On Apr 14, 2014, at 22:45, Colin A. Smith wrote: > On Apr 14, 2014, at 14:12, Christiaan Hofman <[email protected]> wrote: > >> I am not saying the colon is bad in search operations in general.I am saying >> it becomes a bad choice in this specific situation. The next colon is >> basically always basically always there, which makes it weird. > > In my opinion, it’s worse to have a confusing, counterintuitive syntactical > element (“=“), than a weird one (“:”), which I believe is true in this case. > >>>>> We also disagree on whether typing 3-4 additional characters in addition >>>>> to a copy/paste operation is worth it users in terms of the improvement >>>>> in clarity. >>>> >>>> You cannot copy/paste the UID and the prefix at the same time. This is not >>>> an option. >>> >>> Perhaps there was a miscommunication. Typing “CIT=“/“REF=“ vs. >>> “CITING:”/“CITEDBY:” is a difference of 3-4 characters. I would assume that >>> the vast majority of users not using a script would copy and paste the >>> Wok-Uid, which I believe we agreed would include the “WOS:” prefix. That’s >>> why I think with either syntax, users will still have a copy/paste >>> operation, leaving the sole difference being the 3-4 characters of >>> additional typing. >> >> I still think that's far too many, because it is a repeating thing, not a >> one-time thing. It's extra work without any real functional improvement. > > I think typing is by far the least rate limiting step of performing the > search, compared with (potentially) creating a new search group, copying the > Wok-Uid, then pasting it into the search field.
You are forgetting the very important issue of frequency. The whole point is that searching is the thing you do many times. Setting up the search group is something you do only a limited time. Typically once, ever. And that's why the search term needs to be efficient, much more than setting up the search group settings. Also, typing again is something you do many times. While figuring out how to do it is done only once, basically. That's why being efficient in typing search terms is much more important than easily being able to figuring out what it means. > Even if it was rate limiting, I don’t think the extra time is worth the loss > in clarity. Quite frankly, I just used it very few times for testing. And I was very quickly getting annoyed a lot by the very long text I had to type in the old format. > > It’s worth pointing out that the Google engineers had almost the exact same > choice as we’re making. When they chose to use “related:” to be the syntax > for finding pages similar to another page, they could have chosen “rel:”, as > is currently implemented in BibDesk. In the choice between 3 vs. 7 > characters, 7 clearly won out. I’d guess that the “related:” Google operation > is used several orders of magnitude more times every day than all of the WOK > search operations will ever be performed in the lifetime of BibDesk. If there > were a compelling reason to shorten the syntax anywhere, it would be there. > > Cheers, > > Colin The fact that they did does not mean it's the best choice. Moreover they have a different context, like user base and versatility of search operations. Anyway, I am not yet rejecting your ideas. I do think they have merit, and am trying out how they feel. But as I said, taking frequency of use really has to be taken into account. Christiaan ------------------------------------------------------------------------------ Learn Graph Databases - Download FREE O'Reilly Book "Graph Databases" is the definitive new guide to graph databases and their applications. Written by three acclaimed leaders in the field, this first edition is now available. Download your free book today! http://p.sf.net/sfu/NeoTech _______________________________________________ Bibdesk-develop mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/bibdesk-develop
