David Haywood's Homepage
MAME work and other stuff
August 2, 2011 Haze Categories: General News. 33 Comments on Stupidity in Moderation

Give me one good reason why this helpful post I made was deleted?

Compiling MAME from scratch can take near on an hour. I suggest 2 lines to change it, and achieve the same in 5 seconds, thus making testing a LOT easier, and it gets deleted. That’s how paranoid the mods over there are over MAME and MESS being associated with each other, and the reality that they’re exactly the same thing.


Really, I’m 100% serious, that was deleted. It’s indefensible. I think China has more freedom than this, yet they claim it’s not political? That was an entirely legitimate way of helping. Next step they’ll probably make that impossible just to spite everybody. I’m trying my best to help MAME, even metaphorically sell it to new potential developers and users (a big part of the FME push below is for that purpose), but this kind of crap WILL drive legitimate contributors away.

This is also why forking would be impossible, it would be censored in the same ways, then declared a failure.


You can follow any responses to this entry through the RSS 2.0 feed.

I know what you mean. I also got some (IMHO helpful) posts of mine moderated away, yet other obvious trolls get away with their repeated stupid and offensive postings…

very sad indeed


Those of you with multicore CPUs should pass -jN to make, where N is (number of cores) + 1. Hyper-threading means you have (effectively) twice as many cores, though if you don’t have enough RAM, setting too high of a number may cause slowdowns! I had to upgrade to 8GB of RAM to reliably use -j9.
On my Core i7 Quad 2.2GHz (Sandy Bridge) machine, I do -j9 and get a clean build of MESS in 4:18 and MAME in 6:37
On my old Penryn Core 2 Duo 2.5 GHz, MESS took about 20 minutes with -j3 to cleanly build. Certainly not an hour, though if you’re using an old Pentium 4, chances are it will take that long ;)

I have a Core2 at around 3Ghz, and 8GB RAM.. Compiling takes ~45mins – 1 Hour, Debug builds with Symbols are worse still.

It seems to be an I/O limit, but I’ve defragged the drives to no avail.

The simple fact is I made a factual post, that would save a decent amount of time in checking something, and it was wiped off for purely political reasons, you couldn’t make this stuff up.

(and yes that’s using the -j options)

No, I’m not disputing that it’s slow for you (and yes I/O if often a major bottleneck) … but it seems many people do overlook the -j options.

As for moderation at the forums … I don’t want to comment on that in public.

I/O is often a major bottleneck

My Core i7 (Sandy Bridge) can do a full -j9 build with symbols in 2:30 (150 seconds, not minutes, of which the last 30 seconds are only linking mess.exe), and my memory usage never exceeds 4gb. So with 8gb you should be pretty well equipped. I agree it’s probably I/O in your case. Try creating a 1gb RAMdisk and building in there. Then it should be bound only by your CPU and you should drop into the <20 minutes area. See how fast you can get


A Ramdisk doesn’t exactly sound ideal for development…

I don’t really mind the build time being that long anyway, I’ve expressed in the past that I wish the build time encompassed both projects, because with a fully merged tree the increase is marginal, but that’s another issue altogether.

It’s only really annoying when you have to do a full rebuild, especially rebuilding 2 projects, which sadly is still a little too often but tolerable if I just stick with one.

The basis of my advice was simple, everything needed to test the behavior of mausuke with the latest saturn / STV code was already in the MESS build (it’s all shared), and thus didn’t require a MAME recompile (and likewise, if I’m testing Saturn, I don’t need a MESS recompile) and I was trying to save Kale a little time in recompiling MAME just to check something he could check much more quickly.

Megadrive / MegaCD progress has ground to a halt, not because I don’t want to work on it, but because the code was ripped out of MAME and I can’t quickly turn it on if I’m working in MAME to test / fix something anymore, I have to compile MESS, usually from scratch because I don’t touch it that often, but again, that’s a separate issue, and the powers that be have decided that it’s best that way, and again, any attempt to discuss why it isn’t gets deleted.

