Owner of 'magic spreadsheet' tried to stay in the Lotus position until forced to Excel Welcome once again dear reader, to the sanctuary of sympathy we call Who, Me? where tales of technical derring-do are shared alongside stories of derring-at-least-you-tried. This week meet a reader we'll Regomize as "Randi" who worked in financial services. Among his colleagues was a product manager who jealously guarded his …
To do it properly, you need to combine the triangulation of IP pings with Jake's use of echo ^G (hang on, wasn't that the - apocryphal? - origin story of IP ping, that the receiving machine actually went "ping"?).
Bonus points if someone is also following the cables and tugging on them to see what jolts loose:
"Ping"
"5 metres"
"Ping!"
"3 metres"
"PING!!"
"1 metre"
"That's not possible! It's inside the room with us!"
[FX: BOFH tugs hard on ethercable, box shifts on false ceiling, teeters on foam tile the falls, dragging BOFH up partition wall]
"Aaaargh!"
[FX: ceiling collapses, wire tentacles drop, entangle PFY who is yanked, screaming, into void as server rack crashes down on other side of rafter. PHB fires wildly and terminates anyone in his sights]
(TV edit: removed the following scene, where techie leans over server, says "It's dead" just as the optical drive tray shoots open, impaling his dangling id badge)
Yes. "ping" is not an acronym, and never was.
It was released as public domain software first (1984ish), later changed to the BSD license (1985ish). It was included in 1986's 4.3BSD. Obviously we used it in 4.2BSD as we were building 4.3.
Note that some variations of ping include an audible option. Handy occasionally. Usually annoying.
Also, see: https://www.catb.org/jargon/html/P/ping.html
I agree. I also recall using the Linux eject command in earlier days for identifying a machine in a stack that had been mislabelled as it also had a function to close the tray again if it was a motorised tray instead of the cheapo clicky thing they fitted to laptops - looping that worked.
yeah, just don't do what I did and use a Monty Python film quote piped to the sound card. "Help, help, I'm being repressed!" was the specific file used, except the passer by only heard the "Help" part, relatively infrequently (I think it was every 30 minutes with cron) causing security to wonder where the trapped person was... Luckily, the security guard was a mate from outside work and didn't report it to anyone in charge. Just our mutual friend so he got the chance to give me a good ribbing about it...
Been there done that, spent weeks hunting servers ( not just PCs ) hidden under desks when my former $TELCO employer moved from an office building to another.
The whole idea was to catch them all and put them in racks in the relevant room in order to cull down the quantity of shadow IT equipment. ( not that it mattered much, they started to reproduce again once in the new building. It took network MAC lock out to get rid of anything not official. )
"Ehm, more like: I can ping this important server machine, but I have no idea where on the premises it is physically hiding. Tracing cables through the ducts has not helped to resolve its location, yet."
Accidentally immured?
There was a (apocryphal?) yarn about an IBM 370/VM than ran for more than a decade after being accidentally bricked up.
If your awol server were connected via cat4/5/6 twisted pair ethernet then it would not be much more that 200m from the last switch. Single mode fiber I supposed it could be in the next state.
One option I have used when just pulling the connection was unwise, was to degrade the connection eg if it was 1Gb lock it to 100Mb; if 100Mb slip an old 10Mb hub in. If anyone is aware of the renegade's location they will generally betray that in order to fix the performance problem. Things you learn in BOFH School }(:->
The Union Rep.
Hardware refresh & the union rep wouldnt tell me where "his" machine was, in one of the more older builldings on the site.
Despite being told it was'nt his in the first place, he was concerned about confidential files (Past, present & future) regarding his members & litigation over unfair dismissals, injury claims & the like being accessed by the company. The news that as the machine was owned by the company, anything stored on it was theirs to read, moreso I if I was that bored I could also access the files stored on it. The files would just be copied accross to the new machine & the old one wiped*.
We took it off the domain, expecting a ticket to come in - Nada!
I edited the boot.ini file over the network so it would fail to boot - Natch!
Then came the day he must have restarted it & it was no longer on the network - Three years later, when I left with still no clue as to it's location, or indeed how the union rep got his very important files off the locked down NT4 machine with no USB access.
*The thought occurs to me some 20 years later as I write this, I should have just told him the profiles & data had already been dragged across the network onto the replacement machine & when would he like it installed.
