IBM's 18-month company-wide email system migration has been a disaster: sources

And every time a big user moves away from Lotus Notes (and other such dinosaurs, such as Cobol), it gets harder as this knowledge becomes more niche. The people doing it spend the rest of their career at that combany and that's it.

Task was successful because it increased productivity.

Pardon my ignorance; what does RH mean in this context?

ex-IBMer here.

I was at IBM research (as an engineer), and thankfully I can say most of my direct colleagues were great and highly competent

...when it came to actual research. I think we had a good engineering core team too, but whenever we had to interact with "big IBM" (outside research) we mostly bumped into incompetent people, and career(money)-focused managers.

Fun fact, due incompetent management 90%+ of my team left in the past 5 years.

Since then I had to work with other IBMers (both engineers and managers/team leads) on different projects and although some of them I'd consider knowledgeable, none of the engineers were able to create high quality output.

But I guess this is the same in any monstrously huge organizations.

I would have been surprised if IBM could tackle such a complicated engineering problem like this migration.

IBM is huge... It's not really a single company, but like a bunch of independently operating companies that share (some) resources, and of course their name.

We have fairly high hiring standards in our office, but when it comes to interacting with people from other offices, I've worked with people I'd never hire. I work on the IBM Cloud platform, which seems to be one of the most innovative places at IBM currently (compared to GBS/GTS, the latter became Kyndryl recently).

Ex-IBMer here. I used to work for what was then called IBM Software Group in the early 2000s. I worked with some brilliant engineers, some of whom had spent their entire careers at IBM and achieved some amazing things. I worked with some brilliant managers as well. But the place felt rotten to the core.

I don't think there was a problem with the people so much as with the system. IBM, as you might expect for a company of that size and age, felt incredibly bureaucratic. Whatever you wanted to achieve, there was a process that made it hard to achieve it. Whatever good idea you had, there was a committee of people who would dilute and ruin that idea or simply forbid you from pursuing it. Whatever industry standard tool you wanted to use, you couldn't use it because there were IBM tools that you were required to use instead, tools that slowed you down and frustrated you (I'm thinking of stuff with Lotus or Rational in the name). Whenever you needed to make a decision, it felt like no one was empowered to use their initiative to make that decision. Frankly, it felt like doing good work meant having to go under the radar because IBM wasn't set up to allow you to do good work.

I did have good times there and I hear good things about the way that IBM has dragged itself into the 21st century since I left but I regret having spent so many of the best years of my career at that company.

I’m 41, and know all those things, and my father’s company ran for most of my life on AS/400. But they replaced those systems over a decade ago. My whole professional career IBM has been trying to commit corporate suicide in a stumbling awkward way. It’s a testament to how long you can keep spending down brand credibility. I struggle to call them important today. Maybe historically interesting. But not important.

IBM made RHEL?

IBM was the Google or Microsoft of the 80s. Like HP, in realigning the company to survive they destroyed what made them great. I worked with many smart IBM folks… without exception they were laid off or driven out to meet some random metric.

The Red Hat acquisition and spin out of the services garbage might make them more relevant again.

The data about patent filing does not support your claim about innovation.

IBM literally optimizes for the patent filing metric (and others generally do not), and that metric is not at all correlated to "innovation".

Most people would also likely argue that data would not support your claim about robustness either.

The graph of IBM's revenue over the past few decades does not tell a happy story: https://www.statista.com/statistics/265003/ibms-revenue-sinc...

It doesn't get better if you go further back. IBM has the same revenue now that it did in 1996.

Robust usually means strong and vigorous. I think it would be a challenge to find someone who believes the past decade at IBM has been that, particularly the Rometty era (IE 22 consecutive quarters of revenue decline)

It would be interesting to hear the theory of what happened between developing the interesting engine tech under the hood, suitable for flexible collaboration and active databases, and end user riots. Was it that marketing had to decide between promoting the bundled mail/calendar apps and promoting it as a custom app platform, and latter withered away?

Yes, but you presumably hire people who are good at their jobs.

IBM hires... bodies.

And I'd imagine internal IBM IT is... anyone they can't use to bill a customer with.

