Re: [Inkscape-devel] Fedora Core 3 is no more, Weeee!
Thanks for the info. I have moved to kubuntu 6.06 and someone has kindly made a .deb file so I am in 0.44 heaven now.
The thing I get from this thread is that Linux distros move so quickly, they are having their libraries and other wobbly bits coded so rapidly that it's impossible to keep pace with them all, let alone just one of them.
This is due to the obviously limited resources and voluntary nature of those (like you guys) writing OSS applications to run on Linux.
In an ironic way this speed of improvement of Linux is a strength for Windows. I demonstrate this by means of Inkscape 0.44 and how one can with relatively no trouble develop for Windows and release any old time and users like me can pick up the binary and run it. No mess. No fuss.
That's the contrast for me I suppose. I'll stick with Linux thanks :) That means sticking with these weird version situations.
I wonder if the situation will ever reach a balance, where the vital libs don't move so quickly because they are finished, done, complete and so Linux won't change so massively internally and will approach a kind of binary stability like win has now?
Perhaps I've got the wrong idea completely!
Anyway, thanks for your time.
D.
Donn wrote:
In an ironic way this speed of improvement of Linux is a strength for Windows. I demonstrate this by means of Inkscape 0.44 and how one can with relatively no trouble develop for Windows and release any old time and users like me can pick up the binary and run it. No mess. No fuss.
The Windows case is not that perfect. Inkscape had to drop support for Windows 95/98/ME because one of the dependencies (gtk+) was not working anymore on those operating systems (and nobody bothered enough to fix the regression at the gtk level).
On Tue, 08 Aug 2006 04:30:13 +0200, Donn wrote:
In an ironic way this speed of improvement of Linux is a strength for Windows.
Yep, this has often been debating in the community, but there is no real political will to change things. It would mean largely standardising the distributions and what they ship: right now there is no political institution that can really bring the distros into line except the LSB and they are not really up to the task for desktop systems.
Stuff like autopackage was designed to solve part of the problem but the real issue is that too many people perceive distros being in firm control as a good thing.
thanks -mike
Stuff like autopackage was designed to solve part of the problem but the real issue is that too many people perceive distros being in firm control as a good thing.
I always felt that the singularity of Windows was it's real strength. I have had to work hard on my perceptions to try and see Linux's variations as a strength. Now I see this in the light of freedom; freedom to code, freedom to try something; freedom to build or fork other code. So, that is a strength, but somehow it's like a racing-boat with a team of rowers all using their own paddles and rowing to their own beat. Some of them are sitting sideways, some backwards! While the Linux team scuds along almost randomly, the highly paid b0rg shoot past.
So, I suppose I should just be glad that so much amazing software actually works at all! I should be, and I am, grateful for it all - Inkscape included.
I will say that Windows is great software. It has incredible power and flexibility (according to one's pocket!) and has very fine software to run on it. Whoa - slow down, don't lynch me yet! :)
The only thing I wish is that Windows had the *balls* to stand up and actually compete. Not like it does now, fixing the fight, like the only boxer in the ring! It's like a "cargo cult" boxing match: it has all the form of a system that allows competition, but none of the substance. It has coaches that talk the talk and it has lights and a stadium and a crowd. It has all the hype and the movement and all the fans. It has the world at it's knees and it charges for every ticket. Then the bell goes "Ding! Ding! Ding!" and the announcer yells, "The fight is on!" and only one boxer comes out and struts about...
The crowd goes wild. Why? There's no real fight.
Bah! Don't give up on Linux Inkscape, even if the distros pull in mad directions!
D.
On 8/8/06, Donn <donn.ingle@...400...> wrote:
I always felt that the singularity of Windows was it's real strength. I have had to work hard on my perceptions to try and see Linux's variations as a strength.
Well that's a pretty common response :) The truth is IMHO a bit more subtle, though this is rather offtopic hopefully the Inkscape devs will indulge us!
There's a lot of value in being 100% compatible, but comparitively little value in being 95% compatible. DJB said this, not me, but it's true so I'll repeat it anyway. Right now Linux distros are all 95% compatible, which causes things to explode messily every so often, and instead of fixing it people just write it off by saying "well that's the price of freedom".
Freedom is powerful but should be used responsibly. A lot of the differences that autopackage and autoconf (configure scripts) abstract don't give a competitive edge or advantages for the user. They're different simply through being apart. This sort of thing destroys a lot of value of compatibility with no corresponding increase in innovation, diversity or freedom.
So right now the variations aren't a strength, they're a weakness. Differences as strength are things like SUSE patching the old GTK file picker to suck less - preserving the compatibility whilst adding useful end user features. Differences as strength are things like Ubuntu using sudo instead of su (badly needed I say).
Differences as weakness are DocBook XSL being in /usr/share/xml in some distros, /usr/share/docbook in others, and /usr/share/stylesheets/docbook in a few random distros hardly anyone uses. It's things like some distros whacking symvers on libpng and others not. When this happens we lose.
thanks -mike
others not. When this happens we lose.
Heh. Sounds like someone should write a Linux source-code virus. It would break into all repositories of source-code and Linux distros and enforce a common directory structure and standards! The Linux Standards Virus!
I hear what you say. It seems insoluble. Perhaps it's for the best that one or two distros win out and define "Linux".
D.
