Desktop Browser Market Share (~%)
Google Chrome 70.6%
Microsoft Edge 11.2%
Mozilla Firefox 6.5%
Apple Safari 5.7%
Opera 2.0%
If your browser of choice isnt on the list you are l33t and **** and cool as a penguins ****
Maybe!
I use Brave which I guess is a Chromium browser, I refuse to use Chrome its a disease, Edge is marginally ok but is mostly used to download a better browser like Brave or Firefox. But I thought your previous reply was about the 2000's not now.
Trashbytes wrote on Today, 05:54:I use Brave which I guess is a Chromium browser, I refuse to use Chrome its a disease, Edge is marginally ok but is mostly used to download a better browser like Brave or Firefox. But I thought your previous reply was about the 2000's not now.
It was about browsers back then.
And I was much like you.
IE was a none starter, back then it was about Firefox and Opera.
ElectroSoldier wrote on Today, 06:06:It was about browsers back then. And I was much like you. IE was a none starter, back then it was about Firefox and Opera. […]
Trashbytes wrote on Today, 05:54:I use Brave which I guess is a Chromium browser, I refuse to use Chrome its a disease, Edge is marginally ok but is mostly used to download a better browser like Brave or Firefox. But I thought your previous reply was about the 2000's not now.
It was about browsers back then.
And I was much like you.
IE was a none starter, back then it was about Firefox and Opera.
Firefox was a slow starter for me, didn't really get usable till around 2007 when it finally decided to adopt the IE style of rendering web pages.
IE was really only ever used for schooling and Uni where you could use others but it wasn't guaranteed the web pages would render looking right unless you used the right version of IE. IE even cocked this up with its different versions...the web was certainly messy back then.
Still ..if I could go back to the 2000's internet . .I would, the modern web is ...so sterile and corporate.
Do you remember Geo Cities and Tripod? Pepper Ridge Farms does ! (I still remember Altavista, Lycos, Excite and Dogpile search engines which were all pre-Google, I should add Yahoo here as its also pre-google too)
AOL CD's that gave you free Internet time and if you collected a bunch of them you could just keep making fake accounts with each disc for free Internet.
Thats why I stay away from the browser wars because none of them were ever very good. Even good ol Netscape Navigator wasnt all, at least not as all that as people made it out to be.
Sure it was less, much less broken that IE but it still wasnt "fixed"
if the pages you viewed most worked on your browser of choice then it was thee browser of choice for you and all the others stink because they did.
Find the one that works for you and roll with it but that doesnt mean the others dont work great for other people and the sites they visit.
It is as true now as it was back then.
Yes I remember Geocities. It was at one point a good way to have you own page on the net.
I cant remember what happened to it. Went away like AventureMail, Myspace, MSNIM... Even IRC.
I think once all this AI "stuff" calms down and the hyperscalers take control of the market the web will change from what it is now to two versions.
A version that is AI friendly, so the AI bots can scrap and browse the sites data to present to their users as an answer and a version that must be viewed by the user.
I remember the AOL CDs. Never used any of them.
Here we had a phone service where you could nominate I think it was 2 numbers that were always free to call, so most people chose their ISP to get free internet.
Where I worked at the time we dealt a lot with ISDN2 and because I was working thats what I went with for about 2 years then where I lived at the time a company called Telewest started the UKs first broadband service in 2000 i think it was called blueyonder. Never went back after that. it started at 128k over ISDN then 512k over cable, up to 1mbit, 2mbit 4mbit... 10mbit and now its 2gig.
The main issue I have with AI search results is that its an Ouroboros situation, where the AI will eventually only be feeding upon AI search results and content.
I already don't trust AI search results and try my best to avoid using it but ...there is just so much AI rubbish polluting the results, even the stripped back non AI google search is polluted with it, finding anything has become a exhausting task of .. is this AI slop or is this an actual human generated web page with actual real world information on it.
If this is what's causing the ram shortages I really start to question the sanity of these AI bros.
ElectroSoldier wrote on Today, 05:32:Its not a case of a 30 cent CD, its all the work that went into making what was on it.
Obviously they develop ONE OS then they cripple down various versions to differentiate the price.
Or do you really think that client windows is limited to 20 SMB connection because at the 21th they explode? 😁
And whatever, when they sell millions of copies of a product, the R&D cost is so diluted that is practically next to zero. Each license sold ,by a big company like MS, is free money, with zero effort.
That is.