Then came the day he must have restarted it & it was no longer on the network - Three years later, when I left with still no clue as to it's location, or indeed how the union rep got his very important files off the locked down NT4 machine with no USB access.
From some of the experiences that I had doing IT support, there would still be a few ways of getting data off that machine.
Depending on the configuration, there could be a SCSI card in there, so an external CD burner or Zip drive could be an option. There may even have been an internal CD burner, or he could have used a parallel port Zip drive.
Hell, even good old floppies were still viable in the days of NT4. Even dozens of word documents could fit on a handful of disks.
This post has been deleted by its author
"There was a (apocryphal?) yarn about an IBM 370/VM than ran for more than a decade after being accidentally bricked up."
Here is a real version of that yarn, reported in this very organ a mere 22 years ago!
tracert / traceroute
Or at the very least look at your routing tables for the corresponding gateway. If your device can find a route to the destination, you can follow it.
Get onto the final segment, then the infrastructure will know the MAC - it has to know it otherwise it can't communicate.
If you don't recognise the final segment, do it iteratively by gateway.
That can be improved: A server hidden behind DNAT or somewhere in a VNET, or a local NAT most hypervisors offer, so you don't even have a MAC address you can follow. And traceroute (including tracetcp and traceudp) won't help too 'cause half of the hops don't even respond. That can take some time to trace it down, believe me. And don't get me started on nested-virtualization which is slowly becoming standard.
We had a contract involving high-performance machinery on an application requiring high availability. Big project updating (very) old equipment with modern and lots of automation: involving a change in manning levels for operators and their maintenance crews.
During the transition from old to new, the equipment kept suffering unexpected trips and shut-downs often at inconvenient times including weekends. In order to get a handle on what was going on, we put detailed logging equipment on the instrumentation with a remote-access PC where I could dial in and observe recent historical data. We could then attend site with knowledge and reset things accordingly. No real reason for the trips was identified until 'secret' logging of the 'Hand/Auto' control switch was included. All the trips were associated with operation under (unnecessary) 'Hand' control. Unsurprisingly, when the operators became aware of this phenomenon the problems disappeared.
After a long, uninterrupted period of correct operation we removed the logging equipment but I left an old laptop installed to give the illusion of 'supervision'. It had a note and screensaver saying not to update the W95 operating system.
The problem we have faced over the years is suppliers deliberately making their software only work with a particular version of one OS and version of IE and so on.
One application we had worked on Android tablets so it was only "that" version of Android and "this" version of Chrome. When we found that they were writing their software to version check for no good reason, their days were numbered!
For a while, we stopped all updates on those tablets while we waited for new versions of the app. When they came, they were for 6-month-old versions of Chrome that we could not get.
They are gone now.
Not to defend sharp practice, but there could be a reason for that behaviour.
Medical device certification (and pure software can be a 'medical device') to IEC 62304 is rather exacting and expensive, and will specify the exact versions of software that can be used - because they are the ones that went through the long and expensive testing process. Changes are avoided so that recertification is not required. There are various levels of certification, and if medical decisions are going to be made on the strength of data produced by an app(lication), then the certification requirements are rather stringent. (Medical decisions being: if you get it wrong the patient is harmed or dies)
I worked on a project once where the idea was to use the patients' fitness monitor records of pulse rate to provide data for the management of the treatment. One of the rocks on which the project foundered was the fact that almost all such monitors are not certified as 'medical devices', so you had no audit trail to certify you were likely getting accurate data. And while those monitors were actual physical devices, the same rules apply to the software used to process the data.
If the application being used was to support medical decisions, it probably had testing and certification requirements that made updates a painful, long, and expensive process.
On the other hand, it could just have been sharp practice.
... is that such "magical" bits and bobs are always superfluous to the organization. And usually, so is the secretive person attached to it.
In fact, so much so that when I run across such things as a contractor, I make a bet with the company owner that it's junk, and we can safely "accidentally" delete it during the upgrade (after making suitable backups just in case, of course.), and that the company will continue running without change.
In over thirty years, I have never lost that bet.
That’s interesting. Most people hated Enrage with the fury of ten thousand suns. I spent actual real money on Eudora rather than use Enrage on my personal machines. Anyone still using Enrage must be on Office 2008 or earlier (2007 in Windows years). Not that Outlook in 2011 was any prize. Damn. I still have an eMac that will run Office 2008, but that’s about it. What hardware is m’man running?