On 8/8/06, Donn <donn.ingle@...400...> wrote:
I hear what you say. It seems insoluble. Perhaps it's for the best that one or two distros win out and define "Linux".
It *can* be solved through politics and diplomacy. But will it? Who knows. Does Inkscape even have more Linux users than Windows users? I have no idea.
It *can* be solved through politics and diplomacy. But will it? Who knows. Does Inkscape even have more Linux users than Windows users? I have no idea.
Would a poll on the website help? Download stats ... I suppose repos would mask the stats; who knows how many Linux users get Inkscape from their repos? Hmm, a poser.
Would it be ethical to build-in a "call home" poll. Perhaps during first-run the user can yea/nay a vote back to your servers?
D.
On 8/9/06, Donn <donn.ingle@...400...> wrote:
Would it be ethical to build-in a "call home" poll. Perhaps during first-run the user can yea/nay a vote back to your servers?
Quite a few apps gather information on how they're being used and send it home these days, it can be used to get interesting usability related data. You could piggyback off that I suppose. It really should be some generic GTK+ based framework though.
thanks -mike
On Wed, 9 Aug 2006, Donn wrote:
Date: Wed, 9 Aug 2006 01:11:53 +0200 From: Donn <donn.ingle@...400...> To: Mike Hearn <mike@...869...> Cc: inkscape-devel@lists.sourceforge.net Subject: Re: [Inkscape-devel] Fedora Core 3 is no more, Weeee!
It *can* be solved through politics and diplomacy. But will it? Who knows. Does Inkscape even have more Linux users than Windows users? I have no idea.
Would a poll on the website help? Download stats ... I suppose repos would mask the stats; who knows how many Linux users get Inkscape from their repos? Hmm, a poser.
Would it be ethical to build-in a "call home" poll.
Yes and No.
Once it was fully disclosed and users could easily opt out it would then at least be ethical but less useful.
Putting something like this in release builds would be a bad idea because as you (Donn) have seen distributions will keep these packages around for several years and certain users will likely leave them unchanged even longer.
Perhaps during first-run the user can yea/nay a vote back to your servers?
However the nightly builds are supposed to be all about testing and providing feedback so callback tools and frequent reminders to upgrade would not be at all unreasonable. (Again it would need to be made clear up front.)
Here is another thing that might be worth exploring: http://www.cs.wisc.edu/cbi/
I recall the bug isolation project providing special builds of the GNU Imange Manipulation program to help gather more detailed information, perhaps they could be persuaded to do the same for Inkscape.
Sincerely
Alan Horkan
Inkscape http://inkscape.org Abiword http://www.abisource.com
Alan Horkan wrote:
On Wed, 9 Aug 2006, Donn wrote:
Would it be ethical to build-in a "call home" poll.
Yes and No.
For me I think it would only be ok to do this if it was a voluntary opt in process.
Currently the phone home ability of software has really bad press as a result of Sony's rootkit and Microsoft's WGA with upgrade or we shut you down customer support proposals. TBH I think open source not needing this kind of thing is a huge plus and makes people feel more in control of their own systems.
Being able to choose to contribute feedback is different from a piece of software which does things on behalf of people other than the person who installed it.
If youre just looking for which os people are running - can this be determined by what they are downloading ?
There is an ubuntu website where people are volutarily posting that they are users. This might be a better approach? http://ubuntucounter.geekosophical.net/
Janet
There is only one reason that the situation is so good for running bleeding edge inkscape on windows. Somebody wanted it, or thought it was needed. So they sat down and did it. 1000s of people benefit from this work. It has dramaticly lowered the level of expertise needed to build inkscape on windows. Bob and whoever else took part in this work, YOU ROCK! Thanks.
Linux is a different world. In general users are more tech savvy (because historicly they have had to be). It is pretty normal for a user to compile software. Most often the dependancies are installed or in package repositories, so things don't ship with it.
Worst case scenario someone could make an Inkscape live CD. But 6 months really isn't too long to wait for a distro supported package for someone who doesn't want to compile it.
I guess that is the reason distros don't have packages newer than 6 months. They have to support the software they ship. If they were to constantly update to the newest packages, their distros would be to unstable to support. (unstable meaning changes fast, not crashes often.)
Aaron Spike
Linux is a different world. In general users are more tech savvy
Not for much longer. I'm a case in point! Fear me. :)
Worst case scenario someone could make an Inkscape live CD. But 6 months really isn't too long to wait for a distro supported package for someone who doesn't want to compile it.
Funny thing is that 6 months is almost the time it takes to outmode any given distro and force you to upgrade or face extinction by repo package starvation. It's a kinda catch 22.
I know, it seems I want it all. Stable distro, latest packages, convenient binaries and not having to pay a cent. I do get the picture. I moved onto a newer distro (it took some doing eh?).
I guess that is the reason distros don't have packages newer than 6 months. They have to support the software they ship. If they were to constantly update to the newest packages, their distros would be to unstable to support. (unstable meaning changes fast, not crashes often.)
<not-serious-mode>You can quit the army if you declare that you are insane. But if you do so, they'll know that you are being very clever and quite shrewd in the attempt to stay alive and so they will not believe you and you will be forced to remain in the army.</not-serious-mode>
;) D.
participants (6)
-
Aaron Spike -
Alan Horkan -
Donn -
Janet Hawtin -
Mike Hearn -
Nicu Buculei (OCAL)