And yeah had they charged less they might have made more
They actually did. As usual unofficially. They considered OEM versions that used to cost 1/4 or less than retail (theoretically valid only if sold with a whole PC), first valid if billed together with a MB, then just an HDD, at the end a floppy disk drive was enough. And is not an urban legend it was confirmed by MS top managers, I heard that from the mouth of the CEO of MS in my country, during the presentation to IT professionals of Office XP.
You claimed early 2000s consumer chipsets supported more than 4Gb of RAM
So you invent what I say, just to make your point?
Where I said that? I said that support for more than 3GB by 32bit machines was a thing since NT4, and there wasn't any insurmountable technical challenge.
They just didn't bother to block consumers, because their business was protected by the RAM price, and because they haven't yet any x64 release to sell.
"Casually" they blocked the PAE thing in xp32, with SP2, just few months before the official release of xp64, and when x64 betas where already widespread in the wild.
That is the whole point. You have to actively hunt for the specific drivers that dont choke on high memory buffers.
There is nothing to "hunt", like I said ATI /AMD was smooth sailing, Nvidia disabled the PAE support after 333.11, Intel never released PAE friendly drivers for Vista/Seven and newer, but most of their XP/2003 drivers were OK and worked well in Vista/7 (at the price of aero glass effects).
Two row of info. It's all you need to know to use a PAE enabled 32bit OS, valid in 2000s, still valid today.
As for Blaster in 2003, millions of homes were hooked directly to USB
Oh captain obvious, I replied to you, who talked at the present "try to use an unpatched system", is pretty obvious we talked about today situation.
But even back then most of the famous threats impacted people with disabled firewalls, or running older releases that didn't have a firewall at all, the Administrative account by default did the rest.
"I know more than you and the browser I use proves it"
Yet another FACT that you try to dismiss as forum's boast.
Ask ANY Opera user how much viruses they caught back then, and educate yourself about the difference between technical facts and forum boasts
Trashbytes wrote on Today, 06:56:If this is what's causing the ram shortages I really start to question the sanity of these AI bros.
They don't care about sanity, they care about making money. And they ar making it, for now.
The sad part is that the AI bubble will eventually burst, but the damage they are doing will be persistent
The Solutor wrote on Today, 07:40:Trashbytes wrote on Today, 06:56:If this is what's causing the ram shortages I really start to question the sanity of these AI bros.
They don't care about sanity, they care about making money. And they ar making it, for now.
The sad part is that the AI bubble will eventually burst, but the damage they are doing will be persistent
Well a few of them are . .mostly nVidia, the scalers might be making some too but nothing like they were promised, the AI Bros are making squat, if anything they are burning through billions they don't have. But I'm sure nVidia will keep propping them up.
The AI bubble bursting wont be anything like the Dot Com Bubble, itll be more of a whimper with a slow drawn out deflation.
Trashbytes wrote on Today, 07:46:Well a few of them are . .mostly nVidia
There is a lot of direct and indirect money made thanks to the AI bubble, starting from the RAM price, which is the topic of this thread, ending in electrical utilities that can rise the prices with the excuse of datacenters, to stay silent about politics, an argument I don't even want to start.
A not so small minority of the humankind making money at the expense of the majority, a definition that may be applicable to traditional capitalism, but that now reached the next level.
The Solutor wrote on Today, 08:01:Trashbytes wrote on Today, 07:46:Well a few of them are . .mostly nVidia
There is a lot of direct and indirect money made thanks to the AI bubble, starting from the RAM price, which is the topic of this thread, ending in electrical utilities that can rise the prices with the excuse of datacenters, to stay silent about politics, an argument I don't even want to start.
A not so small minority of the humankind making money at the expense of the majority, a definition that may be applicable to traditional capitalism, but that now reached the next level.
The proliferation of Data Centers is very worrying, they are taking land used for food production and turning it into Data Centers, destroying both power infrastructure and water supplies all so they can have a bit more compute.
We are all going to pay for that in the long term.
Trashbytes wrote on Today, 08:21:The proliferation of Data Centers is very worrying, they are taking land used for food production and turning it into Data Centers, destroying both power infrastructure and water supplies all so they can have a bit more compute.
And that's as big as it is, will be nothing if compared to the social damage, we have already seen the damages done by the social network, with the spread of populsm and fascism, with the downgrade of general culture, with the average QI of new generations diminishing for the first time in centuries compared to the previous generations.
If social networks did this in little more than a decade, what AI can do in the close future?
We are all going to pay for that in the long term.
For sure.
A guy named Carl Marx used to say that capitalism needs to constantly grow to stay afloat, and when growing slows down, new needs has to be invented (which worked for most part of the industrial revolution, with cars, appliances, TVs, PCs, Smartphones...), but when growing proves impossible the usual way is to do a reset, which means an extensive war.