It’s up to them really, I’m trying to make things easier for people, and give advice to make things easier, they seemed determined to censor it all.

I’ll simply work on what’s convenient to me, and the FME stuff is a nice break, and chance to do something productive and interesting in the MAME framework / source with significantly less political crap to deal with (and yes, I say that with a straight face)

Basically it seems ‘Taboo’ to mention / highlight

a) Any situation whereby the current sharing of code / complete drivers can save a significant amount of time by allowing console titles to be run in MAME, or vice versa.

b) Any advantages which could be gained from merging the projects / drivers in any shape or form which would allow console titles to be run in MAME to aid developers or vice versa.

Basically talk about MAME stuff in MESS, or MESS stuff in MAME, for whatever reason, no matter the benefits and your post goes poof.

Anybody with half a grain of sense can see that’s utterly absurd.

Uh, I think you read a lot into that. Belmont just deletes anything he doesn’t like, for any reason, it doesn’t have to do with some mame/mess thing. Half my comments get deleted from the shout box.

Yeah, THAT’s the way to run a discussion board! Good going Belmont! /sarcasm

Rather than of mucking with the mess.lst file just throw hardware at the lengthy compile problem. Since the late 90’s hardware has gotten so cheap you don’t need to follow the old philosophy of the more money you spend now the longer the lifespan you will get out of your computer. Also you don’t even need to buy an expensive scsi drive versus an ide drive to get the best in disk performance.

A 2.8GHz Celeron from 2005 takes a little over 3 hours to compile MAME from scratch.


The year is 2011 not 2005 anymore and all parts of a computer (cpu, hard drive, memory, etc.) are x times faster for the same cost as 2005. If you want to develop or run cutting edge software then you need cutting edge hardware. This rule has existed since the dawn of the computer age.

I don’t have the budget to upgrade my PC every week, what’s ‘cheap’ to you might not be cheap to me.

That’s not the issue here anyway, the fact is I was pointing out a way that a compile time of whatever could be reduced to a few seconds to test exactly the same thing, and that post was deleted for no good reason.

BUT .. I’m not a fan of this ‘you must be cutting edge’ snobbery.

is it only me or 6 years != every week?

a 2yrs old 2.66GHz dual core duo macbook compiles MAME from scratch in less than 10 minutes. I guess my next desktop computer will be in that league even if I go for cheap parts…
2 cores + 1 or 2 GB of RAM it’s all you need to cut down dramatically compile time. beyond that you go for performance during emulation, and then it’s up to you how much you are willing to spend…

Ahh.. some attitudes will never change.

My machine is a 3ghz C2D with 8Gb ram, it takes ~45 mins to compile MAME.

This is in line with my other machines, which are of similar spec. I’m of the opinion that several of the compile times posted are more ‘my dick is bigger than yours’ fabricated times, or are absolute best case scenarios.


what bothers me are idiotic policies resulting in helpful information being deleted, and absolute censorship of everything which connects MAME and MESS and states that THEY ARE EXACTLY THE SAME THING and may as well be a single entity.

Anyway what I’ve been working on is a script to just pull both from the MAME/MESS gits and auto-merge it all, then put up a complete source / binary. Dunno if I’ll ever get round to finishing that, because really a more practical / sensible solution would be better, but sometimes you’re driven to such things.

Note, I have no problem with MAME having high CPU requirements for purely technical reasons.

What I have a problem with is your solutions always being so backwards.

It’s OK to censor and remove posts which suggest a way of saving a bit of time, because the correct solution would be to just buy a new machine??

That’s what’s being implied by some of these posts, and is quite frankly the most head screwed on backwards retarded answer that anybody could give.

I have always been of the mind that a dev should not have a cutting edge box. My work one is nearly 5 years old. These days anything after 2005 should be easily up to snuff to build mega projects in 10-20 mins tops. If it takes longer there is something structurally wrong with the project. Especially with dual/quad core cpus.