My IT group was upgrading various departments in the hospital to PCs: Win3.X, Ethernet, Novell 3.X, and SaberMenu. I encountered one user who, when I showed up with her brand-new PC, angrily refused to let me touch her IBM 3278 terminal. There was a second 3278 under her desk, cabled to the first terminal, configured to act as a communications controller or something like that. (I never learned the fine details of IBM SNA.)
Not quite as bad, but last year my employer took PCOMMS (the IBM mainframe terminal emulator) off our standard build. We still have a mainframe doing mainframe things (which in fairness it does very well), but we no-longer assume that every employee will need it.
Progress, I guess? most staff use nice sensible designed GUIs and almost no-one has to screw around with command-line stuff or JCL. Moving off of mainframe reliance took a hell of a training burden away, the CLI is just not intuitive to new users at all.
Older Doctor used a DOS program for preparing patients' bills "written by his colleague in 1989" that ran on an old Windows Millenium machine hooked up to an old B&W laser printer that I finally convinced him was unsafe to use circa 2019. He promptly went out and bought a "refurb" box with Windows 7 installed but only because another old friend suggested he use DosBox on his "new" computer to run the old billing proggie but for some arcane reason it would not compute.
In order to keep the WinME box as far away from the internet as possible, I set up a 2-port KVM switch with the web-facing Win7 (he did use email and an online calendar!) on one side and the old WinME box on the other with the old printer attached directly to via crossover cable - the newer computer had its own combo printer/scanner/fax inkjet (he still used faxes as indeed, many pharmacies still do at least here in the USA).
This all despite the fact he was supposed to be completely online so that clients could access their billing etc. While he was dragging his feet in this regard, he managed to dodge professional censure, at least until the WinME computer finally booted its last. After two months of basically working for free as he could not charge anymore, he finally retired!
The mention of the Pet reminds me of the time, back in the 1980’s when I visited an oil industry offshore support vessel. This one had high pressure/capacity pumps that could be connected up to an offshore platform with flexible hoses for specialist downhole work. It had to hold station close-in for this (else hoses would break or, far more seriously, it could collide with the platform). The vessel has the necessary arrangement of bow and stern thrusters for manoeuvring; for positional reference, it has a taught wire connected to the platform, with a Commodore Pet to monitor the distance and feed commands to the thruster controls. Crude by today’s standards (nor would it be allowed in the North Sea nowadays) - but it worked!
I worked on a project for a division of our company which used an Apple][ to connect a VAX to a PABX. The VAX emulated the operator console to answer and route calls, but the original console used some weird 10kbps (not 9600) serial protocol that the VAX couldn't generate. The Apple had a very programmable comms card, so acted as a protocol converter.
I remember that it ran very warm, so sat on the floor behind the VAX with the lid off. Ran like that for years.
classic boffin behavior. It normally goes like this. Boffin gets a pot of cash to buy a new bit of kit, HPLC, GC, etc they then buy bit of kit that comes with a PC connected to it which they use, and use and use. Eventually PC needs replacing (the bit of kit will run for years and cost a fortune so its fine) trouble is they haven't purchased any form of support contract, beyond the basic so the software used to run the very expensive bit of kit now won't run on a modern OS and to get the new version will cost actual cash which the boffin doesn't now have as his grant pot was used up years ago! Seen it soooooooo many times ;o)
Boffins are constrained by the grant agencies and their supporting institutions. For instance, the USA's NIH offers shared instrument grants with awards over $1M but they pay only for hardware. Maintenance, infrastructure, and personnel costs are supposed to be borne by the institution, which provides a support letter saying yes, we will support this instrument. Three years later when the original included maintenance contract runs out, institution says you have to self support via user chargeback fees, which in most cases fall far short of what's needed.
I've had excellent success with DOSBox, even using it to run PLC software that requires COM1 or COM2 (ain't any higher numbers, right?) on a Windows machine with a COM5 USB-to-serial dongle. Generally only software that does seriously low-level stuff or uses DOS programs that it expects to be present but aren't, like XCOPY.
If you do a lot of DOS stuff with different DOSBox profiles, take a look at DOSBox Game Launcher (DBGL). GUI for setting up and using different profiles for each game, er, program. Java-based so runs on anything.
