|
1. I have about 400 headers, and there are much larger projects than that. Want to start?
2. As software evolves, #includes get added. Want to maintain that?
3. As I mentioned yesterday, you can remove an #include for a header that is included transitively. But doing so can break the compile if a file that you #include removes an #include that gave you this transitive embedding.
The only way to make this painless is with a tool that analyzes dependencies so that it can update #include directives automatically.
|
|
|
|
|
Greg Utas wrote: The only way to make this painless is with a tool that analyzes dependencies so that it can update #include directives automatically.
I didn't know that such a tool exists. Thanks for the tip.
The difficult we do right away...
...the impossible takes slightly longer.
|
|
|
|
|
It seems to me no matter how headers are #include'd there is an order w/ a top and a bottom. Perhaps it is tree-like. In which case each branch may have different needs at a common root. This would complicate things but can still be automated per FbD's idea as near as I can discern. Specifically if a root of one branch does not require a particular file but another branch w/ the same root does, the automation keeps track of this and #include's it for both branches.
|
|
|
|
|
A similar issue I have with C# is that I use CSC (the command-line C# compiler) with response files -- some response files include other response files, etc.
As my code has evolved, I may have some code files included in a response file when they are no longer needed.
Every once in a while I think about writing a tool which will eliminate references to code files which are no longer needed in a particular build.
|
|
|
|
|
I do not know C# or what a "response file" is. If they are anything like C++ #include's I do not see why a simple awk program which would execute the suggested procedure would not solve the problem at the press of a button.
|
|
|
|
|
Well, right, definitely not AWK.
|
|
|
|
|
May I please inquire why not awk. It is the only text processing language I know. Would you recommend another.
|
|
|
|
|
I think it depends on your target audience.
If you're targeting folks that primarily develop on unix, awk may be the way to go, as most distributions have it installed, or it's easily addable.
With WSL under windows, you *can* add it, but if you're not already set up for it it's more of a pain.
If my target was Windows, I'd consider another programming environment, like C#.
It I truly wanted it to work easily on any platform, I'd write it in C++, and compile it for common platforms + distribute the source .
That being said, sometimes I do use C# on linux because Mono is often shipped with linux distributions. It can be the quick and dirty route to make something easily usable on common platforms.
To err is human. Fortune favors the monsters.
|
|
|
|
|
C# of course.
I don't use UNIX.
|
|
|
|
|
I do not use unix either. I work on Windows. I recall not having too much trouble installing awk for Windows. I suppose if I were to use a general purpose language I would rely on a regular expression library which C++ provides. I wrote a nifty little awk program which indents my source after Visual Studio mangles the indentations the result of my insisting macros embed the statement terminator ';' as it looks ugly otherwise which unfortunately causes VS IntelliSense to mangle the indentations. As I recall I wrote a little tip here re/ same.
|
|
|
|
|
Amazing l just now noticed Edge utiIizes a font in the address bar for which capitaI i and smaII I are indistinguishabIe Just as here
|
|
|
|
|
|
OTOH, after the page has loaded (which is a little bit too late, I guess) the browser removes the capitalization...so while silly.com (lowercase "el") brings you to some sort of Hello world blog, siIIy.com (two uppercase "eyes") brings you to some Chinese page (and the title apparently translates to "site note found"), but the address bar then clearly shows "siiiy.com"
I don't really see a clean way out of this, unless domain names suddenly have to be recognizable dictionary names (which would break half of them)...
|
|
|
|
|
A partial way out which is why I am amazed by this is our browsers are supposed to protect us as best they can. The least they can do is choose a font which does not help malicious actors.
|
|
|
|
|
meagreProgrammer wrote: The least they can do is choose a font which does not help malicious actors.
Devil's advocate:
The user gets to set his own font preferences at the OS level. That can include some specific accessibility requirements. If the browser ignored those choices and started arbitrarily using some alternate font for a control such as the address bar, you would annoy those users - why should the browser's controls be any different than the OS itself? All apps running on the OS should respect what's been selected there.
