OK, let me try to motivate the reasons for this poll. First of all, I'm putting this poll in the dev section because I'm mostly interested in what the devs think about this.
My personal opinion is that we should be cautious and make it not too easy to delete things from the repo.
As far as I understand, the repo was created in the first place to give devs full control, because the way things were done on Craigs apps.openpandora.org was not satisfactory: most devs couldn't maintain their own entries themselves, which caused a lot of frustration. In my opinion it is a good idea to give the devs a lot of control, especially to update and fix their entries themselves instead of having to rely on third parties who may or may not be available 24/7. But perhaps the current repo gives them a little too much control.
In particular, the power to decide to disable or delete your entries in the repo unilaterally seems somewhat dangerous to me. Of course usually this power will not be abused, and there can be valid reasons to delete things, so I'm not against deletion as such.
But consider the following scenario: because of some personal conflict, a dev ragequits the Pandora community, and in his rage, he removes all his contributions from the repo just to be nasty. Or maybe there is a flamewar on the forum about some controversial topic, e.g. some aspect of the design of the Pandora 2, and things get out of hand. The losing side of the argument might decide to pressure the community into accepting their views by removing their contributions from the repo (probably just temporarily), as a kind of "PND boycott" measure.
Maybe I'm being paranoid and those things will never happen. But at the moment nothing prevents this from happening.
The way I see it, uploading a PND to the repo implies that you are releasing something to the public. Of course it depends on the license what exactly you are releasing, but at least the binary executables are at that point "published" (well, for non-commercial software anyway). I don't think this act should be retractable. If you really want to make a "temporary" release (which can be fine, e.g. to get people to betatest something you want to sell later), you shouldn't use the community repo, you should use your own website or use your own way to distribute things.
Consider the way Wikipedia works. As is the case with the repo, anyone can make a contribution, and for example write a new article about some topic. This is a very nice way of doing things, because it makes it really easy to contribute to the community. However, on Wikipedia, you cannot just delete an article you started, whenever you want. You can try to remove or vandalize your own article, but the history is kept and the Wikipedia community can undo that. As soon as you press the commit button, your text will be saved permanently to the records, even if you later change your mind and decide you didn't want to contribute after all. This doesn't mean that articles cannot be deleted from Wikipedia: if some article is considered to be irrelevant, there is a deletion procedure to deal with that. Reasons for deletion are considered and weighed against reasons for not deleting. But this decision is made in a democratic way, not unilaterally by whoever started the article.
So what I'm suggesting is that repo package maintainers should not get the power to decide unilaterally to delete their packages. Instead, there should be a public button (well, for registered users) to report a PND to the mods/admins and nominate them for deletion. Of course the maintainer can press that button himself if needed. If the reasons for deletion are sufficiently uncontroversial (e.g. the upload contained illegal things, or the package is an obsolete duplicate of some other package), the deletion can happen instantly. But if things are not that clear, a public discussion should be started on the forum, and if necessary, a vote.
Many people consider the repo to be their "backup" in case their SD card fails or something like that. I think that's a good idea - I also don't make backups of the GNU/Linux distribution I'm using, because I know this software will not just disappear from the internet. Currently, the repo doesn't offer me the same kind of reliability, because whether or not some PND will still be there next year (or next week for that matter) depends not just on the ability of ED to keep his server running (this I could live with, although a few mirrors wouldn't hurt - but that's another discussion), but also on the mood of the package maintainers.
My personal opinion is that we should be cautious and make it not too easy to delete things from the repo.
As far as I understand, the repo was created in the first place to give devs full control, because the way things were done on Craigs apps.openpandora.org was not satisfactory: most devs couldn't maintain their own entries themselves, which caused a lot of frustration. In my opinion it is a good idea to give the devs a lot of control, especially to update and fix their entries themselves instead of having to rely on third parties who may or may not be available 24/7. But perhaps the current repo gives them a little too much control.
In particular, the power to decide to disable or delete your entries in the repo unilaterally seems somewhat dangerous to me. Of course usually this power will not be abused, and there can be valid reasons to delete things, so I'm not against deletion as such.
But consider the following scenario: because of some personal conflict, a dev ragequits the Pandora community, and in his rage, he removes all his contributions from the repo just to be nasty. Or maybe there is a flamewar on the forum about some controversial topic, e.g. some aspect of the design of the Pandora 2, and things get out of hand. The losing side of the argument might decide to pressure the community into accepting their views by removing their contributions from the repo (probably just temporarily), as a kind of "PND boycott" measure.
Maybe I'm being paranoid and those things will never happen. But at the moment nothing prevents this from happening.
The way I see it, uploading a PND to the repo implies that you are releasing something to the public. Of course it depends on the license what exactly you are releasing, but at least the binary executables are at that point "published" (well, for non-commercial software anyway). I don't think this act should be retractable. If you really want to make a "temporary" release (which can be fine, e.g. to get people to betatest something you want to sell later), you shouldn't use the community repo, you should use your own website or use your own way to distribute things.
Consider the way Wikipedia works. As is the case with the repo, anyone can make a contribution, and for example write a new article about some topic. This is a very nice way of doing things, because it makes it really easy to contribute to the community. However, on Wikipedia, you cannot just delete an article you started, whenever you want. You can try to remove or vandalize your own article, but the history is kept and the Wikipedia community can undo that. As soon as you press the commit button, your text will be saved permanently to the records, even if you later change your mind and decide you didn't want to contribute after all. This doesn't mean that articles cannot be deleted from Wikipedia: if some article is considered to be irrelevant, there is a deletion procedure to deal with that. Reasons for deletion are considered and weighed against reasons for not deleting. But this decision is made in a democratic way, not unilaterally by whoever started the article.
So what I'm suggesting is that repo package maintainers should not get the power to decide unilaterally to delete their packages. Instead, there should be a public button (well, for registered users) to report a PND to the mods/admins and nominate them for deletion. Of course the maintainer can press that button himself if needed. If the reasons for deletion are sufficiently uncontroversial (e.g. the upload contained illegal things, or the package is an obsolete duplicate of some other package), the deletion can happen instantly. But if things are not that clear, a public discussion should be started on the forum, and if necessary, a vote.
Many people consider the repo to be their "backup" in case their SD card fails or something like that. I think that's a good idea - I also don't make backups of the GNU/Linux distribution I'm using, because I know this software will not just disappear from the internet. Currently, the repo doesn't offer me the same kind of reliability, because whether or not some PND will still be there next year (or next week for that matter) depends not just on the ability of ED to keep his server running (this I could live with, although a few mirrors wouldn't hurt - but that's another discussion), but also on the mood of the package maintainers.
Last edited by a moderator:
