Techie told 'Bill Gates' Excel is rubbish – and the Microsoft boss had it fixed in 48 hours The Register knows that tech support is a vocation that induces frustration, which is why each Friday we offer a new edition of On Call – the reader-contributed column that details real-life support stories so you can at least enjoy misery in company. This week, meet a reader we'll Regomize as "Brad" who told us about his …
It actually goes back to limits of the 8086 architecture that weren't resolved until the x86_64 chips were developed is one of the main reasons why x86 chips are considered inefficient in comparison with other architectures.. Why IBM chose Intel's at the time inferior chips over Motorola or others is just one of those things, but, at the time, the decision makers were not engineers but probably keen golfers!
"Why IBM chose Intel's at the time inferior chips over Motorola or others is just one of those things, but, at the time, the decision makers were not engineers but probably keen golfers!"
According to this article by a chap who was there at the time, it wasn't a bunch of "keen golfers" doing the decisions.
So why aren’t we all using 68K-based computers today?
The answer comes back to being first to market. Intel’s 8088 may have been imperfect but at least it was ready, whereas the Motorola 68K was not. And IBM’s thorough component qualification process required that a manufacturer offer up thousands of “production released” samples of any new part so that IBM could perform life tests and other characterizations. IBM had hundreds of engineers doing quality assurance, but component qualifications take time. In the first half of 1978, Intel already had production-released samples of the 8088. By the end of 1978, Motorola’s 68K was still not quite ready for production release.
And unfortunately for Motorola, the Boca Raton group wanted to bring its new IBM PC to market as quickly as possible. So they had only two fully qualified 16-bit microprocessors to choose from. In a competition between two imperfect chips, Intel’s chip was less imperfect than TI’s.
The whole article is a delightful read.
@Sandtitz - Good read. My dad brought home a Ti 99/4 back in the early 80s and was the first home computer I remember using (had previously brought home a Commodore Pet but as that was a work's machine I wasn't allowed to touch it). Had the option to "boot" into an equation calculator or basic when you first turned it on and whilst supposedly compatible with the later Ti99/4a (no equation calculator and chiclet keys replaced with a "proper" keyboard) I could never get parsec (Ti99/4a cartridge) to run on it. Ti-invaders and Blasto were great games and I still remember typing OLDCS1 to load from cassette. The extended basic cartridge really improved the in-built basic and it took some clever routines to get around the hardware limitations. The amount of hardware add-ons available in the states (though sadly not over here in the UK) could turn the machine into a monster.
Without knowing Parsec specifically, the TI99/4a uses a TMS9918a video processor whereas the TI99/4 uses the earlier TMS9918, which lacks the most sophisticated of the graphics modes — the one that doesn't require any real-time trickery to support a full bitmap mode and offers 8x1 attributes rather than 8x8.
Thanks for the link, but IBM's corporate culture was already sclerotic by then, even if it didn't reach it's full degree of uselessness for another decade. The deal with Microsoft for the OS was another example of rushing in where's others fear to tread.
At least the deal with Intel was covered by the dual-supplier contract which gave AMD a leg up. I shudder to think how things would have developed if Intel had not been forced to compete on what became the dominant platform!
If AMD had not been around then someone else would have got there but not until Intel made masses of cash. There wouldn't have been enough incentive for Intel to compete and so they would have become complacent and someone else (potentially even ARM) would have beaten them by the late 90s I think.
Incidentally, can someone ensure this happens to nVidia sometime soon please?
The second supplier contract was given to AMD, Siemens, NEC, Harris, and many others. It was actually standard practice at the time.
What I do not get is Why IBM did not go with the 8086 in the first place. Yes, in principle making a systme with an '88 is cheaper, when the system is small, say, the controls for a traffic light. But when the system is a full fledged general purpose PC, the difference is negligible...
Well, a wasted opportunity.