A microprocessor validation lab at an old job still had a Windows 3.1 machine hooked up to a very expensive ion beam device circa 2015. The vendor never bothered to upgrade/rewrite the software so it sat in a room with no network connectivity while everything else ran Windows 7/Server 2008
I had a major Telco asked for assistance on a server consolidation. The biggest challenge was the kit that was mission critical but either had no owner in the organisation or an owner who basically stated that it was so important the entire fate of the company rested on that particular server. We took the approach of consolidating all the servers for which we had reconciled Owner, application function, business alignment. We then started to switch off (Physically) the servers we could not identify the owner, the real business requirement etc.
We had a couple of servers that did pop up as a requirement (They were important on a quarterly basis) and get these assigned with new owners etc etc.
we went back to any owners who had stated the mission critical nature of their servers 14 months in to advise the servers would be decommissioned and removed - One guy literally screamed that his server was so important that the company would be out of business in a matter of weeks if we did this to his server - I was somewhat perplexed as the server has been switched of for over 14 months at this stage and when we advised him of this fact, he seemed some what deflated and meekly said he would agree to the server being decommissioned.
The switch off approach did work, there were some servers used on an infrequent basis but we caught most of these early on and the vast majority that were switched off were never switched back on and there was very little issues caused (Just some minor inconveniences).
We had a consultant with a laptop that was encrypted wih McAfee - which needs to check in with the server every 30/60/90 days (can't remember) otherwise it will lock until you give it a code.
Claims that he used this laptop almost everyday, but all of a sudden, it came up with the lock screen. We couldn't unlock as the server had been switched off and scrapped two years earlier!
Been thinking about this. I use a FOSS encryption tool or two as required. But I won't use one which isn't well supported with plenty of download sites and I always keep a copy of the installer somewhere, too.
Because there is no interoperability between any programmes, even if they use the same encryption method. So the software becomes a risk- the risk of not being able to access the data- should the programme not be available any more.
This, surely is a barrier to new programmes. They do appear, but personally I wouldn't use them until I'm reasonably sure of them. I had thought about encrypting two copies, if it ever came to it, with two different programmes, should I be needing to secure something seriously important to toke away from home, for example.
Mostly rely on Vera btw.
It's not uncommon. And that is the scary part.
I really wish I could elaborate, but to do so would earn me a tap on the shoulder and a quick van-ride to a place I'd probably want to avoid.
Though one system I can mention is a Linux system that was set up to connect to another company to send files. It replaced an old Windows machine that was unstable (both in terms of the hardware, O/S, and software it was running). 6 years(!) after I left the company, one of the guys I used to work with called up in a panic because files were not being transferred. Turned out someone had decided the trusty old system was no longer used and just turned it off. A quick check of the logs would have shown otherwise, but hey-ho....
Spreadsheets are an unsuitable tool for scientific or business analysis, as they are too prone to error. Some examples.
Gene scientists were using one to analysis gene sequences and entered something such as ‘apr1’, which the spreadsheet helpfully converted to a DATE April 01, thus rendering incorrect results on any calculation.
The UK Gov department responsible for calculating the Coronavirus growth rate or R number used a spreadsheet. This was calculated from a number of sources imported into the spreadsheet from CVF files. The master spreadsheet had less rows than the sources, so the excess data was discarded. thus rendering an incorrect R value.
The solution is to use a database in the backend with the frontend designed to look like a spreadsheet. But most managers insist on using spreadsheets as it's the only thing computery in college. And they are unable to consult with their own IT people as that would be beneath them.
For a personal anecdote here one. The finance manager (using the industry standard spreadsheet) imported records from some ancient flatfile database. This came as one unique record number per line, followed by a number of fields. Anyways, once he managed to import the data, he then sorted on the record number, except the spreadsheet left the fields in the original order. Later on people started to complain that they were getting erroneous results on end of month totals.
Who knew /s
Spreadsheets are great for certain tasks, but they're not databases or the other various things they can be made to stand in for, And IMHO beyond a certain (perhaps subjective) level of complexity they can't be trusted. The level of complexity is the point where the user no longer knows what's going on and can see if something is screwy.
But too often ordinary users have leant how to use a spreadsheet and it's their automatic go to. Especially if the employer doesn't want to spend money on a proper toolkit/training.
