Return to Answer

"I would really like to learn this stuff". If you are long-term serious:

• Go to college, specialize in software engineering. Take every compiler class you can get. Those people providing the classes are better educated and more experienced than you; its good to have their expert perspectives used to present the information to you in ways you'll never get from reading code.

• Stick with math classes through high school and continue in college for all 4 years. Focus on non-standard math: logics, group theory, meta-mathematics. This will force you to think abstractly. It will enable you to read the advanced theory papers on compiling and understand why those theories are interesting and useful. You can ignore those advanced theories, if you forever want to be behind the state of the art.

• Collect/read the standard compiler texts: Aho/Ullman, etc. They contain what the community generally agrees is fundamental stuff. You might not use everything from those books, but you should know it exists, and you should know why you aren't using it. I thought Muchnick was great, but it is for pretty advanced topics.

• Build a compiler. Start NOW by building a rotten one. This will teach you some issues. Build a second one. Repeat. This experience builds huge synergy with your book learning.

• A really good place to start is to learn about BNF (Backus Normal Form), parsers, and parser-generators. BNF is effectively universally used in compiler land, and you can't realistically talk to your fellow compiler-types if you don't know it.

If you want a great first introduction to compiling, and the direct value of BNF not for just documentation but as a tool-processable metalanguage, see this tutorial (not mine) on building "meta" compilers (compilers that build compilers) based on a paper from 1964 (yes, you read that right) ["META II a syntax-oriented compiler writing language" by Val Schorre. (http://doi.acm.org/10.1145/800257.808896)] This IMHO is one of the single best comp-sci papers ever written: it teaches you to build compiler-compilers in 10 pages. I learned initially from this paper.

What I wrote about above is a lot from personal experience, and I think it has served me pretty well. YMMV, but IMHO, not by much.

"I would really like to learn this stuff". If you are long-term serious:

• Go to college, specialize in software engineering. Take every compiler class you can get. Those people providing the classes are better educated and more experienced than you; its good to have their expert perspectives used to present the information to you in ways you'll never get from reading code.

• Stick with math classes through high school and continue in college for all 4 years. Focus on non-standard math: logics, group theory, meta-mathematics. This will force you to think abstractly. It will enable you to read the advanced theory papers on compiling and understand why those theories are interesting and useful. You can ignore those advanced theories, if you forever want to be behind the state of the art.

• Collect/read the standard compiler texts: Aho/Ullman, etc. They contain what the community generally agrees is fundamental stuff. You might not use everything from those books, but you should know it exists, and you should know why you aren't using it. I thought Muchnick was great, but it is for pretty advanced topics.

• Build a compiler. Start NOW by building a rotten one. This will teach you some issues. Build a second one. Repeat. This experience builds huge synergy with your book learning.

• A really good place to start is to learn about BNF (Backus Naur Form), parsers, and parser-generators. BNF is effectively universally used in compiler land, and you can't realistically talk to your fellow compiler-types if you don't know it.

If you want a great first introduction to compiling, and the direct value of BNF not for just documentation but as a tool-processable metalanguage, see this tutorial (not mine) on building "meta" compilers (compilers that build compilers) based on a paper from 1964 (yes, you read that right) ["META II a syntax-oriented compiler writing language" by Val Schorre. (http://doi.acm.org/10.1145/800257.808896)] This IMHO is one of the single best comp-sci papers ever written: it teaches you to build compiler-compilers in 10 pages. I learned initially from this paper.

What I wrote about above is a lot from personal experience, and I think it has served me pretty well. YMMV, but IMHO, not by much.

"I would really like to learn this stuff". If you are long-term serious:

• Go to college, specialize in software engineering. Take every compiler class you can get. Those people providing the classes are better educated and more experienced than you; its good to have their expert perspectives used to present the information to you in ways you'll never get from reading code.

• Stick with math classes through high school and continue in college for all 4 years. Focus on non-standard math: logics, group theory, meta-mathematics. This will force you to think abstractly. It will enable you to read the advanced theory papers on compiling and understand why those theories are interesting and useful. You can ignore those advanced theories, if you forever want to be behind the state of the art.

• Collect/read the standard compiler texts: Aho/Ullman, etc. They contain what the community generally agrees is fundamental stuff. You might not use everything from those books, but you should know it exists, and you should know why you aren't using it. I thought Muchnick was great, but it is for pretty advanced topics.

• Build a compiler. Start NOW by building a rotten one. This will teach you some issues. Build a second one. Repeat. This experience builds huge synergy with your book learning.

• A really good place to start is to learn about BNF (Backus Normal Form), parsers, and parser-generators. BNF is effectively universally used in compiler land, and you can't realistically talk to your fellow compiler-types if you don't know it.

If you want a great first introduction to compiling, see this tutorial (not mine) on building "meta" compilers (compilers that build compilers) based on a paper from 1964 (yes, you read that right) ["META II a syntax-oriented compiler writing language" by Val Schorre. (http://doi.acm.org/10.1145/800257.808896)] This IMHO is one of the single best comp-sci papers ever written: it teaches you to build compiler-compilers in 10 pages. I learned initially from this paper.

What I wrote about above is a lot from personal experience, and I think it has served me pretty well. YMMV, but IMHO, not by much.

