Posts by Bronek Kozicki

Re: I did my studying back then

The thing about OO is that it is being used to either replace or supplement composition with excessive levels of abstraction. Which comes from poor education - the abstractions are useful, e.g. for dependency injection or compilation firewalls, but not nearly as often as OO evangelists made us believe. You can have a perfectly testable non-trivial program, which for most of dependencies employs simple composition (possibly sprinkled with generics), rather than interface classes etc. tools of OOP.

The missed education bit was to learn why and where abstractions are useful, e.g. open-closed principle, dependency injection etc. Instead, programmers are handed tools and commandments (e.g. define your interfaces! use design patterns for everything!) without clear explanation of the purpose, and the results are poor indeed. Because, as it turns out, simplicity is a virtue - and it got lost along the way.

Re: That is fine provided that there is an exception mechanism

I do not understand why downvotes. It is not as if you are asking for higher allowance because you are picky.

... amen to that. Can't live without my keyboard, but the apps are nowhere close to where I want them to be.

Re: Power systems

As someone writing the SW, I am very keen to have a choice of HW platforms where my code will be running on. It is not a wanton wish. Everyone makes assumption that intel architecture will be "forever" because it is most popular, but do not forget that use of 32bit x86 is declining, and 64bit extensions were brought by AMD (not intel) and it was despite intel, not thanks to them. All the good work that intel is currently doing notwithstanding, we must not allow ourselves to become hostages to one platform.

Also, SW must scale across many cores. Sometimes there will be few, sometimes tens, sometimes thousands (Volta GPU) - and there are different ways to useful employ these. It is healthy to have some exploration in this space - should I make use of modern C11 memory model? Should I use consume semantics? Or perhaps try transactional memory? These choices set the scalability limits to the software I am/will be writing, and they should not be dictated by only one R&D department.

EDIT: I do agree that cloud department at IBM is little lazy - I wanted to hire a small VM to try POWER8 myself but was confused as to what I would be buying. The sales ppl never got back to me ...

Re: How long before...

(with thanks to #define INFINITY -1)

Directly from kernel maintainers, as reported in LWN:

If people are acting outside of the social contract, perhaps their code could be rewritten and replaced.

Radcliffe agreed, saying that ripping out the code would provide a warning to the next person that considered doing this kind of thing.

Re: No need to go as far as XZ1

I've read that starting the camera app on XA1 took few seconds - has this been fixed? Asking because missus needs a new 5-inch phone with good camera, and she likes Sony ones.

Re: New tropes...

Agile - do away with the subject matter "experts" and let the developers learn the subject, from the users.

Re: How is SQL Injection Still a Thing?

Actually SQL is not that bad and it definitely is not equivalent of COBOL. The problem is not SQL; it is programmers who think that building SQL programmatically by string concatenation is acceptable. It is not and never was. That's what stored procedures with parameters passing are for, or failing that pre-compiled statements, but of course that requires a little more varied code on the caller side. In effect there are two programming models here: 1) actual database queries and manipulation language in SQL and 2) database interface with procedures/pre-compiled statements, parameters etc. required to use 1) from other languages.

Programmers often perceive 2) as an optional element but it is not and as you have demonstrated above, it is easier to rant about 1) than learn to use it properly.

FWIW I do not use SQL now and have not used it for years, but frankly how can anyone aspire for "better languages" without understanding of what went before?

Re: Open source

ABAP , T-SQL as two of many examples. Basically as a programmer you are given access to language documentation, programming, and runtime environment, but you cannot peak underneath and are not allowed to influence its future evolution.

Faster wi-fi ... as long as you use our routers, our phones, our tablets, etc?

I suspect not, more likely they will share their implementation of 802.11ac firmware with whatever they have in their access points already. Which is standard conforming and works well with all types of devices, nothing proprietary here. However good (or bad) implementation can greatly impact actual achievable throughput and latency, simply because these two are tied to timing, which itself is tied to performance of the code itself (i.e. the standard is quite forgiving because it has to, but the implementation can aim for the best performance)

FWIW I do have some networking kit from TP-Link (e.g. CAP1750 access point, some managed switches, wifi controller etc) and it is good piece of kit - or excellent if you look at the (very small) price tag. They do not spend money on fancy features or loud marketing, and sometimes the hardware may be less robust (looking at old "nano" powerline adapters ...), but it is refreshing compared to Netgear, D-Link, Linksys etc. usual suspects.

Re: Well at 3nm it's a case of...

...the oxide thickness is usually 1/10 that so 1-2 atoms thick. I suspect this will go up, as the current has to flow somewhere. Curious what the new process will look like ...

Re: Aukward moment

FWIW, Howard Hinnant wrote nice C++ library for parsing timezone database files. It also handles conversion to/from GPS and TAI clock . The size of the thing is good indication that this is not simple at all, for those interested link (it is modern C++ and integrates with std::chrono , do not blame me if you have not been following standardization ...)

Re: Root KSK roll

I do not quite understand. It does not appear that the new KSK is somehow linked to the new signing algorithm ed25519 , or at least the article is silent on this. From what I read, it is a new DNSSEC signing key for the root servers, so as long as your (new or not so-new) DNS server has the new DNSSEC keys installed, it should just work. The RFC 8145 seem to be about enabling the root (and not so root) DNS servers to build up a knowledge about who has the new key and who has not, and it does not need to be installed everywhere. Neither seem related to the new key signing algo.

Unless the requirement of the new key is the use of the new signing algo, which would be a pretty important point to make in the article (and yet it is no there)?

Re: Cmon, have guts

... Iran? Sudan?

I think every developer worth his/hers salt wrote some truly awful software in their young age. Bill Gates was unlucky to have that software used by virtually everyone, for a very long time. If people all around the world started using code I've written 25 years ago I would probably die of embarrassment.

Re: Not Qualified

If I was on her place I would keep that visible. The fault is on the board of directors for putting her in a position show was not qualified to. Since that information is was public, there is no suspicion of her trying to pass the qualifications she did not have.

Re: making the planet darker than fresh asphalt!

... or simply because all these starts and landings leave lots of soot anyway

Re: TBH

Actually locking the color scheme of LIVE version would be even more important - imagine the mayhem if a user decided that TEST version scheme is so nice, he wants it on LIVE one. Then few weeks down the line proceeds to use LIVE as if it was TEST, because it looks like one.

Re: Streaming an operating system?

The pre-requisite here is that you need to trust both the network and the server, because otherwise your OS image becomes untrusted. I am not entirely sure I would trust mobile networks more than I (have to) trust Google or Apple.

Re: Typical problem of many large organizations

@Peter Getherhole exactly, you have won "compartmentalization" word in the bullshit bingo. The sad truth is that many C-level executives only know the words, but do not know the meaning.