The average user has no idea what a database is, so they don't turn to one when they have a requirement that needs one.
Whereas they are familiar with spreadsheets and if the requirement seems to be like a big spreadsheet, that's what they'll use.
I do wish all the "helpful" formatting (convert to date etc) was turned OFF by default though.
Annoyingly they also use their f***ing spreadsheet when a simple WP table would be the best thing.
Having just spent a jolly hour copying my part of an exam invigilation time table from a spreadsheet which happens to the omit any half days when there are no exams ( so the cells aren't even aligned).
It is literally a block of half day cells for each of the 8 weeks, each with a long list of names vertically for each person working that week, and dates horizontally with Y or N in the cells for each half day for each of us. And it won't fit on an ordinary sized monitor because the spreadsheet just continues. You have to scroll across your line to find which days to go into work, then enter it into your own calendar of choice There are no formulae. Just a set of untidy tables. And endless opportunities for errors
Python to the rescue.
Export as .csv.
Now read the .csv with python.
https://docs.python.org/3/library/csv.html?highlight=csv
Your python program extracts everything referring to your name and creates a .txt with an overview of your working days.
I do hope names are standardized, howver.
Once created it saves you the hassle of digging through the original file.
And just when the spreadsheet gets too complicated what does some bright spark do? Start using Access of course, which is a whole new world of pain!
That said, for some years I had quite a nice gig with a client reimplementing such home brewed databases using PHP and MySQL. These guys were consultants who charged more by the hour than I did by the day or possibly week, monitoring multi-billion pound projects, but their IT was like turning up in a Hummer only to open the door to find a couple of guys frantically pedalling to keep it going.
One classic example. A new partner had joined the firm and brought with him his very special Access database which was probably why he was hired. They needed to keep using the Access database while I converted it, but one small problem - the database screen still contained the logo of the previous company, not a good look at meetings!
So I was asked if I could do a quick hack to replace the logo with one from my client. Easy to do, but when I deleted the old logo what did I find? The logo of yet another company where I assume the software was written in the first place. No wonder no-one knew how the hell it did what it did.
Anon because...
> <i<Spreadsheets are an unsuitable tool for scientific or business analysis, as they are too prone to error.</I>
The problem is with the programmer…
If you have grown up with say C and so have a real understanding of the need for precision etc. then it is possible to write error free spreadsheets. However, the typical spreadsheet is thrown together by less expert users following an Agile development process…
> The UK Gov department responsible for calculating the Coronavirus growth rate or R number used a spreadsheet.
Would not be surprised if actual DB linkage was too expensive and would have involved a lot of effort. Says he remembering the HMRC lost CDs which contained a full data dump rather than the just the required data because getting EDS to produce the required dataset would have cost…
"he then sorted on the record number, except the spreadsheet left the fields in the original order."
To be fair, that's user error, not a problem with the spreadsheet per se.
He knew enough to know he could sort data, but clearly had no idea how to sort data in a spreadsheet.
"Spreadsheets are an unsuitable tool for scientific or business analysis, as they are too prone to error."
Anything used wrongly is 'prone to error', FFS. Just admit you don't know how to use spreadsheets and don't want to learn.
"The solution is to use [...] a spreadsheet"
LOL.
Once upon a time a company begins with T had a neat product for optimising Reuters market data feed. It was doing quite well and installed in a number of banks.
Then they added a little extra feature. An excel plugin which would automatically update a cells value when a selected share price changed. Word got around and sales went through the roof. Nearly every trader in the city of London had their own custom spreadsheet. I heard a couple of anecdotes about traders demanding the most powerful available PC just to run their humongous spreadsheet.
Speaking as a VBA dev at a big financial company -
Excel can do damn-near anything. It's a powerful, flexible tool that lets users freely decide on how calculations work and how results are presented. It allows changes on the whims of PHBs that you just can't do in so many other packages, and the internet is full of useful, free, guidance on how to use it. You can build complex models, you can build ETL packages, you can make fancy graphs, it can link to databases, fold, spindle and mutilate almost any dataset.
Arguably, Excel *should* do a whole lot less than it does.
So long as it remains able to do so very many things - even if it's not the fastest, the cleanest, or the most auditable - it's going to keep being asked to do them.
