On Friday 30 May 2003 22:30, Stanislav Malyshev wrote: > DA>> Well, that's hardly the fault of the lyx developers - you can't > DA>> expect them to maintain packages for different versions of the > DA>> same distro. This is the distro's job. Althouh it'd be nice to > > Well, if we talk the "I don't owe you nothing" way, definitely yes.
No, we're talking about in whos department a task is. Or who is best suited to carry it out. Maintaining packages that are made for a specific distro in a specific way is by definition the work of the distro's developers. As long as the distro wants to keep a centralized package repository, anyway, and AFAIK redhat does. Making special packages (of important apps!) that work with old versions of a distro is definitely the work of the distro's devs, since they have the most knowledge and resources to go about it. > But if > we talk about getting a tool that is usable as Office replacement, I don't > see how maintaining package for just one (or handful of) pet distro can > help that. I can understand that if LyX developers don't care - they just > are working on their project and if someone has problem with that he can > go and make his own LyX better. That's fine by me - but this is certainly > not a replacement for Office. To be replacement for Office one has to > obtain a different mindset - thinking of little users first. That's much > less fun, so I have no any blame on LyX guys. I just conclude LyX is not > fit for a thing you say it is fit. That's certainly not their problem - > that's my problem. And I am sorry that I have this problem - still > after years I have been told that I won't have it very soon now... Well, I was only talking about the capabilities of lyx itself, not the availability of binary packages. Why don't you request redhat to maintain such packages as you need, since you're a user/customer of theirs? > > DA>> On a tangent, this whole idea seems wrong to me. Many versions > DA>> of the same distro that require separate packages to be > DA>> maintained for each, and upgrading to a new version of the > DA>> distro is a big bother (and might cost you access to packages > > Well, they could just have generic Linux binary. After all, all Linuxes > running now are compatible with plain old glibc2. Why Mozilla can do this > and LyX can not? I agree with you that a generic binary would be useful seeing that major distros like redhat don't offer lyx packages. Why not ask the lyx people? Compiling a wholly static binary shouldn't take that much effort or time... > Or they could go and make their code compiling in older gcc's. That's not > that hard - believe me or not, people can do even greater projects without > two-level templates. But I understad I'm preaching to the choir, sorry for > that. Well, there's nothing I can do about that wrt lyx... > > DA>> range of releases, or both. If they don't, that's a major > DA>> obstacle for Joe User. > > That's what I am telling you. So far developers didn't go too far towards > little Joe. I'm hearing a lot of talk for last 5 years about this, but > the frustrating thing is that talk does not turn into deeds. Or, more > precisely, it does, but the conversion rate is still way below what I > (and, more important, Joe - because I can, with certain amount of > frustration, do without it, he can not) would like. Well, there's been a lot of progress in the last 5 years. Lyx was at 0.13 five years ago. Hebrew support in latex was rather worse than today from what I gather, and we didn't have things like qt/kde3 or gtk/gnome2 with their Hebrew support either. Five years ago, this would've been a very good list of things that had to be done before linux could enter the Hebrew desktop. In fact, almost a dream list: - A desktop environment with a full suite of apps that all have Hebrew interfaces and full Hebrew support as good or better than windows'. good, popular API(s) that let people write x11 apps with good Hebrew support without knowing Hebrew. - Read/write support for msoffice documents, including Hebrew support, and a full-featured msoffice lookalike replacement. (koffice is getting openoffice's msoffice filters too, so the two won't have to be always tied together) - A modern word/doc processor replacement for the ugly msword model, with Hebrew support, with professional dtp features and, again, Hebrew support. - Hebrew spellchecker and such (WIP but progressing nicely). - Support (financial & publicity) from the government (for alternatives to msoffice, if not for linux as such). Government plans to use open source in one way or another. As for making distros friendlier, well, look at Gentoo as an example since it's a fairly new/young distro. It solves the upgrade path problem very nicely. You never have to reinstall your whole system, or in fact take it offline for upgrading. Things 'just work' when installing or upgrading packages, for Joe User as well. Also, packages have the necessary dependencies listed, and you can install any package on any gentoo system noa matter how old if the minimal deps are satisfied. Plus, packages can be added to the central repository more easily since it's a community project; if gentoo was missing a lyx package and you wanted to add one, you could become a gentoo dev to maintain it. I don't know what redhat offers in that regard, maybe they've gotten better since I looked at binary distros a couple of years ago. -- Dan Armak Matan, Israel Public GPG key: http://cvs.gentoo.org/~danarmak/danarmak-gpg-public.key
pgp00000.pgp
Description: signature
