Jump to content

deanishe

Member
  • Posts

    8,759
  • Joined

  • Last visited

  • Days Won

    522

Everything posted by deanishe

  1. I very much doubt an API will be made available. The data stored on the server are encrypted. They're encrypted on your machine before upload, and the service can't access them. That might change, as one of the deeply stupid things Panic did when making Panic sync was to derive the encryption key from your sync account password, so it's possible to build an API. Indeed, in face of the sheer amount of stupidity that went into designing Panic Sync, it's perfectly possible they'll implement decryption-on-the-server and add an API.
  2. Please post the actual workflow if you'd like help. Without it, we're basically just guessing, which wastes your and our time.
  3. Paste this URL into Alfred (it's a custom search with keyword "gy"): alfred://customsearch/Search%20GoYangLib%20for%20%27%7Bquery%7D%27/gy/utf8/nospace/http%3A%2F%2Fwww.goyanglib.or.kr%2Fcenter%2Fdata%2Fsearch.asp%3Fa_q%3D%7Bquery%7D
  4. Why do you need to press ↩ between entering keyword and query?
  5. If you result points to an image file, you can Quicklook it with SHIFT.
  6. Yes, it's possible. Quicksilver does it. Here's the relevant plugin.
  7. You should probably mention that users first need to install Node (and therefore possibly Homebrew). Also that npm install --global may well break existing software.
  8. I know, and I'd rather Alfred didn't do that. It makes it so that you can't select a directory directly, only browse one. It's also inconsistent: if I pass a filepath, Alfred selects it. If I pass a directory path, Alfred enters it. If I want to browse the directory, I could append the trailing slash, just like I would if I were typing the query into Alfred.
  9. Most of it, yeah. The workflow I use also has a "slugify" mode, which can't be done (easily) within Alfred. In any case, it's generally easier to write two lines of code that add a bunch of Alfred elements, and you can't show a "live preview" in a Script Filter, either.
  10. Could you also make queries via AppleScript work the same way as they do in Alfred? I mean, if I enter "~/Desktop" into Alfred, it filters the contents of ~/ by keyword "Desktop". If I enter "~/Desktop/" (with a trailing slash), Alfred shows the contents of ~/Desktop. If I pass a path via AppleScript, and it's a directory that exists, Alfred adds the trailing slash automatically and shows me the contents of ~/Desktop. It'd be useful if Alfred didn't mess with the query, so I can enter "~/Desktop" to remain in ~/, but with the folder "Desktop" selected, or I can enter "~/Desktop/" (with a trailing slash) to enter the directory in Alfred.
  11. Alfred Preferences > Features > Contacts > Advanced > [X] Use Spotlight metadata for searching contacts That should do what you want.
  12. The main thing that sets Alfred apart from Spotlight is its speed. That come largely from not including loads of stuff in the default search. Typing a prefix should be expected except for the corest of core searches (apps, System Preference panes, not a lot else…) It's very hard for Alfred (or any other 3rd-party app) to override bugs in Finder (or any other app)…
  13. Can the code not be more-or-less copy-pasted? Is some massive JS build-pipeline really required? How big, exactly, would that binary be? I mean, does it include Node? More generally, the "global install" method (in this case, using "npm install -g") is not a good idea. In terms of Alfred workflows, it means your workflow doesn't sync across machines like well-behaved (self-contained) workflows. In more general Node terms (though the same applies to Ruby and Python), it sets your workflow up to break/be broken by other software, as you can only install one version of a library at a time globally. You probably don't want to open yourself up to many years of support requests/bug reports because your globally-installed program requires a different version of library X to program Y.
  14. It's probably this issue: https://github.com/deanishe/alfred-ssh/issues/4#issuecomment-304690963 I must have forgotten to release a fix
  15. Node is a pretty hardcore dependency for such a simple workflow. Could this work as JXA? That way, it would run on every Mac.
  16. Have you tried Tampermonkey or does that not work any more?
  17. Try repos workflow:delcache to reset the cached list of versions.
  18. Can't you release the theme you made for the PinPlus GUI as a userscript?
  19. The way I usually handle this (my libraries support it "natively") is to show the cached data and update them in a background process if they're out of date. If an update process is running, I set rerun=0.5 in the Script Filter feedback. Often, you're working with the new dataset before you've finished typing, or the results are updated a second later. I think it's much more elegant than setting a launch agent in most cases. With a launch agent, it's tough to find a balance between freshness of data and needlessly hammering the server. And the ideal update interval is going to be different for different users.
  20. Hmm. On my machine, the command returns when I close the window it opened (i.e. I can leave other MacVim windows open and the application running).
×
×
  • Create New...