You can subscribe to this list here.
2007 |
Jan
|
Feb
|
Mar
|
Apr
|
May
(52) |
Jun
(30) |
Jul
(17) |
Aug
(9) |
Sep
(4) |
Oct
(7) |
Nov
(11) |
Dec
(19) |
---|---|---|---|---|---|---|---|---|---|---|---|---|
2008 |
Jan
|
Feb
(1) |
Mar
(37) |
Apr
(28) |
May
(15) |
Jun
(28) |
Jul
(7) |
Aug
(125) |
Sep
(116) |
Oct
(85) |
Nov
(14) |
Dec
(6) |
2009 |
Jan
(11) |
Feb
(4) |
Mar
(5) |
Apr
|
May
(9) |
Jun
(5) |
Jul
(4) |
Aug
(40) |
Sep
(1) |
Oct
(19) |
Nov
(43) |
Dec
(45) |
2010 |
Jan
(76) |
Feb
(95) |
Mar
(3) |
Apr
(23) |
May
(39) |
Jun
(54) |
Jul
(6) |
Aug
(13) |
Sep
(12) |
Oct
(59) |
Nov
(53) |
Dec
(43) |
2011 |
Jan
(43) |
Feb
(44) |
Mar
(25) |
Apr
(23) |
May
|
Jun
|
Jul
|
Aug
|
Sep
(5) |
Oct
(1) |
Nov
(2) |
Dec
|
2013 |
Jan
|
Feb
|
Mar
|
Apr
|
May
|
Jun
|
Jul
|
Aug
|
Sep
(6) |
Oct
|
Nov
|
Dec
|
From: Dean M. C. B. <dmb...@fr...> - 2008-08-07 06:41:42
|
Hi Guys, I've only recently been able to try and test the localhost test which uses the server. And what I found out is if you're running it in Linux without an X server installed (or xterm), it won't work. I've tried modifying it to use Python directly instead, but I'm not having any luck with that. Divye, have you tried not using xterm to spawn a new process after the fork to run the server? Perhaps writing a bash script that gets executed instead in execlp would be another option? HTH -- Dean Michael Berris Software Engineer, Friendster, Inc. |
From: Divye K. <div...@gm...> - 2008-07-02 09:20:04
|
Glyn, Yes, those patches are related to that issue. I have fixed 3 issues in the patch: 1. Headers such as "Content-Type: 117" were being stored as ("Content-type", "117\r") in the headers map. 2. The body was not being properly stored because an empty line was being looked for after the headers response whereas parsing the response yielded a line containing "\r", thus causing the actual body content to be parsed. 3. Formatted I/O functions were being used on the socket_stream whereas we were transferring unformatted data (HTTP doesn't know anything about the payload). I have changed the code to use unformatted I/O functions. Tests for issues 1 and 2 have been committed in the localhost_tests suite (content_length_header_test, text_file_query, binary_file_query) If the tests can be verified to pass, then we can close the issue. Divye On Wed, Jul 2, 2008 at 12:34 PM, Glyn Matthews <gly...@gm...> wrote: > Divye, > > Is this related to the reported patch: > http://tinyurl.com/4cjwo7 > > ? > > Glyn > > > 2008/7/2 Dean Michael C. Berris <dmb...@fr...>: >> >> Thanks Divye, >> >> I'll go take a look when I get the time. I think it would be worth >> checking whether Windows is sending/receiving the file in wide character >> mode. >> >> -- >> Dean Michael Berris >> Software Engineer, Friendster, Inc. >> >> > -----Original Message----- >> > From: cpp...@li... >> > [mailto:cpp...@li...] On >> > Behalf Of Divye Kapoor >> > Sent: Wednesday, July 02, 2008 8:19 AM >> > To: C++ Networking Library Developers Mailing List >> > Subject: [cpp-netlib-devel] [commit] Patch for header and >> > body processing inhttp::client >> > >> > Hi, >> > I've removed the extra \r being inserted with every http header and >> > ensured proper newline processing for the body of the message. The >> > only problem I'm facing is that on Windows/MSVC the actual size of a >> > text file and the size of the transferred content don't match whereas >> > the binary file transfer is perfect. (look at the text_file_query >> > test) I have not been able to trace the reason for this. Is the error >> > in my test code or in the implementation of http::client? Strangely, >> > the same test succeeds on linux. Is it some line ending issue? >> > >> > Now, all the tests should run automatically (nearly) on both linux and >> > windows. If you face any issues, please mail! >> > >> > Divye Kapoor >> > >> > >> > -- >> > Whether the chicken crossed the road or the road moved beneath the >> > chicken depends upon your frame of reference. >> > My official web site: http://people.iitr.ernet.in/shp/061305/ >> > >> > -------------------------------------------------------------- >> > ----------- >> > Sponsored by: SourceForge.net Community Choice Awards: VOTE NOW! >> > Studies have shown that voting for your favorite open source project, >> > along with a healthy diet, reduces your potential for chronic lameness >> > and boredom. Vote Now at http://www.sourceforge.net/community/cca08 >> > _______________________________________________ >> > Cpp-netlib-devel mailing list >> > Cpp...@li... >> > https://lists.sourceforge.net/lists/listinfo/cpp-netlib-devel >> > >> >> ------------------------------------------------------------------------- >> Sponsored by: SourceForge.net Community Choice Awards: VOTE NOW! >> Studies have shown that voting for your favorite open source project, >> along with a healthy diet, reduces your potential for chronic lameness >> and boredom. Vote Now at http://www.sourceforge.net/community/cca08 >> _______________________________________________ >> Cpp-netlib-devel mailing list >> Cpp...@li... >> https://lists.sourceforge.net/lists/listinfo/cpp-netlib-devel > > > ------------------------------------------------------------------------- > Sponsored by: SourceForge.net Community Choice Awards: VOTE NOW! > Studies have shown that voting for your favorite open source project, > along with a healthy diet, reduces your potential for chronic lameness > and boredom. Vote Now at http://www.sourceforge.net/community/cca08 > _______________________________________________ > Cpp-netlib-devel mailing list > Cpp...@li... > https://lists.sourceforge.net/lists/listinfo/cpp-netlib-devel > > -- Whether the chicken crossed the road or the road moved beneath the chicken depends upon your frame of reference. My official web site: http://people.iitr.ernet.in/shp/061305/ |
From: Glyn M. <gly...@gm...> - 2008-07-02 07:05:03
|
