Planet Four Talk

Talk improvements, post your suggestions here

  • DZM by DZM admin

    Hello Planet Four Zooites!

    My name is Darren--DZM here on Talk--and I've just joined the Zooniverse team. I'm working on a few projects, first and foremost an overhaul of the Talk system. I’m also going to try to be a regular on these boards going forward, so I wanted to say hello and get to know everyone here on Planet Four!

    If you have any additional ideas for how Talk can be improved, I’d love to know. I’ll also be looking for any other ways in which I can help out, so I’m always interested in hearing from you all—questions, concerns, ideas, anything.

    I should note that I’m not an astronomer, and I probably know more about Barsoom than the real Mars, but I’m definitely interested in learning from you Martian masterminds! Just note that any actual science contributions that I may attempt to make are purely amateur. 😃

    Anyway, I’ll be around and checking back, so thoughts, comments, anything like that are always welcome. Nice to be here!

    Mars is amazing!

    DZM

    Posted

  • p.titchin by p.titchin

    Hi Darren, from me, welcome aboard! You have joined the best of the zooniverse discussion boards~ salaried as well! If you have a sense of humour, then I am sure you will fit in with us unpaid 'citizen scientists. Mars is amazing, and I wish you all the best in your new job. ~ Pete

    Posted

  • Kitharode by Kitharode moderator

    Hi Darren. Nice to meet you. Yes, Mars is amazingly weird and wonderful. A bit like like some of us really. 😉

    Since you ask, I'd like a way to empty my mailbox. Living out in the wilds means slow broadband for me and my overflowing mailbox takes forever to load. Cheers.

    If you really are new to astronomy then you should definately get one of these. Enjoy.

    http://www.facebook.com/photo.php?fbid=455575746998&set=pb.316550651998.-2207520000.1364013903&type=3&src=http%3A%2F%2Fsphotos-g.ak.fbcdn.net%2Fhphotos-ak-prn1%2F35959_455575746998_6922755_n.jpg&size=720%2C628

    Posted

  • wassock by wassock moderator

    Hi darren and welcome. 2 things for starters:

    If we, as moderators, move a post to somewhere else the "last post by" tag stays put on the old board. If you are quick you can see an example - go to the borads and look at Chat/Hirise images you' see the last post was by Meg. Click into the board and you can't find the post, because its been moved to Science/The Objects/Mars.

    The other thing is that we tend to have some fairly long threads here. If you are at the most recent end of a discussion, or worse in the middle of one, and you follow a link from a post, then when you click the back arrow, you return to the first post in the thread rather than where you were. Irritating if you are on the last post, annoying if you're in the middle of a 12 page thread but can't quite remember where.

    Oh Actually 3 things - I've got an unopened mail showing on my profile and try as I might I can't clear it - There's no title to click on to open it but a blue dot to the side - I did open and reply to it when it came in, but can't get the notification to vanish now.

    Cheers

    Posted

  • DZM by DZM admin

    @Kitharode, I had one as a kid! Used to bring it out into the yard and look at stuff. I never quite got the wherewithal to really learn what I was looking at, though. I just liked seeing the North Star up close. I think it's still at my old house, just mostly serves as decoration these days... I should try to reclaim it! Although you can't stargaze well from Chicago, sadly. #lightpollution

    Thank you all for your suggestions and your welcomes! @p.titchin, the best of the boards, you say? Better not hear Galaxy Zoo hear you talking that way, I don't want any Zooniverse civil wars! 😃

    I will add your suggestions to the Big Ol' Talk Ideas Document, or what I'm calling the BOTID. That'll be a guide for us in putting together the new system. It's still in its earliest stages, but I will be sure to keep everyone posted.

    Glad to be here!

    Posted

  • Kitharode by Kitharode moderator in response to DZM's comment.

    What you mean is, you had a telescope similar to the one in the link. The one in the link can't see the Pole star, or any other star for that matter !! (look closely).

    Did you ever notice that the Pole star was a variable? Sometimes brighter sometimes dimmer. Not many people know that. 😉

    Posted

  • mschwamb by mschwamb scientist, translator in response to DZM's comment.

    Welcome to Planet Four Talk Darren.

    Cheers,

    ~Meg

    Posted

  • DZM by DZM admin

    @Kitharode ... I had a telescope, but you're right, mine was not designed by M.C. Escher. I didn't look closely enough! #fail

    Thank you for the welcome, Meg! I've gotten more welcomes here so far than any other board... so maybe you really are the best board? 😃

    Posted

  • ElisabethB by ElisabethB in response to Kitharode's comment.

    Hi Kitharode

    You can delete pm's.Click on a pm and bottom left there is a message in blue (delete this message). But be careful, I think if you click on delete the whole pm thread gets deleted.

    Posted

  • ElisabethB by ElisabethB in response to DZM's comment.

    You do realize that some zooites travel all over the place ! 😄

    Posted

  • DZM by DZM admin

    I do! You in particular, I see!

    But, hey, @p.titchin laid claim to being the best board, and because I am fantastically self-centered, I'm going to measure the truth of his comment by which board offers me the most welcomes! 😄

    In all seriousness, it's great to get around and meet everyone. You all seem like awesome people and I'm excited about doing SCIENCE with you all!

    Posted

  • wassock by wassock moderator in response to ElisabethB's comment.

    Hi Elizabeth, twas I with the pm prob. Which is basically that I cant actually open it, its stuck as an unread mail, but ckick away as I might it doesnt wanna open

    Posted

  • mschwamb by mschwamb scientist, translator in response to wassock's comment.

