Pretty new..
- Inicie sesión o regístrese para enviar comentarios
So, technical support forum, that's new to me,..
never needed such a thing before, so, thanks!
Looking at system requirements for Trisquel, it simply states;
a computer built in 2005 and newer.
It was bought in 2007, but, I don know how long it had been in stock for example..
First thing on the list to check would be CPU, I thought as represented by Windows it wouldn't be one that would even be enough to support Ubuntu, so I chose Lubuntu based on that thought, not being knowledge.. now I've gotten rid of windows for six weeks, read up on the ish and the jinx, and am now looking to replace the windows replacement Lubuntu with Trisquel! Du-uh!
The question!
Can I 'safely' put Trisquel on this system without worrying I'm
not giving Trisquel the hardware it deserves?
(Simply put; does the desktop meet the system requirements for Trisquel?)
https://trisquel.info/en/wiki/system-requirements
Now running on the replacer, it shows in the System Information Summary it's not even one processor. Windows might have shown that to, but I've never noticed..
Turns out there's two CPUs.. euhm, haha, now it got lost on me..
It's got: 2x Genuine Intel(R) CPU 2140 @ 1.60GHz 1600,00MHz
That's not even intel Celeron I guess? Hope you're not missing I don't even actually exactly know what I'm talking about. :D Because looking at the next informational item on the list, the 1.60 GHz is quite a little short on the stated 1.80 GHz.
But than again, there's two.. oooo, sweet mysteryyy, what will this mean??
Moving on to RAM, it's a 32 bit, 1012 MB, so that would suffice I guess..
The video card, if I'm not making a mistake here, is a;
Intel Corporation 82945G/GZ Integrated Graphics
Controller (rev02) (prog-if OO [VGA controller])
(I simply note this because it says Graphics)
It's a 160GB hard disk, with 150 GB free space..
That's the hardware part, looking at the current OS after installing, it turns out, it ís ubuntu 16.04.3 LTS, and only the desktop environment is: LXDE (Lubuntu). Judging by this I'd say it's perfectly possible to replace the current OS with Trisquel, but it's nice to take this chance for assurance..
The BIOS does support boot with flah drive, that's how 'I' (haha) succeeded in replacing windows with Ubu, maybe that tells something to..?
Well,.. really looking forward to this new adventure.. I'm not just going to put it on the desktop, I really feel like exploring the world of free software some more, and I think Trisquel is the one to start that with.. to me it is, the decision and/or choice is made, guess now I am to go and understand it.. let's see where the white bunnies take me this time!
Anybody feel like playing a little rabbit? You don't have to run, just show me the hole by answering the system requirements question please, I'll jump in myself from there on out..
Thanks for creating thé hole in thé tree!
Kind regards!
Your hardware is fine. However, doubling the RAM would not cost much and would make a real difference in the number of Web pages and applications you could simultaneously open.
You can test Trisquel's live system before installing (or not).
> It's got: 2x Genuine Intel(R) CPU 2140 @ 1.60GHz 1600,00MHz
I believe the “2x” says the processor has two cores. (This means it's better at doing two things at once.) The model appears to be Pentium, which shouldn't fall behind Celeron, as far as I know. It looks fine to me.
The rest should suffice as well, but like Magic Banana wrote, more RAM might improve your experience, as many programs nowadays consume a lot of memory.
Trisquel should work for you, and if you find that the default setup isn't smoot enough, you can install a different environment on it, too. There's also a version called “Trisquel Mini” to download here, which comes with LXDE preinstalled, so it's to Trisquel what Lubuntu is to Ubuntu. (I'm not sure whether it's going to be continued in the future, though.)
At any rate, good luck and enjoy Trisquel! Run free, run GNU!
With 1 GB (even 2GB) of RAM, I would *really strongly* suggest Xfce, LXDE, or LXQt desktop environments instead of KDE or GNOME or a GNOME derivative. I gather that Trisquel Mini comes with LXDE, so I would suggest Trisquel Mini. After the installation, you can additionally install LXQt[*] and/or Xfce and try them all before settling on one.