Aonther thing I do not get (and another wasted opportunity), is that Intel told IBM of the impending 87186. If intel were wise, they would have used in the OG XT chips compatible with what they knew would end up inside the 187, even if not buying them from intel (rememeber, second supplier was not something you forced the chipmaker to do, it was par for the course at that era).
This meant that the PCjr and the klatter X86s were more expensive than the 80186 and 80188 clones of the era.
The original PC uses 1-bit RAM chips, as did almost everything of the era. So the cost of a 16-bit bus would have been that the cheapest model had twice as many RAM chips as if it had an 8-bit bus. That’s not a negligible cost difference, and not an easy one to explain the motivation behind if it led to negative comparisons with the Apple II — then, very temporarily, one of the business machines to beat thanks to VisiCalc.
Not just that the price of RAM was high and would need twice as many chips, it was so expensive the 5150 originally came with only 64KB of RAM. Expanding it up even to the full 640KB was prohibitively expensive except for the most well off institutions or government departments. And IBM had no idea of how fast things were going to move in the future. They expected to get something in the order of 10 years from the 5150/PC/XT range before needing to bring out a new and better model. I think they told buyers it would have 10 years of support in the early days. Ceratinly that was the case with the AT/286 range, and why Compaq beat them to the first 386 model.
I don't think that the availability of the CPU was an issue here.68000 were available in production numbers since late '79. By 1980, I used 68k based systems from Sage and Cromemco and even a pre-series model of the HP-9816 desktop computer...
What made a (cost) difference was that with the 8088, it was possible to use a large number of existing (and thus cheap) 8 bit peripheral chips, while offering up to 10x the amount of RAM than the then prevailing 8 bit CP/M systems and micros like the Apple II, Tandy TRS-80 and Commodore CBM 3000/4000/8000 were capable of. For more of a side show project at IBM, that seemed good enough and with the subsequent evolution of 80286/80386/486 CPUs by Intel, any initial advantage that using an 68k chip quickly vanished, even considering that Motorola very quickly dropped the price of the 68k chips and IBM could have easily gotten a similar sweetheart deal as Apple got for using the CPU in the Lisa and then Macintosh series...
And at the same time as IBM in Boca Raton developed the IBM PC, another IBM department was developing the IBM 9000 lab/instrumentation/process control system, based on the Motorola 68000, trying to compete with IBM and DEC in that market.
I'm pretty sure the 640kb limit was first resolved by the 80286's almost-unused 16-bit protected mode, in which segment registers first mutated into selectors, loading bases and offsets via descriptors from a local or global table; it offered a logical address space of 1gb from a physical pool of up to 16mb of RAM, though each segment was still at most 64kb in length in part because all pointer registers were still 16 bits.
Then the 80386's 32-bit protected mode offers flat addressing of up to 4gb of memory.
> Jake wrote: You do realize that Gates&Co. had nothing to do with the 640K "limit", right? That so-called "limit" was part of the IBM hardware spec...
> Charlie Clark wrote: It actually goes back to limits of the 8086 architecture that weren't resolved until the x86_64 chips were developed...
Jake is right, it was a "feature" of IBM PCs. Back in 1987 I was using Concurrent CP/M on non-IBM compatible, 8086-based machines and they had about 960KB free out of the 1MB fitted.
The maximum memory supported by Concurrent CP/M on an IBM PC is only 544KB according to http://bitsavers.trailing-edge.com/pdf/digitalResearch/concurrent/Concurrent_CPM-86_Users_Guide_Aug82.pdf
My original 5150 runs 704K, with the "extra" memory coming from space allocated for my nonexistent EGA card. I know a couple people who added another 64K to that, but I had no real need. Hacking around with memory & hardware was a normal part of the personal computing world back in the day. Modifying and/or fooling the BIOS was/is trivial. Remember, IBM published the extremely detailed " IBM PC Technical Reference Manual " (36 bucks), making these kinds of hacks fairly easy. Having access to the brains at The Homebrew Computer Club probably colo(u)rs my memories of how easy this stuff actually was ... or wasn't.
