Re: [boost] Review Queue Needs Attention
judging by the review schedule (almost no review managers), that would mean boost isn't very interested in new libraries in general.
It would seem so.
I would recommend that authors try find a review manager before they request a review.
Review managers are generally recruited from people that already have a library that has been approved in boost, that means you luke (hint). The wizards could open it up to anyone. Its been considered, but rejected in the past. I would like to focus attention on something different. That would be creating a "non-stable" branch of boost, where the proposed libraries would live for a while, get some exposure. Since most libraries are header file only, it should not be a problem. While not all will compile on all platforms, and certainly not have perfect documentation, it would be a place for future boost authors to elicit feedback. The experimental libraries would not affect the core boost libraries in any way. I'm absolutely convinced that it would encourage lots more participation. There is so much more to do, we haven't even scratched the surface of what is possible. The core boost authors have put in place a wonderful place to come and share ideas, but we are letting it flounder. Ugh.
I would like to focus attention on something different.
That would be creating a "non-stable" branch of boost, where the proposed libraries would live for a while, get some exposure.
Since most libraries are header file only, it should not be a problem. While not all will compile on all platforms, and certainly not have perfect documentation, it would be a place for future boost authors to elicit feedback.
The experimental libraries would not affect the core boost libraries in any way.
I'm absolutely convinced that it would encourage lots more participation.
What's the difference between a non-stable branch as you're suggesting and the existing sandbox? The only difference I could see is to establish some kind of pre-review process. But that would just move the problem to a different spot, no? Regards Hartmut ------------------- Meet me at BoostCon http://boostcon.com
Hartmut Kaiser wrote:
What's the difference between a non-stable branch as you're suggesting and the existing sandbox? The only difference I could see is to establish some kind of pre-review process. But that would just move the problem to a different spot, no?
To my mind, there are two key differences: 1. The sandbox does not have a release cycle. There are no sandbox packages released. 2. Somewhat connected to #1. The sandbox is not tested regularly and widely. The most testing level we can expect is local testing by the library authors and enthusiast users. These two key drawbacks has always prevented me from using anything from sandbox in real world applications. If Boost does release a non-stable branch, it should receive at least periodical testing on key compilers and a release cycle (be that the same cycle that core Boost libraries have or not). Otherwise I don't see much sense in it.
Tom Brinkman-2 wrote:
judging by the review schedule (almost no review managers), that would mean boost isn't very interested in new libraries in general.
It would seem so.
I would recommend that authors try find a review manager before they request a review.
Review managers are generally recruited from people that already have a library that has been approved in boost, that means you luke (hint).
The wizards could open it up to anyone. Its been considered, but rejected in the past.
I would like to focus attention on something different.
That would be creating a "non-stable" branch of boost, where the proposed libraries would live for a while, get some exposure.
There had already a lot of review managers that do not have a library accepted, for example Thread-Safe Signals Frank Hess Stjepan Rajko Finite State Machines Andrey Semashev Martin Vuille Singleton (fast-track) Tobias Schwinger John Torjo Switch Steven Watanabe Stejpan Rajko Physical Quantities System Andy Little Fred Bertsch binary_int Scott Schurr and Matt Calabrese Pavel Vozenilek Xpressive Eric Niebler Thomas Witt Typeof Arkadiy Vertleyb and Peder Holt Andy Little Singleton Jason Hise Pavel Vozenilek State Chart Andreas Huber Pavel Vozenilek Promotion Traits (fast-track) Alexander Nasonov Tobias Schwinger Output Formatters Reece Dunn John Torjo FC++ Brian McNamara & Yannis Smaragdakis Mat Marcus Fixed-Point Decimal Bill Seymour Jens Maurer Math Constants Paul A. Bristow Jaap Suter ... and there are other on the review queue AutoBuffer Thorsten Ottosen Robert Stewart Task Oliver Kowalke Vicente Botet without forgotten the review Wizard John R. Phillips and Tom Brinkman Accumulators Eric Niebler John R. Phillips Function Types (Re-review) Tobias Schwinger Tom Brinkman Generic Image Library Lubomir Bourdev Tom Brinkman Wave Hartmut Kaiser Tom Brinkman As you can see review manager is not restricted to library authors. I call any invested booster to request the possibility to be the review manager of a library if he knows the domain. Since most libraries are header file only, it should not be a problem. While not all will compile on all platforms, and certainly not have perfect documentation, it would be a place for future boost authors to elicit feedback. The experimental libraries would not affect the core boost libraries in any way. I'm absolutely convinced that it would encourage lots more participation. There is so much more to do, we haven't even scratched the surface of what is possible. The core boost authors have put in place a wonderful place to come and share ideas, but we are letting it flounder. I don't see clearly which will be the criteria for including a library in this non-stable group. Any suggestions? Which test platforms will be available to test the non-stable libraries, the same as the trunck? Will these libraries be delivered? Doesn't the Sandbox pays already the role of non-stable libraries? Best, Vicente -- View this message in context: http://old.nabble.com/Review-Queue-Needs-Attention-tp26500094p26507135.html Sent from the Boost - Dev mailing list archive at Nabble.com.
