Got a software development and deployment story? The call for papers for Continuous Lifecycle 2018 closes in a couple of weeks, meaning you’ve still got the chance to tell your peers, and us, exactly how software development and delivery should work. Continuous Lifecycle aims to cut through the hype, and show what really happens when real tech pros - like you - take ideas …
I'm currently involved with a project that is (more-or-less) going to use Docker/Moby containers to replicate some of our applications/servers for testing purposes. The concept of quickly spinning stuff up is not a bad one, though to be honest I'm still not entirely convoinced that a container pretending to be a server is a good idea compared to a proper virtual machine of some description.
But unfortunately for me (as token sys admin who actually knows how the current applications/servers work) I've lost that battle. It's not the end of the world. What I'm more worried about is that we have a pretty hard project deadline of December to get ready for MiFID2 and we haven't really got anything concrete yet. It hasn't helped that the developers (with their limited system knowledge) have been trying to write the applications and sort out the container stuff at the same time.
I'm still yet to find the person who approved this approach without anyone being able to demonstrate that it was actually possible. No-one had even thought to knock up some skunkworks-type demo to prove the concept.
Oh, and for maximum buzzword, there's also talk of various container managers and CI stuff to jeep it all ticking along. None of which have been tested yet of course. It's alright for the US lot to come up with these bright ideas, but they don't have a rathe rnasty deadline waiting for them.
"You might have been give a blank slate and a huge budget to reimagine software development and deployment from the ground up"
Has this ever hapened to anyone, anywhere, ever? If this is the sort of rubbish they'll be spouting, I'll stay well away and carry on working in the real world.
The article is labelled "Team Register", yet uses "we" to speak of the event. As in language such as
The call for paper closes on October 20, and shortly after that, our esteemed programme committee will dive into the proposals, looking for the meat, and discarding the marketing waffle.
I think we should be told. Whose is the authorial voice asserting ownership of the event? What is the relationship of The Register to this event?
I've sort of taken over a project's web side because my predecessor essentially did a "rage quit" when it became painfully obvious that he was obstructing me, getting nothing done, and wouldn't even help when asked a simple, direct question. His typical answers were in the form of an 'RTFM', with condescension topping. I started getting involved when it was obvious that I was the one that needed to make certain security-related modifications [including adding new tables and columns in the database]. Being a Django system it was WAY more complicated than it needed to be, and all written in Python.
Once 'predecessor' was no longer passive-aggressively obstructing, I put a 'C' program in place of a bunch of inefficient python code and reduce calculation time [on every upload] from >2 minutes down to about 10 seconds [on average]. The old code was causing server timeouts on large data files. And EVERY web developer HATED the idea of using a C program [particularly "that guy"]. There were a boatload of unfixed bugs, unusable admin screens, and "just plain wrong" data being displayed, and every time I test something I find a whole new pile of nitty/irritating bugs that have gone unfixed, like, forever (including one I fixed recently, which was causing an infinite loop on the server, but only when done with the 'rent-a-server', not with the code I have loaded locally - grrr...).
My focus has been on integrating the web side with what's going on with the firmware and phone application. After having modified all of these [for that purpose], I supopse I own them ALL now. No biggee, I need the cash, and now there's plenty for me to do. But the irritations of Django (and the poorly written code) have made it a LOT less fun...
Worst thing about this: my predecessor apparently HATED comments in the code. He objected when I mentioned that I was commenting my changes with my name in the comment, in particular so that _I_ could find where I'd touched things. He complained, said something about it messing up the code [to have COMMENTS???], with respect to cosmetic issues I think, and my response was "NOT NEGOTIABLE".
I'm just glad the bureaucracy didn't go with what HE wanted, though he DID try to backstab me [it took months before "the right person" would let me have the user/pass to the production server so I could update it]. no complaints since then - many bugs fixed, people can get their work done without bugs in the way, etc. and the web site doesn't have "the old logos" on it.
Working on a great project at the moment - with my "newly" coined term of deployment called "fluid".
Basically it involves government setting legislation but neglecting to pass on the details leading you being in the position of having to develop something (in order to meet legislation) but not actually being sure whether what you are doing is right or wrong.... so every second week involves a major change in direction with corresponding panic, fear, chaos etc.
Does make life interesting though.
A 'Continuous Lifecycle' of vacuous talking shops with zero relevance to the real world.
Still if all the buzzword bingo wallahs are herded into a conference hall for a few days to compare the size of each other's containers, it gives the rest of us some time to get some actual development and deployment done without them getting in the way.
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2025
