🌐 Mailman 3 April 2000 - Python-list - python.org

Starship update: April 18
by akuchlin＠mems-exchange.org May 9, 2000

May 9, 2000

This is an update on the current status of starship.python.net; here is bad news, very bad news, and a little bit of good news. First, the bad news: The hard drive on starship is toast; it makes nasty rattling noises when the power is turned on. The very bad news is that there do not seem to be any backups, because no one ever volunteered to set them up. The most recent data we have is the data that was on an older disk drive, which went out of service back in 1998. The good news: a new disk has been put into the machine, a clean install of Linux done, and the machine should be online again tomorrow. Creating user accounts will take longer, of course. These newly created accounts will all have empty directories, since the data hasn't been restored. Setting up the Web server again will take longer still. A data recovery firm has been contacted to reconstruct the data from the hard drive; their service costs around US$300-1000 if the data is recovered successfully, and is free if nothing is recovered. We'll probably wind up asking for donations to cover the cost, if they manage to recover the data. In the meantime... * If you can recover your starship data through other means (google.com's caching, your Web browser cache, copies on your own computer, etc.), please do so immediately, before the caches get expired. * If you had downloadable files on Starship (such as software) which you no longer have copies of, you might want to post to comp.lang.python or the appropriate mailing list and ask if anyone grabbed a copy recently. * If you downloaded someone else's software from Starship before the crash, DO NOT DELETE IT; your copy might be the only one left. This is particularly true for packages that are no longer under active development; you might consider e-mailing the author to check if they still have a copy. Further updates will be posted as events warrant. -- A.M. Kuchling http://starship.python.net/crew/amk/ Our sister died. But we never took revenge. We knew there'd be trouble one day, her being mortal. It's not as if we never said anything. We still miss her. We still mourn. -- Stheno, in SANDMAN #60: "The Kindly Ones:4"

4 3

RE: Using python on the web
by Brent Fulgham May 9, 2000

May 9, 2000

A couple of questions: > We build medium-busy, CGI-intensive sites in Python. We were briefly > quite excited about Zope, but once we realized it locked us into Yet > Another Programming Language, we lost interest. DHTML is particularly > problematic, 1. I'm not familiar with Zope. Is DHTML specific to Zope? It seems to be proudly advertised on a lot of "webware" brochures I see these days. > The biggest hassle for us has been the weak MySQL support in > Python. The spec keeps changing and the interface is raw. 2. Have you tried using PostgreSQL? MySQL may be fast, but it's really little more than a filesystem represented as a database. It has no transaction support whatsoever which tends to lead to problems when you have a community-style site that has multiple parties trying to add messages, etc. Have you had much success using it in a production environment? I realize Slashdot and I think Sourceforge do, but they have their share of troubles. Of course, the Postgres Python bindings may be just as bad from your perspective. :-) Thanks, -Brent

11 11

[UNIX] Detaching from parent process
by Florian Weimer May 8, 2000

May 8, 2000

I'm trying to implement a CVS loginfo script with Python. The script has to fork to the background, because the parent process has to release the repository lock. But it seems as if Python doesn't detach completely from the parent process, because CVS exists only after the lengthy operation has completed. Currently I use the following code: if os.fork() == 0: os.setsid() if os.fork() == 0: os.close(0) os.close(1) os.close(2) os.open("/dev/null", os.O_RDWR) os.dup(0) os.dup(0) # Lengthy operation claiming the repository lock follows. -- Florian Weimer Florian.Weimer(a)RUS.Uni-Stuttgart.DE University of Stuttgart http://cert.uni-stuttgart.de/ RUS-CERT +49-711-685-5973/fax +49-711-685-5898 http://ca.uni-stuttgart.de:11371/pks/lookup?op=get&search=0xC06EC3B5

2 1

BAD starship.python.net news!
by Manus Hand May 6, 2000

May 6, 2000

This from Mike McLay. Everyone, let's pony up what it takes to recover our starship! Manus (who will backup more and more often than he has until now) > From: Mike McLay (mclay(a)nist.gov) > Subject: Re: starship.python.net?? > > The prognosis isn't good. The old drive is toast and backup software > was never installed. I've located a data recovery company that will > provide a quote on the chances of recovery for $125. The cost of > recovery, if it is possible, will be between $300-$1000. They can > turn around the recovery in a few days time. Someone will need > to fund the recovery. So far I have a $200 pledge. > > I bought a new drive and installed it in the server last night. The > system will be back online tomorrow by noon.

5 6

RE: Stackless/microthreads merge news
by Brent Fulgham May 6, 2000

May 6, 2000

I've now visited and read the information on both Stackless Python and Microthreads, and I must say they both seem like excellent features. Has any thought been given to making the generic Python use the Stackless implementation? It seems that if the stackless implementation is binary-compatible with "regular" Python, there shouldn't be too much of a problem. But are there other issues I haven't seen addressed in this forum? At any rate, both of these projects seem to have produced some excellent tools. I look forward to playing with the microthread implementation in particular. :-) Thanks, -Brent

13 27

Do I always have to write "self." ?
by Louis M. Pecora May 5, 2000

May 5, 2000

I am two weeks plus into learning Python and have started on a first real project. Amazing language. BUT, now that I've gotten going with classes, I'm finding I hate seeing all those "self." prefixes on the class variables. Python is generally very clean, but the classes I write use a lot of "per-instance" variables, hence seem to require self. As a result the expressions can become cluttered and harder to read than C++ in this particular case -- imagine that! Is there any relief here? Any (simple) way around writing that self. every time? Thanks. Love Python so far otherwise.

22 56

Python in the future
by j vickroy May 5, 2000

May 5, 2000

Hello all, I'm relatively new to Python and am trying to gain support for it at my work place. To that end, I'm giving a brief (overview) presentation. One of the items I have been asked to consider is where will Python be 5-10 years from now. I certainly don't know where C++ or Java will be in that time frame! Any ideas on where to research the future plans for Python would be appreciated. Thanks for your time, - jv

16 16

raw_input in loop
by A[r]TA May 4, 2000

May 4, 2000

Heyz, I want to write a loop in which there's something like raw_input but which doesn't stop the loop until the enterbutton is hitten. How do I do that??? A[r]TA -- We shall not cease from exploration, and the end of all our exploring will be to arrive where we started and know the place for the first time. - T. S. Eliot

5 4

HTML parser example, anybody?
by haering＠sunhalle8.informatik.tu-muenchen.de May 4, 2000

May 4, 2000

Hi! I want to transform my own HTML templates (with a few template tags, special comments, and so on) into final HTML for the browser. I would like to use Pyhton's htmllib, but from the docs I can get no clue on how to get it done. I seem to have to use a writer and a formatter class ... I would greatly appreciate an example of HTML -> HTML using htmllib/sgmllib. Can anybody provide pointers/code snippets? Gerhard

6 5

Q: Recursive embedded Python interpreters?
by mdefreitas＠sikorsky.com May 4, 2000

May 4, 2000

I am working on a helicopter simulation user-interface that takes commands such as: > start_sim > freeze_sim > altitude = 100.0 > end_sim I would like to add scripting capabilities to this user-interface, so I am attempting to extend and embed Python. I am adding a new command (python) to my user interface that will run a python script with optional arguments: > python script.py arg1 arg2 My first step was to extend python with a module ui that python scripts can import to execute various user interface commands. This was fairly easy. I am now trying to embed the python interpreter into my user interface. I anticipate a possible problem when a python scripts executes the ui command to run another python script. I anticipate re- entry issues. I assume I have to create a new interpreter for each nested call to the python interpreter. I really dont understand the mechanics of how this works, but from comp.lang.python archives, and from the Demos/pysvr example that came bundled with the python1.5 distribution this is what I think I have to do: static int nest_level = 0; static PyThreadState *gtstate = 0; void interp(char *script) { if (nest_level == 0) { nest_level++; // init topmost interp Py_Initialize(); PyEval_InitThreads(); gtstate = PyEval_SaveThread(); } else { PyEval_ReleaseLock(); // nested interp needs to relinquish lock? } PyEval_AcquireLock(); PyThreadState *tstate = Py_NewInterpreter(); FILE *fd = fopen(script, r); PyRun_SimpleFile(fd, script); Py_EndInterpreter(tstate); PyEval_ReleaseLock(); nest_level--; if (nest_level == 0) { PyEval_AcquireThread(gtstate); // top-most interp... clean up gtstate = 0; Py_Finalize(); } else { PyEval_AcquireLock(); // need to reacquire lock if nested? } } Is this right? Did I miss anything? Also how do I pass arguments to the (possibly nested) scripts. Thanks in advance for any help. Sent via Deja.com http://www.deja.com/ Before you buy.

2 9