    Hi,

    I'm sorry if I knew about this and didn't do anything earlier. Can you send me the details in a PM and I'll see if the development team can manually delete the message? Might take a while since it's a low priority to do item, but I'll inquire what they can do.

    Cheers,

    ~Meg

    Posted

  • Kitharode by Kitharode moderator

    Meg: Thanks for the advice. Most useful.

    Wassock: Meg's reply was to my enquiry about emptying my mailbox.

    Darren: Welcome, welcome, welcome, welcome, welcome, .... Planet Four rules OK !! 😉

    Posted

  • DZM by DZM admin

    Thank you, Kitharode! :3

    Posted

  • mschwamb by mschwamb scientist, translator in response to Kitharode's comment.

    Actually I'm referring to the unopened message that won't open so the mailbox icon always says there is new mail that Wassock is experiencing... Kitharode there is the delete button now for the messages which is a new feature that got added a few months back if I recall. So you're going to have to delete your own inbox, Kith (sorry!)

    Cheers,

    ~Meg

    Posted

  • wassock by wassock moderator in response to mschwamb's comment.

    Hi Meg, its a message from Paul Johnson from 2 months ago, theres no title so I cant open or redpond to it, though im sure I did at tye time. Not high on my priority list either

    Posted

  • mschwamb by mschwamb scientist, translator in response to wassock's comment.

    I bet It's the no title that's causing the problems. I've sent an email to the development team to see if they can go into the Planet Four Talk database and manually add a title or delete the message for you.

    Cheers,

    ~Meg

    Posted

  • DZM by DZM admin

    If it is indeed the lack of a subject line that is causing an inability to delete, I'll make sure that's noted as a bug for the Talk rebuild, too.

    Please let me know if that turns out to be the issue; thank you!

    Posted

  • JeanTate by JeanTate in response to wassock's comment.

    Oh Actually 3 things - I've got an unopened mail showing on my profile and try as I might I can't clear it - There's no title to click on to open it but a blue dot to the side - I did open and reply to it when it came in, but can't get the notification to vanish now.

    If it's any consolation, I had that problem too. Got reported, got Dev team onto it (apparently), got feedback ... and nothing happened (the errant PM is still there). All took place quite a few months' ago now ...

    Posted

  • Kitharode by Kitharode moderator in response to JeanTate's comment.

    Interesting. I had a 'won't go away' message marker ages ago that hung around for about two months. It did eventually disappear and the problem has not occured since.

    Ah, the joys of life on Mars ... 😉

    Posted

  • DZM by DZM admin

    I've put one of our developers on trying to fix the no-subject issue, the same one who fixed the link colors.

    Hopefully should get it functional/repaired soon!

    Posted

  • JeanTate by JeanTate

    @Kitharode @DZM: It was, indeed, a 'no subject' PM (and it's still there); it did not originate (?) in Planet Four, but I didn't think that made any difference ... until now. IIRC, no one else in GZ Talk (I think that's where it got the most discussion) mentioned having such PMs, but here at least two do, so perhaps it does have a Zooniverse project-specific aspect?

    I also just noticed something else: the envelope notification counter is (was) wrong. Some time ago it would go grey, even though there were 'unopened' PMs (if only the 'unopenable' one), but that's been fixed for a while. However, my counter has been "1" for a long time, which I assumed was that errant PM; not so, there were others I hadn't 'opened' (not that it matters, for me, I try not to use v2 Talk PM at all, given how unreliable it has been; now that Planet Hunters has gone to v2 Talk though ...).

    Posted

  • mschwamb by mschwamb scientist, translator

    From the Zooniverse blog - more about Darren and his role in the Zooniverse.

    Cheers,

    ~Meg

    Posted

  • DZM by DZM admin

    Thanks for posting that, Meg! I'm posting it everywhere else on Talk, too.

    Feel free to check it out, and I'm happy to answer any additional questions that anyone might have! 😃

    Posted

  • wassock by wassock moderator

    Darren, And another thing -

    It would be good to have some sort of 'chat' facility at the moment you can see who's on but the only way of talking is to mail them or post on a discussion thread, neither of which is ideal for a bit of a chat, specially if there's more than one interested. And doing it in a discussion thread generally ends up so far off the theme that the thread has to be restarted to get it back on track.

    Posted

  • DZM by DZM admin

    @wassock, yep, that's been suggested by several folks! It's in the ideas document.

    I'm glad you mentioned it, though... the more people who ask for something, the more likely it is to happen.

    Thank you very much!

    Posted

  • Kitharode by Kitharode moderator in response to DZM's comment.

    ... and if only for some of wassock's epic threads, maybe a 'back to top' button?

    Posted

  • DZM by DZM admin

    @Kitharode, by back to top, do you mean the top of the current page, or the beginning of the thread?

    Posted

  • DZM by DZM admin

    Also, check out that no-subject PM issue... it should be fixed now.

    Posted

  • wassock by wassock moderator

    Suspect kith means just back to the top of a single post, but top of the page would be good.

    Another thing that has been asked for before is to have the most recent post at the top, rather than the bottom as is now.

    Still no joy with the PM, screen shot below, nit is the one with the blue dot

    Posted

  • Kitharode by Kitharode moderator in response to DZM's comment.

    To top of page for me. Happy to use page number buttons to return to start of thread. Ta.

    Posted

  • DZM by DZM admin

    Drat, it's not fixed? Alright, let me see what I can do. Thanks!

    Posted

  • DZM by DZM admin

    I'm going to request something insane... Someone please send me a private message with no subject!

    If it's sitting in my inbox, I'll be as annoyed as the rest of you until I can get it to go away. 😃'

    EDIT: Is it fixed now?

    Posted

  • mschwamb by mschwamb scientist, translator in response to wassock's comment.

