| Summary: | gbonds -- new ebuild | ||
|---|---|---|---|
| Product: | Gentoo Linux | Reporter: | Paul Thompson <set.mailinglist> |
| Component: | New packages | Assignee: | Stewart (RETIRED) <blkdeath> |
| Status: | RESOLVED WONTFIX | ||
| Severity: | enhancement | CC: | blkdeath, brad, gnome, rak088+gentoobugzilla |
| Priority: | High | ||
| Version: | 1.2 | ||
| Hardware: | x86 | ||
| OS: | Linux | ||
| URL: | http://snaught.com/gbonds/ | ||
| Whiteboard: | |||
| Package list: | Runtime testing required: | --- | |
| Attachments: |
ebuild app-office/gbonds/gbonds-0.7.4.ebuild
updated ebuild tarball of new ebuild and patch |
||
|
Description
Paul Thompson
2002-08-20 22:36:45 UTC
Created attachment 3250 [details]
ebuild app-office/gbonds/gbonds-0.7.4.ebuild
Created attachment 11689 [details]
updated ebuild
Bumped to next stable version. Added IUSE. (since when the first ebuild
was submitted, it didnt exist:) Ive been running this for the better
part of a year now.
we've been looking at this and wondering how many people will actually use it. i myself can't really vouch for it because i don't have any US bonds ;). i wonder how many people will actually use this to keep track of their US bonds. the latest version is 0.7.5 and there's also a gnome2 version at 1.90.0 right now. Well, as to use, who can say; if its not in portage, then probably no gentoo user:) Its basicly a clone of the proprietary software offered by the treasury dept., and uses their data to calculate value. I have to admit to becoming discouraged as an ebuild submiter in recent times. You are basicly saying 'we' (the insiders) dont use or understand this package, so we cant certify it, and are suspicious of its value. Ah well, its valuable to me, and its in my ever growing local portage tree of ebuilds; the most recent of which I dont bother to submit, because it appears the pipeline is too constricted. (I have a newer ebuild for this guy, with a patch I wrote to fix something, but ...) Sorry, if this turned into a whinge, but I have been turned away from an enthusiastic ebuild submiter to one who just writes them for myself and submits bugs for the most part. Thanks for your efforts. paul, i understand your discouragement with the submission system. i mean i would feel the same way if i had a bug sitting in bugzilla for about 9 months. unfortuantely, gnome@gentoo.org is rather stretched at the moment, and i've only just started going through and commenting on older bugs that are still sitting in our queue. i've got both gtk1/gtk2 versions in my overlay for testing right now. but i'm not sure when or if i'm going to add it to the tree. although i am encouraged that they seem to be in active development unlike some other older submissions in out queue. It is not that we dont value your ebuild submissions, but see in the greater picture of the ever expanding ebuild tree and the amount of maintenance that comes with it. Is it really so weird we are picky about ebuilds we want to support actively ? I checked with a few other major distros and none of them packages gbonds. These are the sort of packages great for your local overlay, but just don't deserve the amount of work it takes to have and hold them in the main tree imho. Hey, I understand how things are working, which is why I decided to punt. It would be nice, if I could have responsibility for an ebuild I submit-- mark it permanately unstable if you can't ratify it, but thats not how things are structured. Im not bitter-- but nearly all of the ebuilds I have submited since things got constipated have been ignored or rejected, not because someone said; this ebuild is poorly written, or doesnt work for me, but essentially because I am not a developer, and thus able to take responsibility for it, or argue credibly for some aspect of it. So, Ill stick to bugs, and keep my ebuilds to myself for a while at least, and keep an eye out for other outlets; Ive already contributed ebuilds to external sources, that later entered the developemnt circle. And I want to make it clear that I am thankful; its just that the way things are, based upon my experience, there is little motivation for me to submit an original ebuild. Or even a signifigant redesign of an existing one, I am sorry to have gone on so long, but I like gentoo... despite ebuilds being simple to make, mine seem too esoteric:) It's not that you are not a developer or can't take 'ownership' over it here, my main reason still is that it will not be used enough. We market for the masses in the end, there are a zillion small useful apps out there that get used by only a few people. imho we shouldnt be adding stuff from that category, it makes the tree unreasonable big for no good reason. thanks for your understanding. And once again sorry that it took so long to get around to this. And may i suggest you to give the ebuild to the gbonds ppl, so they can add it to their homepage. One minor niggle with your ebuild; it should be marked as ~x86 initially. If you'd prefer, I can make the change myself for inclusion in the tree. I'll even merge Gnome to verify it merges and runs correctly, but I'll have to rely on you to take maintainership with regard to functionality, bugs, and new versions. i'm reassigning this to you because you reopened it. the gnome team won't have the resources to maintain this one. you may want to look at the gnome2 version on their webpage. if you use it, look at possibly using the gnome2 eclass. also get paul to assign the copyright to gentoo (see copyright header), remove redundant deps and do the regular qa checks on it. Hi I will not assign copyright to Gentoo Inc. The 'redundant' dependancies are valid; eg. zlib is no doubt satisfied by several of the earlier deps, but it _is_ a dep. None of the deps are unrequired; this wont run or build without all of them. A script generates this. (and the redundancy is much preferable to the unrequired, or missing requirements I commonly find and report in ebuilds) I dont run gnome, or gnome2, so I cannot vouch for them. I do run gbonds, and it seems to average about an hour or two a year in effort to maintain my personal ebuild. finaly, I have a patch to reformat the entry and cell spacing based upon font size to avoid their eclipse. I havent pushed it upstream yet, but should I include it in my new ebuild submission? Paul on the issue of ~x86 and maintainability, this is not problematic. Created attachment 14185 [details]
tarball of new ebuild and patch
Here is an updated ebuild; changed to ~x86, the patch I am using
to better size entries is included, but the application of the
patch is commented out. It passed lintool, and built. Ive been
using it for some time without problems. If there remain issues
I havent addressed let me know.
Paul
whoohoo, the system works! Fuck you for luring me back, you evil bastard. just a note (i dont want to blow this back up) but dependencies are considered redundant if they are specified in the profile (take a look in /usr/portage/profiles/default/packages to get an idea). the packages specified in that file are considered part of the most baseline install. of your dependency list the following are defined in the default profile- zlib ncurses glibc gettext db in addition i understand you say that the dependencies are all required for gbonds to run, that's fair enough, but portage maintains this sort of sanity so you dont have to. for example, gnome-vfs-1 requires gnome-libs-1, so you specifying it seperately is where the redundancy comes in. and lastly, please attach as text/plain, rather than zips/tars etc, makes it easier for the devs. Mike; The depenancy issues you mention, and the attachment policy would all be interesting things I would like to discuss, ordinarily. However, all things considered, I have clearly given up on this avenue of contribution to gentoo, and if the bridge isnt burnt, then the match I tossed has impotently gone out. Could someone please set this to 'will not fix' or something, or at least take me off the CC? Sadly, it doesn't look like we'll see gbonds in the tree just yet. |