Looks like recently we are trying do do both things, with new invented needs like AI, and with the current international situation.
I hope a third way will materialize, but I'm not optimist at all.
The Solutor wrote on Today, 08:58:A guy named Carl Marx used to say that capitalism needs to constantly grow to stay afloat, and when growing slows down, new needs has to be invented (which worked for most part of the industrial revolution, with cars, appliances, TVs, PCs, Smartphones...), but when growing proves impossible the usual way is to do a reset, which means an extensive war.
Looks like recently we are trying do do both things, with new invented needs like AI, and with the current international situation.
I hope a third way will materialize, but I'm not optimist at all.
There is a kinda "third" way but its the one no one will especially like - the west came closest in the 20th century, a kind of regulated compromise oriented welfare granting mixed market property rights economy with democracy, checks and balances, rule of law, anti trust, unions, separation of powers, civil society and freedom of speech/press. a kind of pluralism without one dominant ideology, it would have been true for USA and Scandinavian/Canada at the same time, different tunings of the same mix. That mix means no one is getting exactly what they want, its always compromising - that leads to a tug of interest groups trying to pull society towards one or other dominant leaning.
It looks a bit like there is instability now, pendulum swinging wildly right/left (or whatever passes for that now), division and ideology, disruptive technology in there too with huge effects on people and culture. No idea where it leads in next 20-30 years, but it doesn't look good
Meanwhile, no one will get the RAM they want either..
spiroyster wrote on 2026-09-14, 18:57:Windows XP 32-bit limits processes to 2GB unless built with laa (Large Address Aware, which is a compiler flag), in which (as an app) you can access 3GB, no more. If you want/need more within a single process, you have to use file memory mapping, not RAM (which is kinda application version of using a swap file). Swapping, RAM mapping (address translation) etc is all handled by the OS since we have something called protected memory, and have done for a number of decades now.
Well, specially on this forum, you should probably mention Address Windowing Extension, basically like DOS' EMS/XMS for Win32 😜 . it does allow allocating and addressing more than 2^32 from a single process, and it is neither file mapping nor swapping (if anything it is the opposite, as AWE-allocated memory cannot be swapped out!).
The Solutor wrote on Yesterday, 21:11:Those are mostly retrofit technical explanations to justify a nasty commercial practice. Wrong in a number of ways, purposely wrong given Chen is anything but an incompetent.
I do agree, capping client 32-bit Windows at <4GB is purely arbitrary. Exactly the same story as removing 16-bit support from 64-bit windows, or all these CPUs from Windows 11. It is all arbitrary, any technical reasons you can point are mostly irrelevant when such limits can be lifted trivially even by third parties.
E.g. if the argument is 3rd party driver compatibility, then watch how they did it recently with VBS, or how did they it did with PAE/ the NX bit itself in the same time period!
They offer the toggle to users. They have a denylist/allowlist of drivers that are known to be incompatible/compatible.
They could have simply done the same to allow >4GB on client 32-bit Windows.
So it is just a way to push users to the direction they want. Whether this direction is a good thing or not, your mileage may vary.
Now all irrelevant with the death of 32-bit OSes anyway...
gerry wrote on Today, 12:17:There is a kinda "third" way but its the one no one will especially like
Well that worked, pretty well here for almost half century, our beloved constitution wasn't written by Americans, like it happened in other places, we had the fortune that the country was freed only partly by the American troupes, the other half of the job was done by antifascist partisans, hence the constitution tuned out pretty balanced.
But that lasted until we started to ape Americans, and their naive model of two parties only, and also the social conquests had to be stepped down just because the competition from countries of that once was called third world. You can't pay a worker 1000 euro, if there is a competitor that pay 300 euro for the same job, at the same time you can't blame people from China, India, Vietnam and so on for trying to have a better life.
In short there isn't any obvious solution to the problem, unless the entire world was regulated by a super partes authority, an authority with enough power to impose the regulation to any country, from Mozambico to the USA. And we know that this isn't going to happen anytime soon, sadly Star Trek is just SF.
The Solutor wrote on Today, 07:23:ElectroSoldier wrote on Today, 05:32:Obviously they develop ONE OS then they cripple down various versions to differentiate the price. […]
Obviously they develop ONE OS then they cripple down various versions to differentiate the price.
Or do you really think that client windows is limited to 20 SMB connection because at the 21th they explode? 😁
And whatever, when they sell millions of copies of a product, the R&D cost is so diluted that is practically next to zero. Each license sold ,by a big company like MS, is free money, with zero effort.
That is.
