Weekly Edition
Return to the Development page
| |||||||||||||
Poettering: [Google code search is] A Big Loss
Lennart Poettering covers a Google
announcement that Google Code Search will shut down in January. "I
think it must be of genuine interest to the Free Software community to have
a capable replacement for Google Code Search, for the day it is turned
off. In fact, it probably should be something the various foundations which
promote Free Software should be looking into, like the FSF or the Linux
Foundation. There are very few better ways to get Free Software into the
heads and minds of engineers than by examples -- examples consisting of
real life code they can find with a source code search engine. I believe a
source code search engine is probably among the best vehicles to promote
Free Software towards engineers. In particular if it itself was Free
Software (in contrast to Google Code Search)." (Thanks to Paul
Wise)
(Log in to post comments)
Bad News Posted Oct 17, 2011 20:06 UTC (Mon) by wahern (subscriber, #37304) [Link]
Google Code Search rocks, even though the corpus hasn't been expanded in years. Maybe Rob Pike will be allowed to release his source code (other than RE2), though probably not much of it would be useful outside of Google's environment. A volunteer and donation supported project could be put together to build something on top of EC2.
What makes Code Search so awesome is the ability to do pure regular expression searches instantly over the entire corpus. That requires lots of memory and cpu operating in parallel.
Bad News Posted Oct 17, 2011 22:22 UTC (Mon) by error27 (subscriber, #8346) [Link]
What kind of regular expressions do you search for?
I only have used it to see how functions are called. And the last couple times I tried to use it, it was broken. It's broken for me now actually.
Bad News Posted Oct 18, 2011 19:26 UTC (Tue) by wahern (subscriber, #37304) [Link]
I meant one can use regexes to search the corpus. I haven't yet needed to search for a regex. This works, though: ~\s*m\/foo\/ lang:perl
Bad News Posted Oct 19, 2011 18:37 UTC (Wed) by deepfire (subscriber, #26138) [Link]
RE2's author (if we're talking about the regular expression engine) is Russ Cox, not Rob Pike.
Covering the cover? Posted Oct 17, 2011 20:43 UTC (Mon) by mmcgrath (subscriber, #44906) [Link]
Why is LWN covering Lennart Pottering covering a Google code announcement?
Covering the cover? Posted Oct 17, 2011 20:49 UTC (Mon) by raven667 (subscriber, #5198) [Link]
My uninformed guess is that there wasn't much to add by writing their own copy on the story. A brief paragraph and a link allows the editors to work on other, more important, stories and updates. All those LWN weekly reports, security update notices, etc. don't research and write themselves.
Covering the cover? Posted Oct 17, 2011 21:08 UTC (Mon) by wingo (subscriber, #26929) [Link]
Hah! You mean, because Jon has a crush on Lennart? :)
More seriously, Lennart offers some positive things that free software folks can do, which is I guess the real story here. Keeping a 20GB index in memory should be possible for $10K/year, which we should be able to raise as a community, no?
Covering the cover? Posted Oct 18, 2011 0:11 UTC (Tue) by mezcalero (guest, #45103) [Link]
Well, they call it Lennart's Weekly News for a reason ;-)
Poettering: A Big Loss Posted Oct 17, 2011 23:13 UTC (Mon) by andikleen2 (guest, #52506) [Link]
Agreed on it being a big loss. I found codesearch extremely useful
and kept pointing people to it.
It was also often quite useful in kernel development to point how
Hopefully there will be a usable replacement. Maybe koders is motivated
Poettering: A Big Loss Posted Oct 18, 2011 10:46 UTC (Tue) by jwakely (subscriber, #60262) [Link]
> It was also often quite useful in kernel development to point how
> much an API is actually used.
And useful for the C++ standard committee to show how often an identifier is used, to get a feel for how much code might be broken by introducing a new reserved word. Obviously only a fraction of the world's code is visible to the search engine, but it has proven useful in the past to show that one possible spelling for a keyword would cause more problems than another.
Poettering: A Big Loss Posted Oct 17, 2011 23:47 UTC (Mon) by Duncan (guest, #6647) [Link]
That title is... unfortunate... in recent news context. =:^(
Glad my first intuit was incorrect!
Poettering: A Big Loss Posted Oct 18, 2011 4:07 UTC (Tue) by geuder (subscriber, #62854) [Link]
> That title is... unfortunate...
n++
Poettering: A Big Loss Posted Oct 18, 2011 4:15 UTC (Tue) by alison (✭ supporter ✭, #63752) [Link]
Indeed, I felt mildly sick when I read the item title. (No exaggeration: my stomach turned over.) Please use more discretion in picking headlines!
