Closed marianfoo closed 2 years ago
Hi,
sounds like a fair comment but I'm not sure what's the best way to deal with this situation. I guess there are multiple options such as:
Does any of these ideas cover what you mean with dotabap.org? If not, can you please explain in more detail what you had in mind?
Hi,
I don't mean the calculation of the trend data or the ranking. I am actually only interested in the display of the respective information (e.g. recently updated or stars) and the resulting order in "recently updated". If this information is only retrieved once a month, the data that is displayed is up to 4 weeks out of date. As an example: I use dotabap.org to see if there are new repos in the list and which repos have been updated. Since the list is always sorted by "recently updated", I have this information at a glance. The list is also updated once a day.
I would then wish for the following: Displayed information (last update, stars, forks, pulls) is retrieved separately and more often (ideally once a day). Continue to calculate trend data as desired (monthly, weekly). More sorting: e.g. for "recently updated" a filter according to "recently updated". for "new", sort by "creation date". possibly also sorting by stars, forks, pulls I also only recently realised that all three pages (trends,new,updated) are sorted by ranking, which is not so clear.
"New Artifacts" is also a bit misleading, I think "Newly added Artifacts" makes it a bit clearer.
All good points! I could see a daily job that updates the items. One problem here might be the BigQuery API Key that doesn't allow such a frequent access to the table. All other providers should remain within the free quota even when we update the stats daily.
I don't think I'll have time for this in the next few weeks as Devtoberfest and TechEd keep me quite busy. But I'll keep the issue open in case someone wants to tackle this.
Hi @IObert the new GitHub action seems to have worked. But unfortunately the rebuild was not triggered. Should the rebuild be done in the gh-update action or should the new data also be written directly to the docs folder?
I think it makes sense to write directly into docs
at the end of the gh-update action
The commit today looks good.
All good points! I could see a daily job that updates the items. One problem here might be the BigQuery API Key that doesn't allow such a frequent access to the table. All other providers should remain within the free quota even when we update the stats daily.
Does it make sense to expand this action to the other providers? Should PyPi be excluded from this?
I think it makes sense to extend this for npm. I'd exclude PiPy for now was it uses a paid API key.
And yes, it looks great! Many thanks for your great work 🤩!
No worries :) I think with #18 we can close this issue as it is resolved, right?
Yes :)
The page 'Recently updated artefacts' shows all artefacts that have been updated recently. However, since the update process is not triggered often enough, the data is increasingly out of date. Is it possible to start this process more often, or to move it to a separate repo like dotabap.org and thus keep the 'trends.json' and 'allItems.json' up to date?