeah. I actually voted NO on both of these, simply because I think until a complete overhaul of search is implemented, it's a bit moot the exact ways in which it fails.

>328 brightcopy: Is that on the list already?

329> And I voted Undecided on that because it's about priorities, right? ;)

330. Ditto. The inadequacies of talk search don't bother me, I use google:

site:librarything.com/topic intitle:"group" search terms

Edited: shorter version from brightcopy - thanks!

In other words, ditch the sites own search and have it open a new window (or something, don't get hung up on that part) that uses the google search instead of the built-in one. If necessary, it could change www.librarything.com to whatever language site you are on.

ETA: I should note that this would probably (in my estimation being outside of the LT codebase) be something on the order of a 10 minute change. I've already coded it for myself in a greasemonkey script, and that took less than 10 minutes!

331> Oh wow, I didn't know you could specify further with the site: search. That is beautiful!

I mean the personal recommendation list linked from the profile.

Just today I came across one with an average rating of 3.14 on 750 copies (more than 200 ratings).

>335 _Zoe_:
I'd like to see this work both ways, or at least have an adjustable definition of how low the exclusion point is.

(If this post isn't clear enough, you'll have to start a new thread. This one is taking forever to load.)

Okay, continued here.

Just today I came across one with an average rating of 3.14 on 750 copies (more than 200 ratings).

What book?

>338 timspalding: Tantalize

LibraryThing says I won't like that book, and I believe it.

LibraryThing says with very high certainty that I'll love it. I haven't read the book, but I can proclaim with confidence that LibraryThing is wrong.

I dunno, Zoe. You read some crap books.

:runs

Nah, I don't think you know as much about my reading tastes as you claim. LT is too focused on ownership.

No, I was just vilifying!

I was just taking advantage of an opportunity to get a point across ;)

What we've stumbled upon is that, actually, LibraryThing does not have a recommendation system. It's just Tim playing Mechanical Turk, nosing through your library and sometimes giving good recommendations, sometimes intentionally recommending a book he knows you'll hate just to get your dander up.

>346 brightcopy:

But I actually like that as a recommendation system. Call me strange, but I'd much rather be recommended books that I'd find interesting not necessarily that I'd like/love all the time (for me those are not the same thing).

For example, if LT only recommended books that it thought I'd love, it would mostly show Victorian fiction. As much as I do like it, I'd like some recommendations outside that limitation, but not so crazy as to start recommending L.A. Candy or something.

This thread is continued here: http://www.librarything.com/topic/99431

> 306

Explaining my "no" vote.

It's tempting, really tempting. But what do you do, for example, about books that are co-authored? I can see an ongoing battle between those who want to set the canonical author at "Author A", those who feel "Author B"'s contributions were greater, and those who want to give them equal time and put in "A and B".

349:
I'm not claiming to have a fully-worked out design for the feature, but I do agree it would need to be able to handle multiple authors. Perhaps multiple fields, like the book level "Other Author" fields. Roles, perhaps, e.g. Editor for collections. Obviously people who only contributed to specific editions wouldn't be appropriate -- translator, illustrator, etc. Don't vote against it just because it's not easy or fully specified (unless you think it's absolutely impossible to make it work).

Look at message number 2: "Lists". 111 people thought that was clear enough to vote for it.

349> Wouldn't this same line of reasoning (because people will disagree with what the values should be and get into edit wars) equally applicable to Canonical Title? Yes, sometimes people do disagree and yes, sometimes there are (generally small) edit wars. But I think that's minor compared to how it addresses a real existing problem.

36> I put this in Private Comments, but would love to have a separate line, maybe with the ability to keep it private.

And, it would be even better to have a field for where purchased, and linked back to the book store page.

53> I'm all for this one! Been wanting it for a long time. It should also link to the bookstore on LibraryThing's site.

65> Sorry, but I don't understand this. Could anyone, _Zoe_ especially, explain this? I'm really curious to comprehend.

Currently, clicking on the green plus brings us to the search results on the add books page, where we have to choose a particular edition. I'd like it just to add a stripped-down version of the work immediately, with all fields blank except for title and author.

115> If we could link to GetGlue, we could show everyone the days we're reading a given book...

Jim, it would be helpful if you could quote a little but of text from the message that you're replying to. Jumping back 300 messages is a bit of a pain.

349> I didn't think Message 2 was clear at all, so I didn't vote on it.

357> Which will lead to a poll, if it's not on the next topic... I was just following other's lead by xxx>, which I noticed early on. It wasn't till later that I noticed replies with italics.

66> Or, vice versa, where if I put something in my current reading collection then the start date is entered and when I move it out (or maybe click an "I'm finished" link) it'll put in the date finished.

356> I think linking to GetGlue would be great.

360> re: 66 - that was the problem back when Tim first suggested tying dates to Currently Reading. For my own usage - I read a bunch of books and then come on the site and enter dates; if I happen to be reading one that takes a while, it will go into CR, but easily 75% of my reading never does. And even the ones that do, basically all of them go into CR some time - some days - after I've started, and get taken out of CR some days after I finish. So if it auto-entered dates, I'd have to edit them every time. I'd rather enter the dates myself, it's simpler.

I'd really rather that "Currently reading" were auto-populated (a "smart" collection) with things "recently finished" and "recently started". Say, the ten most of each.

>360 arcane_scholar: Ideally there could be a few options so that it could go either way. I might check the boxes "add to Currently Reading when I enter a started date" and "remove from Currently Reading when I enter a finished date", and you might check the options "set start date at today when I add a book to Currently Reading" and "set stopped date at today when I remove a book from currently reading".

>362 lquilter: I'd like to see a separate Recently Finished collection (and maybe Recently Started as well).

I'd rather just have 'currently reading' turn into 'recent reading', which covers both stopped and started nicely.

I would love to see them autopopulated, though. And something actually /done/ with this data sitewide. I hate it a lot when they give us toys and then take them away again.

Can I have a pony instead? I don't like lobster.

Actually, I'm allergic to real ponies. So how about a virtual pony?

That would be great especially if buying used books through a broker like Alibris so you can track the actual vendors who have the book. This would also make a great memory of places traveled. Leave the field as a fill in, not a drop-down.