Divye, Is this related to the reported patch: http://tinyurl.com/4cjwo7 ? Glyn 2008/7/2 Dean Michael C. Berris <dmb...@fr...>: > Thanks Divye, > > I'll go take a look when I get the time. I think it would be worth > checking whether Windows is sending/receiving the file in wide character > mode. > > -- > Dean Michael Berris > Software Engineer, Friendster, Inc. > > > -----Original Message----- > > From: cpp...@li... > > [mailto:cpp...@li...] On > > Behalf Of Divye Kapoor > > Sent: Wednesday, July 02, 2008 8:19 AM > > To: C++ Networking Library Developers Mailing List > > Subject: [cpp-netlib-devel] [commit] Patch for header and > > body processing inhttp::client > > > > Hi, > > I've removed the extra \r being inserted with every http header and > > ensured proper newline processing for the body of the message. The > > only problem I'm facing is that on Windows/MSVC the actual size of a > > text file and the size of the transferred content don't match whereas > > the binary file transfer is perfect. (look at the text_file_query > > test) I have not been able to trace the reason for this. Is the error > > in my test code or in the implementation of http::client? Strangely, > > the same test succeeds on linux. Is it some line ending issue? > > > > Now, all the tests should run automatically (nearly) on both linux and > > windows. If you face any issues, please mail! > > > > Divye Kapoor > > > > > > -- > > Whether the chicken crossed the road or the road moved beneath the > > chicken depends upon your frame of reference. > > My official web site: http://people.iitr.ernet.in/shp/061305/ > > > > -------------------------------------------------------------- > > ----------- > > Sponsored by: SourceForge.net Community Choice Awards: VOTE NOW! > > Studies have shown that voting for your favorite open source project, > > along with a healthy diet, reduces your potential for chronic lameness > > and boredom. Vote Now at http://www.sourceforge.net/community/cca08 > > _______________________________________________ > > Cpp-netlib-devel mailing list > > Cpp...@li... > > https://lists.sourceforge.net/lists/listinfo/cpp-netlib-devel > > > > ------------------------------------------------------------------------- > Sponsored by: SourceForge.net Community Choice Awards: VOTE NOW! > Studies have shown that voting for your favorite open source project, > along with a healthy diet, reduces your potential for chronic lameness > and boredom. Vote Now at http://www.sourceforge.net/community/cca08 > _______________________________________________ > Cpp-netlib-devel mailing list > Cpp...@li... > https://lists.sourceforge.net/lists/listinfo/cpp-netlib-devel > |
From: Dean M. C. B. <dmb...@fr...> - 2008-07-02 02:15:49
|
Thanks Divye, I'll go take a look when I get the time. I think it would be worth checking whether Windows is sending/receiving the file in wide character mode. -- Dean Michael Berris Software Engineer, Friendster, Inc. > -----Original Message----- > From: cpp...@li... > [mailto:cpp...@li...] On > Behalf Of Divye Kapoor > Sent: Wednesday, July 02, 2008 8:19 AM > To: C++ Networking Library Developers Mailing List > Subject: [cpp-netlib-devel] [commit] Patch for header and > body processing inhttp::client > > Hi, > I've removed the extra \r being inserted with every http header and > ensured proper newline processing for the body of the message. The > only problem I'm facing is that on Windows/MSVC the actual size of a > text file and the size of the transferred content don't match whereas > the binary file transfer is perfect. (look at the text_file_query > test) I have not been able to trace the reason for this. Is the error > in my test code or in the implementation of http::client? Strangely, > the same test succeeds on linux. Is it some line ending issue? > > Now, all the tests should run automatically (nearly) on both linux and > windows. If you face any issues, please mail! > > Divye Kapoor > > > -- > Whether the chicken crossed the road or the road moved beneath the > chicken depends upon your frame of reference. > My official web site: http://people.iitr.ernet.in/shp/061305/ > > -------------------------------------------------------------- > ----------- > Sponsored by: SourceForge.net Community Choice Awards: VOTE NOW! > Studies have shown that voting for your favorite open source project, > along with a healthy diet, reduces your potential for chronic lameness > and boredom. Vote Now at http://www.sourceforge.net/community/cca08 > _______________________________________________ > Cpp-netlib-devel mailing list > Cpp...@li... > https://lists.sourceforge.net/lists/listinfo/cpp-netlib-devel > |
From: Divye K. <div...@gm...> - 2008-07-02 00:19:12
|
Hi, I've removed the extra \r being inserted with every http header and ensured proper newline processing for the body of the message. The only problem I'm facing is that on Windows/MSVC the actual size of a text file and the size of the transferred content don't match whereas the binary file transfer is perfect. (look at the text_file_query test) I have not been able to trace the reason for this. Is the error in my test code or in the implementation of http::client? Strangely, the same test succeeds on linux. Is it some line ending issue? Now, all the tests should run automatically (nearly) on both linux and windows. If you face any issues, please mail! Divye Kapoor -- Whether the chicken crossed the road or the road moved beneath the chicken depends upon your frame of reference. My official web site: http://people.iitr.ernet.in/shp/061305/ |
From: Divye K. <div...@gm...> - 2008-07-01 20:55:28
|
Hello all, I have made 2 separate commits. Commit 40 adds the python server to the repository and commit 41 adds the associated tests. Currently, the tests highlight many deficiencies in the http::client code. Most notably, the headers have extraneous '\r' characters and the body of the message gets stripped of its newlines. I will be committing a patch for this soon. One problem with the tests is that the server process that is spawned in a separate terminal has to be manually closed and that produces a "fatal error" report while running the tests on linux. I am trying to resolve this problem. Any ideas? Divye Kapoor -- Whether the chicken crossed the road or the road moved beneath the chicken depends upon your frame of reference. My official web site: http://people.iitr.ernet.in/shp/061305/ |
From: Dean M. B. <mik...@gm...> - 2008-07-01 06:34:33
|