As a boss used to tell me if you cant time it you cant test it. If you cant test it you cant fix it. If you are changing something (or not) just because you do not want to time it and figure it out you are wasting time…

You have a way to take a 40 min build to 5 mins and have the numbers to back it up. To ignore it is just being silly. That is either ‘not invented here’ or they really just that myopic.

Well if you read the suggest it isn’t a ‘magic’ way.

MAME and MESS contain exactly the same code for Saturn and STV emulation.

Enabling a STV game in MESS is as simple as adding a single line to a file to enable the driver, and then testing there. Behavior etc. will be exactly the same as in a MAME compile, and saves compiling a MAME build from scratch if you don’t have one, or having to move the files across etc.

However, to even suggest that is forbidden.

As for the rest, merging the projects results in an overall reduction of compile time because at least half the codebase is the same between projects, and built twice if you’re building both projects. A single, slightly longer compile is better than 2 compiles which add up to significantly more. The alternative is of course that both exist as separate projects in the same distributed tree, with shared code being compiled only once, but again, attempts to discuss that get zapped.

It’s either paranoia over something, or pure political stupidity.

Also yes, I would agree with what’s said about not needing cutting edge but for other reasons too.

You need to keep a grasp on the USER experience. Just because you’re some top paid dev doesn’t mean you can go out and buy the top of the range machine, and produce something so badly coded, inefficient and bloated that it requires a machine of similar spec to work. With emulation there are good technical reasons for high requirements (although in some areas MAME is just taking the piss, see the recent performance problems with simple banking dragging a top end machine to it’s knees) However with regular software, designed to perform simple tasks more often than not the problem is the developer.

That’s what bugs me about some attitudes, it’s always made out to be the fault of the USER, not the actual person writing the application.

What’s been said here has similar vibes, as a user of the MAME/MESS framework to emulate hardware I’m seeing people told it’s their fault for not having a good enough machine, and that instead of allowing sensible suggestions to save a bit of time they are being told to upgrade their PC instead.

It’s just the same as the whole RTFM stuff when somebody can’t get something to work. Pass the blame to the user, over and over again instead of looking at actual flaws.

MAME does have some flaws in the build system, the dependencies are still such that you often need clean compiles, that is a flaw, I’ve never worked on a commercial project where I’ve had to do a rebuild from scratch as often as MAME! Maybe if that wasn’t the case, it would be more of a non-issue anyway, but even with the recently introduced ‘make depend’ (which was long overdue anyway) it still fails more often than not.

The facts however remain:
Building the project as one is quicker than building 2 projects due to the amount of shared code.
The projects are essentially the same thing due to the vast amount of common hardware, and often to test something as it would appear MAME/MESS side is just a case of adding a single line

Both these facts simply get censored by the current development teams.

To me MAME still restricting it’s domain to Arcade games is like a 16 year old kid still sucking on a dummy and wearing nappies, it should have outgrown that long ago. The current mamedev policy? more like an abusive parent telling them they have to keep doing it!

Please use your talents to help with an optimised MAME build over @ open pandora .. Would love to have a solid Mame build for the Pandora :)

I honestly blame the ‘buy a better machine’ problem on a comment Knuth made “premature optimization is the root of all evil”. People focus on that small bit and ignore the rest of his comment of timing it and looking to see what is wrong. You have identified it and proven it. At this point it is just chest thumping over how quickly they can throw you out.

I recommend a good book called Debugging by David J. Agans. I read the book and basically handed out copies to all my fellow programmers. It was like a step by step guide on how to fix and identify problems. What a good programmer/debugger just does by reflex others need to be told. This book does a really good job at it. Many times people are breaking step 3 of ‘stop thinking and look’. Why? it is fun to speculate what is wrong. It also lets you get things out of your hair quickly without having to really do any work. Blame the user is a goto excuse. But many times if the dev just shut their mouth for 2 seconds and listened to what is going on they would see ‘hey something is not right here’ OR they could help someone out. Since they blame the user “buy better hardware” they break step 9 “If you didnt fix it it aint fixed”.

