Database available in other formats
Dear Sirs, First thig first a big THANK YOU for your invaluable work. I was overwhelmed by the complexity of your database, because it requires also external pieces of code to be safely read. I was wondering if does exist the database in different formats that may be larger but more readable and easy usable. An SQL database could be really "for everyone", because can be queried by everything from everywhere by any language. If I may make a humble suggestion, a JD date in floating format that equips any record with the initial and final date of the rule/event could be of great help. With a single query, one could recall all the records that pertain a certain moment in time also with different and overlapping rules. Thanks for any reply and thank for your work that is making me sweat the classical " seven shirts". Stay well Sergio Bonfiglio ITALY
On 4/28/20 8:12 AM, Sergio Bonfiglio wrote:
I was overwhelmed by the complexity of your database, because it requires also external pieces of code to be safely read.
Yes, it's in its own .zi format. Typically I edit it with a text editor so I suppose that counts as an external piece of code. Most people either do that, or process it with the code supplied as part of tzdb, or use one of the already-written tz compilers <https://data.iana.org/time-zones/tz-link.html#compilers>. Perhaps one of those will work for you.
An SQL database could be really "for everyone", because can be queried by everything from everywhere by any language.
MariaDB, MySQL, and Oracle Database use tzdb in some form; perhaps they're already doing something along the lines of what you're asking for.
If I may make a humble suggestion, a JD date in floating format that equips any record with the initial and final date of the rule/event could be of great help.
JD should be relatively easy to generate automatically from the existing format. I doubt whether the source data should have JD directly, though, as JD is inconvenient for humans and anyway the .zi format is now read by so many programs that any changes to it would need to be carefully considered and justified. But JD could be good for database queries, along the lines of your suggestion.
Dear Paul, thanks for the reply. I have found that there could be some versions already available of the TZ database that could suit my needs. In order to make a due reply to your suggestions I'd tell something about your responses and I'd like too to submit you some topics that I need to confront with. The first topic is this: the UNIX date that is supported in many compilers has some limitation in order to the dates preceding Jan 1 1970. The second topic is that some of the compilers you (greatly) list onto your page are not updated at the last version of data, like for instance the Delphi version - right the one I'd like to use - which is - strange to say - includes the data WITHIN the code, that is something really against the most simple rules of Information Technology that is to keep data separated from the code. So for what this is the solution of this problem I think I will go to some other resource, that is another that I haven't found listed into your page (unless I have resoundingly missed it) that is this one: https://timezonedb.com/ These great guys provide a MySQL version of the TZ database that seems also up-to-date (I have to check it out but it is promising) and I think you should nicely include this resource into your list (unless, I repeat, I have missed it and this is just in the list already. In this case I apologize for that). I have an idea about making a software to keep the TZ database updated. I will talk you in the future If I decide to carry on this work. The third issue is what I have with Italian zones listed in the TZ, because as you correctly state in the description notes of the "europe" file, In Italy there are 3 zones, not only ROME but also Palermo (Sicily island complete) and Cagliari (Sardinia island complete), which differ from each other, also if slightly in time. How can I add these two zones to the data ? May I kindly ask you to include these two zones in some the future releases of the database ? Thanks again for your great help and your invaluable work. Sergio Bonfiglio Il giorno gio 30 apr 2020 alle ore 22:39 Paul Eggert <eggert@cs.ucla.edu> ha scritto:
On 4/28/20 8:12 AM, Sergio Bonfiglio wrote:
I was overwhelmed by the complexity of your database, because it requires also external pieces of code to be safely read.
Yes, it's in its own .zi format. Typically I edit it with a text editor so I suppose that counts as an external piece of code. Most people either do that, or process it with the code supplied as part of tzdb, or use one of the already-written tz compilers <https://data.iana.org/time-zones/tz-link.html#compilers>. Perhaps one of those will work for you.
An SQL database could be really "for everyone", because can be queried by everything from everywhere by any language.