[*] LXQt is the continuation of LXDE. LXDE is being phased out, while LXQt is being phsed in by the same development team. Currently, LXDE is still usable (not really phased out), and LXQt has caught up and become quite usable already. I use LXQt and I'm quite satisfied so far. As for tomorrow, LXQt will sure surpass LXDE. But then, future Trisquel Mini editions (if any) would probably come with LXQt or Xfce, instead of LXDE.
In short, I would install Trisquel Mini, and then - optionally - additionally install LXQt and Xfce on top of it, and try out all 3. Also I would heed Magic Banana's suggestion, that try to upgrade RAM to 2 GB. (1 GB is more than enough for the system itself. But if you use modern web browsers, office suits, and the like, it may not be enough.)
> With 1 GB (even 2GB) of RAM, I would *really strongly* suggest Xfce, LXDE,
> or LXQt desktop environments instead of KDE or GNOME or a GNOME derivative.
I have 2GB of RAM. I have tried GNOME 3 and indeed is a problem. However, I did not have any issues with GNOME Fallback when I was using Trisquel 7 or MATE on Trisquel 8. 1GB might be pushing it though. I would recommend a simple window manager rather than a full desktop environment, but since the OP is a beginner you're probably right that LXDE is their best bet.
> I gather that Trisquel Mini comes with LXDE, so I would suggest Trisquel
> Mini. After the installation, you can additionally install LXQt[*] and/or
> Xfce and try them all before settling on one.
I wonder what is the difference between installing Trisuel Mini versus installing regular Trisquel and installing LXDE?
> But if you use modern web
> browsers, office suits, and the like, it may not be enough.)
Yes, choice of programs is at least as important as choice of desktop environment. I try to use as many cli programs as possible, but for a beginner I'd recommend
- abiword as a word processor
- I only use a word processor when formatting is important. Writing can just be done in a text editor (I use vim, but pluma and gedit are good graphical alternatives) and pasted into abiword at the end.
- gnumeric for spreadsheets
- mpv as a media player (or vlc if you need something a little more fully featured)
- claws for email (or neomutt if you like cli programs)
- lifrea (maybe) as a news reader. I say maybe because while I've never had trouble with it, I haven't gone out of my way to try many other graphical news readers.
- ??? for image manipulation (Can anyone suggest a low-RAM-usage alternative to GIMP?)
- audacity for editing audio
- gajim for chat (or mcabber if you like cli programs)
- linphone for sip (does anyone know of a cli sip client that can run as a daemon and receive incoming calls?)
I don't know if this is true as a rule, but I have found that as I have begun to avoid using my web browser for more than internet searches and do everything else (email, chat, news) with desktop clients, I have run out of RAM much less often. However, this has coincided with my replacing several graphical programs with cli ones, so it's hard to say how much of my reduced RAM usage has anything to do with how much I use a web browser. It also surely depends on the browser. I absolutely recommend disabling JavaScript, though. Freedom and security issues aside, a JavaShit heavy site can easily use up almost all of my RAM, especially Screwgle Drive (which should of course be avoided for other reasons anyway).
> I wonder what is the difference between installing Trisuel Mini versus installing regular Trisquel and installing LXDE?
I don't know to be honest. I know that it comes with LXDE, and I gather that it is a rather minimalist edition. So I *infer* that default software selection and their default setups would probably be in line with a minimalist strategy. I.e. perhaps your software suggestions (or similar) might already be in place with the Mini edition. As I don't know normal edition neither do I mini edition, all I can say that it would be prudent for a low-spec computer to start with mini edition and then slowly add on top of it, instead of starting with the full blown edition and then go the other way around.
> Writing can just be done in a text editor (I use vim, but pluma and gedit are good graphical alternatives) and pasted into abiword at the end.
I would say please give JuffEd a try. I also do most of my writing in plain text editors. When I was transitioning from KDE to LXQt I was in search of some good sophisticated plain text editor (as it is kind of my habitat :), particularly good sessions support, and finally settled down on JuffEd. Its name sounds somewhat... amateurish, but the stuff is very good. Almost as sophisticated as kate, yet with much less resource requirements.
> - claws for email (or neomutt if you like cli programs)
> - lifrea (maybe) as a news reader. I say maybe because while I've never had trouble with it, I haven't gone out of my way to try many other graphical news readers.
