Reply to: Death March projects

[sadkingbilly]

Would you classify it as a cutting edge thing? Hard to find people to do it, because there are not enough people skilled up in it yet?

Or is it an older technology? Hard to find people to do it, because the number of people doing it is thinning out over time?

AI take:
'A shortage of mainframe engineers is primarily caused by the large number of experienced professionals nearing retirement and the declining number of new professionals entering the field. This skills gap is worsened by the misconception that mainframes are outdated, leading to a lack of training in universities and a preference for newer technologies like cloud computing.
Aging workforce and retiring talent

• A significant portion of the current mainframe workforce is aging, with many experienced professionals planning to retire in the next five to 10 years.
• When these experts leave, they take with them a lifetime of knowledge and experience that is difficult to replace quickly.

Lack of new talent and training

• Many younger professionals are not considering mainframe careers due to the perception that it is an outdated technology.
• Newer technologies like cloud computing are getting more attention, and university curricula have fewer mainframe-specific courses.
• The result is a shortage of new talent trained in the necessary skills and languages, like COBOL, that underpin mainframe systems.

Misconceptions and outdated image

• There is a persistent myth that mainframes are obsolete and incompatible with modern development practices.
• This negative branding discourages students from learning mainframe skills, even though the technology is still critical to many large organizations.

Continued reliance on mainframes

• Despite the focus on cloud, most Fortune 500 companies still rely heavily on mainframes for their reliability, security, and efficiency in handling high-volume transactions.
• The continued importance of mainframes for core business operations, especially in hybrid cloud environments, means that the demand for skilled engineers remains high. '

so, something like that. I really CBA.
Means I can work, or not, whenever I like.
Still not in applications though, that's for pond life.

[willendure]

Java's just another 'language'. mainframe(Z/series) is an ecosystem and is constamtly being updated.

Would you classify it as a cutting edge thing? Hard to find people to do it, because there are not enough people skilled up in it yet?

Or is it an older technology? Hard to find people to do it, because the number of people doing it is thinning out over time?

[sadkingbilly]

What I mean is - skillsets in stuff that was new decades ago. Java starting to move into that category even.

Java's just another 'language'. mainframe(Z/series) is an ecosystem and is constamtly being updated.

[willendure]

out of date? and still essential? can't be both pal.
and there's a lot more to mainframes than 'coding'.
but I expect nothing more from code monkeys.

What I mean is - skillsets in stuff that was new decades ago. Java starting to move into that category even.

[sadkingbilly]

Some contractors make a living from having a out of date but niche skill that is still essential - such as mainframe coding.

out of date? and still essential? can't be both pal.
and there's a lot more to mainframes than 'coding'.
but I expect nothing more from code monkeys.

[willendure]

We still need to pay the bill as long as the person who signs the timesheets is aware of the situation; that's all I really care about as long as my mental health is good shape. 3 and a bit weeks to go before an extended Christmas break.

Sometimes that is good enough, but in the main it isn't. As a contractor I will accept a difficult or boring or stressful job and not complain - but I would much rather be doing one of the more interesting and fun ones and I have had a few of those, so I know they exist.

Main reason for finishing this one really, as I am not scared about never finding another contract, and I know that worst case I can find another one just like it!

Some contractors make a living from having a out of date but niche skill that is still essential - such as mainframe coding. Others make a living from having cutting edge skills that are still fresh enough to command a premium. The latter kind of contractor is what I have always been. So I need time to brush up on my skills, consult my crystal ball on what tomorrows market is going to need, and go after it.

Thats what keeps it fun and interesting for me anyway. Proud that I can still navigate a many 100s of Ks of lines of Java and keep a level head and fix stuff, but jeez, its bloody dull. Something more interesting next, if I can get it!

[BlueSharp]

2 weeks to go. In some ways things are in a lot better shape now, we have a spec that covers a decent amount of future work, and things have settled down a bit in terms of time demands that completely short circuited the design process. The development work has largely been succesfull, just followed a very messy process, and lots of rework from testing. In that sense, the work should be a bit easier going forward, but I somehow don't trust this client to ever give me a straightforward ride! The main issue is that this client runs what I call fake agile - they pretend to be agile, but really do waterfall, but things never work out to allow the steps of the waterfall to complete. Often my step has been in negative time, development work starts before the spec is decided and ready for technical design.

Extension is up in 2 weeks, and AFAIK they still need somebody, but I need a rest.

Main skills are AWS design, and ability to deal with lots of unknowns and lots of patience. There is probably someone out there that might even enjoy this kind of work and be able to do it much better than me. Around the 500/day mark. When I tell the agency I am not interested in any extension, they will ask if I know anyone that might be interested, and if that is you, drop me a PM and I can tell you more and forward the job description and agency details and so on.

Death march - but its a job, init? In this market thats still worth something!

We still need to pay the bill as long as the person who signs the timesheets is aware of the situation; that's all I really care about as long as my mental health is good shape. 3 and a bit weeks to go before an extended Christmas break.

[willendure]

Not a code monkey, but so-called "fragile" is definitely a thing - see it everywhere in orgs that formerly (read: still) used waterfall and, at some point, thought it might be a good idea to improve things, but couldn't be fecked to understand what or why.

