Gitlab Social Sciences: Difference between revisions

Line 94:

Longer answer: GitLab distinguishes read-access within 3 layers, and you can choose for each project individually.

~~TODO~~: ~~Explain 3 layers~~ with ~~correct terms~~.

Each project can be either

* ''private'': you and hand-picked users can read it,

* ''internal'': anyone with a GitLab account can read it

* ''public'': anyone on the Internet can read it

For more information, see https://gitlab.socsci.ru.nl/help/public_access/public_access#visibility-of-projects.

=== What can/should be stored on GitLab? ===

Short answer: Scripts, documentation, theses, figures etc. '''No data please!''' For data, use https://data.donders.ru.nl/.

Longer answer: Git and GitLab are meant primarily for smaller amounts (Megabytes).

Longer answer: Git and GitLab are meant primarily for smaller amounts (think Megabytes, not Gigabytes).

Most features like comparing versions, only make sense with text-based files.

Most features, like comparing versions, only really make sense with text-based files.

Git keeps track of every ~~single~~ change you make in its history.

Git keeps track of every change you make in its history, and never really forgets.

It and never really forgets~~, unless you explicitly tell it to~~.

That means Git repositories only grow larger, and even if you "delete" things, you can go back and un-delete them.

That's a main feature, you ~~have~~ an "unlimited undo".

That's a main feature, you get an "unlimited undo" for all kinds of files.

~~But that feature~~ makes ~~also~~ impractical for data - you can't really delete them, comparing versions of them makes little sense~~, and~~ potentially you later want to ~~invite more people or~~ make the project available to ~~everyone~~ - but ~~then you also would have to share the~~ data ~~since they're in the repository history~~.

Finally, ~~the~~ server ~~simply~~ has limited capacity.

Keeping an unchangeable history makes it impractical for data - you can't really delete them, comparing versions of them makes little sense.

For all these reasons, ~~for data,~~ use https://data.donders.ru.nl/.

And potentially you later want to make the project available to more people - but maybe without sharing all data.

Finally, our server also has limited capacity.

For all these reasons, please use https://data.donders.ru.nl/ for your data.

@@ Line 94: / Line 94: @@
 Longer answer: GitLab distinguishes read-access within 3 layers, and you can choose for each project individually.
-TODO: Explain 3 layers with correct terms.
+Each project can be either
+* ''private'': you and hand-picked users can read it,
+* ''internal'': anyone with a GitLab account can read it
+* ''public'': anyone on the Internet can read it
+For more information, see https://gitlab.socsci.ru.nl/help/public_access/public_access#visibility-of-projects.
 === What can/should be stored on GitLab? ===
 Short answer: Scripts, documentation, theses, figures etc. '''No data please!''' For data, use https://data.donders.ru.nl/.
-Longer answer: Git and GitLab are meant primarily for smaller amounts (Megabytes).
+Longer answer: Git and GitLab are meant primarily for smaller amounts (think Megabytes, not Gigabytes).
-Most features like comparing versions, only make sense with text-based files.
+Most features, like comparing versions, only really make sense with text-based files.
-Git keeps track of every single change you make in its history.
+Git keeps track of every change you make in its history, and never really forgets.
-It and never really forgets, unless you explicitly tell it to.
 That means Git repositories only grow larger, and even if you "delete" things, you can go back and un-delete them.
-That's a main feature, you have an "unlimited undo".
+That's a main feature, you get an "unlimited undo" for all kinds of files.
-But that feature makes also impractical for data - you can't really delete them, comparing versions of them makes little sense, and potentially you later want to invite more people or make the project available to everyone - but then you also would have to share the data since they're in the repository history.
-Finally, the server simply has limited capacity.
+Keeping an unchangeable history makes it impractical for data - you can't really delete them, comparing versions of them makes little sense.
-For all these reasons, for data, use https://data.donders.ru.nl/.
+And potentially you later want to make the project available to more people - but maybe without sharing all data.
+Finally, our server also has limited capacity.
+For all these reasons, please use https://data.donders.ru.nl/ for your data.

Gitlab Social Sciences: Difference between revisions

Navigation menu

Search