www.digitalmars.com         C & C++   DMDScript  

digitalmars.D - Google Summer of Code 2013(?)

reply "alex" <info alexanderbothe.com> writes:
Hi everyone,

I've just read that there are only 5 days remaining for 
organization applications
http://www.google-melange.com/gsoc/homepage/google/gsoc2013

Will Digitalmars do it a further year? Or are there too many 
resources spent on managing DConf?

Personally, I'd like to participate in the event a second time - 
not only for the $5k (+$500 for the mentor) to get - hoping that 
we can improve our toolchains another time (and/or finally move 
away from the goddamn conservative GC!)

I dunno whether Andrei or Walter really want to process through 
all this stuff again, but it was kinda nice to see the 
Digitalmars rising up once more :-)
Mar 24 2013
next sibling parent reply "alex" <info alexanderbothe.com> writes:
I forgot to specify the topic I'd like to work on:

To implement CTFE of D code under .Net - whereas the D code 
should become compiled into CIL (Common language runtime 
Intermediate Language, an assembler-like language), so one could 
execute the 'final' program directly in a .Net environment.
I know this goal can never be reached within one summer of code - 
and it's also rather experimental / an entire compiler back-end 
that has to be implemented (The front-end is mostly there already 
;-)) + perhaps needs a bunch of people for such a huge thing, but 
well..why not?
Mar 24 2013
parent reply "Adam Wilson" <flyboynw gmail.com> writes:
On Sun, 24 Mar 2013 08:07:24 -0700, alex <info alexanderbothe.com> wrote:


 I forgot to specify the topic I'd like to work on:

 To implement CTFE of D code under .Net - whereas the D code should  
 become compiled into CIL (Common language runtime Intermediate Language,  
 an assembler-like language), so one could execute the 'final' program  
 directly in a .Net environment.
 I know this goal can never be reached within one summer of code - and  
 it's also rather experimental / an entire compiler back-end that has to  
 be implemented (The front-end is mostly there already ;-)) + perhaps  
 needs a bunch of people for such a huge thing, but well..why not?


I just want to throw some caution into the wind here. Many people have  
tried emitting CIL from D source, and have had varying levels of success.  
Ranges in particular seems to a pain point as the CIL has no way to  
express that concept natively. I am not saying this isn't a good idea, as  
it's the only realistic way to do CTFE in the .NET world, just something  
to be careful when approaching.


happen.

-- 
Adam Wilson
IRC: LightBender
Project Coordinator
The Horizon Project
http://www.thehorizonproject.org/
Mar 24 2013
parent reply "alex" <info alexanderbothe.com> writes:
On Sunday, 24 March 2013 at 20:22:56 UTC, Adam Wilson wrote:

 I just want to throw some caution into the wind here. Many 
 people have tried emitting CIL from D source, and have had 
 varying levels of success. Ranges in particular seems to a pain 
 point as the CIL has no way to express that concept natively. I 
 am not saying this isn't a good idea, as it's the only 
 realistic way to do CTFE in the .NET world, just something to 
 be careful when approaching.


Ranges.. in .Net there's something beautiful called List, which 
imho could somehow be used for wrapping e.g. array slices. I 
mean, even if it's a workaround it's still considering the 
functionality



 make this happen.


Yeah, would love that, too. But then there's also a question of 
effort vs. gain - is another "compiler" really needed? Is it 
worth the 2 years of CTFE implementation brainfuck? Hmm.
Mar 24 2013
parent reply Timon Gehr <timon.gehr gmx.ch> writes:
On 03/24/2013 09:42 PM, alex wrote:

 On Sunday, 24 March 2013 at 20:22:56 UTC, Adam Wilson wrote:

 I just want to throw some caution into the wind here. Many people have
 tried emitting CIL from D source, and have had varying levels of
 success. Ranges in particular seems to a pain point as the CIL has no
 way to express that concept natively. I am not saying this isn't a
 good idea, as it's the only realistic way to do CTFE in the .NET
 world, just something to be careful when approaching.


 Ranges.. in .Net there's something beautiful called List, which imho
 could somehow be used for wrapping e.g. array slices. I mean, even if
 it's a workaround it's still considering the functionality



 this happen.