Hi Divye, I'm not sure about zip files, but the Boost Software License can cover the file as well. IANAL though, so I suggest you wait for other people's input. Thanks! On Tue, Jul 1, 2008 at 2:31 PM, Divye Kapoor <div...@gm...> wrote: > Hello all, > I spent the whole of yesterday clearing out a bug in the python web > server on linux that was because of an incorrect line ending style. > Sheesh! It was a huge waste of time. Anyways, I'm ready to check in > the python code. The only thing is, some of the tests require testing > against a transfer of a binary file. The file I'm currently using is a > zip file. What is the policy of such a check in and where should the > License for this file go? > > Divye Kapoor > > -- > Whether the chicken crossed the road or the road moved beneath the > chicken depends upon your frame of reference. > My official web site: http://people.iitr.ernet.in/shp/061305/ > > ------------------------------------------------------------------------- > Check out the new SourceForge.net Marketplace. > It's the best place to buy or sell services for > just about anything Open Source. > http://sourceforge.net/services/buy/index.php > _______________________________________________ > Cpp-netlib-devel mailing list > Cpp...@li... > https://lists.sourceforge.net/lists/listinfo/cpp-netlib-devel > -- Dean Michael C. Berris Software Engineer, Friendster, Inc. |
From: Divye K. <div...@gm...> - 2008-07-01 06:31:38
|
Hello all, I spent the whole of yesterday clearing out a bug in the python web server on linux that was because of an incorrect line ending style. Sheesh! It was a huge waste of time. Anyways, I'm ready to check in the python code. The only thing is, some of the tests require testing against a transfer of a binary file. The file I'm currently using is a zip file. What is the policy of such a check in and where should the License for this file go? Divye Kapoor -- Whether the chicken crossed the road or the road moved beneath the chicken depends upon your frame of reference. My official web site: http://people.iitr.ernet.in/shp/061305/ |
From: Dean M. C. B. <dmb...@fr...> - 2008-06-25 11:30:02
|
Hi Divye, Thanks. Please clean check in the Python code as well so that we can use it for local testing. This will be very valuable. Please make sure that the code you check in will be under the Boost Software License. You can look at the code to see an example of how the header comments should look like. Have a great week, and I certainly hope to hear from you next week! -- Dean Michael Berris Software Engineer, Friendster, Inc. > -----Original Message----- > From: cpp...@li... > [mailto:cpp...@li...] On > Behalf Of Divye Kapoor > Sent: Wednesday, June 25, 2008 4:28 PM > To: C++ Networking Library Developers Mailing List > Subject: Re: [cpp-netlib-devel] HTTP Integration > > Hello guys, > Sorry for the delay. I'm out of town right now and I don't > have much time. I'm up for some documentation writing on the > messaging framework side of things but I can start only after > next Monday. I won't be looking at my mails for a week so > I'll get back to you with my queries on Monday. I have > already started on the docs and have a quickbook document on > hand. I'm not experienced at doc writing so I hope you will > help me out on that front. > On the testing front, what's the next step? Should I clean > up that rough set of tests with a local python server so that > it can be used for testing the http::client interface? > > I'll get back to you next week, > > Divye > > > -- > Whether the chicken crossed the road or the road moved > beneath the chicken depends upon your frame of reference. > My official web site: http://people.iitr.ernet.in/shp/061305/ > > -------------------------------------------------------------- > ----------- > Check out the new SourceForge.net Marketplace. > It's the best place to buy or sell services for just about > anything Open Source. > http://sourceforge.net/services/buy/index.php > _______________________________________________ > Cpp-netlib-devel mailing list > Cpp...@li... > https://lists.sourceforge.net/lists/listinfo/cpp-netlib-devel > |
From: Divye K. <div...@gm...> - 2008-06-25 08:28:30
|
Hello guys, Sorry for the delay. I'm out of town right now and I don't have much time. I'm up for some documentation writing on the messaging framework side of things but I can start only after next Monday. I won't be looking at my mails for a week so I'll get back to you with my queries on Monday. I have already started on the docs and have a quickbook document on hand. I'm not experienced at doc writing so I hope you will help me out on that front. On the testing front, what's the next step? Should I clean up that rough set of tests with a local python server so that it can be used for testing the http::client interface? I'll get back to you next week, Divye -- Whether the chicken crossed the road or the road moved beneath the chicken depends upon your frame of reference. My official web site: http://people.iitr.ernet.in/shp/061305/ |
From: Dean M. B. <mik...@gm...> - 2008-06-24 03:34:06
|
On Mon, Jun 23, 2008 at 7:59 PM, Glyn Matthews <gly...@gm...> wrote: > Dean, > > 2008/6/23 Dean Michael Berris <mik...@gm...>: >> >> Ah. So what do I do from the client side? > > With TortoiseSVN (which I'm looking at now), you can add properties using > the dialog (right-click in the directory, select "Properties" in the menu, > select the Subversion tab and click the box that says "Properties". They > should be empty. Otherwise, from a command line, follow the instructions > given in the link I posted. If you need to know more about how properties > work in subversion, take at look at the boost repository and check the > properties for different files/directories. I expect we'd add something to > individual files like the following: > Oh, goodie. :D > $> svn propset pion-version '0.5.6' boost/network/http/parser.hpp Ayt, I'll set pion-version to the revision number. I've taken some time to write the unit test as well to keep the parser interface simple(r). I'll check in whatever I have in a while, but that unit test will definitely fail. Update you all later. -- Dean Michael C. Berris Software Engineer, Friendster, Inc. |
From: Glyn M. <gly...@gm...> - 2008-06-23 14:27:59
|