    I actually prefer the most recent post on bottom so I can read chronologically starting from the top with the older responses to the newer. It's much harder to read a conversation's history for me if you have the most recent on the top especially if since I'm typically not involved in the conversation from the beginning.

    Cheers,

    ~Meg

    Posted

  • p.titchin by p.titchin in response to mschwamb's comment.

    I agree with Meg, I think they have to be in chronological order to easily follow the discussion, but a 'top of page' button would be handy in the long threads. ~Pete

    Posted

  • wassock by wassock moderator in response to p.titchin's comment.

    I can see that Meg, I'm just lazy and want to see the most recent post straight away. Just out of interest how do your e-mail threads arrange themselves?

    Posted

  • DZM by DZM admin

    I definitely agree that there needs to be a button, even a tiny one, that brings you to the most recent post! 😃

    Posted

  • mschwamb by mschwamb scientist, translator in response to DZM's comment.

    Hi,

    There already is. Is everyone aware of it? In case not, here's what I'm talking about:

    On the recent pages when you click on the discussion comment it takes you to the place in discussion scrolled to that comment. On the discussion subboards if you click on 'last post' when you are in the subboard and its lists all the threads with the most recently updated at the top (http://talk.planetfour.org/#/boards/BPF0000005) it's actually a link though with the color scheme and link scheme it might not look like one) on the right where each thread is listed and it says who made the post and when that takes you to the recent post. I use it all the time.

    Cheers,

    ~Meg

    Posted

  • mschwamb by mschwamb scientist, translator in response to wassock's comment.

    I don't thread my email (only on my phone and it's not my primary email device). So emails go in chronological order, newest on the bottom (though to be fair I have a bunch of folders and filters sorting my email but the same format occurs in those folders). So I guess that preference in email carries over to here as well.

    Cheers,

    ~Meg

    Posted

  • wassock by wassock moderator

    Yes I use the most recent post button, like I said just sloth wishing not to have to go to the effort of that one click 😉
    What would be really great would be a "take me to the most recent post in this thread that I havent read" button

    What would help tho is if the last post link worked from the general boards page. At the mo you get when the latest post was on each board but you have to enter the board before you can get to the latest post

    Btw the stuck pm problem seems to be fixed, Thanks DZM

    Posted

  • Kitharode by Kitharode moderator

    Hi Darren. Might be getting a bit deep here, but if I don't ask I won't know. 😉

    I was browsing the Zooniverse Education blog and revisited the Zoo Tools post. Brilliant idea. Do you know about them? Have you any idea if Planet Four is on the list to get them? What I'm most interested in is a measuring tool to measure objects in the images we see here. As I say, might be asking a bit much (?)

    Zoo Tools: http://education.zooniverse.org/2013/11/26/zootools-going-deeper-with-zooniverse-project-data/

    Posted

  • DZM by DZM admin

    Hi Kitharode,

    I'm still figuring out exactly how Tools work(s), myself. I'll tackle this on Monday with the developers. Thanks for the idea!

    Posted

  • angi60 by angi60

    Hi Darren. Just wanted to say welcome. I'm another one of the mad regulars, but I've been away for a couple of weeks. The others have already mentioned the issues I would have pinpointed, but I'm sure I'll think of more at some point! Hope all goes well for you. ~ Angi

    Posted

  • DZM by DZM admin

    Thank you for the welcome, Angi! 😃

    Posted

  • wassock by wassock moderator in response to DZM's comment.

    Hi there, woild it be sensible to shift this thread up to the interface board? Or I could open a new board therenfor you to collect suggestions, its a bit hidden away in this board.

    Posted

  • wassock by wassock moderator

    Posts moved from chat "Hello Planet 4"

    Posted

  • DZM by DZM admin

    Thank you! 😃 Sorry I missed your post before!

    Posted

  • Kitharode by Kitharode moderator in response to DZM's comment.

    Hi Darren. Was wondering if anything might be done about the 'Insert Image' function. It's not clear to visitors how it works.

    Thanks to wassock and JM we now know that we need to put images in Dropbox (or similar) then grab the link from there to place in the Insert Image box. To make the image appear in the discussion we then need to change www. to dl. (for whatever reason). Anyone who doesn't know this, or asks about it, will simply think that it doesn't work or they're doing something wrong (as I used to).

    Also, embedded images 'overflow' the screen and are only partly on view. Again, without being told that one needs to zoom out (Ctrl- on my laptop) visitors might not be able to make sense of the image. For an example, please see my last post in the 'Fan Tool Gone Haywire' discussion which is just underneath this one on the 'Interface Help' board.

    We could put instructions on the FAQ page, but visitors might not see it there. Sorry to add to your job list, but you did ask ... ** 😉 **

    Posted

  • rwilliams208 by rwilliams208

    it would be helpful to know the direction of sunlight in the pix. despite the statement that the dark co2 follows 'channels' in the surface, it would appear to me that they might actually be in situ sublimation of the ice from the top of 'elevations' rather than depressions, with the darkness reflecting underlying darker surface. also it would be helpful to see the complete marking rather than partials, since the top of a fan would appear as a blotch, and a blotch might have formed from two fans if wind direction reversed or shifted. some blotches appear to have been formed by multiple direction fans, not just reversals. also would be nice to know the science, if sub limation from elevations is posited, assuming that cold air would be heavier and fill the depressions or channels while elevations might warm faster? also does mars atmosphere get layered as it does on earth where elevations might actually get colder air over the warm air in the depressions or channels

    Posted

  • rwilliams208 by rwilliams208

    why do pictures repeat? is this a check of consistency in marking? or is it a test of mental competence? please provide picture id for commentary purposes?

