Using gtkmm in inkscape
Hi, someone pointed this status page out to me: http://www.inkscape.org/status/status_20031231.php
So I thought I would offer some advice, quoting that page:
A second and much more critical issue has been raised regarding Gtkmm. As planned since early in the project, we began incorporating Gtkmm into Inkscape following the 0.36 release. Quickly it became evident that this introduced severe dependency issues.
Gtkmm 2.x is not available commonly on most Linux distros.
It is easily available for debian and RedHat/Fedora, from official distro sources: http://www.gtkmm.org/download.shtml
A Mandrake developer assured me that they would ship it, but I don't have a Mandrake installation and I don't know how to check that. I'm fairly sure that it's easily available for Gentoo.
SUSE have no public bug tracking or feedback system that I know of, but hopefully that will get better. With no contributions system, you will have difficult installing Inkscape on SUSE whether or not you need gtkmm.
What other distros are you interested in?
On Windows it adds to the DLL download weight.
I'm not sure what significance the download size has these days. By the way, gtkmm is easily installable on Windows: http://www.gtkmm.org/download.shtml
Most distressingly, it isn't yet ported to OSX.
It builds, and works, and Julian Missig built a darwinports package: http://mail.gnome.org/archives/gtkmm-list/2003-December/msg00122.html
I don't know much about MacOS X these days, but I suspect that you want it as a Fink package. That should not be at all difficult to do. Please tell ask on the gtkmm-list if you need help with that. I would appreciate being CCed on any communication with the Fink people.
Murray Cumming www.murrayc.com murrayc@...167...
Murray.Cumming@...166... wrote:
Hi, someone pointed this status page out to me: http://www.inkscape.org/status/status_20031231.php
Thank you, Murray, for clearing that up! I think we will probably release at least one release without gtkmm dependence as it is currently not needed, but we have some good gtkmm code waiting in the wings.
njh
Murray.Cumming@...166... wrote:
It is easily available for debian and RedHat/Fedora, from official distro sources: http://www.gtkmm.org/download.shtml
A Mandrake developer assured me that they would ship it, but I don't have a Mandrake installation and I don't know how to check that. I'm fairly sure that it's easily available for Gentoo.
SUSE have no public bug tracking or feedback system that I know of, but hopefully that will get better. With no contributions system, you will have difficult installing Inkscape on SUSE whether or not you need gtkmm.
What other distros are you interested in?
Of course it is available. The question is, "Is it already there?" The average blue-haired old lady might be puzzled a bit by a dialog: "You must have gtkmm-2.4-x.rpm installed for Inkscape to work" 'RPM Hell' is becoming a major part of the Download Factor.
On Windows it adds to the DLL download weight.
I'm not sure what significance the download size has these days. By the way, gtkmm is easily installable on Windows: http://www.gtkmm.org/download.shtml
This is the other major part. Every little bit hurts. If we need it, include it. (We already did for a week or so.) But don't frivolously add dependencies. This is true especially for DLLs, as the libraries are supplied monolithically; you cannot choose which elements of a given library that you want.
Most distressingly, it isn't yet ported to OSX.
It builds, and works, and Julian Missig built a darwinports package: http://mail.gnome.org/archives/gtkmm-list/2003-December/msg00122.html
This is excellent. The only issue I can think of, is that we should keep track of what version of GCC made the gtkmm and sigc++ libs, and make sure that the version we use has the same ABI.
Bob
Murray.Cumming@...166... wrote:
Most distressingly, it isn't yet ported to OSX.
It builds, and works, and Julian Missig built a darwinports package: http://mail.gnome.org/archives/gtkmm-list/2003-December/msg00122.html
I don't know much about MacOS X these days, but I suspect that you want it as a Fink package. That should not be at all difficult to do. Please tell ask on the gtkmm-list if you need help with that. I would appreciate being CCed on any communication with the Fink people.
That's probably a good course.
I touched base with the libsigc++ and gtkmm maintainers. The former got libsigc++ 1.2 pushed into stable the next day for us, since it had already in untable for a while and seemed to have no problems.
gtkmm, on the other hand, has no 2.x support. The maintainer himself had no need for it, so hadn't done things to bump up from 1.x. (seems reasonable and to be expected to me)
He did say that:
I will take a look at it later today, but can't promise anything. In the meantime, you can either try to do a package yourself, or install it (at least for now) manually in /usr/local, or post a package request into our a package tracker (or a combination of all :-)
So... that's where fink gtkmm 2.x stands at the moment. I hadn't quite gotten to adding a package request to the package tracker. If our own mac people don't come up with anything, we'll probably go ahead and ask on the gtkmm list and let that mailing list "own" getting a fink package of 2.x.
Personally... I feel that OS X and Win32 are the two main drivers for package adoption. I think that if they are on OS X and Win32, then linux will be covered. But at the same time, we don't want to cut off either of those just by adopting a new library.
Jon A. Cruz wrote:
Personally... I feel that OS X and Win32 are the two main drivers for package adoption. I think that if they are on OS X and Win32, then linux will be covered. But at the same time, we don't want to cut off either of those just by adopting a new library.
Jon, we _DON'T_ want lots of users just yet. Look what happens when you get lots of users: they complain about features not being available; they complain that the interface isn't exactly like their favourite program XYZ; they clutter mailing lists up with questions answered already in the FAQ or user manual(which we haven't even written yet); and you spend the whole time worrying about fixing code to compile on borland visual gcc 2.96.
Can we stop treating this as a commercial project whose main aim is to maximise user base? I've been on such projects before, and they usually die a horrible death.
njh
On Mon, 2004-01-12 at 16:40, Nathan Hurst wrote:
Jon, we _DON'T_ want lots of users just yet. Look what happens when you get lots of users: they complain about features not being available; they complain that the interface isn't exactly like their favourite program XYZ; they clutter mailing lists up with questions answered already in the FAQ or user manual(which we haven't even written yet); and you spend the whole time worrying about fixing code to compile on borland visual gcc 2.96.
Can we stop treating this as a commercial project whose main aim is to maximise user base? I've been on such projects before, and they usually die a horrible death.
Well, I still share Jon's concerns as far as developer recruitment goes.
Anyway, I'm a firm believer in "if you build it, they will come." Maybe not as fast and as furiously as if it were agressively marketed, but IMO we need some breathing room to gain (and keep) our footing anyway.
We're already getting mentioned in the same breath as Evolution and the GIMP on Slashdot. Let's not tempt fate. ^_-
-mental
participants (5)
-
unknown＠example.com -
Bob Jamison -
Jon A. Cruz -
MenTaLguY -
Nathan Hurst