One thing Excel doesn't do is source code control/change control/version control. There are add-ons and external tools, but the typical end-user building a big Excel workbook doesn't use them, so blithely modifies individual cells, copies incomplete ranges, makes mistakes with formulas, and specifies incomplete ranges in functions. Referencing other spreadsheets and external data in a controlled manner is also harder than it should be. Auditing big workbooks is 'interestingly' difficult. And don't get me started on using a common Excel workbook amongst users in different countries.
Excel is just not bad enough for people to throw it away and use tools more suited to the tasks at hand.
To the extent that geneticists have renamed several gene to work around the problem. Presumably because changing the human genome and scientific literature is less trouble than getting Microsoft to fix a hateful spreadsheet application.
I used to work for a Japanese company, and my colleagues from there used Excel all the time. Engineering calculation to do? Excel. Financial calculation to do? Excel.
Letter to write? Excel.
Presentation slide to prepare? Excel.
Gannt chart for a multimillion pound chemical plant installation to prepare? Excel.
Not that they didn't have correct Office tools for all these things - they just all used Excel. Baffling.
“ The UK Gov department responsible for calculating the Coronavirus growth rate or R number used a spreadsheet. This was calculated from a number of sources imported into the spreadsheet from CVF files. The master spreadsheet had less rows than the sources, so the excess data was discarded. thus rendering an incorrect R value.”
Hmmm, I might have said “…the outsourcing crowd stupidly brought in by the government were using an out of date MS Office version due to penny pinching which was clearly not fit for purpose despite the huge amount of money thrown at them for providing this service, as piss-poor as it was”….
Story vaguely reminds me of a part-time tech support job I had shortly after the KT extinction. Most of the machines were XP with a smattering of older 2000 units still in active service. If memory serves, the entire network had been compromised and my job was to go around to each one and make sure that Windows Update was run. The XP machines had been pressed into service as Warez mirrors while the 2000 machines had been left alone because they apparently weren't worth the bother. Anyway, my travels took me to a dusty corner of one building where I'd never been before and I found an NT4 system. Being the dutiful helper I started trying to run Windows Update on it to make sure it was as updated as could be. For reasons lost to the mists of time, I happened to mention finding an old NT4 system to my supervisor at the time who promptly told me not to bother with that unit. It was being intentionally abandoned in the hopes one day it would keel over and give them an excuse to replace it.
When I first used it in a mainframe only office, the only PC they had was a terminal to the Sperry, that came with a copy of 1-2-3.
After a few hours messing with it, I showed it to a colleague who immediately saw it could do 3 days of work in an hour.
When I left, 9 months later, they had 23 PCs between 50 staff. Which is why I am a bit "meh" about the hype about progress this millennia. I don't think we'll see that rate of growth ever again.
Anyone remember Freelance ?
Office 97 works with Windows 11 too - except or Access due to missing ODBC. Office 2000 works too except for outlook requiring some old IE 4.x-components - which I probably could have delivered, but was too lazy. Deltaforce 1, from CD, works too. A surprisingly long list of 32 bit software still works.
I call BS on your calling BS.
The bit you're reacting to appears to be this: "any operating system released this century wouldn't run [1-2-3] at all."
The people involved could have mistakenly believed it wouldn't run, and acted accordingly.
Or Randi knew at the time that it could run but did away with it anyway -- or more likely, didn't bother to look into the question --because his goal was to get the guy onto modern kit, not to enable his continuing obstruction; but Randi now, in retelling the story, is misremembering.
Or it could have been an older version of 1-2-3 than your folks have, which truly couldn't have run on a modern OS.
So there are any number of ways for that one brief assertion, even should it be false, not to invalidate the entire story.
We still get end users demanding static IPs because their software requires it (from really obscure subject specialist outfits comprising one man and his dog).
One was even compiled on-site after the IP addresses were put into the code because that was the copy protection method being used, the installation engineer didn't seem happy when I explained how NAT & VLANs could be used to duplicate the environment multiple times* on a single box costing less that a years support contract.
* Yes it'd be a nightmare to maintain or modify quite apart from being naughty.
