Modern browsers on legacy PC's (Split thread from PSU hall of shame)

[Heihachi_73]

Re: Modern browsers on legacy PC's (Split thread from PSU hall of shame)

I am guesstimating that XP probably has 1-2 years more of use on the web with Firefox.... or maybe 3 if we are lucky.

The only way Windows XP will be killed off from the net is if the HTML standard itself is scrapped entirely for something brand new and completely incompatible with anything older, while 32-bit x86 programs are discontinued at the same time.

[RJARRRPCGP]

Re: Modern browsers on legacy PC's (Split thread from PSU hall of shame)

but even 1.5GB is still enough for simple browsing. Can be used by old persons to talk with their children via Skype...

Even a good while ago, back in 2013, IIRC, Windows was struggling with Skype, with 2 GB of RAM...

Much less dare to use major web applications on a 64-bit browser with < 5 GB of RAM! Especially Chrome! You can expect the dreaded "Aw, Snap!" error message (or similar)...

[momaka]

Re: Modern browsers on legacy PC's (Split thread from PSU hall of shame)

Been playing with that Pale Moon for a while, many bloated sites like my webmail or fleebay seem to run almost normally in that. That's while running Opera 12.18 simultaneously for other websites all on that 3@3.3GHz 478 Prescott

Yup, that's the way to do it, as I've mentioned before. I do the same, but with Opera and Firefox (actually mostly Firefox nowadays). As long as you have at least 1 GB of RAM on a XP machine (or 2 GB on 7), you'll be fine.

The only way Windows XP will be killed off from the net is if the HTML standard itself is scrapped entirely for something brand new and completely incompatible with anything older, while 32-bit x86 programs are discontinued at the same time.

I suppose there is some truth to that. But eventually new programs won't be made in 32-bit, and so you will be forced to use only legacy stuff. When that happens to the browsers, it won't be too long before the old browsers become too obsolete (given how often browsers are updated nowadays).

[junktv]

Re: Modern browsers on legacy PC's (Split thread from PSU hall of shame)

Old ff up to 24 many sites don't work for these days.

[eccerr0r]

Re: Modern browsers on legacy PC's (Split thread from PSU hall of shame)

If M$ didn't patch the SMB bug that WCRY was exploiting, would XP die faster?

[Behemot]

Re: Modern browsers on legacy PC's (Split thread from PSU hall of shame)

I've just run into an unofficial fork of Pale Moon which has the latest stuff for XP/2003, New Moon. So far found only portable build, but to hell with it, it works!

[stj]

Re: Modern browsers on legacy PC's (Split thread from PSU hall of shame)

so share a link!

[Behemot]

Re: Modern browsers on legacy PC's (Split thread from PSU hall of shame)

Try search engine?

Building own build thread https://forum.palemoon.org/viewtopic.php?f=5&t=16241
portable version https://www.sendspace.com/file/z54hl0 from K-Meleon forums (http://kmeleonbrowser.org/forum/read...,142494,142581)
Even found installer for 27.5.0 https://mega.nz/#!0OJAiCoL!gMHDffJ8H...FHx7FlDZCUXLXM

[GeorgeBlitz]

Re: Modern browsers on legacy PC's (Split thread from PSU hall of shame)

Windows performs badly on an i7 and no, it doesn't get better as you subtract hardware.

I was using a browser on an old i386 box running SCO Unix, X11R5 and Konky last week. It works fine. CNN's JavaScript-overload main page loads in a couple of seconds. I checked Weather.com. It was a pig, but it's a pig on anything. That's a 386. 16MHz. 8MB of RAM.

Use Firefox. Install NoScript and AdBlock Plus. When a page says "You're Using AdBlock, Please Disable It", just say "Good Call! No thanks!" and close that site because they suck: they want to play an ad on your browser that may very well install a virus because morons like MSN don't check those ads for malware. Not quite as bad on Win 10, but still really REALLY bad.

Windows is like if Scarlett Johansson lost her clothes and passed out drunk in an alley in downtown Chicago: violation is inevitable, and it doesn't do much besides look good.

[goontron]

Re: Modern browsers on legacy PC's (Split thread from PSU hall of shame)

^ Preaching to choir i see?

I do miss the SunOS days some times....

[retiredcaps]

Re: Modern browsers on legacy PC's (Split thread from PSU hall of shame)

Windows performs badly on an i7 and no, it doesn't get better as you subtract hardware.

I just finished building an used Windows 10 machine for someone. I found general browsing in Windows 10, latest edition, with chrome, ublock origin on an Intel E8500, 3GB DRAM acceptable.

On my "test" machine, it has the same E8500 configuration, but I run Lubuntu and I find for my purposes that is acceptable even in 2017.

So my expectation is that i7 should probably run way faster than a E8500?

When a page says "You're Using AdBlock, Please Disable It", just say "Good Call! No thanks!" and close that site because they suck: they want to play an ad on your browser that may very well install a virus because morons like MSN don't check those ads for malware.

You also have to worry about crypto miners now stealing your CPU cycles without consent. I have that blocked on my router running LEDE software and ublock origin.

https://arstechnica.com/information-...ryptocurrency/

For sites that detect my ublock origin and display that disable adblock message, I just never visit that site again. There is very FEW original content today. Most sites are rehashing or simply cut and paste of material stolen from somewhere else.