Part of growing up as a developer is loosing the attitude that your code is the best ever. That is probably not true. A seasoned developer will look back on code they did 2-5 years ago and think ‘what was I doing’. If you are saying that about what you did 2-5 years ago what will you say 2-5 years from now?

Part of also growing up as a developer is accepting criticism from people you do not want to hear from. Just blanking out Hazes comments is just silly. It is something I expect out of a teenager on his forum. Not grown adults.

Also the not merging the two projects together will eventually hurt both of them. Arcade is done. Oh there are still a couple coming out here and there. But basically they do not make those systems anymore. Those that do come out are basically glorified PC’s. There is still a largish body of games that could be emulated. But that eventually will all end up in MAME. Then what? At that point it is just bug churn (which is boring for most devs). Putting the two projects together will let MAME survive and continue to improve (just like adding all of those gambling games will good work on adding those). To ignore this fact is silly. To ignore it just because Haze came up with it is silly too. He just got to the conclusion none of you are seeing yet (or willing to see). I have seen this behavior in the project for a long time. It is like for every good feature the whole dev team needs to go thru the kubler ross stages of grief. Much like C++ before it and the huge macro system that was put into place to NOT have to do C++ (bargaining). Right now the merging is in the anger/denial stages.

A good development system (I have built a few myself) should be ready to rock when cool features like this show up. Instead of ‘go away that will destroy everything’. If it was a bad feature you guys should be able to convince Haze (and he is not the only one saying it at this point) that it is bad. You should be working with him to figure out how to stage the code (as I believe big bang releases are always bad ideas). Or at least some good reasons on why ‘not now’ and ‘here is how we can do it later’. You guys may actually be doing this (but given the way haze is reacting I would say no). But guess what no one knows outside of mamedev. Because all we get is the occasional post every 6 months to a year about some game that is being worked on and very little else or Haze getting ticked about posts being censored.

Keep up this attitude and instead of ending up with a fork or a better product you guys will end up with 2 dead projects. I have seen too many projects flame out due to the exact same things you guys are doing right now.

Re: Pandora builds
I can’t see any current version of MAME being fast on a mobile platform, there is far too much core code which mean performance isn’t going to match older versions, in the majority of drivers most of your overhead is just running MAME and the CPU emulation even before anything is displayed. As I said, even things like bankswitching are now so complex they manage to introduce slowdown, when at the most raw level all you need to do for the most basic of implementations is change a pointer.

Re: Gambling games.
I don’t think they will be as beneficial to MAME as the console / MESS stuff would have been. They’ll improve some basic device code for sure, probably a couple of CPUs, and hopefully the artwork / layout system but basically the range of tech used on them is tiny, and while they will do rather thorough tests of the components they do use it’s a drop in the ocean compared to home systems and brings a lot less to the table.

While the machines appeared in arcades, or locations where arcades used to be they’re less interesting than home systems, especially to a casual MAME user who is just going to see a bunch of flashy lights. The consoles and titles released are more interesting, and extend the arcade legacy; it’s interesting to see how games got ported, what sacrifices were made, and in some cases how things actually ended up going the other way (starting out on consoles but ending up with arcade ports) To me the various home ports of Outrun for example are the children of the arcade version, they might not touch it in terms of quality but the legacy of the arcade machine, and the against-all-odds porting of it to various 8-bit systems is clearly part of the history of the game and part of our culture. The differences between how games were ported in Europe vs. in Japan etc… How Japanese tech was clearly designed for more direct, closer porting of arcade games, using similar concepts (the near perfect Sharp X68k ports showing that) whereas much of the Europe/US tech descended from business style machines, with framebuffers, and accelerated blitters (and how putting that type of tech in consoles often just resulted in massive failures like the Jag) Having that information in one place allows you to make direct comparisons.

MAME is as much as anything a cultural project, and one which gives you a firm grasp on influences and trends in an industry which where Arcades once pioneered new tech but has since seen the roles dramatically reversed as console tech gradually crept into the arcades. Following that progression from both sides would be a natural part of the project, right now you get a single sided, and rather skewed view of it.