While scouring the data centre for kit that could be taken off charge I came across a machine, still powered up, of great antiquity with dust on its keyboard. It turned out that it held data for a now-obsolete system that needed to be retained for 12 years for legal reasons. The software which used the data was so tightly bound to the NT4 OS it was still running that it hadn't been possible to dump it onto a VM so here it was, spinning away in case the machine didn't reboot if it was restarted. It wasn't on any network (not even a network card) and there was another 4 years to go until it could be wiped with no legal blowback. I found a couple of hard drives that matched the ones in that box, cloned the machine onto them and stacked them next to the holy relic and left it to live out the rest of its life in peace.
"it held data for a now-obsolete system that needed to be retained for 12 years for legal reasons"
I once worked for a place that had three mission-critical applications, each running on its own Unix server (well, Unixish; most of them were Xenix), plus there was a fourth server that backed up the other three.
I had taken to naming the servers after mythological figures. Charon (the ferryman across the River Styx), along with its main function, served as our Internet gateway. Vishnu ("The Preserver") was the backup box. Shakti ... well, to be honest, that one was named more for the band [1] than for the goddess.
Came the time that all three of those applications got replaced by two new ones -- all at the same time, but that's another story. Their servers got replaced by Apple kit and NetWare. [2]
We needed to keep two of the old applications available, though, for people to refer back to, essentially read-only. Given the much reduced load, we coalesced them onto one Xenix box.
I called that machine Baal -- for old, superceded systems, the name of an old, superceded god. [3]
[1] Shakti plays jazz fusion that's heavily influenced by Indian classical music. It originally consisted of John McLaughlin, Zakkir Hussein, L. Shankar, and T. H. "Vikku" Vinayakram. They have toured in various incarnations over the years. As it happens, they're on a world tour now, but not coming anywhere near me, alas.
[2] Far from ideal, especially for this old Unix hand, but those were the requirements for the two high-end, off-the-shelf applications we bought.
[3] Ba'al is the Caananite god that the Israelite prophets are always inveighing against in the Hebrew Bible, aka Old Testament.
There was an old MSDOS machine used to control card access to the doors in the place where I got my first permanent job. Do not touch, it's far too important to alter in any way even if it's totally unsupported, etc. One day it crashed anyway and I was asked to take a look. I can't remember what I did to fix it, but afterward I did suggest removing the games (Lemmings Christmas Edition and a handful of others). I was assured that those games couldn't possibly exist on that PC.
the whole estimating process in a shipyard I worked in mid 90's was run on 1-2-3. It grabbed the costs of resources, steel, calculated the man hours required to do a long list of standard jobs, etc it was a thing of beauty! Trouble was it was written and looked after by one estimator and the entire business depended on it! When I left they were in the process of moving to a new solution.
I've never let it come in the way of upgrading my PC but Palm Desktop is the golden source of my contacts even today. The software came free with my Palm m105 PDA in circa 1998, worked fine with the Palm Tungsten|TX that I bought in circa 2007, and has been working on all five Windows laptops that I've used during this period. On more than one occasion during this period, I've tried to migrate to Microsoft Outlook, Google Whatever, Samsung Whatever and half a dozen contact management tools but somehow I find myself going back everytime to the 25 year old tried-and-trusted warhorse.
I fondly remember my various Palm devices - my first being around 2003. (Yes...) Rock-solid, nearly impossible to crash, very reliable. Many happy hours coding on my Sony Clie during my lunch break using OnboardC and a whopping 128 MB MemoryStick (original, not Pro, Duo, or Micro).
Hmm. I wonder if I could get an emulator for my phone to play some of those old games. Kyle's Quest in particular.
You are not alone. There are plenty of people still using Psion Organisers of one type or another.
I don't think it is possible to get Compact Flash cards with small-enough capacity to work in the Psion Series 5 now - anything larger than 128 Megabytes is not guaranteed to work.
Reminds me of the "Operations Manager" at a place I used to work, job title in inverted commas because phrases involving piss-ups and breweries don't fully convey how useless this guy was.
He couldn't work Outlook or shared calendars, so every bit of scheduling was done via an Excel "calendar", and he would constantly go in the huff because engineers like me would put appointments in our own calendars and then he'd ask us to do something and we'd say we couldn't because we had something else scheduled.
He'd also schedule jobs 100 miles apart with an hour gap and not give people time to have lunch and various other similar easily avoidable errors, but his utter refusal to adapt or change and the directors refusal to get involved in any sort of conflict caused me and every other engineer to leave in a 6 month period, which I later heard was attributed to me "rabble rousing and turning everyone against him" :-D
We had a SQL DBA who insisted that his machine could never be restarted, since it took him half an hour to get everything open and laid out the way he required it each time. That lasted until Wannacry hit, after which point forced restarts overnight (if the machine hadn't been restarted already) sorted it.