And yeah had they charged less they might have made more
They actually did. As usual unofficially. They considered OEM versions that used to cost 1/4 or less than retail (theoretically valid only if sold with a whole PC), first valid if billed together with a MB, then just an HDD, at the end a floppy disk drive was enough. And is not an urban legend it was confirmed by MS top managers, I heard that from the mouth of the CEO of MS in my country, during the presentation to IT professionals of Office XP.
Nobody is talking about the inbound SMB file sharing license limit. That has always been commercial tiering. CALs were sold as licencing packs.
You are talking about something they sell as a product! I would have expected a hexagenerian to know about that and not cite it as part of crippling the OS.
If you actually want to talk about network limits tied to stability and security, look at what Microsoft did in XP SP2: they capped outbound half open TCP connections to 10 in tcpip.sys (Event 4226). Why? Because rampant worms and early P2P clients were generating thousands of simultaneous embryonic SYN requests, exhausting kernel non-paged pool memory and melting cheap consumer modems/routers. That was an engineering safety cap, just like the 4 GB HAL clamp.
The Solutor wrote on Today, 07:23:That is what you said, with a quote from me in the middle of it. […]
That is what you said, with a quote from me in the middle of it.
You said many chipsets that support 4Gb or more.
Name them!
I wont even confine you to the chipsets that would be used by these small and medium size businesses, that would like to have used a consume OS as a server OS...
The scope is even wider than that, lets start in the NT4 era...You cant, because they didnt exist!
Intel 440FX: Capped at 1 GB.
Intel 440LX: Capped at 1 GB (512 MB SDRAM).
Intel 440BX: The legendary enthusiast chipset was hard capped at 768 MB unbuffered / 1 GB registered.
Intel 440GX: High-end dual Xeon workstation chipset, capped at 2 GB.The only chipsets from the NT4 era that addressed 4Gb were the Intel 450GX (Orion) and 450NX which were massive 4way/8way enterprise server architectures running on machines like the Compaq ProLiant 6000 and costing tens of thousands of dollars. The exact "enterprise mambojambo" you claimed small businesses never needed was literally the only hardware on earth that could physically take that much RAM in the NT4 era.
I think you only know the NT4 era from retro computing and then mostly in retro VMs because that is the only place you can get those old machines to support that much RAM because the hardware back then sure didnt!
Nvidia driver 333.11 was released in May 2014
A month after Windows XP reached End of Life, and ten full years after XP SP2 released in 2004.
Citing a 2014 Quadro/Tesla enterprise driver branch to argue how things worked in the early 2000s proves the exact point I made .You are retrofitting hindsight and cherry picking modern/late era drivers on modern platforms, while ignoring the complete lack of over 4Gb consumer hardware and consumer driver validation that existed when SP2 was actually written in 2004.You mentioned 333.11 earlier, I dismissed it then because I thought it was a typo because it actually proves my point in more ways than one.
If a user in 2004 had to wait until 2014 to get an Nvidia driver branch that reliably tolerated high memory 32bit PAE addressing without crashing, that literally proves consumer drivers in the XP era were not ready, not tested, and not stable for high address PAE when SP2 was authored.
333.11 was an Enterprise Quadro/Tesla ODE Branch, Not Consumer XP
Look at what 333.11 actually was
It was an "Optimal Driver for Enterprise" (ODE) release for Quadro, NVS, GRID, and Tesla hardware (R331 branch).
Its release notes explicitly state "Discontinued support for Windows XP."
It was built for 32bit Windows 7, Vista, and Server 2008.
To run it on a consumer card under an uncapped 32bit PAE OS, you have to use an enterprise workstation driver branch from 2014, manually INF mod it to accept a consumer GeForce device ID, and run it on modern hardware.
This directly confirms what I said earlier: stable 4Gb and above DMA execution lived on enterprise certified driver codebases, not the consumer GeForce drivers that shipped on the CD with a GeForce 4 or FX 5200 in 2003.Again this comes from a retro computing lens, not as we lived it back in the early 2000s
Your idea is that driver crashes above 4 GB were a myth invented by Microsoft to sell Server licenses, and that drivers always worked.
Well your own example of 333.11 proves you wrong.
With that one example you have to land on a hyper specific, curated, enterprise derived driver branch from 2014 to stop the machine from falling over.So you have ZERO technicals to back up your idea about microsoft dumbing down the OS to force small/medium size businesses into buying server OSs.
The only crippling of the OS, in SMB, was actually part of the Microsoft billing structure! Client Access Licences. And even then that comes up as a request in the OS as to you having one, how many its for and only comes up in server OS and then only as part of a compliance review! So they would come to your business and inspect your licences, which aint going to happen!And then we come to the cherry on top
MSBlast
You apply a solution from now to a problem back then...