Note that games often barfed on the various mods, but most of the important business software of the day ran just fine with the extra memory. Also note that DOS would happily use as much contiguous conventional memory (the RAM between the LMA and the first populated portion of the UMA) as the system reported. DOS itself had no 640K limit baked into it.
Later still, memory tricks allowed up to about 950K. I find it absolutely amazing that the "640K DOS limit" piece of incorrect trivia is still being parroted as fact after all these years ...
On the other hand, I personally remember Steve Jobs saying that "128K ought to be enough for home users", at a meeting of the Homebrew Computer Club in late 1983, as he was demonstrating the original 128K Mac, just before the public unveiling. At the time, he had a point ... people were running flight simulators in 64K!
And of course there were cards like Tall Tree JRAM that could take the PC up to 2 megs ... but that didn't come out until the PC had been on the market for 8 or 9 months. I rode my bike over to Elwell Court in Palo Alto to get mine direct from Tall Tree ... which I only remember because the "shortcut" alongside Adobe Creek under Hwy 101 flooded out due to high tide and I had to take the long way home, over San Antonio Road. Took a while to convince myself to fully populate that board ...DRAM was spendy!
Yeah. IBM defined the memory map, but didn't enforce it. And you could move things like the mapped BIOS around as long as everything you ran called it properly rather than just jumping to hard-coded addresses.
But as you say, RAM was expensive. We had a 5150 at home but never got it past 512KB because that was enough for our purposes and it just wasn't worth spending more money on it.
And I agree re the Technical Reference; that was a great manual. I learned a lot about 8088 assembly from the BIOS listing, and wrote some hacks that made use of it. Most of the IBM PC manuals were good. IIRC, the Advanced BASIC manual had an example for every keyword, which was very useful.
The 5150 had some fun hardware hacks, too. The cassette port was good for some hacky I/O — I built Steve Ciarcia's lightpen, which used the cassette port. And we had a CGA connected to a monochrome monitor, so we eventually soldered some resistors to it to change the burst (NTSC color) signal to different amplitude levels, giving us greyscale (well, green-scale).
Jake is right, it was a "feature" of IBM PCs.
Correct. The 8088 had 20 address lines and could address 1MB of memory. (That "20 address lines" bit came back to bite everyone in the ass with the 80286 and the infamous real-mode-doesn't-force-A20-low A20 line issue.)
It was IBM who decided to limit RAM at 640KB and make the upper memory area portion of the memory map reserved for BIOS and other mappings. That had nothing to do with the 8088.
The reason was simple: The original IBM PC was a gap filler with an intended 5 year lifespan with explicit design constraint of using standard parts
8088 was chosen because it was available (68k wasn't), it was a COTS part (which ruled out the IBM RISC cpu) and it was better than the competing TMS processor
Let's not forget these originally shipped with 64kB of ram and were primarily used as serial terminals with a basic word processor
Even a ESP32 microcontroller would have trouble booting up in only 64kB
Many Many years ago when the RS/6000 and AIX were not so old I discovered a bug in the C compiler. I spent many hours isolating the lines of code that caused the bug and then sent an example of the code to IBM so that they could understand what was happening and, hopefully, correct the bug. I received a polite acknowledgement receipt of the information and acknoledgement of the fault and then a few weeks later a significant bill for the time they had taken to check the information. I also received a new copy of the corrected compiler (eventually) and a new bill for the making and sending of the correction. We had a support contract with IBM but apparently correcting bugs that we discovered was not in there. We didn't pay it of course. A chat with our account manager and it was soon solved.
I ran into an issue with a load balancer appliance a couple of years ago. The details escape me right now, but there was something weird happening with the packets for a new service that I was adding to the cluster - they were corrupted or inadvertently dropped or something similar. The vendor TAC engineer requested a call to troubleshoot the issue online, and the engineer was so intrigued that he went and got another colleague to help out while we were still on the call. They were talking amongst themselves - without muting themselves so that I could listen in to their reasoning, props for that - but couldn't find an immediate explanation. They eventually said that "Listen, this problem will most likely go away if we reboot the box. But we'd prefer NOT to do that, but instead dig a bit deeper into this. Are you OK with that?" I said sure, and they brought in a development engineer and another couple of friends of his - so at one point there were five or six people looking at my box at the same time.
