You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
We already have more than 250 ignore rules for version patterns of 20[0-9]{6} and alike (e.g. dates). It seems viable to instead identify projects that DO officially use dates in versions, make a whitelist for them and ignore all other dates globally.
The whitelist maintenance cost involve considerable amount of work at start to identify and review all legal cases, after which the upkeep would be easier (basically, only processing new entries infrequently). This should/could overweight the constant cost of processing reports of fake date versions and user annoyance.
NB:
Total metapackages with 20[0-9]{6}: 12126
Total metapackages with 20[0-9]{6}$: 9717
Total metapackages with '^20[0-9]{6}$: 5819
This is the upper estimate of work required. We can start with the narrower pattern. Actual cost would be smaller, as we only need to review non-unique metapackages and exclude e.g. rolling versions. Also the most of these are texlive-* metapackages. In most conservative case, only couple hundreds of metapackages need to be reviewed.
The text was updated successfully, but these errors were encountered:
We already have more than 250 ignore rules for version patterns of
20[0-9]{6}
and alike (e.g. dates). It seems viable to instead identify projects that DO officially use dates in versions, make a whitelist for them and ignore all other dates globally.The whitelist maintenance cost involve considerable amount of work at start to identify and review all legal cases, after which the upkeep would be easier (basically, only processing new entries infrequently). This should/could overweight the constant cost of processing reports of fake date versions and user annoyance.
NB:
20[0-9]{6}
: 1212620[0-9]{6}$
: 9717'^20[0-9]{6}$
: 5819This is the upper estimate of work required. We can start with the narrower pattern. Actual cost would be smaller, as we only need to review non-unique metapackages and exclude e.g. rolling versions. Also the most of these are
texlive-*
metapackages. In most conservative case, only couple hundreds of metapackages need to be reviewed.The text was updated successfully, but these errors were encountered: