I was copying out an link address to a design in the new lab view. It just took me a second to realize that I’m wishing for your elegant solve on copying out the design sequence.
I simply wish the OK button to be like your dismiss button for sequence copying and just fade away in the background without us having to click on it to make it disappear. This will help a lot.
Another small note. These two links, the design URL and the one to all the players designs URL are not underlined, making it harder to see they are links.
I love that you got the shortcuts for meltplot and dotplot shortcuts really working.
(Background details: Can be called with shortcut (S) and minimised with (M). If there is no reaction, just click in the lab design and the all the keyboard shortcuts should be activated. This goes as a rule for all shortcuts.)
Before the shortcuts for meltplot/dotplot has never really worked, because one had to open the dot plot/meltplot manually first to get the keyboard shortcuts working, which kind of looses the point.
Found one small thing. I can manage to get both the minimized and big plot showing at once. No biggie. For my sake it can wait till after the other bigger issues. Basically the function works, which it didn’t before. So big thx from here.
I believe you haven’t touched that function, as this is an old returning problem and has bugged us for long. In the hope it can help I can add that Johan’s microRNA and MS2 switch labs are working fine not only when active, but also when closed. It is the older labs that are touched by this problem of not allowing eterna flash users to stay in flash as soon as the lab goes to the archive. Hope this can help you pinpoint the difference between past and new labs.
Jieux brought up in chat that the base pair numbers were gone in the base picking tool in the updated version. NOVA made their version for puzzles and made it simplified as to make it easy for players all new to eterna. However for lab work, I very much want these numbers and information back directly in the puzzle where I need it.
Yes, I can see the number of base pairs in the lab list - if or after the design has been submitted. But I will like that kind of info directly in the design when I’m looking at it, not having to go look it up elsewhere or counting manually when designing or voting.
@Eli: as I just mentioned in chat, I don’t think I will be changing any of this. In puzzles, if there’s a constraint, then the information is there. In labs, you are one hotkey (s) or one click away from the information. In conclusion, I see no good justification to going back to cluttering the interface.
As for the puzzlemaker, hoglahoo had a good idea about the publishing dialog, and I may try to implement that one.
@Nando, I do understand you want it clean. But this specific clutter is useful info. Its far nicer each time one don’t have to click to get to something. It quickly adds up to a lot of clicking. I think the old interface was more useful at this specific point.
if I have to weigh between being able to click the pairs tools more easily, or having the numbers of pairs information immediately accessible, the choice has to be the ease of use for pair tools.
The one easy to fall in trap is to add and add and add more information in an interface and clutter it. This palette is an action bar. It should have tools, and markers to show which one is the currently active one. This by itself is already taking a very large portion of the space. Trying to squeeze statistics there is unwise, specially when it’s not really necessary.
I think the Nova guys made a wise choice here, and I’m going to stick to it, unless someone can prove to me that it’s isn’t as good as I think.
@jandersonlee: so “useful” and indispensable that it takes 5 days (15 if I count the public preview, 30 if I count the closed previews, which Eli had access to) for anyone to notice they’re missing…
This said, there might be enough space where you’re suggesting, so I’ll look into it, but no promises.
The numbers are replaced by the “pairs” selector, which, in turn, is much less visible than highlighting the nucleotides two at a time. It also shouldn’t matter where you click between the two nucleotides. That is, whether you click the icon at a height right between the center of the circles or at a height just above the lower border of the icon.
With that, the pair icons become completely irrelevant. Their only remaining use would be the highlight, which was much better done in the old interface (the highlight was about 9 times the current size).
Is it a case of the pair selectors acting as separate buttons on an action bar and the highlights having to be constrained within that button?
/* rant
As for the 5/15/30 days… I really shouldn’t bother, but… It wasn’t even Eli who spotted this. It might have taken him a whole lot longer to be honest. But it doesn’t matter either. All it takes is just that one occasion when we would like to know. Eli picked it up for the same reasons I would have. I found it important and useful before. Which means I don’t want to lose it on account of someone just saying “it shouldn’t be that way, because we’ve (re)designed it this way”.
He also took it up to make sure it gets mentioned. Because, let’s face it, most people won’t come here to post. Even if they noticed it 30 days ago. On that note though, I don’t think a lot of people were actually interested in either the closed or the public preview. Especially since we’ve seen this interface before. Most people don’t care until they have to actually cope. You can say it’s wrong all you want but that attitude is not going to change the situation.
It’s much more befitting a group of scientists to plan ahead and go live with feedback and time allocated. To work together to create something everyone is happy about.
rant */
On a different note, I have regularly wished that I could choose between having base pair numbers shown or their percentages.
The reason why, is that when I design, - then I usually aim for a good range for base pair percentages as I have seen it in other similar shaped labs. However those similar labs don’t always hold exactly the same number of base pairs. Which makes the base pair percentages a more useful measure of similarity than base pair count.