 Yeah, would love that, too. But then there's also a question of effort
 vs. gain - is another "compiler" really needed? Is it worth the 2 years
 of CTFE implementation brainfuck? Hmm.


The front end is the hard part. The D AST => CIL transformation is 
rather easy.
Mar 24 2013
next sibling parent "alex" <info alexanderbothe.com> writes:
On Sunday, 24 March 2013 at 20:49:06 UTC, Timon Gehr wrote:

 On 03/24/2013 09:42 PM, alex wrote:

 On Sunday, 24 March 2013 at 20:22:56 UTC, Adam Wilson wrote:

 I just want to throw some caution into the wind here. Many 
 people have
 tried emitting CIL from D source, and have had varying levels 
 of
 success. Ranges in particular seems to a pain point as the 
 CIL has no
 way to express that concept natively. I am not saying this 
 isn't a
 good idea, as it's the only realistic way to do CTFE in the 
 .NET
 world, just something to be careful when approaching.


 Ranges.. in .Net there's something beautiful called List, 
 which imho
 could somehow be used for wrapping e.g. array slices. I mean, 
 even if
 it's a workaround it's still considering the functionality



 help make
 this happen.


 Yeah, would love that, too. But then there's also a question 
 of effort
 vs. gain - is another "compiler" really needed? Is it worth 
 the 2 years
 of CTFE implementation brainfuck? Hmm.


 The front end is the hard part. The D AST => CIL transformation 
 is rather easy.


The front end is more or less finished - and perhaps it's also 
just needed to satisfy the most common needs such as string 
operations and manipulations..because CTFE should be used mostly 
for putting together mixin strings, am I right?
Mar 24 2013
prev sibling parent "Adam Wilson" <flyboynw gmail.com> writes:
On Sun, 24 Mar 2013 13:49:06 -0700, Timon Gehr <timon.gehr gmx.ch> wrote:


 On 03/24/2013 09:42 PM, alex wrote:

 On Sunday, 24 March 2013 at 20:22:56 UTC, Adam Wilson wrote:

 I just want to throw some caution into the wind here. Many people have
 tried emitting CIL from D source, and have had varying levels of
 success. Ranges in particular seems to a pain point as the CIL has no
 way to express that concept natively. I am not saying this isn't a
 good idea, as it's the only realistic way to do CTFE in the .NET
 world, just something to be careful when approaching.


 Ranges.. in .Net there's something beautiful called List, which imho
 could somehow be used for wrapping e.g. array slices. I mean, even if
 it's a workaround it's still considering the functionality



 this happen.


 Yeah, would love that, too. But then there's also a question of effort
 vs. gain - is another "compiler" really needed? Is it worth the 2 years
 of CTFE implementation brainfuck? Hmm.


 The front end is the hard part. The D AST => CIL transformation is  
 rather easy.


I imagine that he could use his existing lexer/parser from Mono-D and  
reporpuse it.

-- 
Adam Wilson
IRC: LightBender
Project Coordinator
The Horizon Project
http://www.thehorizonproject.org/
Mar 24 2013
prev sibling next sibling parent reply "David Nadlinger" <see klickverbot.at> writes:
On Sunday, 24 March 2013 at 14:48:30 UTC, alex wrote:

 I've just read that there are only 5 days remaining for 
 organization applications
 http://www.google-melange.com/gsoc/homepage/google/gsoc2013

 Will Digitalmars do it a further year? Or are there too many 
 resources spent on managing DConf?


I was meaning to write this post myself right now:

In my opinion, we can't afford to miss the chance to participate 
in the Summer of Code for a third time. For this to be possible, 
though, we ned to throw together an application, and pronto.

Also, as far as I can see, in both of the previous two years, the 
work by one student was not actually incorporated into the 
upstream projects or made available in an easily usable way: 
Cristi Cobzarenco's SciD work from 2011) and Antti-Ville 
Tuunainen's garbage collector improvements from last year. (Note 
aside: It seems like in both of the cases, David Simcha was the 
mentor, but this could just be a coincidence.)

We need to quickly come up with and commit to a strategy to make 
sure this does not happen again. Otherwise, this will look very 
bad on an application for this year.

