GSoC 2010 - Boost.Phoenix
Hi List, As the subject indicates I am interested in the Boost.Phoenix project for this years summer of code. At first, a little information on me. I am currently studying at University of Erlangen-Nuremberg in Germany. I am learning C++ since 2005. I have experience with Expression Templates. I used them as part of my bachelors thesis: http://www10.informatik.uni- erlangen.de/Publications/Theses/2009/Heller_BA09.pdf Some months ago I discovered the Boost.Proto library. I was trying to implement a DSEL for a GUI description language. This attempt can be found here: http://gitorious.org/phobos/main/trees/c014102b5aefb2592eef625047340fa92d854... However this approach was not continued, and I developed a spirit parser which gives more flexibility in that particular domain. I like the idea of Boost.Phoenix as it is right now. I think an implementation with Boost.Proto seems only natural. I searched the boost SVN repository but i couldn't find the prototypes you mentioned. A link to these would be helpful to be able further look into this project proposal. Regards, Thomas Heller
Thomas Heller wrote:
I like the idea of Boost.Phoenix as it is right now. I think an implementation with Boost.Proto seems only natural. I searched the boost SVN repository but i couldn't find the prototypes you mentioned. A link to these would be helpful to be able further look into this project proposal. Hi Thomas,
There are two prototypes: - Eric Niebler made one here : https://svn.boost.org/trac/boost/wiki/BoostPhoenix3 - I have a few on my HD currently but none of them are online; I'll fix that during the upcoming week. There are everythign but fully functionnal but explore variation on how to handle the extensibility of the library. -- ___________________________________________ Joel Falcou - Assistant Professor PARALL Team - LRI - Universite Paris Sud XI Tel : (+33)1 69 15 66 35
On Friday 19 March 2010 10:42:23 Joel.Falcou@lri.fr wrote:
Thomas Heller wrote:
I like the idea of Boost.Phoenix as it is right now. I think an implementation with Boost.Proto seems only natural. I
searched the
boost SVN repository but i couldn't find the prototypes you mentioned. A link to these would be helpful to be able further look into this project proposal.
Hi Thomas,
There are two prototypes: - Eric Niebler made one here : https://svn.boost.org/trac/boost/wiki/BoostPhoenix3 - I have a few on my HD currently but none of them are online; I'll fix that during the upcoming week. There are everythign but fully functionnal but explore variation on how to handle the extensibility of the library.
Hi Joel, Thank you for the wiki url. I will dive into that. Some general questions: - The interface should be (more or less) the same as phoenix V2? - The main extention point for a user should be phoenix::functions? - What exactly to you want to have in the proposal? function/class interfaces? general description of phoenix? Regards, Thomas
Thomas Heller wrote:
Hi Joel,
Thank you for the wiki url. I will dive into that. Some general questions: - The interface should be (more or less) the same as phoenix V2?
It has to be the same to minimize disruption. If anythign need to be changed for accomodating some inenr changes, it'll have to be discussed.
- The main extention point for a user should be phoenix::functions?
we wanted to have more extension scenario, like simplifying how to make new actor (like let's say I want to add new control structure or new palceholder-like actor with special member). The problem is in fact hidign proto internals in a pretty way so user wntign to extend don't have to master Proto.
- What exactly to you want to have in the proposal? function/class interfaces? general description of phoenix? Good question. Let's see what Joel or Harmut have to say on the matter.
-- ___________________________________________ Joel Falcou - Assistant Professor PARALL Team - LRI - Universite Paris Sud XI Tel : (+33)1 69 15 66 35
On 3/21/2010 1:07 AM, joel falcou wrote:
Thomas Heller wrote:
Hi Joel,
Thank you for the wiki url. I will dive into that. Some general questions: - The interface should be (more or less) the same as phoenix V2? It has to be the same to minimize disruption. If anythign need to be changed for accomodating some inenr changes, it'll have to be discussed. - The main extention point for a user should be phoenix::functions? we wanted to have more extension scenario, like simplifying how to make new actor (like let's say I want to add new control structure or new palceholder-like actor with special member). The problem is in fact hidign proto internals in a pretty way so user wntign to extend don't have to master Proto. - What exactly to you want to have in the proposal? function/class interfaces? general description of phoenix? Good question. Let's see what Joel or Harmut have to say on the matter.
To be honest, I am not entilrely sure I understand the question. Perhaps I lost the context here. Could you elaborate? Do you mean to ask about the scope of the proposal? Regards, -- Joel de Guzman http://www.boostpro.com http://spirit.sf.net http://www.facebook.com/djowel Meet me at BoostCon http://www.boostcon.com/home http://www.facebook.com/boostcon
On Sunday 21 March 2010 04:06:25 Joel de Guzman wrote:
On 3/21/2010 1:07 AM, joel falcou wrote:
Thomas Heller wrote:
Hi Joel,
Thank you for the wiki url. I will dive into that. Some general questions: - The interface should be (more or less) the same as phoenix V2?
It has to be the same to minimize disruption. If anythign need to be changed for accomodating some inenr changes, it'll have to be discussed.
- The main extention point for a user should be phoenix::functions?
we wanted to have more extension scenario, like simplifying how to make new actor (like let's say I want to add new control structure or new palceholder-like actor with special member). The problem is in fact hidign proto internals in a pretty way so user wntign to extend don't have to master Proto.
