The Daily WTF: Curious Perversions in Information Technology

2006-04-04 Reply Admin

.... workstation is remotely administered and most local settings are reset to company whenever I login... talk about corporate totalitarianism....
---------------------------------------------------------------

I am absolutely sure I typed "are reset to company defaults whenever I login", where did the word 'defaults' dissapear?? stupid forum software

GoatCheez · 2006-04-04 Reply Admin

Anonymous:
GoatCheez:
Databases like we have now were in their infancy back then that's for sure. Oracle might've had something, but not sure, as I really didn't/do much database stuff

A lot of people have made mention of Oracle, Paradox, MS SQL Server, Sybase and others but perhaps GoatCheez was referring to FOSS solutions. PostgreSQL only became mature recently and MySQL is still a work in progress (although v5 brings it considerably closer) to the big boys.

Knowing that this company didn't seem to want to make capital investments, licenses for one of the proprietary RDBMS (and the server that hosts on it) probably didn't seem like a Good Idea.

Yes, I was... not sure why so many people had to comment on this, as i even admit that i didn't and don't do much of any database work...

2006-04-04 Reply Admin

Anonymous:
<snip lots="" of="" remarks="" on="" france="" and="" french="" people="">I'm sorry you had to end up in a big company with employees that look like they work for the government.
But I can assure you that in a reasonnably sized company (in fact, usually the smaller the better, I think), people acutally WORK for a living, instead of waiting for the paycheck to come.

And not everyone in France hates foreigners, far from it.

But I agree with you, some people in France should think about how they would do if they were suddenly paid for the real amount of work they provide... (french administration employees mainly).

I'm sorry you feel that way about our country anyway.
</snip>

My sincere apologies to all French readers I have offended with my ridiculously overgeneralized blanket statement.

During my field trip through the French corporate catacombs I have met many great Frenchies who didn't hesitate a bit to help me out when the help was most needed during the lean times. These were mostly (but not always) people who spent some time abroad and have felt themselves on their own hide what does it looks like to be a stranger in a strange land.

The problems I was whining about were mostly caused by office politics, which is ugly, dirty and underhanded whatever the country. I'd say the French corporate culture is a bit more rigid, compartmentalized and hierarchical than what I expected, and it can at times be a cold, cruel place for someone with subpar French language skills.

But the main reason for my moving on that outweighs all other is the general economic climate. The rigid work laws and stagnating economy have apparently made all innovative IT work flee the country in panic, and with it left many bright, foreign language-speaking French IT guys and girls. Those who tend to remain in France are mostly the people who are too lazy (or proud?) to learn English, or simply don't want to leave their homecountry, which unfortunately further marginalizes French IT community and isolates it from the English-speaking world.

2006-04-04 Reply Admin

Anonymous:
Anonymous:
Anonymous:
CSV files sent in by others are basicaly unusable because my Excel uses semicolon as separator and comma is reserved for decimal point.

WTF? They localized a de-facto standardized file format?

(Ok, so I'm not seeing the WTF in the original story---lots of ISO-standard stupidity, but pretty common everywhere.)

Actually, the CSV "format" used by Excel is directly dependent upon your regional settings. If you choose semicolon as the system-wide "list separator", that's how fields will be separated. Idem for the choice of decimal point symbol.

If you export a table from Excel as CSV, then change list separator from default "," to ";" (Control Panel -> Regional Settings), the same file can no longer be read by Excel. You have to manually search/replace semicolons, colons and commas to correspond to wha's defined in your current locale.

Of course, nothing prevents me from resetting locale settings from French back to standard, but my workstation is remotely administered and most local settings are reset to company whenever I login... talk about corporate totalitarianism

I have this problem as well. Do you by any chance live in Denmark or has other countries "," as floating point separator?
Anyway the trick is to change the floating point separator from "," to "." (like US) in excel. Off course I cant say if this is reset at login. This way all CSV formats works and the only punishment is to look at Pi as 3.14

This CAPTCHA really is broken!