    Posted

  • Kitharode by Kitharode moderator in response to rwilliams208's comment.

    Hi. You might be better off posting your questions and ideas in 'Citizen Science' or 'Chat', so that people can join in at length. This thread is primarily for the 'tech team' to gather interface ideas/ problems.

    I can tell you, however, that sunlight comes from the bottom right of the images you are marking. Cheers. 😉

    Posted

  • Kitharode by Kitharode moderator in response to rwilliams208's comment.

    PS: Each image needs to be classified by numerous people (50, I think) before being stored for analysis. Because they are randomly displayed you might well get to see an image more than once.

    I'm not sure what you mean by "provide picture id for commentary purposes".

    Posted

  • wassock by wassock moderator

    I think thaf the idea is that you should not get the same image twice. That being said there are now 5 years of pics inn the system so it possible to get the same area twice but from different years. If you do get the same image twice please flag the details for the team on the interface help board as a new thread.

    Wrt your other questions I'm not s sure I understand some of them, if you could start a new thread and rephrase them we'll do our best to answer. But the spiders are channels, took me a long time to be totally convinced as well, and the question of which way is up/where is the sun is a non trivial issue to sort out even though pretty much all of us have suggested it as a good idea.

    Posted

  • michaelaye by michaelaye scientist

    Actually, we try to avoid having the same user with the same login to get the exact same Planet4 image more than once. If you guys feel that this happens significantly often, please let us know!

    Posted

  • JeanTate by JeanTate in response to michaelaye's comment.

    If you read through the various Talks, for different Zooniverse projects, 'repeats' seem to happen in (just about) all of them*. In various Talks there are inputs from various Dev Team (a.k.a. Admin) people about a) it doesn't happen, b) it happens only very rarely, c) we treat this very seriously, d) sorry, but yes, it is happening and we still haven't fixed it up, e) ...

    Initially I thought it was just one or two projects, but now I think it's a generic problem, likely at the very heart of the 'engine', of how Subjects are chosen for Users, for all Zooniverse projects. Yes, there are many assurances that "we're on it!", but posts like rwilliams208's, from less than a day ago, are pretty solid evidence that it hasn't been solved yet (and for every rwilliams208 - who notices and writes about it - there are, very likely a dozen to ~100 others who don't^).

    *perhaps the most distressing to read are those in Notes from Nature; each classification takes quite a bit of time to do, and it surely a downer to read, once you're done with one, that you had gotten that very same Subject before!

    ^as I discovered in Quench; there were very few reports of repeats, but when the data was crunched (and made available to us ordinary zooites) the rate of repeats was seen to be high (shockingly high to me). (Links available upon request)

    Posted

  • mschwamb by mschwamb scientist, translator in response to rwilliams208's comment.

    Hi,

    To re-echo the above from Michael, Waasock, and Kithrarode, we show an image now to 30 people before we retire it from showing. Typically the images shouldn't repeat. It does happen in very rare instances from our understanding based on discussions with the Zooniverse development so you shouldn't have seen the same image twice very often. If you have seen a few please note the APF ids and we'll contact the development team about it.

    Cheers,

    ~Meg

    Posted

  • DZM by DZM admin

    I got a message from one of our developers today asking me to report projects where repeats had been noticed.

    The more data that I can give them, the better... so if you want to hit me with examples/links to threads/APF IDs, etc., now might be an excellent time!

    EDIT: Also, @kitharode, we're definitely aware that image functionality is confusing and limited, and it's high priority to fix and/or implement in next-gen Talk!

    Posted

  • michaelaye by michaelaye scientist

    Actually, to save your time and the health of our database, in these cases, I would recommend to just close the browser tab WITHOUT pressing "Finished". That way we don't get an entry of your duplicate and you don't waste your time doing it again. I know it's not optimal, but the best we can do until the underlying problem is solved. (I honestly don't know what the problem is, but it must be something deep in the system, I agree).

    Posted

  • JeanTate by JeanTate in response to michaelaye's comment.

    From my own experience that's ... impractical.

    Rare is the image - in any Zooniverse project I participate in - which is so distinctive that I will be certain to recognize it when I get it again to classify [1] In fact, I wasn't even aware of how common repeats were until I started commenting on every image I classified (and, just so that you know, I no longer classify/click in any Zooniverse project except RGZ; at least, not regularly/hours on end) [2]

    but it must be something deep in the system, I agree

    Based on the feedback given by the Development team (not here), it's highly likely there is more than one cause. IIRC, the (a?) cause of 'nearly back-to-back repeats' was found (and fixed), but the 'long cycle repeats' remain mysterious.

    [Meg] It does happen in very rare instances from our understanding based on discussions with the Zooniverse development

    Yeah, when I first reported it (not this project), that's the answer we got too; it took some concrete examples - provided by zooites - before they (collectively? just the Dev Team member assigned to this?) agreed that it was happening far more often than they'd originally stated. So perhaps there's also a problem in that the tools the Dev Team are using to spot repeats is itself flawed? That the Zooniverse project databases contain a much higher level of duplication than anyone really recognizes? If you're interested, I can give you links to what we finally learned, in a different project (Quench). What I think is really disturbing is that that level of repeats/duplicates may exist in many, perhaps nearly all, projects, and no one is really aware of it.

    [1] repeats within a few minutes, or back-to-back aside. These do happen, but in my experience are the exceptional repeats, not the common ones.

    [2] Stardate M83 is an exception. The PI very quickly accepted that repeats were happening, said he didn't know why, that it seemed to be peculiar to that project, and suggested zooites not continue classifying if they didn't like getting repeats

    Posted

  • JeanTate by JeanTate in response to DZM's comment.

