~sircmpwn/public-inbox

5 3

[gddo] Backlinks

Details
Message ID
<C8SGRG9J6R7V.2HWDZYTFOPYMQ@nitro>
DKIM signature
pass
Download raw message
I stumbled across something while reading the gddo-server code:
https://git.sr.ht/~sircmpwn/gddo/tree/c4471710cd7c451d14cb969946908fb44b63b65b/item/gddo-server/main.go#L307

Any idea what this is meant to achieve?
Details
Message ID
<3117e835-f46a-499b-b9e7-fd695d580b1c@pixelhero.dev>
In-Reply-To
<C8SGRG9J6R7V.2HWDZYTFOPYMQ@nitro> (view parent)
DKIM signature
pass
Download raw message
I think it is intended to make indexing easier for Google.
Details
Message ID
<C8SGZWQCZAXU.FXKA7TF3BT4J@taiga>
In-Reply-To
<C8SGRG9J6R7V.2HWDZYTFOPYMQ@nitro> (view parent)
DKIM signature
fail
Download raw message
DKIM signature: fail
I'm guessing it's a poor substitute for <a rel="nofollow">, which is
standard practice for websites to signal to crawlers that a link is
user-generated content and not endorsed by the website. Prevents SEO
spamming.
Details
Message ID
<C8SH0AYTXQKO.39R4TCP83T2HJ@taiga>
In-Reply-To
<C8SGRG9J6R7V.2HWDZYTFOPYMQ@nitro> (view parent)
DKIM signature
fail
Download raw message
DKIM signature: fail
On the whole, the robot detection thing is pretty bad. I have it
disabled in production. It would be better to use things like
rel="noopener nofollow" on all user-generated links.
Details
Message ID
<C8SH525BYDEI.3I14OFIOJGVYY@nitro>
In-Reply-To
<C8SH0AYTXQKO.39R4TCP83T2HJ@taiga> (view parent)
DKIM signature
pass
Download raw message
On Mon Jan 25, 2021 at 2:02 PM EST, Drew DeVault wrote:
> On the whole, the robot detection thing is pretty bad. I have it
> disabled in production. It would be better to use things like
> rel="noopener nofollow" on all user-generated links.

It seems like the server does not increment the popularity counter of a
package for robot requests. Would removing the robot detection cause a
problem?
Details
Message ID
<C8SH6OCOHWHC.2KK1H7Z9VODT4@taiga>
In-Reply-To
<C8SH525BYDEI.3I14OFIOJGVYY@nitro> (view parent)
DKIM signature
fail
Download raw message
DKIM signature: fail
Ah, that makes sense. I think we might want to remove the popularity
from the UI and just use it to silently rank results in the background.
We could also use the number of packages which depend on a package as a
measure of its popularity, rather than hits on the web.
Reply to thread Export thread (mbox)