# how to start looking at the question archive starting at a known place in the middle?

there's probably already a question on this, but i couldn't find it.

i've been out for a few days, and would like to review what has happened while i was away. i know the number of the last question from just before i was away, and i can get to that. but i don't know how to get to that point and then proceed "upward" in the archive, so that i don't miss anything (except by inattention).

for that matter, it would be useful to learn how to invert the order of the questions, so (on request), i can look at a list with oldest first, and then skip down to a specified point. i'm sure i'm missing something obvious here ... (and, obviously, i'd like to be able to shift back to "newest first" once i get caught up.)

• I'm quite sure that there is no such simple way. I do it like this: I start at the page 2, go to the known point (known by the date of my last access to the site), and then I return to the most recent ones. Going in this direction has this good point: If a new question is posted, the list is shifted by 1 and you meet one question twice. If you start the other direction, you miss one question. – yo' Feb 27 '13 at 22:00
• @tohecz -- i've used that tactic before, but if it takes too long to back up, by the time you get to to top of a page, and are ready to shift back a page, there are often enough additions that if you just back up, some messages are skipped. one has to remember to refresh the current page to avoid that, and even then it's not a sure thing. but better than nothing, of course. thanks. unless someone comes up with a more surefire solution, if you enter that as an answer, i'll accept it. – barbara beeton Feb 28 '13 at 13:09
• somewhat similar to Speravir RSS approach, i use feedly with TeX.SX rss feed. Feedly is online news aggregator which works with google reader and on mobile OS as well. I can get combined TeX.SX tag feed also and even add extra tags. One need to install feedly on each browser first. – texenthusiast Mar 1 '13 at 4:55

Recently, the network implemented a date range specification as part of the "Ninja Search Options". From the linked post, here's an extract on the implementation:

This is now implemented, here are some examples:

• created:2012 (created between 2012-01-01 and 2012-12-31)
• created:2012-04 (created between 2012-04-01 and 2012-04-30)
• created:2012-04-05 (created on 2012-04-05, 24 hr UTC range)

The above date formats can be applied to a range as well with .., like this:

• created:2012..2012 (created between 2012-01-01 and 2012-12-31)
• created:2012-03..2012-04 (created between 2012-03-01 and 2012-04-30)
• created:2012-01-05..2012-01-10 (you get the idea)

Those are for when a question is created. For the last time it was active (same as the active tab) you can use lastactive: the same way as created: above.

So, you could find the specific question you remember seeing last and visit it using the link:

https://tex.stackexchange.com/q/<number>


Then, identify the date of that question and perform a search based on that creation date. This will list posts in newest-to-oldest form (technically not what you're after). Here's how to search for questions tagged amsmath between Jan 2, 2013 and Mar 1, 2013:

is:question [amsmath] created:2013-01-02..2013-03-01

• marvy! this is just the ticket! (i thought i had missed over 700 questions, but the search scooped up 1165 results; some overlap, because i was here for part of the last day pre-outage, but even more than i guessed. well, that's my weekend gone.) thanks -- you get the checkmark. – barbara beeton Mar 1 '13 at 21:19

You could try out Stack2RSS or another solution (see Stack Apps), that make use of the Stack Exchange API.

Notes:

• The date must be given in Unix time. For Windows users an online converter should be useful: https://duckduckgo.com/?q=convert+unix+timestamp+online. I used here the 20th February 2013, 0:00:00. (There is also a todate.)

• The maximum pagesize is the pagesize=100, as given above, i.e. 100 questions in this query (the default is 30); page=1 could be omitted, but not higher page counts, of course.

• One could also write tagged=amsbook;amsmath, but unfortunately this will only find questions tagged with both tags, not the sum of all questions tagged with either of them. Leave out the &tagged=…-part for getting all questions.

• If you want to get only unanswered questions, what in StackExchange means questions with no or no upvoted answer, add &isanswered=false to the end of the address above. Does not work together with tagged!

Regarding combining multiple tags: In RSS feed for New questions? Anna Lear points to Tags AND Tags OR Tags by Jeff Atwood (this is not RSS related).

• interesting suggestion, but unless i'm uninformed (a distinct possibility) that would only send things to my mailbox, which is already overstuffed. and since i prefer to scan the list for everything (not just ams-tagged items), it doesn't really cut down on the load, although it does subdivide it into identifiable chunks. (by my count, i missed about 700 messages while i was out, so maybe identifiable chunks would be helpful; if i do manage to attend tug 2013, i expect to be away for 3 weeks, no net contact. a lot to catch up then!) – barbara beeton Feb 28 '13 at 21:54
• The given address can be put into a feed reader or even read in a browser, at least in Firefox, Opera and IE 10. The only inconvenience would be the manual change of page count and date. Well, and perhaps that for every tag an own tab must be opened. But on Stack Apps there are also solutions based on Python or PHP or … – Speravir Feb 28 '13 at 23:14