2006-04-04 Reply Admin

Anonymous:

I have this problem as well. Do you by any chance live in Denmark or has other countries "," as floating point separator?
Anyway the trick is to change the floating point separator from "," to "." (like US) in excel. Off course I cant say if this is reset at login. This way all CSV formats works and the only punishment is to look at Pi as 3.14

This CAPTCHA really is broken!

I think comma (,) is used as decimal symbol all over continental Europe. Additional trouble is caused by the thousands separator, in US thousands are separated by comma, in France by space, and in some other countries like Spain by dot (.) ... and then there are date/time values, for which each and every country, county and city block seem to have their own format incompatible with everything else.

I have to validate and store into the central database data files that come from all EU countries, plus some from overseas... the number of problems that arise from something as innocent as reading dates, strings and numbers from flat files is simply astonishing.

2006-04-04 Reply Admin

Anonymous:
Anonymous:

I have this problem as well. Do you by any chance live in Denmark or has other countries "," as floating point separator?
Anyway the trick is to change the floating point separator from "," to "." (like US) in excel. Off course I cant say if this is reset at login. This way all CSV formats works and the only punishment is to look at Pi as 3.14

This CAPTCHA really is broken!

I think comma (,) is used as decimal symbol all over continental Europe. Additional trouble is caused by the thousands separator, in US thousands are separated by comma, in France by space, and in some other countries like Spain by dot (.) ... and then there are date/time values, for which each and every country, county and city block seem to have their own format incompatible with everything else.

I have to validate and store into the central database data files that come from all EU countries, plus some from overseas... the number of problems that arise from something as innocent as reading dates, strings and numbers from flat files is simply astonishing.

I suspect that , is actually metric-system standard and . is MS posing stupid US standards over our heads.
There is a vital difference between the floating point separator and the rest: It is a part of the data. The rest is formatting.
And as we all know: formatting should not be stored in the data!!!!! So if anyone sends you files with these problems, you should really tell them to send you unformatted data. And I dont think excel exports the formatting either (its MS, who knows for sure)

CAPTCHA simply refuses to accept first time????

2006-04-05 Reply Admin

Anonymous:

I have this problem as well. Do you by any chance live in Denmark or has other countries "," as floating point separator?

All of Europe uses the comma as decimal separator, and afaik a significant part of the rest of the world.

2006-04-05 Reply Admin

Anonymous:
All of Europe uses the comma as decimal separator, and afaik a significant part of the rest of the world.

This is false. The UK is part of Europe, despite what Daily Mail readers may claim, yet uses the full-stop ('.') as a decimal seperator, and the comma (',') as a thousands seperator.

Stoffel · 2006-04-05 Reply Admin

marvin_rabbit:
ferrengi:

Wait a minute...
You were using a "new " windows 3.1 machine with 13 MEG of RAM during the dot bomb era?!?!
What year was this? I thought the dot com era can be traced back to about 1998. This doesn't make any sense.

A friend of mine went to DC and accepted a job as a programmer at a 'quasi-governmental' organization. (Home loans... one of the '-Mae's.) This was around early 1999 or so. He was given a brand new Dell desktop computer to use... installed with [wait for it!] Windows 3.11.

Plus, they had a policy that he couldn't get Windows '95 until he had a training course on it. And the next course wasn't going to be for about 4 months. His response was, "Well you better give me a course on 3.11, because I haven't used it in 3 years!"

One of the most distressing things, is that I reckon that they couldn't buy machines with Windows 3.11 installed on them anymore. This HAD to have been delivered with Win 98 or at least 95! So somebody had the job of uninstalling it and doing an install of 3.11.

Exactly my first thought--people who don't find this credible have never worked in a cheap private shop or a normal public shop.

