SQL on ?

Having installed Wordpress on a Pi B, the standard packages work fine. They installed cleanly and with no issues. I used the recipe at

The only issue being that as laid out there the php tests don't work. The php lines need to be closed with "?>" (without the inverted commas)! Rendering is slow and how it would perform in a real loaded enviroment I am not so sure though Google will find you pages for people who have optimised the standard instal for speed.

MArtin

--
This email is free from viruses and malware because avast! Antivirus protection is active. 
http://www.avast.com

Owncloud works very satifactorily with sqlite, much easier to manage in my opinion too.

I think Wordpress defaults to mysql though, try it and see how it goes, you've lost nothing.

--
Chris Green

I run mysql on a pi without problems

In article ,

Some time (er, 2 years?) back I compiled up my usual LAMP stack for a Pi which was a slightly tuned kernel, apache, php & MYSQL... Then stuck wordpress on it.

I was plesantly surprised with the result. It was usable - although I didn't load any big plugins into wordpress.

That was on a Rev 1 Pi too (256MB of RAM). It's more CPU bound than SD/Memory though.

I think it would be fine for a small home/hobby setup especially on a Rev 2/B+ with a lot more RAM.

Gordon

You're probably not going to want to hear this, but if you post with Unicode characters that are not in ISO8859-1 to an English Usenet group, then expect to see your name and/or the title horribly broken in the followups with the followups misthreaded as the various newreaders decide that the title is in the same thread or not at random because each newsreader will break the Unicode in different ways. Bear in mind that Usenet can now be regarded as a "legacy" system, with perhaps most newsreaders only able to display a limited character set. And while ISO8859-1 contains a number of non-English characters, the s-with-cedilla unfortunately isn't one of them. You'll have to settle

So far only one out of five replies (the only Windows user?) has mangled the '?', and this post should make it one out of six.

Also the trn user, so that was 2 out of 4, or 50% before I posted, which was par for the course, but that would get worse as people followed the followups.

This tin user (a descendent of trn I believe) is seeing the '?' OK, the Linux command line works pretty well with UTF, I get to see Chinese, Russian, Arbic etc. in all their glory! :-)

--
Chris Green

I'm the (a) trn user... Maybe time to move on, but it's done me well for the past decade or 2...

What's mildly irritating is that I see the weirdo characters OK here as I type this into vim, but get the inverse question marks, etc. in the trn viewer...

Gordon

So you have installed SQL and it behaves well? I?m also thinking of/for OwnCloud.

B. Alabay

--
http://www.thetrial.de/ 
???????????????

Yes, and we live in 2014, usenet can?t be *so* legacy that it touches

only for shy ?s :-)

Windozy ? yes, but (I hope) Raspberry users deal /a little bit/ with linux, unix, darwin or something else and therefore know how to adjust utf.

B. Alabay

--
http://www.thetrial.de/ 
???????????????

BTW, this tiny little ? found its way via tin (2.3) and vim on Darwin.

Burp.

B. Alabay

--
http://www.thetrial.de/ 
???????????????

Globalization says hellouw :-)

Vim should be unicode-aware out of the box, trn ? well, tin is really not bad.

B. Alabay

--
http://www.thetrial.de/ 
???????????????

On Thu, 30 Oct 2014 07:47:38 +0000 (UTC), Ba?ar Alabay declaimed the following:

SQL is a semi-standardized language for querying databases -- you don't "install" SQL.

SQLite3, MySQL/MariaDB, PostgreSQL, Firebird, Interbase, Oracle, Sybase are all relational database systems.

The odds are good that SQLite3 is already part of an install -- it's a file-server (each client program is directly accessing the database data file) rather than a client-server (clients send requests to a daemon server process which has full control over the data file(s) in use).

No idea of what constitutes "weak" -- I used to run MySQL on a Win98 machine with a 733MHz processor (where the server had to be started as part of my login startup programs, as Win98 didn't support running it as a system service).

For an r-pi -- the big killer may be wear&tear on the SD card. Relational databases are heavy I/O processes.

--
	Wulfraed                 Dennis Lee Bieber         AF6VN 
    wlfraed@ix.netcom.com    HTTP://wlfraed.home.netcom.com/

I meant MySQL instead SQLite.

So you mean, it?s no problem to run (e. g.) Owncloud with MySQL instead SQLite on a Pi?

