4 ms·
Thanks for your suggestion, I always love reading detailed descriptions of what people would like from GitHub. However, now I've got a bit of a question for you
by kneath 16y ago
Thanks for your suggestion, I always love reading detailed descriptions of what people would like from GitHub. However, now I've got a bit of a question for you...
My opinion is that every extra field I add to a page means I've personally failed. It's so easy to correlate databases to HTML that people often go feature-crazy adding fields to solve their problems and you end up with a product like JIRA.
However, as a designer I'm directly opposed to this idea. Every field I add to a web application torments me. I think about it constantly and scheme for hours trying to come up with any alternative to adding another field. I've been known to dream about "simply adding another field" for weeks. I need to make sure that this field is really going to positively impact my users. That it's more important than any other field already on the page.
So, would this inactive/status flag provide a significant benefit over editing the name/description/readme? I guess that's the question I'd like you guys to think about.
- PStamatiou 16y agocould something like this be done automatically? No commits in X months leads to auto-abandoned + fires off email to owner asking them to commit something to unabandon it if they want.
- kneath 16y agoIt's hard to say what active means. For example http://github.com/technoweenie/acts_as_versioned/commits/master http://github.com/technoweenie/acts_as_versioned/commits/mas... hadn't had any commits in more than a year until today, but now it's up-to-date with rails 3.
- gus_massa 16y agoThe (current) last update of TeX is from March 2008. The idea is that it is stable, not abandoned.
- Groxx 16y agoYou sure? I thought work was shifting to ǐTéX (ring!).
- _delirium 16y agoThat's partly because Knuth's TeX is more the core of a piece of software than the software itself; actual distributions of TeX get much more frequent maintenance, and would bitrot without it. For example, the last commit to texlive (the TeX distribution most Linux users use) was 20 minutes ago: http://www.tug.org/svn/texlive/trunk/ http://www.tug.org/svn/texlive/trunk/
- IgorPartola 16y agoAnother aspect of this is whether the project is used by anyone. If a project like wget loses the maintainer, it's a big deal and we should shout on street corners about it until someone steps up. If a project is simply a piece of code that is of no use to anyone, even the original author, then we should at least rank it lowest. So maybe for the former it makes sense to add the extra field so that people can be notified instantly and someone could step up. Then again, IIRC wget did lose its mainrainer recently and someone stepped up very quickly since the project has other communication channels.
- Groxx 16y agoWhile I do like the idea, and it could certainly be nice to have as a user... I've got to vote against it. Abandonware can be pretty easily detected once you look at the page (name/description/readme/last update(!)), this would just put the "abandoned" label a layer higher. I haven't looked through the GitHub API, but is there a way to pull the description? I see "abandoned" or something similar in there more frequently than anywhere else; a browser plugin could fetch that data and display the tag moderately accurately. Or color based on last activity, etc. and that way you don't have to add a field. edit: on second / third thought, maybe that's a possible solution: show the last-updated date when multiple repos are listed. I know I'd find that useful. What defines an "update" is debatable... commits? Issues? Wiki? All of the above? Something else?
- togasystems 16y agoFrom a user's perspective, I find that the date of the last commit as well as the number of open issues are tell tale signs of the the project's state.
- Slackwise 16y ago> So, would this inactive/status flag provide a significant benefit over editing the name/description/readme? You can always cram more data into another field, but now you've lost easy access to valuable metadata--this would be a hack, rather than a solution. Proposing for users to use a hack rather than a clean, semantically different, UI element is not good on your part as a designer. A "Status" field benefits users as repository pages can have a prominent notification of the current status. You can also use it as a flag for searches to avoid having to wade through various abandoned projects before finding one that is being maintained. And for the GitHub developers, it will provide a valuable statistic. Now they can track the progress and usage of repositories without potentially using assumptions like "Hmm, this repo has not be committed to for X days--could be abandoned?" or having to parse through the description/readme for keywords like "abandoned" or "no longer maintained", etc. So yes, I would say that this simple metadata field would significantly benefit the users and developers.