(Page content is another mess altogether, outside the control of the OS and best left for another discussion.)
All that said, I'm not disagreeing with you...the browser is the doorway to potentially very nasty stuff, so yes, it should go out of its way to validate everything. I'm just not sure how you're supposed to guess the user's intent...to re-use my previous example, maybe I did mean to go to siIIy.com (using two capitalized "eyes") rather than silly.com (two lowercase "els").
OTOH, as domains aren't supposed to be case-sensitive, and a link being followed contains capitalized letters as part of the domain...warn the user and show him the domain as all lowercase? Or at least, have the hover text show the lowercase version? But if the user types in a domain name with capitalized letters...then he only has himself to blame.
Food for thought for sure...
|
|
|
|
|
dandy72 wrote: OTOH, as domains aren't supposed to be case-sensitive, and a link being followed contains capitalized letters as part of the domain...warn the user and show him the domain as all lowercase? Or at least, have the hover text show the lowercase version? But if the user types in a domain name with capitalized letters...then he only has himself to blame.
DNS is case insensitive by RFC, so domain names are case insensitive. What isn't case insensitive is email addresses before the @ sign, but all major mail servers treat them as case insensitive.
|
|
|
|
|
obermd wrote: DNS is case insensitive by RFC
Yep, that's true for the old RFC. However with aggressive DNSSEC the query will randomly capitalize the domain; (wWw.cODeProJecT.coM for example) and the upstream authoritive server should preserve the capitalization in the response.
I think capitalization is used in other areas such as the NSEC record, but I don't remember the details.
|
|
|
|
|
The domain name is still case insensitive. This capitalization is designed to increase the entropy of the encrypted DNS record, not add to the number of domain names.
|
|
|
|
|
obermd wrote: What isn't case insensitive is email addresses before the @ sign
I was not aware of that.
"I'm Bob@company.com, but make sure you capitalize the first B, otherwise it's gonna go to that jerk in accounting".
Good thing that is NOT being respected. I'm not familiar with the whole RFC process...would an RFC this old (now that there's plenty of implementations out there) allow for an addendum/correction that goes against what's been previously stated...? Just to make it formal...especially since, as you say, "all major mail servers" treat them as case-insensitive...
|
|
|
|
|
Thanks - I'm going to turn this into a security tip for my company.
|
|
|
|
|
Oddly, I can tell the difference, just barely. If you screen capture Il (the difference here is obvious when I type it in the editor but rendered in the final post and in the preview, they look the same) you'll notice the antialiasing (or is it aliasing?) is different for uppercase I and lowercase l. Maybe somehow I'm seeing that subtle difference. 
|
|
|
|
|
@GregUtas
Where's the CCC? 
"I have no idea what I did, but I'm taking full credit for it." - ThisOldTony
"Common sense is so rare these days, it should be classified as a super power" - Random T-shirt
AntiTwitter: @DalekDave is now a follower!
|
|
|
|
|
Bit early for Greg 
Life should not be a journey to the grave with the intention of arriving safely in a pretty and well-preserved body, but rather to skid in broadside in a cloud of smoke, thoroughly used up, totally worn out, and loudly proclaiming “Wow! What a Ride!" - Hunter S Thompson - RIP
|
|
|
|
|
...Present company not included, of course
🤣🤣🤣🤣🤣 Whoever wrote that (who was it, guys?) has a nice, dry, self-deprecating SOH, I love it.
And I can relate (about the kind of developer I am, I mean) x
Paul Sanders.
If I had more time, I would have written a shorter letter - Blaise Pascal.
Some of my best work is in the undo buffer.
|
|
|
|
|
I am fond of your quote of Mons. Pascal. I am thankful I learned of it. In case you have not already done so may I suggest looking up the olde English translation of the original French as the verbiage is quite impressive at least to me.
|
|
|
|
General 
News 
Suggestion 
Question 
Bug 
Answer 
Joke 
Praise 
Rant 
Admin 
Submit an article or tip