Way back when I still had a job of consequence and meaning in IT, I worked for a nation-wide retail company. While I was acquainting myself with the server room, I hears a forlorn 'Fweeep' noise, like some woefully underpowered Piezo buzzer thats been stood on. It would go off about every 30 minutes or so, and I eventually located a very sad looking server hiding in the bottom of a rack, clearly out of step with every other bit of kit in the place. I kept hearing and asked one of the more sage engineers there - "Ah, that's the HR server... we must never touch it.". This thing had clearly had a very long and hard life and a little more digging revealed that not only was it probably installed by Noah himself, but it had been flooded during a pipe burst, and refused to die, literally run over by a transit van the site was ram raided by some local entrepreneurs through the wall (note: if you are going to have a ground floor computer room make sure that none of the walls are glass and open onto the car park) and extensively fogged with whatever that anti-theft fogging stuff is after a second break-in. The general consensus was that were it to be turned off 1. it wouldn't come back on, and 2. that would lead to something terrible happening to HR, possibly the non-payment of that months wages.
I did track down the HR bod who was supposed to own it, but they denied all knowledge and said it was something vital to do with the Accounts dept. Accounts said it wasn't theirs but it was critical to Distribution... who told me it was HR. I worked there for about 5 years in all, accompanied by that sad little 'Fweeep' noise. Someone had obviously loved it enough to stick it on its own UPS, but not enough to link it to any backup, or patching of the esoteric OS that was on it, or replicate this doomsday box anywhere. Company is gone now - assume someone turned that box off by mistake one day, but everytime my train goes through that business park on the way into London, I swear I can hear a faint 'Fweeep' on the wind.
Apologies for excess length.
...no longer manufactured or supported, of course, but still completely functional and fit for purpose. Think "spectrum analyzer, oscilloscope, etc". These things don't wear out and frequently last a decade or more in active use. If only the control/display software ran on the current version of Windows (development stopped, I think, with XP)
Example: Fluke multimeters with logging capability. Very handy items, EXCEPT for this one thing: their logging data is inaccessible unless you have a PC running Excel. That's right, you can't get the data off the device with just a serial transfer, you need their proprietary, licensed (yep, that'd be a network connection as well) application, as well as Excel. I'll let you guess what Corporate IT's response to that request was. A lab computer, generic user/password and connected to OUR network??? So, thanks, Fluke, your poor proprietary design has made the logging feature of your meters inaccessible. If only you had allowed export of the logging data as CSV, we'd still be using that feature.
Sadly, Fluke makes very good (if overpriced) multimeters.
This is just someone's clever idea that didn't quite work out the way they planned. (although they still got paid for it, so maybe it did?)
Their proprietary, licensed and not inexpensive app does the data download and calls Excel to display the graphs. "Annoyed" does not quite convey my reaction. We have not purchased any more meters with the logging "feature"
I wouldn't know about modern Fluke ... most of my go-to kit of that genre are indeed Fluke, but I bought them in the mid late '80s, and have felt no need to buy more since. Why bother, they still work fine, and electricity hasn't changed in the last 40 years.
I do know that I will not purchase anything that uses proprietary data formats unnecessarily, and I flat out refuse to purchase anything that won't function unless it can call home over the Internet for no reason other than because it's fashionable.
I'd be annoyed, too. Have a beer.
I was sent to a 12-person local brand of a financial company to assist in the email migration to O365.
Despite reminders of housekeeping weeks before our arrival, there are a few oldtimers who just ignored such emails.
Anyway, there was this old timer of almost 15+ years in the company who looked constipated when we came to her desk. She has a "don't touch my emails" look in her face, despite ours, her supervisor's, and the HoD's reassurance that she will not be let go if the email is missing.
The real reason is that she has accumulated over 30 GB of deleted items and a 50 GB inbox over the years. Keeping them there, just in case she needs to look up stuff. To backup her mailbox, it took almost a day. My colleague and I skipped lunch and dinner to ensure that nothing happened during the backup process. Gosh, finally home at 3 a.m.
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2024
