Menu
▾
▴
gnuplot-beta
|
From: Ethan A M. <merritt@u.washington.edu> - 2006-01-11 19:35:00
|
Any chance that one of you can clean up some of the dubious listing on the "gnuplot development" page? The existence of a 3.8k snapshot is hardly breaking news, and should not be highlighted at the top left. I'd love to see it replaced by a pointer to a 4.1 snapshot, but that's a separate question. The "Operating System" section lists us as All 32-bit MS Windows (95/98/NT/2000/XP), OS Portable (Source code to work with many OS platforms) This seems misleading to me. There is no 32-bit restriction in the code, and the focus is not on MS Windows no matter how you look at it. I'd opt to list only "OS portable". If we're going to mention windows in particular, then might as well also tick off the categories "All POSIX" and "Classic" (includes Atari). -- Ethan A Merritt Biomolecular Structure Center University of Washington, Seattle 98195-7742 |
|
From: Petr M. <mi...@ph...> - 2006-01-11 20:52:43
|
> Any chance that one of you can clean up some of the > dubious listing on the "gnuplot development" page? I don't see anything you mention below on http://gnuplot.sourceforge.net/development/index.html > The existence of a 3.8k snapshot is hardly breaking in binaries/ ? I've removed that 38k-gpbin.zip, as it was actually just a dummy version. > The "Operating System" section lists us as > All 32-bit MS Windows (95/98/NT/2000/XP), > OS Portable (Source code to work with many OS platforms) ??? --- PM |
|
From: Petr M. <mi...@ph...> - 2006-01-11 22:14:56
|
> http://sourceforge.net/projects/gnuplot/ > > The page title is "gnuplot development". The "3.8k" message should be removed. Added should be: You can follow current activity in gnuplot 4.1 development at http://gnuplot.sourceforge.net/development/ >>> The "Operating System" section lists us as >>> All 32-bit MS Windows (95/98/NT/2000/XP), >>> OS Portable (Source code to work with many OS platforms) Also see there: Programming Language : C, C++, Visual Basic Could someone edit all these information? --- PM |
|
From: Clark G. <ga...@di...> - 2006-01-11 23:29:03
|
On Wed, 11 Jan 2006 11:34:50 -0800, "Ethan A Merritt" <merritt@u.washington.edu> said: > Any chance that one of you can clean up some of the > dubious listing on the "gnuplot development" page? > > The existence of a 3.8k snapshot is hardly breaking > news, and should not be highlighted at the top left. > I'd love to see it replaced by a pointer to a 4.1 snapshot, > but that's a separate question. Done. > The "Operating System" section lists us as > All 32-bit MS Windows (95/98/NT/2000/XP), > OS Portable (Source code to work with many OS platforms) > > This seems misleading to me. There is no > 32-bit restriction in the code, and the focus is not > on MS Windows no matter how you look at it. I'd opt to > list only "OS portable". If we're going to mention I'm not so sure I agree. We have generally had semi-supported Windows distributions, but I don't think there is anyone doing regular regression testing on Windows 286 anymore. If you'd like to take that on, we can add that. :-) But it isn't a matter of there being a Windows focus -- of course we don't focus on it! But we do tend to have a useable version that someone keeps reasonably maintained in binary distro, so it is reasonable to attract users for that purpose. Maybe no one does any Windows regression testing or binary distribution anymore -- would anyone care to 'fess up here? > windows in particular, then might as well also tick > off the categories "All POSIX" and "Classic" (includes Atari). POSIX: done Do we keep a reasonably modern binary build for Atari? We do tend to do this for Windows, and we have source tarballs for *nix. I don't recall the current status of Mac binaries (looks like the link is broken), but that could be a possible addition if some claims it. I certainly don't see deleting Windows unless we really disavow all knowledge. I think we tend to have a *lot* of users consider gnuplot only because there is a Windows binary distro. I have added the "All POSIX", though ... I think that is fair advertising and likely to be something people will search for. I think the standard of "gets at least some regression testing fairly regularly and/or is a specific design objective" is the standard to use here, and as such I think it is fair to say: - Linux/BSD/*nix/POSIX - source that will work in a lot of misc places - Windows (only suitable to include if we also do binary) - maybe MacOS (if someone would confirm that this is a platform they regularly build on, I'll add this ... in the past, this has been pretty touch-and-go) - otherwise as someone specifically claims blame (Do we still care about OS/2 or DOS? We do have 4.0 builds available for these.) We could use a more inclusive net -- I still keep the historical stuff, including the HPLX port (which I actually use on occassion :-) -- on the ftp site, but I don't know that it is fair to include that as a "supported" platform for development. Maybe a more complete list of older "unsupported" platforms could be organized, but I don't think that's the point. For "development" purposes, I think my definition above is pretty reasonable, though I'd be happy to hear from those who have differing views to consider. --ckg -- Clark Gaylord Blacksburg, VA USA ga...@di... |
|
From: Petr M. <mi...@ph...> - 2006-01-12 15:22:03
|
> - Linux/BSD/*nix/POSIX > about OS/2 yes, OS/2 is actively supported Maybe VMS should be included as well. It was working "some time ago". --- PM |
|
From: Hans-Bernhard B. <br...@ph...> - 2006-01-12 14:53:00
|