David
Mar 24 2013
parent reply Dmitry Olshansky <dmitry.olsh gmail.com> writes:
24-Mar-2013 19:25, David Nadlinger пишет:

 On Sunday, 24 March 2013 at 14:48:30 UTC, alex wrote:

 I've just read that there are only 5 days remaining for organization
 applications
 http://www.google-melange.com/gsoc/homepage/google/gsoc2013

 Will Digitalmars do it a further year? Or are there too many resources
 spent on managing DConf?


 I was meaning to write this post myself right now:

 In my opinion, we can't afford to miss the chance to participate in the
 Summer of Code for a third time. For this to be possible, though, we ned
 to throw together an application, and pronto.


Sure, and AFAICT for that to happen we need at least an ideas page. 
Shouldn't be hard to put together.
For one idea I'd love to see more D support in embedded/ARM world e.g. 
an easy to use STM32 quick-start package would be nice.


 Also, as far as I can see, in both of the previous two years, the work
 by one student was not actually incorporated into the upstream projects
 or made available in an easily usable way: Cristi Cobzarenco's SciD work
 from 2011) and Antti-Ville Tuunainen's garbage collector improvements
 from last year. (Note aside: It seems like in both of the cases, David
 Simcha was the mentor, but this could just be a coincidence.)


It might be only my perception but one common case in both of these 
projects is using the only recently available features (RTInfo hook) and 
otherwise hitting rough places of compiler (e.g. defunct post-blit back 
then).

That being said I think we (D GSoC-ers) can testify that overcoming D's 
implementation limitations (~bugs) proves to take not a small amount of 
time allocated for the project. But this year it should be easier, heh?


 We need to quickly come up with and commit to a strategy to make sure
 this does not happen again. Otherwise, this will look very bad on an
 application for this year.


Maybe focus on bite-sized deliverables?
Also maybe having explicitly listed minor/major goals, a plan-B in case 
something doesn't work out etc.


-- 
Dmitry Olshansky
Mar 24 2013
next sibling parent Jacob Carlborg <doob me.com> writes:
On 2013-03-24 17:01, Dmitry Olshansky wrote:


 Sure, and AFAICT for that to happen we need at least an ideas page.
 Shouldn't be hard to put together.
 For one idea I'd love to see more D support in embedded/ARM world e.g.
 an easy to use STM32 quick-start package would be nice.


There's always the ideas from previous years.

-- 
/Jacob Carlborg
Mar 24 2013
prev sibling next sibling parent reply Iain Buclaw <ibuclaw ubuntu.com> writes:
On 24 March 2013 16:01, Dmitry Olshansky <dmitry.olsh gmail.com> wrote:


 24-Mar-2013 19:25, David Nadlinger =D0=BF=D0=B8=D1=88=D0=B5=D1=82:

  On Sunday, 24 March 2013 at 14:48:30 UTC, alex wrote:

 I've just read that there are only 5 days remaining for organization
 applications
 http://www.google-melange.com/**gsoc/homepage/google/gsoc2013<http://ww=






w.google-melange.com/gsoc/homepage/google/gsoc2013>

 Will Digitalmars do it a further year? Or are there too many resources
 spent on managing DConf?


 I was meaning to write this post myself right now:

 In my opinion, we can't afford to miss the chance to participate in the
 Summer of Code for a third time. For this to be possible, though, we ned
 to throw together an application, and pronto.


 Sure, and AFAICT for that to happen we need at least an ideas page.
 Shouldn't be hard to put together.
 For one idea I'd love to see more D support in embedded/ARM world e.g. an
 easy to use STM32 quick-start package would be nice.


This could be taking up with someone willing to dedicate time in GDC /
LDC.   I would be willing to mentor anyone who wants to join in on getting
GDC on ARM ready for production, but I must admit this is rather short
notice for me. ;-)


--=20
Iain Buclaw

*(p < e ? p++ : p) =3D (c & 0x0f) + '0';
Mar 24 2013
next sibling parent reply Jacob Carlborg <doob me.com> writes:
On 2013-03-24 22:28, Iain Buclaw wrote:


 This could be taking up with someone willing to dedicate time in GDC /
 LDC.   I would be willing to mentor anyone who wants to join in on
 getting GDC on ARM ready for production, but I must admit this is rather
 short notice for me. ;-)


"short notice"? It's still -10 C and snow here in the mornings. The 
summer is no way near to be seen :)