I mean you know I dont know where to take that one.
You say you mention NAT and firewall protections, but none of them remove the problem with an unpatched system they just take it out of harms way, which is kinda beside the point. Microsoft needed to fix the problem not paper over it or mask it behind a wall of protection.
Thats the difference between being competent and professional and getting by with guile.Its exactly like forcing WinXP onto a modern motherboard with an unsupported ACPI table (anything beyond ACPI 2.0). You can slipstream patches, bypass the 0x000000A5 stop error, and trick the installer into declaring the machine an "ACPI Multiprocessor PC." But the NT5.1 kernels ACPI driver still doesnt have the hooks or logic to properly parse power states, sleep vectors, or modern interrupt routing defined a decade later. Device Manager might look clean, and the system might boot, but you havent actually solved the architectural incompatibility youve just masked it with a hack. And a dirty hack at that.
Hiding a broken stack behind a router or masking an ACPI mismatch with an installer patch might satisfy an enthusiast tinkering on a test bench, but it doesnt change the underlying technical reality of the platform.
Which is what Im talking about.
javispedro1 wrote on Today, 13:14:Exactly the same story as removing 16-bit support from 64-bit windows
Well that's a little more complicate story, it has to do with AMD that made a mistake in switching context between various CPU modes, and MS that took this an excuse to simplify Windows asking AMD to not fix the bug.
While I don't remember all the details, look like the problem ins't fixable, even using a kernel driver given the Windows scheduler would need a major rewrite to properly handle 16/32/64 bit at the same time. I discussed this with someone really expert on such low level things few years ago, the problem is to find the (very interesting) discussion.
Now all irrelevant with the death of 32-bit OSes anyway...
One of the most stupid move ever, given 32 bit means smaller (mostly faster) installation, even in current days. 32 bit only CPUs are mostly extinct, but modern servers that run dozen of virtual machines are a current thing. Long story short is that to run the same number of 32 bit VMs that can run on 10 servers, you need 16 servers to do the same job.
1.6x storage, 1.6x RAM, 1.6x Electrical power and so on...
@electrosoldiers
Seriously learn to quote properly. This way is impossible to discuss with you.
The Solutor wrote on Today, 13:24:gerry wrote on Today, 12:17:There is a kinda "third" way but its the one no one will especially like
But that lasted until we started to ape Americans, and their naive model of two parties only, and also the social conquests had to be stepped down just because the competition from countries of that once was called third world. You can't pay a worker 1000 euro, if there is a competitor that pay 300 euro for the same job, at the same time you can't blame people from China, India, Vietnam and so on for trying to have a better life.
I think you have that backwards. If one company is paying 1000 euro and another is only paying 300 , people will flock to the company paying more. I thin you mean that a company wont pay worked 1000 euro when there are workers who will take the job for only 300.
Which happens in places like canada when you have an influx of foreign workers willing to take a lot less and so companies fire everyone and hire lower wage workers.
or you have government programs where they pay half the wage for a foreign worker but not for a native born worker. so companies hire only foreign "temporary " workers.
Also quoting Marx is hilarious as his ideas are complete failures.
The Solutor wrote on Today, 13:54:Well that's a little more complicate story, it has to do with AMD that made a mistake in switching context between various CPU modes, and MS that took this an excuse to simplify Windows asking AMD to not fix the bug.
While I don't remember all the details, look like the problem ins't fixable, even using a kernel driver given the Windows scheduler would need a major rewrite to properly handle 16/32/64 bit at the same time. I discussed this with someone really expert on such low level things few years ago, the problem is to find the (very interesting) discussion.
This is also an excuse pretty much like the 4GB limit issues . There are two "reasons" that I have heard:
- The CPU change: amd64 in "long mode" cannot have a 16-bit code segment. But you can simply switch back from long mode temporarily. I was using a linux kernel module that did exactly that at the time (and not a "rewrite"), so it was obviously a workable solution. (Plus, emulation...)
- They wanted to make USER/GDI handles longer and thus this would break 16-bit applications or at the very least sharing and passing data between 16-bit and 32/64-bit applications. Well, there are various 3rd party NTVDM , so you can go and see for yourself how much of a problem creating this 16-bit handles abstraction was. So also workable.
I am not saying that the challenges do not exist. They definitely do. Drivers crashing the system because assuming addresses are <4GB were (and are!) definitely a thing. But 3rd parties worked around them, so for MS, they would have been trivial to fix. However, they used them as excuse to push their customers (to newer machines/software).