Hello, please forgive my bad english - I am not a native speaker.
The number of review requests is far outstripping the number of volunteer review managers.
is the lack of review managers the bottleneck of the review process right now?
I`m relatively new to boost, but judging from the "notes for review managers" on the website that doesn`t sound like a lot of work. I'm planning to submit a library for review myself, and the prospect that that'd take 2 years isn't very encouraging, so I'd probably volunteer just for this reason alone.
I would like to help too. I would do the work of a review manager for a library too. But I am new to boost. Is there in the boost documentation a page, which describes what a review manager must do. E.g. how to start a review, how to collect the answers to the result, other things to know? Greetings Detlef Wilkening
Am Thursday 26 November 2009 21:57:44 schrieb Detlef Wilkening:
I`m relatively new to boost, but judging from the "notes for review managers" on the website that doesn`t sound like a lot of work. I'm planning to submit a library for review myself, and the prospect that that'd take 2 years isn't very encouraging, so I'd probably volunteer just for this reason alone.
I would like to help too. I would do the work of a review manager for a library too. But I am new to boost. Is there in the boost documentation a page, which describes what a review manager must do. E.g. how to start a review, how to collect the answers to the result, other things to know?
Hi, I'm following the discussions about potential changes in the boost organization and I'm happy to see that a lot of people here wants to fix those problems ( review management, separation of library states (or libraries), etc.). I'm a game developer but I'm only using boost in my home projects. At work I'm currently not using C++ but have been for years in previous jobs. I'm not a C++ guru yet, just to be clear. We didn't use boost in my previous game company position, for a lot of reasons, one being the complexity of the implementations that a lot of my coworkers (and myself for a long time) couldn't follow -- making debug hard as if you don't trust a libary and the errors occurs in the library code, you're not sure if the bug is yours or from the library... . It wouldn't have been a problem as learning is part of the job, but there were deadlines. That have already been said so I just add a quick comment about that. (there were technical reasons too relative to the compilers and embedded systems we were working on) --- About potentially replacing SVN by another source control system, I thought that it would be helpful to point you to the OGRE (C++ OO graphic engine) lead blog where he reported his data while evaluating Git and Mercurial (Bazaar was also cited). It might be a good documentation about the current state of these solutions as a lot of analysis is already done for you. - the most interesting comparison : http://www.stevestreeting.com/2009/11/06/dvcs-score-card/ - you can find other discussions on this page : http://www.stevestreeting.com/category/development/ - the community poll thread : http://www.ogre3d.org/forums/viewtopic.php?f=4&t=53129 To sum-up, currently the choice of the Ogre lead would be Mercurial, the reasons being windows support that is poor for git, git error reporting being problematic or non-existant and mercurial being easier to learn (git asking a bit of internal guts knowledge to be well used). Now, I don't want to start a flamewar too, I'm pointing here those analysis as interesting documentation to get a faster decision if the source control system really have to be changed. There are too much slow processes in the C++ communities. I hope that some boost libraries will be reviewed or aproved for a near-complete-or-stable stage soon and I'm using some from of them currently with the last boost version. It would be very usefull to get the libraries separately from a source control system, making upgrading a specific library to fix a bug in the user application easier. Joël Lamotte. On Thu, Nov 26, 2009 at 23:56, Stefan Strasser <strasser@uni-bremen.de>wrote:
Am Thursday 26 November 2009 21:57:44 schrieb Detlef Wilkening:
I`m relatively new to boost, but judging from the "notes for review managers" on the website that doesn`t sound like a lot of work. I'm planning to submit a library for review myself, and the prospect that that'd take 2 years isn't very encouraging, so I'd probably volunteer just for this reason alone.
I would like to help too. I would do the work of a review manager for a library too. But I am new to boost. Is there in the boost documentation a page, which describes what a review manager must do. E.g. how to start a review, how to collect the answers to the result, other things to know?
http://www.boost.org/community/reviews.html#Review_Manager _______________________________________________ Unsubscribe & other changes: http://lists.boost.org/mailman/listinfo.cgi/boost
participants (7)
-
Andrey Semashev -
Detlef Wilkening -
Hartmut Kaiser -
Klaim -
Stefan Strasser -
Tom Brinkman -
Vicente Botet Escriba