MariaDB, MySQL, and Oracle Database use tzdb in some form; perhaps they're already doing something along the lines of what you're asking for.
If I may make a humble suggestion, a JD date in floating format that equips any record with the initial and final date of the rule/event could be of great help.
JD should be relatively easy to generate automatically from the existing format. I doubt whether the source data should have JD directly, though, as JD is inconvenient for humans and anyway the .zi format is now read by so many programs that any changes to it would need to be carefully considered and justified. But JD could be good for database queries, along the lines of your suggestion.
On 5/1/20 8:49 AM, Sergio Bonfiglio wrote:
tz-link.html mentions TimeZoneDB under "Time zone boundaries" but it'd be good to also mention their CSV and SQL downloads. I installed the attached proposed patch to try to do that.
in Italy there are 3 zones, not only ROME but also Palermo (Sicily island complete) and Cagliari (Sardinia island complete), which differ from each other, also if slightly in time. How can I add these two zones to the data ?
Because these zones are identical to Europe/Rome for timestamps after 1970, they're out of scope for tzdb proper <https://data.iana.org/time-zones/theory.html#scope>. However, we do have a file 'backzone' for out-of-scope data (this file is not normally used in tzdb installations). So you can send in a patch for that file if you like, preferably in 'git format-patch' form. See Europe/Belfast for an example of what might appear in 'backzone'.
Thanks, I'll check them out. Sergio Il giorno ven 1 mag 2020 alle ore 20:51 Paul Eggert <eggert@cs.ucla.edu> ha scritto:
On 5/1/20 8:49 AM, Sergio Bonfiglio wrote:
tz-link.html mentions TimeZoneDB under "Time zone boundaries" but it'd be good to also mention their CSV and SQL downloads. I installed the attached proposed patch to try to do that.
in Italy there are 3 zones, not only ROME but also Palermo (Sicily island complete) and Cagliari (Sardinia island complete), which differ from each other, also if slightly in time. How can I add these two zones to the data ?
Because these zones are identical to Europe/Rome for timestamps after 1970, they're out of scope for tzdb proper <https://data.iana.org/time-zones/theory.html#scope>. However, we do have a file 'backzone' for out-of-scope data (this file is not normally used in tzdb installations). So you can send in a patch for that file if you like, preferably in 'git format-patch' form. See Europe/Belfast for an example of what might appear in 'backzone'.
Dear Paul, I don't know what to do with the file 0001-Mention.... etc. that you sent me. Must I apply to a git clone or something ? Would you apply to the official page, so you sent it to me for a sort of approval ? Excuse my ignorance, please. Sergio Il giorno ven 1 mag 2020 alle ore 20:51 Paul Eggert <eggert@cs.ucla.edu> ha scritto:
On 5/1/20 8:49 AM, Sergio Bonfiglio wrote:
tz-link.html mentions TimeZoneDB under "Time zone boundaries" but it'd be good to also mention their CSV and SQL downloads. I installed the attached proposed patch to try to do that.
in Italy there are 3 zones, not only ROME but also Palermo (Sicily island complete) and Cagliari (Sardinia island complete), which differ from each other, also if slightly in time. How can I add these two zones to the data ?
Because these zones are identical to Europe/Rome for timestamps after 1970, they're out of scope for tzdb proper <https://data.iana.org/time-zones/theory.html#scope>. However, we do have a file 'backzone' for out-of-scope data (this file is not normally used in tzdb installations). So you can send in a patch for that file if you like, preferably in 'git format-patch' form. See Europe/Belfast for an example of what might appear in 'backzone'.
On 5/3/20 5:48 AM, Sergio Bonfiglio wrote:
I don't know what to do with the file 0001-Mention.... etc. that you sent me. Must I apply to a git clone or something ?
You could do that, yes. Or you could simply clone the current development version <https://github.com/eggert/tz>, as it has the patch already applied. Eventually the development version should evolve into the next release.
participants (2)
-
Paul Eggert -
Sergio Bonfiglio