    Suggestion: go to all the Zooniverse projects (or at least those which use v2 Talk), and search for threads with 'repeat' or 'duplicate' in them. If you've got the time, you can also try searching in the comments too. An hour or so's sleuthing will tell you if what's being posted is, in fact, evidence of repeats (and not something else, such as overlapping images).

    I think you'll find that more than one zooite will have noticed this behavior, and commented on/discussed it. However, the way they describe it will likely vary, by project, so it won't always be obvious. If it's any help, some time ago I went and looked at all the current Space projects (the ones which use Talk), and found evidence of repeats being reported in all of them.

    Posted

  • mschwamb by mschwamb scientist, translator in response to JeanTate's comment.

    Hi Jean,

    This duplicate issue appears to be a different issue for other projects. The Planet Four science team doesn't have the background on those builds or the history of what took place for us to have a constructive dialogue about other Zooniverse projects and their duplicate issues. I think it's best to have the discussion about duplicates for each project on its own Talk since the frequencies are different and the causes may be different depending on the requirements for how the data is classified.

    I can tell you for Planet Four, we actively looked into this issue. At launch there was a bug that enabled duplicate classifications but that was corrected once spotted a month after launch. For Planet Four, I worked on checking this issue when we saw other duplicate classifications during the spring and summer of this year with our development team liaison. Duplicates represent was 0.18% of all classifications, which is incredibly small and the frequency per cutout is very low at the present time.

    Many images on Planet Four look similar so sometimes it look like its a duplicate image but it may not be the same image. So if you or anyone else think yous see frequent duplicates on Planet Four, please list the APF ids so we can be certain and double check in the database that there are duplicate classifications. If it turns out this is occurring for some reason at a high rate, I can assure you the Planet Four science team will work with the Zooniverse development team to get the issue sorted.

    If you have concerns on duplicate classifications for Planet Four, let's continue that discussion in this thread.

    Cheers,

    ~Meg

    Posted

  • DZM by DZM admin

    On our end, a developer reports that we've just updated and deployed the Zooniverse library, which should take care of most/all repeat issues here on Planet Four, at least.

    I'll keep an eye out on other projects, and I'm currently looking into Notes from Nature.

    DZM

    Posted

  • JeanTate by JeanTate in response to mschwamb's comment.

    Thanks Meg.

    I can tell you for Planet Four, we actively looked into this issue.

    I think that makes the suggestion I was going to make, based on what you wrote in an earlier post*, moot. I cannot know for certain, but in at least one other Zooniverse project, when a Science Team member analyzed the 'click data' themselves, they found repeats had occured at a higher rate than the Development Team had found.

    I think it's best to have the discussion about duplicates for each project on its own Talk since the frequencies are different and the causes may be different depending on the requirements for how the data is classified.

    Fair enough. I'll limit my comments (etc) here to just P4 objects, from now on.

    *"It does happen in very rare instances from our understanding based on discussions with the Zooniverse development so you shouldn't have seen the same image twice very often."

    Posted

  • JeanTate by JeanTate in response to mschwamb's comment.

    Hi Meg,

    This didn't click with me immediately:

    Duplicates represent was 0.18% of all classifications, which is incredibly small and the frequency per cutout is very low at the present time.

    As understand the architecture of how the Zooniverse engine works*, duplicates/repeats should be zero, exactly. To be sure, Zooniverse projects are not like bank accounts (much less commercial airliners!), and dups/repeats may not be particularly problematic when it comes to analyzing clicks; however, given that what is not supposed to happen - period - does, are the reasons why it does understood?

    *from what Arfon Smith wrote, in a Zooniverse blog post quite a while ago now; of course, individual projects might have pieces outside this, which might affect things

    Posted

  • wassock by wassock moderator

    Jean, given that there was a bug early on, now fixed, it is a given that there will be some duplicates in the data set.

    Posted

  • p.titchin by p.titchin

    I spot the odd one, but infrequently enough that I don' find it annoying, but rather tend to just greet them as an old friend! 😃 Pete

    Posted

  • mschwamb by mschwamb scientist, translator in response to JeanTate's comment.

    In practice yes, but what you code a computer to do and what it does are too different things. On paper it's perfect, in reality the system's very close to zero duplicates, hence why when I looked into it with Chris Snyder, the Zooniverse project manager, this summer it was decided this was so small and the cause was not apparent, that this was going to be dropped from further investigation. The development team before that did push several patches where they thought they had found a a place that maybe possibly could be contributing to the duplicate bug and it wasn't fixed that we left it since it was such a small fraction of the classification rate. Fully understanding why it was happening means you could fix it. If the developers did they would have it fully sorted, but the Zooniverse system is quite complex on the backend and it's a small team that built the system and maintains it while also building other projects.

    My understanding is that there was another attempt another fix the issue in the early fall for other projects on the backend where it is happening more often and that update to the Zooniverse backend library was pushed to P4 when Darren inquired.

    As Wassock mentioned, early on there was a bug in the backend found during launch of our project and that corrected, so we already have duplicate classifications we need to deal with when doing our analysis for the first paper.

    My point is, that when ever it's been brought to our attention of duplicates occurring in increasing numbers, we've notified the development team and have worked together to try and find the cause and minimize it's impact on this project. The development was aware of it and I was pointing out examples from the classification database and there were active development team people working on it. For this project, the rate was so small and after several attempts at fixes, it was decided this was okay to move forward and use the development time for other things that were higher priority in the Zooniverse.

    Cheers,

    ~Meg

    Posted

  • DZM by DZM admin

    Meg put it well.