After a couple of hours, they had developed a theory that there was some obscure race condition in their logic causing the issue. Within two days I had a patched version of the software in my account and we never saw the issue again after that. I've never seen that level of support from another vendor; before or since. And credit where credit is due: It was Radware.
I'd agree with you in general, but it's much loved by millions of users, so it's not going anywhere fast. But, since the 2015 rewrite using .NET, it has got a lot more powerful, which is both a blessing and a curse. It's also pretty good for report generation: I happen to know that quite a few very large companies rely on it to produce the reports they send to customers. It's not without it's problems, but in this area, it's probably better than a custom solution.
I've come across far too many corrupt databases to ever want to use Access for anything serious and that's even before trying to work with its form of SQL, which makes it pretty much unusable for anything else. Sqlite would generally be the better choice in any similar situation. But Pandas and Juypter notebook were already available and widely used in science and finance. The tech world wasn't helping with its various attempts at data grabbing disguised as solutions for unknown problems.
Access gets too much bad press. I'd never use it now for the sort of production systems I work on, but back in the day, when I created modest-sized databases for modest-sized teams, it was great. Nothing came close in terms of development ease and speed. I used it for a few years, but in just a few weeks, I went from total novice to having a decent professional understanding of relational BDs. It was a great training platform in prep for working on a larger scale, on critical stuff.
For several years, my go-to solution was VB6 front end and Access DB, and customers were always happy. I've not used it for ages, but the vast majority of corruption issues were caused by sub-optimal programming and configuration/maintenance, e.g., incorrect use of opp locking, sketchy networks, shared folder configs, not running compact/repairs, and unwanted server-side file caching and disk write issues - like the dreaded 'delayed write error' (shudder). But these were easily fixed. Granted, a proper DB engine would cope with all that, but Access was a fraction of the price and extremely accessible for both customers and less experienced techies (like me).
I have a real fondness and nostalgia for Access, which I admit is largely unwarranted. However, it's hard to forget your first love.
Up until 2017 we were still using Access as a backend for the entire company's timesheet system, 1700 employees would use a simple website every month to log their time and it all went into an Access DB. Every so often it would get corrupted but it was so simple it took about 15 mins to fix. Finally replaced by a full stack system that the HR dept paid about £40k for and it didn't do half the things the naff old Access DB based app could do, the new system wouldn't allow half day holidays to be booked, the company that sold it had never come across any customer that needed half day holiday bookings!!! They took about 3 weeks to recode it after we threatened to sue them for selling a product that promised XYZ and never delivered. It took 3 years to catch up to the original Access app!
Originally a company I worked with was using SuperCalc with CPM everywhere, when Excel appeared they all moved to the Windows operating system.
Excel had problems that everyone worked around, that's how everyone starts using new apps and a new operating system. So Excel was very effective for Microsoft.
I'm not so sure. I've lost count of the number of times I, and others in the office, suddenly had Excel fail to open a file that had been fine the previous time.
My solution was to open them with Open/LibreOffice & just save them in the appropriate Excel format without doing anything else.
This worked *every* time and nobody ever lost data by my doing it.
I see this a lot with Word docs as well. Passing a document file among collaborators in different establishments, especially education where no one has any up to date software, is bound to end in failure at some stage. Sooner rather than later if you are using tracked changes. At least everyone has moved on from the land of .doc to .docx. And yes, the 100% reliable fix is to read and rewrite using LibreOffice.
God help you if your Word document has equations in it. There is a 50% chance that LO will mangle them and a 25% change that it will screw up the rest of the document, just for fun.
I had to borrow a Windows laptop from $FORMER_WORKPLACE solely to avoid this problem.
While I mostly agree, part of the problem is that Excel has too many features that make people think it's something it's not.