[GeorgeBlitz]

Re: Modern browsers on legacy PC's (Split thread from PSU hall of shame)

So my expectation is that i7 should probably run way faster than a E8500?

Moore's Law is over w/ regard to PC CPUs. A Rizen chip will buy very little noticeable performance over an old rusty 2700K Sandybridge. 50% tops, and that on stuff no one does.

The biggest bang for buck these days are the PCIe FLASH SSDs. Nothing will work with mechanical drives any more except backups. Why? Because you start your 40K .EXE program, it pulls in 400,000 200MB .DLLs spread all over the disk drive. A mech drive spends the next 30s to 2m seeking and waiting for the correct sector to stumble under the read head. Since that wastes so much memory, part of that has to be paged to disk (more seeking, waiting).

Microsoft took the concept behind shared libraries and decided that 'if shared libraries work in Usage Case A, just imagine if everything uses them!'. Forget that shared libraries only make sense if multiple programs will use the same library often enough to offset the performance hit.

Problem is, every OS retailer gives their dev crew the latest/greatest hardware, 10Gb Ethernet, and 802.11ac WiFi. They code for that, and they only optimize once performance suffers noticeably on an overpowered machine. Well, if performance suffers noticeably on a 6-core, 12-thread, 4GHz box, then just imagine what it will behave like on a 1.8GHz Socket-P laptop...

Microsoft: put me in charge of Windows development. I'll give my dev crew 486DXes with 4MB RAM and a 300 baud acoustic coupler running PPP. My crew will write software that makes every user smile every time they start their computer.

I'd do that, too, but I'm too comfortable where I am.

[Behemot]

Re: Modern browsers on legacy PC's (Split thread from PSU hall of shame)

76.2939453125 TB of data to start a 40KB .exe? Seeking for two MINUTES? Oh come on, tell me about hyperbolizing?

[GeorgeBlitz]

Re: Modern browsers on legacy PC's (Split thread from PSU hall of shame)

76.2939453125 TB of data to start a 40KB .exe? Seeking for two MINUTES? Oh come on, tell me about hyperbolizing?

For the quadrillionth time, I do NOT hyperbolize.

[Behemot]

Re: Modern browsers on legacy PC's (Split thread from PSU hall of shame)

Oh, I see. Either one of those crazy "alternative" ppl, or some math genius with his own mathematics theory (so pretty much a crazy alternative person ).

[GeorgeBlitz]

Re: Modern browsers on legacy PC's (Split thread from PSU hall of shame)

Oddly related to OP: If anyone does mobile apps on the side, avoid those 'We Work On Android OR iOS' crap tools like Telerik.

If you have an app that seems ridiculously slow for what it's supposed to do, it's probably a Telerik or other JS-based app.

JavaScript has ruined everything. Why do I need to download, then cache, the entire Dojo v2.0.3.2.1.14.64.1 toolkit, just so a web site can display a pie chart of "Major Causes Of Web Performance Issues"?

Learn to use Chrome's builtin "Developer Tools". It will show you exactly why that page sucks roadkill mandrill butt.

[goontron]

Re: Modern browsers on legacy PC's (Split thread from PSU hall of shame)

^ You do realize the Dalvik VM Android uses IS essentially Java, right?


The DVM uses a register-based architecture that requires fewer and more complex instructions. Dalvik programs are written in Java using the AAPI (Android application programming interface), compiled to Java bytecode, and converted to Dalvik instructions as necessary.

[GeorgeBlitz]

Re: Modern browsers on legacy PC's (Split thread from PSU hall of shame)

^ You do realize the Dalvik VM Android uses IS essentially Java, right?


The DVM uses a register-based architecture that requires fewer and more complex instructions. Dalvik programs are written in Java using the AAPI (Android application programming interface), compiled to Java bytecode, and converted to Dalvik instructions as necessary.

Uh huh. Java and JavaScript are two completely different things. Their only commonality is that they're both interpreted (byte compile requires an interpreter - the DVM).

So, you have byte-compiled Java, and a VM using that to outright interpret JavaScript. That's a guaranteed pig, no testing required. Don't believe me? Write a loop timer in JS that performs enough matrix multiplication to be quantifiable on a given HW device. Measure it accurately, to the microsecond. Write the same loop for Java and measure it (not bad!). Now, write the same loop for NDK and try to measure it. Can you guess why there's such a massive difference? That's what I'm on about.

JavaScript is great for prototyping. It's necessary for all those completely unnecessary transition effects that "responsive web sites" torture us with, and necessary for some interactive charting or graphic activities. But it's a pig. It will remain a pig until someone builds a computer with a pipelining JS coprocessor with dedicated RAM on a dedicated bus. I'm not going to hold my breath awaiting that marvel.

Just because a 4GHz processor makes a given task seem Really Fast does not mean that it is, in fact, not a pig.

[Behemot]

Re: Modern browsers on legacy PC's (Split thread from PSU hall of shame)

JS is good for forms and other stuff. Think it was primarily intended for that some time ago.

On the other hand, Java is pretty poor beast as well. If phone CPUs did not have Java instructions for acceleration for some time now, it would still be unusable.

[goontron]

Re: Modern browsers on legacy PC's (Split thread from PSU hall of shame)

^ Yup. Thats why im glad Android moved to ART. ELF FTW!