Claws-mail is a dual purpose program, with both email and news reader capabilities. So I would suggest using just claws-mail for both purposes.
> I would say please give JuffEd a try.
I just installed from the Trisquel 8 repo. Creating or opening a file causes a seg fault, so I wouldn't call myself an expert in JuffED at this point :) but it does indeed look similar to Kate, which I used before switching to vim.
> Claws-mail is a dual purpose program, with both email and news
> reader capabilities. So I would suggest using just claws-mail for
> both purposes.
Yes, I forgot Claws had that feature. It's been a while.
1 gb of RAM is enough, I used Trisquel with gnome-fallback with 1 gb of RAM in the past and it was just fine. Using lightweight programs is highly recommended. Firefox with few addons is not recommended if you are used to having lots of tabs open or simply browsing heavy websites (diaspora* comes to my mind). And, no.. I don't have a decent alternative for it. I like netsurf although it screws up the layout of many websites. It's gold with wikipedia :)
Without addons the recent versions of FF take around 150 mb of RAM when opened, you might want to try that one provided the proprietary foxy logo (which is not software) is not an issue for you.
welcome to da forum o/
> And, no.. I don't have a decent alternative for it. I like netsurf although it screws up the layout of many websites.
Have you tried midori and qupzilla (aka falkon)? I find midori the best in terms of "hogtionality" (resource hogging vs. functionality balance). Qupzilla is twice as worse than Midori, and Firefox is twice worse than Qupzilla, in terms of hogtionality.
Www compatibility (or better put, ability to cope with broken web sites) goes in reverse direction. Firefox is the best, then comes Qupzilla (subtly worse than FF), then Midori (again subtly worse than Qupzilla).
So, the hogtionality ladder goes like this: For a subtly better compatibility, twice the resources are hogged. (Midori -> Qupzilla -> Firefox)
Currently I have 26 tabs open in Midori, that has been running for a day or so, and its RAM usage is 286188 K. Now I have just freshly started Firefox (one tab, blank page) and its RAM usage is 270572 K.
Edit: I ought to add that, everything is disabled in Midori, while everything is enabled in Firefox (ESR). "Everything" = Java, JS, Flash, images. (No extension is used in FF)
>Have you tried midori
Bugori, one of the worst software I ever used. It's also abandoned, unmaintained software (and for a good reason I would say). :P
>Qupzilla
Yes, but for some reason I did not like it (the version in Debian Stable, 1.8.9). I just reinstalled it. Will re-test it later today.
> Bugori
:)
Yes it is somewhat buggy but not so much as to render it unusable, at least for me. It should be normal, given that it is orphaned for a few years. Still, I use it for 99% of my web browsing. I'm just being careful not to touch its sore points.
My most favorite bug: Whenever you type the letter "b" in URL entry line as the first character, it crashes. Only crashes on the letter "b", only when it is the first character typed in URL line. Copy-pasting such a line doesn't cause it to crash, though.
Seeing that it is a lean and efficient browser, and it is orphaned (that is, ripe for adoption) I wish GNU would adopt its maintenance, and offer Midori instead of IceCat.
> My most favorite bug: Whenever you type the letter "b" in URL entry line as the first character, it crashes. Only crashes on the letter "b", only when it is the first character typed in URL line. Copy-pasting such a line doesn't cause it to crash, though.
Just checked again, and the bug has simply vanished. It was not thre either when I first installed Midori. So I infer that it was something to do with URL auto-completion. Apparently the offending URL timed out and purged from the list, so did the bug.
Re: Qupzilla
> Yes, but for some reason I did not like it (the version in Debian Stable, 1.8.9). I just reinstalled it. Will re-test it later today.
Qupzilla is in a rather fast development cycle, so Debian stable might not be hte best way to evaluate it. I'm on Debian testing and currently have Qupzilla 2.2.3 installed (upgraded last month). I see that 2.2.5 is already available on the repository.
I thought so. I will compile the latest source. Am curious :)
EDIT: heh, unfortunately it requires a higher version of qt than the one shipped in Stable :/
Btw, I know now why I didn't like it: the font rendering is beyond awful. It also doesn't respect my GTK theme :(
Have you a fix for the font rendering issue? Fonts are terrible, just terrible. Was it not for teh fonts I would enjoy this browser, it's not as fast as netsurf but fast enough and it doesn't screw the layout of most websites I tried so far.