With ${day_job} hat on, I see lots of "databases" done in Excel that really, REALLY, should have been done with a database. When you end up doing lots of data verification manually because, for example, there are multiple values in a cell (e.g., in a network, a switch has two uplinks, so the "parent" cell contains "device_name_1[return]device_name_2") which means it's impossible to simply run queries against stuff. Another (related) one I had to work with had A end and B end of links, and a separate cell containing an (I assume manually generated) list of nodes involved from A to B - and of course, no easy way to link to that "database" from others where the links were used, and no easy way to pull up views such as "what are the fibres in cable X used for" (sorted by fibre rather than something else).
Had I been involved in the early days I'd have been going to the IT people and demanding we had free access to a decent database - which would need to combine the ease of use that the old Filemaker had, with the flexibility of SQL behind the scenes. I'd have built a system where the user saw much the same thing via a view that gave the same front end layout as their Excel sheet, but which allowed all the data validation and error checking any such application really needs to have. For example, in the above example, the Excel sheet has now way to see, other than manual validation, if you have used the same fibre in cable X for two different routes.
Plenty of colleagues spend their lives replacing Excel solutions developed by business users with more polished / corporate solutions. Though many times the Excel resurfaces. Its not an usually an Excel problem its normally a lack of decent development methodology.
For volume I used to use SQL server or Access (back when Excel choked on 1M lines.) but Excel is open all day on my machine.
What better alternatives are there?
Surely that should be "Pedant's?"
No. It should be Pedants'. The Pedant's Handbook is a book for one pedant. The Pedants' Handbook is a book for pedants (plural).
Since hardly anyone here uses apostrophes correctly - a crime against humanity IMO - a simpler and less controversial title would have been "The pedantry handbook" or "The handbook for pedants".
This pedant is now returning to the endless quest to make sure "its" and "it's" get used correctly.
If it really were a book published as "The Pedants Handbook", then that is how it should be written- adding an apostrophe anywhere would be incorrect. See many discussions on the correct naming of Presidents Day, and other similarly named public holidays in the US. Anyway, everyone should be saving up their apostrophe pedantry for the official day, August 15th.
This post has been deleted by its author
Oh, must I hang my head in shame? Will they ever let me back beyond the pale to tow the line down at The Old Panda and Gun?
But I also think the grocers' apostrophe is generally observed in its absence titles and there's some even obscure rule as to why this is so: it is not the handbook that belongs to any single peasant or group and therefore not a genitive. Some languages (Swedish and I think French) have forms that make this clear in the phrases like "he puts on his hat". You needed to know that, didn't you? And that on a Friday!
Cheers!
The boss had a degree in psychology rather than technology, and so was better at managing people than technical issues.
"worthy
noun [ C ] humorous
UK /ˈwɜː.ði/ US /ˈwɝː.ði/
a person who is important, especially in a small town:
The front row of chairs was reserved for local worthies."
https://dictionary.cambridge.org/dictionary/english/worthy#google_vignette
To be fair, I've only ever heard it in the plural.
"My boss gave me a lot of leeway so long as I kept the PCs running," Brad told On Call – perhaps because that worthy had a degree in Psychology, not tech.
I grok'd this as Brad's having the misfortune of psychology degree and that his manglement boss unusually cut Brad a fair bit of slack - the blind leading the vision impaired as it were. Brad probably worked cheaper too.
On reading the alternative parse, I admit that it is more probable.
I was wondering what the serfs in a retail chain would be using Excel for. I am guessing it was just to send a "machine readable" structured report to headquarters (and I presume before decent POS systems and integration with back office reporting.) I would have thought a custom screen form to xml application would be less prone to problems than using a spreadsheet.
If the part of that sentence after the hyphen had been put between double quotes, it would have been clearer it was Brad talking, most probably not about himself. Now it looks as if the author is speaking about Brad, since how is he to know Brad's boss? So you need to read it twice to understand.