Basically I’m doing the basic Fruit Machine emulation because I feel it needs doing, the lack of information is frustrating, but also proof that they need doing for their own sake. Again one thing that stands out is the difference influences tho. The early games were sometimes influenced by things like Pacman, blatant rip-offs of the concepts. Some later ones are themed around Mario and Sonic instead (the home versions of said characters) again showing how the influences flipped over time.

This type of thing I find interesting, how everything connects together, it just happens to be that from a technical viewpoint it makes sense as well, although that’s no coincidence.

I know some people have a very narrow minded view on such things, and firmly believe that anything that wasn’t in an arcade is an inferior version, not worthy of being preserved, documented, played, or anything else, but that’s no different to the me having a dislike for vs. fighting games, I wouldn’t base policy around it. In some cases ports were superior too, sure the Spectrum version of OutRun might have been terrible compared to the arcade, but the Amiga version of Rodland is arguably better and some of the Laserdisc titles had interesting ports, Dragons Lair on the Amstrad CPC is clearly influenced by the arcade, but presents a very different game simply because you couldn’t have an FMV game on such a platform. Without the arcade originals, these ports would never have existed. If you watch this video of Amstrad CPC games you’ll see at around half of them are interesting arcade ports, some technical marvels when you consider the hardware they’re ported to and look at arcades running on something equivalent! The rest, worthy offerings.

Pipe Dream / Pipe Mania I believe is one that went the other the other way
The original copyright? a UK company, the original version, on the home systems. MAME fails to represent that right now, if you want to go digging you have to look elsewhere, which is a shame. In this case I imagine the original (from memory) is a fairer game, because it didn’t have to take your money. Tetris was also a staple of many systems, and a popular arcade, but started it’s life somewhere entirely different and evolved in different ways on different platforms. Arcades have influenced home systems in terms of both hardware and games. Home systems have influenced Arcade systems, again in terms of both hardware and games. The view from MAME is one-sided at the moment, only tells half the story.

What about the games that started off in the arcades, but then got home only sequels? Capcom were good at this. Final Fight 3 was bootlegged as an arcade, but really intended for the home market only. Super Ghouls and Ghosts was released on the SNES and Saturn. Strider Returns (although after the near perfect Strider port the less said about SR the better) Moving way from Capcom, Tempest 2000 a worth sequel on an ill-fated platform. Contra sequels, the 3rd Streets of Rage? Again MAME only tells half the story.

Then you have games which were released to the arcades in a crippled form. The MegaPlay version of Sonic The Hedgehog has strict zone time limits, the slow zones removed, special stages removed, and is generally a much less complete version of the game. The complete version (identical to the Genesis ROM) did appear on MegaTech, but then you have to credit feed a timer just to play it. Once again, you only really get half the overall picture right now due to the limited scope. The Thunderforce series also only made a brief appearance in the arcade with Thunder Force AC, which combines elements of from the home games but only gives a small window into the series as a while.

There are others. The UK industry of the 80s seemed to do arcade games in a similar way to the home ones. Hunchback Arcade was also popular as a Home game on 8-bit platforms like the C64 The same company also blatantly ripped off Donkey Kong as Logger much as happened on the 8-bits with games like Kong which was at the time unlicensed. They did later release a licensed version however. This is all interesting stuff IMHO.

Odd cases? Psycho Pigs UXB is clearly based on Butasan but said original game was designed for the Japanese market only, so it’s curious as to how it ended up being ported over here.

Kirby’s Avalanche and Dr. Robotniks Mean Bean Machine were the westernised home ports of Puyo Puyo again an interesting evolution (and the Arcade version was even on hardware based on the Genesis). Pac Attack is an exclusive home version of Cosmo Gang – The Puzzle (it was also released in Japan on the SNES with that name)

That’s just looking at the games, an angle I chose because it’s easier to show than to try to explain the hardware connections (beyond the obvious ones like Saturn is the same as ST-V). The story of how the hardware evolved and eventually conquered the arcades is an interesting albeit more technical one too, and one MAME should be telling.

