rhu: (Default)
2009-05-28 03:35 pm

New way of sharing my music

Since several of you have complained that you don't read sheet music and my computer-generated samples are piano-only, I threw together a little Flash program that displays the lyrics in sync with the music. The UI is dead simple.

For now, I just did it for the first movement of Kabbalat Shabbat. Let me know if this is useful; if it is, I'll generate pages for the other movements.
rhu: (Default)
2009-05-28 03:35 pm

New way of sharing my music

Since several of you have complained that you don't read sheet music and my computer-generated samples are piano-only, I threw together a little Flash program that displays the lyrics in sync with the music. The UI is dead simple.

For now, I just did it for the first movement of Kabbalat Shabbat. Let me know if this is useful; if it is, I'll generate pages for the other movements.
rhu: (Default)
2009-05-05 08:46 pm
Entry tags:

Toolchain FAIL

Inspired by Eric's looming success with Kickstarter, I started to work on what I hope will be the final push to get my siddur (paryerbook) to the point where I could see whether enough people would be willing to buy a copy for it to be worth my ordering some.

Technical details cut for length. )

This is not acceptable.

So now I'm stuck seeing how quickly I can write a program that will do just enough typesetting and image placement to reconstruct my 172-page book so that I can actually, you know, print the silly thing. This is not a wheel that I really have time to re-invent right now.
rhu: (Default)
2009-05-05 08:46 pm
Entry tags:

Toolchain FAIL

Inspired by Eric's looming success with Kickstarter, I started to work on what I hope will be the final push to get my siddur (paryerbook) to the point where I could see whether enough people would be willing to buy a copy for it to be worth my ordering some.

Technical details cut for length. )

This is not acceptable.

So now I'm stuck seeing how quickly I can write a program that will do just enough typesetting and image placement to reconstruct my 172-page book so that I can actually, you know, print the silly thing. This is not a wheel that I really have time to re-invent right now.
rhu: (Default)
2007-03-13 02:19 pm
Entry tags:

SQL gripe

SQL is a standard language for expressing database design. It has the following useful constructs:

* A datum may be assigned the value NULL, which means (roughly) "there is no meaningful value to put here." NULL is not considered equal to any other value, even another NULL.

* A column (such as "the mayors of all the cities") may be constrained to have unique values ("no person can be the mayor of more than one city at the same time.")

But if you mark a column as unique, it won't allow more than one row to have NULL as its value, even though testing two NULLs for equality returns false. So there's no way to express "If there is a value here, it must be unique; but any number of rows may have an unknown or empty value at the same time."

Surely I'm not the first person to run into this problem. But I can't find any good solutions --- neither in my SQL books nor via Google.

Edited to add: Thanks to [livejournal.com profile] abbasegal for suggesting setting up a unique nonclustered index on a schemabound view.
rhu: (Default)
2007-03-13 02:19 pm
Entry tags:

SQL gripe

SQL is a standard language for expressing database design. It has the following useful constructs:

* A datum may be assigned the value NULL, which means (roughly) "there is no meaningful value to put here." NULL is not considered equal to any other value, even another NULL.

* A column (such as "the mayors of all the cities") may be constrained to have unique values ("no person can be the mayor of more than one city at the same time.")

But if you mark a column as unique, it won't allow more than one row to have NULL as its value, even though testing two NULLs for equality returns false. So there's no way to express "If there is a value here, it must be unique; but any number of rows may have an unknown or empty value at the same time."

Surely I'm not the first person to run into this problem. But I can't find any good solutions --- neither in my SQL books nor via Google.

Edited to add: Thanks to [livejournal.com profile] abbasegal for suggesting setting up a unique nonclustered index on a schemabound view.