    We're definitely working to minimize/eliminate repeat classifications, and we've just deployed another library update that should help, but there's a point at which obsessing over a tiny, tiny, tiny fraction of classifications becomes counter-productive when there are so many other things that require our developers' attention. 😃

    However, that said, please do keep reporting them when you see them. The greater a body of confirmed repeats that we have, the better we know what causes them to happen!

    Posted

  • p.titchin by p.titchin in response to DZM's comment.

    As someone on this sight for 2 years, I really do not see this as a problem you should be using your budget on. I spot them when I call them up and see I have already commented on the image in the past. So what? As I commented already on this, I see them as old friends, and take the time to see what others have commented on the image. Please everybody, the number is so small, let's put it to bed, and let the 'sight gurus' spend there time on more important problems.~Pete.

    Posted

  • Kitharode by Kitharode moderator

    Hi Darren. Added a comment to 'Boulders' (in Citizen Science) and it shows as 'last post by Kitharode 6 hrs ago'. However, on the main Discussions board it says the last post in Citizen Science was by wassock several days ago. Do you see the same problem, or is it something at my end?

    Posted

  • wassock by wassock moderator

    I am seeing same as you Kitharode

    Posted

  • DZM by DZM admin in response to Kitharode's comment.

    Yep, I'm seeing it. That's an obvious bug, and might give us some insight.

    I'll show a developer as soon as I can find one. Thanks!

    Posted

  • Kitharode by Kitharode moderator in response to DZM's comment.

    Hi Darren. I know you've got this covered, but I just wanted to show it wasn't a one-off. On the main board in 'Chat', 'The Objects' it shows 'last post by Mote 12 hours ago', but go into the Objects and it shows 'last post 4 hours ago by Kitharode'.

    I'm using Internet Explorer. Cheers, Kith.

    Posted

  • DZM by DZM admin

    Why are you using Internet Explorer? 😃

    Nevertheless, I've put this through as a bug report and I hope a dev will take note of it soon. It's not the biggest issue, but yeah, we want people to be able to see recent discussion!

    Posted

  • Kitharode by Kitharode moderator in response to DZM's comment.

    I'm only using IE when I want to do some classifying. I generally use Firefox, but that won't let me use the marking tools properly as explained in 'Fan Tool Gone Haywire' elsewhere on this board.

    I understand the 'last post' bug is not a priority, so no worries.

    Posted

  • DZM by DZM admin

    Might want to give Chrome a try. Although, going forward, I really want to prioritize making our projects work equally well on all browsers! 😃

    Posted

  • mschwamb by mschwamb scientist, translator in response to DZM's comment.

    Hi Darren,

    I just want to jump in here to give context since you might not be aware of the background.

    This issue isn't browser specific (P4 behaves the same in all browsers for the most part). This bug Kitharode talks about is different. It has been seen occasionally in all types of browsers (Safari and Firefox, etc), but it's intermittent and doesn't impact a majority of users. I've never been able to replicate it, and Michael on the P4 science team has seen it once. Chris Snyder has spent a considerable amount of time trying to track it down, but never was able to get the bug to repeat on windows machines or mac. Without a cause, Chris has been unable to apply a fix. Unfortunately Kith seems to keep hitting it, and we're not sure why. The development team has put alot of time into tracking it down, but without being able to replicate, there's not much they can do. If more people have symptoms then that might give enough info to track this weird race condition bug down, but so far we don't have the numbers impacted.

    Cheers,

    ~Meg

    Posted

  • DZM by DZM admin

    Thanks for all the information, @mschwamb... These issues that only ever seem to plague one or two users are always the trickiest, aren't they?

    Sorry, @Kitharode... Hopefully this either stops happening, or we get more people reporting the same issue.

    Posted

  • Kitharode by Kitharode moderator in response to DZM's comment.

    Thanks. It's no longer a problem now I know that IE works ok for classifying. With luck nobody else will get the problem and you can forget about for now.

    Posted

  • wassock by wassock moderator

    HI DMZ - pesky back arrow problem again:

    If you use the search feature to try and find a thread containing a phrase or word in a discussion (probably applies to other searches too) you get a list of the posts. If you click onto one of the posts it displays fine but if that's not the one you were looking for then 'back arrow' returns to the blank search page.

    Weirdly in order to do the search again you need to enter the search term and then click the category - "discussions" BUT only when you then unclick 'discussions' do the posts appear again.
    Bit of a pain when trying to search for a specific thread.

    Posted

  • DZM by DZM admin in response to wassock's comment.

    Yeah, we definitely know that the back arrow has a wild mind of its own. It has to do with the software upon which this version of Talk was built, and is sort of heavily ingrained in the system.

    It's probably not fixable without rebuilding Talk from the ground up... which, incidentally, we ARE doing!

    Posted

  • ajah by ajah

    I'd like to suggest adding Contrast/Brightness controls if possible.

    Some form of Zoom control (other than me having to globally magnify the browser) would be nice too, but definitely brightness/contrast, to improve the ability to identify subtleties that might otherwise go unnoticed.

    thanks,

    -a

    Posted

  • DZM by DZM admin in response to ajah's comment.

    Thanks, @ajah .

    Are you talking about within Talk, or within the classification interface?

    Posted

  • alex8192000 by alex8192000

    Wow, these blotches are so off the chain!

    By Victoria

    Posted

  • IsabelKershaw by IsabelKershaw

    Still having a problem marking multiple fans on my iPad. Crashes about every 4th image. Throws me out and into a new image. Hoped it would've been sorted by now as has been a year since I last tried.

    Posted

  • mschwamb by mschwamb scientist, translator in response to IsabelKershaw's comment.