"I would really like to learn this stuff". If you are long-term serious:

• Go to college, specialize in software engineering. Take every compiler class you can get. Those people providing the classes are better educated and more experienced than you; its good to have their expert perspectives used to present the information to you in ways you'll never get from reading code.

• Stick with math classes through high school and continue in college for all 4 years. Focus on non-standard math: logics, group theory, meta-mathematics. This will force you to think abstractly. It will enable you to read the advanced theory papers on compiling and understand why those theories are interesting and useful. You can ignore those advanced theories, if you forever want to be behind the state of the art.

• Collect/read the standard compiler texts: Aho/Ullman, etc. They contain what the community generally agrees is fundamental stuff. You might not use everything from those books, but you should know it exists, and you should know why you aren't using it. I thought Muchnick was great, but it is for pretty advanced topics.

• Build a compiler. Start NOW by building a rotten one. This will teach you some issues. Build a second one. Repeat. This experience builds huge synergy with your book learning.

• A really good place to start is to learn about BNF (Backus Normal Form), parsers, and parser-generators. BNF is effectively universally used in compiler land, and you can't realistically talk to your fellow compiler-types if you don't know it.

If you want a great first introduction to compiling, and the direct value of BNF not for just documentation but as a tool-processable metalanguage, see this tutorial (not mine) on building "meta" compilers (compilers that build compilers) based on a paper from 1964 (yes, you read that right) ["META II a syntax-oriented compiler writing language" by Val Schorre. (http://doi.acm.org/10.1145/800257.808896)] This IMHO is one of the single best comp-sci papers ever written: it teaches you to build compiler-compilers in 10 pages. I learned initially from this paper.

What I wrote about above is a lot from personal experience, and I think it has served me pretty well. YMMV, but IMHO, not by much.

"I would really like to learn this stuff". If you are long-term serious:

• Go to college, specialize in software engineering. Take every compiler class you can get. Those people providing the classes are better educated and more experienced than you; its good to have their expert perspectives present the information to you in ways you'll never get from reading code.

• Stick with math classes through high school and continue in college for all 4 years. Focus on non-standard math: logics, group theory, meta-mathematics. This will force you to think abstractly. It will enable you to read the advanced theory papers on compiling and understand why those theories are interesting and useful. You can ignore those advanced theories, if you forever want to be behind the state of the art.

• Collect/read the standard compiler texts: Aho/Ullman, etc. They contain what the community generally agrees is fundamental stuff. You might not use everything from those books, but you should know it exists, and you should know why you aren't using it. I thought Muchnick was great, but it is for pretty advanced topics.

• Build a compiler. Start NOW by building a rotten one. This will teach you some issues. Build a second one. Repeat. This experience builds huge synergy with your book learning.

• A really good place to start is to learn about BNF (Backus Normal Form), parsers, and parser-generators. BNF is effectively universally used in compiler land, and you can't realistically talk to your fellow compiler-types if you don't know it.

If you want a great first introduction to compiling, see this tutorial (not mine) on building "meta" compilers (compilers that build compilers) based on a paper from 1964 (yes, you read that right) ["META II a syntax-oriented compiler writing language" by Val Schorre. (http://doi.acm.org/10.1145/800257.808896)] This IMHO is one of the single best comp-sci papers ever written: it teaches you to build compiler-compilers in 10 pages. I learned initially from this paper.

What I wrote about above is a lot from personal experience, and I think it has served me pretty well. YMMV, but IMHO, not by much.

"I would really like to learn this stuff". If you are long-term serious:

• Go to college, specialize in software engineering. Take every compiler class you can get. Those people providing the classes are better educated and more experienced than you; its good to have their expert perspectives used to present the information to you in ways you'll never get from reading code.

• Stick with math classes through high school and continue in college for all 4 years. Focus on non-standard math: logics, group theory, meta-mathematics. This will force you to think abstractly. It will enable you to read the advanced theory papers on compiling and understand why those theories are interesting and useful. You can ignore those advanced theories, if you forever want to be behind the state of the art.

• Collect/read the standard compiler texts: Aho/Ullman, etc. They contain what the community generally agrees is fundamental stuff. You might not use everything from those books, but you should know it exists, and you should know why you aren't using it. I thought Muchnick was great, but it is for pretty advanced topics.

• Build a compiler. Start NOW by building a rotten one. This will teach you some issues. Build a second one. Repeat. This experience builds huge synergy with your book learning.

• A really good place to start is to learn about BNF (Backus Normal Form), parsers, and parser-generators. BNF is effectively universally used in compiler land, and you can't realistically talk to your fellow compiler-types if you don't know it.

If you want a great first introduction to compiling, see this tutorial (not mine) on building "meta" compilers (compilers that build compilers) based on a paper from 1964 (yes, you read that right) ["META II a syntax-oriented compiler writing language" by Val Schorre. (http://doi.acm.org/10.1145/800257.808896)] This IMHO is one of the single best comp-sci papers ever written: it teaches you to build compiler-compilers in 10 pages. I learned initially from this paper.

What I wrote about above is a lot from personal experience, and I think it has served me pretty well. YMMV, but IMHO, not by much.