wysiwyg is handy... but anything bigger than a
few lines of code and you'll start to feel limited if you stay in bash instead of going for
c/c++... same with documents...
I expect any significant writing I do to be in a structured markup language like XML.
Most of our departmental documentats are in DocBook.
They've been revised several times; they've been run through several revisions of stylesheets; they have not been revised for the sake of style.
TeX's typesetting is great, but the markup language is mostly presentational.
LaTeX is okay, but I expect (SG|X)ML + (CSS|XSL|DSSL) to be a better solution.
I want a TeX-like back end to CSS and/or XSL (FO).
It's sad that all this time after the introduction of TeX, Adobe InDesign can advertise paragraph-level justification as a new feature.
(Although I am intrigued by the supposed use of the hz program.)
they no longer care about 'other' Linux distributions, but assume
market leadership
I'm not a communist, but I don't trust for-profit companies.
The best thing about Red Hat is that they deal almost exclusively in free software.
(Their secure web server isn't free, but I imagine that will change, with the passing of the RSA patent.)
For better or worse, they've unleashed RPM on the world.
They've done extensive QA on gcc and many other packages.
I disagree with the decision to ship Red Hat 7.
Version 7 should have included a released version of Linux 2.4, the new gcc, and the corresponding glibc.
But that's their business model: periodic releases, where ++.0 > x.++.
Making a new version of Red Hat Linux is like printing money.
You don't do it too often, or you'll devalue it (and I think they may have done that with version 7).
On the other hand, you don't wait too long, because there's an opportunity cost.
you can vertically "letterbox" 5:4
on
Nvidia's NV20
·
· Score: 1
2560x2048 is the wrong aspect ratio.
2560x2048 and 1280x1024 are 5:4, which is only a little different from the more conventional 4:3.
To get your square pixels, adjust the horizontal size on your monitor.
I have a Voodoo3 and a 21" monitor.
An older version of the drivers didn't correctly support 1280x960.
At 1024x768, I didn't feel like I was getting my money's worth out of the monitor.
At 1600x1200, things didn't look right--there are too many pixel size dependencies in Windows and on the web.
I settled on 1280x1024 with a modest letterbox down the sides.
I agree that 4:3 is preferable for a conventional monitor, but I wouldn't call 5:4 "wrong."
For my thesis, I created a Web query system called ParaSite.
Thanks for the links.
This looks more compelling than the WebQL product.
The first example in the WebQL manual demonstrates a regex to extract phone numbers.
One of your first examples appears to implement the Google technique.
I find the latter more interesting.
I wouldn't exactly call that a simple interface!;-)
Like they say about Perl: it makes the easy things easy, and the hard things possible.
The average SQL user could probably learn WebQL syntax, although regexes can be complicated.
(I didn't realize how complicated until I read Mastering Regular Expressions.)
On the other hand, writing a web crawler in Perl may be beyond his reach.
That said, it wouldn't take much work to cook up a little language like this to wrap a Perl web crawler.
I certainly wouldn't pay $500 for this proprietary package.
I can't imagine someone making a "screen-scraping" search engine that returns bits of data and not just a link.
They will probably get sued by the owners of the purloined content.
I don't see any indication that Caesius intends to start such a search engine.
WebQL is just a web crawler.
If someone did, the primary defense would be fair use.
Some search engines already display an abstract in the search results.
On the other hand, I think eBay won a case against (or bullied into submission) a site that crawled their auctions.
U.S. courts don't seem to like deep linking, let alone data extraction.
Something about the God-given right to banner ad impressions.
Next thing you know, a U.S. Marshall will break down my door because I'm using the Internet Junkbuster proxy.
I did post anonymously, right?
I grab the html and apply various forms of on the fly text processing
I downloaded the WebQL Business Edition manual. Here's an abbreviated version of the first example query:
select
text("(\(206\)\s+\d{3}-\d{4})","","","T")
from
http://foo/bar.html
where
approach=sequence("1","10","1","XX")
The select clause accepts a variety of functions, of which text() seems to be the most useful.
You can see that the first argument is a regex designed to match phone numbers.
The from clause is an URL.
The where clause primarily takes the approach "descriptor," which can crawl or guess new URLs.
So basically, it doesn't do anything a Perl script can't. It just presents a simpler interface.
