Before I worked at TSS doing in-home PC repair, I did a short stint at Radio Shack and got the Compaq experience you described. I remember going to the introduction event for the Compaq roll-out into Radio Shack stores, I was sitting in a large hotel ballroom with like 100 other Radio Shack sales folks. The Compaq guy was asking questions from the audience and I was literally the only person there who could answer questions like "what motherboard interface did the Pentium 2 use" (it was the first cpu to use a slot (Slot 1) versus a socket which was a big deal back then). I had dropped into the event straight after class at ITT, so I was just wearing casual attire and didn't think anything of it. Apparently that was a mistake, one of the higher up sales guys got offended and complained to my manager. I was offended that apparently no one else there knew anything technical at all, lol. Thankfully I left that job not long after.
I was spending all my time learning computers, and I was starting to build systems for family and friends. Once I was lent out to another Radio Shack store that was short handed and the other salesmen kinda lamented not being able to afford the Compaq a customer was buying. Afterwords, I tried to explain to him that I could easily build him a nicer system for a fraction of what we sold Compaq's for. He was still not interested, I think he was way more broke than I realized. I grew up broke too, but I also had my eyes set on improving my situation. Was kind of a wake-up-moment that most people around me had minimal ambition about this.
I also remember having to push cell phones at RS then too, I think we pushed a lot of Sprint phones. I was still fairly young and was just learning about phreaking, but I remember my boss telling people Sprint phones couldn't be cloned like analog cell phones could back then (I disagreed, but I had to agree it was harder at least). It was apparently a thing that your analog cell phone could be cloned over the air and suddenly your bill would be crazy that month. I only ever heard rumors about that kind of thing happening.
> I thhink he was way more broke than I realized. I grew up broke too, but I also had my eyes set on improving my situation. Was kind of a wake-up-moment that most people around me had minimal ambition about this.
I feel like I'm missing something, because it's not clear at all why that would be your takeaway. Why would "not trusting a temporary employee that I don't usually work with to allocate my home computer budget" somehow imply "broke and unwilling to do anything about it"? Without more detail, that seems like a kinda weird inference to make.
It was the sadness in his voice and the look in his eyes that aligned with his words. I certainly experienced plenty of skepticism directed at me based on my age in my youth, this was not that.
These old ads are great, really takes me back to that era. In the mid to late 1990's I was a pimply faced youth in community college and picked up a part time job doing "PC repair". I couldn't put my finger on it, but the outfit carried itself like it was a major corporation despite being a small office, it was called TSS (Technology Service Solutions). The majority of the other guys were always working in the field so I wouldn't get to know them until many months later. When I got my first paycheck (physical paper), it said IBM on it. I asked my boss why this was and he nonchalantly said "we just use IBM's payroll system". I was very confused but didn't ask any further questions.
What I came to learn later was that TSS was IBM's PC service division which it had sold off to Kodak just a few years prior. I spent most of my time doing contracted in-home PC repair for all the major PC brands of the era, including Gateway. I'd get a service call (on a RIM bullfrog two-way pager), call the customer and tell them "Hi, I'm with <major brand computer>, give me directions to your house and I'll fix your system". Most of the consumer PC manufacturers were unequipped to provide any kind of in-field repair service on their products, so they hired TSS for this. If you bought any kind of extended warranty on your PC then you were paying for TSS to show up on-site at your home to fix it. Apparently this was better than regional central repair hubs.
Let me tell you, there were some seriously crazy stories from the front line. Most of our work was through Sears for Packard Bell (cheapest machines sell higher volume and therefore require the most service, not even counting for PB's lower quality). We serviced Acer, Sony Vaio, Gateway, Dell, and other brands, and did other point-of-sale and other contract work. I remember servicing the Gateway Destination system - that 35" CRT was huge for even a TV at that time. But being up close I realized it actually kinda sucked to use as a computer monitor. People tried to set them up in their living rooms, so imagine using your computer in the living room instead of an office, and required sitting next to the 35" display meant you saw the CRT up close which just meant the low resolution was blown up larger and you could see the physically larger pixels up close.
