Profile_bird

Hey there! jonpryor is using Twitter.

Twitter is a free service that lets you keep in touch with people through the exchange of quick, frequent answers to one simple question: What's happening? Join today to start receiving jonpryor's tweets.

Already using Twitter
from your phone? Click here.

Default_profile_0_bigger
jonpryor

  1. So it looks like the extension is needed, despite it working as I'd expect under xsltproc and the fragment technique being listed in a book.
  2. Damn; http://www.w3.org/TR/xslt: In particular, it is not permitted to use the /, //, and [] operators on result tree fragments.
  3. I shouldn't need an MS XSLT extension to convert a result tree into a node set, which works w/o extensions under Mono, xsltproc, etc.
  4. .NET's System.Xml looks brain-damaged compared to Mono's (msxsl:nodeset()? wtf?). Porting mdoc to .NET will be less trivial than I thought.
  5. .@trampstr .NET XslCompiledTransform is much faster than XslTransform: 1.74s for .NET vs. ~1.5s for Mono. I'll have to update mdoc now...
  6. I feel like I've lost hours today trying to make sure VS actually does a clean build so my tests run properly...
  7. Things I didn't know: in SQLite, primary keys must be INTEGER, not "int", to permit auto-generating the primary key.
  8. @JustinEtheredge Yes, but their collection support is type-specifc, and thus can't support user-defined types. :-(
  9. @JustinEtheredge So they're catching up to C# (though C# collection initializers are still better), though underscores in numbers is nice.
  10. Though, for comparison's sake, xsltproc processes the same files in 0.218s.
  11. Some say mono is slow. For XSLT, not true. 'mdoc export-html' on mono: 0:01.52 (< 2s); .NET: 3:07.68 (> 3 minutes)
  12. @KevinHazzard FxCop can't, as IL doesn't have that info. StyleCop could do it...
  13. @KevinHazzard Won't happen, as it's a de-facto breaking change (which will cause mountains of code to break).
  14. @KevinHazzard Plus, I suspect that there's very rarely a "future-proof" true/false condition, so enums are "safer"...
  15. @KevinHazzard Yes, at least until Resharper always requires use of named parameters for boolean parameters...
  16. wrt #openSUSE 11.2, Evolution is nice (Google calendar integration!), but #metacity sucks: window Z order is odd (can't raise windows), etc.
  17. @ermau I'm still thinking Cadenza, but what do you think of Intermezzo - "short melodies that connect the main parts of a composition."
  18. @sandyarmstrong ...and will likely waste the dev's time, as if/when these patches are submitted upstream they'll need further review.
  19. @sandyarmstrong So Google is wasting Google's time (by re-fixing already fixed bugs)...
  20. @sandyarmstrong It's because Linux isn't using the most up-to-date code, and is reimplementing the wheel.