This message is not meant as a personal attack. The intention is to offer criticism of the way a vague question and its apparently non-specific replies, have produced less than satisfying 'results' - for everyone.
A broad question can be good. I ask them too(!) 'Good', in the sense that its open-ness allows for a wide-range of replies. At the same time, the 'problem' with such a width, is that many options proffered will not suit the specific case. You possess sufficient understanding of 'the problem' to judge that point. Based (only) upon what was told in the posted-question, does anyone else? Thus, can such replies be considered "errors" or "mistakes", particularly when they are volunteered - and doubly-so when they do answer some version of an understanding, of some interpretation, of the vague wording, of the original question? Don't all problem-solving approaches, such as brain-storming, reject more ideas than they 'keep' - by definition? The difficulty with open-ended questions lacking 'anchors', such as 'which database', is that respondents must make assumptions, eg my earlier response. Obviously, it was useless to you. It was rendered a complete waste-of-time (both yours and mine), by the single word "PostgreSQL"! When could I realise that? When did you? Arguing later, about the appropriateness (or otherwise) of such assumptions, does not lead to a solution. Similarly, the adding of previously-unmentioned, but material constraints. How do they affect the respondent? Positively or negatively? What does the manner in which these are expressed do for your reputation? Note also the final exclamation mark of my first response, indicating the expectation that the provided list be heavily-subject to *your* interpretations of 'open' and 'ad-hoc' - and my ability to only make vague assumptions of what you really want... (it was NOT a question-mark!) All access to the DB (in this application user-access) should be controlled by a coordination of VIEWs and ACLs - and thus read-only to such an application. @Alan offered a serious warning about "risk". Meantime, I had presumed such as unnecessary. Which view was correct? Which unnecessary or unhelpful? What in the original question proves such? A competent DBA would be horrified by the assumption that a user won't ever access the DB merely because (s)he doesn't "knows SQL", ie "security by obscurity"! Further, the very lack of access and security controls is a good reason not to use SQLite; but the reply to such a suggestion might say that SQLite/single-user is only a reduced-environment dev/test-tool and thus irrelevant to the end-user environment. Both 'sides' of which are based upon yet another assumption! (NB not a question requesting, or requiring, a written answer!)) Similarly, you defined F/LOSS for me, even though I had indicated recognition of the $free aspect of the spec: "costs ... to include". As you probably realise, many others define the term differently, hence "for varying definitions" - allowing that some will accept Oracle on the grounds of offering a $free advantage, but also that the corporation certainly does not follow "open source" principles required/desired by many others. You have presumed to explain "ad hoc" to someone who trains the full-range of users/professionals in the use and management of Databases - and thus understands it (also) as a term which has a single 'dictionary definition' and yet has come to mean slightly different things to different people. Similarly, you appear to question warnings from a(nother) person who has sufficient experience to have quite literally 'forgotten more about DBs than you (appear to) have learned'. Were such statements made with assumptions about the respondent? How could you know about our knowledge/experience - one way or the other? How could anyone else, about you? Please notice how the first use of the word "spreadsheet", which can (now) be assumed to be material in your thinking, does not occur until the *fourth* exchange. Hence @Peter's response moving the conversation in a quite different direction! (and hopefully one that is more helpful to you) Is it possible that you have criticised responses, rather than allowing for the fact that your respondents have necessarily made assumptions? Would it be more reasonable to assess that any 'fault' is in the assumption, and that such assumption is directly related to the wide-open expression of the question? Have your replies appreciated the freely-given advice and motivation behind them, or have they (perhaps) had the effect of de-motivating the very people who seek to be helpful (to you)? Please note: The questions (above) are Socratic and rhetorical. No reply is requested or required. They ask you to think about maintaining constructive relationships. You certainly do not need to explain yourself to us (nor us to you). Per the opening comment, the idea behind this message is that we become better at helping each other... -- -- Regards, =dn -- https://mail.python.org/mailman/listinfo/python-list
