Hello, I'm a CS student at the University of Patras in Greece, and I'm really interested in doing my thesis on programming language theory, compilers JIT optimization etc. I find your project really interesting and I'd like you to advice me if there is any potential for my thesis here. A contribution maybe. Is it feasible? Thanks in advance for all answers and considerations, George -- George 'papanikge' Papanikolaou
Hi there, I finished my thesis recently implementing a vectorizing optimization. So I guess it is feasible :). My advice would be to build the VM and make some baby steps in the test environment. This might sound weird, but the development is very different to what you do at university (at least in austria) and also very different to most projects I know. I think there are many possible ways you could improve PyPy. It is hard to give some advice on exactly what you should implement. After all you need to be motivated until you finish your thesis. It would probably be better to think about a topic (VM X does Y), see if PyPy implemented Y (you can of course ask us, it is hard to get around in such a big project in the beginning), (if not) reflect if it gives any benefit for the VM and then implement it. Here is a list of potential projects (might get you started): http://pypy.readthedocs.org/en/latest/project-ideas.html What really would help PyPy (might be more engineering than a 'researchy' topic) is to implement the Python 3 spec (e.g. 3.3,3.4,3.5). You can also join IRC and we will see if we can help to choose a topic. Cheers, Richard On 10/30/2015 06:46 PM, George Papanikolaou wrote:
Hi, First of all thanks for answering, yes I'll pop on IRC soon. My thesis will be for an undergraduate degree (https://en.wikipedia.org/wiki/Engineer%27s_degree#Greece) so I guess I can go for an engineering project more than a research-y one. How hard would implementing the 3.0 spec be? It sounds tedious but pretty standard. I'll get started right away with your advice in order to get familiar with the VM, the environment and coding nuisances in general. Regards, George On Sat, Oct 31, 2015 at 11:57 AM, Richard Plangger <planrichi@gmail.com> wrote:
-- George 'papanikge' Papanikolaou http://www.5slingshots.com/
On Sat, Oct 31, 2015 at 6:37 PM, George Papanikolaou <g3orge.app@gmail.com> wrote:
Important stuff todo, and not tedious at all. Python 3.3 introduced the `yield from` construct and other changes in generator semantics. The asyncio library depends on them and many Pypy users would love to use it. Cheers, Luciano
-- Luciano Ramalho | Author of Fluent Python (O'Reilly, 2015) | http://shop.oreilly.com/product/0636920032519.do | Professor em: http://python.pro.br | Twitter: @ramalhoorg
Hi there, I finished my thesis recently implementing a vectorizing optimization. So I guess it is feasible :). My advice would be to build the VM and make some baby steps in the test environment. This might sound weird, but the development is very different to what you do at university (at least in austria) and also very different to most projects I know. I think there are many possible ways you could improve PyPy. It is hard to give some advice on exactly what you should implement. After all you need to be motivated until you finish your thesis. It would probably be better to think about a topic (VM X does Y), see if PyPy implemented Y (you can of course ask us, it is hard to get around in such a big project in the beginning), (if not) reflect if it gives any benefit for the VM and then implement it. Here is a list of potential projects (might get you started): http://pypy.readthedocs.org/en/latest/project-ideas.html What really would help PyPy (might be more engineering than a 'researchy' topic) is to implement the Python 3 spec (e.g. 3.3,3.4,3.5). You can also join IRC and we will see if we can help to choose a topic. Cheers, Richard On 10/30/2015 06:46 PM, George Papanikolaou wrote:
Hi, First of all thanks for answering, yes I'll pop on IRC soon. My thesis will be for an undergraduate degree (https://en.wikipedia.org/wiki/Engineer%27s_degree#Greece) so I guess I can go for an engineering project more than a research-y one. How hard would implementing the 3.0 spec be? It sounds tedious but pretty standard. I'll get started right away with your advice in order to get familiar with the VM, the environment and coding nuisances in general. Regards, George On Sat, Oct 31, 2015 at 11:57 AM, Richard Plangger <planrichi@gmail.com> wrote:
-- George 'papanikge' Papanikolaou http://www.5slingshots.com/
On Sat, Oct 31, 2015 at 6:37 PM, George Papanikolaou <g3orge.app@gmail.com> wrote:
Important stuff todo, and not tedious at all. Python 3.3 introduced the `yield from` construct and other changes in generator semantics. The asyncio library depends on them and many Pypy users would love to use it. Cheers, Luciano
-- Luciano Ramalho | Author of Fluent Python (O'Reilly, 2015) | http://shop.oreilly.com/product/0636920032519.do | Professor em: http://python.pro.br | Twitter: @ramalhoorg
participants (3)
-
George Papanikolaou -
Luciano Ramalho -
Richard Plangger