Hello Moshe, 2008/6/23 Moshe Matitya <Mos...@ka...>: > I am looking for information on cpp-netlib, in order to evaluate it for > possible use in a new project. I have been trying to find documentation of > the library (high level and class level), but I haven't come across anything > apart from two rather short pages (the Requirements specs and the > Architecural doc) on the wiki. Is there anything beyond this available? > > This project is still in a relatively early stage of development, so documentation is currently sparse. > For a good example of what I'm talking about, see Apache MINA [< > http://mina.apache.org/>] (which is for Java). All of its features [< > http://mina.apache.org/features.html>] can, of course, be implemented by a > developer working directly with Java's networking and concurrency API's -- > but MINA provides a higher-level abstraction in which all the "plumbing" is > already taken care of. Similarly, we could implement our needs by working > directly with Boost.Asio, but I'm hoping that perhaps cpp-netlib already > provides the sort of framework we're looking for. Does it? > Asio will provide you with most of your requirements, apart from the concurrency model. However, it does provide good examples of how to customize a concurrency model using Boost.Threads. I think the only thing in your requirements that is lacking in Boost is the thread pool, but there is a good one available here: http://threadpool.sourceforge.net/ cpp-netlib is intended to provide application layer support on top of Boost.Asio (HTTP, SMTP, SNCP eventually etc.). > One other thing -- where can the sources be downloaded from? Is there a > released tarball available (I haven't been able to find it), or do I need to > check it out of SVN? > > There are no releases available. The subversion repository can be found here: https://cpp-netlib.svn.sourceforge.net/svnroot/cpp-netlib Regards, Glyn |
From: Moshe M. <Mos...@Ka...> - 2008-06-23 13:55:28
|
I am looking for information on cpp-netlib, in order to evaluate it for possible use in a new project. I have been trying to find documentation of the library (high level and class level), but I haven't come across anything apart from two rather short pages (the Requirements specs and the Architecural doc) on the wiki. Is there anything beyond this available? Here is a summary of what we're looking for. I would very much appreciate it if someone here could tell me if cpp-netlib is appropriate for these requirements (or most of them): * Support for both UDP and TCP * Support for both synchronous and asynchronous I/O * A unified API for different transport types (TCP / UDP) * A chainable message filter interface * Asynchronous handling of tasks (separate from the network I/O) In particular, we are interested in a customizable concurrency model (single thread / thread pool / multiple thread pools), which allows the library user to specify parameters such as the following: * The size of the thread pool * The size of each thread's task queue * What policy to use for the task queue's size (fixed size vs. grow as needed) * What data structure to use for the queue (vector, list, etc.) We'd like an API that lets the user simply set these parameters, without having to write the code that acually implements it all. For a good example of what I'm talking about, see Apache MINA [<http://mina.apache.org/>] (which is for Java). All of its features [<http://mina.apache.org/features.html>] can, of course, be implemented by a developer working directly with Java's networking and concurrency API's -- but MINA provides a higher-level abstraction in which all the "plumbing" is already taken care of. Similarly, we could implement our needs by working directly with Boost.Asio, but I'm hoping that perhaps cpp-netlib already provides the sort of framework we're looking for. Does it? One other thing -- where can the sources be downloaded from? Is there a released tarball available (I haven't been able to find it), or do I need to check it out of SVN? Thanks for your help. Sincerely, Moshe Z. Matitya Kayote Networks, Ltd. |
From: Glyn M. <gly...@gm...> - 2008-06-23 12:00:07
|
Dean, 2008/6/23 Dean Michael Berris <mik...@gm...>: > On Mon, Jun 23, 2008 at 5:25 PM, Glyn Matthews <gly...@gm...> > wrote: > > Hi, > > > > 2008/6/23 Dean Michael C. Berris <dmb...@fr...>: > >> > >> > * Might it be better to add the pion-net version number to > >> > the SVN properties? > >> > >> Our SVN properties? I'm not very familiar with this, can you elaborate? > > > > http://svnbook.red-bean.com/en/1.1/ch07s02.html > > > > We can add a property denoting the version of pion-net that we're working > > from. The subversion system will be able to detect conflicts with > > properties, so we won't accidentally commit changes relating to different > > versions of pion-net. This would be far less error-prone than having to > > check the source code comments. > > > > Ah. So what do I do from the client side? With TortoiseSVN (which I'm looking at now), you can add properties using the dialog (right-click in the directory, select "Properties" in the menu, select the Subversion tab and click the box that says "Properties". They should be empty. Otherwise, from a command line, follow the instructions given in the link I posted. If you need to know more about how properties work in subversion, take at look at the boost repository and check the properties for different files/directories. I expect we'd add something to individual files like the following: $> svn propset pion-version '0.5.6' boost/network/http/parser.hpp HTH, G |
From: Dean M. B. <mik...@gm...> - 2008-06-23 11:55:56
|
Hi Glyn! On Mon, Jun 23, 2008 at 7:49 PM, Glyn Matthews <gly...@gm...> wrote: > Dean, > [snipped] > > I added all of the above to the task list. IMO, sourceforge is very slow > and it has a confusing interface... > Yeah, it is slow and has a confusing interface. No wonder the Boost guys moved out of the Sourceforge tracker. :-P At any rate, I've just branched from http_integration revision 37 to http_integration-149381 -- I should be able to get this done tonight my time, and be able to commit during the day. If anybody wants to take on some of the other tasks, please sound off now so that we can discuss the approaches and possible solutions to difficulties you'll encounter. Thanks Glyn, hope to bring good news soon. :) -- Dean Michael C. Berris Software Engineer, Friendster, Inc. |
From: Glyn M. <gly...@gm...> - 2008-06-23 11:50:02
|