> Have you a fix for the font rendering issue? Fonts are terrible, just terrible.
There is no problem with fonts at all in my version 2.2.3
Edit: And I was not aware of it in the earlier version (of late stable). Maybe it was problematic when stable has just been rolled out (when you tried it), and resolved later. You said you have re-installed it from stable repo. Perhaps it is resolved now in your version too?
Using the awesome thing a search browser is (I'm looking at you GrevenGull :P ) I managed to fix both the issues.
I did notice before that the only other qt5 program I used had the same issue (GTK theme and fonts) but was not aware it was an issue with qt5..
So, the fix for the GTK theme is:
Since libqt5libqgtk2 isn't available in the repo (installing this package if available should fix it) a workaround is needed ->
Install 'qt5-style-plugins' and edit /etc/environment to add 'export QT_QPA_PLATFORMTHEME=gtk2' . Reboot..
The fix for the fonts is:
Create a file named /etc/fonts/local.conf with the following content ->
attached txt file*
-----------------
Qupzilla is very good so far, I am quite enjoying it. Tx for metioning it and thus stimulating my autismo. \o/
Adjunto | Tamaño |
---|---|
Qt5 Fonts Fix.txt | 635 bytes |
> Qupzilla is very good so far, I am quite enjoying it. Tx for metioning it and thus stimulating my autismo. \o/
You are welcome. :)
As I use LXQt (though installed both LXQt and Xfce) I depend on whatever Qt environment LXQt provides by default. That is possibly why I didn't run into the font problem you have experienced.
Edit: Just tried Qupzilla in Xfce and no font problems there either. So it must be related to the packages and setup LXQt provides - not the actual LXQt runtime environment.
>Just tried Qupzilla in Xfce and no font problems there either
Debian Stable (current and also the old Jessie) with Xfce is apparently allergic to Qt.. The fix for Qt4 is installing the convenient program qt4-qtconfig and then setting GTK+ as 'GUI style'. The fix for Qt5 is more convoluted but fortunately it can be fixed as described above. I hope the next stable will get it right :/
what about Palemoon or Basilisk internet browsers?
> what about Palemoon or Basilisk internet browsers?
I have a very simple program hunting strategy. If it is not on Debian, expect some blatant or between-the-lines issue with that program.
Palemoon and Basilisk are not in Debian repositories. If I am not desperate, then I look no further, as I trust the community. If I am desperate, then I will install *anything* from source or binary. :)
So I have not tried them, and won't, unless they appear in Debian repositories. (May be already hosted in "unstable" distribution, I don't know. I use "testing".)
Ironically, Midory is included in Debian stable (Stretch) but kicked out of testing (Buster), the reason being it is too old, orphaned, and serious security vulnerabilities pending. Unless Midori gets adopted by an outreaching golden-hearted family, it will likely stay out of Debian.
I stole it from the stable distribution, and use it in testing. Will do so as long as I can. ツ
Re: Midori
> The files in the archive were from 2015. So it is most likely too insecure already.
That is one of the reasons why Midori is removed from future releases of Debian.
[QUOTE]
The final stable release of Midori still uses the unmaintained WebKit1 instead of webkit2gtk and therefore the browser suffers from numerous known security vulnerabilities.
[/QUOTE]
https://bugs.debian.org/864951
OTOH, I use it in totally passive mode. No active content processing is enabled. No exposures, no vulnerabilities. :)
> There are things that happen in the HTTP layer itself, also in cryptographic layers etc.
When there are so many (~all) systems with active content enabled, it must take a really curiously demented hacker (or agent or whatever) to attack passive layers.
So I am hiding behind the flock. ツ
> Let's not forget that they attack every possible layer, not just what is easy.
By now, I think I should have already given my general approach to usability - security - privacy equation.
The thing is, I am not specifically targeting any one of them in depth. I am *compromising* for the best balance between them, using light (easy to implement) but radical means.