-- 
/Jacob Carlborg
Mar 25 2013
parent reply Iain Buclaw <ibuclaw ubuntu.com> writes:
On 25 March 2013 09:00, Jacob Carlborg <doob me.com> wrote:


 On 2013-03-24 22:28, Iain Buclaw wrote:

  This could be taking up with someone willing to dedicate time in GDC /

 LDC.   I would be willing to mentor anyone who wants to join in on
 getting GDC on ARM ready for production, but I must admit this is rather
 short notice for me. ;-)


 "short notice"? It's still -10 C and snow here in the mornings. The summer
 is no way near to be seen :)


It has been a long and colder March than usual. :o)

-- 
Iain Buclaw

*(p < e ? p++ : p) = (c & 0x0f) + '0';
Mar 25 2013
parent Jacob Carlborg <doob me.com> writes:
On 2013-03-25 10:08, Iain Buclaw wrote:


 It has been a long and colder March than usual. :o)


It can snow in the beginning of May here :(

-- 
/Jacob Carlborg
Mar 25 2013
prev sibling parent reply Martin Nowak <code dawg.eu> writes:
On 03/24/2013 10:28 PM, Iain Buclaw wrote:

     Sure, and AFAICT for that to happen we need at least an ideas page.
     Shouldn't be hard to put together.
     For one idea I'd love to see more D support in embedded/ARM world
     e.g. an easy to use STM32 quick-start package would be nice.


 This could be taking up with someone willing to dedicate time in GDC /
 LDC.   I would be willing to mentor anyone who wants to join in on
 getting GDC on ARM ready for production, but I must admit this is rather
 short notice for me. ;-)


I'm actually interested in doing this. So can we make that happen?
Mar 26 2013
parent reply Iain Buclaw <ibuclaw ubuntu.com> writes:
On 26 March 2013 16:13, Martin Nowak <code dawg.eu> wrote:


 On 03/24/2013 10:28 PM, Iain Buclaw wrote:


     Sure, and AFAICT for that to happen we need at least an ideas page.
     Shouldn't be hard to put together.
     For one idea I'd love to see more D support in embedded/ARM world
     e.g. an easy to use STM32 quick-start package would be nice.


 This could be taking up with someone willing to dedicate time in GDC /
 LDC.   I would be willing to mentor anyone who wants to join in on
 getting GDC on ARM ready for production, but I must admit this is rather
 short notice for me. ;-)


 I'm actually interested in doing this. So can we make that happen?


We certainly can do. Which compiler do you have in mind for using?

I have two ARM boards at hand, but will need to get round to buying a
monitor to connect up at least one of them to start off the port and
testing process. :-)

-- 
Iain Buclaw

*(p < e ? p++ : p) = (c & 0x0f) + '0';
Mar 26 2013
next sibling parent reply Martin Nowak <code dawg.eu> writes:
On 03/26/2013 06:22 PM, Iain Buclaw wrote:

 We certainly can do. Which compiler do you have in mind for using?


I mostly thought of porting GDC and LDC to linux glibc and bionic.
A Windows ARM port would be interesting too, but could be out of reach.
Support for bare metal targets is interesting too (I've already used GDC 
for STM32) but way less important.


 I have two ARM boards at hand, but will need to get round to buying a
 monitor to connect up at least one of them to start off the port and
 testing process. :-)


I have two android devices, some STM32 boards and getting a raspberry or 
so would not be a problem. I guess qemu could be used too.
Mar 26 2013
parent reply Johannes Pfau <nospam example.com> writes:
Am Wed, 27 Mar 2013 01:50:40 +0100
schrieb Martin Nowak <code dawg.eu>:


 On 03/26/2013 06:22 PM, Iain Buclaw wrote:

 We certainly can do. Which compiler do you have in mind for using?


 I mostly thought of porting GDC and LDC to linux glibc and bionic.
 A Windows ARM port would be interesting too, but could be out of
 reach. Support for bare metal targets is interesting too (I've
 already used GDC for STM32) but way less important.


Android support would be very nice but I'd make it a secondary goal.
As we have shared libraries now the only big remaining problem is that
Android doesn't have real TLS so we'd have to make the GCC EMUTLS
implementation work with our GC. For glibc android targets this issue
doesn't exist.

Would be nice if we can get the dmd test suite results to match those
for x86. It will be a little annoying though as many floating point test
are wrong and simply assume x86 precision.