    Hi,

    I'm sorry that's still happening. I'll let the development team know and see if there is anything they can do.

    Cheers,

    ~Meg

    Posted

  • Blendouble by Blendouble

    Could we have a leader board on the planetfour.org website that tallies a basic verified statistic from each mode per user.

    The verified stats possibilities could be: (1) how many images have been identified with craters, (2) how many craters users have been pinpointed, and (3) how many craters have been outlined?

    Having a leader board would seriously help the motivation for monotonously putting dots and circles on an image over and over again. It would be like a game then, with level-ups. I know people who, at times, love to do monotonous work if it improves a level statistic about their profile/character.

    Respond Posted May 21 2015 8:50 PM

    Posted

  • Blendouble by Blendouble

    It would be nice to have a little control in picking an area to identify craters.

    I think I saw a Google Maps type map of Mars somewhere. If we could have a color graded overlay of the user data for areas of the map, or an overlay of the map that we have personally covered, we would be able to pick areas that are still within the set of picture sections under study, but of more personal interest.

    I second the zoom feature request by @ajah too.

    We would also get a better sense of accomplishment, seeing an overlay of the area of mars that we have personally viewed and contributed data towards.

    At stat on a leader board of how much area we have verifiably covered would be nice too.

    Posted

  • wassock by wassock moderator

    Hi Blendouble, the craters thing is an offshoot project. The images come from a prospective site for a future lander. But the project is actually more about how users feel about the different tools than about mapping the surface. Your comments about incentivisation will hopefully be 8f use to the project managers, but it is a fairly short term thing so I wouldn't expect anything to happen for the craters project just yet. Someone else has been looking at what drives people to take part in citizen science

    Posted

  • Audriusa by Audriusa

    Sometimes a clearly huge fan originates outside the visible photo area. Possibility to scroll up, down left or right even by a limited distance would help to mark these as well. If fans are few and big, otherwise it may be nothing to mark on a photo.

    Posted

  • p.titchin by p.titchin

    These cropped images are just that. Hard to see how you could scroll beyond them .If it is interesting at an edge, I go on the HiRise and find the image on the colour strip, and examine the larger area. if it seems of note, then I either comment on it on 'talk', or just on the image comment box itself. Others classifying the adjacent images will mark to the edge, so I think it all gets covered. ~ Pete

    Posted

  • mschwamb by mschwamb scientist, translator in response to Audriusa's comment.

    Hi Audriusa ,

    Thanks for the feedback. As Pete notes, we have overlap between each of the subimages so even though you see half a fan someone else classifying the neighboring image will see all of it. Just do your best there. We're working on combining classifications from the overlap regions, so marking half a fan is helpful to us. Just do your best at the edges.

    Cheers,

    ~Meg

    Posted

  • Noharrr by Noharrr

    Collections were my main way of keeping track of my classifications. I had hoped they would be restored once the new platform settled in, bu that has not been the case.

    All possibility of "collecting" images seems to have been lost. Am I mistaken? If collections are gone, I'd say that's a big negative for the new format at ZOO.

    Thanks for your consideration.

    Posted

  • mschwamb by mschwamb scientist, translator in response to Noharrr's comment.

    Hi Noharrr,

    Thanks for the post. You should absolutely be able to continue making collections in Planet Four Talk. Collections you made in Planet Four still exist, you just need to go to http://talk.planetfour.org/#/users/Noharrr to view them for your account. The new system is a different platform so for now it's only projects on the new system whose collections are showing up on the Zooniverse home page, but I believe there is a future plan to eventually make sure collections from older projects show up.

    Cheers,

    ~Meg

    Posted

  • pimtewes@gmail.com by pimtewes@gmail.com

    I see a lot of white lines in red background, any idea what it si?

    Posted

  • mschwamb by mschwamb scientist, translator in response to pimtewes@gmail.com's comment.

    Hi,
    If you mean white lines in an image like this one, than you're likely seeing yardangs. They're wind blown scuplted terrains. You can learn more about them here.

    Cheers,

    ~Meg

    Posted

  • Gabbleduck1 by Gabbleduck1

    Please optimize site for iPad!
    Really difficult to use, keeps crashing!
    I do not have a computer anymore!

    Posted

  • mschwamb by mschwamb scientist, translator in response to Gabbleduck1's comment.

    Hi,

    I'm sorry the site isn't optimized for ipad. Our companion project on the new Zooniverse platform is fully ipad compatible. You can find it at http://terrains.planetfour.org

    Cheers,

    ~Meg

    Posted

  • ruben.piacentini by ruben.piacentini

    Congratulations for your Zooniverse work! Some suggestions for the improvement of the representation of the objects:
    -If the object is all defined in the inner part of the figure, without interception with the borders, they can be represented by the two geometrical figures ("abanico", actually a triangle + a semicircle and "mancha" an elipsoid). However when the object to be registered is not completely in the interior of the image, its representation is quite complicated. Can you include more geometrical figures or gives the possibility to star these figures outside the Mars image?

    • Can the Mars images by expanded, in order to see more details?
    • Is the color of the object real or it is a false color representation?

    Posted

  • mschwamb by mschwamb scientist, translator in response to ruben.piacentini's comment.

    Hi Robert,

    Thanks for the feedback. We have overlap between our different subimages we show on the website (it's easier in data reduction to allow for only two tools than a larger number), but if we get whatever part of a fan on the edge you can, someone will likely mark the whole thing on the adjacent tile and we can combine that information.