Article is probably mistaken.
on
Nvidia's NV20
·
· Score: 1
Dr. Merkwürdigliebe answered my question in another post: "I think they may be confusing it with the DDR-memory clockspeed which will be 500 MHz."
According to The Register, the NV20 will support "250MHz double data rate (DDR) SDRAM."
Where's the bottleneck?
on
Nvidia's NV20
·
· Score: 1
[The article] claims [the NV20] will be faster [than the GeForce2] as graphics get more complex
I don't know the details, but we see stuff like this in benchmarks all the time. If you're CPU limited, a faster video card won't buy you much. Likewise, texture compression and DDR RAM on a fast, wide bus won't buy you much if you're fill rate limited.
Granted, going from 2 to 7 times is a little unusual in practice, but it's perfectly normal for pre-release boasts.
What he said. Ping pong and Wednesday breakfast are no substitute for a simple recognition of your value, especially in this market. If you train someone, he's worth more. A raise is cheap compared to the cost of hiring and training someone new (who, in the meantime, will be doing a less effective job). If you don't recognize it, another companywill.
I'm convinced that turnover is the number one productivity killer at our company (complacency is number two). It's hard enough communicating with other departments. When your emails bounce because someone quit... well, that's frustrating.
We hired an inexperienced buffoon as a junior member. Didn't know Windows, Unix, programming, or anything. He read an O'Reilly book and got hired as a Java network programmer six months later. Makes me wonder what I'm still doing here.
why such a fast RAMDAC?
on
Nvidia's NV20
·
· Score: 1
the NV20 is said to have a ramdac speed of
over 500MHz
The current 350 MHz can support a decent refresh at 1600x1200. Does a consumer graphics card really need more than that?
While the images on the web are JPEGs, I reckon the library has the original in TIFFs. If another image format becomes feasible, they can always convert.
Indeed, I'm working on a crisis for a customer who runs a large web site with three Linux/Apache/PHP app. servers hitting one database server (via ODBC) running our DBMS.
the military vote still has to come in,
and that is generally Republican
In the first presidential debate, George Bush criticized our readiness and morale, saying we're over extended.
Gore scored points by immediately responding that we have not only the best military in the world, but in the history of the world.
Gore's comments sounded more presidential, like he knew that the little Hitlers of the world were listening.
The great unwashed can already buy cheap DVD players
Exactly. I don't get where this elitist backlash comes from. Steve says, "when everyone has a DVD player, we won't feel special anymore." He's been reading too much Widescreen Review.
Ask the betamax or laser disc aficionado whether he enjoys the exclusivity of his club.
After a crisis of conscience about the DeCSS suit, quickly followed by American consumerism and apathy, I bought a DVD player (a Sony, no less) for about $200. Maybe that makes me one of the unwashed.
I want to see more DVDs at the video store.
I want more titles on DVD.
I want simultaneous release on VHS and DVD.
I understand the purists concern that mass-market DVDs may make for cheap DVDs (poor video transfers, poor auding mixing, no extras, etc.).
That doesn't mean they feel threatened by the PS2.
So if you don't want your software to change with each new release, specify a version number!
Don't do this unless you have to. The version clause makes your software more useful.
Compare it to the clause of the LGPL that allows relicensing under the GPL. True, it lets someone make a more restrictive fork. Its intent is to allow linking with GPLed code.
If you prefer GPL 2 to 2.1 (or 3, etc.), continue to release your own code under version 2, with the version clause. Include a note in the README explaining your wishes.
If the BSD projects can avoid license forks, versioning the GPL shouldn't be a big deal.
The only change I support is amending the requirement to offer source code by mail order. Even that is trivial: it costs nothing to include a second CD with source code.
Bruce, I find your comment on linking puzzling:
"The GPL[2] concept of 'linking' has aged." The GPL derives its force from copyright law. Its "viral" nature comes from the legal concept of a derived work. RMS's opinion of DLLs, plugins, or daemons is irrelevant.
The GPL is an instance of copyleft--an attempt to diminish, not extend, the scope of copyright.
RMS is going to be pissed, but not because of the title of the article. The author actually gets it right in the text.
RMS will be pissed because the author attributes open source to the Free Software Foundation:
The GNU project and others had already duplicated, using an
open-source development model, much of the functionality of UNIX....
The goal was to produce a complete open-source UNIX-like OS...
Because of its pioneering of the open-source philosophy,... the FSF has the respect of many.