Dean, 2008/6/23 Dean Michael Berris <mik...@gm...>: > On Mon, Jun 23, 2008 at 6:00 PM, Dean Michael Berris > <mik...@gm...> wrote: > > > > No problem. I'm starting on them now. I'm having trouble staying > > logged into sourceforge from where I'm at though (darned load > > balancers) so this may take a while. > > > > Okay, now I need help from either Glyn or Mike because I cant' seem to > stay logged into sourceforge long enough to do this. I'd like to > create the following tasks under the HTTP Integration subproject: > > - Convert pion::net::HTTPParser to templated boost::network::http::parser > (Dean) > - Write isolated http::parser unit-tests that: > * create http::message instances from valid HTTP Message samples > * validates HTTP strings > - Isolate URI handling routines into a boost::network::uri type POD > and create a boost::network::uri::parser nested type > - Write isolated uri::parser unit-tests that: > * create boost::network::uri instances > * tests/specifies boost::network::uri interfaces > - Implement synchronous Boost.Asio based http::client using > http::parser from pion-net > - Write unit tests for synchronous http::client > - Implement HTTP 1.1 presistent connection support in http::client > - Write unit tests for persistent connection http::client > - Implement boost::network::http::message::cookie_map > - Write unit tests for http::message::cookie_map that > * create http::message::cookie_map instances > * specifies/tests http::message::cookie_map interfaces > I added all of the above to the task list. IMO, sourceforge is very slow and it has a confusing interface... G. |
From: Dean M. B. <mik...@gm...> - 2008-06-23 10:41:25
|
On Mon, Jun 23, 2008 at 6:00 PM, Dean Michael Berris <mik...@gm...> wrote: > > No problem. I'm starting on them now. I'm having trouble staying > logged into sourceforge from where I'm at though (darned load > balancers) so this may take a while. > Okay, now I need help from either Glyn or Mike because I cant' seem to stay logged into sourceforge long enough to do this. I'd like to create the following tasks under the HTTP Integration subproject: - Convert pion::net::HTTPParser to templated boost::network::http::parser (Dean) - Write isolated http::parser unit-tests that: * create http::message instances from valid HTTP Message samples * validates HTTP strings - Isolate URI handling routines into a boost::network::uri type POD and create a boost::network::uri::parser nested type - Write isolated uri::parser unit-tests that: * create boost::network::uri instances * tests/specifies boost::network::uri interfaces - Implement synchronous Boost.Asio based http::client using http::parser from pion-net - Write unit tests for synchronous http::client - Implement HTTP 1.1 presistent connection support in http::client - Write unit tests for persistent connection http::client - Implement boost::network::http::message::cookie_map - Write unit tests for http::message::cookie_map that * create http::message::cookie_map instances * specifies/tests http::message::cookie_map interfaces > I'll branch accordingly later. Especially when you have the tracker numbers available to name branches. :D > > I'll update everyone as necessary. > Now I hope either Glyn or Mike can add the tasks above and others that need doing so that we can start assigning these tasks to ourselves and actually track the progress we make. Thanks and I certainly hope to hear from you soon! -- Dean Michael C. Berris Software Engineer, Friendster, Inc. |
From: Dean M. B. <mik...@gm...> - 2008-06-23 10:01:02
|
On Mon, Jun 23, 2008 at 5:25 PM, Glyn Matthews <gly...@gm...> wrote: > Hi, > > 2008/6/23 Dean Michael C. Berris <dmb...@fr...>: >> >> > * Might it be better to add the pion-net version number to >> > the SVN properties? >> >> Our SVN properties? I'm not very familiar with this, can you elaborate? > > http://svnbook.red-bean.com/en/1.1/ch07s02.html > > We can add a property denoting the version of pion-net that we're working > from. The subversion system will be able to detect conflicts with > properties, so we won't accidentally commit changes relating to different > versions of pion-net. This would be far less error-prone than having to > check the source code comments. > Ah. So what do I do from the client side? >> >> > Could we get cracking by adding some requests to the tracker? >> > The best way I see of making this work is to create a new >> > branch from http_integration for each tracker ID, and to >> > merge the changes once each part of pion has been integrated. >> >> I'll do this later today. I'll break tasks into bite-sized pieces and >> take on some of them as I go along as well. > > Great, thanks. No problem. I'm starting on them now. I'm having trouble staying logged into sourceforge from where I'm at though (darned load balancers) so this may take a while. I'll branch accordingly later. I'll update everyone as necessary. -- Dean Michael C. Berris Software Engineer, Friendster, Inc. |
From: Glyn M. <gly...@gm...> - 2008-06-23 09:25:15
|
Hi, 2008/6/23 Dean Michael C. Berris <dmb...@fr...>: > > > * Might it be better to add the pion-net version number to > > the SVN properties? > > Our SVN properties? I'm not very familiar with this, can you elaborate? > http://svnbook.red-bean.com/en/1.1/ch07s02.html We can add a property denoting the version of pion-net that we're working from. The subversion system will be able to detect conflicts with properties, so we won't accidentally commit changes relating to different versions of pion-net. This would be far less error-prone than having to check the source code comments. > > Could we get cracking by adding some requests to the tracker? > > The best way I see of making this work is to create a new > > branch from http_integration for each tracker ID, and to > > merge the changes once each part of pion has been integrated. > > I'll do this later today. I'll break tasks into bite-sized pieces and > take on some of them as I go along as well. > Great, thanks. Glyn |
From: Dean M. C. B. <dmb...@fr...> - 2008-06-23 01:08:26
|