I did an internship where I took a digital design position at a company that didn't want to spring for a VHDL compiler for the FPGA code they needed, and besides nobody knew it anyway. My project was to design an FPGA hooked up between a PIC microcontroller, ethernet controller and RAM. The FPGA itself had to route the packets from the ethernet controller into the RAM, the PIC told it where to put the packets. All done in schematic capture.

vasco · 2006-04-06 Reply Admin

GoatCheez:
Well, lets see here... I got my Pentium 90 back in like... oh god... what was it.... pre 95 at least... was top of the line with an ATI Mach64 2MB PCI video card and a 800MB hard drive. Came with Windows 3.1. It had 8MB of RAM... This story does not make much sense. To have the machines they had would've meant that they were searching around landfills for their new pc parts.... umm.... yeah... Windows95 was out when this happened, so it must've been at least '96. Databases like we have now were in their infancy back then that's for sure. Oracle might've had something, but not sure, as I really didn't/do much database stuff... hmm.... the 13 megs though is really really odd... All memory back then was done in pairs, so what pops into my mind in that configuration is 4+4,2+2,.5+.5 in an 8 slot motherboard. It's much more likely that 12MB was being misread as 13MB due to the whole powers of two thinggy, although math doesn't back that up.
Something smells fishy....

I guess it was a 640 + 384 Kb on the motherboard plus 3 4 Mb modules...
Could it be a Compaq Deskpro 386 ? 486 ?

vasco · 2006-04-06 Reply Admin

RJTECH:
Databases like we have now were in their infancy back then that's for sure. Oracle might've had something, but not sure, as I really didn't/do much database stuff... hmm....

Not sure how to read this, but be aware that when the web went live, databases like we use today existed for more than 20 years, and that is not what I call infancy in IT! OK, there is more possible today, but the database features I used in 1994 for danymic web content generation were not drastically different from what is used today. BTW, good old WAIS helped us indexing everything just fine...

A tree have very good search times (even a directory tree) but in all simple algorithms insert and delete are very expensive operations.

Balanced trees, AVL trees and all that crap you can find in a 1990 IT University course can look very good in a project design, but don't live up to what you could get with a honest to God 1970 COBOL program.

For this task you DON'T need a RDBMS, anything that have indexed files (being it COBOL, Berkeley DB or dBASE III+) will do.

If one have to use archaic techs it's a good idea to choose the right ones.<o:p></o:p>

vasco · 2006-04-06 Reply Admin

ammoQ:
Martin:

Also I seem to remember that win95 was not released in 95, but later.

Wrong. I'm sure I had Win95 in '95.

I also remember that many people where using 3.11/NT 3.51 a long time after win95 simply because ram was so expensive and win95 could not really run with less than 16 megs (many tried and went insane).

My notebook had 8 MB and it worked fine with Win95.

And now you're insane... QED.

ammoQ · 2006-04-06 Reply Admin

vasco:
ammoQ:
Martin:

Also I seem to remember that win95 was not released in 95, but later.

Wrong. I'm sure I had Win95 in '95.

I also remember that many people where using 3.11/NT 3.51 a long time after win95 simply because ram was so expensive and win95 could not really run with less than 16 megs (many tried and went insane).

My notebook had 8 MB and it worked fine with Win95.

And now you're insane... QED.

But not because of that. I was already insane because of Windows 3.1.

2006-04-07 Reply Admin

Deja vu time. Back about 95/96 I was working for a small science/engineering consultancy firm that was quite switched on, (lots of smart scientists and engineers) .

We had nice smooth internet access with Pegasus Mail running on a single Netware 3.12 server (running on a Pentium 75 with lots of RAM) . Email worked well & we had hacked up a system so that engineers working on site could email a spreadsheet lists hours. The hours and extras were automatically totaled, time allocated against project numbers, currency issues sorted automatically (we worked worldwide) and invoices generated & accounting system updated electronically. We were proud of the system.

The company then got taken over by a large enterprise. First week, they killed our server (which was running with ~100 days uptime) & ripped & replaced with a set of 3 Compaq servers running Windows NT 3.51 & MS Mail (POS if anybody remembers that) which cost ~20x more.

Since they didn't have a computer for every employee (but what seemed like a photocopier or printer for each body), the policy was to print every memo & have a secretary deliver a one sheet of paper to each of the 300+ employees rather than send out an email. It hurt to see a pack 500 sheets of paper used to announce that the CEO would be away on Wednesday & that Bob would look after the company until Thursday.