I spent months a couple of years ago trying to implement a new system of "back end access" for a major NHS supplier a couple of years ago to the point where I felt I had been working for them the whole time in testing and development . They finally scrapped the whole thing, then went away and came back last month with a whole shiny new plug and play "it just works" system , the quality of which is yet to be verified.
First impressions are its a lot better though.
So frustrating was that prospect that Brad sent "a quite angry email" to billg@microsoft.com – assuming that was the best way to contact Microsoft founder Bill Gates. "I expressed how upset I was at this obvious bug and essentially accused Bill of beta testing his software on the public," Brad told On Call.
In other words, "Brad" discovered the SatNav approach to software releases waaaay back then!
I shocked, shocked to learn that Micros~1 development practices have lasted that long.
While it is funny to read, you actually CAN get the newest with quite some fixes included. Backside: Thing you may not want are in there too. The ultra-newest is reserved for MS-employees.
I've used "billg@microsoft.com" on any site that asked for an email address but didn't bother verifying it, for probably decades at this point. So it must receive an absolute shitload of spam.
Sorry Bill, I always assumed it wasn't a real address.
That email address is arguably one of the most famous in the world.
There also used to be jeff@amazon.com, but since ol' Bezos no longer runs Amazon, I think it's just a catch-all for the executive customer care team now. Emailing Andy Jassy works (but it's not andy@) too.
And yes, managing to get the email address of the top dog is a great tool to get something done, but more important is that you should continue to use decent language when you email them. Inevitably it's read by either one or a group of executive assistants, and by emailing with decent language, explaining the problem, why it's got you absolutely fuming, how you expect them to fix it and how to get in touch with you, you are most likely going to get a favourable response.
Whether it's Bill Gates, Jeff Bezos (a few times over the almost 3 decades I've been an Amazon customer), the CEO of RSA Insurance (ending a years-long saga)... if you're nice, the problem can be solved. If you're being an utter dick, less so.
:-)
It's always the social engineering side of things that gets neglected. Back when I was an PFY we were having issues with a number of PCs (I forget what it was) and I'd tried to get help via the website. In the end I tried the customer services number and got a little bit of info, so rang back and used that to get more. After a few calls of this social privilege escalation I acquired an office number and rang that instead of CS. But it was a voicemail which then gave me the mobile number for the big boss.
I thought about this for all of 30 seconds before dialling, it turned out he was on holiday and gave a return to work date. A few days after he returned from holiday I called as the problem still wasn't resolved, and we ended up chatting about his holiday before getting down to resolving our PC issue. Magically an engineer was waiting for our office to open the next morning.
The entire thing was funny in a "how did that even work?" kind of way. But it did scare my boss what I could do starting with a tiny nugget of information.
Once I had a cable TV installation crew break a transom window in my flat (the guy got shocked and flew from the ladder into the window). They said they would get in touch to fix it, but I tried contacting the company several times only to being told over and over that they would "look into it".
I got fed up and one morning faxed the company's CEO complaining about it and had a (his?) secretary call me that same afternoon scheduling a glazier's visit the next day.
I've probably mentioned this before. I needed a SPARC architecture manual but that existed nowhere on Sun's (before Oracle) website. In frustration I emailed Scott McNealy, pointing out that Intel put all their tech documents online. A week or so later I got a copy of the manual in the post. It does sometimes work talking to the Big Boss
1.72 MB was typical. Cram more sectors in one track. If you interleaved the sectors and used you could go up to 1.85 MB, but not all disk drives could handle them. 1.72 MB was the "lowest common" of those which all drives, and DOS, can read. https://duckduckgo.com/?q=1.72+MB+disk&ia=web
They are not that difficult to copy, but you cannot format those with Windows (maybe you can now, but I cannot test since no floppy... not even USB floppy...)
This post has been deleted by its author
A thing that happened to me around 2000. I worked for a municipal government. At the time, departments had been purchasing their own PCs, but it was decided to consolidate them.