> -----Original Message----- > From: cpp...@li... > [mailto:cpp...@li...] On > Behalf Of Glyn Matthews > Sent: Monday, June 23, 2008 2:54 AM > To: C++ Networking Library Developers Mailing List > Subject: Re: [cpp-netlib-devel] HTTP Integration > > Hi, > > > 2008/6/16 Michael Dickey <mi...@mi...>: > > > Thanks, Jonathan and Divye, for offering your help with > this. I've > also been corresponding off-list with Dean, and he > expects to be able > to help out with this conversion as well. > > I tried dumping the things off the top of my head into > the Wiki that > Glyn setup at > http://cpp-netlib.wiki.sourceforge.net/Http+Integration+Plan > > Please let me know if this is a helpful place to start, > or what else I > can spec out or explain. > > > > Yes, I think this is a good start. Some small points: > > * Might it be better to add the pion-net version number to > the SVN properties? Our SVN properties? I'm not very familiar with this, can you elaborate? > * As for documentation - I'm not sure its such a good idea to > add the docs to the sourceforge page, since we'll need to > convert it to boostbook anyway. Couldn't we use quickbook to > generate the docs and then keep a pointer to the latest > version on-line? > I like quickbook better. Divye, are you up for some documentation writing especially on the messaging framework side of things? That interface is pretty much set, and I don't think we are going to change the interface(s) there anytime soon. Let me know if you need more information to be able to get started. > Could we get cracking by adding some requests to the tracker? > The best way I see of making this work is to create a new > branch from http_integration for each tracker ID, and to > merge the changes once each part of pion has been integrated. I'll do this later today. I'll break tasks into bite-sized pieces and take on some of them as I go along as well. > > Does that sound OK? I'd like it as much as possible for > people to start working on this, and to use the sourceforge > tools (tracker, wiki, mailing list) to generate more > discussion. We're at an important stage with this and I > would like us to gain some more momentum so we can make some > cool software. > Glyn This soulds OK to me. -- Dean Michael Berris Software Engineer, Friendster, Inc. |
From: Glyn M. <gly...@gm...> - 2008-06-22 18:53:47
|
Hi, 2008/6/16 Michael Dickey <mi...@mi...>: > Thanks, Jonathan and Divye, for offering your help with this. I've > also been corresponding off-list with Dean, and he expects to be able > to help out with this conversion as well. > > I tried dumping the things off the top of my head into the Wiki that > Glyn setup at http://cpp-netlib.wiki.sourceforge.net/Http+Integration+Plan > > Please let me know if this is a helpful place to start, or what else I > can spec out or explain. > Yes, I think this is a good start. Some small points: * Might it be better to add the pion-net version number to the SVN properties? * As for documentation - I'm not sure its such a good idea to add the docs to the sourceforge page, since we'll need to convert it to boostbook anyway. Couldn't we use quickbook to generate the docs and then keep a pointer to the latest version on-line? Could we get cracking by adding some requests to the tracker? The best way I see of making this work is to create a new branch from http_integration for each tracker ID, and to merge the changes once each part of pion has been integrated. Does that sound OK? I'd like it as much as possible for people to start working on this, and to use the sourceforge tools (tracker, wiki, mailing list) to generate more discussion. We're at an important stage with this and I would like us to gain some more momentum so we can make some cool software. Glyn |
From: Dean M. B. <mik...@gm...> - 2008-06-20 06:30:06
|
Hi Guys, Sorry I had been out for the last few days due to some health issues. I'd love to get this plan rolling along. I'll start with the pion-net conversion of the lower level pieces (as Mike has posted in the Wiki) and then update everyone as necessary. Have a good one guys. On Tue, Jun 17, 2008 at 3:17 AM, Michael Dickey <mi...@mi...> wrote: > Thanks, Jonathan and Divye, for offering your help with this. I've > also been corresponding off-list with Dean, and he expects to be able > to help out with this conversion as well. > > I tried dumping the things off the top of my head into the Wiki that > Glyn setup at http://cpp-netlib.wiki.sourceforge.net/Http+Integration+Plan > > Please let me know if this is a helpful place to start, or what else I > can spec out or explain. > > Seems to me that the immediate tasks at hand would be: > > a) Update the http::parser so that it uses the new methods and static > constants that were moved over into cpp-netlib (?) > b) Add documentation (to the sourceforge wiki?) for all the new > network/protocol/http stuff > c) Write unit tests for http::parser and the new http::message methods > (some may already exist in pion-net's test suite) > d) converting the URI parser(s) to use Boost.Spirit and craft HTTP 1.1 > requests associated with the URI (from Dean) > e) Incorporate http::parser into http::client so that it is used to > parse server responses > f) Update http::client so that it can use http::request objects to > build & send the request > g) Write new unit tests for http::client > ... > > Take care, > -Mike > > > On Jun 11, 2008, at 3:52 PM, Michael Dickey wrote: > >> Sure, I'd be happy to compile a plan and help coordinate others. I >> think there is still 20-30k lines of code or so to convert over, so >> it's not a small task. >> >> Who on this list would be able to help? >> >> -Mike >> >> >> On Jun 6, 2008, at 4:41 AM, Glyn Matthews wrote: >> >>> Hi Mike, >>> >>> >>> I think its understandable if you can't devote as much time as you'd >>> wish to the project, I have the same problem and I'm sure Dean does >>> too. But there are some students who appear on this list who I >>> think are eager to get some work done, so maybe it could be >>> worthwhile delegating to them. >>> >>> How far along is the integration? What steps do you still need to >>> take? Do you think its a good idea to lay out a small plan (if you >>> have time) so that others can get involved? >>> >>> >>> Regards, >>> Glyn >>> >>> 2008/6/6 Michael Dickey <mi...@at...>: >>> I'm really sorry for disappearing... The past several weeks have >>> just been extremely busy between new hires, customers, funding and >>> working towards a Pion beta launch later this month.. I've been >>> checking periodically for new messages on this list every now and >>> then, but just realized that my mail client was filing them into a >>> different folder.. >>> >>> I just wanted to drop in to say that I'm still very interested in >>> cpp- >>> netlib. I was intending originally to set aside at least a few hours >>> a week to work on the code conversion, but I've found even that to be >>> difficult lately. A lot of other things are starting to wrap >>> themselves up, and between that and the new additions to Atomic Labs, >>> I'm hoping that I will have a more sane schedule soon. >>> >>> Take care, >>> -Mike >>> >>> On May 14, 2008, at 8:57 AM, Dean Michael C. Berris wrote: >>> >>>> Ping Mike? >>>> >>>> -- >>>> Dean Michael Berris >>>> Software Engineer, Friendster, Inc. >>>> <dmb...