Thanks for your suggestions, I cc''d this to the developer''s list. Curt On 10/18/07, Roger Pack <rogerpack2005 at gmail.com> wrote:> > Hi Curt. Thought I''d give some feedback stuffs on one-click after the > latest announcement on the ruby mailing list. > Some random thoughts: > One-click rocks! > It might be worth looking into compiling it with VC8, as apparently it is > speedier (see http://www.ruby-forum.com/topic/59411#new which seemed to > say so at the bottom). > > A suggestion--when it installs I''m still a little lazy and wish that there > were an option ''just overwrite everything in my ruby directory!'' maybe with > a warning ''this is dangerous'' and ''make sure to turn off any ruby programs'' > and ''unspecified results,'' or what not. I tried doing that recently (seeing > if ploughing over it resulted in a working distro) and it seems to work > fine. > > A note--it still says still says 185-24 in the folder under "program > files" (I believe). > > And a small suggestion--when I first started using one click I got > confused as to what 185-24 meant in comparison to the ruby core patch > level. "Were they the same?" "What patch level was 185-24 using, anyway?" > That type of thing. It might be helpful to newbies like me if the numbering > sequence were modified slightly, such as 186-26-111 or 186-111-26, so that I > don''t get so confused. > > Anyway that''s it! Thanks for your work. Worked well (didn''t override my > path, and I had pre-existing Ruby). I''m glad to finally have a working ruby > again :) > -- > -Roger Pack > I like belief. http://www.google.com/search?q=free+bible > >-------------- next part -------------- An HTML attachment was scrubbed... URL: http://rubyforge.org/pipermail/rubyinstaller-devel/attachments/20071018/c6686936/attachment.html
---------- Forwarded message ---------- From: Roger Pack <rogerpack2005 at gmail.com> Date: Oct 18, 2007 12:49 PM Subject: Re: one click installer thoughts To: curt at hibbs.com Cc: RubyInstaller Developers <rubyinstaller-devel at rubyforge.org> Oops looks like the fastest win32 compiler is ruby 1.8.6 (2007-03-13 patchlevel 0) [i386-mingw32] (gcc 4.2.1-dw2-2) also cc''ed to devs--will probably bounce :) Take care.> It might be worth looking into compiling it with VC8, as apparently it is > > speedier (see http://www.ruby-forum.com/topic/59411#new which seemed to > > say so at the bottom). > > >-------------- next part -------------- An HTML attachment was scrubbed... URL: http://rubyforge.org/pipermail/rubyinstaller-devel/attachments/20071018/9173a2a1/attachment.html
On 10/18/07, Curt Hibbs <curt.hibbs at gmail.com> wrote:> > > ---------- Forwarded message ---------- > From: Roger Pack <rogerpack2005 at gmail.com> > Date: Oct 18, 2007 12:49 PM > Subject: Re: one click installer thoughts > To: curt at hibbs.com > Cc: RubyInstaller Developers > <rubyinstaller-devel at rubyforge.org > > > Oops looks like the fastest win32 compiler is > ruby 1.8.6 > (2007-03-13 patchlevel 0) [i386-mingw32] (gcc 4.2.1-dw2-2) > > > also cc''ed to devs--will probably bounce :) > Take care. > >Thanks Curt for the update. The thing regarding gcc-4.2.1 is that is Technology Preview in MinGW. What does mean? The stable and inmediate candidates version of GCC under windows (mingw) are 3.x, no plan for inmediate jump/mark as stable the 4.x branch. Since we are trying ot get a better product build under Windows, we should avoid using too bleeding edge stuff. Trying to trace a bug when you don''t know if it''s the compiler, ruby or the steps you''re doing. Anyway, there is a performance compromise like always, and I lested those in the ruby-core thread, but copy & paste here for the reference (since cannot find the ruby-core number for it:P Quote of what I replied to _why about the compilers and the options we have for Windows: ===So, to summarize we have a few paths: - Keep using VC6 and find workarounds to compile extensions with other compilers (maybe with better documentation for this topic). Pro: we are already there. Con: bad performance and issues with the compiler- - Implement some sort of distutils for ruby Pro: will be good for ruby at the end. Con: will take time and much more to be integrated as part of ruby-core. - Switch to VC8/9. Pro: - Better performance, compiler freely available (in the shape of Visual Studio Express or Windows SDK). - Known x64 support availability, for future developments/switch to 64bits OS -- (There is a Ruby 1.9 build against VC8_x64.) Con: - Requires hacking and fixing of upstream libraries (dependencies) which aren''t prepared for it. (gnu tools like readline, iconv, gettext, and so on as example). - Switch to MinGW Pro: will ease the path of getting dependencies build (like the ones cited before). It also link against already known