I'd be glad to help wherever I can though.


 I have two ARM boards at hand, but will need to get round to buying
 a monitor to connect up at least one of them to start off the port
 and testing process. :-)


 I have two android devices, some STM32 boards and getting a raspberry
 or so would not be a problem. I guess qemu could be used too.
 


qemu + debian for arm might be a good idea. I wonder whether ARM
emulated by qemu is faster than e.g. a Raspberry Pi. Compiling GDC on
the raspberry itself takes over 12 hours IIRC so you probably want to
build cross compilers or use a faster system.
Mar 27 2013
next sibling parent Iain Buclaw <ibuclaw ubuntu.com> writes:
On 27 March 2013 08:14, Johannes Pfau <nospam example.com> wrote:


 Am Wed, 27 Mar 2013 01:50:40 +0100
 schrieb Martin Nowak <code dawg.eu>:


 On 03/26/2013 06:22 PM, Iain Buclaw wrote:

 We certainly can do. Which compiler do you have in mind for using?


 I mostly thought of porting GDC and LDC to linux glibc and bionic.
 A Windows ARM port would be interesting too, but could be out of
 reach. Support for bare metal targets is interesting too (I've
 already used GDC for STM32) but way less important.


 Android support would be very nice but I'd make it a secondary goal.
 As we have shared libraries now the only big remaining problem is that
 Android doesn't have real TLS so we'd have to make the GCC EMUTLS
 implementation work with our GC. For glibc android targets this issue
 doesn't exist.

 Would be nice if we can get the dmd test suite results to match those
 for x86. It will be a little annoying though as many floating point test
 are wrong and simply assume x86 precision.

 I'd be glad to help wherever I can though.


 I have two ARM boards at hand, but will need to get round to buying
 a monitor to connect up at least one of them to start off the port
 and testing process. :-)


 I have two android devices, some STM32 boards and getting a raspberry
 or so would not be a problem. I guess qemu could be used too.


 qemu + debian for arm might be a good idea. I wonder whether ARM
 emulated by qemu is faster than e.g. a Raspberry Pi. Compiling GDC on
 the raspberry itself takes over 12 hours IIRC so you probably want to
 build cross compilers or use a faster system.


Through testing I've done in the past, QEMU+ARM is reliable enough that you
can compile + test GCC under it.  Other than that, we can go for using
cross compilers.

-- 
Iain Buclaw

*(p < e ? p++ : p) = (c & 0x0f) + '0';
Mar 27 2013
prev sibling parent Iain Buclaw <ibuclaw ubuntu.com> writes:
On 27 March 2013 11:06, Iain Buclaw <ibuclaw ubuntu.com> wrote:


 On 27 March 2013 08:14, Johannes Pfau <nospam example.com> wrote:


 Am Wed, 27 Mar 2013 01:50:40 +0100
 schrieb Martin Nowak <code dawg.eu>:


 On 03/26/2013 06:22 PM, Iain Buclaw wrote:

 We certainly can do. Which compiler do you have in mind for using?


 I mostly thought of porting GDC and LDC to linux glibc and bionic.
 A Windows ARM port would be interesting too, but could be out of
 reach. Support for bare metal targets is interesting too (I've
 already used GDC for STM32) but way less important.


 Android support would be very nice but I'd make it a secondary goal.
 As we have shared libraries now the only big remaining problem is that
 Android doesn't have real TLS so we'd have to make the GCC EMUTLS
 implementation work with our GC. For glibc android targets this issue
 doesn't exist.

 Would be nice if we can get the dmd test suite results to match those
 for x86. It will be a little annoying though as many floating point test
 are wrong and simply assume x86 precision.

 I'd be glad to help wherever I can though.


 I have two ARM boards at hand, but will need to get round to buying
 a monitor to connect up at least one of them to start off the port
 and testing process. :-)


 I have two android devices, some STM32 boards and getting a raspberry
 or so would not be a problem. I guess qemu could be used too.


 qemu + debian for arm might be a good idea. I wonder whether ARM
 emulated by qemu is faster than e.g. a Raspberry Pi. Compiling GDC on
 the raspberry itself takes over 12 hours IIRC so you probably want to
 build cross compilers or use a faster system.