So having the base pair percentage would allow me to shoot closer to the range I want, either from looking at past labs or from experience. Instead of having to judge what base pair percentages estimates to what base pair count.
We get both percentage and number of base pairs in the lab list. I wish to be able to get both these informations in the lab design - not at once - but by choice. Preferably by being able to change settings between seeing percentage or base pair count and have it shown directly in the base picking tool.
Demonstration in lab tutorial of use of base percentages for shooting for a good range of putting ones design in and how I have been finding it:
@Eli: I tested jandersonlee’s idea on my dev instance, see for instance http://nando.eternadev.org/game/puzzl… and since it looks somewhat acceptable, it will be pushed to prod soon. This said, I remain completely unconvinced about the usefulness of this information, and whether it is primordial or not to have it directly available in the solving/designing interface. I still feel this is a bad idea.
Your additional request about its format (counts vs percentages) will go on my feature request list. At the moment, there are way too many other “construction sites” and I really need to focus on them in the near future.
@Nando, I like your test out of jandersonlee’s idea. Big thx for listening and even picking up the idea with base percentages. This is sweet news as it is something I have been wishing for, for a long time. I’m apologizing for the clutter.
Past lab
Old browser
Possible oversight in the UI for labs that are Waiting for synthesis results. like http://eterna.cmu.edu/game/browse/573… http://prntscr.com/6zk01m
Could you remove the box that shows how many Votes and solutions that a player have left, Once the round has closed that box is no longer needed ?
hi Mat, and thanks for posting this here, this is indeed preferable to PMs
About the Flash-based design browser (what you call ‘old’), I’m quite uncertain of whether the applet has ever been informed of the status of the round. You are very correct that the information is useless, once the round is closed, and that it looks quite silly and ugly, specially when the chat is missing too. But I need to see with John how this can be properly signalled to the applet. I’ve noted this on my todo list.
The web-based browser (which you call ‘new’) is all under John’s control. This is a bit off-topic here, but I’ll inform him.
On http://eterna.cmu.edu/game/puzzle/583…, when entering sequence
AUUACGUACGACAUGAGGAUCACCCAUGUAAUUCGAUCAAUCAAGCUUUCUAUUUAAUGAAACAUACGUACUAGGUAACAUGAGGAUCACCCAUGUGGGUAUGAGU
All conditions appear normal, and state 1 and 2 look good, but that little face on the left top says that the MS2 hairpin does not form. Is that intended? Or shouldn’t one of the two states at the right at least look a little bit red?
actually, since it appeared again in
GUUGCGUGCGACAUGAGGAUCACCCAUGUACUUCGCUCAGUCACGCUUUCUGUUUACUGCCCUAUUCGUACUUGGUUACAUGAGGAUCACCCAUGUAGGUUUGUGU
I presume this is intentional and not a bug. Is this smiley face the replacement for the “NO” state before? IF so, it’ll be hard for new folks to grasp the idea, I think, but that’s ok, leave it, I like the graphics.
While I don’t know anything about the final UI for Eterna-cell and the future, scripting capabilities will eventually be integrated into the game.
If it has not already been proposed or given consideration, it may be useful to code another popup dialogue box or window (or link it to the current pairing probabilities plot window) that contains information about the RNA sequence, based on a few default parameters as well as other statistics that a player may want generated with custom scripts. This could include scoring information, base pair counts, % of base pairs, relative distribution of base pairs throughout the secondary structure, secondary structure composition, energy barriers between targets, etc.
The basic infrastructure of the Flash applet only has support for a few simple constraints. For most puzzles, this is plenty enough, but specially in labs, there are cases where this is insufficient. For instance, in the specific context you mention, while it’s very easy to specify that both MS2 hairpins should form in a specific state, it is not possible to require that *neither* MS2 hairpins be formed.
We had two possible paths to fix that problem. Either “complexify” the applet to make it support a variety of new exotic, but rare contraints. Which comes at a high price: the code becomes more prone to bugs, code maintenance and implementing new features becomes harder (because one need to be careful not to break the existing feature set). In short, the code becomes more monolithic, and “heavier” (not only to manage, but the binary size also grows with routines very rarely used)
The other route was to flexibly expand the applet’s capabilities by making use of EternaScripts. This seemed to me a lot more future-oriented and light-weight. It also comes at a price: first encounters may be disconcerting for players. So far though, it didn’t seem to cause too much trouble. I would guess that the red outline prompts users to hover the mouse over the constraint and read the error message.
Some improvements could certainly be considered. For instance, in the case of a scripted constraint with structural requirements (like the one you’re referring to), we could imagine that the script could eventually point out which bases are “faulty” in the puzzle, by way of some highlighting maybe. It’s on my todo list, don’t expect it very soon though (sorry)