- What exactly to you want to have in the proposal? function/class interfaces? general description of phoenix?
Good question. Let's see what Joel or Harmut have to say on the matter.
To be honest, I am not entilrely sure I understand the question. Perhaps I lost the context here. Could you elaborate? Do you mean to ask about the scope of the proposal?
Let me try to rephrase one part of my question. What do you expect from me as a GSoC Student? I think I read through all the previous discussions, articles and documentation about phoenix and proto. Let me try to sum everything up: - Some basic building blocks (actors, statements?, functions?, ...) - Easy extendable without exposing to much proto specific stuff I hope i got that right, more or less. Right now i am waiting to take a look at Joel's (Falcou) prototypes to get some more insights on what you already have done. The other part of my question was more like: What should be actually put into the formal proposal. Just an overall description? Something like pseudo/C++ interfaces and/or first implementation ideas? Regards, Thomas
On 3/21/2010 3:56 PM, Thomas Heller wrote:
- What exactly to you want to have in the proposal? function/class interfaces? general description of phoenix?
Good question. Let's see what Joel or Harmut have to say on the matter.
To be honest, I am not entilrely sure I understand the question. Perhaps I lost the context here. Could you elaborate? Do you mean to ask about the scope of the proposal?
Let me try to rephrase one part of my question. What do you expect from me as a GSoC Student?
I think I read through all the previous discussions, articles and documentation about phoenix and proto. Let me try to sum everything up: - Some basic building blocks (actors, statements?, functions?, ...) - Easy extendable without exposing to much proto specific stuff I hope i got that right, more or less.
That is correct. I think this one will be unique in that we don't expect you to implement everything and I doubt if that is even possible given the time frame. We, Joel and I (Man, that will be confusing having tow Joels), intend to have close collaboration with the student. We expect you to be working closely with us in implementing the above.
Right now i am waiting to take a look at Joel's (Falcou) prototypes to get some more insights on what you already have done.
The other part of my question was more like: What should be actually put into the formal proposal. Just an overall description? Something like pseudo/C++ interfaces and/or first implementation ideas?
I think an overall description would suffice at this point. Regards, -- Joel de Guzman http://www.boostpro.com http://spirit.sf.net http://www.facebook.com/djowel Meet me at BoostCon http://www.boostcon.com/home http://www.facebook.com/boostcon
Joel de Guzman wrote:
I think this one will be unique in that we don't expect you to implement everything and I doubt if that is even possible given the time frame. Exactly We, Joel and I (Man, that will be confusing having tow Joels) Could be worse, my 2nd name could have been Eric :p
In anyway, don't overdo the proposal I think, it can't anything else than some rough work plan and structuration IMHO. -- ___________________________________________ Joel Falcou - Assistant Professor PARALL Team - LRI - Universite Paris Sud XI Tel : (+33)1 69 15 66 35
On Sunday 21 March 2010 09:09:07 joel falcou wrote:
Joel de Guzman wrote:
I think this one will be unique in that we don't expect you to implement everything and I doubt if that is even possible given the time frame.
Exactly
Great, that was one of the points which gave me some headaches
In anyway, don't overdo the proposal I think, it can't anything else than some rough work plan and structuration IMHO.
Alright, that sounds good. I will definetly apply! I am very happy to discuss further details with you. I think i can get a first draft of my proposal by the end of the day.
Hi Thomas and Joels,
In anyway, don't overdo the proposal I think, it can't anything else than some rough work plan and structuration IMHO.
Alright, that sounds good. I will definetly apply! I am very happy to discuss further details with you. I think i can get a first draft of my proposal by the end of the day.
I think in the past, we've had students actually submit or send links to their applications in order to get feedback before submitting them officially. Maybe the Joels would be willing to look at an early draft or two? Andrew Sutton andrew.n.sutton@gmail.com
On 3/21/2010 11:06 PM, Andrew Sutton wrote:
Hi Thomas and Joels,
In anyway, don't overdo the proposal I think, it can't anything else than some rough work plan and structuration IMHO.
Alright, that sounds good. I will definetly apply! I am very happy to discuss further details with you. I think i can get a first draft of my proposal by the end of the day.
I think in the past, we've had students actually submit or send links to their applications in order to get feedback before submitting them officially. Maybe the Joels would be willing to look at an early draft or two?
Sure. Email 'em in, Thomas. Regards, -- Joel de Guzman http://www.boostpro.com http://spirit.sf.net http://www.facebook.com/djowel Meet me at BoostCon http://www.boostcon.com/home http://www.facebook.com/boostcon
On Monday 22 March 2010 02:39:41 Joel de Guzman wrote:
On 3/21/2010 11:06 PM, Andrew Sutton wrote:
Hi Thomas and Joels,
In anyway, don't overdo the proposal I think, it can't anything else than some rough work plan and structuration IMHO.
Alright, that sounds good. I will definetly apply! I am very happy to discuss further details with you. I think i can get a first draft of my proposal by the end of the day.
I think in the past, we've had students actually submit or send links to their applications in order to get feedback before submitting them officially. Maybe the Joels would be willing to look at an early draft or two?
Sure. Email 'em in, Thomas.
Regards,
Took a little longer than expected. I appreciate your help. Attached is a first draft of my application. Thanks, Thomas
participants (5)
-
Andrew Sutton -
Joel de Guzman -
joel falcou -
Joel.Falcou＠lri.fr
-
Thomas Heller