At the same time, we would move from (a bunch of Windows versions) to Windows NT on the client machines, Token Ring to Ethernet, and Banyan Vines servers to Windows NT.
Lots of work to get the images together and test versus all the software.
Rolled it out.
Turns out the audio driver would corrupt a Microsoft Access file if sounds were produced while it was open. Much work ensued to remove the audio drivers until a fix could be found. Months later.
Considering all the changes involved, it's amazing that it went as smoothly as it did. That was the only major problem after all.
Excel will always be my least favourite spreadsheet application because of one seriously stupid design flaw that's been around forever, specifically the fact you can't open two separate spreadsheets that share the same name at the same time when every other spreadsheet application can. Why it doesn't allow for distinguishing two files based on file path and not name, I dunno.
I once tweeted MS about it and got a cheeky "are you using Windows or Mac?" response back, when Excel for Windows has had this issue since '95.
Running a:\Setup.exe /a <drive>:\<directory> copied and exploded the contents of the disks to a destination (usually you’d choose a mapped network drive) from where you could run setup.exe at the client PC and never load a disk.
Installed in ‘admin’ mode. And with an optional answers file if I recall, one of my first pc support roles.
The backup software Duplicity, which is the default software for Ubuntu (Deja-dup is just a GUI for the CLI Duplicity) has a quite helpful mailing list. I've asked for help a couple of times, and always had good success. One of the most common responders is Kenneth Loafman, who wrote a lot of it and has been maintaining it for years. His responses are nearly always polite and helpful. (The one time I saw he wasn't entirely polite was when someone essentially called him incompetent and ignorant of how software works; his surprisingly-restrained reply was entirely justified.)
Not support as such, but I found that going to top works. I bought a game a few years ago when it was on budget rerelease. The key didn't work despite it being a shrinkwrapped dvd from the retailer. When I contacted support they basically accused me of piracy, even though I supplied pictures of the case, key and proof of purchase.
The publisher was a UK PLC, so their directors are listed at companies house. The support guy had pissed me off, so fired off an email guessing the MDs address making my opinion very clear.
Shortly after the same support person got back to me with a grovelling apology and a new key.
What a load of bullshit. We all know Windows has a lot of problem features... and im sure there are thousnads of people who complain daily to Bill about one thing that pisses them off.
Typical example of yet another article from American media trying to brainwash the public how wonderful corporate leaders are. Of course its all Bill's will , no mention of the team who actually put the time and effort to fix the bug.
"Brad frantically conducted more tests and found the cause: a spreadsheet created on a PC could be read on a Mac, but once the Mac opened it, Windows users would see only gobbledygook."
Not gobbledygook but opening a complex (especially if it includes macros) Excel spreadsheet created on a PC, on a Mac, is still anything but perfect. Word and Powerpoint are fine. Not Excel!
Many moons ago I was an 'individual contributor' at a very large PC company and did some early experiments on a PC variation Microsoft (and Bill, who was still in charge) were keen on at the time. Wrote it up into a little presentation (TLDR: not ready yet) for some tech managers and a VP who I supposed had not much else to do at the time. As usual someone asked for a copy of said presentation which was duly emailed over around 4pm and thought not much of it. Came back the next morning to a reply from the VP-- he'd sent the presentation to the CEO who forwarded it to Bill Gates around midnight, who then replied back to the CEO around 4am and the CEO bounced it back to the VP before 6. The next week a large contingent of Microsoft people descended on us festooned with PowerPoint replies to said presentation.
Bill could light fires under people if he wanted to.
Apple was also very good - but only if they wanted! - to handle mails directed to the executives. One of more than one positive experience:
When they were first time shipping 27 iMac they experienced broken screens due to lacking experience in shipping that large pieces of glass. When a younger me sent an email to Steve Jobs pleading to not shipping me a broken iMac I had waited so long for, I did get an reply but an UPS update: Shipment halted for inspection.
Maybe this was a coincidence but I had further positive exchanges with Apple Executive Relations when having sent an email to an executive to fix a real problem.
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2024
