Catch up on stories from the past week (and beyond) at the Slashdot story archive

 



Forgot your password?
typodupeerror
×

Comment Re:Why make it RHEL compatible? (Score 2) 39

Someday that might be possible. For now RHEL is the enterprise standard for nearly all commercial and proprietary software packages, of which there are many out there that many enterprises rely on. Oracle Corporation produces hundreds. It is also the standard base for a great deal of internal software as well. The needs of both to have a long term stable base is why Redhat has a business. It will take a great deal of trust and experience for a new standard to supplant that, and the more incompatible things are the harder it is for anyone to migrate or support both. That is why so many vendors are trying to be RHEL compatible, at least for now.

Comment Re:Whose first? (Score 1) 75

It is hard to overestimate what an incredibly bad idea requiring a website to conduct age verification to display the sort of material that could be found in a public library would be. Mandatory age verification for services that provide or specialize in material that is overtly pornographic is probably the intent though, to the degree that those services are within the reach of UK law.

Comment Re:Not a crash (Score 1) 73

This is a large scale international network. Every node should be programmed to reject data from other nodes that does not meet basic integrity constraints rather than initiate some sort of nationwide shutdown. Of course if the data is quite important appropriate alarms and alerts should be raised so that the originator can supply properly formatted data or other corrective action can be taken.

Comment Re:Not a crash (Score 1) 73

An entire system going into a fail safe mode in response to a data entry error or data format validation error is indistinguishable from a crash. In fact it is actually worse than a crash. A *system* should never do that under any conditions. If the Internet operated that way it would shut down within milliseconds, and somehow I doubt the resulting inability to communicate with anyone or anything anywhere would make people happy just because the system wide outage was credited to a "fail safe" mode.

Comment Re:Goldman Sachs is an investment bank (Score 1) 59

Payments are one thing, paying interest is another. Paying interest on demand deposits is no mean trick. Banks fail because they cannot manage to do that on a regular basis. Deposits at "Apple" will not be FDIC insured either unless Apple affiliates with an FDIC member bank. Somehow I don't think Apple plans to fail the way SVB did or be associated with such a disgrace to their reputation either.

Comment Re:It was already gone from the Supreme Court (Score 3, Informative) 399

While the overwhelming consensus among legal scholars is that what Biden did is constitutional and legal the current court is extremely partisan and their expected to ignore that and rule against it.

That is far from the truth. The Constitution does not give the President the authority to dispose of a trillion dollars of federal assets without congressional approval, and the idea that the law for making minor adjustments does is preposterous.

Comment Re:Marketing (Score 1) 124

It is not a DSL so much as a restricted dialect of Python. I suppose they could use the technology to make true DSLs though, and it no doubt reduces confusion to refer to this one as a different programming language rather than as a general purpose drop in replacement for Python (which it most definitely is not).

The main reason I wouldn't consider it a domain specific language is because it is suitable for a number of different domains.

Comment Re:Non-Commercial Use Only? (Score 1) 124

No doubt they are hoping to commercialize it and don't expect (almost) anyone to use it for anything other than an academic exercise until then. It is such a great idea though, someone should consider making an open source (i.e. OSI compliant) implementation or something along similar lines.

Comment Re:Support C-API? (Score 2) 124

According to the documentation it is ASCII only internally for now, but supports Python extensions like NumPy with certain limitations. It has its own C FFI as well.

I strongly suspect that if you want maximum performance you should use the native C FFI rather than the support for Python extensions (although that apparently all works as long as you stick to ASCII strings), because there is a conversion across the Codon to Python boundary and Python extensions are probably supported from the Python side. Of course whether the conversion overhead is even noticeable would depend on what you were doing.

Comment Re:Only ASCII Strings instead of Unicode (Score 3, Interesting) 124

UCS-2 support has had the unfortunate effect that Microsoft has tended to be very late in supporting UTF-8. It barely supports it in Windows for example, and only on newer versions. SQL Server apparently introduced UTF-8 support in 2019. Java and Javascript have similar issues, although not as serious as the ones Windows has, because both have supported UTF-8 encoding for input and output for a long time.

As far as Python goes, it is pretty easy to see why those focused on high performance string processing applications like genomics would want to support 8-bit characters natively, using a compilation option or something. It no doubt makes a significant performance impact. In C you can have whatever character set you want if you stick to native encoding, and for string processing that is fast. Mandatory Unicode with code point semantics is bad news on the performance and efficiency front if what you really want to do is process 8 bit characters. UTF-8 has only taken over much of the world because a great deal of low level software is entirely unaware (i.e. transparent to the fact) that it is in use at all.

That goes for UTF-16 in a variety of contexts designed for UCS-2 as well. Codepoint semantics are relatively slow or inefficient in every programming language that supports them, and most don't and can't without breaking backward compatibility. So Python is ahead of its time in a way that gives it a performance disadvantage for some applications even when compiled and optimized to the hilt, unless someone takes non-standard narrow use expedients like this, or rewrites quite a bit of code to use byte support instead, which of course has a different set of compatibility issues.

Comment Linux Kernel 6.2 (Score 1) 40

There is no such thing as "Linux 6.2", although some distributions have releases with that version number. This is Linux Kernel 6.2. The difference is important. A great many devices use the Linux kernel, but not a Linux operating system in the normal sense of the term. Every Android phone in the world for example. Linux kernel, Android user space.

Comment Re:Spend is a verb not a noun (Score 1) 100

Once it's ratified by the dictionaries, it IS proper English. There's no need to make yourself look like a fuddy-duddy by complaining about such trivia. The world moves on.

On the contrary there is extensive commentary on what is proper usage that does not even begin to concede the issue when something is added to a dictionary, and the consensus is that "spending" is to be preferred over "spend" when a noun is required. If you are in an informal context in Britain that may be different, but here we are talking about a published article not a post it note in London.

Slashdot Top Deals

He has not acquired a fortune; the fortune has acquired him. -- Bion

Working...