@fr...> >>>> +639287291459 >>>> >>>> >>> ------------------------------------------------------------------------- >>>> This SF.net email is sponsored by: Microsoft >>>> Defy all challenges. Microsoft(R) Visual Studio 2008. >>>> http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/ >>>> _______________________________________________ >>>> Cpp-netlib-devel mailing list >>>> Cpp...@li... >>>> https://lists.sourceforge.net/lists/listinfo/cpp-netlib-devel >>>> >>> >>> >>> >>> ------------------------------------------------------------------------- >>> Check out the new SourceForge.net Marketplace. >>> It's the best place to buy or sell services for >>> just about anything Open Source. >>> http://sourceforge.net/services/buy/index.php >>> _______________________________________________ >>> Cpp-netlib-devel mailing list >>> Cpp...@li... >>> https://lists.sourceforge.net/lists/listinfo/cpp-netlib-devel >>> >>> ------------------------------------------------------------------------- >>> Check out the new SourceForge.net Marketplace. >>> It's the best place to buy or sell services for >>> just about anything Open Source. >>> http://sourceforge.net/services/buy/index.php_______________________________________________ >>> Cpp-netlib-devel mailing list >>> Cpp...@li... >>> https://lists.sourceforge.net/lists/listinfo/cpp-netlib-devel >> >> >> ------------------------------------------------------------------------- >> Check out the new SourceForge.net Marketplace. >> It's the best place to buy or sell services for >> just about anything Open Source. >> http://sourceforge.net/services/buy/index.php >> _______________________________________________ >> Cpp-netlib-devel mailing list >> Cpp...@li... >> https://lists.sourceforge.net/lists/listinfo/cpp-netlib-devel >> > > > ------------------------------------------------------------------------- > Check out the new SourceForge.net Marketplace. > It's the best place to buy or sell services for > just about anything Open Source. > http://sourceforge.net/services/buy/index.php > _______________________________________________ > Cpp-netlib-devel mailing list > Cpp...@li... > https://lists.sourceforge.net/lists/listinfo/cpp-netlib-devel > -- Dean Michael C. Berris Software Engineer, Friendster, Inc. |
From: Michael D. <mi...@mi...> - 2008-06-16 19:17:49
|
Thanks, Jonathan and Divye, for offering your help with this. I've also been corresponding off-list with Dean, and he expects to be able to help out with this conversion as well. I tried dumping the things off the top of my head into the Wiki that Glyn setup at http://cpp-netlib.wiki.sourceforge.net/Http+Integration+Plan Please let me know if this is a helpful place to start, or what else I can spec out or explain. Seems to me that the immediate tasks at hand would be: a) Update the http::parser so that it uses the new methods and static constants that were moved over into cpp-netlib (?) b) Add documentation (to the sourceforge wiki?) for all the new network/protocol/http stuff c) Write unit tests for http::parser and the new http::message methods (some may already exist in pion-net's test suite) d) converting the URI parser(s) to use Boost.Spirit and craft HTTP 1.1 requests associated with the URI (from Dean) e) Incorporate http::parser into http::client so that it is used to parse server responses f) Update http::client so that it can use http::request objects to build & send the request g) Write new unit tests for http::client ... Take care, -Mike On Jun 11, 2008, at 3:52 PM, Michael Dickey wrote: > Sure, I'd be happy to compile a plan and help coordinate others. I > think there is still 20-30k lines of code or so to convert over, so > it's not a small task. > > Who on this list would be able to help? > > -Mike > > > On Jun 6, 2008, at 4:41 AM, Glyn Matthews wrote: > >> Hi Mike, >> >> >> I think its understandable if you can't devote as much time as you'd >> wish to the project, I have the same problem and I'm sure Dean does >> too. But there are some students who appear on this list who I >> think are eager to get some work done, so maybe it could be >> worthwhile delegating to them. >> >> How far along is the integration? What steps do you still need to >> take? Do you think its a good idea to lay out a small plan (if you >> have time) so that others can get involved? >> >> >> Regards, >> Glyn >> >> 2008/6/6 Michael Dickey <mi...@at...>: >> I'm really sorry for disappearing... The past several weeks have >> just been extremely busy between new hires, customers, funding and >> working towards a Pion beta launch later this month.. I've been >> checking periodically for new messages on this list every now and >> then, but just realized that my mail client was filing them into a >> different folder.. >> >> I just wanted to drop in to say that I'm still very interested in >> cpp- >> netlib. I was intending originally to set aside at least a few hours >> a week to work on the code conversion, but I've found even that to be >> difficult lately. A lot of other things are starting to wrap >> themselves up, and between that and the new additions to Atomic Labs, >> I'm hoping that I will have a more sane schedule soon. >> >> Take care, >> -Mike >> >> On May 14, 2008, at 8:57 AM, Dean Michael C. Berris wrote: >> >>> Ping Mike? >>> >>> -- >>> Dean Michael Berris >>> Software Engineer, Friendster, Inc. >>> <dmb...@fr...> >>> +639287291459 >>> >>> >> ------------------------------------------------------------------------- >>> This SF.net email is sponsored by: Microsoft >>> Defy all challenges. Microsoft(R) Visual Studio 2008. >>> http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/ >>> _______________________________________________ >>> Cpp-netlib-devel mailing list >>> Cpp...