Remember RMS's reply to Jorrit, regarding an LGPLed "open source" 3D engine: "I don't support the Open Source Movement, so I can't have a discussion with you in the name of open source."
Of course, since the open source release is GPLed, porting it to other OSes is perfectly
legitimate.
Qt is in the same boat. TrollTech GPLed Qt for Unix, but has yet to release the Windows port. I'm not aware of any effort to fork a Windows port from the GPL release.
Harmony for Windows, anyone?
GNU GhostScript has always trailed Alladin by a year or so. I'm not aware of any extra features in the GNU version.
I haven't heard much from CUPS since it was first announced.
Their strategy was to GPL the engine and sell proprietary printer drivers.
MySQL had a liberal Unix license and an expensive Windows license for a while. When they GPLed the Unix version, I wondered whether someone would fork a Windows port.
Then the Windows version was GPLed.
Quake has been GPLed, but id will still sell you a license to use it in a proprietary product.
I doubt that id is worried that the GPL release of Quake will cut into any of their business.
In fact, they're limiting the Quake 3 licensees for the first year, to make an "exclusive club."
A GPL dual license is a gamble. Someone could add a compelling new feature to the GPL release that can't be integrated into the "value-added" proprietary version. There seems to be a strong resistance do doing so, however. Doing so would require a fork, because the company won't accept a contribution unless you assign copyright.
This lessens the contribution to the free software community, but I'm not about the smell the gift fish.
I'm lucky; in my two jobs as a software tester, I've actually gotten written descriptions from the developer of what he thinks he did.
Requirements? Come on.
I've read about software requirements, but I've never seen them.
Often, the rationale in the developer description is, "Because [the president] said so."
Our group had been planning for a long time to rewrite the bug database.
Unbeknownst to us, a new developer's "sink or swim" induction was to rewrite it.
Was he privy to any of our requirements? No.
Now, we're reluctantly using it.
Tom DeMarco, Tim Lister, Ed Yourdon, Larry Constantine; Grady Booch, James Rumbaugh, Ivar Jacobson.
Software engineering scholars? No, science fiction authors.
the subrelease quality of most software on Freshmeat is not sufficient - and Mr. Public does not want to compile his own software.
Chilliware is marketing shovelware, the rip-off-at-half-the-price crap that screams at you pitifully from the bargain bin with the garish colors of laundry detergent boxes.
Why would I pay "$79 or less" for a contact manager or a desktop publisher.
A 117-person company with a hundred apps.? I dare you to call Support with a question on app. #42.
I bought DeScribe for OS/2 almost ten years ago, because it was a native app.
That was probably its only marketable feature.
I didn't buy StarOffice when it was proprietary.
I didn't buy Applix Office.
I'm not going to buy Chilli Office.
Will I buy Corel Office? Yes, maybe.
My experience with proprietary software is that you should buy big.
Buy from the company that's going to be around tomorrow.
Bend over when it's time to upgrade.
Why does Chilli see an opportunity?
Because free software projects rarely compete for users.
They're big on "scratch an itch" development, decent on "all bugs are shallow" development, but really weak on testing, release engineering, and documentation.
There are some potential solutions:
start a company like Red Hat or Helix and sell garish boxes with the word "official" stamped on them
brainwash someone into having an itch for a really easy to install version of, say, GNU Cash for Windows, Mac, and 31 flavors of Unix
have a bake sale so that Damian Conway's wife can take a year off to package all his software
Why are we making him out to be the free
software nazi?
Exactly. Jorrit, bless his heart, kept saying, "but, but you wrote the LGPL...." RMS is not a lawyer! I doubt that he'd even be called as an expert witness in a GPL infringement case.
Yes, his obtuse answers are frustrating; but can you imagine the volume of email he gets on this topic?
He once wrote that it's like putting a fire hose in your mouth.
Slashdot Mirror
I expect any significant writing I do to be in a structured markup language like XML. Most of our departmental documentats are in DocBook. They've been revised several times; they've been run through several revisions of stylesheets; they have not been revised for the sake of style.
TeX's typesetting is great, but the markup language is mostly presentational. LaTeX is okay, but I expect (SG|X)ML + (CSS|XSL|DSSL) to be a better solution. I want a TeX-like back end to CSS and/or XSL (FO).