Yeah, if they just stuck with waterfall and committed to doing that well things would mostly be fine. Or go agile, but get rid of the waterfall part completely - involve the BAs in backlog refinement, review and learn and adapt and so on. Can't see that working well with an off shore team though, hard enough to have a discussion about something simple like what a field is named, never mind trying to get into the higher level discussions around refining the OODA loop. But worst of all is to be in between and not do either approach well.

[jamesbrown]

that I call fake agile

Not a code monkey, but so-called "fragile" is definitely a thing - see it everywhere in orgs that formerly (read: still) used waterfall and, at some point, thought it might be a good idea to improve things, but couldn't be fecked to understand what or why.

[CoolCat]

I am in exactly the same position as you! Indian devs all ready to go, they have started in fact, and the spec isn't even half way ready! I've been chasing the a-hole who is writing the spec for months and he kept promising I would have it next monday morning, next monday morning. I think he has actually made a start on it now. Project MUST be delivered end of Sept. Senior stakeholders making a BIG FUSS. Funny thing is, we needed the senior stakehodlers to do some actual work the other day, to get us some info on the clients to set them up with accounts, and they could not be arsed and tried to fob it off on me! Well, it is sunny at the moment, the poor buggers need to WFH and sit in the garden enjoying a lemonade most likely.

Just heard they want to extend my contract as it runs out in about 8 weeks. I switched my mic off I was laughing so hard. Haven't told them where to stick it yet! What a farce.

sounds like network rail, or the MOD, or ...

[willendure]

2 weeks to go. In some ways things are in a lot better shape now, we have a spec that covers a decent amount of future work, and things have settled down a bit in terms of time demands that completely short circuited the design process. The development work has largely been succesfull, just followed a very messy process, and lots of rework from testing. In that sense, the work should be a bit easier going forward, but I somehow don't trust this client to ever give me a straightforward ride! The main issue is that this client runs what I call fake agile - they pretend to be agile, but really do waterfall, but things never work out to allow the steps of the waterfall to complete. Often my step has been in negative time, development work starts before the spec is decided and ready for technical design.

Extension is up in 2 weeks, and AFAIK they still need somebody, but I need a rest.

Main skills are AWS design, and ability to deal with lots of unknowns and lots of patience. There is probably someone out there that might even enjoy this kind of work and be able to do it much better than me. Around the 500/day mark. When I tell the agency I am not interested in any extension, they will ask if I know anyone that might be interested, and if that is you, drop me a PM and I can tell you more and forward the job description and agency details and so on.

Death march - but its a job, init? In this market thats still worth something!

[krytonsheep]

No one lost their job and the super genius responsible left for a better job soon after.

I've seen that time and time again. You'll get someone who is very outspoken, but lacking in any deep technical understanding. Managers listen to them, waste a gazillion pounds. The guy leaves, will put on his CV he did the equivalent of building a SpaceX rocket single handedly, and gets a ridiculously good job because people buy into the bulltulip.

The things I've seen people won't believe. The best analogy I can give is like someone makes a text input box in Javascript for a payment system, writes on their CV they made Stripe, and then gets a job as the governor of the BoE.

[willendure]

It taught me that after a point you have to stop caring and just take the money. I also got loads of cool trips to the US out of it so it was still a win for me.

Indeed. Sometimes you find yourself in these situations and its not your fault and there is nothing you can really do about it, and even trying to sort things out is just going to make your life miserable. Current contract, the hiring manager that interviewed me was hopeless and did not really give any indication of what I was really letting myself in for. Perhaps deliberately. So yeah - gimme the cash and I will see what comes up next! Worst case scenario it will be equally bad, so I've nothing to lose.

[TheDude]

I joined a project for a large US firm. The plan was to replace all of it's huge data silos with a huge data lake.

So far, so good.

They also wanted to add a frontend so that legacy SQL developers could write queries that would be transformed into Apache Spark jobs. The plan being retain business knowledge whilst allowing legacy devs to be productive on a distributed computing platform.

Not do good. Spark has a very different execution model to Oracle, Sybase etc. Simply porting legacy stored procs is a recipe for disaster however this wasn't a fatal flaw - with some training users could write efficient jobs even though such training negated one of primary benefits of the new ssystem.

The absulte killer was the front end required users to manually link the 'functions' they had written. Modify a function and every job that uses it has to be manually re-linked. A one line change to a function could literally require days of manual re-linking.

The thing is that the person that proposed the system was regarded as an untouchable super genius and he had the buy in of several influential managers.

I explained all of these issues in great detail, provided calculations that even automating the re-linking task involved traversing huge dependency graphs and would still be slow. I also proposed a package based dependency solution that would eliminate linking.

Of course I was ignored but I was also absolutely correct and the project was canned six months later at a cost of tens of millions. No one lost their job and the super genius responsible left for a better job soon after.

It taught me that after a point you have to stop caring and just take the money. I also got loads of cool trips to the US out of it so it was still a win for me.

[BlueSharp]

The penny is dropping that I was not talking out of my rear (for a change).

The business processes I warned them about being not fit for purpose are turning out to be, well not fit for purpose. Either way the team is delivering again, even if the business process is not working. But the work at risk was signed off, so arse is covered.

But in the great contractor scheme of things thats not my problem.

41 days to go.