 Through testing I've done in the past, QEMU+ARM is reliable enough that
 you can compile + test GCC under it.  Other than that, we can go for using
 cross compilers.


BTW, QEMU+ARM emulation under x86/x86_64 still takes about 8 hours for GCC.
:-)

-- 
Iain Buclaw

*(p < e ? p++ : p) = (c & 0x0f) + '0';
Mar 27 2013
prev sibling parent reply Martin Nowak <code dawg.eu> writes:
On 03/26/2013 06:22 PM, Iain Buclaw wrote:

 We certainly can do. Which compiler do you have in mind for using?


Have you send Andrei a mail to add you as mentor?
I'll add a project idea to the wiki http://wiki.dlang.org/GSOC_2013_Ideas.
Mar 27 2013
parent Iain Buclaw <ibuclaw ubuntu.com> writes:
On 27 March 2013 15:19, Martin Nowak <code dawg.eu> wrote:


 On 03/26/2013 06:22 PM, Iain Buclaw wrote:


 We certainly can do. Which compiler do you have in mind for using?


 Have you send Andrei a mail to add you as mentor?
 I'll add a project idea to the wiki http://wiki.dlang.org/GSOC_**
 2013_Ideas <http://wiki.dlang.org/GSOC_2013_Ideas>.


Not yet, if this is going to be considered as a project that we are going
to move forward with, I'll definitely pick this one up. :-)

-- 
Iain Buclaw

*(p < e ? p++ : p) = (c & 0x0f) + '0';
Mar 27 2013
prev sibling parent "David Nadlinger" <see klickverbot.at> writes:
On Sunday, 24 March 2013 at 16:01:45 UTC, Dmitry Olshansky wrote:

 Sure, and AFAICT for that to happen we need at least an ideas 
 page. Shouldn't be hard to put together.
 For one idea I'd love to see more D support in embedded/ARM 
 world e.g. an easy to use STM32 quick-start package would be 
 nice.


I would be more than happy to support any effort on getting an 
LDC ARM (and possibly Android) port done.

I have a dev board lying around (4x1.4 GHz Cortex-A9, 1 GB RAM, 
Android/Arch Linux) which could be used for testing — and which I 
could provide somebody interested SSH access to (via my home 
internet connection, though, so probably a bit laggy at times).

David
Mar 26 2013
prev sibling parent reply Andrei Alexandrescu <SeeWebsiteForEmail erdani.org> writes:
On 3/24/13 10:48 AM, alex wrote:

 Hi everyone,

 I've just read that there are only 5 days remaining for organization
 applications
 http://www.google-melange.com/gsoc/homepage/google/gsoc2013

 Will Digitalmars do it a further year? Or are there too many resources
 spent on managing DConf?

 Personally, I'd like to participate in the event a second time - not
 only for the $5k (+$500 for the mentor) to get - hoping that we can
 improve our toolchains another time (and/or finally move away from the
 goddamn conservative GC!)

 I dunno whether Andrei or Walter really want to process through all this
 stuff again, but it was kinda nice to see the Digitalmars rising up once
 more :-)


Yes, we do want to participate to this year's edition, I was about to 
send a request for mentors.

Andrei
Mar 24 2013
parent "alex" <info alexanderbothe.com> writes:
On Sunday, 24 March 2013 at 15:39:37 UTC, Andrei Alexandrescu 
wrote:

 On 3/24/13 10:48 AM, alex wrote:

 Hi everyone,

 I've just read that there are only 5 days remaining for 
 organization
 applications
 http://www.google-melange.com/gsoc/homepage/google/gsoc2013

 Will Digitalmars do it a further year? Or are there too many 
 resources
 spent on managing DConf?

 Personally, I'd like to participate in the event a second time 
 - not
 only for the $5k (+$500 for the mentor) to get - hoping that 
 we can
 improve our toolchains another time (and/or finally move away 
 from the
 goddamn conservative GC!)

 I dunno whether Andrei or Walter really want to process 
 through all this
 stuff again, but it was kinda nice to see the Digitalmars 
 rising up once
 more :-)


 Yes, we do want to participate to this year's edition, I was 
 about to send a request for mentors.

 Andrei


Nice! Thank you!

I already wondered why nobody ever mentioned the new GSoC in the 
NG :)
Mar 24 2013