It's sad that all this time after the introduction of TeX, Adobe InDesign can advertise paragraph-level justification as a new feature. (Although I am intrigued by the supposed use of the hz program.)
I'm not a communist, but I don't trust for-profit companies. The best thing about Red Hat is that they deal almost exclusively in free software. (Their secure web server isn't free, but I imagine that will change, with the passing of the RSA patent.) For better or worse, they've unleashed RPM on the world. They've done extensive QA on gcc and many other packages.
I disagree with the decision to ship Red Hat 7. Version 7 should have included a released version of Linux 2.4, the new gcc, and the corresponding glibc. But that's their business model: periodic releases, where ++.0 > x.++. Making a new version of Red Hat Linux is like printing money. You don't do it too often, or you'll devalue it (and I think they may have done that with version 7). On the other hand, you don't wait too long, because there's an opportunity cost.
2560x2048 and 1280x1024 are 5:4, which is only a little different from the more conventional 4:3. To get your square pixels, adjust the horizontal size on your monitor.
I have a Voodoo3 and a 21" monitor. An older version of the drivers didn't correctly support 1280x960. At 1024x768, I didn't feel like I was getting my money's worth out of the monitor. At 1600x1200, things didn't look right--there are too many pixel size dependencies in Windows and on the web. I settled on 1280x1024 with a modest letterbox down the sides.
I agree that 4:3 is preferable for a conventional monitor, but I wouldn't call 5:4 "wrong."
Thanks for the links. This looks more compelling than the WebQL product.
The first example in the WebQL manual demonstrates a regex to extract phone numbers. One of your first examples appears to implement the Google technique. I find the latter more interesting.
Like they say about Perl: it makes the easy things easy, and the hard things possible. The average SQL user could probably learn WebQL syntax, although regexes can be complicated. (I didn't realize how complicated until I read Mastering Regular Expressions.) On the other hand, writing a web crawler in Perl may be beyond his reach.
That said, it wouldn't take much work to cook up a little language like this to wrap a Perl web crawler. I certainly wouldn't pay $500 for this proprietary package.
I don't see any indication that Caesius intends to start such a search engine. WebQL is just a web crawler.
If someone did, the primary defense would be fair use. Some search engines already display an abstract in the search results. On the other hand, I think eBay won a case against (or bullied into submission) a site that crawled their auctions. U.S. courts don't seem to like deep linking, let alone data extraction. Something about the God-given right to banner ad impressions. Next thing you know, a U.S. Marshall will break down my door because I'm using the Internet Junkbuster proxy. I did post anonymously, right?
I downloaded the WebQL Business Edition manual. Here's an abbreviated version of the first example query:
The select clause accepts a variety of functions, of which text() seems to be the most useful. You can see that the first argument is a regex designed to match phone numbers. The from clause is an URL. The where clause primarily takes the approach "descriptor," which can crawl or guess new URLs.So basically, it doesn't do anything a Perl script can't. It just presents a simpler interface.
According to The Register, the NV20 will support "250MHz double data rate (DDR) SDRAM."
I don't know the details, but we see stuff like this in benchmarks all the time. If you're CPU limited, a faster video card won't buy you much. Likewise, texture compression and DDR RAM on a fast, wide bus won't buy you much if you're fill rate limited.
Granted, going from 2 to 7 times is a little unusual in practice, but it's perfectly normal for pre-release boasts.
What he said. Ping pong and Wednesday breakfast are no substitute for a simple recognition of your value, especially in this market. If you train someone, he's worth more. A raise is cheap compared to the cost of hiring and training someone new (who, in the meantime, will be doing a less effective job). If you don't recognize it, another companywill.
I'm convinced that turnover is the number one productivity killer at our company (complacency is number two). It's hard enough communicating with other departments. When your emails bounce because someone quit ... well, that's frustrating.
We hired an inexperienced buffoon as a junior member. Didn't know Windows, Unix, programming, or anything. He read an O'Reilly book and got hired as a Java network programmer six months later. Makes me wonder what I'm still doing here.
The current 350 MHz can support a decent refresh at 1600x1200. Does a consumer graphics card really need more than that?
Sounds like Grand Prix Legends. Although Papyrus recently released a Direct3D patch, it seems to work best in Glide.
While the images on the web are JPEGs, I reckon the library has the original in TIFFs. If another image format becomes feasible, they can always convert.