runtime library (msvcrt.dll) which is available in every stable OS since Windows 2000 -- AFAIK. Con: This will also require maintain builds of upstream libraries, even are more simpler to them build. The support for x64 platforms is still under development -- is not part of stable either candidates of current MinGW releases... which is a problem (is another branch of project looking to get integrated). Ok, these are the scenarios, and their good and bad points. for those who every time just reply that "we should stick to MS compilers" in the field I haven''t found (excluding the 64bits part) no differences from using MinGW. On the contrary, and thankfuly to the "simple installer" they have, get the correct dev environment is simpler than Microsoft tools. == -- Luis Lavena Multimedia systems - Leaders are made, they are not born. They are made by hard effort, which is the price which all of us must pay to achieve any goal that is worthwhile. Vince Lombardi
Good summary, Luis! On 10/18/07, Luis Lavena <luislavena at gmail.com> wrote:> > On 10/18/07, Curt Hibbs <curt.hibbs at gmail.com> wrote: > > > > > > ---------- Forwarded message ---------- > > From: Roger Pack <rogerpack2005 at gmail.com> > > Date: Oct 18, 2007 12:49 PM > > Subject: Re: one click installer thoughts > > To: curt at hibbs.com > > Cc: RubyInstaller Developers > > <rubyinstaller-devel at rubyforge.org > > > > > Oops looks like the fastest win32 compiler is > > ruby 1.8.6 > > (2007-03-13 patchlevel 0) [i386-mingw32] (gcc 4.2.1-dw2-2) > > > > > > also cc''ed to devs--will probably bounce :) > > Take care. > > > > > > Thanks Curt for the update. > > The thing regarding gcc-4.2.1 is that is Technology Preview in MinGW. > > What does mean? > > The stable and inmediate candidates version of GCC under windows > (mingw) are 3.x, no plan for inmediate jump/mark as stable the 4.x > branch. > > Since we are trying ot get a better product build under Windows, we > should avoid using too bleeding edge stuff. > > Trying to trace a bug when you don''t know if it''s the compiler, ruby > or the steps you''re doing. > > Anyway, there is a performance compromise like always, and I lested > those in the ruby-core thread, but copy & paste here for the reference > (since cannot find the ruby-core number for it:P > > Quote of what I replied to _why about the compilers and the options we > have for Windows: > > ===> So, to summarize we have a few paths: > > - Keep using VC6 and find workarounds to compile extensions with other > compilers (maybe with better documentation for this topic). > Pro: we are already there. > Con: bad performance and issues with the compiler- > > - Implement some sort of distutils for ruby > Pro: will be good for ruby at the end. > Con: will take time and much more to be integrated as part of ruby-core. > > - Switch to VC8/9. > Pro: > - Better performance, compiler freely available (in the shape of > Visual Studio Express or Windows SDK). > - Known x64 support availability, for future developments/switch to > 64bits OS -- (There is a Ruby 1.9 build against VC8_x64.) > Con: > - Requires hacking and fixing of upstream libraries (dependencies) > which aren''t prepared for it. > (gnu tools like readline, iconv, gettext, and so on as example). > > - Switch to MinGW > Pro: will ease the path of getting dependencies build (like the ones > cited before). It also link against already known runtime library > (msvcrt.dll) which is available in every stable OS since Windows 2000 > -- AFAIK. > Con: This will also require maintain builds of upstream libraries, > even are more simpler to them build. > The support for x64 platforms is still under development -- is not > part of stable either candidates of current MinGW releases... which is > a problem (is another branch of project looking to get integrated). > > > Ok, these are the scenarios, and their good and bad points. > > for those who every time just reply that "we should stick to MS > compilers" in the field I haven''t found (excluding the 64bits part) no > differences from using MinGW. On the contrary, and thankfuly to the > "simple installer" they have, get the correct dev environment is simpler > than > Microsoft tools. > > ==> > -- > Luis Lavena > Multimedia systems > - > Leaders are made, they are not born. They are made by hard effort, > which is the price which all of us must pay to achieve any goal that > is worthwhile. > Vince Lombardi > _______________________________________________ > Rubyinstaller-devel mailing list > Rubyinstaller-devel at rubyforge.org > http://rubyforge.org/mailman/listinfo/rubyinstaller-devel >-------------- next part -------------- An HTML attachment was scrubbed... URL: http://rubyforge.org/pipermail/rubyinstaller-devel/attachments/20071019/6195df62/attachment.html