Poettering: A Big Loss Posted Oct 18, 2011 7:09 UTC (Tue) by flewellyn (subscriber, #5047) [Link]
Since the LWN editors just echoed the title used by Lennart Poettering, I think your feedback should be aimed at him.
Poettering: A Big Loss Posted Oct 18, 2011 7:34 UTC (Tue) by alison (✭ supporter ✭, #63752) [Link]
I believe that readers of Poettering's original posting were less likely to fear that he had died. Who knows, perhaps we are afflicted by email, software patents and Facebook even in the afterlife now? I certainly wouldn't be surprised to see Adwords on the clouds in heaven.
Poettering: A Big Loss Posted Oct 18, 2011 7:39 UTC (Tue) by khc (subscriber, #45209) [Link]
When I originally saw the post on Planet Gnome, I thought someone else had died.
Poettering: A Big Loss Posted Oct 18, 2011 7:40 UTC (Tue) by price (subscriber, #59790) [Link]
No, the unintended meaning comes about only when the person's name is added. "$person: A Big Loss" sounds like $person may *be* the loss, joining the subjects of certain other stories this month. "A Big Loss", which was Poettering's title, is much more general about what has been lost.
The combination is especially unfortunate when seen in, e.g., LWN's Twitter feed, with just the headline and a link.
(I had the same reaction as Duncan and alison.)
Poettering: A Big Loss Posted Oct 19, 2011 3:32 UTC (Wed) by flewellyn (subscriber, #5047) [Link]
Ahh, yes. I see where the issue comes from.
I note the helpful change of the LWN title, to reflect that. :-)
Poettering: A Big Loss Posted Oct 18, 2011 10:27 UTC (Tue) by salimma (subscriber, #34460) [Link]
Likewise. 'Poettering: "A Big Loss"' would have been less easily misinterpreted
Poettering: A Big Loss Posted Oct 18, 2011 11:05 UTC (Tue) by nowster (subscriber, #67) [Link]
"Google Code Search will shut down in January"
Just enough time for the functionality to be added to systemd, then?
Prior art searching Posted Oct 18, 2011 12:17 UTC (Tue) by SecretEuroPatentAgentMan (guest, #66656) [Link]
Specialised search engines like this appear useful for prior art searches, and while I have not had to do this myself I have kept Google code search in mind should the need arise. Also EPO (European Patent Office) in their Guidelines for examination refers to use of (not named) external search engines.
Not reviewing existing software, is a frequent critisism against patent examiners, now it just got a bit harder. And it is not as if documentation appeared to have the highest priority either.
Code search engines. Posted Oct 19, 2011 16:39 UTC (Wed) by przemoc (subscriber, #67594) [Link]
There are other services beside Google Code, but sadly no one is near as broad, efficient and flexible as GC. Some of them:
http://koders.com/ http://codefetch.com/ http://www.sourcecodeonline.com/ https://github.com/search
There are also narrowed services like:
Did I miss any important one?
Code search engines. Posted Jan 19, 2012 5:29 UTC (Thu) by tcf628 (guest, #82399) [Link]
You can try http://www.symbolhound.com/codesearch
There is a new source code repository search feature, in addition to the symbol-inclusive web search option.
-Tom (co-founder of SymbolHound)
Poettering: [Google code search is] A Big Loss Posted Oct 19, 2011 17:37 UTC (Wed) by wookey (subscriber, #5501) [Link]
Hmm. I'd never heard of this, and struggle to imagine what a code search service looks like. It lets me search for other people's code, right? But how do I specify what sort of code I'm looking for? 'conacatenating lines in perl'? 'what the hell can I put after the dot in python whatever.'? (That last one would be my favourite question).
Poettering: [Google code search is] A Big Loss Posted Oct 19, 2011 22:06 UTC (Wed) by bjartur (guest, #67801) [Link]
It's a full-text regex, keyword and tag search of publicly available source code. Think remote execution of find and grep over a large collection of source code.
Poettering: [Google code search is] A Big Loss Posted Oct 25, 2011 16:23 UTC (Tue) by penguin42 (subscriber, #72294) [Link]
Potentially, if google were just to add some tags to their main search then it might be as good; i.e. something like code: or the like or be able to search in specific file types and actually search for what you asked for.
|
|||||||||||||