Ethan A Merritt wrote: > Any chance that one of you can clean up some of the > dubious listing on the "gnuplot development" page? Clark did some, now I did some more. Most importantly, I removed the bogus entries claiming our license were GPL or LGPL. > The existence of a 3.8k snapshot is hardly breaking > news, and should not be highlighted at the top left. Clark apparently just deleted that entry... But keep in in mind that it's not SF's fault if we produce so few news that completely outdated news still hasn't dropped below their radar? > This seems misleading to me. There is no > 32-bit restriction in the code, No intentional one, sure. But as a matter of fact the 16-bit builds haven't worked for quite a while now --- the last time I managed to squeeze gnuplot in that small box was for release 3.7.3. I tried to do something about that in the run-up period to the 4.0 release, but eventually had to give up when it became clear that the necessary changes were too big to contemplate that shortly before a release. I even held a little poll in the newsgroup to find out if anyone cared --- looked like nobody did, so we abandoned that. As it is right now, a 16-bit gnuplot doesn't build even with *all* terminal drivers disabled. The break-neck problem is that graphics.c alone now contains more than 64K of code. I managed to produce a build only by effectively breaking graphics.c in two and applying the 'outboard postscript includes' patch. It still crashes rather lot, though, and I don't think it is worthwile tracking those down ;-( > and the focus is not on MS Windows no matter how you look at it. Ours isn't, but if you look at what the users have "voted with their feet", their message is quite clear: The win32 binary packages have seen about as many downloads as the source and all other binary packages combined. > I'd opt to list only "OS portable". I added a more-or-less comprehensive list of all our supported platforms. Since the list is now much longer, the perceived focus on Win32 should now be less pronounced. We can't do anything about Win32 being listed before "portable", though --- well, not without pestering SF staff about it. |
|
From: Ethan A M. <merritt@u.washington.edu> - 2006-01-12 16:23:37
|
On Thursday 12 January 2006 06:54 am, Hans-Bernhard Broeker wrote: > Ethan A Merritt wrote: > > > This seems misleading to me. There is no > > 32-bit restriction in the code, > > No intentional one, sure. But as a matter of fact the 16-bit builds > haven't worked for quite a while now Heh. I wasn't even thinking of 16-bit. To me it read like a warning that the code wasn't 64-bit safe. > > and the focus is not on MS Windows no matter how you look at it. > > Ours isn't, but if you look at what the users have "voted with their > feet", their message is quite clear: The win32 binary packages have > seen about as many downloads as the source and all other binary packages > combined. I question that interpretation of the download stats. Downloading a binary is about the only way Windows users can get the package at all. On the linux side, gnuplot is included as a regular part of every major distribution. Really the only people who need to download from SourceForge are the ones interested in building the cvs version. -- Ethan A Merritt Biomolecular Structure Center University of Washington, Seattle 98195-7742 |
|
From: Hans-Bernhard B. <br...@ph...> - 2006-01-12 18:46:31
|
Ethan A Merritt wrote: > On Thursday 12 January 2006 06:54 am, Hans-Bernhard Broeker wrote: >>Ethan A Merritt wrote: >>>This seems misleading to me. There is no >>>32-bit restriction in the code, >>No intentional one, sure. But as a matter of fact the 16-bit builds >>haven't worked for quite a while now > Heh. I wasn't even thinking of 16-bit. > To me it read like a warning that the code wasn't 64-bit safe. Well, since there's no heading for 64-bit Windows, nor for "all MS Windows" in SF.net's project categorization system, I don't think that can readily be helped. >>>and the focus is not on MS Windows no matter how you look at it. >>Ours isn't, but if you look at what the users have "voted with their >>feet", their message is quite clear: The win32 binary packages have >>seen about as many downloads as the source and all other binary packages >>combined. > I question that interpretation of the download stats. > Downloading a binary is about the only way Windows users can get the > package at all. Only if one equates "Windows users" with "computer illiterates" --- which may have dangerously much truth to it, but still isn't exactly true ;-) It's not exactly rocket science to build gnuplot on Windows. It's just a little uncomfortable having to find and install all those libraries first. Right now the expected roadblock for Win32 builds would be the considerable mismatch between the way Boutell packages his Win32 GD binary distribution, and our pre-built makefiles. None of our prepared makefiles or configure scripts expects to find GD installed as -lbgd, with all the underlying libraries (freetype, jpeg, png, ...?) swallowed into that 3 MB blob of a DLL. > Really the only people who need to > download from SourceForge are the ones interested in building the cvs > version. Ahem... CVS users don't show up in any of those download stats. We're talking here about downloads of a 4.0.0 release source tarball vs. those of Win32 binaries made from it. |
|
From: Ethan M. <merritt@u.washington.edu> - 2006-01-12 19:04:28
|
On Thursday 12 January 2006 10:47 am, Hans-Bernhard Broeker wrote: > > Ahem... CVS users don't show up in any of those download stats. > We're talking here about downloads of a 4.0.0 release source tarball > vs. those of Win32 binaries made from it. That was my point. A low count for people downloading the 4.0 release tarball may not mean anything, given that the same source is already in their Redhat/Suse/Mandriva/etc distribution. If I want to rebuild from the 4.0 for some reason, I can type "rpm --rebuild gnuplot-4.0.0-7mdk.src.rpm". That won't show up in any SourceForge stat. In fact, I may do exactly that in an attempt to pin down that x11 double-free error. -- Ethan A Merritt Biomolecular Structure Center University of Washington, Seattle WA |