Indeed, I'm working on a crisis for a customer who runs a large web site with three Linux/Apache/PHP app. servers hitting one database server (via ODBC) running our DBMS.
Guess which component is failing?
In the first presidential debate, George Bush criticized our readiness and morale, saying we're over extended. Gore scored points by immediately responding that we have not only the best military in the world, but in the history of the world.
Gore's comments sounded more presidential, like he knew that the little Hitlers of the world were listening.
Exactly. I don't get where this elitist backlash comes from. Steve says, "when everyone has a DVD player, we won't feel special anymore." He's been reading too much Widescreen Review. Ask the betamax or laser disc aficionado whether he enjoys the exclusivity of his club.
After a crisis of conscience about the DeCSS suit, quickly followed by American consumerism and apathy, I bought a DVD player (a Sony, no less) for about $200. Maybe that makes me one of the unwashed. I want to see more DVDs at the video store. I want more titles on DVD. I want simultaneous release on VHS and DVD. I understand the purists concern that mass-market DVDs may make for cheap DVDs (poor video transfers, poor auding mixing, no extras, etc.). That doesn't mean they feel threatened by the PS2.
Don't do this unless you have to. The version clause makes your software more useful.
Compare it to the clause of the LGPL that allows relicensing under the GPL. True, it lets someone make a more restrictive fork. Its intent is to allow linking with GPLed code.
If you prefer GPL 2 to 2.1 (or 3, etc.), continue to release your own code under version 2, with the version clause. Include a note in the README explaining your wishes.
If the BSD projects can avoid license forks, versioning the GPL shouldn't be a big deal.
Bruce, I find your comment on linking puzzling: "The GPL[2] concept of 'linking' has aged." The GPL derives its force from copyright law. Its "viral" nature comes from the legal concept of a derived work. RMS's opinion of DLLs, plugins, or daemons is irrelevant.
The GPL is an instance of copyleft--an attempt to diminish, not extend, the scope of copyright.
Gosh, what have I been doing at work for the last three years without these programs?
RMS is going to be pissed, but not because of the title of the article. The author actually gets it right in the text. RMS will be pissed because the author attributes open source to the Free Software Foundation:
Remember RMS's reply to Jorrit, regarding an LGPLed "open source" 3D engine: "I don't support the Open Source Movement, so I can't have a discussion with you in the name of open source."A GPL dual license is a gamble. Someone could add a compelling new feature to the GPL release that can't be integrated into the "value-added" proprietary version. There seems to be a strong resistance do doing so, however. Doing so would require a fork, because the company won't accept a contribution unless you assign copyright. This lessens the contribution to the free software community, but I'm not about the smell the gift fish.
That's nothing. Yahoo is putting Larry Wall's daughter through college.
Our group had been planning for a long time to rewrite the bug database. Unbeknownst to us, a new developer's "sink or swim" induction was to rewrite it. Was he privy to any of our requirements? No. Now, we're reluctantly using it.
Tom DeMarco, Tim Lister, Ed Yourdon, Larry Constantine; Grady Booch, James Rumbaugh, Ivar Jacobson. Software engineering scholars? No, science fiction authors.
Chilliware is marketing shovelware, the rip-off-at-half-the-price crap that screams at you pitifully from the bargain bin with the garish colors of laundry detergent boxes. Why would I pay "$79 or less" for a contact manager or a desktop publisher. A 117-person company with a hundred apps.? I dare you to call Support with a question on app. #42.
I bought DeScribe for OS/2 almost ten years ago, because it was a native app. That was probably its only marketable feature. I didn't buy StarOffice when it was proprietary. I didn't buy Applix Office. I'm not going to buy Chilli Office. Will I buy Corel Office? Yes, maybe. My experience with proprietary software is that you should buy big. Buy from the company that's going to be around tomorrow. Bend over when it's time to upgrade.
Why does Chilli see an opportunity? Because free software projects rarely compete for users. They're big on "scratch an itch" development, decent on "all bugs are shallow" development, but really weak on testing, release engineering, and documentation. There are some potential solutions:
Exactly. Jorrit, bless his heart, kept saying, "but, but you wrote the LGPL ...." RMS is not a lawyer! I doubt that he'd even be called as an expert witness in a GPL infringement case.
Yes, his obtuse answers are frustrating; but can you imagine the volume of email he gets on this topic? He once wrote that it's like putting a fire hose in your mouth.