@li... >>> https://lists.sourceforge.net/lists/listinfo/cpp-netlib-devel >>> >> >> >> >> ------------------------------------------------------------------------- >> Check out the new SourceForge.net Marketplace. >> It's the best place to buy or sell services for >> just about anything Open Source. >> http://sourceforge.net/services/buy/index.php >> _______________________________________________ >> Cpp-netlib-devel mailing list >> Cpp...@li... >> https://lists.sourceforge.net/lists/listinfo/cpp-netlib-devel >> >> ------------------------------------------------------------------------- >> Check out the new SourceForge.net Marketplace. >> It's the best place to buy or sell services for >> just about anything Open Source. >> http://sourceforge.net/services/buy/index.php_______________________________________________ >> Cpp-netlib-devel mailing list >> Cpp...@li... >> https://lists.sourceforge.net/lists/listinfo/cpp-netlib-devel > > > ------------------------------------------------------------------------- > Check out the new SourceForge.net Marketplace. > It's the best place to buy or sell services for > just about anything Open Source. > http://sourceforge.net/services/buy/index.php > _______________________________________________ > Cpp-netlib-devel mailing list > Cpp...@li... > https://lists.sourceforge.net/lists/listinfo/cpp-netlib-devel > |
From: Divye K. <div...@gm...> - 2008-06-13 17:32:50
|
Sorry for the really late reply. I have been travelling of late and I have been "disconnected" from the internet. I'm still there and willing to help. Divye On 6/13/08, Zhang Jonathan <jon...@gm...> wrote: > HI, > "Who on this list would be able to help?" > > I hope i can do something for cpp-netlib. I want to see your plan. > > Jonathan > > > > ------------------------------------------------------------------------- > Check out the new SourceForge.net Marketplace. > It's the best place to buy or sell services for > just about anything Open Source. > http://sourceforge.net/services/buy/index.php > _______________________________________________ > Cpp-netlib-devel mailing list > Cpp...@li... > https://lists.sourceforge.net/lists/listinfo/cpp-netlib-devel > > -- Whether the chicken crossed the road or the road moved beneath the chicken depends upon your frame of reference. My official web site: http://people.iitr.ernet.in/shp/061305/ |
From: Divye K. <div...@gm...> - 2008-06-13 17:30:23
|
The test suites in the http_integration branch work "as is" by simply running bjam in the tests directory or the root of the branch by using the Visual Studio configured command line (provided bjam is in your path). The tests compile and run, though they may run indefinitely if internet access is not available because if the hosts are not resolved, the code does not exit cleanly. Divye On 6/13/08, gershon bialer <ger...@gm...> wrote: > Does this work running the test suites? I can't figure out how to get the > test suites running on either Visual Studio, and I want to try running them > on Vista. > > Gershon Bialer > > > On Fri, Jun 13, 2008 at 11:12 AM, Michael Dickey <mi...@mi...> > wrote: > > > For cygwin (I believe MinGW) as well, you need to add "-mthreads" to > > your CPPFLAGS and LDFLAGS (compile and link commands) for the > > Boost.Thread library (Boost.ASIO uses it for certain things). You > > might want to try that out too. > > > > Take care, > > -Mike > > > > > > On Jun 12, 2008, at 11:15 PM, Dean Michael C. Berris wrote: > > > > > Hi Allister! > > > > > >> -----Original Message----- > > >> From: cpp...@li... > > >> > [mailto:cpp...@li...] On > > >> Behalf Of Allister Levi Sanchez > > >> Sent: Friday, June 13, 2008 1:41 PM > > >> To: C++ Networking Library Developers Mailing List > > >> Subject: [cpp-netlib-devel] cpp-netlib on vista + MinGW > > >> > > >> Hi, > > >> > > >> I tried to build cpp-netlib (http_integration branch) on > > >> Windows Vista (MinGW compiler) as follows: bjam --toolset=gcc debug > > >> > > >> The build failed however because I don't have the pthread > > >> library. I see that one might be able to use the > > >> pthreads-win32 (http://sourceware.org/pthreads-win32/) > but it > > >> seems rather dated. I'm guessing someone else has met this > > >> issue already. Could you tell me what's the best way to deal > > >> with this? > > >> > > > > > > I personally have not yet tried building in Windows using MinGW > > > though, > > > but if you can get the pthreads library that comes with MinGW or > > > Cygwin, > > > that might bring you farther. Right now though I don't think cpp- > > > netlib > > > explicitly requires threading, but Boost.Asio might -- in this case > > > you > > > might have better mileage by trying to build a minimal application > > > using > > > Boost.Asio first before trying to build with MinGW. > > > > > > Others might have other ideas too. :) > > > > > > -- > > > Dean Michael Berris > > > Software Engineer, Friendster, Inc. > > > > > > > ------------------------------------------------------------------------- > > > Check out the new SourceForge.net Marketplace. > > > It's the best place to buy or sell services for > > > just about anything Open Source. > > > http://sourceforge.net/services/buy/index.php > > > _______________________________________________ > > > Cpp-netlib-devel mailing list > > > Cpp...@li... > > > > https://lists.sourceforge.net/lists/listinfo/cpp-netlib-devel > > > > > > > > > > ------------------------------------------------------------------------- > > Check out the new SourceForge.net Marketplace. > > It's the best place to buy or sell services for > > just about anything Open Source. > > http://sourceforge.net/services/buy/index.php > > _______________________________________________ > > Cpp-netlib-devel mailing list > > Cpp...@li... > > > https://lists.sourceforge.net/lists/listinfo/cpp-netlib-devel > > > > > ------------------------------------------------------------------------- > Check out the new SourceForge.net Marketplace. > It's the best place to buy or sell services for > just about anything Open Source. > http://sourceforge.net/services/buy/index.php > _______________________________________________ > Cpp-netlib-devel mailing list > Cpp...@li... > https://lists.sourceforge.net/lists/listinfo/cpp-netlib-devel > > -- Whether the chicken crossed the road or the road moved beneath the chicken depends upon your frame of reference. My official web site: http://people.iitr.ernet.in/shp/061305/ |