• Hal (unregistered)

    So the WTF here is that in -> 2006 <- a "legacy" (it was already ASP and not .NET therefore probable had been around for a while) project did not have appropriate version control, branching, and CI pipe line?

    I mean this would be laughable mismanagement if it was 2020. However back in '06, this was probably a project that began in the late 90's at which time sure there was know how to do it right and VSC/CVS/SVN were all things but probably the vast vast majority of web projects people tinkering a way and making lots .bkp, .old, .ori files.

  • AM (unregistered) in reply to Hal

    Gotcha.

    So we need to export the database versioned by filename, save everything to an archive versioned by filename, and version control that archive.

  • (nodebb) in reply to Hal

    Even by 2006 standards this is pretty well in the "clueless and unaware" territory.

  • Anonymous (unregistered)

    Back in '06, I worked a summer at a company that did printer leasing. Their primary business was the installation, maintenance, and supplies logistics of photocopiers and photocopier-sized printers, for dozens of private and government offices throughout the area.

    I was tasked to maintain their existing support-ticket and storefront portion of their website. It was originally written by the company owner, in ASP. Sometime in the 00's, however, he had to move it to a new server and didn't have a license for ASP (the way he tells it, I dunno), and so instead of IIS+ASP, he said he used a tool that machine-translated(!) all the ASP code into PHP(3!), so he could set up IIS+PHP+MSSQL on the new server.

    The owner's attempts to maintain that illegible mess of machine-translated code had made the codebase rife with the kinds of devetestuction-isms named in this article. I was worried for a few paragraphs that I had submitted this story in my sleep.

    Week 1 of the job was spent with me getting him to agree (successfully) that it needed to be rewritten from scratch.

  • Bruce W (unregistered) in reply to DocMonster

    Even by 2006 standards this is pretty well in the "clueless and unaware" territory.

    Agreed. As someone who maintained a legacy ASP system in 2006 using VSS, this is crap.

  • MiserableOldGit (unregistered) in reply to Hal

    I agree, I suspect it was written in the late nineties, by a self taught ASP coder. There was lots of that stuff created then, sudden explosion in expectation from small and medium firms for web/ecommerce/intranet solutions and not enough people to do it. Things like ASP made it easy enough for a competent keyboard basher to throw together pages, but their limitations become very obvious if they were not in a team that could help them with not reinventing the wheel, and most weren't.

    I'm wary of saying that's the WTF, although it certainly produces lots of crap like this. TRWTF is that code (and "sauce kontwole") like that was still kicking around in 2006 (and probably still is now). It aint broke so don't fix it is not as safe as it seems to management.

    Having said all that, I was working with someone a couple of years ago who worked this way, it was pitiful to watch.

  • (nodebb)

    I was in IT back in the '90s and I was a TDWTF reader in 2006. This was laughable when it came out. Classic ASP is very amenable to version control as it's no different than PHP from a deployment perspective.

  • (nodebb) in reply to MiserableOldGit

    Having said all that, I was working with someone a couple of years ago who worked this way, it was pitiful to watch.

    I'm fighting to reform two people working like this today. They seem to believe that if they dig themselves a deep enough hole, then the hole itself becomes their excuse for not changing.

  • MiserableOldGit (unregistered) in reply to Jaime
    They seem to believe that if they dig themselves a deep enough hole, then the hole itself becomes their excuse for not changing.

    Strictly speaking, that's sort of true, in my depressing experience.

    Classic ASP is very amenable to version control ....

    Also true, it is just files, but back in the nineties there were plenty of self taught web front end people, maybe they'd had a course or two, but it was very possible for some isolated individual to not even know that there easy solutions to version/source control or a whole wealth of other things we would see as mandatory and obvious by the 2000s.

    Mind you, this one looks like a disorganised mess, if you have to do something "manual" that's not the way to do it!.

  • (nodebb) in reply to MiserableOldGit

    Not knowing fundamentals of your craft is pretty much the definition of "bad at your job". A good developer that grew up without ever communicating with another developer would re-invent source control within a few years. That's how it was invented the first time around.

    I ran into some of the work of these individuals a few years ago. I was asked to see what I could do about an application reliability problem. After asking a few question, it seems that the authors were running a large number of flaky batch jobs on the database server and the server needed to be rebooted every once in a while to fix things, sometimes in the middle of the work day. OK, easy, I ask them to move the batch programs to another server. "We can't do that, there's a million of them and many of them access files on this server. Just finding the source code for all the programs and recompiling would be a herculean task". Yes, paths are hard-coded in compiled programs.

    OK, then let's cluster the database server. I come to find out that database server names are hard-coded in hundreds of programs sprinkled throughout the entire company. The only way they can ever build a new database server is to build it with a temporary name, shut down the old server, and rename the newly built server to the old name. Windows clustering does not take kindly to renaming, so that plan is out the window.

    Never mind, you people cannot be helped.

  • Hasseman (unregistered) in reply to Jaime

    Remember those day of VB1 (1996). Was alone on a project with VB1 (and later VB3). Indeed, one of the first thing I did was Version Control. Did find some sort of Add-In to VB1 (more of intervention with Window messages to implement it). Never worked with version control before this (but keeping directory copies around ...). Had read alot about it ...

  • (nodebb) in reply to Hasseman

    Remember those day of VB1 (1996).

    Apparently, I remember them better than you. In 1996, I was teaching VB 4. Classic VB tolerated source control, but since a large portion of the .frm files were designer-generated, merges could become nasty. The biggest problem with building a deployment pipeline around classic VB was that there was no command line compiler.

  • David Mårtensson (unregistered) in reply to Jaime

    I concur, VB was released late summer 1991, I was an intern as Microsoft Sweden the same spring and got to see the internal beta :)

  • MiserableOldGit (unregistered) in reply to David Mårtensson

    Yes, I started coding VB in the late nineties and we were all arguing about VB5/6 and the booby prize was being asked to convert some shitty program that had been auto-upgraded from VB3 to VB4 to 32 bits. Neither fun nor pretty.

Leave a comment on “Classic WTF: The Developmestuction Environment”

Log In or post as a guest

Replying to comment #:

« Return to Article