    In terms of viewing more area, it would be hard to do that in the web interface, but on Talk we have links to the full HiRISE data that the subimages we show on the site come from. You can explore those be hitting the 'View HiRISE Link'

    The color is close to what you'd see on Mars. One of the filters is infrared which our eyes are not sensitive too but is good for composition study. The color transform is described here

    Cheers,

    ~Meg

    Posted

  • everlasthitman by everlasthitman

    I have a question we humans ever live on mars

    Posted

  • everlasthitman by everlasthitman

    will we find other planets that has life on it

    Posted

  • mschwamb by mschwamb scientist, translator in response to everlasthitman's comment.

    I think you'll find we the Planet Four Team talked about this a few days ago. You can read about it here.

    Cheers,

    ~Meg

    Posted

  • rhw007 by rhw007

    Would like Zoomverse by Planet Four to highlight "unusual" buttes or other "interesting" terrain. To take this up especially along the 40th latitude both north and south since THIS latitude in summer can reach temperatures as high as 70 deg F in summer AT THE SURFACE. Which can cause 'unusual' geo-morphic changes more so than ANY other latitude. Thank for consideration in advance,

    Posted

  • mschwamb by mschwamb scientist, translator in response to rhw007's comment.

    Hi,

    Thanks for the feedback.There are definitely lots of things that can be marked on Mars, but the main question is there a scientific reason to do it and is there someone who use the data and make sure the work of the volunteers gets applied to a scientific question. For the Planet Four team the focus is on the seasonal fans and that processes. For our latest project Planet Four: Ridges, we're looking for polygonal ridges at mid-latitude because there are a few ways they can form and so finding more means we have a large sample to compare to locations of hydrated minerals to see if underground water was the dominant mechanism for their formation, giving insight into Mars' past climate.

    If we did find a use for marking other terrain, we'll definitely think about building that project.

    Cheers,

    ~Meg

    Posted

  • mariakatosvich by mariakatosvich

    I spot the odd one, but infrequently enough that I don' find it annoying, but rather tend to just greet them as an old friend.

    Posted

  • jbx56 by jbx56

    Speaking of overlap... On Planet Four: Terrains 6751938, the image would be a 'yes' on Planet Four: Ridges - but there's no way to notate it as such. Probably inevitable with this many images to process, and this may not happen all that often. But if, in the search for spiders and craters, we come across ridges, perhaps there should be button for that too??

    Posted

  • mschwamb by mschwamb scientist, translator in response to jbx56's comment.

    Hi,

    We know there are spiders and araneiforms in the images shown for Planet Four, so we don't ask you to identify them. What we're after is the shape and direction of the seasonal fans. For Planet Four: Terrains we use a different camera that has a larger field of view and is covering more area that we don't know necessarily has spiders. Comments about what you're seeing in the image are very welcome. Talk would be the right place to do that.

    Does that make sense?

    Cheers,

    ~Meg

    Posted

  • TPS1BCW by TPS1BCW

    Hi Meg,

    Kitharode said to advise you about my outer image edges marking multiple copies of the icons. The center placing of the icons markers works correctly.  About 20% of the way in from the images edges are where the icons multi-copy on P4 for me. Kitharode mentioned that there is other project work ongoing that may or may not be the problem cause.  

My access is via Google Chrome, version Version 66.0.3359.181 (Official Build) (64-bit)    Google Chrome is up to date

    FYI & Thanks : )
    Bob

    Posted

  • mschwamb by mschwamb scientist, translator in response to TPS1BCW's comment.

    Thanks for letting me know. The development team might not be able to fix the issue since we're slowly moving to the Zooniverse project builder, but if the issue continues do let me know and I'll see what we can do.

    Thanks,
    `Meg

    Posted

  • geert3 by geert3

    I've classified a few images on planetfour in the past. I just got a mail and tried some more, but for some reason the "diamond shaped sides" of the fan do not highlight and when I click them a new fan is created at that spot, instead of widening the first fan. This prevents correct classification. Is this a regression, am I the only one having this problem? Am I doing something wrong?
    Thanks

    Posted

  • p.titchin by p.titchin

    Can happen when you are not getting the "mouse" indicator right on the lateral diamond. They seem to be more sensitive towards the edges of the image. I've been doing it for years, with literally thousands of classifications and have only found it an occasional minor annoyance. I can't offer any other advice, other than keep trying. }pete

    Posted

  • mschwamb by mschwamb scientist, translator in response to geert3's comment.

    Hi,
    Thanks for the feedback. If you give your operating system version and kind we can look into it. Are you using a tablet?

    Thanks,

    ~Meg

    Posted

  • Noharrr by Noharrr

    I just received a request from ZOO that Planet 4 needs help.

    However, when entering the Zooniverse main website:

    1. Planet 4 does not come up on the "all projects" list;
    2. Planet 4 does not come up when run through the "most help needed" filter;
    3. Planet 4 does not come up on the "Space" project tab.

    Now, I wonder why Planet 4 needs help?????????

    Someone needs to get Planet 4 back on the Zooniverse menu, stat.

    I enjoy working on Planet 4, but my only way to access it seems to be through my personal home page and "my recent projects." This does not exactly help P4 attract new volunteers! (PS: Mars COSMIC and Mars Terrains seem to be the only Mars projects currently coming up in the ZOO menus.)

    Just thot someone might like to know . . .

    Posted

  • mschwamb by mschwamb scientist, translator in response to Noharrr's comment.

    Hiya,

    Thanks for letting me know. I think that's because the Zooniverse is preparing to shut down the old platform that Planet Four runs on and so they jumped ahead of themselves and took the project off of those lists.

    Planet Four needs help because we want to get these images classified before the platform shutdowns and so that we can complete the analysis linking local dust storms to the number of fans and blotches

    Cheers,

    ~Meg

    Posted