So yea, showing up to peoples houses meant going to a lot of random peoples houses which was mostly fine, but occasionally terrible. I had multiple instances of working on computers in a house having brownouts. "So wait, the computer reboots when the lights in the house flicker? And dispatch has sent me out with a new display for you... yea, you're gonna have to get the power company out here...". One time, I was sent to a house that was barely a shack in the woods. Main room had a light bulb hanging from the ceiling. That call in particular stands out in my mind, nice people, a 70 year skinny old dude in overalls and a 400' bearded lady in a shack. I couldn't do anything to make their Gatway machine work better as it was the power lines (lightbulb was visibly flickering the whole time I was there, that poor PC never stood a chance), so I showed her how to play her music CD's in PC (between power sags that crashed the machine) and she was tickled pink. One of my trainers had a couple horror stories: one dudes house was an underground bunker. My coworker was sent to replace a modem. After, the client couldn't find his Solitaire on his desktop, so he locked the door and demanded my coworker fix his Solitaire game before he could leave. My coworker obviously had nothing to do with that issue, but he recreated the shortcut to the desktop and was set free.
Eventually I jumped ship from TSS to IBM proper, it was a fun adventure. But that year and a half slinging in-home service calls in the evenings was an interesting time.
It's cute to see folks who have completely drunk the in-house koolaid try to communicate to outsiders, and we're better off for it. I really struggled to understand OS/400 coming in with a working knowledge of Unix/Linux, concepts as simple as creating a text file document in OS/400 don't translate at all.
Yes. I could not fathom why anyone would want to run anything besides AIX on them though. You can run all the other OS's on cheap commodity hardware, the draw for me was having an AIX system to work with.
I was familiar with both the "laptop AS400" and the PowerPC thinkpad - these were different hardware unrelated to each other. I believe the sales demo machine was a CISC architecture demo machine that predated PowerPC, but I'm not 100% confident here. I never laid eyes on the AS400 sales demo machine but we were told about it in IBM AS/400 training school. I also had one of the PowerPC laptops running AIX for a while, nice little machines for the day.
As a youth a few decades ago I had an opportunity to play with one of these systems at home (and I actually still have it boxed up somewhere). I managed to acquire a 9401 Model 150 that fell off the back of a truck, it came with a 5250 terminal and a heavy box of documentation. Even as a kid who had started playing with linux in high school I really struggled to get it to IPL - the system was seriously weird from anything I had known before. Strangely enough, within two years of that I ended up working for IBM and went on to be trained on servicing AS/400 and iSeries systems. What was fascinating as an outsider coming in to IBM was just how hard they clutched their pearls on the AS/400 system. I also serviced RS/6000 (pSeries) and RS/6000 SP clusters so I saw two different sides of the house operate. I was a typical pimply faced youth eager to learn, and the pSeries guys, both in the field and in IBM training schools were happy to feed me all the info I could digest. When I needed software for testing stuff on my own personal AIX box, stuff would get passed my way no questions asked. On the iSeries side, they would quite literally tell me "no" and shut the door in my face if I was asking too many questions. Despite the fact that OS400 can't be ran on any other hardware than IBM's, IBM guards the SLIC or OS "golden tapes" very jealously. The class for learning how to service iSeries was four weeks long back then, and our brains got crammed full of tons of information. But if I asked questions outside the scope of servicing the hardware I'd still get the door slammed in my face ("that's a different class kid, your manager will have to approve that").
I only worked there for about five years and I am grateful for the experience. I'm also glad I left when I did to avoid becoming "institutionalized" in my thinking as I had witnessed others had become. I have to assume that the pearl clutching came from IBM's experience with Amdahl and its prized mainframe (zSeries). They had become so afraid of competition that they were even afraid us lowly CE's would start freelancing and selling consulting services on the side on OS/400 work. As a result of that, you have a walled garden which is very hard for fresh talent to penetrate and IBM likes it fine that way.
Anyone running ripgrep on a an HPC cluster against a large cluster filesystem needs to stop and redesign their workflow. This generates high amounts of small I/O which is the Achilles heel of any large cluster filesystem. You are exporting your workload onto the metadata mechanisms of the filesystem rather keeping it within the higher bandwidth capable memory subsystem on your cluster. It doesn't take but a couple users running these types of jobs simultaneously to bring a high-bandwidth filesystem to its knees. Just stop it already.
Similarly, I've often wondered if this is the root cause of why GitHub has become fragile of late. All of a sudden you have operations on billions of tiny files that are amplified by AI usage. And object graphs are so fragmented by nature that it's hard to e.g. prefetch a page into memory and ensure that common git operations mainly touch that page's set of objects.
> All of a sudden you have operations on billions of tiny files that are amplified by AI usage. And object graphs are so fragmented by nature that it's hard to e.g. prefetch a page into memory and ensure that common git operations mainly touch that page's set of objects.
In theory, that's what git's packfiles are for: loose objects get packed together into a file, and if the file is well-arranged then objects that tend to get read together will also end up getting paged-in together (see also: https://www.kernel.org/pub/software/scm/git/docs/technical/p...).
