• Allie C (unregistered)

    That, already, is the real WTF.

    Speaking as someone who was once (very early in my career) an APEX "certified professional", this is 100% true. Although, fair being fair, I'm willing to attribute most of it to the fact that this platform was the first place I ever did any sort of application development, did not have any senior devs to point out my bad choices. It was a glorious mess of PL/SQL stored procs, hacked-in AJAX and jQuery, and don't forget the Python scripts running on a one-off RHEL server!

  • (nodebb)

    Anything you cannot test should never ever find a way into production. The world would actually be a way nicer place without all those demons from the past, may it be overloaded Excel sheets or Access Databases, funny up-sell features like APEX or any other of those "cut the developer/tester out" solutions. But eh, I'm pretty sure there's already something new on the horizon to provide some artificial way to screw things up in the least intelligent way possible...

  • LZ79LRU (unregistered)

    I SHALL NOT have you besmirch the glorious name of Excel. For it is a most useful tool and a most joyous of IDEs for everything from notes to calculations to game development. Yes, it can be misused horribly but the same can be said for any tool. Just look at what you can do in C++.

    But it is not the tool to blame for these things. For the axe cares not if it chops through logs to warm man or through the flesh of his neck. It is the hatchetman and his captains that are to blame for any misdeeds.

    And in the hands of someone capable Excel is the most marvelous of tools for a huge variety of tasks.

  • (nodebb) in reply to MaxiTB

    Anything you cannot test should never ever find a way into production.

    Sounds like a mature position. My company prefers to choose products that make testing impossible, so they can simply blame the lack of testing on the vendor.

  • Steve (unregistered)

    Okay, I'll bite... How does a WYSISYG designer differ from a WYSIWYG one?

  • Duke of New York (unregistered)

    This is what I see all the time from novice developers, and therefore would certainly expect it from "no-code" developers. The thought of a loop that doesn't execute multiple times throws them into chaos and confusion. Why, it's as foolish as slicing bread with a chainsaw!

  • Fizzlecist (unregistered)

    Maybe it's because I see this sort of thing a lot that it doesn't seem that big a WTF to me. It's been too long since I did any PL/SQL but in DB2 it could pretty quickly & simply reduced to a simple cursor query & loop

  • (nodebb) in reply to Steve

    WYSISYG = What You See Is Sorta (what) You Get?

  • xorium (unregistered) in reply to Steve
    Comment held for moderation.
  • (nodebb)

    Someone requested that it support multiple rows, so boom: we add a branch. This solution shows a radical lack of understanding regarding loops though, since a loop that only executes one iteration is still a loop.

    Could also be based in a fundamental mistrust in the code base.

    With our code base we seriously lack tests, because the code base has grown out of a relatively small simulation software into a behemoth, that suffers from poor choices made at the tinkering stage, but now can't risk changing the results of preexisting models.

    Combine that with excessive use of global variables - going back to times when Fortran didn't have better ways of sharing large data structures across program components than common blocks, or more likely, to developers who gained their earlier experience with such versions of Fortran - and often it is an unacceptable risk to change and generalize a code path over adding an if/else.

    It is awful, but until the current push for more testing is finished well, I can perfectly see me running into situations where the technical debt may force me to pile onto it in this manner.

    Not sure if that's a reasonable explanation here though.

  • kolik (unregistered) in reply to MaxiTB

    "Anything you cannot test should never ever find a way into production" - tell that to God, who for all we know pushed string theory to production

  • (nodebb)

    That, already, is the real WTF.

    I think I have to disagree with the article and Allie C. here. To me the real WTF is the conflation of these two things:

    low-code business application space much of Oracle APEX is implemented in Oracle's PL/SQL database language

    what

    Addendum 2023-08-02 04:39: darn it, just imagine a line break in the middle there

Leave a comment on “The Apex of Development”

Log In or post as a guest

Replying to comment #:

Log Out

« Return to Article