Coming back to the point.. (the time sheets)... the new company accountants were in the head office in another city... and didn't do email nor could they possibly use our smooth system which had been well proven. So the instructions for our engineers was for them to print out their spreadsheets and fax them into our office. The faxed items were then collated by a secretary on a Monday, who would then type them into the approved form which was an dBase app from what I remember. The newly typed forms would be printed out and faxed to the head office... who would then have a secretary type the information into the accountancy system. Needless to say, the head office system couldn't actually handle multi-currencies (or timezones) so the actual invoices were typed out by secretaries using information printed from the accounting system. We eventually found a clueful exec in the head office & managed to cut some of the electronic <--> paper transfers, but by the time I was asked to implement project charge codes (so they could account for & charge for each email sent) and shortly after they renamed all project directories (which had human readable names on our Novell system) to 8 digit codes, I left. Funny thing was that the company basically went under & was brought out by a larger American company.. and friends still working there told me that systems just got worse.

2006-04-10 Reply Admin

Nice post. While I've been lurking in the thread, post after post has had me thinking "these kids think everything in IT is new". An awful lot of stuff in IT isn't new at all. A surprising amount is so old, it was available to your grandfather.

And the new stuff isn't always better. Comparing typical desktop systems in the early 90's to today:

The graphics today beat the absolute crap out of anything we had back then, except on the most powerful, specialised systems.
Certain types of consumer grade hardware are much more reliable now. (There doesn't seem to be as much difference for server hardware, although I don't look at hardware mtbf numbers anymore.)
Installing new hardware was more complicated, but that was OK because a much higher percentage of users had a clue.
A surprising amount of stuff is actually pretty much the same, apart from the better graphics.
Some things today are significantly worse. To pick one random example, in the early 90's reading email with Pine over a SLIP connection to a T1 seemed way more responsive, and was definitely easier to manage efficiently, than reading email with Thunderbird over ADSl today. (And I'm not dissing Thunderbird, BTW.)
Yeah, we had databases back then. Even free ones. ndbm was available and was even built into Perl 4 from, oh, at least '93 or '92. Berkeley DB was and is free, don't know when it came out but I used it in '94. Ingres was open-source full RDBMS and nearly free, available since sometime in the 70's, being replaced by BSD-licensed Postgres in 1982, and ... anyway you get the point.
The XML spec came out in early '98, and expat was written the same year. Of course it took a couple of years before people went gaga over it, but by early 2000 your CV had to have a few X's in it.

Anyway, to get to the main point: apart from the badly chosen indexing algorithm which doesn't scale well for a problem of this size, and the occasional silly use of paper, nothing in the story is particularly WTF-worthy for, say, 1998 or thereabouts. (Which, incidentally, is when NetZero started offering free Internet access; they stopped in 2001). A few point by points:

Canada: no idea why they wanted hosting in Canada instead of UK, although a lot of European companies were very wary of US hosting due to differences in privacy laws. However transatlantic bandwidth was already acceptable in '98
Internet: in '98 it was still far from universal to give all employees HTTP internet access. It created an unnecessary security risk, it was a distraction, it cost money (lots more money than "good" protocols like email), and it wasn't yet obvious to management that it was a great research tool. (In fact, with a few exceptions like hardware specs and vendor knowledge bases, the value as a research tool was rather debatable.) Certainly the company I was working for in '98 had email and FTP access for everyone, but only gave HTTP access to users who could present a "business case"--and then they internal billed the relevant department for any bandwidth increase.
Develop offline, then upload from a special server: umm, that's what everyone does. All the professionals, anyway. Develop and test in your LAN, move to staging server for signoff testing, upload.
Human mediated indexing: The basic idea works. It was Yahoo!'s main index up to the end of 2002, and they still maintain it. You will find many people claiming it produces much better results than spiders. See also DMoz.
Internet PC--> Printout --> Type back in: OK, this bit is weird. Possibly it is meant as some sort of airgap for security? But even then, why not use a floppy to eliminate two layers of human handling? It seems a little ultraparanoid.
Create index at each insert: Since inserts are presumably much less common than queries (or else your business model is screwed), this is perfectly sensible.
New Windows 3.1 machine: A Windows 3.1 machine is not so surprising. As late as 2000 I saw a survey that estimated that some large percentage of government PCs were still Win 3.11 ! Certainly it was still supported in '98, so if you had a licensse, why not use it? Put it this way: Win 3.1 came out in '92, so using it in '98 was a bit like using Win 2K in 2006. Which a lot of people do.
13 MB: A lot of posters have already pointed out how this was both possible, and likely to surprise a semi-skilled user.
"Don't shut your computer down like that": The Boss might be getting persnicketty, but he was right. By '98 we are well past hard drives that needed manual "park" commands, but systems that could use rotational momentum to autopark themselves after an abrupt power cutoff (as they do today) were not universal. In other words, unless you knew exactly what model hard drive you had, there was a good chance that shutting down in that way would leave your drive in a state where a hard bump could destroy the read/write head. In addition, cycling the power fast has a small but non-zero chance of damaging a switch-mode power supply. (And finally, of course killing the power like that can screw up the file system if there are any file operations in progress. Those sorts of errors are usually recoverable though. Usually.)
200,000 files on remote server: he never explains what the web platform was, but since everything else was on windows, one wonders if the webserver was too? Older versions of windows could perform very badly indeed with very large numbers of files in a directory...
"Since they only updated one record at a time, a move script to update only fifteen items would result in over a million move commands": Okaaaay ... quite apart form reinventing the wheel by not using a database: they haven't heard of merge algorithms, invented back in the 1960's!? This does make them look a little foolish.

Richard Nixon · 2006-04-10 Reply Admin

Anonymous:
Nice post. While I've been lurking in the thread, post after post has had me thinking "these kids think everything in IT is new". An awful lot of stuff in IT isn't new at all. A surprising amount is so old, it was available to your grandfather.

Great way to start a post - very inciteful. That's a sure way to have a friendly discussion!

sincerely,
Richard Nixon

tofu · 2006-04-10 Reply Admin

GoatCheez:
'96. Databases like we have now were in their infancy back then that's for sure. Oracle might've had something, but not sure

ROFL! What are you, like 12? You need to get some perspective.

2006-04-10 Reply Admin

Richard Nixon:
Anonymous:
Nice post. While I've been lurking in the thread, post after post has had me thinking "these kids think everything in IT is new". An awful lot of stuff in IT isn't new at all. A surprising amount is so old, it was available to your grandfather.

Great way to start a post - very inciteful. That's a sure way to have a friendly discussion!

sincerely,
Richard Nixon

Huh? By complimenting the post, or by implying that I'm starting to feel old?

Richard Nixon · 2006-04-10 Reply Admin

Anonymous:
Richard Nixon:
Anonymous:
Nice post. While I've been lurking in the thread, post after post has had me thinking "these kids think everything in IT is new". An awful lot of stuff in IT isn't new at all. A surprising amount is so old, it was available to your grandfather.

Great way to start a post - very inciteful. That's a sure way to have a friendly discussion!

sincerely,
Richard Nixon

Huh? By complimenting the post, or by implying that I'm starting to feel old?

If you can't be bothered to register, I cannot verify that you are, in fact, the original poster to whom I was replying. Therefore, I will not continue this discussion. You figure out why I think the person I replied to is a jackball on your own. Good luck!

sincerely,
Richard Nixon

GoatCheez · 2006-04-11 Reply Admin

tofu:

ROFL! What are you, like 12? You need to get some perspective.

Again... wtf... I say that I didn't/don't do much database stuff, and yet you think that I think I know what I'm talking about when I comment on things like that, even though I admit that I don't know what I'm talking about. By saying the kinds of databases we have now were in their infancy back then, I was saying that a common object model for database queries wasn't supported across all databases that existed back then, and the FOSS solutions back then sure as hell didn't even come close to what the commercial offerings were. Bottom line: I am in idiot when it comes to databases, so just ignore what I say because I have no idea what I'm talking about, and even I know that.

2006-04-12 Reply Admin

Brillant

Search-dot-Bomb

Leave a comment on “Search-dot-Bomb”