Ouch.

They have a 1.8% market share in cloud, to be fair

(probably about 99.5% of which is their own usage internally)

They have their mainframes, too. Not a big market, but highly lucrative, AFAIK. I read something about the mainframe division contributing 10% of their revenue but 25% of their profits.

Red Hat is all gsuite and runs all of its own systems like HR, etc.

Maybe IBM should hire one of those big IT companies that specialises in consultancy and service design.

to their own servers

"They were transitioning to IBM-owned servers," one source told us. "That's where it broke down."

Wouldn't a sane approach be to settle on one system first? Or to upgrade one of these to the target one? Migrating all of them at the same time sounds incredibly complicated, with no fallback option if something goes wrong.

IBM squired Lotus many years ago and has maintained the Notes / Domino product line ever since. The basic architecture has never really changed.

I did a few projects with it, it worked really well but we didn’t use it for email.

That's the joke

They are far more of a services and support company than a tech company these days. Their current tech is mostly used by their services teams to implement solutions.

https://www.statista.com/statistics/274823/ibms-global-reven...

Only novel tech they have been working on that I’m aware of is quantum computers…

As I've said before in comments here on HN and Twitter, there is no real definition of a "tech company" or "tech industry", the word "tech" defines the operating model. This is the dirty little secret VCs don't like to tell you because it means they can call Tesla a tech company (its a car and battery manufacturer) to fetch tech-like valuations or Facebook a tech company (its a media company). Tech operating models typically net high gross margins (i.e. Facebook) which is one of its major defining characteristics.

I therefore posit there is no such thing as a "tech industry", but rather businesses that sit on a spectrum of operating models from:

- Back office IT supported

- Tech enabled

- Tech Led

Personally I would define IBM as a Diversified IT, IT Services and Software company that is has an operating model that sits somewhere between tech enabled and tech led.

It's my understanding that they are largely a consultancy now, helping their clients move things onto their cloud products. I imagine there is quite a bit of vendor lock-in. Seeing as the need for cloud migration is ever-growing in every single industry, I don't see IBM becoming totally irrelevant any time soon. But they are definitely past "Day 1".

IBM is a consulting company. Only IBM Research and their mainframe division do anything interesting in tech. And their research division seems to be mainly there for marketing/image purposes (helping consultants sell overpriced “solutions” to clueless top management).

That makes more sense anyway. Why should your individual identity be hardwired to your current location or department? If you changed jobs within the company you had to get a new email address and login?

Same thing happened at one of the largest employers in the state. They have several entities all with different domains. Now everyone's login is different from their email domain. Even though no matter which domain you use in the To: field, It will still be routed properly.

You are speaking truth my friend.

>the majority of it appears to be outsourced to a third party in India

IBM should just be able to run down the block and talk to them then.

MS has always been stronger in B2B. Consumer facing.. well there's been some missteps.

My experience was that MS understands how IT departments work (from small to large) and are fantastic at providing resources to customers if it means you'll end up using more MS products. At a prior job they paid for a local consulting group to configure the Office 365 tenant, connect it with the existing on-prem infrastructure and train the staff on migrating mailboxes.

Office365 as a platform appears to suffer from terrible unreliability. My last role was in a google apps shop, and i miss it every single day. (it's been 3 years).

That's where I am too. We've been moving to 365 over the past year, and during the pandemic I finally cut the cord to the old technology (corporate IMAP/SMTP which I then glue up with clients myself).

And... sorry everyone, but BY FAR the best, cleanest, most performant and reliable work-provided email/messaging/calendaring solution I've had during my almost 2.5 decades of being a full time Linux user is... Microsoft Office 365. And frankly the competition isn't even close.

I don't love everything about it. But... OMG, it works. It just works in my browser and I don't need to do anything to get all the features that IT departments for decades were only able to make work on their own blessed windows images.

Same story. We have some weird behavior with Microsoft services but rarely and don't suffer from lack of resources and instability due to inflexible self hosted infrastructure.

Doesn't the web version of Outlook work?

I believe NTFS has support for 32767 char filenames (but 255 per level), but yes, Windows Explorer does not. Maybe in Win 11?