The fact that I use Midori with all active content disabled serves me in several fronts:
(1) It automatically filters out the crop (internet junk) from the cream (valuable information). My attention span is better used on worthwhile information.
(2) It automatically filters out non-free content and malware.
(3) It increases snappiness of internet use, at the same time decreasing bandwidth waste.
Whenever I run into a situation where I *have to* use active content, then Qupzilla is waiting there, with all the bells and whistles enabled.
You may say, what use is protecting oneself 99% of the time, if one is exposing himself 1% of the time? For one, Midori in passive use does not protect me fully (but I make a compromise there). Second, I am shrinking my exposure as much as possible, in line with my approach to security. This is not an all or none game, rather it is a multi faceted compromise.
Edit: In addition to (3), it also decreases the tax put on system resources.
> Of course if one doesn't get out of home and has thick window bars one doesn't need to worry about getting a sunburn but what kind of life is that? :)
1% of the sunbath I am getting with Qupzilla is enough for me for a healthy dose of vitamin-D synthesis. ツ
You obviously do not know what you are talking about. You apparently think that "all CSS" is some kind of "infection". Like a disabled telemetry, in your mind (lol). Nobody browses the Web blocking *all* CSS, unless he/she is masochist.
Let us take your website as an example: https://anchev.net
Its CSS is more than 10,000 lines long (!): https://anchev.net/home.css
Without it, your site becomes much uglier: http://dcc.ufmg.br/~lcerf/anchev_no-css.html
I blanked all pictures. Although you told me in https://trisquel.info/forum/web-browser?page=5#comment-127512 that I have "a serious mental disorder" for distributing my scripts under the GPLv3+ (that allows copying), I am afraid your judgment may suddenly change when it comes to copying your pictures.
Anyway, *some* (not "all") CSS can allow third parties to track the visitor. Like Google fonts taken from fonts.googleapis.com. However, if the website uses Google Analytics, that does not make much of a difference for most users, who will execute that proprietary JavaScript and be tracked (how disgusting!). You know, like on your site: https://anchev.net/home.js
FWIW (not that I owe anyone an explanation): this website was made on a quick notice, using a ready made template.
Yet you took the time to subscribe to Google Analytics and to add their proprietary JavaScript to your website.
The majority of the visitors are using Mac OSX, Windows, Google Chrome, (not hardened) Firefox, Android and iPhone to browse it, so by visiting this particular site they are surely not more or less exposed than they otherwise are.
You really love the perfect solution fallacy: https://en.wikipedia.org/wiki/Nirvana_fallacy#Perfect_solution_fallacy
So, because you believe Microsoft or Apple or ... (for users of GNU/Linux... there is still Spectre... and the cables are owned anyway!) already tracks the visitors of your website, it would somehow be OK to make them automatically download and execute proprietary JavaScript that only aims to track them with the help of Google!
Well, no. That is just disgusting.
Regardless of that there will be a new version of the site which will come online when it is ready.
"The baby is infected and is infecting the other babies, including the favorite baby. (...) When one sees that the house is burning one doesn't sit and discuss - one acts instantly": https://trisquel.info/forum/pretty-new#comment-128408
That is just taking one of your posts in this thread. Many of your posts are similar: blaming the FSF, Mozilla, etc. for minor (if not non-existing) issues such as a *disabled* telemetry component. But, somehow, you should not be criticized for *enabled* Google Analytics on your website. Something you could disable in literally one minute.
I removed your page from my site. As expected, you consider at the same time that not asking "you[r] permission to copy anything from [your] site" is unethical but that sharing my code under the terms of the GNU GPLv3+ "is not a moral stance but a serious mental disorder".
This thread is not about "What do you Magic Banana think about person X and his site".
Indeed. I have just posted a thread called like that: https://trisquel.info/fr/forum/what-i-think-about-george-anchev-heyjoe-and-his-site
Thank you for the suggestion.
And you really have the nerves to talk back against MBs post... i personally would just have been ashamed and stayed silent.
> [Qupzilla] has no mechanism to control loading of 1st and 3rd party resources (as uM and uBO)
I don't trust Mozilla and all the derivative downstream works that are based on Mozilla technologies. I would rather use Qupzilla - even Midori - without uM and uBO and what not (lack of ability), than use Mozilla based products (lack of goodwill) - let alone being a hog of a browser.