From a business point of view it’s interesting too, Cave, now famous for their manic shooters started off with a couple of titles elsewhere, they did some driving games on the Saturn with Atlus (in addition to Steep Slope Sliders which had an arcade port) Without those releases, and the connections and contacts made you have to wonder where Cave would be today, would it have worked out the same? It’s a part of their history, and the same can be applied to many other manufacturers.

You can’t really deny the connections, it doesn’t make sense to. It’s all part of our heritage, all part of our history and all interconnected. The coin slot and machine location are irrelevant.

I don’t know… it never took me 45 minutes to compile MAME, except on my EEEPC… anyway, I was only referring to the ‘update each week’ comment.

concerning the censorship on some forums, I completely agree: it’s plainly stupid (and, like you, I have been censored there many times with whole threads being removed without a word, even if usually for different reasons).
however, what you fail to say in your comments is that censorship is not as strict as you depicted it: some days the same comment would have simply been ignored and not removed.
of course, this makes things worse in some sense: basically the fate of your posts does not depend on some precise rule, but it depends on mod’s temper of the day… but this also happens pretty often in real life so I sort of learned to live with it…

Well there are possible reasons

1) Antivirus – it’s certainly faster without one, even Security Essentials has overheaded, Avast followed AVG in making things noticeably slower.

2) Slower drives – I’ve purposefully chosen to have slightly slower drives in this system because they generate far less heat, and are more reliable. I found heat to be the #1 killer of drives so would rather go with cooler, slower but more reliable.

3) TSVN – It’s possible TSVN monitoring all the files is slowing it down, but can’t be sure on that one. I know TSVNcache does still like to crash every now and again.

4) Other factors – The modern 2011 version of MSN is a pain, it so often crashes and eats up an entire CPU core worth of CPU time, just leaving it overnight can cause that to happen (it seems to be related to the ad engine…). I stopped using it for a few months it became so bad. Half the people I know using this version have had similar issues, people have abandoned it for the abomination that is Facebook chat as a result. If that happens and I don’t notice then it doubles the build time yet again to well over an hour!

Either way, I’m not going to change 1/2/3 because I consider them essential. Improvements to the build system are needed, and ideally they would allow any combination of systems (home or arcade) to be included in the final build, and hopefully encourage some additional sub-division and name-spacing (pinball/mechanical, casino, appliances etc.) without the risk of things being split off to further additional projects. That would be beneficial to everybody.

All this talk of MAMEUI / MESSUI being discontinued and pulled out of the MESSSVN completely only makes me think that if things had been merged when I said the unofficial builds would be taking up the slack and maintaining the whole of the GUI for the needs of their projects now. Given that MESS stuff wasn’t included in MAME and there are no real unofficial MESS clones I imagine the MESS specific parts of that will now just die.

I changed to aMSN a few years ago, not only because MS banned older PCs from using official MSN beyond 8.1 (or Windows Live Messenger as they call it), but it also doesn’t have the bull$hit ads which Microsoft most certainly do not need in a program in the year 2000. Just because they are a billion dollar corporation doesn’t mean that all ‘free’ programs HAVE to somehow generate revenue. By contrast to Haze’s entire CPU core being wasted on MSN, I can leave aMSN running 24/7 on a nearly six year old single core Celeron machine without any sign of it chewing mass amounts of RAM or CPU time. Currently, it’s using a whole 35MB of RAM in the background, with 2 (tabbed) windows open and displaying a rather large animated GIF (800KB) as an avatar.

yeah I tried aMSN for a while, but it’s ugly and clunky and they seemed determined to import all the worst parts of the changes from the new MSN versions, but make them worse, and not have full compatibility with other proper versions.

I gave up on it when one version update seemed to screw my preferences and end up with half my contacts in some pseudo ignore/block state without me knowing about it, and without them knowing about it either. When a communication tool fails in it’s primary goal, to allow communication between two parties, it has no place on my computer. The official client might be crap, annoying, and getting worse with every update, but at least it does that part.

Haze, could you contact me of let me know how I can pass you some fruit machine information that would be very useful to the MameDev team.