Is there a difference between MySQL?s and SQLite?s usage of the SD? It?s a 32 GB card, so it?s not yet cramped.

B. Alabay

--
http://www.thetrial.de/ 
???????????????

I'm using an Alix board as linuxbox with my personal website (serendipity blog + postgresql). Has to be said that /boot is in the internal CF card, but the rest of / is on an axternal USB HD. It's not the fastest webserver in the world, but it works. The Pi is more powerful (and has more ram) than the Alix, so I don't see the problem there.

Bye Jack

--
Yoda of Borg am I! Assimilated shall you be! Futile resistance is, hmm?

On Fri, 31 Oct 2014 08:00:03 +0000 (UTC), Ba?ar Alabay declaimed the following:

That question I couldn't really answer -- especially as I lack an R-Pi (I've got a half dozen assorted PICs, Arduino Uno/Mega256/Due/Esplora, Propeller, overpriced BASIC Stamps, and a few TIVA C boards, and while the Due [ARM M3] and TIVAs [ARM M4] could support a minimal RTOS, they aren't really meant for dynamic application loading.

MySQL does seem to have gotten fatter over the years... The 64-bit Win7 mysqld process is running in a 470MB working set -- though how much of that is actually loaded into RAM I can't say (Did find a "commit charge" of

980MB).

I expect a 32-bit build (especially if it uses mostly ARM Thumb 16-bit instructions) will be much smaller in active memory footprint.

It's not the space needed (that is mostly going to be down to your dataset) -- it's the heavy rewrite of blocks when you change a record; and if you change a record key field, you also have the rewrite of part of the index. MyISAM format creates three files for each "table"; index, fixed width data, and variable width data.

If the tables are mostly static data -- with an initial batch loading [turn off indexing until the data is loaded, then let it build the index in one run] all you'll be doing is reading, and that should not affect the lifespan of the card much. But updating even a small (

Mysql running on one of my R-pi's uses 34 mb; climbing to 42 when I did a statistics run on a 15k line table with three joins (two inner) against supporting tables. It uses 317 mb virtual space; but it is mostly empty.

Indeed. The emacs binary is 38% as big as the x86 counterpart, around 22% of the x86_64.

For the right sd card this shouldn't be a big problem. Disks have wear too.

I went for SD and SSD cards/units all over three years ago, and constantly went for high end media; mostly sandisk but also some samsung. I haven't had a single failure on more than 30 devices, some constantly updating.

This is not something I can say about disks. (the ones with the rotating oxide).

But SD cards are not made alike; there are huge differences between them. Some are barely good enough to store photgraphs on, with a single digit edit rounds and then permanently withdrawn as original media from the camera.

-- mrr

Slow response to this thread, but I had to do a lot of playing first...(:-/)

I'm also a trn user, and was seeing a lot of garbage in the original post, and I finally got irritated enough to dig into the code a bit.

Turns out that when trn checks for control characters in an article it assumes everything might be 7-bit with parity, so it strips off the top bit before checking! As a result it turns every extended byte between 0x80 and 0x9F into a control character which it then escapes! Hence the various garbage characters in any utf-8 or ISO-8859 article text.

I ended up with a simple fix, which anyone who's compiled their own should be able to quickly replicate. In util.h there are two macros:

#define AT_GREY_SPACE(s) ((*(Uchar*)(s) & 0x7F) = ' ' && *(s) != '\177')

which I switched to:

#define AT_GREY_SPACE(s) (*(Uchar*)(s) = ' ' && *(s) != '\177')

and I'm now getting my utf-8 text unadulterated. (My Terminal window is utf-8. I can easily switch it to 8859 if needed.)

Of course I'll be in trouble if I ever encounter any post that is actually 7-bit-with parity, but I wonder if this can ever happen these days? In the olden time of UUCP, I suppose that was standard, but I think everything is 8-bit these days.

Ironically, the only posts that are now getting munged are Ed Davies'! Whatever transformation his reader did seems to have put an 'underline-on' code into the Subject header (and also hides the newlines in trn's article list...).a [So I've edited the Subject to avoid this.]

In any case, I rather feel that it is good manners *not* to use either utf-8 or 8859 (or Windows-1252) in articles. There are just too many variant standards around. [I'm having the same irritation with the web these days. Sometimes I have to try 3 or 4 browsers before I find one that shows the site properly!]

-- Pete --