Lack of goodwill is *qualitatively* more dangerous than lack of ability.
> Lack of goodwill is *qualitatively* more dangerous than lack of ability.
Here again, it is not a uni-dimensional case of ability vs. goodwill.
(1) Obese software are generally better ripe for smuggling in some malware, than slim and lean ones.
(2) Obese organizations are generally better ripe for infiltration. (1) and (2) go usually in lockstep: If you have 1, then you have 2 too, and vice versa.
(3) For the reasons above, large software are more likely to be targeted by institutional intelligence attack (I hoped we would discuss this in the other thread but it is drawned among the other things)
(4) Obese software means complexity, which translates into larger exposure to threats.
(5) Obese software are resource hogs.
If you think about (3), then it becomes all too plain why "lack of goodwill" mostly occurs in obese software circles. Add other drawbacks to that, and you can see why I prefer keeping away from obese software. It is a practical, easy, and radical approach, than putting a monstrous of software with millions of LOC under microscope and trying to secure it. It is practically impossible, and possibly a good waste of your time.
You already know one of my strategical approaches to *practical security* (shrink exposure, get rid of ballast, simplify all). It is quite easy and straightforward to implement. And quite an efficient one.
So, there is more to it than just "lack of ability vs. lack of goodwill".
As for IceCat being endorsed by FSF, I believe it was just a tactical error. As humans we all are prone to errors.
I guess the rationale behind it was "if we take a successful browser with good www compatibility, and if we prune it (mission impossible) then we would be able to offer something *desirable* for the masses, instead of some theorethically perfect niche browser with a niche user base."
Your warning RMS of the issues with IceCat is indeed a *BIG* service to FOSS. Because it will hopefully prompt FSF to think it all over again. I don't expect a quick decision any time soon, because it inherently entails radical measures to be taken. I *do* really hope GNU adopts Midori and starts offering it instead of Mozilla derivatives.
> https://trisquel.info/en/forum/web-browser?page=6#comment-128395
Mission impossible, and a tactical approach. They should question the base instead.
Edit: Oops - I meant the post one above, regarding IceCat. As for uzbl, I will look into it.
> (3) For the reasons above, large software are more likely to be targeted by institutional intelligence attack
Add to that (2.5) obese software have usually broader user base, offering a better ROI (return on "investment") for intelligence attacks.
> We still can. We actually did and it is marked for further investigation. (simplicity)
That is a pending issue, but what I meant was something else: The possibility, ways, and remedies of institutional or intelligence attacks on FOSS. That issue needs to be laid on the table and given a good in-depth look.
> Sure but it is not always practical, e.g. in a work scenario...
Yes, but we still need to shrink exposure as much as we can. If your exposure is too risky to mix it with your personal environment, then you simply separate them (to separate laptops if need be). There is no perfect compromise. This is also the same in everyday life.
> What measures? What is there to wait for?
Let me be emphatic with the FSF. We have invested such a money and man-months in a browser to prune it, then someone tells us that it is leaking information, and more importantly, it may be doing this on purpose. I.e. we got an *expensive* baby in our arm that bites. It is all too easy to tell them "just drop the baby". Then what? Before doing that, they have to evaluate the alternatives, the costs involved, success probability, etc. If I were FSF, I would simply put IceCat to "maintenance mode", and adopt Midori (it *is* an excellent option) and iron out its bugs and start pushing Midori in less than a year, offering IceCat in the mean time (don't forget that they are running a large ship, not a canoe - they cannot quickly change direction on whims).
What would you expect? RMS publicly stating that IceCat is crap, that it is removed from GNU archives and endorsement list, that everyone should just quit using it and instead just use what the heck they want..? Dropping IceCat and adopting something else is very, *very* serious and expensive affair. And as I have said, FSF is not a small canoe that can change directions on short notice.
We need to wait and see what comes eventually off of your warning RMS about IceCat. Just patience.
> Ok, lay it. Let's look into it in the other thread.
I will, sometime.