Going off topic I only replace my computer when it dies. My laptop computer from 2006 is still going strong though it is a total dog while in the same period I have gone through at least five desktop computers and those were replaced because they died and I figured something on the motherboard fried itself, because they run everything so fast and this in turn generates copious amounts of heat.

coder> I’ve contacted you, if you didn’t receive anything then check spam folders etc. :-)

KE> I’ve actually found people to replace laptops far more often than desktops, especially these days where most people consider the desktop to be just something that sits there and they use in an emergency.

The laptops get more use, but are also prone to overheating due to the compact cases, and a build up of dust + easily damaged by something being spilled on them, which people manage at an alarming rate. It’s also very easy to leave your laptop somewhere stupid, or fall asleep watching something, end up throwing covers over it, and have it overheat and die that way.

Maybe sensible owners, that doesn’t apply to, but people buying cheap throwaway laptops usually aren’t sensible (and the recent trend I’m seeing is that eventually they just decide the PC hardware is trash, even if it’s their fault, and buy a more expensive MAC which they then take far better care of, because it was more expensive.)

Of course this doesn’t mean people have cutting edge hardware, most of these throwaway laptops are pieces of junk, people buy *new* ones then wonder why they can’t play The Sims 3 and similar games properly.. (which again leads to an overall hatred for PC / Windows when really it was the fault of the buyer for buying cheap garbage)

I still have a feeling it won’t be long before the primary desktop/laptop platform is Apple, I don’t like them, at all, but suicidal Windows Activation stuff (which only affects genuine customers when they fuck things up then try to reinstall) and this kind of throw-away junk PC hardware is putting an entire generation off PCs. The fact that on modern OSs there’s so much crap going off in the background, and again stupid copy protection schemes has also near enough killed the PC off as a gaming platform for legitimate customers (People want to be mobile, on laptops, forcing games to have require internet access, especially always on internet access is as dumb as you can get, a customer might accidentally buy one game like that, but they’ll NEVER buy another, -especially- non-geeks who can’t work around it)

I’ve even overheard Macs being recommended as a better gaming rather often in the last year or two! Games were one of the only big things Windows had going for it and keeping people there historically. It would probably be a good time for Linux to grab the market really, but they won’t, they’ll let the chance slip away because the OSS community, much like MAME and many other free projects simply can’t get their act together when it comes to user experience, and all working TOGETHER on the same thing.

But yeah, it surprises me you’ve had more issues with desktops, but again, maybe if you buy cutting edge you get the same problem as buying cheap garbage, it’s all being pushed a little TOO hard. Faster spinning drives, more head, SDDs have their own share of problems, over complex motherboard designs with more to go wrong, early generation chipsets with bugs in both hardware and software. Even when buying NEW I wouldn’t go for cutting edge for those reasons, I’d go with a proven and reliable design which had been around for a while, not only saving me money, but giving me something I could hopefully trust. Sure, sometimes that comes at a cost (in this case, slower compile speeds probably due to slower HDDs) but I’d take that option any day.

I haven’t been going with cutting edge components just middle of the road components. The hard drive and dvd drive work fine except the motherboard or video card end up frying themselves. To keep improving performance the chips are running hotter and hotter. (I believed you mentioned this.) I have been concerned with heating problems since Intel released the Pentium 60/66 and a heat sink became necessary but what do you do? The only choices are Intel and AMD anymore. Cyrix and the other clone manufacture which I cannot remember the name of anymore don’t count and I have no desire to try and design my own CPU.

I’m not even remotely ready to try a SSD drive until the issue with the number of write cycles flash memory can handle is resolved. I’m going to sit on the sidelines with a bag of popcorn and wait for the people to complain when their SSD drive doesn’t save data anymore and they learn why it’s called the bleeding edge.

By continuing to use the site, you agree to the use of cookies. more information

The cookie settings on this website are set to "allow cookies" to give you the best browsing experience possible. If you continue to use this website without changing your cookie settings or you click "Accept" below then you are consenting to this.