As for the other points, I don't want to derail this topic so I will hopefully address them in the security thread in troll lounge - in due conversations they will come up by themselves, anyway. :)
But with the "expense" I meant both financial and man-hour costs. Just because you do some work for free doesn't mean that the work has no associated costs. IceCat didn't grow on trees, after all. As for my using FSF in place of GNU, it was my bad. I use them interchangeably, while they are not the same. OTOH, you were right in that there are incremental steps that can be taken, such as working on a better user.js and things like that, instead of waiting for the big strategical decision over IceCat.
>I don't need to wait, I need a clean browser - today. Not when someone is in the
>mood for it. I don't want to depend on anyone, dependency is not freedom.
"I don't want to depend on anyone" is not compatible with "I don't need to wait, I need a clean browser - today. Not when someone is in the mood for it."
Code a browser yourself if you don't want to depend on anyone. Fork one if you want to pretend you don't depend on anyone and still have some cred. At least, contrubitue to make/fork/improve one.
Behan put it best: “Critics are like eunuchs in a harem; they know how it's done, they've seen it done every day, but they're unable to do it themselves.”
I speculate you don't even know how it's done (neither do I, but at least I have respect for those who do and give their time freely to help me, their imperfections notwithstanding).
I mention Palemoon, because it is fork of early versions of Firefox. I love this old interface-UI-GUI.
[SuperTramp83]
> Without addons the recent versions of FF take around 150 mb of RAM when opened,
[/SuperTramp83]
[me]
> Now I have just freshly started Firefox (one tab, blank page) and its RAM usage is 270572 K.
[/me]
Just in case anyone wonders why there is such a big difference, it is probably and mostly due to different "memory crowdedness" of our systems. The more crowded RAM becomes, the more aggresively kernel starts paging out code segments (and shrinks data cache, which is out of our scope in this context). IOW, contention for RAM increases linear to the crowdedness. The more acute the competition, the less memory is allocated to programs, which translates to increasingly more active pages occupy RAM space, others paged out. So the same program can ocuppy much different RAM (consume) in different environments.
Crowdedness is a function of cumulative memory needs of all programs versus available physical memory.
E.g. on a system with 16GB RAM and running no other program than Firefox (except a minimalist barebones DE), the same Firefox' initial startup RAM usage may be perhaps 1 GB.
Conversely, on a system with 1GB of RAM and with a very "rich" (bloated) runtime environment, initial RAM consumption of the same Firefox may be perhaps less than 100 MB.
I have 4G of RAM and concurrently using such and such programs. SuperTramp83 uses perhaps 2GB of RAM and uses a different set of programs. So, our systems' memory crowdedness factors are different.
This is one reason why RAM consumptions cannot be benchmarked between different systems (or different times). I use program X and it consumes 100M. An hour later the same program consumes 200M (decrased crowdedness). You use program Y and it consumes 300 M. An hour later the same program consumes 150M (increased crowdedness). It conveys nothing. When I (or you) use *both* programs at the same time and record their *concurrent* memory consumptions, then this is meaningful benchmark-worthy data.
I always wondered indeed how come different hardware with same programs use different amount of RAM memory.. My laptop has 3 gb of RAM. Had no idea about 'crowdedness'. A fresh FF (no addons) when loaded uses exactly 140 mb of RAM here. Once the three addons I use are installed it goes up to 230.
Nexine, I'd like to apologize, on behalf of the Trisquel forum regulars, for the hijacking of your thread. I hope you successfully managed to get Trisquel installed, and that the first 8 comments or so (the ones that were on-topic) were helpful. If you are still having any teething problems, please feel free to start a new thread, with a title that reflects those problems.
One thing I'd add is that replacing your old magnetic hard drive with a new SSD (Solid State Drive) makes a huge difference, a much bigger difference in my experience than increasing RAM from 1GB to 2GB (which I did as well). I run Trisquel 7 on an Aspire One mini-laptop ("netbook") that I bought in 2010. I basically stopped using it a couple of years ago, as it had become painfully slow, and I managed to get a slightly newer laptop for $50. When I put in an SSD, on the advice of my peer programming buddy, it made a *huge* difference, and the Aspire One is back to being my main computer (I sold all the other prior to relocating).
- Inicie sesión o regístrese para enviar comentarios